163
CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ENDOTHELIAL CELLS IN TYPE 2 DIABETES MELLITUS Structural, morphological and molecular analysis DOCTORAL THESIS for obtaining the academic degree of Doctor of Philosophy (Ph.D.) submitted by Monika Hämmerle, MD within the thematic program: Cell communication in health and disease (CCHD) supervised by Prof. Dr. Dontscho Kerjaschki & Dr. Brigitte Hantusch Clinical Institute of Pathology Vienna, August 2012

CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  • Upload
    others

  • View
    7

  • Download
    0

Embed Size (px)

Citation preview

Page 1: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

 CHARACTERIZATION OF LYMPHATIC 

VESSELS AND LYMPHATIC ENDOTHELIAL CELLS IN TYPE 2 DIABETES MELLITUS 

 

Structural, morphological and molecular analysis      

DOCTORAL THESIS for obtaining the academic degree of 

 

Doctor of Philosophy (Ph.D.)   

  

   

  

            

submitted by Monika Hämmerle, MD within the thematic program: Cell communication in health and disease (CCHD) 

   

supervised by Prof. Dr. Dontscho Kerjaschki & Dr. Brigitte Hantusch 

Clinical Institute of Pathology Vienna, August 2012 

Page 2: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

   Acknowledgements  

First  of  all,  I  would  like  to  thank  my  supervisors  Prof.  Dontscho  Kerjaschki  und  Dr.  Brigitte 

Hantusch for giving me the opportunity to do my PhD in the research laboratory of the Clinical 

Institute of Pathology and who supported me throughout the years. 

 

I would like to thank my cooperation partners at the Department of Rheumatology, Carl‐Walter 

Steiner,  for  excellent  technical  assistence  in  FACS  sorting  and  at  the  Department  of  General 

Surgery, Dr. Christoph Neumayer, for guaranteeing me that I could use my material as fresh as 

possible.  Moreover,  I  would  like  to  thank  Dr.  Stefan  Thurner  and  especially  Dejan  Stokic  for 

helping me with the bioinformatical data analysis. 

 

I thank all my friends and colleagues from the CCHD PhD program, especially my lab mate Tom 

without whom the day would not have been so much fun. 

 

I  would  like  to  announce  a  big  thank  to  Bernhard  Höfle,  who  triggered  my  enthusiasm  for 

science. 

 

Last but not least, I would like to thank my family and friends for their incessant support,  love 

and motivation. 

 

 

 

 

 

Die Wissenschaft, richtig verstanden, heilt den Menschen von seinem Stolz; 

denn sie zeigt ihm seine Grenzen. 

Albert Schweitzer 

Page 3: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  I 

   Summary  

Background  ‐  Small  vessel  disease  of  kidney,  nerves,  retina  and  skin,  referred  to  as 

microangiopathy,  is  a  major  cause  of  morbidity  in  type  2  diabetes  mellitus  (T2DM).  While 

characteristic changes in blood capillary walls and endothelial dysfunction of blood vessels are 

well studied in type 2 diabetes, examination of lymphatic endothelial cells (LECs) and lymphatic 

vessels  (LVs)  is  scarcely  done. However,  complications  seen  in  type  2  diabetes,  e.g.  increased 

risk for infections, wound healing defects and obesity, may be related to lymphatic dysfunction. 

Therefore,  we  aimed  at  comprehensively  analyzing  potential  morphological  and  structural 

differences  of  lymphatic  endothelial  cells  and  lymphatic  vessels  in  the  skin of  type 2 diabetes 

mellitus  patients.  Further,  we  wanted  to  identify  gene  expression  signatures  that  are 

deregulated  in  human  dermal  lymphatic  vessels  to  define  mechanisms  that  are  linked  with 

microvascular complications observed in type 2 diabetes. 

Methods  –  By  immunohistochemistry,  basement membranes  of  lymph  vessels were  analyzed 

and  blood  and  lymph  vessel  densities  of  diabetic  versus  normoglycemic  skin  was  evaluated. 

Further, we identified signs of inflammation, e.g. macrophage infiltration and TNFα expression. 

We  compared  the  gene  expression  profiles  of  ex  vivo  isolated  dermal  LECs  retrieved  from 

normoglycemic  versus  type  2  diabetic  patients  using  microarrays  and  subsequent  intensive 

bioinformatical analysis. The up‐ or downregulated expression of selected candidate genes was 

confirmed by quantitative real‐time PCR and immunofluorescence stainings. Further, we focused 

on two differentially regulated genes and performed macrophage adhesion, transmigration and 

chemotatic assays as well as siRNA‐mediated knockdown experiments to identify their specific 

function in lymphatic endothelial cells.  

Results  ‐ Neither  prominent  alterations  in  extracellular matrix  (ECM) protein  deposition,  nor 

morphological BM changes of lymphatic capillaries and collecting LVs were found in the skin of 

T2DM patients. This excluded the occurrence of diabetic lymphangiopathy comparable to that of 

blood vessels. However, the evaluation of  lymph vessel counts revealed a prominent enhanced 

lymph vessel density  in type 2 diabetic patient's skin. Further, we traced a strong macrophage 

infiltration  in  the  dermis  of  type  2  diabetic  patients.  These  macrophages  produced  vascular 

endothelial growth factors VEGF‐A and VEGF‐C, as well as the pro‐inflammatory cytokine TNFα. 

Page 4: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  II 

Transcriptomal analysis of ex vivo isolated diabetic versus non‐diabetic LECs resulted in a list of 

180 differently expressed genes. Consistent with earlier studies, we identified several genes that 

have already been linked to genetic susceptibility for type 2 diabetes, including HP, APOD, HHEX, 

CD55, ANXA1, LMNA  and FABP4.  Essentially, we  observed multiple  changes  related  to  altered 

LEC proliferation, adhesion and migration. Further, in line with increased TNFα abundance, we 

observed expression changes of CXCL10, VCAM1, CYR61, CXADR, SDC1 and AQP3. TNFα treatment 

of cultured LECs led to deregulated expression of selected genes, recapitulating the array results, 

indicating that TNFα is one major contributor to diabetes‐specific gene expression signatures in 

lymphatic  endothelial  cells.  CXCL10  was  confirmed  as  one  important  candidate  gene  only 

expressed  in  chronically  inflamed  lymphatic  vessels,  contributing  to  adhesion  and 

transmigration  of  macrophages  and  possibly  intending  to  resolute  the  dermal  inflammation. 

Further,  the  fatty  acid  transporter FABP4  was  specifically  upregulated  in  LECs  and  lymphatic 

vessels  in  type  2  diabetes  in  comparison  to  blood  endothelial  cells  (BECs)  and  blood  vessels. 

FABP4 was shown to regulate LEC proliferation and permeability  in vitro, and pointed out  the 

crucial role of lymphatic vessels in fatty acid transport and metabolism. 

Conclusion  ‐  These  data  reveal  gene  sets  highlighting  the  dramatically  altered  milieu  skin 

lymphatic vessels have to cope with during type 2 diabetes mellitus. Further, we discovered that 

skin lymphatics show a chronic subacute inflammatory phenotype characterized by macrophage 

recruitment  and  de  novo  lymphangiogenesis.  We  provide  evidence  for  a  paracrine  crosstalk, 

mainly  via  TNFα  and  CXCL10,  fostering  macrophage  recruitment  to  LECs  as  one 

pathophysiological process that might contribute to persistent inflammation and consecutively, 

aberrant lymphangiogenesis in the skin. 

 

Page 5: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  III 

 

  Zusammenfassung  

Hintergrund  ‐ Mikroangiopathie, eine Erkrankung der kleinen Gefäße der Nieren, der Nerven, 

der Retina und der Haut, ist eine häufige Komplikation im Verlauf des Typ 2 Diabetes mellitus. 

Während diabetische Veränderungen der Blutgefäße  in der Haut sehr gut charakterisiert sind, 

ist  über  mögliche  morphologische,  strukturelle  und  molekulare  Veränderungen  von 

Lymphgefäßen  wenig  bekannt.  Trotzdem  besteht  die  Vermutung,  dass  Komplikationen  des 

Diabetes  wie  erhöhtes  Infektionsrisiko,  Wundheilungsstörungen  und  Veränderungen  des 

Fettstoffwechsels auch die Folge einer Lymphgefäßdysfunktion sind. Das Ziel dieser Arbeit war 

es,  potentielle  Veränderungen  des  Lymphgefäßsystems  in  der  Haut  auf  morphologischer, 

zellulärer und molekularer Ebene zu beschreiben. Darüber hinaus war es das Ziel, deregulierte 

Genexpressionsmuster  zu  erkennen,  um  sie  mit  den  oben  genannten  Komplikationen  in 

Verbindung zu bringen. 

Methoden  –  In  dieser  Arbeit  wird  eine  umfassende  immunhistochemische  Analyse  der 

diabetischen  Haut,  inklusive  einer  Analyse  der  Basalmembranen  der  Gefäße,  der  Gefäßdichte 

und  von  Zeichen  einer  Entzündung,  präsentiert.  Zusätzlich  wurde  mit  Hilfe  der  Genechip 

Microarray‐Technologie  und  nachfolgender  intensiver  bioinformatischen  Analyse  das  mRNA‐

Expressionsprofil der diabetischen im Vergleich zu nicht‐diabetischen Lymphendothelzellen, die 

ex vivo aus Patientenhaut isoliert wurden, analysiert. Die Expression wichtiger Kandidatengene 

wurde mit Hilfe  von quantitativen PCR‐Analysen  sowie  Immunfluoreszenzfärbungen bestätigt. 

Im  weiteren  Verlauf  haben  wir  funktionell  auf  zwei  Gene  fokussiert.  Es  wurden 

Makrophagenadhäsions‐, ‐transmigrations und –chemotaxis‐Experimente durchgeführt, um ihre 

Rolle  in  der  Interaktion  mit  Makrophagen  zu  analysieren.  In  siRNA‐mediierten  knockdown 

Studien wurde versucht, die spezifische Funktion dieser Gene in lymphatischen Endothelzellen 

herauszufinden. 

Ergebnisse  ‐ Signifikante Veränderungen der Basalmembranen von, sowie erhöhte Expression 

von  Extrazellulärmatrixproteinen  rund  um  diabetische  Lymphgefäße  konnten  nicht  gefunden 

werden,  was  die  Existenz  einer  sogenannten  diabetischen  Lymphangiopathie  ausschloss. 

Dennoch  zeigte  ein  Vergleich  der  Lymphgefäßdichte  von  diabetischer  und  nicht‐diabetischer 

Haut eine signifikant erhöhte Dichte der Lymphgefäße beim diabetischen Patienten. Zusätzlich 

konnte  in  der  diabetischen  Haut  eine  starke  Infiltration  mit  Makrophagen  nachgewiesen 

Page 6: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  IV 

werden.  Diese  Makrophagen  produzierten  vaskuläre  Wachstumsfaktoren  wie  VEGF‐A  und 

VEGF‐C,  sowie  das  pro‐inflammatorische  Zytokin  TNFα.  Der  transkriptionelle  Vergleich  des 

mRNA‐Profils  von  diabetischen  und  nicht‐diabetischen  Lymphendothelzellen  führte  zur 

Identifikation  von  180  differentiell  regulierten  Genen.  Neben  Genen,  die  als  mögliche 

Suszeptibilitätsgene  für  die  Entwicklung  des  Typ  2 Diabetes  gelten, wie  z.B. HP, APOD, HHEX, 

CD55, ANXA1,  LMNA  und FABP4,  wurden  Transkripte  gefunden,  die mit  der  Proliferation  von 

Lymphendothelzellen,  sowie mit der Adhäsion und Migration von  inflammatorischen Zellen  in 

Gefäßen  assoziiert  sind,  was  mit  dem  immunhistochemischen  Befund  korrelierte.  Es  wurden 

besonders prominente Expressionsunterschiede von CXCL10, VCAM1, CYR61, CXADR, SDC1 und 

AQP3 detektiert. Diese konnten durch eine Stimulation von Lymphendothelzellen mit TNFα  in 

vitro spezifisch rekapituliert werden. CXCL10 wurde als ein wichtiges Chemokin identifiziert, das 

wahrscheinlich nur im Rahmen einer Entzündung auf dermalen Lymphgefäßen exprimiert wird 

und das eine wichtige Rolle bei der Adhäsion und Transmigration von Makrophagen und damit 

möglicherweise  bei  der  Auflösung  einer  lokalen  Entzündung  spielt.  Es  war  auch  eine 

Gensignatur von Lipidtransportern  in dLECs dereguliert. Darunter war speziell die Expression 

von  FABP4  nicht  nur  in  diabetischen  Lymphendothelzellen  und  Lymphgefäßen  signifikant 

erhöht,  sondern  grundsätzlich  spezifisch  für  Lymph‐  im  Vergleich  zu  Blutendothelzellen.  Ein 

Einfuss von FABP4 auf das Verhalten von Lymphendothelzellen, wie endotheliale Proliferation 

und Permeabilität wurde gezeigt. Die verstärkte Expression von FABP4 in Lymphendothelzellen 

hebt  damit  die  essentielle  Rolle  der  Lymphgefäße  im  Rahmen  des  Lipidtransports  und 

Fettsäurestoffwechsels hervor. 

Schlussfolgerung  ‐  Diese  Arbeit  beleuchtet  die  Genexpressionsveränderungen  von 

Lymphendothelzellen der Haut  im Rahmen des Typ 2 Diabetes mellitus und zeigt mit welchen 

metabolischen Veränderungen diese zu kämpfen haben. Die erhöhte Lymphgefäßdichte  ist mit 

einer  starken  Makrophagendichte  assoziiert,  und  es  scheint  hier  eine  enge  Kommunikation 

zwischen  diesen  Zellen,  vor  allem  mittels  der  Chemokine  CXCL10  und  TNFα  stattzufinden. 

Umgestellt: Zusammenfassend lässt sich sagen, dass es über einen parakrinen Mechanismus zu 

einer  vermehrten  Rekrutierung  von  Makrophagen  in  der  diabetischen  Haut  kommt,  was  zu 

chronischer  Entzündung  und  in  der  Konsequenz  zu  vermehrter  Lymphangiogenese  führt. 

Dadurch  konnten  wir  zeigen,  dass  dermale  Lymphkapillaren  aktiv  an  den  bekannten 

Phänomenen  der  verzögerten  Wundheilung  und  persistierenden  Entzündungen  im  Typ  2 

Diabetes beteiligt sind. 

Page 7: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  V 

 

  Contents 

 

ACKNOWLEDGEMENTS  1 

SUMMARY  1 

ZUSAMMENFASSUNG  3 

CONTENTS  5 

LIST OF FIGURES  9 

LIST OF TABLES  11 

ABBREVIATIONS  12 

1. AIMS AND RESEARCH OBJECTIVES  1 

2. THE LYMPHATIC VASCULAR SYSTEM  3 

2.1 Embryonic lymphatic vascular development  4 

2.2 Lymphatic vessel anatomy and structure  8 

2.3 Molecular markers for lymphatic endothelial cells  9 

2.4 Lymphatic vascular function  14 

2.5 Perspective  16 

3. TYPE 2 DIABETES MELLITUS  17 

3.1 Hereditary factors in T2DM  17 

3.2 Insulin action and resistance  18 

Page 8: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  VI 

3.3 Role of adipose tissue in insulin resistance and type 2 diabetes mellitus  18 

3.4 Vascular complications in diabetes mellitus  20 

3.5 Skin symptoms and clinical signs of T2DM  21 

3.6 Diagnosis of T2DM  22 

3.7 Therapy of T2DM  23 

4. LYMPHATIC VESSELS, TYPE 2 DIABETES MELLITUS AND LIPID METABOLISM  25 

4.1 Lymph vessel morphology in T2DM  25 

4.2 Lymph vessels and inflammation  26 

4.3 Lymph vessels with special emphasis on lipid metabolism  27 

5. DNA MICROARRAYS  28 

5.1 Advantages and problems of using DNA microarrays  28 

5.2 Application of the microarray technique in T2DM research  29 

6. MATERIAL AND METHODS  31 

6.1 Patients and skin samples  31 

6.2 Antibodies  32 

6.3 Buffers and solutions  33 

6.4 Micropreparation of lymphatic endothelial cells from human skin  35 

6.5 Cell staining procedure for subsequent fluorescent activated cell sorting  35 

6.6 Fluorescence­activated cell sorting  36 

6.7 RNA isolation  36 

6.8 RT­PCR and gel electrophoresis  37 

6.9 Sample preparation for microarray experiment  37 

6.10 Bioinformatical and statistical analysis  38 

6.11 Quantitative Real­time PCR  38 

6.12 SDS­PAGE and Western Blot  39 

6.13 Immunohistochemistry and Immunofluorescence  40 

6.14 Evaluation of lymph and blood vessel density and counting of macrophages  40 

Page 9: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  VII 

6.15 Tissue fixation and processing for electron microscopy  41 

6.16 Primary Human Dermal Endothelial Cell Culturing  41 

6.17 siRNA­ mediated gene knockdown  42 

6.18 LEC Proliferation assay  43 

6.19 Protein Co­Immunoprecipitation (Co­IP)  43 

6.20 Chromatin immunoprecipitation (ChIP)  43 

6.21 TNFα stimulation of LECs  46 

6.22 Scratch wounding assay  46 

6.23 Enyzme­linked Immunosorbent Assay (ELISA)  46 

6.24 Macrophage adhesion assay  47 

6.25 Macrophage transmigration experiment  47 

6.26 Agarose spot assay  47 

6.27 LEC monolayer permeability assay and TEER measurements  48 

6.28 Statistical methods and analysis  48 

7. RESULTS AND DISCUSSION  49 

7.1 Morphological features of diabetic skin  49 7.1.1 Basement membrane morphology of small blood and lymphatic capillaries in diabetic skin  49 7.1.2 Basement membrane morphology of lymphatic collectors in diabetic skin  53 7.1.3 Increased lymphatic vessel density in the skin of T2DM patients  53 7.1.4 Increased macrophage infiltration in diabetic skin  56 7.1.5 Macrophages produce vascular endothelial growth factors  57 7.1.6 Increased TNFα levels in human diabetic skin  59 7.1.7 TNFα production by CD68+ macrophages  60 

7.2 Ex vivo isolation of LECs from human skin  62 7.2.1 Quality control of isolated LECs  62 

7.3 Bioinformatical analysis of diabetic versus non­diabetic LEC transcriptomes  67 7.3.1 Identification of deregulated pathways and gene functions using Ingenuity Pathway Analysis  70 7.3.2 Verification of LEC specific genes and comparison with other arrays  84 7.3.3 Diabetic LECs exhibited a distinct gene expression profile compared to diabetic BECs  85 7.3.4 Genes linked to altered lipid transport and metabolism, increased oxidative stress and to the pathogenesis of type 2 diabetes mellitus  87 7.3.5 Identification of a gene signature related to wound healing and tissue repair in dLECs  89 7.3.6 Identification of a gene signature related to increased adhesion of inflammatory cells  91 7.3.7 Deregulated genes associated with cellular host defense  93 

7.4 TNFα­induced effects on LEC behavior  96 7.4.1 TNFα responsiveness of LEC genes in vitro  96 

Page 10: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  VIII 

7.4.1 Increased migration of LECs upon TNFα stimulation  99 7.4.2 Macrophage adhesion to LECs is increased by TNFα stimulation  99 

7.5 C­X­C motif chemokine 10 (CXCL10) expression and function in LECs  102 7.5.1 CXCL10 is upregulated and secreted by LECs upon TNFα stimulation  102 7.5.2 CXCL10 mediates macrophage adhesion to LECs  103 7.5.3 CXCL10 induces chemotaxis of macrophages  106 7.5.4 CXCL10 enhances LEC‐transmigration of macrophages  108 

7.6 Characterization of Fatty acid binding protein 4 (FABP4) expression and function in LECs  110 7.6.1 FABP4 is specifically expressed in lymphatic endothelial cells  111 7.6.2 FABP4 expression could be specifically knocked down in LECs  112 7.6.3 FABP4 regulates LECs proliferation  113 7.6.4 FABP4 expression increases LEC permeability  114 7.6.5 FABP4 expression regulates PPARγ expression in LECs  114 7.6.6 Possible interactions of FABP4 with PPARγ  115 

8. CONCLUSIONS AND FUTURE PERSPECTIVES  118 

REFERENCES  124 

CURRICULUM VITAE  144    

Page 11: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  IX 

 

  List of Figures Figure 1: Growth of lymphatic vessels in the mammalian embryo: Injection model of 

Sabin. 4

Figure 2: Development of the vascular systems.  5 

Figure 3: Pathways of how glucotoxicity contributes to diabetic complications in blood vessels. 

20

Figure 4: Confirmation of LEC purity used for cell culture experiments.  42 

Figure 5: Graphical depiction of the human FABP4 promotor.  45 

Figure 6: BM morphology of blood and lymphatic capillaries in human diabetic skin.  50 

Figure 7: Laminin expression of diabetic and non‐diabetic LVs and BVs.  51 

Figure 8: Collagen IV expression of diabetic and non‐diabetic LVs and BVs.  52 

Figure 9: Ultrastructural analysis of LVs.  52 

Figure 10: Lymphatic collectors in diabetic skin do not show any morphological changes.  53 

Figure 11: LV and BV density in diabetic versus normoglycemic skin.  55 

Figure 12: Macrophage infiltration in normoglycemic versus diabetic human skin.  56 

Figure 13: VEGF‐C and VEGF‐A expression of infiltratin skin macrophages.  57 

Figure 14: Quantitative analysis of VEGF‐C and VEGF‐A production by infiltrating skin 

macrophages. 

58 

Figure 15: TNFα expression in skin.  59 

Figure 16: Production of TNFα by macrophages.  60 

Figure 17: Fluorescent activated cell sorting of ex vivo isolated LECs and BECs.  63 

Figure 18: Exclusion of leukocyte contaminations.  63 

Figure 19: Quality control of isolated LECs.  64 

Figure 20: Amplification check of RNA isolated from dLECs.  65 

Figure 21: Amplification check of RNA isolated from ndLECs.  66 

Figure 22: Normalization of microarray chip data.  67 

Figure 23: Bioinformatical analysis using Student's t‐test and RVM.  67 

Figure 24: Hierarchical cluster analysis.  69 

Figure 25: Pathway analysis using IPA.  71 

Figure 26: Deregulated cellular functions in dLECs.  71 

Figure 27: Confirmation of FABP4, APOD and NOX4 by realtime PCR and  88 

Page 12: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  X 

immunofluorescence. 

Figure 28: Confirmation of AQP3, MMP2 and CYR61 expression by realtime PCR and immunofluorescence. 

90

Figure 29: Confirmation of VCAM‐1, CXCL10, CXADR and SDC1 expression by realtime PCR and immunofluorescence.   

92

Figure 30: Upregulation of CXCL10, VCAM‐1 and CYR61 gene expression of cultured LECs by TNFα stimulation.  

97 

Figure 31: Downregulation of CXADR, SDC1 and AQP3 gene expression of cultured LECs by TNFα stimulation. 

97 

Figure 32: Downregulation of FABP4, GALNTL2 and APOD gene expression of cultured LECs by TNFα stimulation. 

98 

Figure 33: No change of MMP2 and NOX4 gene expression of cultured LECs by TNFα stimulation. 

98 

Figure 34: Increased migration of LECs after TNFα stimulation.  100

Figure 35: Adhesion of macrophages to primary LECs in vitro. 101

Figure 36: Expression and secretion of CXCL10 in LECs upon TNFα stimulation.  103

Figure 37: CXCL10 is responsible for macrophage adhesion to lymphatic endothelial monolayer. 

104

Figure 38: VCAM‐1 is dispensable for macrophage adhesion to lymphatic endothelial monolayer. 

105

Figure 39: Secreted CXCL10 is responsible for chemotaxis of macrophages. 107

Figure 40: CXCL10 confers macrophage transmigration. 108

Figure 41: FABP4 is specifically expressed in LECs compared to BECs. 111

Figure 42: siRNA‐mediated knockdown of FABP4 in LECs. 112

Figure 43: FABP4 regulates LEC proliferation. 113

Figure 44: FABP4 expression is not upregulated by lymphangiogenic factors.  113

Figure 45: FABP4 regulates permeability of LEC monolayers. 114

Figure 46: FABP4 regulates PPARγ expression. 115

Figure 47: Co‐Immunoprecipitations of FABP4 and PPARγ. 116

Figure 48: Chromatin immunoprecipitations of PPARγ. 116

Figure 49: Importance of lymphatic vessels in type 2 diabetic skin and the attempt of establish a model showing their contribution, involvement and alterations in the skin of type 2 diabetic patients.  

122

 

 

 

Page 13: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  XI 

 

  List of Tables  

Table 1: Structural differences of blood capillaries, lymphatic capillaris and collecting LVs 

 

Table 2: Lymphatic specific genes and their knockouts.  13 

Table 3: Clinical data of selected diabetic patients and normoglycemic controls.  32 

Table 4: Primary antibodies  33 

Table 5: Secondary antibodies  33 

Table 6: Primers used for RT‐PCR.  37 

Table 7: Taqman gene expression assays used for quantitative real‐time PCR.  39 

Table 8: Primers used for qPCR to check binding of PPARγ to FABP4 promotor region.  45 

Table 9: Genes associated with most deregulated pathways as evaluated by IPA analysis.   72 

Table 10: Genes assciated with most deregulated functions as evaluated by IPA analysis.  73 

Table 11: Transcript levels of deregulated candidate genes functionally clustered in (A)    Inflammatory Response, (B) LEC Adhesion and Migration, (C) LEC growth and Lymphangiogenesis, and (D) Small Molecule Biochemistry. 

78 

Table 12: Transcript levels of LEC specific genes. No significant differences could be seen for LEC‐specific genes between diabetic  LECs (dLECs) and non‐diabetic LECs (ndLECs). 

85 

Table 13: Gene overlap of diabetic LECs (dLECs) and diabetic BECs (dBECs)  86 

Table 14: Transcript levels of deregulated genes already associated with type 2 diabetes.  88 

Table 15: Transcript levels of genes associated with wound healing and tissue repair.  90 

Table 16: Transcript levels of genes linked to increased adhesion of inflammatory cells.  91 

Table 17: Transcript levels of genes associated with cellular host defense.  95 

      

Page 14: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  XII 

 

  Abbreviations ANGPT2     Angiopoietin 2 AQP3        Aquaporin 3 APC(s)      Antigen presenting cell(s) APOD       Apolipoprotein d BEC(s)      Blood endothelial cell(s) BM        Basement membrane BMI        Body mass index BV(s)        Blood vessel(s) ChIP        Chromatin immunoprecipitation CLEC‐2      C‐type lectin‐like receptor 2 CLEVER‐1      Common lymphatic endothelial and vascular endothelial receptor‐1 Co‐IP        Co‐immunoprecipitation CRP        C‐reactive protein COUP‐TF II      COUP transcription factor 2 CXADR      Coxsackie and adenovirus receptor CXCL10      Chemokine (C‐X‐C motif) ligand 10 CYR61      Cysteine rich, angiogenic inducer, 61 DARC        Duffy antigen/chemokine receptor DAVID      Database for Annotation, Visualization and Integrated Discovery DC(s)        Dendritic cell(s) ddH2O      Double destilled water dLEC(s)      Diabetic lymphatic endothelial cell(s) DM        Diabetes mellitus EC(s)        Endothelial cell(s) ECM        Extracellular matrix ER        Endoplasmatic reticulum FABP(s)      Fatty acid binding protein(s) FACS        Fluorescence activated cell sorting FFAs        Free fatty acids FITC        Fluorescein isothiocyanate FOXC2      Forkhead box protein C2 FPG        Fasting plasma glucose GLUT‐4      Glucose transporter type 4 HA        Hyaluronic acid HDMEC(s)      Human dermal microvascular endothelial cell(s) HUVEC(s)      Human umbilical venous endothelial cell(s) ICAM‐1      Interendothelial cell adhesion molecule 1 IFG        Impaired fasting glucose IGT        Impaired glucose tolerance IL        Interleukin IR        Insulin resistance IRS        Insulin receptor substrate LEC(s)      Lymphatic endothelial cell(s) LN(s)        Lymph node(s) 

Page 15: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  XIII 

LV(s)        Lymph vessel(s) LYVE‐1      Lymphatic vessel endothelial hyaluronan receptor 1 MMP2       Matrix metalloproteinase 2 ndLEC(s)      Non‐diabetic lymphatic endothelial cell(s) NEFA(s)      Non‐esterified fatty acid(s) nFκB        Nuclear factor 'kappa‐light‐chain‐enhancer' of activated B‐cells NO        Nitric oxide NOX4        NADPH oxidase 4 NRP2        Neuropilin 2 OGTT        Oral glucose tolerance test PAI‐1        Plasminogen activator inhibitor‐1 PAL‐E       Pathologische anatomie Leiden‐endothelium PAS        Periodic acid‐Schiff PI3K        Phosphoinositide 3‐kinase PKC        Phosphokinase C PLC2        Phospholipase C gamma 2 PPAR        Peroxisome proliferator‐activated receptor gamma PROX‐1      Prospero homeobox protein 1 RMA        Robust multi‐array average ROS        Reactive oxygen species RVM        Relative variance method SDC1        Syndecan 1 SLC        Secondary lymphoid tissue chemokine SLP76       SH2 domain‐containing leukocyte protein of 76 kDa SMC(s)      Smooth muscle cell(s) SNP(s)      Single‐nucleotide polymorphism(s) SOCS        Suppressor of cytokine signaling SOX18      SRY (sex determining region Y)‐box 18 SYK        Tyrosine kinase SYK T1DM       Type 1 diabetes mellitus T2DM       Type 2 diabetes mellitus TEER        Transendothelial electric resistance TLR4        Toll‐like receptor 4 TNFα        Tumor necrosis factor alpha TNFβ        Tumor necrosis factor beta VCAM‐1      Vascular cell adhesion molecule 1 VE‐cadherin      Vascular endothelial cadherin VEGF        Vascular endothelial growth factor VEGFR      Vascular endothelial growth factor receptor vWF        Von Willebrand factor     

Page 16: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  1

  1. Aims and Research Objectives  

The number of patients suffering from type 2 diabetes mellitus (T2DM) is increasing constantly. 

Impairment  of  wound  healing  and  higher  frequence  of  skin  infections  are  common 

complications seen  in T2DM patients. These symptoms widely  rely on pathological  changes of 

big  and  small  blood  vessels  which  is  called  macro‐  and  microangiopathy,  respectively 

(Stehouwer and Schaper, 1996). While endothelial dysfunction of blood vessels (BVs) is a well‐

studied  situation  in  T2DM,  research  on  pathological  changes  of  lymphatic  vessels  (LVs)  has 

rarely been done. Complications seen in T2DM seem to be not only caused by BV dysfunction. It 

was hypothesized that LVs additionally account for these complications, because LV dysfunction 

leads  to wound healing defects (Saaristo et al., 2006),  local  inflammation (Rockson, 2001) and 

obesity (Harvey et al., 2005). These are conditions often associated with the diagnosis of T2DM. 

However,  almost  nothing  is  known  about  potential  morphological,  structural  and  functional 

changes of lymphatic endothelial cells (LECs) and LVs in T2DM. 

 

The aim of this thesis project was to perform a comprehensive illustration of the morphological, 

structural and molecular changes of LVs and LECs in the skin of T2DM patients. To gain insight 

into how T2DM alters the lymphatic vasculature and to elucidate functional implications of LVs 

in the pathogenesis of the disease, three major research objectives were addressed: 

1) Trace potential structural and morphological changes of LVs in the skin of T2DM patients, 

including basement membrane changes, vessel density and signs of inflammation. 

2)  Perform  an  ex  vivo  transcriptomal  comparison  of  diabetic  versus  non­diabetic  LECs 

isolated from the skin of four type 2 diabetic and four normoglycemic patients. 

3) Identification of deregulated functional gene expression patterns of diabetic LECs and 

confirmation of association with the pathogenesis of T2DM by in vitro cell culture assays. 

 

The  first  issue was addressed by extensive histochemical and electron microscopy analyses  to 

trace morphological  alterations of LVs  in T2DM,  such as  altered  composition and  thickness of 

basement membranes, LV shape or dilation, and changes of vacuolisations or organelles in LECs. 

Further,  we  determined  the  number  of  BVs  and  LVs  in  the  skin  of  T2DM  compared  to 

Page 17: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  2

normoglycemic patients and analyzed whether these morphological differences were associated 

with signs of inflammation, e.g. immune cell infiltration. 

Secondly,  in  order  to  discover  cellular  processes  affected  in  LVs  by  T2DM,  we  performed 

microarray analyses of ex vivo isolated LECs from diabetic versus non‐diabetic human skin and 

compared  their  transcriptomal  profiles.  Extensive  bioinformatical  pathway  analysis  and 

literature  search  identified  gene  expression  signatures  linking  LEC  biology with  T2DM.  These 

gene clusters contained well known  factors as well as gene candidates  that were novel  in  this 

context,  and  they  highlighted  the  role  of  LECs  in  processes  related  to  'inflammation',  'lipid 

metabolism' and 'wound healing' as well as 'lymphangiogenesis'. 

Thirdly, strong deregulated candidate genes were selected and confirmed by quantitative real‐

time PCR and immunofluorescent stainings on human diabetic and normoglycemic skin. Among 

these,  C‐X‐C  motif  chemokine  10  (CXCL10)  and  fatty  acid  binding  protein  4  (FABP4),  were 

selected  for  further  functional  characterization. CXCL10 was  shown  to be a  chemotactic  factor 

for macrophages, contributing to the quite uncharted knowledge about the interaction between 

lymphatic endothelium and macrophages. Further, the involvement of FABP4 expression in LEC 

proliferation and endothelial monolayer permeability was analyzed. 

Conclusively, a comprehensive morphological and molecular characterization of LECs and LVs in 

the  skin  of  diabetic  versus  normoglycemic  patients  is  presented  and  is,  to  the  thesis  author’s 

knowledge,  the  first  of  this  kind.  The  data  implicate  that  lymphatic  vessels  play  a  significant 

active role  in the skin alterations of  type 2 diabetes.  It  is hoped that by  identifying patterns of 

dLEC  gene  expressions,  new  therapeutic  targets  for  this  upcoming  worldwide  epidemic  will 

become available. 

Page 18: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  3

 

  2. The lymphatic vascular system  

The  lymphatic  vascular  system  represents  a  network  of  blind‐ended  lymph  capillaries  and 

lymphatic  collectors  which  drain  the  interstitial  fluid  and  transport  it  back  to  the  blood 

vasculature,  thereby  maintaining  tissue  fluid  homeostasis.  Besides,  lymphatic  vessels  are 

important regulators of immune cell trafficking and absorption of dietary fats. The existence of 

the lymphatic vascular system is indispensable for life and lymphatic dysfunction in individuals 

leads  to  chronic  edema  and  impaired  immune  response  (Tammela  and  Alitalo,  2010;  Alitalo 

2011). The lymphatic vasculature was first described by Gaspare Aselli (Aselli, 1627). In 1902, 

Florence  Sabin  proposed  a  model  of  lymphatic  vasculature  development  (Sabin,  1902).  She 

injected ink into pig embryos and showed that lymphatic endothelial cells (LECs) bud from veins 

and form primary lymphatic sacs near the junction of the subclavian and anterior cardinal vein. 

From  these  primary  lymphatic  sacs,  LECs  form  a  primitive  lymphatic  vessel  (LV)  system  by 

endothelial sprouting (see Figure 1) surrounding tissues and organs. The thoracic duct, the main 

deliverer of  lymph fluid, develops from the cisterna chyli, a dilated sac at  the  lower end of  the 

thoracic duct, and remains as a  final connection to the blood circulation (van der Putte, 1975). 

The  Sabin model  suggests  that  the  LV  system  arises  from  the  blood  vasculature.  In  1910,  an 

alternative  model  was  proposed  by  Huntington  and  McClure  (1910)  who  claimed  that  the 

primary lymph sacs are built in the mesenchyme, and only later venous connections are formed. 

At least, this theory is true in birds (Schneider et al., 1999).  

 

Until a few years ago the differentiation between blood vessels (BVs) and LVs within tissues was 

merely  possible  according  to  morphological  and  histological  criteria,  e.g.  thinner  walls  of 

lymphatics,  no  basement  membranes  (BM)  and  lack  of  pericytes.  However,  uncertainty 

remained about distinguishing blood endothelial cells (BECs) from LECs. Especially in the last 10 

to  20  years  progress  in  this  field  accelerated  dramatically  because  of  the  identification  of 

lymphatic  specific  markers  like  vascular  endothelial  growth  factor  receptor  3  (VEGFR­3, 

Kaipainen  et  al.,  1995),  Prospero  homeobox  protein  1  (PROX­1,  Wigle  and  Oliver,  1999), 

lymphatic  vessel  endothelial  hyaluronan  receptor  1  (LYVE­1,  Banerji  et  al.,  1999)  and 

podoplanin (Breiteneder‐Geleff et al., 1999). Along with the discovery of these lymphatic specific 

Page 19: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  4

markers, it became possible to sort and cultivate primary LECs and BECs in vitro. Still, only few 

specific markers for LECs are available and most endothelial cell (EC) markers are expressed on 

both, BVs and LVs. This affirms the close structural and developmental relationship between the 

two vessel systems. 

 

 

 

 

 

 

 

 

 

 

 

 

2.1 Embryonic lymphatic vascular development  

Due  to  the  importance of  the  lymphatic  vasculature  in physiologic  and pathological  condition, 

the development of LVs has been extensively studied. Embryonic blood vessels originate  from 

mesodermal  cells which give  rise  to hemangioblasts and angioblasts,  the precursor  cells of  all 

endothelial cells (ECs; Figure 2 A). Afterwards, angiogenesis occurs that includes formation and 

remodeling of new vessels by endothelial sprouting and splitting. LVs originate from the venous 

vascular system. They start to develop in the sixth to seventh embryonic week in humans and at 

embryonic day (E) 9.0‐9.5 in mice (Tammela and Alitalo, 2010; Oliver, 2004). Oliver (2004) and 

colleagues  (Oliver  and  Srinivasan,  2010)  suggested  a  four  stage  model  of  the  lymphatic 

vasculature  development:  LEC  competence,  LEC  commitment,  LEC  specification  and  LV 

differentiation and maturation (Figure 2 B). During these steps certain genes are specifically up‐ 

and downregulated in LECs, each of them important for these distinct stages of development. 

 

Lymphatic endothelial cell competence and commitment  

In order to gain a certain competence, cells become able to respond to an initial inducing signal 

(Grainger, 1992). The key step for development of LECs is that venous ECs become responsive to 

a lymphatic‐inducing signal. Concerning the developmental competence for embryonic veins, it 

is  believed  that  the  expression  of  COUP  transcription  factor  2  (COUP­TF  II)  is  important  for 

development of the venous vascular system by inhibiting Notch‐signaling, which is required for 

Figure 1: Growth of lymphatic vessels in the mammalian embryo: Injection model of Sabin(Sabin, 1902). Lymphatic ducts growing in the three directions (a) and building of lymphatic sacs(b) from which EC are sprouting to form a primary lymphatic capillary system (c). 

Page 20: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  5

arterial  cell  differentiation  (You  et  al.,  2005).  Further,  COUP­TF  II  might  be  involved  in  the 

repression of lymphatic key transcription factor PROX­1 to prevent too early LEC differentiation 

(Oliver and Srinivasan, 2010). Earlier,  it was suggested that the expression of LYVE­1 by some 

endothelial cells lining the anterior cardinal vein at E 9.0‐9.5 is important for gaining lymphatic 

competence (Jurisic and Detmar, 2009; Cueni and Detmar, 2008; Maby‐El Hajjami and Petrova, 

2008). However, the lack of a specific lymphatic phenotype in Lyve­1 knockout mice voiced the 

concern whether it is really crucial for LEC phenotype development (Luong et al., 2009; Gale et 

al., 2007). 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure  2:  Development  of  the  vascular  systems.  Scheme  of  the  main  steps  of  blood  andlymphatic vascular development starting  from the mesodermal angioblasts, which differentiateinto endothelial cells (A). Stepwise development of the lymphatic vascular system (B); modifiedfrom Oliver and Srinivasan (2010). 

Page 21: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  6

There are speculations that not LYVE­1 alone, but another factor is responsible for the initiation 

of  this  important  step  in  venous  endothelial  cells.  This  factor  was  claimed  to  be  SOX18,  a 

transcription  factor which  is an upstream regulator of  the  lymphatic master regulator PROX­1. 

SOX18 is expressed in a subpopulation of ECs in the anterior cardinal vein and its loss results in 

lack  of  PROX­1  expression  and  termination  of  lymphatic  development.  SOX18  might  be 

transiently required for lymphatic differentiation as it is detected only up to E 14.5 (Francois et 

al.,  2008).  The  signal  leading  to  the  expression  of  SOX18  in  endothelial  cells  is  still  unknown. 

However,  after  gaining  lymphatic  competence,  the  cells  are  able  to  form  a  specific  cell  type 

which  is  dependent  on  the  inducing  tissue  (Grainger,  1992).  This  is  called  lymphatic 

commitment and occurs between E 9.0‐10.5 and  is characterized by the expression of PROX­1. 

PROX­1  is  only  expressed  in  a  subpopulation  of  ECs  in  the  region  anterior  of  the  developing 

forelimb of mice (Wigle and Oliver, 1999) and it  is the most important regulator to confer and 

maintain the LEC phenotype (Tammela and Alitalo, 2010). 

 

Lymphatic specification 

Lymphatic  specification  takes  place  between  E  10.5‐11.5.  in  mice.  After  formation  of  LEC 

precursor cells, ECs start to bud from the anterior cardinal vein on E 11.0 to build the primary 

lymph  sacs.  This  is  stimulated  by  vascular  endothelial  growth  factor  C  (VEGF­C) which  is  the 

most important lymphangiogenic factor (Kukk et al., 1996) and enhancer of lymphatic sprouting 

(Saaristo et al., 2002). During lymphatic specification, additionally to LYVE­1 and PROX­1, other 

lymphatic specific markers, e.g. secondary lymphoid chemokine (SLC), podoplanin, VEGFR­3 and 

Neuropilin 2 (NRP2), are expressed, whereas typical blood vascular markers like CD34 become 

downregulated (Kume, 2010; Tammela and Alitalo, 2010; Wigle and Oliver, 1999).  

 

Lymphatic differentiation and maturation 

The  final  step of LEC development  is  characterized by differentiation and maturation of LECs. 

This  last  stage  lasts  several  days  and  is  a  stepwise  process  (Oliver  and Harvey,  2002;  Alitalo 

2011).  LECs  are  budding  and  sprouting  from  the  lymph  sacs  to  form  a  primitive  lymphatic 

vasculature.  Additional  lymphatic  markers  are  expressed,  e.g.  desmoplakin,  β‐chemokine 

receptor D6 and angiopoietin 2 (ANGPT2) and NFATC1. Shortly before birth, the whole gamut of 

lymphatic specific markers that is found in adult LVs is present (Tammela and Alitalo, 2010). A 

recent  publication  claims  that  besides  the  well  known  factors  semaphorin3a‐neuropilin‐1 

signaling is additionally required for lymph vessel maturation and valve formation (Jurisic et al. 

2012).  

 

 

 

Page 22: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  7

Separation of blood and lymphatic vessels 

Important  for  the  formation  of  an  independent  and  functional  LV  system  is  its  complete 

separation from the blood vasculature. Several genes have been shown to control the separation 

process of LVs from BVs: SRC homology 2‐domain‐containing leukocyte protein of 76 kDa (SLP­

76)  and  tyrosine  kinase  SYK  (Abtahian  et  al.,  2003),  phospholipase  C2  (PLC2)  (Ichise  et  al., 

2009), O‐glycans  (Fu et al., 2008) and podoplanin  (Uhrin et al., 2010). Slp­76­/­  and Syk­/­ mice 

exhibit  vascular  malformations,  e.g.  arteriovenous  shunts,  abnormal  blood‐lymphatic 

connections,  haemorrhage  and  blood  filled  lymphatics  (Abtahian  et  al.,  2003).  Similar 

phenotypes  were  observed  in  mice  lacking  PLC2  (Ichise  et  al.,  2009)  or  O‐glycans  (Fu  et  al., 

2008).  However,  the mechanism  leading  to  these misconnections  is  not  absolutely  clear.  It  is 

hypothesized  that  failed  platelet  aggregation  and  thrombi  formation  in  the  blood‐lymphatic 

connector  regions  are  the  reason,  due  to  a  failure  in  signaling  between  C‐type  lectin‐like 

receptor  2  (CLEC­2),  SLP­76,  SYK  and  podoplanin.  CLEC­2‐mediated  platelet  activation  is 

dependent on the activity of SYK and SLP­76 (Suzuki‐Inoue et al., 2006). Moreover it was shown 

that podoplanin is a direct interaction partner of CLEC­2 expressed on platelets (Suzuki‐Inoue et 

al.,  2007).  Loss  of  CLEC­2‐podoplanin  interaction  leads  to  failed  platelet  activation  and 

aggregation. Confirmation of  this hypothesis was recently provided by Uhrin et al.  (2010) and 

Bertozzi et al. (2010) who showed that podoplanin is responsible for activation of platelets via 

CLEC­2,  which  results  in  activation  of  SLP­67,  and  for  subsequent  closure  of  blood‐lymphatic 

connections. Another important molecule responsible for separation of the two vascular systems 

is T‐synthase, a glycosyltransferase, possibly also by regulating podoplanin expression (Fu et al., 

2008). 

 

Development of lymph nodes 

Another  important  feature  of  lymphatic  development  is  the  formation  of  lymph  nodes  (LNs) 

which  begins  at  around  E  12.5  in  mice  when  connective  tissue  protrudes  into  lymph  sacs. 

Incipient  clusters  of  cells were  shown  to  express  interleukin‐7‐receptor‐α  (IL­7Rα)  and  to  be 

positive for CD4 and CD45, but negative for CD3 (Mebius, 2003). Via IL­7Rα signaling, expression 

of  lymphotoxin‐α1β2  is  induced which  is  necessary  for  lymphoid  neoangiogenesis.  Crucial  for 

formation  of  LNs  is  not  only  the  clustering  of  CD45+CD4+CD3‐‐cells  but  also  the  expression  of 

adhesion molecules  like vascular  endothelial  growth  factor‐1  (VCAM­1) which keeps  incoming 

cells in place for their interaction with hematopoietic and stromal cells. Chemokines expressed 

and necessary for the development of LNs are CCL19, CCL21, CXCL12 and CXCL13 (Tammela and 

Alitalo, 2010; Mebius, 2003). 

Page 23: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  8

2.2 Lymphatic vessel anatomy and structure  

There  are mainly  five different  compartments  of  the  lymphatic  vasculature  system:  lymphatic 

capillaries,  lymphatic  collectors,  lymph  nodes,  lymphatic  trunks  and  lymphatic  ducts,  having 

diameters ranging between 10µm and 2 mm (Rovenská and Rovenský, 2011; Swartz, 2001). The 

LV system starts in the periphery with blind‐ended and thin‐walled lymph capillaries. They have 

a diameter of 10‐50µm. The lymph capillaries build a network, the so‐called lymphatic areolas 

which finally assemble to bigger collecting LVs. These collecting vessels then assemble to mainly 

five main  lymphatic  trunks:  the  lumbal,  the  intestinal,  the  bronchomediastinal,  the  subclavian 

and  the  jugular  trunk. These  trunks  in  turn pass  the  lymph  into  the right and  the  left  thoracic 

ducts.  

The main LV is the left thoracic duct which drains the lymph fluid from the whole body, except 

the  upper  right  body  region,  and  transports  it  back  to  the  blood  vasculature  via  the  left 

subclavian vein. The right thoracic duct drains lymph from the upper right body region into the 

right  subclavian  vein  (Jeltsch  et  al.,  2003;  Fritsch  and  Kühnel,  2001).  Except  these  two  sites 

where  lymph  enters  the  blood,  no  connections  between  the  blood  and  the  lymph  vascular 

system  exist.  The  two  vascular  systems  rather  run  in  parallel  and  in  close  proximity  to  each 

other. Along its way, lymph is filtered by passing through LNs where foreign antigens are taken 

up  and  presented  by  antigen‐presenting  cells  (APCs)  to  initiate  a  specific  immune  reaction. 

Besides, LNs are sites of immigration and reservoir for white blood cells and tumor cells which 

gain access to the blood vasculature via high endothelial venules (Swartz and Skobe, 2001). 

Lymphatic capillaries, blood capillaries and bigger collecting LVs dispose of eminent structural 

differences  (summarized  in Table 1).  In  contrast  to blood  capillaries with narrow and  regular 

lumina, LVs normally have wide and irregular lumina. The wall of lymphatic capillaries is formed 

by a single layer of non‐fenestrated ECs with an incomplete or absent BM (Leak, 1970; Leak and 

Burke, 1968, 1966). Small LVs are not covered by pericytes and do not have valves (Alitalo et al., 

2005). Because a BM is mostly absent from lymphatic capillaries, LECs are directly attached to 

the  extracellular matrix  (ECM) with  elastic  anchoring  filaments  to  prevent  collapse  of  vessels 

due to high interstitial pressure. These anchoring filaments, 6‐10nm in diameter, are composed 

of emilin‐1 and fibrillin (Danussi et al., 2008; Gerli et al., 2000) and tether the LECs to collagen 

fibers  (Leak and Burke, 1968, 1966). Lymphatic  capillaries  show overlapping ECs which build 

valve‐like opening structures and hence regulate permeability  for  interstitial  fluid  (Trzewik et 

al.,  2001).  The  lack  of  interendothelial  tight  junctions  also  contributes  to  their  increased 

permeability  compared  to  blood  capillaries.  Pre‐collecting  and  collecting  lymph  vessels  are 

characterized by a smooth muscle cell (SMC) layer, interendothelial tight junctions and bileaflet 

valves  (Alitalo  et  al.,  2005).  The  contractility  of  the  SMCs  is  important  to  propulse  lymph, 

whereas the valves prevent its backflow (Tammela and Alitalo, 2010). The distance between two 

Page 24: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  9

valves is called the lymphangion (Jeltsch et al., 2003), each of which is a contractile compartment 

pumping  lymph  into  the  next  one.  Important  proteins  for  valve  formation  are  integrin  α9 

(Bazigou et al., 2009) and FOXC2 (Norrmén et al., 2009; Petrova et al., 2004). In functional LVs, 

which have separated regularly from BVs, no blood can be found intraluminally.  

 

 Blood 

capillaries Lymphatic capillaries 

Collecting lymphatic vessels 

Lumina  Regular, narrow  Irregular, wide  Circular, wide 

Endothelial cells Abundant cytoplasm  Scant cytoplasm  ‐ 

Overlapping endothelial cells 

Absent  Present  ‐ 

Valves  Absent Absent Present Tight junctions  Present Absent Present Cell membrane invaginations and cytoplasmic 

vesicles 

Scant  Abundant  ‐ 

Basement membrane  Present Absent Present, but thinnerAnchoring filaments  Absent Present Absent 

Blood  Present Usually absent Usually absentPericytes  Present Absent Present 

Table  1:  Structural  differences  of  blood  capillaries,  lymphatic  capillaris  and  collecting LVs: based on data from Tammela et al. (2005b), Nathanson (2003), Schmid‐Schönbein (1990) and Gnepp and Green (1979). 

2.3 Molecular markers for lymphatic endothelial cells  

LECs exhibit a panel of specific markers which are not expressed in BECs. Some EC markers are 

expressed on both cell types, e.g. CD31 (PECAM1; platelet and endothelial cell adhesion molecule 

1)  and  vascular  endothelial  (VE)‐cadherin.  Others,  like  pathologische  anatomie  Leiden‐

endothelium (PAL­E) and endoglyx‐1 are only expressed on BECs (Wilting et al., 2002; Christian 

et al., 2001). Many studies aimed at characterizing differential gene expression  in BECs versus 

LECs using  transcriptomal  and proteomic  approaches  (Podgrabinska et  al.,  2002; Hirakawa et 

al., 2003; Wick et al., 2007; Roesli et al., 2008), as well as phage display technique (Keller T.M., 

unpublished).  Due  to  the  identification  of  specific  lymphatic  markers,  research  in  this  field 

accelerated  dramatically  in  the  last  20  years.  Table  2  shows  a  summary  of  all  important 

lymphatic specific genes including lymphatic phenotypes in murine gene knockout models. The 

four main lymphatic specific markers LYVE­1, PROX­1, Podoplanin and VEGFR­3 are described in 

more detail below. 

 

 

Page 25: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  10

Lymphatic vessel endothelial hyaluronan receptor­1 (LYVE­1) 

LYVE­1  was  discovered  as  a  homologue  of  the  leukocyte  hyaluronic  acid  (HA)  receptor CD44 

(Banerji  et  al.,  1999),  showing  a  similarity  of  51%  with  CD44.  However,  LYVE­1  expression 

differs largely from to the expression of CD44. While CD44  is widely expressed in different cell 

types,  LYVE­1  has  a  restricted  expression  pattern  in  LECs  and  in  some  BECs  in  the  papillary 

dermis of eczematous and psoriatic skin (Gröger et al., 2007). Like CD44, LYVE­1 is a receptor for 

HA. Upon stimulation with  inflammatory cytokines these receptors are able to bind HA.  It was 

shown that specific changes  in receptor sialylation regulates  the  functional status and thereby 

the  binding  capacity  for  HA  (Nightingale  et  al.,  2009).  Anyhow,  LYVE­1  is  the  first  lymphatic 

specific  gene  expressed  in  ECs  during  lymphatic  development.  It  was  suggested  that  LYVE­1 

expression  is  the  initial  signal  for  ECs  to  gain  lymphatic  competence  and  to  respond  to  a 

lymphatic‐specific  signal  (Jurisic  and  Detmar,  2009;  Maby‐El  Hajjami  and  Petrova,  2008). 

However, Lyve­1­/­ mice  fail  to exhibit any specific phenotype. They do not show any apparent 

defects in embryonic development, lymphangiogenesis and no signs of lymph absorption defects 

like chylous ascites or edema (Luong et al., 2009; Gale et al., 2007). The same is true for CD44­/­ 

and Lyve­1/CD44 double knockout mice. Additionally, they show an unperturbed HA, leukocyte 

homeostasis and dendritic cell  (DC) migration (Gale et al., 2007). Stimulation of LECs with  the 

pro‐inflammatory cytokines tumor necrosis factor α (TNFα) and tumor necrosis factor β (TNFβ) 

results  in  an  immediate  internalization  and  degradation  of  LYVE­1  expression,  which  is 

completely  reversible  by  withdrawal  of  cytokines.  Nevertheless,  uptake  of  HA  or  leukocyte 

trafficking  was  not  changed  in  these  cells  (Johnson  et  al.,  2007),  suggesting  that  some  other 

receptor compensates for the function of LYVE­1 in LECs (Jackson, 2009). 

 

Prospero­related homeobox 1 (PROX­1) 

PROX­1  is  a  nuclear  transcription  factor  expressed  very  early  in  the  development  of  the 

lymphatic vasculature in mice. Around E 9.5, Prox­1 is expressed in the subset of ECs which give 

rise  to  the murine  LV  system  (Tammela  and  Alitalo,  2010).  PROX­1  was  originally  cloned  by 

homology to the Drosophila homeobox gene prospero. It is expressed in a variety of tissues, like 

in the central nervous system, lens, heart, liver and pancreas and transiently in skeletal muscle 

(Oliver et al., 1993). By generating mice which lack Prox­1 expression it could be shown that its 

inactivation leads to embryonic lethality. PROX­1 is necessary for appropriate lens development 

(Liu et al., 2012; Wigle et al., 1999) as well as liver tissue organisation (Sosa‐Pineda et al., 2000). 

Moreover,  Wigle  and  Oliver  (1999)  showed  that  inactivation  of  PROX­1  completely  arrested 

lymphatic vascular development by blocking budding and sprouting of LECs. PROX­1 is not only 

important  for  LV  development  but  also  crucial  to  establish  and  maintain  the  lymphatic 

phenotype  of  ECs.  It  has  been  shown  that  re‐expression  of  PROX­1  in  BECs  can  induce  a 

lymphatic  phenotype  of  these  cells  by  repressing  genes  which  are  associated  with  the  blood 

Page 26: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  11

vascular  phenotype  (Hong  et  al.,  2002;  Petrova  et  al.,  2002).  Conversely,  siRNA‐mediated 

downregulation of PROX­1 in cultured LECs resulted in expression of blood endothelial markers 

like CD34 and endoglin, suggesting that LEC identity is reversible (Johnson et al., 2008). 

 

Podoplanin 

Podoplanin  is  a  38kDa  protein  and  was  originally  discovered  in  puromycin  aminonucleoside 

nephrosis (PAN), a rat model of human minimal change nephropathy characterized by a severe 

flattening of podocytes and proteinuria at a nephrotic  level due to  lowered glomerular protein 

retention  capacity.  In  PAN,  podoplanin  expression  was  downregulated  by  70%  (Breiteneder‐

Geleff et al., 1997) in the glomeruli and it was shown that podoplanin is necessary for controlling 

the shape of podocytes (Matsui et al., 1999), which are  the crucial  functional  filtering cells. By 

using inhibitory antibodies to podoplanin, specific features of PAN could be induced (Matsui et 

al.,  1998).  Having  observed  that  endothelial  cells  were  specifically  stained  with  the  anti‐

podoplanin antibody,  also angiosarcomas were  stained and  it was  shown  that podoplanin  is  a 

specific  marker  for  LVs  (Breiteneder‐Geleff  et  al.,  1999).  Besides  its  expression  in  LECs  and 

podocytes, podoplanin is expressed in various other adult tissues like skeletal muscle, placenta, 

lung and heart  (Martín‐Villar et al., 2005), myofibroblasts of  the breast and salivary glands,  in 

osteoblasts  and mesothelial  cells  (Ordonez,  2006),  alveolar  type  I  cells  (Ramirez  et  al.,  2003; 

Rishi et al., 1995) and in the basal layer of the human epidermis (Schacht et al., 2005). The first 

expression of podoplanin appears  around E 10.5‐11.5  in PROX­1  positive LEC progenitor  cells 

(Oliver, 2004). Compared to LYVE1, podoplanin is indispensable for LV development. Schacht et 

al.  (2003)  showed  that  a  homozygous  mutation  in  the  podoplanin  gene  results  in  lymphatic 

pattern defects, whereas BVs appear to be normal. This defect leads to congenital lymphedema, 

dilation  of  LVs  and  impaired  lymph  transport.  Ramirez  et  al.  (2003)  described  a  podoplanin 

knockout  mouse  which  died  immediately  after  birth  because  of  respiratory  failure  due  to  a 

reduced number of type I alveolar epithelial cells. Furthermore, podoplanin is suggested to be a 

ligand for CLEC­2, a platelet membrane protein (Suzuki‐Inoue, 2009) and therefore implicated in 

the regulation of platelet function. Just recently, Uhrin et al. (2010) confirmed this hypothesis by 

providing evidence that podoplanin can induce platelet aggregation and is especially important 

for the separation of LVs from BVs during embryonic development. An additional evidence that 

podoplanin regulates platelet function is based on the finding that podoplanin‐Fc shedded into 

the  blood  vasculature  induced  disseminated  intravascular  coagulation  (Cueni  et  al.,  2010a). 

Moreover, expression of podoplanin was detected in various epithelial tumors and could be used 

as a prognostic factor because of its ability to induce migration, spreading and invasion of tumor 

cells  (Kapoor, 2012; Fujita and Tagaki, 2012; Cueni et  al., 2010b; Wicki  and Christofori, 2007; 

Wicki et al., 2006). 

 

Page 27: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  12

Vascular endothelial growth factor receptor 3 (VEGFR­3) and its ligands 

VEGFRs  belong  to  the  receptor  tyrosine  kinase  superfamily.  They  are  equipped  with  an 

extracellular  domain  organized  into  seven  immunoglobin‐like  folds,  a  single  transmembrane 

region,  a  juxtamembrane  domain,  a  split  tyrosine‐kinase  domain  that  is  interrupted  by  a  70‐

amino‐acid kinase  insert and a C‐terminal  tail (Koch et al., 2011; Olsson et al., 2006). VEGFR­3 

was one of the first discovered LEC markers. During early development, it is expressed in all ECs, 

but  it  becomes  restricted  to  the  lymphatic  system  during  lymphatic  vascular  development 

(Kaipainen et al., 1995), except for BVs in the thyroid, adrenal glands and pancreas (Tammela et 

al.,  2005a). Additionally,  tumor BVs are positive  for VEGFR­3  (Lymboussaki et  al.,  1998). Mice 

deficient for Vegfr­3 die between E 10‐12.5 due to severe cardiovascular failure accompanied by 

vascular  remodeling  defects,  pericardial  edema  and  necrosis  (Dumont  et  al.,  1998).  Two 

important  ligands  for  activating  and  regulating  VEGFR  activity  are  VEGF‐C  and  VEGF‐D 

(Tammela  and  Alitalo,  2010)  which  both  induce  migration,  proliferation  and  survival  of  ECs 

(Tammela et al., 2005a). Whereas homozygous deletion of VEGF‐C results in a complete absence 

of  lymphatic  vasculature  in  mice  and  heterozygous  loss  leads  to  lymphatic  hypoplasia 

(Karkkainen  et  al.,  2004),  loss  of  VEGF‐D  has  no  effect  on  lymphatic  development  and  was 

suggested  to  be  dispensable  (Baldwin  et  al.,  2005).  However,  blockage  of  VEGF‐D  resulted  in 

impaired  LEC  sprouting  and migration  (Ny  et  al.,  2008).  Overexpression  of Vegfr­3, Vegf­C  or 

Vegf­D  stimulated  lymphangiogenesis  in  a  plethora  of  animal models.  This mechanism  is  also 

important for tumor progression and LN metastasis. Inhibition of the Vegf­C/­D­Vegfr­3 signaling 

reduced LN metastasis in a variety of tumor models (reviewed in Tammela and Alitalo, 2010). A 

co‐receptor of VEGFR­3 is NRP2 which mediates VEGF‐C‐induced lymphatic sprouting (Xu et al., 

2010). NRP2  is  expressed  in veins and LVs and may cooperate with VEGFR­3  to maximize  the 

affnity  of  LECs  towards  the  angiogenic  factors  VEGF‐C  and  VEGF‐D.  Mice  lacking Nrp2  show 

selective  absence  or  reduction  of  small  lymphatic  capillaries,  whereas  the  collecting  vessels 

appear to be normal (Yuan et al., 2002). 

 

Page 28: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  13

 

Gene  Phenotype  Reference Transcription factors

Prox­1 No LVs, chylous ascites (‐/‐), adult‐onset 

obesity (+/‐) (Wigle and Oliver, 1999; Harvey et al., 2005) 

Foxc2 

absent valves, abnormal lymphatic patterning, lymphatic dysfunction (‐/‐), 

vessel and lymph node hyperplasia (+/‐) 

(Kriederman et al., 2003; Petrova et al., 2004; Norrmén et al., 2009) 

Sox18 (ragged)  lymphedema, chylous ascites (‐/‐) (Pennisi et al., 2000)Adhesion molecules

Integrin α9  chylothorax, lymphedema (‐/‐) (Huang et al., 2000)Growth factors/receptors

Angpt1  hyperplastic LVs (transgenic) (Tammela et al., 2005c)

Angpt2 lymphatic hypoplasia, chylous ascites 

(‐/‐)  (Gale et al., 2002) 

Vegf­C no LVs (‐/‐), lymphatic hypoplasia, lymphedema, chylous ascites (+/‐)  (Karkkainen et al., 2004) 

Vegfr­3 cardiovascular failure, chylous 

pericardium (‐/‐)  (Dumont et al., 1998) 

Vegfr­3  lymphedema (Chy mouse) (Karkkainen et al., 2001)Nrp2  lymphedema, reduction of LVs (‐/‐) (Yuan et al., 2002)

Others

Podoplanin lymphedema, LV dilation, diminished lymphatic transport (‐/‐), blood filled 

lymphatics 

(Schacht et al., 2003; Uhrin et al., 2010) 

Slp­76 and Syk abnormal blood‐lymphatic connections 

(‐/‐) (Abtahian et al., 2003) 

Efnb2 retrograde lymph ow, chylothorax, 

absent valves (‐/‐) (Mäkinen et al., 2005) 

Lyve­1  no lymphatic phenotype (‐/‐) (Gale et al., 2007)

Adrenomedullin lymphedema, abnormal lymphatic 

patterning (‐/‐) (Dunworth et al., 2008; Fritz‐

Six et al., 2008) Fiaf  dilated intestinal LVs (‐/‐) (Bäckhed et al., 2007)

Table  2:  Lymphatic  specific  genes  and  phenotypes  of  their  knockouts  in mice.  PROX‐1, Prospero homeobox protein 1; FOXC2, forkhead box 2; SOX18, sex determining region Y‐related high mobility  group box 18; ANGPT1,  angiopoietin 1; ANGPT2,  angiopoietin 2; VEGF,  vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor; NRP2, neuropilin 2; SLP‐76, SRC homology 2‐domain‐containing leukocyte protein of 76 kDa; EFNB2, ephrin B2; LYVE‐1,  lymphatic  vessel  endothelial  hyaluronan  receptor  1;  FIAF,  fasting‐induced  adipose factor; KO, knockout; TG, transgenic. 

Page 29: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  14

2.4 Lymphatic vascular function  

LVs are found throughout the body, except for the brain and avascular or bradytroph tissues like 

cornea,  cartilage  or  epidermis.  Particularly  in  the  dermis, mesenterium  and  LNs  (Fritsch  and 

Kühnel,  2001), which are  in need of  interstitial  fluid drainage,  the density of  lymph vessels  is 

very high. Adult de novo lymphangiogenesis is normally negligible but may occur during certain 

pathological  conditions  like  tissue  inflammation,  wound  healing  processes  and  lymphatic 

metastasis (Tammela and Alitalo, 2010; Cueni and Detmar, 2008). 

 

Drainage of interstitial fluid 

The  most  important  function  of  LVs  is  to  drain  the  interstitial  fluid  to  maintain  body  fluid 

homeostasis. The  lymphatic vascular  system  is a one‐way  transport  system  for  fluid, proteins, 

lipids  and  cells  forming  a  milky  suspension  which  is  absorbed  from  the  interstitium  and 

transported back to the blood vasculature. Via valve‐like openings, tissue fluid is taken up and 

transported  further  by  active  and  passive  lymphatic  pumps  and  valves  (reviewed  in 

Bridenbaugh et al., 2003). Active  lymphatic pumps (intrinsic  forces) are only present  in  larger 

collecting  vessels  and  are  generated by  the  contractions  of  SMCs.  The passive  lymph pump  is 

dependent  on  extrinsic  forces  produced  in  the  surrounding  tissue,  e.g.  skeletal  muscle  and 

cardiac contraction, respiratory or gastrointestinal peristaltic pressure, and mainly relevant for 

lymph  capillaries  due  to  the  lack  of  surrounding  SMCs.  In  general,  lymph  fluid  is  transported 

against  a  hydrostatic  pressure  and  a  protein  concentration  gradient  back  to  the  blood 

vasculature mainly by  the  'left'  thoracic duct. Lymph propulsion  in  the  larger collecting LVs  is 

dependent  on  the  contraction  of  SMCs  and  on  systemic  forces,  like  respiration  (Schad  et  al., 

1978),  blood  pressure  (Parsons  and  McMaster,  1938),  exercise  (Olszewski  et  al.,  1977)  and 

massage (Mortimer et al., 1990). The net flow rate of lymph is approximately 100‐500 times less 

than  blood  flow  rate  (Swartz,  2001).  Lymph  formation  is  dependent  on  interstitial  tissue 

pressure and tightness of the ECM (reviewed in Swartz, 2001). Fluid movement can be described 

by  the  Starling  Equation  and  is  mainly  a  result  of  diffusion  and  filtration  and  dependent  on 

hydrostatic  and  oncotic  forces,  so  called  'Starling  forces'  (Jackson,  2009).  Normally,  all  forces 

together lead to a diffusion or filtration rate of 2 liters of interstitial fluid per day, most of which 

is taken up by blind‐ended lymphatic capillaries. Determinants of lymphatic uptake are molecule 

size, composition and lipophilicity. It was shown that molecules with sizes up to 1µm can enter 

lymphatics  freely  (Bergqvist  et  al.,  1983). Moreover,  cancer  drugs  enter  lymphatic  circulation 

much faster as in form of colloids than as solutes (Hagiwara et al., 1989) and lipids show higher 

uptake  rates  than  hydrophilic  substances  (Jackson,  1981).  Lymphatic  and  interstitial  flow  is 

further  suggested  to  play  important  role  in  tumor  antigen  presentation  and  tumor  immune 

escape (Swartz and Lund, 2012).   

Page 30: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  15

Uptake and transport of intestinal lipids ­ Implications for lipid metabolism 

Lacteals, specialized and blind‐ended lymphatics  in the center of each villus are crucial  for the 

uptake  of  dietary  lipids  packed  as  large  lipoprotein  particles  from  the  intestine  (Schmid‐

Schönbein, 1990). Intestinal lymph is a milky, chylomicron‐rich fluid which is also designated as 

'chyle'.  The  appearance  of  free  chyle  in  the  abdominal  or  thoracic  cavity,  chylous  ascites  or 

chylothorax,  respectively,  is  a  sign  of  lymphatic  dysfunction  (Rockson,  2008).  This  is  seen  in 

mouse  models  with  mutations  in  Prox­1,  Vegfr­3,  Angpt2  and  Sox18.  Information  about 

regulation of uptake of  lipids  into  lacteals  is sparse. Additionally,  it  is not known whether LVs 

can  actively  influence  and  regulate  lipid  uptake  and  transport.  Recently,  Dixon  et  al.  (2009) 

described  an  in  vitro  model  of  the  enterocyte‐lacteal  interface  using  differentiated  LECs  and 

Caco‐2  cells.  By  using  a  fluorescently  labeled  fatty  acid,  it  was  shown  that  its  transport  is 

polarized  from  the  enterocytes  to  the  lymphatics  through  both  transcellular  and  paracellular 

pathways.  This  model  highlights  not  only  the  importance  of  lymphatics  for  lipid  uptake  but 

provides an interesting tool for evaluating drug delivery from the intestine into the circulation. 

 

Interaction with immune cells 

LVs  serve  as  the  principal  conduits  for  APCs  to  reach  regional  LNs  and  to  encounter  foreign 

antigens.  Moreover,  they  take  part  in  the  clearance  of  leukocytes  from  sites  of  inflammation 

(Tammela  and  Alitalo,  2010).  An  inflammatory  condition  itself  stimulates  angiogenesis  and 

lymphangiogenesis  (Alitalo,  2011;  Alitalo  et  al.,  2005),  e.g.  chronic  inflammatory  diseases  like 

psoriasis  (Halin  and  Detmar,  2008)  or  rheumatoid  arthritis  (Zhang  et  al.,  2007).  Pro‐

inflammatory  cytokines  such  as  TNFα  can  induce  expression  of  the  lymphangiogenic  factor 

VEGF‐C, which  is mainly  produced by macrophages  (Baluk  et  al.,  2009,  2005).  Thereby,  these 

cells  are  able  to  stimulate  lymphangiogenesis  (Maruyama  et  al.,  2005).  Toll‐like  receptor  4 

(TLR4)  is  highly  expressed  on  the  surface  of  LECs  and  the  main  mediator  of  LPS‐induced 

activation  of  NFκB  (nuclear  factor  'kappa‐light‐chain‐enhancer'  of  activated  B‐cells).  Via  this 

signaling  pathway,  various  chemokines,  such  as  CCL2,  CCL5  and  CX3CL1  are  released,  which 

contribute to the chemotaxis of macrophages (Kang et al., 2009). On the other hand, migration of 

DCs is regulated via CCR7 expressed on DCs and CCL21 expressed on LECs (Förster et al., 2008; 

Ohl et al., 2004). In Ccr7‐knockout mice, migration of DCs via LVs to LNs is precluded (Randolph 

et  al.,  2005).  Further,  pro‐inflammatory  cytokines,  e.g.  TNFα,  stimulate  the  expression  of  the 

leukocyte  adhesion  molecules  inter‐cellular  adhesion  molecule  1  (ICAM­  1),  VCAM­1  and  E‐

selectin on LECs which in turn stimulate DC adhesion and transmigration (Johnson et al., 2007). 

CCR10‐positive  T‐lymphocytes  are  especially  attracted  by  a  subpopulation  of  collecting  LVs 

expressing  high  levels  of  CCL27  but  low  levels  of  podoplanin  (Wick  et  al.,  2008).  Moreover, 

expression of macrophage‐mannose receptor 1 (MMR­1, Marttila‐Ichihara et al. (2008); Irjala et 

al.  (2001)) and common  lymphatic endothelial and vascular endothelial receptor‐1 (CLEVER­1, 

Page 31: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  16

Karikoski  et  al.,  2009;  Salmi  et  al.,  2004)  regulate  lymphocyte  traffcking  in  LVs.  However, 

detailed knowledge on the specific interaction mechanisms of different immune cell types with 

the lymphatic endothelium is still sparse. 

 

Escape route for tumor cells ­ Lymphatic metastasis 

Tumors  spread  either  via  LVs  or  BVs  to  form  metastases  in  regional  LNs  or  distant  organs, 

respectively. Tumors start to metastasize (i) by induction of peritumoral LV growth and (ii) by 

invading the newly formed as well as pre‐existing afferent LVs. Detection of tumor cells within 

LNs,  especially  in  the  so‐called  sentinel  LN,  is  clinically used  and  important  for  tumor  staging 

and therapy (Lyman et al., 2005). However, the mechanism how tumor cells invade into LVs is 

poorly  understood.  Lymphangiogenic  factors  like VEGF­C  and VEGF­D  induce  the  formation  of 

new  LVs  by  binding  to  VEGFR­3  (Jeltsch  et  al.,  1997)  and  were  shown  to  be  prominently 

upregulated  in  a  variety  of  tumors  (reviewed  in  Stacker  et  al.,  2002). VEGF­C‐induced  tumor 

lymphangiogenesis and tumor growth could be inhibited by blocking VEGFR­3 signaling (He et 

al., 2002). Additionally, neutralizing monoclonal antibodies to VEGF­C, VEGF­D or VEGFR­3 were 

suffcient  to  block  tumor  growth  (Stacker  et  al.,  2001).  Moreover,  chemokines  and  their 

corresponding receptors are suggested to be involved in the interaction of tumor cells with the 

lymphatic  endothelium,  e.g.  CCL21  and  CCR7  (Shields  et  al.,  2010,  2007).  Additionally, 

lipoxygenases  are  implicated  in  tumor  cell  invasion  of  LVs  by  interendothelial  gap  formation 

(Kerjaschki et al., 2011). 

2.5 Perspective  

The analysis of LECs from different anatomic sites and pathological situations would allow for a 

better  understanding  of  the mechanisms  involved  in  lymphatic  vascular  growth,  function  and 

repair. Moreover, there is a strong need for the development of novel therapeutics in metabolic 

as  well  as  malignant  diseases.  Still,  few  studies  have  investigated  the  alterations  of  human 

lymphatic  endothelial  cells  in  human  pathophysiologic  situations.  Depending  on  the  specific 

pathophysiological  microenvironment,  different  molecular  repertoires  are  expected  to  be 

altered,  and  these  could  highlight  potential  disease  markers  and,  consecutively,  therapeutic 

targets. 

 

  

 

Page 32: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  17

  3. Type 2 diabetes mellitus  

There is a dramatic global increase in the prevalence of type 2 diabetes mellitus (T2DM) that will 

reach  an  estimated  number  of  366  million  patients  in  2030  (Wild  et  al.,  2004),  making  it  a 

workwide  epidemic.  Diabetes  mellitus  (DM)  describes  a  metabolic  disorder  mainly 

characterized by abnormal hyperglycemia which leads to subsequent defects in insulin action or 

insulin  secretion  or  both,  and  then  being  associated  with  disturbances  in  fat,  protein  and 

carbohydrate metabolism in the whole body (Alberti and Zimmet, 1998). 

3.1 Hereditary factors in T2DM  

T2DM is known to be a multifactorial disease. Besides environmental factors, genetic elements 

are suggested to contribute to its pathogenesis. Predisposition to the disease is determined by a 

combination  of  both  components,  meaning  that  without  exposure  to  certain  environmental 

factors,  development  of  T2DM  in  genetically  predisposed  people  does  not  necessarily  occur. 

However,  a  positive  family  history  is  a  major  risk  factor  for  T2DM  because  15‐25%  of  first‐

degree relatives of patients develop diabetes  (Pierce et al., 1995).  If both parents are affected, 

the  lifetime  risk  to  develop  T2DM  is  60%  at  the  age  of  60  (Tattersal  and  Fajans,  1975).  The 

concordance rate  for monozygotic  twins  reaches 35‐58% and  for dizygotic  twins  it  is 17‐20% 

(Kaprio et al., 1992; Newman et al., 1987).  

A  series  of  genome‐wide  linkage  scans  was  performed,  which  identified  new  important 

candidate  genes  for  the  development  of  T2DM.  These  are  gene  variants  affecting  insulin 

secretion,  insulin sensitivity, glucagon secretion or adiposity (reviewed in Staiger et al., 2009). 

One  attractive  and  consistent  candidate  gene  is  peroxisome  proliferator‐activated  receptor, 

encoded by PPARG. Individuals with a prolinealanine change at position 12 show a lower body 

mass index and decreased risk for the development of insulin resistance and T2DM (Altshuler et 

al.,  2000;  Deeb  et  al.,  1998).  Other  genetic  polymorphisms  linked  to  development  of  IR  and 

T2DM include Gly927Arg in insulin receptor substrate 1 (IRS­1) and Gly483Ser in PPAR gamma 

coactivator‐1 (PGC1α), which seemed to be not consistent because their initial association could 

not  be  recapitulated  (Stumvoll  et  al.,  2005).  During  the  last  decade,  a  lot  of  single  nucleotide 

polymorphisms (SNPs) ‐ most of  them affecting β‐cell  function (Staiger et al., 2009)  ‐ could be 

identified, which represent potential risk SNPs for the development of T2DM. 

Page 33: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  18

3.2 Insulin action and resistance  

Insulin is a peptide hormone synthesized in the β‐cells of the islet of Langerhans in the pancreas 

and  central  to  regulating  fat  and  carbohydrate metabolism  in  the  body.  It  downregulates  the 

glucose serum level by  inducing glucose uptake  into  liver, muscle and  fat  tissue, which  is  then 

stored  as  glycogen  in  these  tissues.  Insulin  further  stimulates  growth  and  differentiation, 

lipogenesis, glycogen and protein synthesis and inhibits lipolysis, proteolysis and glycogenolysis 

(Saltiel  and  Kahn,  2001).  Deviation  from  the  balanced  interplay  between  insulin  action  and 

insulin  secretion  leads  to  changes  in  blood  glucose  levels  (Stumvoll  et  al.,  2005).  In  T2DM, 

individuals are characterized by resistance of organs to the actions of insulin, thereby causing a 

relative  insulin  deficiency.  Main  insulin  target  organs,  i.e.  adipose  tissue,  liver,  muscle  and 

endothelium  are  affected  by  insulin  resistance  (IR,  Zeyda  and  Stulnig  (2009);  Simonson  and 

Kendall  (2005)).  IR  is often strongly associated with obesity, disturbances  in  lipid metabolism 

and  physical  inactivity  (Stumvoll  et  al.,  2005).  IR  is  suggested  to  be  the  central  pathological 

mechanism  underlying  glucose  intolerance,  dyslipidemia,  hypertension,  hyperuricemia, 

endothelial  dysfunction  and  a  pro‐inflammatory,  pro‐thrombotic  environment  (Simonson  and 

Kendall, 2005).  

On a molecular level, IR and T2DM become manifest in all stages of insulin signaling (Pessin and 

Saltiel,  2000).  These  include  insulin  receptor  and  kinase  activity,  the  concentration  and 

phosphorylation of insulin receptor substrate 1 (IRS­1) and 2 (IRS­2), phosphoinositide‐3‐kinase 

(PI3K)  activity,  glucose  transporter  translocation,  and  the  activity  of  intracellular  enzymes. 

Several  gene  knockout  models  with  targeted  deletion  of  components  of  the  insulin  signaling 

pathway were developed to study IR in rodents, e.g.  insulin receptor knockout, Irs­1 knockout, 

Irs­2  knockout  and  glucose  transporter  4  (Glut­4)  knockout  (Saltiel  and  Kahn,  2001).  While 

single knockouts of  genes often did not disturb glucose homeostasis,  combinatorial knockouts 

did, reflecting the polygenetic nature of T2DM (Saltiel and Kahn, 2001). 

3.3 Role  of  adipose  tissue  in  insulin  resistance  and  type  2  diabetes mellitus  

Adipose tissue plays a crucial role in the pathogenesis of T2DM, as it is a very active organ that 

releases  hormones,  cytokines  and  lipid messengers.  Enlarged  visceral  adipose  tissue  releases 

non‐esterified fatty acids (NEFAs) and inflammatory cytokines, e.g. TNFα and interleukin (IL)‐6 

(Rajala and Scherer, 2003; Ravussin and Smith, 2002; Hotamisligil et al., 1995) that contribute to 

the so‐called metabolic syndrome. NEFAs have been suggested to contribute to IR by inhibition 

of  glucose  uptake,  glucose  oxidation  and  glycogen  synthesis  and  by  increasing  hepatic 

Page 34: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  19

gluconeogenesis (Bergman and Ader, 2000). Additionally, NEFAs can activate pro‐inflammatory 

signaling  pathways  (Stumvoll  et  al.,  2005;  Itani  et  al.,  2002).  TNFα  can  further  enhance 

concentrations  of  circulating  NEFAs  by  stimulating  lipolysis  and  thereby  provoking  its  own 

negative effect on the  insulin signaling pathway (Hotamisligil, 2000). Further,  in the metabolic 

syndrome  TNFα  expression  is  increased  systemically  and  interferes with  insulin  signaling  by 

inhibiting  IRS­1  phosphorylation  and  insulin‐stimulated  glucose  uptake  into  adipocytes 

(Hotamisligil  et  al.,  1994).  Conversely,  deletion  of  TNFα  improved  insulin  sensitivity  in 

genetically  obese  mice  (Hotamisligil,  1999).  Similarly,  IL‐6  inhibits  insulin  signaling  by 

increasing  the  expression of  suppressor of  cytokine  signaling  (SOCS) proteins, which  increase 

IRS protein degradation (Krebs and Hilton, 2003; Senn et al., 2003).  

The finding of increased expression of pro‐inflammatory cytokines in adipose tissue and serum 

was the first link between inflammation, obesity and diabetes (Hotamisligil, 2006; Pradhan et al., 

2001).  Chronic  inflammation  in  adipose  tissue  induced  by  infiltration  of  macrophages  was 

described  in obese humans and mice and was  linked  to  the development of obesity‐related  IR 

(Weisberg  et  al.,  2003;  Xu  et  al.,  2003).  These  macrophages  were  shown  to  be  of  an  anti‐

inflammatory phenotype but they were capable of producing pro‐inflammatory cytokines, which 

contributed  to  the development of  IR  (Zeyda et al., 2007; Zeyda and Stulnig, 2007). Further,  a 

close  connection  between  IR  and  inflammatory  pathways,  e.g.  NFκB  pathway,  has  been 

established  (Stumvoll  et  al.,  2005).  Alternatively,  recent  experiments  suggested  that 

endoplasmic  reticulum  (ER)  stress  is  crucial  for  the  initiation  and  regulation  of  inflammation 

and insulin action as observed in IR (Hotamisligil, 2010).  

Whereas NEFAs and pro‐inflammatory cytokines are upregulated in visceral obesity and T2DM, 

the  concentration  of  adiponectin,  an  adipokine  with  beneficial  metabolic  effects  is  reduced 

(Hotta  et  al.,  2000;  Hu  et  al.,  1996).  Adiponectin  potentiates  the  effects  of  insulin,  namely 

decreased  hepatic  gluconeogenesis,  enhanced  glucose  uptake  in  skeletal  muscle,  fatty  acid 

oxidation  and  inhibition  of  lipolysis  (Stumvoll  et  al.,  2005).  Other  cytokines  released  by 

adipocytes  and/or  the  stromal  vascular  fraction  which  have  been  suggested  to  be  of  a  pro‐

inflammatory phenotype  and  to be  involved  in  the pathogenesis  of T2DM are:  leptin,  resistin, 

visfatin, retinol‐binding protein 4, IL‐1, IL‐18 and monocyte chemotactic protein‐1 (reviewed in 

Tilg and Moschen, 2008). 

Altogether, these recent findings reflect the multifactorial nature of IR and T2DM and highlight 

the  importance  of  adipose  tissue  inflammation,  ER  stress  and  inflammatory  cytokine  serum 

levels in disease pathogenesis. 

 

 

Page 35: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  20

3.4 Vascular complications in diabetes mellitus  

Cardiovascular complications are the main cause of increased morbidity and mortality in T2DM. 

Adults  diagnosed  with  diabetes  have  a  2‐  to  4‐fold  increased  risk  for  development  of 

cardiovascular  events  compared  to  those  without  diabetes  (Fox  et  al.,  2004).  Vascular 

complications  include  micro‐  and  macroangiopathy  (Schalkwijk  and  Stehouwer,  2005; 

Stehouwer et al., 1997). Microangiopathy is the essential cause of nephropathy, retinopathy and 

neuropathy. On the contrary, macroangiopathy mainly affects coronary, carotid and peripheral 

arteries  and  increases  the  risk  for  myocardial  infarction,  stroke  and  diabetic  foot  syndrome 

(Duby  et  al.,  2004;  Goldberg,  2003;  Kikkawa  et  al.,  2003;  Porta  and  Bandello,  2002).  A  huge 

clinical trial study published 1995 demonstrated that hyperglycemia plays an important role in 

the pathogenesis of microvascular complications in non‐insulin dependent DM (Group, 1995).  

Figure 3: Molecular pathways of how glucotoxicity contributes to diabetic complications in bloodvessels (taken from Calcutt et al. (2009)). 

Page 36: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  21

There  are  four major hypotheses  about how hyperglycemia  can  cause diabetic  blood vascular 

complications (Figure 3) on a molecular  level:  (i)  increased polyol pathway  flux,  (ii)  increased 

advanced glycation end‐product (AGE) formation, (iii) activation of protein kinase C (PKC) and 

(iv)  increased hexosamine pathway  flux (Brownlee, 2001). The common element  linking these 

four mechanisms  is  the overproduction of reactive oxygen species (ROS) by the mitochondrial 

electron‐transport  chain  (Du  et  al.,  2000;  Nishikawa  et  al.,  2000).  Besides  ROS  production, 

endothelial  dysfunction  is  characterized by  an  imbalanced production of  vasoconstrictors  and 

vasodilators  with  increased  production  of  endothelin  and  reduced  bioavailability  of  the 

vasodilator nitric oxide (NO), increased vascular permeability due to an increased expression of 

VEGF,  capillary  and  vascular  occlusion  due  to  increased  production  of  TGF‐β  and  fibronectin, 

pro‐inflammatory  gene  expression  and  decreased  fibrinolysis  with  increased  risk  of 

thromboembolic events (reviewed in Brownlee, 2001). Important serum biomarkers for diabetic 

endothelial  dysfunction  include  increased  levels  of  plasminogen  activator  inhibitor‐1  (PAI‐1), 

sE‐selectin, sVCAM‐1, sICAM‐1, C‐reactive protein (CRP) and TNFα (Schalkwijk and Stehouwer, 

2005; Meigs et al., 2004). However, the complete cascade of pathophysiological events that lead 

to vascular alterations by T2DM is still unsolved. 

3.5 Skin symptoms and clinical signs of T2DM  

Patients share common clinical features suggestive for the development of T2DM. These include 

obesity,  hypertension,  dyslipidemia,  non‐alcoholic  fatty  liver  and  acanthosis  nigricans  with 

brownish  hyperpigmentation  of  the  skin.  Moreover,  many  patients  show  a  familiar 

predisposition with  first or second degree relatives diagnosed with T2DM (Pinhas‐Hamiel and 

Zeitler,  2007;  Reinehr,  2005).  Around  80%  of  children  and  adolescents  are  asymptomatic 

(Reinehr,  2005).  Symptoms  of  severe  hyperglycemia  include  polyuria,  polydipsia, weight  loss, 

polyphagia and blurred vision (Association, 2005). Often associated with T2DM is an increased 

susceptibility to infections of the urinary and respiratory tract and soft tissue infections (Joshi et 

al., 1999).  

Additionally, many patients show cutaneous manifestations (van Hattem et al., 2008): 

‐ Yellow nails: Yellow discoloration of hallux nail. 

‐ Diabetic thick skin: Asymptomatic, measurably thicker skin. 

‐ Acrochordons: Small, pedunculated, soft lesions, most often on eyelids, neck, and axillae. 

‐ Diabetic  dermopathy:  Atrophic,  scarring,  hyperpigmented  macules  on  the  extensor 

surface of lower legs. 

‐ Acanthosis nigricans: Hyperpigmented plaques. 

Page 37: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  22

‐ Acquired  perforating  dermatosis:  Dome‐shaped  papules  and  nodules  with 

hyperkeratotic plug. 

‐ Calciphylaxis:  Localized  redness  and  tenderness,  then  subcutaneous  nodules  and 

necrotizing skin ulcers appear. 

‐ Eruptive xanthoma: Crops of yellow papules with an erythematous halo. 

‐ Granuloma anulare: Oval or ring‐shaped lesions with a raised border of skin‐colored or 

erythematous papules 

 

The  can  often  precede  disease  onset  by  many  years.  Moreover,  T2DM  patients  suffer  from 

recurrent  and  long‐lasting  skin  infections  with  colonization  of  bacteria,  viruses  and  fungi 

(Jelinek,  1994).  Therefore,  it  is  suggested  that  there  is  a  disturbed  skin  microenvironment 

caused by an interplay of different mechanisms, including regulatory misconduct in metabolism, 

inflammation, angiogenesis etc.  

3.6 Diagnosis of T2DM  

The  preferred  test  for  diagnosis  T2DM  is  the  measurement  of  fasting  plasma  glucose  (FPG), 

which is cost‐efficient, easy to do and acceptable to patients (Association, 2007). Alternatively, 

one  could  do  a  75‐g  oral  glucose  tolerance  test  (OGTT),  which  is  more  sensitive  and  more 

specific  than  FPG  but  hardly  reproducible  (Association,  2007).  The  use  of  HbA1c  is  not 

recommended  for  diagnosis  at  this  time,  but  rather  for  monitoring  therapy  response 

(Association, 2005), while others identified HbA1c as a predictor of T2DM (Pradhan et al., 2007). 

Criteria for diagnosis of DM are the following (Association, 2007; Rydén et al., 2007): 

1. Symptoms of diabetes (polyuria, polydipsia, unexplained weight loss) and a plasma glucose 

>200 mg/dl (11.1 mmol/l) at any time of day. 

OR 

2. FPG >126 mg/dl (7.0 mmol/l). Fasting is defined as no caloric intake for at least 8 h. 

OR 

3. 2‐h plasma glucose >200 mg/dl (11.1 mmol/l) during an OGTT. 

 

If hyperglycemia cannot be defined as diabetes, it will be categorized as impaired fasting glucose 

(IFG =  IFG 100 mg/dl (5.6 mmol/l)  to 125 mg/dl (6.9 mmol/l)) or  impaired glucose tolerance 

(IGT = 2‐h plasma glucose 140 mg/dl (7.8 mmol/l) to 199 mg/dl (11.0 mmol/l)). Both terms are 

significant  risk  factors  for  the  development  of  manifest  diabetes  and  cardiovascular  diseases 

(Association,  2007;  Rydén  et  al.,  2007).  Not  only  patients  with  symptoms,  but  also  in 

asymptomatic subjects, early diagnosis of T2DM is necessary. Criteria for testing for diabetes in 

Page 38: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  23

asymptomatic  adults,  which  are  at  high  risk  to  develop  diabetes,  are  the  follow  (Association, 

2009, 2007; Rydén et al., 2007; Association, 2005): 

 1. Age >45 years, BMI (body mass index) >25 kg/m2 

2. Age >45 years, BMI >25 kg/m2 and additional risk factors: 

‐ Physical inactivity 

‐ First‐degree relative with diabetes 

‐ High‐risk ethnic population 

‐ Hypertension (>140/90 mmHg) 

‐ HDL (high density  lipoproteins) cholesterol  level <35 mg/dl and/or  triglyceride  level 

>250 mg/dl 

‐ Polycystic ovary syndrome 

‐ IGT (impaired glucose tolerance) or IFG (impaired fasting glucose) 

‐ Vascular disease 

 Although  complex  and  expensive,  the  current  gold  standard  to  measure  insulin  sensitivity  is 

hyperinsulinemic‐euglycemic  clamp  (Simonson  and  Kendall,  2005).  IR  and  T2DM  are  often 

associated with central obesity. This feature can be defined by waist circumference (> 102cm for 

men, > 88cm for women), BMI (> 30kg=m2) or waist‐to‐hip‐ratio  (> 0:9  in men and > 0:85  in 

women).  Moreover,  it  is  important  to  use  laboratory  measurements  of  plasma  triglycerides, 

HDL, cholesterol and CRP (reviewed in Simonson and Kendall, 2005). 

3.7 Therapy of T2DM  

IR  is  the  basis  for  the  development  of  overt  T2DM.  Therefore,  the major  aim  of  therapy  is  to 

improve  insulin  sensitivity  of  peripheral  tissues.  This  can  either  be  done  by  life  style 

modifications or by conventional drug therapy. 

 

1) Life style interventions: 

The  first  step  of  therapy  are  lifestyle  interventions,  including  diet,  weight  loss  and  physical 

activity, which were shown to be effective  in preventing the development of T2DM in subjects 

with  IGT  (Tuomilehto  et  al.,  2001),  to  improve  insulin  sensitivity  in  patients  with  T2DM 

(Ostergård et al., 2007) and to reduce cardiovascular risk factors (Stumvoll et al., 2005). 

 

2) Medications: 

Thiazolidinediones are a class of drugs which enhance insulin sensitivity, reduce glycemia and 

hyperlipidemia,  improve vascular  function and  reduce  the  inflammatory milieu  in T2DM (Yki‐

Page 39: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  24

Järvinen, 2004). Thiazolidinediones act as PPAR agonists. They alter adipose tissue metabolism 

and  lead to redistribution of  tissue triglycerides  from visceral  to subcutaneous stores,  thereby 

reducing  levels  of  circulating  NEFAs  (Ye  et  al.,  2004)  and  hepatic  lipid  content  (Yki‐Järvinen, 

2004).  Additionally,  they  are  able  to  reduce  pro‐inflammatory  cytokines  and  increase 

adiponectin (reviewed in Meriden, 2004). Moreover, they decrease FPG and HbA1c (Balfour and 

Plosker, 1999).  Important side effects are weight gain, uid retention, edema, hemodilution and 

congestive  heart  failure  (Nesto  et  al.,  2004).  Therefore,  they  should  be  used  with  caution 

especially in patients with pre‐existing cardiovascular diseases (Nissen and Wolski, 2007). 

Another  treatment option  is Metformin  (Group, 1998a).  It  is a potent antihyperglycemic drug 

with  extrapancreatic  effects.  It  reduces  intestinal  glucose  resorption  and  hepatic 

gluconeogenesis,  and  increases  muscular  glucose  uptake  (Hundal  and  Inzucchi,  2003; 

Kirpichnikov  et  al.,  2002).  Metformin  has  also  been  shown  to  have  beneficial  effects  on 

cardiovascular  risk  factors  (Hundal  and  Inzucchi,  2003).  Gastrointestinal  complaints,  e.g. 

diarrhea, nausea and vomiting are common side effects (Correia et al., 2008). 

Sulfonylurea  derivates  act  by  increasing  insulin  secretion  from  the  pancreatic  β‐cell  and 

reduce the occurrence of micro‐ and macro‐vascular complications (Group, 1998b). 

Other  treatment  options  include α­glucosidase  inhibitors  (Chiasson  et  al.,  1998).  In  the  late 

stages of disease, when pancreatic insulin secretion is reduced to a level causing absolute insulin 

deficiency, treatment is only possible by exogeneous insulin. 

Moreover,  therapy  also  includes  management  of  associated  cardiovascular  risk  factors,  e.g. 

hypertension,  coronary  artery  disease,  cerebrovascular  disease  and  diabetic  foot  syndrome. 

However, even more important than therapy, is the early diagnosis and prevention of diabetes in 

order to reduce the population's burden of this disease. 

 

However,  till  now,  curative  therapy  of  T2DM  is  still  not  available,  explaining  the  necessity  of 

future  research  to  identify  novel  biomarkers  as well  as  novel  therapy  target  structures. Here, 

especially  compartments  are  interesting,  which  were  fairly  neglected  in  the  past,  e.g.  the 

lymphatic vessel system.   

Page 40: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  25

  4. Lymphatic vessels, type 2 diabetes mellitus and lipid metabolism  

Research has remained sparse in the field of LV changes in conditions of IR and obesity. While 

endothelial  dysfunction  of  BVs  is  a  well‐studied  situation,  almost  nothing  is  known  about 

potential  involvement and morphological changes of LECs and LVs  in T2DM. We hypothesized 

that LVs might experience phenotypical and molecular changes due to their important function 

in  wound  healing,  immune  cell  traffcking  and  fat  metabolism  (Alitalo,  2011;  Tammela  and 

Alitalo, 2010). 

4.1 Lymph vessel morphology in T2DM  

Early microscopy studies showed histochemical and ultrastructural changes of LVs in the skin of 

patients with T2DM. Thickening of BM, dilatations of intracellular spaces and dislocation of the 

lymphatic  endothelium  were  described,  all  of  which  affirming  the  existence  of  a  so‐called 

diabetic  lymphangiopathy  (Kaufmann  et  al.,  1980;  Ohkuma,  1979).  However,  uncertainty 

remained  as  lymphatic  capillaries  could  not  be  distinguished  from  blood  capillaries  without 

unambiguous staining of definite lymphatic markers at that time. Later, increased expression of 

the  lymphatic  markers  podoplanin  and  LYVE‐1  was  shown  for  pancreatic  lymphatics 

concomitant with  an  increased number of  vesicles  and protrusions of  LECs during  insulitis  in 

T1DM. Additionally, an increased number of DCs and T‐lymphocytes was shown to penetrate the 

lymphatic endothelial wall and to infiltrate the pancreatic tissue (Ji et al., 2006; Qu et al., 2003, 

2004), indicating an active role of lymphatic vessels during the autoimmune reaction of T1DM. 

Skeletal muscle LVs of diabetic compared to non‐diabetic mice showed no changes regarding LV 

density,  but  increased  VEGF‐C  and  VEGF‐D  mRNA  expression,  mainly  attributable  to  an 

increased  damage  of  muscle  fibers  (Kivelä  et  al.,  2007).  Moriguchi  et  al.  (2005)  evaluated 

changes  of  the  lymphatic  system  in  an  alloxan‐induced  diabetic  mouse  model.  An  increased 

thoracic duct lymph flow and reduced uptake of dextran into regional LNs was observed. These 

changes were reversible by the application of insulin. Overall, these data were pointing towards 

an  involvement of  lymphatic vessels  in  the pathophysiology of diabetes,  yet dependent on  the 

specific anatomic environment.  

Page 41: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  26

4.2 Lymph vessels and inflammation  

Maruyama  et  al.  (2007)  used  a  corneal  inflammation  model  to  evaluate  lymphangiogenesis 

associated with wound healing in diabetic db/db mice, which show an inactivating mutation of 

the leptin receptor. The study claimed reduced lymphangiogenesis in corneal incision wounds of 

diabetic mice, which was accompanied by a decreased macrophage  infiltration  into the wound 

region.  Both  phenomena,  decreased  lymphangiogenesis  and  reduced macrophage  infiltration, 

were  detected  to  be  beneficial  for  corneal wound  healing.  This  group  has  further  shown  that 

macrophages  are  able  to  contribute  to  the  formation  of  LVs  either  by  production  of  VEGF‐C, 

which  drives  lymphangiogenesis  or  by  conversion  into  LYVE‐1/podoplanin  expressing  cells 

which  reintegrate  into  the  lymphatic  endothelium  (Maruyama  et  al.,  2005).  On  the  contrary, 

another study showed that treatment of skin wounds of diabetic mice with VEGF‐C accelerated 

LV  formation  associated with  faster wound  closure  (Saaristo  et  al.,  2006).  Therefore, de novo 

lymphangiogenesis, though versatile, seems to play a role during diabetic skin wound healing. 

Further,  LVs are  essential  for proper  immunological  functioning,  especially  regarding  immune 

cell  trafficking  and  antigen delivery.  The major  cell  types  transported by  the  lymph  are CD4+ 

memory T cells and APCs (Angeli and Randolph, 2006). APCs, e.g. DCs migrate via LVs to regional 

LNs  where  they  present  antigens  to  naive  T‐cells  (Randolph  et  al.,  2005).  Pro‐inflammatory 

cytokines induce the migration of DCs from the skin towards LVs by diminishing the expression 

of E‐cadherin (Schwarzenberger and Udey, 1996), which loosens the interaction to surrounding 

keratinocytes,  and  by  upregulation  of  adhesion  molecules  (Angeli  and  Randolph,  2006). 

Moreover, DCs produce matrix metalloproteinases to digest components of the ECM and BM and 

to  facilitate  their migration  throughout  the  tissue  (Ratzinger  et  al.,  2002). Defective  lymphatic 

function leads to an inefficient clearance of inflammatory cells and pathogens via the lymph and 

consequently causes an increased risk for tissue infections (Rockson, 2001). Especially patients 

with T2DM are susceptible to all kinds of infections, including skin infections (Muller et al., 2005; 

Shah  and Hux,  2003).  It was  shown  that  the  number  of  peripheral DCs  are  reduced  in  T2DM 

patients (Seifarth et al., 2008), pointing at reduced immune surveillance.  

So far, little is known about the interaction of LECs and inflammatory cells as well as DCs under 

hyperglycemic  conditions.  Cell  culture  experiments  with  human  umbilical  venous  endothelial 

cells  (HUVECs)  showed  a  significant  increased  adhesion  of  leukocytes  after  24  hours  of  high 

glucose exposure (Morigi et al., 1998). The same effect was seen with the incubation of HUVECs 

with  sera  of  diabetic  patients  (Morigi  et  al.,  1998).  Additionally,  the  expression  of  adhesion 

molecules like ICAM­1 and VCAM­1 was significantly increased after incubation with high glucose 

which  was  reversible  by  blocking  these  molecules  (Morigi  et  al.,  1998).  Just  recently,  the 

expression  of  chemokine  ligand  27  (CCL27)  on  podoplanin‐low  expressing  lymphatic 

precollectors was reported. This finding provides one explanation of an exit route for CCR10+ T‐

Page 42: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  27

lymphocytes  from  inflamed  tissue  to  the  regional LNs  (Wick et  al.,  2008). Overall,  the  specific 

mechanisms of  immune cell recruitment and consecutive migration into and along the LVs are 

still poorly understood. 

4.3 Lymph vessels with special emphasis on lipid metabolism  

Another  important  function  of  LVs  is  the  absorption  and  transport  of  fat  and  free  fatty  acids 

(FFAs, Alitalo et al. (2005); Jurisic and Detmar (2009)). Further, there is increasing evidence that 

there  exists  a  close  relationship  between  lymphatic  function  and  fat  metabolism  as  well  as 

adipogenesis (Harvey, 2008; Rosen, 2002): 

(i)   LNs and LVs are often surrounded by fat tissue (Harvey, 2008). 

(ii)  LV‐adjacent adipocytes respond to local immune reactions by increasing lipolysis (Pond and 

Mattacks, 1998, 1995). 

(iii) Chronic stimulation of LNs expands surrounding adipose tissue (Mattacks et al., 2003). 

(iv)  LN development is necessary for development of associated LN fat pads (Eberl et al., 2004). 

 (v)  Haploinsufficiency in the PROX­1 gene, a master regulator of lymphatic development, causes 

adult‐onset  obesity  in  mice.  Cell  culture  experiments  showed  that  incubation  with  lymph  of 

these obese mice led to a differentiation of 3T3‐L1 preadipocytes to mature adipocytes (Harvey 

et al., 2005; Nougues et al., 1988). 

(vi) Another  indication  for  the  involvement  of  lymphatics  in  fat  formation  is  provided  by 

patients with lymphedema. In the late stages of disease they suffer from severe and irreversible 

fibrosis and overgrowth of adipose tissue in the edematous limb (Rockson, 2001). 

(vii) Further, it was shown that blood cholesterol levels can influence lymphatic function. Lim et 

al.  (2009)  discovered  LV  dysfunction  and  degeneration  in ApoE‐/‐ mice  fed  a  high‐fat  diet.  LV 

integrity was  disturbed  and  lymph  transport  declined,  resulting  in  tissue  swelling. Moreover, 

lymphatic collectors lacked SMC coverage and had dysfunctional valves (Lim et al., 2009). 

Hence,  the  hypothesis  that  leakage  of  lymph  or  failed  removal  of  tissue  fluid  and  adipogenic 

factors may cause adipocyte maturation may be supported by these studies. This in turn would 

have therapeutic implications for obesity and lymphedema by promoting lymphatic integrity or 

preventing  the  release of  adipogenic  factors  (Schneider et  al.,  2005). However,  to what extent 

lymph vasculature  regulates  lipid metabolism and adipose  tissue  formation  is  still  speculative 

and needs to be further clarified. Especially, knowing the regulation of uptake and transport of 

adipogenic  factors  from  and  to  the  adipose  tissue  would  be  of  great  importance  to  evaluate 

potential  contributions  of  lymphatics  to  conditions  like  lymphedema  and  obesity.

Page 43: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  28

  5. DNA Microarrays The microarray  technology  emerged more  than  a  decade  ago.  It  was  first  described  in  1987 

(Kulesh et al., 1987) and improved until the first miniaturized microarrays were produced with 

10.000 cDNAs spotted on a microscope slide (Schena, 1996; Schena et al., 1995). This technical 

progress was appreciated with enthusiasm due  to  the possibility  to detect  the expression of  a 

very  high  number  of  genes  at  once  (Lander,  1999;  Brown  and  Botstein,  1999).  In  principal, 

several types of microarrays exist according to the length of the spotted DNA‐probes (Draghici 

et al., 2006): 

1) cDNA arrays: length of the probe is up to a few thousand base pairs 

2) oligonucleotide arrays: short (25‐30mer) or long (60‐70mer) oligonucleotide probes 

The  probes  can  either  be  contact‐spotted,  inkjet  deposited  or  directly  synthesized  on  the 

substrate (Draghici et al., 2006). 

5.1 Advantages and problems of using DNA microarrays  

High  sensitivity,  high  accuracy  and  reproducibility  are  extremely  important  for  reliable 

microarray experiments. The sensitivity is defined as the concentration range (number of mRNA 

copies) in which accurate measurements versus background noise are possible. The accuracy is 

the degree of conformity of the measured versus its real value. High specificity is the ability that 

a probe binds to a unique target sequence (Draghici et al., 2006). Reproducibility is the ability of 

an experiment to be accurately reproduced. All these factors may be influenced by a variety of 

parameters,  including  technical  handling  variations  during  RNA  isolation  and  amplification, 

synthesis of cDNA, washing and hybridization conditions and the amount of DNA bound to the 

material  (Pollock,  2002).  Low  starting  material  is  a  strong  limitation  for  reliable  microarray 

experiments. In order to avoid this problem, additional 

RNA  amplification  steps were  added  (Wick  et  al.,  2004)  in  this  study.  Another  possible  error 

could  be  the  source  and  identity  of  the  cDNA  clones  used  in  the  array  (Pollock,  2002). 

Assessment  of  clone  identity  for  1189  IMAGE  cDNA  clones  revealed  that  only  62%  showed 

sequence identity with the published data (Halgren et al., 2001). An often faced problem is the 

variability of obtained results due to the biological complexity of tissues or diseases, e.g. T2DM 

or schizophrenia (Miklos and Maleszka, 2004). 

Page 44: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  29

However, the great advantage of microarrays is that analysis of 47.000 genes (Human Genome 

U133 Plus  2.0 Arrays,  Affymetrix)  is  possible  at  once.  Especially,  to  obtain  a  first  overview of 

which  genes  are  deregulated,  high‐throughput  microarray  analyses  are  indispensable.  They 

were  shown  to  be  very  helpful  in  differentiating  normal  from  tumor  cells  (Cole  et  al.,  1999). 

Additionally,  tumor  classifications  became  possible,  e.g.  for  diffuse  large  B‐cell  lymphomas 

(Alizadeh et al., 2000), melanomas (Bittner et al., 2000) and breast cancer  (Perou et al., 2000, 

1999). Moreover, large‐scale genotyping became applicable, e.g. the analysis of SNPs (Fan et al., 

2000). 

The  possibility  to  look  at  thousands  of  genes  at  the  same  time  also  bears  problems,  because 

bioinformatical analyses often result in a large list of genes. These are often hard to dissect and 

to  relate  to biological  processes which  are  important  for  the  condition  analyzed. Additionally, 

the  outcome  of  different  bioinformatical  tests  is  completely  diverse  lists  of  genes.  Moreover, 

innocent bystander genes often show the greatest differences while  the most  important genes 

are hidden in the bulk of data (Miklos and Maleszka, 2004). 

Using different online availabe tools, data can be analysed and genes are annotated according to 

their similar biological function and similar localization within the cell. Moreover, sets of genes 

can be related to distinct pathways, thereby highlighting the most important cellular processes 

within a disease. 

5.2 Application of the microarray technique in T2DM research  

T2DM  is a multifactorial disease  integrating environmental  factors and genetic elements  in  its 

pathogenesis.  Therefore,  many  parameters  can  generate  expression  data  retrieved  from 

microarray  experiments.  in  the  context  of  DM  and  obesity,  different  cell  types  have  been 

analyzed  by  genome  wide  expression  analysis  to  identify  deregulated  gene  expression 

signatures. Gene expression analysis of adipose tissue of obese and lean mice provided a wealth 

of data and insight into adipocyte function (Nadler et al., 2000; Soukas et al., 2000). Microarray 

analysis of aortic ECs in a streptozotocin‐induced model of T1DM unravelled transcripts related 

to metabolic dysfunction and inflammation (Maresh and Shohet, 2008). Gene expression profiles 

of  peripheral  blood  mononuclear  cells  of  diabetic  patients  with  and  without  diabetic 

nephropathy  were  analyzed  (Moczulski  et  al.,  2007).  Proximal  tubular  cells  exposed  to  high 

glucose  in  vitro  were  checked  for  characteristic  gene  expression  patterns  (Qi  et  al.,  2007). 

Additionally, especially the analysis of SNPs plays an important role in the identification of geno‐

/haplotypes  and  potential  susceptibility  genes  (Yang  et  al.,  2007).  Several  hundreds  of  genes 

have  been  studied  for  their  role  in  T2DM  (van  Tilburg  et  al.,  2001).  Genome wide‐scans  and 

analysis  of  SNPs  will  help  to  identify  completely  new  genes  and  pathways  linked  to  the 

Page 45: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  30

pathogenesis  of  the  disease  (van  Tilburg  et  al.,  2001).  Although  a  lot  of  new  information  is 

generated  using  these  approaches,  it  has  to  be  kept  in mind  that  T2DM  is  a  complex  disease 

where  it  is often hard  to  integrate gene expression data  into  the  functional context of disease. 

Nevertheless,  these  genetic  approaches  provide  informations  that  represent  the  basis  of 

diagnosis, development of new therapeutic targets and disease prevention. Moreover, they can 

contribute to a better understanding of the etiology and pathogenesis of T2DM (van Tilburg et 

al., 2001). 

 

Page 46: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  31

  6. Material and Methods 

6.1 Patients and skin samples  For  ex  vivo  transcriptomal  comparison,  LECs  were  isolated  from  non‐necrotic,  apparently 

healthy human skin derived from four type 2 diabetic and four normoglycemic patients. Table 3 

shows the clinical characteristics of included patients. The study and the isolation protocol were 

approved by  the  local ethics committee (ethics committee proposal nos. 449/2001; 81/2008). 

All included patients gave their informed consent prior to the isolation procedure. 

 

Diabetic patients 

In  the diabetic group, skin was obtained  from leg amputates. Leg amputations were necessary 

because of severe vascular side effects of T2DM. All four diabetic patients were diagnosed with 

T2DM more than 6 years prior to cell isolation. They suffered from peripheral arterial occlusive 

disease  (PAOD) and were  treated because of hypertension with one or  two antihypertensives. 

Three  out  of  four  patients  had  an  elevated BMI.  All  patients  received  insulin  for  treatment  of 

T2DM.  

 

Normoglycemic controls 

In  the  non‐diabetic  (normoglycemic)  group,  skin  was  taken  either  from  leg  amputates  (2 

samples) or from surgical abdominoplasty (2 samples). Leg amputations were necessary either 

because of a trauma or an ulcus cruris due to venous insuffciency not related to T2DM. Only one 

of  the  normoglycemic  patients  had  a  PAOD  and  received  hypertensive  therapy  with  three 

antihypertensives. The BMI of  included normoglycemic patients was slightly elevated and was 

between 24.7 and 30.4.  

 

Page 47: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  32

 Diabetic patients 

Patient#  1  2  3  4 Age (years)  88  77 56 76 

Sex  Male  Female Male MaleHbA1c (%)  11.4  8.7 12.4 8.7 BMI (kg/m²)  30.8  16.4 26.2 32 

PAOD  Yes  Yes Yes Yes Microangiopathy  Yes  No Yes yes Hypertension  Yes  Yes Yes yes Blood pressure 

(mmHg) 130/65  120/65  120/70  130/80 

Anti­hypertensive therapy 

Yes  Yes  Yes  Yes 

Anti­diabetic therapy 

Yes  Yes  Yes  Yes 

Indication for operation 

Ulcus cruris  Ulcus cruris  Ulcus cruris  Bypass stenosis 

Normoglycemic control patients Patient#  1  2  3  4 

Age (years)  35  26 31 84 Sex  Female  Male Male Male

BMI (kg/m²)  25  30.4 26 24.7PAOD  No  No No Yes 

Microangiopathy  No  No No No Hypertension  No  No No Yes 

Anti­hypertensive therapy 

No  No  No  Yes 

Indication for operation 

Cutis laxa abdominis 

Cutis laxa abdominis  Trauma  Ulcus cruris 

Table 3: Clinical data of selected diabetic patients and normoglycemic controls. HbA1c: glycated hemoglobin, BMI: body mass index, PAOD: peripheral artery occlusive disease. 

6.2 Antibodies  The following antibodies were used in immunohistochemistry, FACS and Western blotting:  

 

Name  Application Catalogue number 

Company 

FITC‐conjugated anti‐CD31  FACS*  555445 Becton Dickinson Pharmingen 

RPE‐Cy5.1‐conjugated mouse monoclonal anti‐CD45 

FACS  PM IM2653  Beckman Coulter 

IgG fraction of a polyclonal anti‐podoplanin antiserum 

FACS, MB†, IF‡,IHC§ 

  Produced in our lab 

mouse anti‐CD31 (Clone JC70A) 

MB, IF  M0823  DAKO Cytomation 

mouse monoclonal anti‐laminin alpha‐chain  IF  M0638  DAKO Cytomation 

mouse anti‐CD68  IF, IHC M0876 DAKO Cytomation rabbit anti‐collagen IV IF  20411 Novotec 

Page 48: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  33

mouse clone 2C3 anti‐Duffy antigen/chemokine receptor gene (DARC) 

IHC   Kindly provided by Dr. I. Colin, INSERM, Paris 

goat anti‐VEGFC  IF  AF752 R&DAnti‐VEGF‐A  IF  AF‐293‐NA R&Drabbit anti‐TNF‐α  IF  ab6671 Abcam mouse anti‐VCAM‐1  IF, CC|| 1244 Immunotech 

rabbit anti‐CXADR  IF  sc‐15405 Santa Cruz (kindly provided by B. Vigl, ETH Zurich) 

goat anti‐FABP4  IF  AF3150 R&Dgoat anti‐CXCL10  IF, CC AF‐266‐NA R&Drabbit anti‐AQP3  IF  ab15117 Abcam rabbit anti‐CYR61  IF  ab24448 Abcam 

Table  4:  Primary  antibodies  used  for  *  FACS:  Fluorescence‐activated  cell  sorting,  †  MB: magnetic beading, ‡  IF:  Immunofluorescence, §  IHC:  Immunohistochemistry,  || CC: Cell culture experiments and ¶WB: Western blotting. 

 

Name  Application Catalogue number 

Company 

R‐Phycoerythrin donkey  anti‐rabbit IgG   FACS*  711‐116‐152 

Jackson ImmunoResearch 

Alexa Fluor 546 goat anti‐mouse  IF‡  A11018  Molecular Probes 

Alexa Fluor 488 goat anti rabbit 

IF  A11034  Molecular Probes 

Alexa Fluor 546 goat‐anti rabbit 

IF  A11010  Molecular Probes 

Alexa Fluor 488 goat anti‐mouse  IF  A11017  Molecular Probes 

Alexa Fluor 488 donkey  anti‐mouse  IF  A21202  Molecular Probes 

Alexa Fluor 594 donkey anti‐goat  IF  A11058  Molecular Probes 

HRP‐conjugated rabbit anti‐mouse 

IHC§, WB¶  JZM035046  Axell 

HRP‐conjugated goat anti rabbit 

IHC, WB  SGZ034047  Axell 

Table  5:  Secondary  antibodies  used  for  *  FACS:  Fluorescence‐activated  cell  sorting,  †  MB: magnetic beading, ‡  IF:  Immunofluorescence, §  IHC:  Immunohistochemistry,  || CC: Cell culture experiments and ¶WB: Western blotting. 

6.3 Buffers and solutions  Laemmli sample buffer: 2x concentrate 

2 ml 1 M Tris‐HCl (pH 6.8) 

4 ml 20% SDS (Sodium dodecyl sulfate; no. 161‐0302, Biorad) 

200mM DTT (DL‐Dithiothreitol, for molecular biology; no. D9779‐5G, Sigma) 

Page 49: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  34

4.7 ml 60% Glycerol (Glycerin 87%; no. 1.04094.1000, Merck) 

6.6 mg Bromphenol Blue (3',3",5',5"‐Tetrabromophenolsulfonephthalein; no. B‐8026, Sigma) 

9.3 ml Aqua bidestillata 

 

Resolving gel: 10 ml 

3.3 ml Aqua bidestillata 

4.0 ml 30% Acrylamide mix/Bis Solution (29:1, 3.3%C; no. 161‐0156, Biorad) 

2.5 ml 1.5 M Tris‐HCl (pH 8.8) 

0.1 ml 10% SDS (Sodium dodecyl sulfate; no. 161‐0302, Biorad) 

0.1 ml 10% APS (Ammonium persulfate; no. 161‐0700, Biorad) 

0.004 ml TEMED (N,N,N,N'‐Tetra‐methyl‐ethylenediamine; no. 161‐0800, Biorad) 

 

Stacking gel: 5 ml 

3.4 ml Aqua bidestillata 

0.83 ml 1 M Tris‐HCl (pH 6.8) 

0.63 ml 30% Acrylamide mix/Bis Solution (29:1, 3.3%C; no. 161‐0156, Biorad) 

0.05 ml 10% SDS (Sodium dodecyl sulfate; no. 161‐0302, Biorad) 

0.05 ml 10% APS (Ammonium persulfate; no. 161‐0700, Biorad) 

0.005 ml TEMED (N,N,N,N'‐Tetra‐methyl‐ethylenediamine; no. 161‐0800, Biorad) 

 

Electrophoresis buffer: 10 l 

144 g Glycine (no. 161‐0718, Biorad) 

30g Tris (Tris‐hydroxymethyl‐aminomethan; no. 1.08382.1000, Merck) 

50 ml 20% SDS (Sodium dodecyl sulfate; no. 161‐0302, Biorad) 

up to 10 l with Aqua bidestillata 

 

Transfer buffer: 10 l 

24.58 g Tris (Tris‐hydroxymethyl‐aminomethan; no. 1.08382.1000, Merck) 

112.08 g Glycine (no. 161‐0718, Biorad) 

2 l Methanol (no. M/4000/17, Fisher Scientific) 

 

Tris­Buffered Saline Tween­20 (TBS­T): 

6.1 g (50mM) Tris 

9 g (150mM) NaCl 

up to 1000 ml with Aqua bidestillata 

500µl Tween‐20 

Page 50: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  35

6.4 Micropreparation of lymphatic endothelial cells from human skin  Ex vivo  isolation of ECs from human skin of diabetic patients and normoglycemic controls was 

performed  according  to  a  combined  enzymatic‐mechanical  protocol  as  previously  described 

(Wick  et  al.,  2008,  2007; Kriehuber  et  al.,  2001).  The whole  procedure was  conducted  on  ice. 

Around 6‐8 hours after surgical amputation, approximately a ten to ten centimeter piece of skin 

was removed and immediately put on ice onto a NIROSTA plate, disinfected with 70% alcohol. 

Excessive  subcutaneous  fat  was  removed  and  skin was  cut  into  2  cm wide  stripes  for  easier 

handling. Thereafter, skin was dermatomized with a clamp and a dermatome with a sterile blade 

mounted  (E.Weck%CO;  10  mm).  The  dermatomized  skin  was  collected  in  ice‐cold  Iscove's 

Modified Dulbecco's Medium (IMDM,  InvitrogenTM, no. 21056‐023) and washed several  times 

with  fresh  IMDM  on  a  shaker  to  remove  excessive  blood  and  fat.  In  the meantime,  Dispase  I 

(Roche Applied Science, 10 x 2 mg sterile, no. 04942086001) was reconstituted in 1 ml of aqua 

bidestillata (Abd) and further diluted with pre‐warmed PBS (cell culture grade, free of Ca2+ and 

Mg2+) to obtain a final concentration of the enzyme of 0.6‐2.4 U/ml. The skin was put into 10 cm 

dishes with  its  epidermal  side  up  and  covered with  a  sterile NIROSTA  grid  and  the Dispase  I 

solution. The dishes were incubated for at least 60 minutes in an incubator at 37°C/5% C02. The 

enzymatic reaction was stopped with pre‐warmed IMDM/10% FCS/1% 500mM EDTA solution. 

After 60 minutes,  the dermis could be easily removed from the epidermis using two tweezers. 

The dermal sheets were put into 10cm dishes containing ice‐cold EBM‐2 medium supplemented 

with  5%  fetal  serum  and  EGM‐2‐MV  SingleQuots  R  (CC‐4147;  Lonza).  Thereafter,  ECs  were 

released  from the dermal sheets by scraping with an  inverted scalpel until  the skin  fragments 

were  disaggregated.  Released  cells  were  sieved  with  a  cell  strainer  (70µm  pore  size,  BD 

FalconTM, Bedford, USA, no. 352350) positioned on 50 ml tubes. The expected cell  loss within 

the sieving process was 10‐15%. Released and sieved cells were pelleted at 4°C at 1600 rpm (in 

443g)  for  8  minutes  and  resuspended  in  20 ml  ice‐cold  PBS/2%BSA/1mM  EDTA  and  sieved 

again. 

6.5 Cell  staining procedure  for  subsequent  fluorescent activated  cell sorting  The  isolated single cells were stained with antibodies directed to CD45, CD31 and podoplanin. 

Cell  staining  was  performed  to  sort  EC  populations  into  LECs  (CD31+PDPN+)  and  BECs 

(CD31+PDPN‐) and  to exclude contaminations with CD45+ cells, e.g.  leukocytes and  fibroblasts. 

The antibodies were diluted  in PBS/2%BSA/1mM EDTA,  and  incubated  for 20 minutes  in  the 

dark. After sieving through a cell strainer, cells were spun down again at 1600 rpm (443g) for 8 

minutes,  resuspended  in 1100µl PBS/2%BSA/1mM EDTA and  transferred  into micronic  tubes 

Page 51: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  36

for subsequent cell staining. 100µl of cells were used for antibody isotype control staining (tube 

B)  and  transferred  into  a  separate  micronic  tube.  The  rest  of  the  cells  represented  the  vast 

majority of the cells and the definite sample (tube A).  

A  three‐step  staining  procedure  was  performed  with  intermediate  washing  steps  and 

centrifugation of cells at 1600 rpm for 8 minutes in between:  

1. Anti­podoplanin antibody (concentration 2.8mg/ml): 1:150 for A; rabbit IgG (concentration 

100µg/ml): 1:150 for B 

2. Donkey anti­rabbit IgG PE (R‐phycoerythrin‐conjugated; JacksonImmunoResearch, no. 711‐

116‐152): 1:200 for A and B 

3.  Mouse  monoclonal  anti­CD31­FITC  1:10  (Fluorescein‐Isothiocyanate‐conjugated;  BD 

Biosciences Pharmingen, no. 555445) and mouse anti­CD45­PC5 1:100 (RPE‐Cy5.1‐conjugated; 

Immunotech,  no.  IM2653)  for  A;  mouse  IgG1­FITC  1:10  (Dako  Cytomation,  no.  X0927)  and 

mouse anti­IgG1­PC5 1:100 (Immunotech, no. IM2663) for B 

 

After  the  last washing  and  centrifugation  step,  cells were  resuspended  in  EGM‐2MV medium, 

transferred on ice and transported to the FACS core facility unit of the General Hospital Vienna.  

6.6 Fluorescence­activated cell sorting  After cell isolation and cell staining, flow cytometric isolation was performed on a BD FACSAria 

Cell Sorter (Becton Dickinson Immunocytometry Systems, San Jose, CA). The flow rate was set to 

10.000 cells per second but normally it was not higher than 3600 cells per second due to lower 

cell numbers. Cells were separated  into LECs (CD31+PDPN+) and BECs (CD31+PDPN‐) by using 

CD45  (=leukocyte  marker)  as  gate  and  podoplanin  and  CD31  (=panendothelial  marker)  for 

sorting.  One  important  control  for  purity  of  sorted  cells  was  the  CD45  negativity,  which was 

checked  in  all  isolated  cell  populations  after  sorting  and  was  shown  as  a  plotted  histogram. 

Immediately  after  sorting,  cells  were  lyzed  with  350µl  of  RLT‐buffer/β‐mercaptoethanol  and 

RNA was isolated. 

6.7 RNA isolation  The RNA isolation was conducted using QIAshredder (Qiagen, no. 79654) and RNeasy Mini Kit 

(Qiagen,  no.  74104).  The  procedure  was  performed  at  room  temperature  according  to  the 

manufacturer's  instructions.  For  RNA  isolation,  cells  were  lyzed  in  350µl  of  RLT‐buffer/β‐

mercaptoethanol,  transferred  to  a  QIAshredder  column  and  centrifuged  for  two  minutes  at 

maximum speed (13200 rpm). 350µl of 70% ethanol was added to the flow through (lyzed and 

homogenisated  cells),  transferred  to  RNeasy mini  columns  and  centrifuged  for  30  seconds  at 

Page 52: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  37

maximum speed. Afterwards,  a DNase digest was performed: 350µl RW1‐buffer was added  to 

the  column  followed  by  30  seconds  of  centrifugation  at  maximum  speed.  80µl  DNase  I/RDD 

buffer  (1:8;  RNase‐free  DNase  Set,  Qiagen,  no.  79254) was  added  to  the  RNeasy  spin  column 

membrane  and  allowed  to  incubate  for  15  minutes  at  room  temperature.  Thereafter,  350µl 

RW1‐buffer was added and the columns were centrifuged for additional 30 seconds at maximum 

speed.  500µl  RPE‐buffer  was  added  and  the  columns  were  centrifuged  for  30  seconds  at 

maximum speed. This step was repeated twice. Additional  two minutes of centrifugation were 

added  to dry  the membrane of  the  column. Elution of  isolated RNA was performed with 40µl 

RNase‐free water. The isolated RNA samples were immediately stored at ‐80°C. 

6.8 RT­PCR and gel electrophoresis  RNA  of  ex  vivo  isolated  dLECs  and  ndLECs  was  used  for  non‐quantitative  RT‐PCR  to  detect 

transcripts  for  actin,  keratin  and  the  endothelial  markers  von  Willebrand  factor  (vWF), 

podoplanin and PROX‐1. Results were visualized using standard agarose gel electrophoresis. 

 

Gene  Forward  Reverse Podoplanin  caacgggaacgatgtggaag cgttggcagcagggcgtaac LYVE­1  gccaggtgcttcagcctggtg cttcagcttccaggcatcgcacgg Prox­1  acaagccgaagcgagaagg aacaagggtggtggctcag vWF  cgctccttctcgattattgg ccggacagcttgtagtaccc 

Keratin  agaccaaaggtcgctactgc agaactgggaggaggagagg Actin  atctggcaccacaccttctacaatgagctgcg cgtcatactcctgcttgctgatccacatctgc

Table 6: Primers used for RT­PCR. 

6.9 Sample preparation for microarray experiment  The approximate amount of  isolated mRNA per patient  sample was 10‐100 ng/µl. Due  to  this 

low  amount  of  starting  material  and  in  order  to  be  able  to  perform  a  reliable  microarray 

analysis, RNA was amplified (Wick et al., 2004). Two amplification steps were added using the 

MessageAmp®  II  aRNA  Amplification  Kit  (Ambion  Cat.No.  AM1751),  according  to  the 

manufacturer's instructions. After each amplification step, RNA amount and purity was checked 

using the Agilent 2100 Bioanalyzer. Then, 1µg of amplified RNA was biotin‐labeled and purified 

using  MessageAmpTM  II‐Biotin  Enhanced  Kit  (Ambion,  Cat.No.  AM1791),  hybridized  to 

Affymetrix  Gene  Chips  (Gene  Chip  R  Human  Genome  U133  Plus  2.0  Arrays,  no.  900466, 

Affymetrix) and chips were scanned using a Gene Chip® Scanner 3000 7G (Affymetrix). These 

arrays represent the expression repertoire of the whole human genome for the analysis of over 

47.000 transcripts on one chip. 

Page 53: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  38

6.10 Bioinformatical and statistical analysis  Bioinformatical  and  statistical  analysis  was  performed  as  previously  described  (Wick  et  al., 

2008,  2007).  After  background  correction  and  normalization  using  the  robust  multi‐array 

average (RMA) method, datasets were analyzed by the relative variance method (RVM, Stokić et 

al.  (2006))  and  Student's  t‐test  to  identify  genes  differentially  expressed  in  diabetic  (dLECs) 

versus non‐diabetic LECs  (ndLECs). Only genes with postnormalization  values  above  the BioB 

control were  included  in  the calculations. Further, a p‐value < 0.05 was considered  to  identify 

genes  that  were  significantly  different  between  dLECs  and  ndLECs.  Database  for  Annotation, 

Visualization  and  Integrated  Discovery  (DAVID,  http://david.abcc.ncifcrf.gov,  Huang  et  al. 

(2009)),  the  NetAffx  tool  available  on  the  affymetrix  website  (www.affymetrix.com)  and  the 

online  available  software  IPA  (Ingenuity  Pathway  Analysis,  http://www.ingenuity.com)  were 

used as a source for pathway analysis and information about involvement of genes in biological 

processes.  In  addition,  intensive  literature  research  was  added  to  finally  group  the  genes 

according  to  their  similar biological  function.  Cluster  analysis  and heat map visualization was 

done  using  the  open  source  software  programs Cluster  2.11  and Tree View 1.6  (available  via 

http://bonsai.hgc.jp/~mdehoon/software/cluster/software.htm).  Moreover,  a  comparison  of 

differentially regulated genes in dLECs and in diabetic BECs (dBECs) from the same patients was 

performed in order to gain knowledge on specific reaction patterns of LECs compared to BECs in 

T2DM. 

6.11 Quantitative Real­time PCR  Differential  transcript  levels  of  the  several  LEC  specific  genes were  evaluated  by  quantitative 

real‐time PCR. Quantitative real‐time PCR analyzes were performed on a Chromo4® Real‐time 

PCR  System  (Biorad).  1  µg  RNA  was  transcribed  into  cDNA  using  Superscript  II  Reverse 

Transcriptase (Invitrogen, no. 18064‐014). The obtained cDNA was diluted up to 100µl in aqua 

bidestillata (Abd). For one reaction 3µl of cDNA plus 6µl of Abd was mixed together with 1µl of 

the Taqman gene expression assay plus 10µl of Taqman Gene Expression Master Mix (Applied 

Biosystems; no. 4369016). 

Taqman gene expression assays were purchased from Applied Biosystems: 

 

Gene  Probe number  Gene  Probe number GAPDH  hs99999905_m1 GALNTL2 hs00365065_m1CRIP1  hs00832816_g1 VCAM‐1 hs01003369_m1PDPN  hs01089982_g1 CXCL14 hs01557413_m1CD44  hs00153304_m1 APOD hs00155794_m1IGFBP3  hs00181211_m1 PTX3 hs01073991_m1CXADR  hs00154661_m1 FABP4 hs00609791_m1

Page 54: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  39

AQP3  hs00185020_m1 PPARG hs01115513_m1MMP2  hs00234422_m1 GATA3 hs00231122_m1CCL27  hs00171157_m1 CXCL10 hs00171042_m1CYR61  hs00155479_m1 SDC1 hs00896424_g1 NOX4  hs00418356_m1 PTGS2 hs01573471_m1

Table 7: Taqman gene expression assays used for quantitative real­time PCR. 

 

The  following  temperature  protocol  was  used:  Steps  3  and  4  were  repeated  40  times  with 

intermediate plate reading. 

1) 50°C ‐ 00:02:00 

2) 95°C ‐ 00:10:00 

3) 95°C ‐ 00:00:15 

4) 60°C ‐ 00:01:00 

Values  were  calculated  applying  the  ddCt  method  as  described  (Pfaffl,  2001)  and  statistical 

analysis was done using Microsoft® Office Excel 2003. 

6.12 SDS­PAGE and Western Blot  Confluent  ECs  were  homogenized  in  Laemmli  sample  buffer  under  shearing  of  genomic  DNA 

with  an  inverted  pipette  tip.  For  protein  denaturation,  samples  were  heated  at  95°C  for  5 

minutes. 25µl of samples were loaded on a 12% bisacrylamide/acrylamide gel. The resolving gel 

and stacking gel were prepared according to Sambrook/Fritsch/Maniatis (Molecular Cloning. A 

Laboratory Manual,  2nd  edition,  1989).  PageRulerTM  Prestained  Protein  Ladder  (no.  SM0671, 

Fermentas Life Sciences),  a 3‐color  ladder with a mixture of 10 proteins covering a molecular 

weight range from 10 to 170 kDa (5µl/well) was used as marker. The gels were run at 120V for 

the first 30 minutes and then with 160V until the sample buffer front left the gel. Protein blotting 

was done overnight at 24V using nitrocellulose transfer membranes (BA83, Whatman Protran). 

Afterwards,  the  nitrocellulose  membrane  was  dried  between  30  minutes  and  overnight  and 

blocked with 5% milk powder in TBS‐T. The membrane was washed three times for 10 minutes 

in  TBS‐T  buffer  at  room  temperature  (RT)  and  incubated  with  indicated  antibodies  in  TBS‐

T/5%BSA  for  1  hour  at  RT  or  overnight  at  4°C  (depending  on  the  antibody  used).  After  first 

antibody hybridization, the membrane was washed three times for 10 minutes in TBS‐T buffer 

and incubated with HRP‐conjugated secondary antibodies for 1 additional hour. Afterwards, the 

membrane was washed for 1 hour to reduce background as much as possible. For detecting the 

chemoluminescent signal, ECL plus (GE) was used. Exposure times varied between 30 seconds 

and  10  minutes  depending  on  the  intensity  of  the  signal.  Densitometric  evaluation  was 

performed with the LumiImager software. 

Page 55: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  40

6.13 Immunohistochemistry and Immunofluorescence  Human skin was fixed in 4% paraformaldehyde at room temperature or in 20% sucrose at 4°C 

overnight for embedding in paraffin or geltol, respectively. For deparaffinization, 2‐5µm sections 

were  incubated  at  60°C  for  30  minutes  or  overnight,  subsequently  dewaxed  in  xylene  and 

rehydrated  in  a  series  of  descending  concentrations  of  ethanol  (100%,  96%,  70%,  50%  and 

water) to completely remove the paraffin. Antigen retrieval was performed in an autoclave for 

60 minutes or in a microwave (600 W, 3 x 5 minutes) in 10mM citrate buffer, pH 6. In the case of 

frozen sections, 5µm were cut using a cryotome and used immediately or stored at ‐20 to ‐80°C. 

For immunostaining, sections were thawed, dried at room temperature for 10 minutes and fixed 

in  ice‐cold  acetone  or  1% paraformaldehyde  (PFA)  for  20 minutes.  Thereafter,  the  individual 

staining  procedure  was  started.  Blocking  of  sections  was  performed  for  20  minutes  at  room 

temperature with 10% goat or donkey serum. Afterwards sections were washed shortly in PBS 

and first antibody solution was added. Depending on the first antibody, incubation was done for 

1 hour at RT or overnight at 4°C. Next, the sections were washed 3 times 5 minutes in PBS and 

secondary  antibody  dilution  was  applied.  For  fluorescence  microscopy,  fluorescently  labeled 

secondary  antibodies  were  used.  Immunohistochemical  stainings  were  performed  by  using 

either  Avidin  Biotin  Complex  (ABC),  Alkaline  Phosphatase  (AP)  or  Horseradish  Peroxidase 

(HRP)  staining  according  to  standard  protocols.  Pictures  were  taken  with  a  VANOX  AHBT3 

microscope (Olympus) or a laser scanning microscope (LSM 5 Exciter, Zeiss). 

6.14  Evaluation  of  lymph  and  blood  vessel  density  and  counting  of macrophages  In  order  to  quantify  the  number  of  BVs  and  LVs  in  skin  samples  of  T2D  and  normoglycemic 

patients,  5µm  sections were  immunostained with  anti‐Duffy  antigen/receptor  for  chemokines 

(DARC)  and  anti‐podoplanin  antibodies  for  visualization  of  BVs  and  LVs,  respectively.  The 

number of CD68‐positive macrophages was determined in the dermis of four diabetic and four 

non‐diabetic patients by staining with mouse anti‐human CD68 antibody (DAKO Cytomation; no. 

M0876). Non‐overlapping fields were captured with an Olympus VANOX AHBT3 microscope at 

200x magnification and the number of vessels was counted per medium power field, the number 

of CD68‐positive cells was counted per 100µm2. Mean numbers of vessels/macrophages were 

calculated per  group of  patients and  the  significance of difference was evaluated using paired 

Student's t‐test. 

Page 56: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  41

 6.15 Tissue fixation and processing for electron microscopy  Dermatomized skin was cut into 1mm2 small pieces and fixed in 4% paraformaldehyde + 0.1% 

glutaraldehyde  in  0.1 M  cacodylate buffer  for 24  to 72 hours. Thereafter,  skin was washed  in 

cacodylate  buffer  2  times  for  10 minutes  and  contrasted using  osmium  tetroxide  for  2  hours. 

After washing  in  aqua  bidestillata  and  dehydrating  in  a  series  of  ascending  concentrations  of 

ethanol (70%, 80%, 96%, 100%) the skin was put into 100% propylenoxide for 10‐15 minutes 

(twice)  and  into  a  1:1  solution  of  propylenoxide  and  embedding  medium  EPON  812  for 

additional 60 minutes. Afterwards, skin was put into pure EPON812 overnight. On the next day, 

tissue  was  poured  into  fresh  resin  and  this  was  polymerized  at  60°C  for  two  to  three  days. 

Ultrathin sections (80nm) were cut using a Reichert‐Jung Ultracut E Mikrotom and collected on a 

copper  grid.  The  grids  were  stained  with  uranyl  acetate/methanol  and  lead  citrate.  Pictures 

were taken at different magnifications using a transmission electron microscope (JEOL 1010). 

6.16 Primary Human Dermal Endothelial Cell Culturing  Human dermal microvascular endothelial cells (HDMECs) were purchased from Promocell (no. 

C‐12260). Pure LEC populations were obtained by magnetic bead sorting using the rabbit anti‐

podoplanin  antibody.  LEC  purity  was  checked  by  Western  blotting  for  the  presence  of 

podoplanin and CD31 and for the negativity of CD146 as BEC specific marker (Amatschek et al., 

2007).  Further,  nuclear  PROX‐1  as  well  as  podoplanin  expression  was  confirmed  by 

immunofluorescence  stainings  on  LECs  (see  Figure  4).  Cells  were  grown  in  endothelial  basal 

medium  (EBM‐2)  supplemented  with  5%  fetal  serum  and  EGM‐2‐MV  SingleQuots  (CC‐4147; 

Lonza)  in  an  incubator  at  37°C  and 5% CO2.  For  all  cell  culture  experiments,  LECs were used 

between passages 5‐8 and starved overnight in EBM‐2/0.5% FCS, unless otherwise indicated. 

Page 57: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  42

 

6.17 siRNA­ mediated gene knockdown  siRNA transfection of LECs was done by using RNAiFect (Qiagen, no. 301605) in 24 well plates. 

5*104 LECs were seeded the day before transfection in EGM‐2MV medium containing serum and 

antibiotics and incubated at 37°C and 5% CO2. 20 µmol of siRNA (FABP4: nos. s4964 and s4965; 

PPARg: nos. s10887 and s10888 from Ambion) was diluted in the appropriate volume of Buffer 

EC‐R (supplied within the kit) to give a final volume of 100µl. 6 µl of RNAiFect were added and 

samples  were  incubated  for  15  minutes  at  room  temperature  to  induce  complex  formation. 

300µl of fresh EGM‐2MV medium and 100µl of the complexes were added to the adherent cells. 

Medium  was  replaced  after  6  hours  to  avoid  cell  toxicity  by  the  transfection  reagent.  Gene 

silencing was monitored after 48 hours using western blot analysis. 

 

Figure 4: Confirmation of LEC purity used  for cell culture experiments. Purity of LECs wasconfirmed by western blotting by checking the presence of podoplanin and CD31 and the absenceof CD146.  Immunofluorescence double staining with  the  lymphatic specific markers podoplaninand PROX‐1 and CD31 further confirmed the purity of LECs in vitro. 

Page 58: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  43

 6.18 LEC Proliferation assay  2*104  cells  of  scrambled  or  FABP4‐specific  siRNA‐transfected  LECs were  seeded  into  24‐well 

plates with EGM‐2MV medium. The number of cells was counted every 48 hours and plotted as a 

line diagram against time on the y‐axis using Microsoft®Excel 2003. 

6.19 Protein Co­Immunoprecipitation (Co­IP)  To  show  a  direct  interaction  of  FABP4  with  PPARγ  in  LECs,  co‐immunoprecipitations  were 

performed  using  the  ProFoundTM Mammalian  Co‐Immunoprecipitation Kit  (no.  26149,  Pierce) 

according  to  the manufacturer's  instructions. Both antibodies  against FABP4 and PPARγ were 

examined  for  the  ability  to  pull  down  the  antigen  together with  the  other  protein.  LECs were 

stimulated with 20µM rosiglitazone (no. 350‐125‐M025; Enzo Life Sciences) for 24 hours prior 

to  cell  lysis  in  order  to  enhance  PPARg  levels.  For  immobilization  of  antibody,  50µg  of  either 

FABP4 or PPAR was coupled to the AminoLink® Plus Coupling Resin. LECs were lyzed with IP 

Lysis Buffer (0.5% Triton‐X100 in PBS, 10% protease inhibitor). Lysates were pulled through a 

syringe with  a  22‐gauge  needle,  put  on  the  resin  and  gently mixed  overnight  at  4°C.  The  day 

after, co‐immunoprecipitates were eluted using 50µl of Elution Buffer (supplemented with 5µl 

1M Tris‐HCl pH 9.5) provided within the kit. Eluates were concentrated using centrifugal filter 

units  (Amicon  Ultra  ‐  0.5ml  10k  Ultracel®,  no.  UFC501024,  Millipore).  30µl  of  eluates  were 

mixed with 6µl of 6x Laemmli sample buffer and applied to the gel  for SDS‐PAGE analysis. For 

long time storage of resin, 200 µl of 1X Coupling Buffer plus 0.02% sodium azide was added to 

the spin columns, which then can be reused up to 10 times. 

6.20 Chromatin immunoprecipitation (ChIP)  To  study  activation  of  the  FABP4  promotor  by  PPARγ  ChIP  was  performed  as  previously 

described (Gal‐Yam et al., 2008).  

Cell culturing and antibody coupling: 

LECs were seeded into 14 cm dishes, grown to confluence and stimulated with or without 20µM 

rosiglitazone for 24 hours. The day before cell harvesting, sheep anti‐mouse IgG magnetic beads 

(Dynabeads R M‐280, no. 112.01D, Invitrogen) were washed three times and resuspended with 

1 ml of PBS‐BSA (bovine albumin serum, no. A7030, Sigma, 5mg/ml). Beads were incubated with 

8µg/ml  of  anti‐PPARγ  antibody  (no.  ab41928,  Abcam)  or mouse  IgG  (ChromPure mouse  IgG, 

whole molecule; no. 015‐000‐003, JacksonImmunoResearch) overnight at 4°C at 10 rpm.  

 

Page 59: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  44

DNA‐protein crosslinking and chromatin isolation: 

Confluent cells were collected in EGM‐2MV, pelleted at 1600 rpm for 5 minutes, resuspended in 

5  ml  EGM‐2MV/1%  formaldehyde  (no.  47608,  Sigma)  for  DNA‐protein  crosslinking  and 

incubated  for 10‐15 minutes. Cells were quenched with 1.37 M glycine  for 5 minutes, washed 

with 10ml of ice‐cold PBS plus 1% protease inhibitors (complete EDTA‐free, no. 11873580001, 

Roche) and centrifuged for 5 minutes at 1600 rpm. Pellet was lyzed in 1ml SDS lysis buffer (1% 

SDS, 10mM EDTA, 50mM Tris‐HCl (pH 8.1)) plus 1% protease  inhibitors and put on  ice  for 10 

minutes.  Chromatin  was  sonicated  10  seconds  for  25  times  (100%,  continuous,  low)  and 

checked by  gel  electrophoresis. Therefore,  20µl  of  chromatin was  filled up with H2O, 10µl 5M 

NaCl  was  added  and  boiled  for  15  minutes  at  95°C.  RNA was  digested  by  adding  1µl  RNAse 

(DNAse  free)  and  incubation  for  30  minutes  at  37°C,  proteins  were  digested  by  adding  1µl 

proteinase K (10mg/ml) and  incubation  for 20 minutes at 45°C. Chromatin was centrifuged at 

full  speed  for  1  minute,  supernatant  was  collected  and  DNA  was  purified  using  a  QIAquick 

column  (Qiagen  PCR  Purification  Kit,  no.  28106).  DNA  concentration  was  measured  by 

Nanodrop 2000 (Thermo Scientific) and 2µg of DNA was loaded onto a 2% agarose gel to check 

for  the  size  of  DNA‐fragments,  which  should  be  between  500  and  1000  base  pairs  for  IP 

experiments.  

Chromatin immunoprecipitation and PCR analysis: 

30µg of chromatin was used for IP and diluted 10‐fold with IP dilution buffer (0.01% SDS, 1.1% 

Triton  X  100,  1.2  mM  EDTA,  16.7  mM  Tris‐HCl  (pH  8.1),  167  mM  NaCl)  up  to  2ml.  80µl  of 

prepared beads coupled with anti‐PPARγ antibody were added and incubated overnight at 4°C 

(10  rpm).  On  the  following  day,  beads  were  washed  5  times  with  1  ml  RIPA  buffer  (50mM 

Hepes‐KOH  (pH  7.6),  0.5M  Lithiumchloride,  1mM  EDTA,  1%  NP40,  0.7%  Deoxycholic  acid, 

ddH2O),  2  times  with  TE  (10mM  Tris‐HCl  (pH=8.0),  1mM  EDTA)  and  eluted  with  2  x  200µl 

elution buffer (0.1M NaHCO3, 1% SDS). 16µl 5M NaCl was added to de‐crosslink and incubated 

for four hours at 65°C. 50µl of input DNA was mixed with 16µl 5M NaCl and 400µl elution buffer. 

Thereafter, 8µl EDTA, 16µl 1M Tris‐HCl and 2µl proteinase K were added to both samples and 

incubated  overnight  at  45°C.  DNA  was  recovered  by  phenol/chloroform  extraction  and 

ethanol/glycogen precipitation and resuspended in 100µl of H2O. Potential PPAR binding sites in 

the  FABP4  promotor  region was  identified  using  SybrGreen  quantitative  real‐time  PCR  using 

KAPA SYBR FAST qPCR master mix (Kapa Biosystems, no. KM4104). The location of the primers 

(see Table 8 for primer sequences and Figure 5  for primer locations) was chosen according to 

their  proximity  to  PPARγ  responsive  elements  (PPREs)  which  are  known  to  be  important 

binding sites for the transcription factor PPARγ in the FABP4 promotor region regulating FABP4 

expression. 

 

 

Page 60: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  45

Gene  Forward  Reverse FABP4­1  tttaagccagttcctctgagagg gaggtagaattgatggaccaatggFABP4­2  aagagtgtatacatacacagtcttga agaatggtaggcaggtttgcFABP4­3  tcttggcaagccacagtaataaatgc gcacttttgctgtgagatttcccFABP4­4  gctttgggaagtcattgctgatcc ccacaggatgctatgggagtgaFABP4­5  gagagggtatgtttcctcttcaacc gtacagcatagacatgcacacaccFABP4­6  gcagcacatgtgaacaacttctgtc ggtgtgtggtcacaagctaccMYT1  gaaccactggcagagaaagc gggtgagaggtgtcccag

Table 8: Primers used for qPCR to check binding of PPARγ to FABP4 promotor region. 

 

The  following  temperature  protocol  was  used.  Step  2,  3  and  4  were  repeated  40  times  with 

intermediate plate reading. The annealing temperature for the primer pairs FABP4‐1, FABP4‐5 

and FABP4‐6 was 59.8°C, for FABP4‐2, FABP4‐3 and FABP4‐4 61.8°C: 

1) 95°C ‐ 00:03:00 

2) 95°C ‐ 00:00:03 

3) 59.8°C or 61.8°C ‐ 00:00:30 

4) 72°C ‐ 00:00:01 

 

 

Figure 5: Graphical depiction of the human FABP4 promotor. Forward and reverse primers are  labeled  red.  Restriction  sites  of  different  restriction  enzymes  are  labeled  black.  PPARresponsive elements (PPREs) and CCAAT‐enhancer‐binding proteins (C/EBP) binding sites are labeled green. 

Page 61: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  46

6.21 TNFα stimulation of LECs  In  order  to  analyze  the  gene  expression  responsiveness  of  LECs  to  TNFα,  cells  were  starved 

overnight in EBM‐2/0.5% FCS and stimulated with or without 1, 10 and 20 ng/ml of TNFα (R&D; 

no. 210‐TA‐010)  for 6, 12 and 24 hours.  In neutralization experiments, LECs were cultured  in 

EBM‐2/0.5% FCS containing 10 ng/ml TNFα with or without 25µg/ml TNFα antibody (Abcam; 

no.  ab6671).  Altered  gene  expressions  induced  by  TNFα  were  checked  using  subsequent 

quantitative real‐time PCR and Western blotting. 

6.22 Scratch wounding assay  In  order  to  check whether  TNFα  has  an  influence  on wound  closure  of  LEC monolayers  and 

therefore also migration of LECs, we starved and stimulated LECs with or without 10ng/ml of 

TNFα for 24 hours. Thereafter, an artificial wound was created using a 200µl pipette tip. Non‐

adherent cells were washed away and wound closure was monitored with an inverted live cell 

microscope  (AxioVert  200M,  Zeiss)  by  taking  pictures  of  the  same  section  every  hour  for  24 

hours.  Wound  area  was  measured  using  AxioVision  4.7  and  calculated  as  amount  of  wound 

closure (in %) from timepoint t=0. 

6.23 Enyzme­linked Immunosorbent Assay (ELISA)  Supernatants  of  stimulated  LECs  were  collected  and  concentrated  20‐fold  from  2  ml  to 

approximately  100µl  using  centrifugal  filter  units  (Amicon  Ultra  ‐  0.5ml  10k  Ultracel  R,  no. 

UFC501024,  Millipore).  80µl  of  supernatants  were  used  for  coating  of  96‐well  ELISA  plates 

overnight and checked for the secretion of CXCL10. Primary anti‐CXCL10 antibody (no. AF‐266‐

NA, R&D Systems) was diluted 1:200  in PBS/1%BSA and 100µl was added  to each well. After 

incubation  for  2  hours  at  room  temperature,  the  plate  was  washed  three  times  with  300µl 

PBS/0.1% tween‐20 and 100µl of secondary antibody (diluted 1:3500  in PBS) was pipetted to 

each  well.  Secondary  antibody  solution  was  incubated  for  an  additional  hour  at  room 

temperature.  Thereafter,  plate  was  washed  and  50µl  of  substrate  solution  (3,3',5,5'‐

Tetramethylbenzidine  Liquid  Substrate,  no.  T4444,  Sigma)  was  added  and  incubated  for  20 

minutes under light protection. The reaction was stopped using Stop Reagent for TMB Substrate 

(no.  S5814,  Sigma).  The  color  development  was  read  at  450nm  in  a  microtiter  plate  reader 

(Synergy HT; Bio‐Tek) within 10 minutes. 

Page 62: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  47

 6.24 Macrophage adhesion assay  5*104  LECs  were  seeded  into  24  well  plates,  starved  overnight  in  EBM‐2/0.5%  FCS  and 

stimulated with or without 10 ng/ml TNFα for 24 hours. Additionally, LECs were pre‐incubated 

with  inhibitory  antibodies  to  CXCL10  (no.  AF‐266‐NA,  R&D  Systems)  and  VCAM‐1  (no.  1244, 

Immunotech). THP‐1 macrophages were grown in RPMI‐1640 medium containing 10% FCS (no. 

10108, Gibco) and 1% Pen/Strep (no. 15140, Gibco). Before addition to LEC monolayer, THP‐1 

macrophages were  labeled  using  Cell  trackerTM  green  CMFDA  (1:5000,  no.  C2925,  Invitrogen, 

Molecular  probes).  5*105  THP‐1  macrophages  (+  10ng/ml  PMA  (no.  P8139,  Sigma))  were 

diluted  in 500µl EBM‐2/0.5% FCS and added  to LECs monolayer. After 3 hours, non‐adherent 

cells were washed away and adherent THP‐1 cells were photographed using an inverted life cell 

microscope (AxioVert 200M, Zeiss). 

6.25 Macrophage transmigration experiment  2*104  LECs  were  seeded  on  the  underside  of  fibronectin‐coated  transwell  filters  (no.  3421; 

Corning) and grown to confluence. Then, cells were starved overnight in EBM‐2/0.5% FCS and 

stimulated with 10ng/ml TNFα (with or without an inhibitory antibody to TNFα or CXCL10) for 

24 hours. THP‐1 macrophages were grown in RPMI‐1640 medium containing 10% FCS and 1% 

Pen/Strep.  Before  addition  to  LEC  monolayers,  THP‐1  macrophages  were  labeled  using  Cell 

trackerTM green CMFDA (1:5000, no. C2925, Invitrogen, Molecular probes). 5*105 THP‐1 cells (+ 

10ng/ml  PMA)  were  diluted  in  500µl  EBM‐2/0.5%  FCS  and  added  to  the  upper  chamber. 

Fluorescence of transmigrated macrophages was measured in the lower chamber at 1, 3 and 6 

hours using a fluorescence plate reader (485‐530nm; Synergy HT, Bio‐Tek). 

6.26 Agarose spot assay  Wiggins  and  Rappoport  (2010)  presented  a  novel  chemotaxis  assay  based  on  the  invasion  of 

cells into agarose spots containing chemoattractants. Due to the fact that CXCL10 is secreted by 

LECs  upon  TNFα  stimulation,  we  wanted  to  know  if  supernatants  containing  CXCL10  are 

chemotactic for THP‐1 macrophages. Therefore, we applied this assay to our needs. 4% Agarose 

solution (no. 18300‐012, Gibco) was shortly boiled and diluted  in PBS  to obtain a 1% agarose 

solution. Thereafter, 20µl of agarose solution was mixed with 20µl of cell culture supernatants, 

which  was  previously  concentrated  (see  above).  Two  10  µl‐spots  of  agarose‐supernatant 

solution were pipetted in each well of a 24‐well plate and allowed to cool for 10 minutes at 4°C. 

THP‐1  macrophages  were  labeled  using  Cell  trackerTM  green  CMFDA  (1:5000,  no.  C2925, 

Page 63: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  48

Invitrogen, Molecular  probes),  diluted  in  RPMI1640/0.5%FCS/1% Pen/Strep,  stimulated with 

10ng/ml PMA and 500.000 cells were added to each well of the 24‐well plate. After 24 and 48 

hours, pictures were taken with an inverted live cell microscope (AxioVert 200M, Zeiss). 

6.27 LEC monolayer permeability assay and TEER measurements  2*104 LECs were seeded onto fibronectin coated transwell filters (no. 3421, Costar) and grown 

to confluence. Then, cells were starved overnight  in EBM‐2/0.5% FCS and 150µl FITC‐dextran 

solution (10µg/ml; no. FD70S‐100MG, Sigma) was applied to the upper and 500µl of medium to 

the  lower  chamber.  Diffused  FITC‐dextran  was  measured  spectroscopically  in  the  lower 

chamber  at  λ=485‐530nm  using  a  fluorescence  plate  reader  (Synergy  HT,  Bio‐Tek). 

Transendothelial  electric  resistance  (TEER)  measurements  were  done  using  Millicell®‐ERS 

electrode  (Millipore).  Resistance  was  measured  in  triplicates  and  calculated  as  per  cm2  of 

transwell  filters.  Transwell  filters  without  LECs  were  used  as  background  control  and  fresh 

medium was added to the wells prior to each measurement. 

6.28 Statistical methods and analysis  Evaluation  and  compilation  of  data  was  carried  out  by  using  Microsoft  Office  Excel®  2003. 

Significance was assessed using Student’s T‐Tests after determination of  the variance equality 

using  an  F‐Test.  Graphics  were  made  using  Adobe  Illustrator  CS3.  A  p‐value  of  <  0.05  was 

considered as statistically significant. * indicates a p‐value < 0.05, ** indicates a p‐value < 0.01, 

*** indicates a p‐value < 0.001. 

 

Page 64: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  49

  7. Results and Discussion 

7.1 Morphological features of diabetic skin  

Changes of vessel morphology and structure in T2DM have been reported earlier, especially for 

BVs. These include BM thickening due to increased expression and deposition of ECM proteins. 

Moreover,  important  features of  tissue morphology  in diabetic skin  include changes  in BV and 

LV density as well as associated signs of inflammation.  

In  the  first  part  of  my  thesis,  I  performed  an  extensive  histological  examination  of  dermal 

lymphatic capillaries in T2DM skin specimens to trace potential morphological alterations. 

 

7.1.1  Basement  membrane  morphology  of  small  blood  and  lymphatic capillaries in diabetic skin  

BM thickening is a common well described feature of diabetic BVs. Expression and deposition of 

ECM proteins like type IV collagen and laminin is generally increased in DM (Asselot‐Chapel et 

al., 1996; Roy et al., 1994, 1996). Around physiological normal LVs, ECM protein expression and 

deposition  is  generally  low  and,  if  present,  LVs  show no  regular  continuous  BM,  but  a  rather 

fragmented  BM‐like  structure  (see  also  section  2.1.4,  Vainionpää  et  al.,  2007;  Petrova  et  al., 

2002; Podgrabinska et al., 2002). Especially, the expression of some laminins was shown to be 

completely  absent  from  LECs  (Vainionpää  et  al.,  2007; Wigle  et  al.,  2002).  In  order  to  take  a 

closer  look  at  potential  BM  alterations  of  skin  lymphatic  capillaries  in  T2DM,  we  co‐stained 

paraffin sections of human skin of T2DM and normoglycemic patients with  the routinely used 

Periodic acid‐Schiff  (PAS) stain  that  labels  the glycoproteins  forming the BM,  together with an 

anti‐podoplanin antibody.  Figure 6 A  shows  the onionskinned‐like  thickening of BM  in BVs of 

human  diabetic  skin  (black  arrow)  compared  to  normoglycemic  skin,  while  LVs  were  PAS‐

negative  both  in  diabetic  and  normoglycemic  condition  (black  arrowheads).  By  staining 

consecutive  sections  of  diabetic  skin with  antibodies  directed against  podoplanin  and  smooth 

muscle  actin  (SMA),  lymphatic  capillaries  were  identified  due  to  their  negativity  for  smooth 

muscle actin (Figure 6 B; Tammela et al., 2005b). Next, we stained frozen sections of human skin 

with  anti‐laminin  antibody  to  visualize  eventual  deposition  alterations  of  this  ECM  protein 

(Figure 7). This staining showed a more prominent deposition of laminin in diabetic (Figure 7 A) 

Page 65: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  50

compared to non‐diabetic BVs (Figure 7 D), whereas laminin was neither expressed in diabetic 

(Figure 7 B) nor in non‐diabetic human skin LVs (Figure 7 E). 

 

Moreover,  double  labeling  with  antibodies  to  podoplanin  and  collagen  IV  showed  similar 

amounts  of  collagen  IV  expression  by  both,  LVs  and  BVs,  indicating  no  significant  expression 

difference between diabetic (Figure 8 A‐C) and normoglycemic skin (Figure 8 D‐F). 

Additionally,  we  performed  ultrastructural  analysis  of  LVs  using  electron  microscopy,  which  

 

 

Figure 6: BM morphology of blood and  lymphatic capillaries  in human diabetic skin. A:Double labeling with PAS and anti‐podoplanin antibody shows thickening of BM of diabetic BVs(black  arrow)  but  PAS  negativity  for  LVs  in  diabetic  and  normoglycemic  skin  (blackarrowheads).  B:  Staining  of  consecutive  sections  of  diabetic  skin  identified  SMA‐negative,  butpodoplanin‐positive vessels as a lymphatic capillary; Size bar: 20µm. 

Page 66: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  51

visualizes  cellular  alterations  like  membrane  and  organelle  changes  at  high  resolution. 

Longitudinal pictures from the LVs confirmed that there are no obvious morphological changes 

of  lymphatic capillaries in diabetic versus normoglycemic skin and that LVs of diabetic skin do 

not dispose of a BM lining (Figure 9). Vielleicht auch noch erwähnen: we did not trace luminal 

expansions of LVs or LEC vacuolisations with this detailed technique, indicating no grave lymph‐

edematous condition in the skin specimens.  

 

To  conclude,  in  contrast  to  early  observations  tracing modifications  of  lymphatic  vessel  basal 

lamina  in  T2DM  (Ohkuma,  1979)  and  aberrant  lymphatic  vessels  newly  grown  in  carcinomas 

(Vainionpää et al., 2007), we could not detect laminin and type IV collagen expression in T2DM. 

Figure 7: Laminin expression of diabetic and non­diabetic LVs and BVs. A‐F: Double labeling with antibodies against laminin (A, D) and podoplanin (B, E); Merged images: C, F; Upper panel: diabetic skin, lower panel: normoglycemic skin; Size bar: 20µm.  

Page 67: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  52

 

 

Figure  9:  Ultrastructural  analysis  of  LVs.  Electron  microscopy  of  LVs  of  diabetic  and normoglycemic skin revealed no changes in LEC morphology, e.g. appearance of BM; Lu: vessel lumen; Stars mark endothelial cell layer; Magnification: 10.000x. 

  

Figure  8:  Collagen  IV  expression  of  diabetic  and  non­diabetic  LVs  and  BVs.  AF:Double labeling with antibodies against collagen IV (A, D) and podoplanin (B, E) to showcollagen IV positive LVs in diabetic (upper panel) and normoglycemic skin (lower panel).Size bar: 20µm.  

Page 68: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  53

7.1.2 Basement membrane morphology of  lymphatic collectors  in diabetic skin  

In contrast to lymphatic capillaries, the larger collecting LVs are characterized by the presence of 

a BM and smooth muscle cell layers (Tammela et al., 2005b). We identified smooth muscle actin 

(SMA)‐positive LVs with PAS in order to show whether they are altered in the diabetic condition. 

However,  similar  to  lymphatic  capillaries,  lymphatic  collecting  vessels  did  not  show  any 

dramatic BM changes, as it is observed for diabetic BVs (Figure 10). 

 

Overall, from these histological data we conclude that there are neither prominent alterations in 

ECM protein deposition, nor morphological BM changes of  lymphatic capillaries and collecting 

LVs in the skin of T2DM patients. This finding is  in contrast to reports by Ohkuma (1979) and 

Kaufmann et al. (1980), where histochemical and ultrastructural changes of LVs were described 

in  the skin of patients with T2DM, namely  thickening of BM, dilatations of  intracellular spaces 

and dislocation of the lymphatic endothelium. All of these changes were accepted as proof for a 

diabetic lymphangiopathy. However, it is questionable if the described changes can be attributed 

to LVs, as BVs and LVs could not be distinguished at that time due to the lack of specific markers 

for lymphatic vessels.  

 

7.1.3 Increased lymphatic vessel density in the skin of T2DM patients  

However,  the  density  of  dermal  capillaries  itself  might  be  an  indicator  of  pathologic  vessel 

alterations,  as  it  was  described  in  lymphedema  or  acute  inflammation  (Seyama  et  al.,  2010; 

Kerjaschki et al., 2004). Hence, we quantified the density of LVs and BVs in the skin of diabetic 

versus normoglycemic patients by immunostaining of paraffin sections using antibodies against 

DARC to visualize BVs (Wick et al., 2008) and podoplanin for the detection of LVs (Breiteneder‐

Figure  10:  Lymphatic  collectors  in  diabetic  skin  do  not  show  any  morphological changes.  A:  Labeling  of  diabetic  skin  with  PAS  and  anti‐podoplanin  antibody;  B:  Double labeling of consecutive sections of diabetic skin with anti‐SMA and anti‐podoplanin antibodies to identify the indicated vessel as a lymphatic collecting vessel; Size bar: 20µm. 

Page 69: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  54

Geleff  et  al.,  1999).  Figure  11  shows  representative  images  of  human  diabetic  and 

normoglycemic  skin  containing  podoplanin‐positive  LVs  (left  panel,  black  arrows)  and DARC‐

positive  BVs  (right  panel).  Counting  and  quantification  of  LVs  revealed  a  significant  higher 

number  of  LVs  in  diabetic  compared  to  normoglycemic  skin  (p‐value  =  0.04).  Similarly,  BV 

density was  slightly  increased  in  diabetic  compared  to  normoglycemic  skin,  but  this was  not 

significant (p‐value = 0.7).  

Under physiological conditions LECs scarcely proliferate. De novo  lymphangiogenesis seems to 

occur  rather  upon  inflammatory  or  cancerous  signals  (Cueni  and  Detmar,  2008).  Hence,  we 

supposed that the higher LV density in the skin of T2DM patients could be the result of a chronic 

inflammatory stimulus present  in  the dermis. Moreover, higher LV density  in diabetic patients 

might  represent  a  reaction  of  the  skin  and  the  skin  vasculature  to  a  perturbed  metabolic 

situation.  In  agreement with  studies  in  human patients with metabolic  diseases,  enhanced LV 

densities  were  previously  detected  in  pathological  situations  such  as  cholesterol‐rich 

atherosclerotic  lesions  (Kholova  et  al.,  2011),  and  in  chronic  venous  insufficiency  ulcers 

(Fernandez  et  al.,  2011).  In  contrast  to  our  findings,  a  decreased  LV  density  associated  with 

reduced macrophage number and impaired wound healing was found in skin wounds (Saaristo 

et  al.,  2006)  and  in  a  corneal  suture  model  assay  (Maruyama  et  al.,  2007)  of  db/db  mice. 

However,  although  established  murine  models  were  used,  artificial  and  acute  wounds  were 

created  which  might  not  reflect  the  chronic  inflammatory  condition  of  the  skin  of  patients 

included  in  our  study.  Furthermore,  our  morphological  characterization  did  not  include  skin 

wounds or skin ulcera but rather (apparently) healthy skin.  

Other  studies  showed  that  accelerated  lymph  vessel  formation  was  crucial  for  resolution  of 

airway obstruction (Baluk et al., 2005), and to restore gut homeostasis in Crohn’s disease (von 

der  Weid  et  al.,  2011).  However,  in  chronic  inflammation,  LVs  are  suggested  to  be  more 

detrimental.  Several  reports  indicate  that  inflammatory‐driven  lymphangiogenesis  leads  to 

dysfunctional  vessel  formation,  representing  an  overproliferative  phenomenon  with  reduced 

lymph  drainage  capacity  and  enhanced  leakyness  (Baluk  et  al.,  2009,  Alexander  et  al.,  2010). 

Consistent  with  these  findings,  in  pancreatic  island  inflammation,  blocking  of 

lymphneoangiogenesis was  beneficial  for  island  transplants  (Yin  et  al.,  2011).  Although  some 

reports  indicate  beneficial  effects  of  increased  lymphangiogenesis  in  inflammatory  conditions 

(Saaristo et al. 2006, Maruyama et al. 2005), other publications report dysfunctional lymphatic 

vessels  especially  in  chronic  inflammatory  diseases  leading  to  prolongation  of  inflammation 

(Alexander  et  al.  2010;  von  der  Weid  et  al.,  2011).  Conclusively,  the  question  whether 

exaggerated  lymphatic  vessel  formation  might  be  dysfunctional  but  still  be  useful  to  resolve 

inflammation and function as a bystander in tissue remodelling and wound repair in T2D has to 

be explored.  

Page 70: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  55

Our findings also raise the question why BV density was not enhanced in T2DM skin. Consistent 

with  our  observations,  recent  studies  in  murine  models  showed  that  acute  inflammation 

strongly  enhances  lymph  vessel  density  but  not  that  of  blood  vessels  (Kim  et  al.,  2009; 

Huggenberger  et  al.,  2011).  Further,  reduced blood  vasculature was  observed  in  diabetic  skin 

due to hypoxia and ischemia, which regulate VEGF production (Galiano et al. 2004; Enholm et al., 

1997; Thangarajah et al., 2009). Further, CXCL10, which was upregulated in dLECs, is known to 

have an inhibitory effect on angiogenesis (Belperio et al., 2000; Bodnar et al., 2009). 

 

 Figure 11: LV and BV density in diabetic versus normoglycemic skin. Staining for LVs (black arrows) with anti‐podoplanin antibody and quantification of LV density shows higher LV density in diabetic skin; Right panel: Staining for BVs with anti‐DARC antibody and quantification of BV density  reveals  slightly  elevated  but  not  significantly  different  number  of  BVs  in  diabetic compared to normoglycemic skin; Size bar: 20µm.  

Page 71: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  56

Overall, our results suggest that lymphatic vessels follow a different proliferative stimulus than 

blood vessels in T2D skin, which might be due to significant differences in tissue‐residing factors 

and mechanisms that drive lymphangiogenesis versus angiogenesis. 

7.1.4 Increased macrophage infiltration in diabetic skin  

Considering the above mentioned murine wound healing study in db/db mice (Maruyama et al., 

2005), we evaluated the number of CD68‐ positive macrophages in normoglycemic and diabetic 

human skin. Macrophage infiltration, measured as the number of macrophages per 100µm2 was 

highly pronounced in diabetic compared to normoglycemic human skin, representing a 3.5‐fold 

increase (Figure 12, p‐value < 0.001).  

Macrophages  produce  growth  factors which  in  turn  support  re‐epithelialization  and  promote 

angiogenesis  during  wound  healing.  When  they  are  missing,  retarded  of  wound  healing  is 

observed  (Martin  &  Leibovich,  2005).  The  process  of  wound  healing  is  strongly  disturbed  in 

non‐healing  diabetic  wounds  (Mahdavian  Delavary  et  al.,  2011).  There  is  retarded  wound 

healing when macrophages show increased cytokine production and are dysfunctional (Khanna 

et al., 2010). Pierce et al. (2001) report macrophage dysfunction which is characterized by the 

enhanced production of  inflammatory cytokines. Moreover, Zampell et al. (2012) show a more 

than  3‐fold  increase  in  macrophage  number  in  lymph  stasis,  a  condition  of  lymphatic  vessel 

dysfunction. Anti‐TNFα treatment of skin wounds depleted macrophages and increased wound 

healing in diabetic mice (Goren et al., 2007). Hence, increased macrophage infiltration could be a 

sign of enhanced production of  inflammatory cytokines, but also of  their reduced clearance by 

newly grown, but dysfunctional lymphatic vessels.   

 

Figure  12:  Macrophage  infiltration  in  normoglycemic  versus  diabetic  human  skin.Immunohistochemical  labeling  of  human  skin  sections  with  anti‐CD68  antibody  andquantification of macrophage number per 100µm2; Size bar: 20µm. 

Page 72: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  57

7.1.5 Macrophages produce vascular endothelial growth factors  

As increased macrophage tissue infiltration is a critical sign of skin inflammation, we next aimed 

to analyze their potential of production angiogenic and proinflammatory substances, supporting 

a pro‐lymphangiogenic and pro‐inflammatory milieu in the skin of T2DM patients. 

It is well accepted that macrophages can produce vascular endothelial growth factors (VEGFs), 

e.g. VEGF‐A and VEGF‐C, and thereby are able to drive lymphangiogenesis (Machnik et al., 2009; 

Kerjaschki, 2006; Maruyama et al., 2005).  

 

 

 

 

 

Figure  13:  VEGF­C  and  VEGF­A  expression  of  infiltratin  skin  macrophages.  A:  Doublelabeling of human diabetic (upper panel) and normoglycemic (lower panel) skin with antibodiesto CD68 and VEGF‐C showing that macrophages in diabetic and normoglycemic skin are able toexpress  VEGF‐C;  Size  bar:  20µm.  B:  Double  labeling  of  human  diabetic  (upper  panel)  andnormoglycemic  (lower  panel)  skin  with  antibodies  to  CD68  and  VEGF‐A  showing  thatmacrophages in diabetic and normoglycemic skin are able to express VEGF‐A; Size bar: 20µm. 

Page 73: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  58

Therefore, we  evaluated VEGF‐C  and VEGF‐A  expression by  infiltrating  tissue macrophages  in 

human  diabetic  and  normoglycemic  skin.  As  shown  in  Figure  13  A,  both,  'diabetic'  and 

'normoglycemic'  macrophages  were  expressing  VEGF‐C,  which  is  a  strong  lymphangiogenic 

factor  (Jussila  and  Alitalo,  2002).  Similarly,  VEGF‐A  production  was  visible  in  infiltrating 

macrophages of diabetic and normoglycemic skin (Figure 13 B).  

Quantitative analysis (Figure 14) revealed that 76% and 90% of macrophages expressed VEGF‐C 

in  nondiabetic  and  diabetic  skin,  respectively  (p‐value  =  0.11).  Further,  85%  and  82%  of 

macrophages expressed VEGF‐A in non‐diabetic and diabetic skin, respectively (p‐value = 0.31).  

 

 

Due to the fact that there was a highly increased number of infiltrating macrophages in diabetic 

skin, which produced  these  (lymph‐)angiogenic  factors, we assumed  the generation of a  steep 

VEGFA  and  VEGF‐C  interstitial  tissue  gradient,  which  could  be  a  potent  driver  of 

lymphangiogenesis in the skin of T2DM patients. Due to this result, it is tempting to hypothesize 

that  increased  macrophage  infiltration  could  be  directly  involved  in  the  generation  of  the 

strongly increased LV density and of the slightly elevated BV number in T2DM skin. 

 

 

 

 

Figure  14: Quantitative  analysis  of  VEGF­C  and  VEGF­A  production  by  infiltrating  skinmacrophages. 

Page 74: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  59

7.1.6 Increased TNFα levels in human diabetic skin  

In  skin  diseases,  e.g.  psoarisis,  TNFα  expression  has  been  identified  as  a  key  agent  driving 

persistent  inflammation,  and  its  expression  was  mainly  assigned  to  T‐cell  and  macrophages 

(Bonifati and Ameglio, 1999). Further,  low grade vascular  inflammation and elevated  levels of 

pro‐inflammatory cytokines are predictive  for the development of cardiovascular diseases and 

T2DM  (Haddy  et  al.,  2003).  Especially,  TNFα  levels  are  increased  in  subjects with  endothelial 

dysfunction,  as  this  pro‐inflammatory  cytokine  downregulates  endothelial  NO  expression 

(Yoshizumi  et  al.,  1993)  and  causes  acute  local  vascular  inflammation  (Chia  et  al.,  2003). 

Moreover,  it  is  well  established  that  TNFα  levels  are  significantly  increased  in  wounds  of 

diabetic mice (Goova et al., 2001). In order to trace if TNFα expression is  increased in diabetic 

skin,  we  performed  respective  immunofluorescence  stainings.  As  shown  in  Figure  15,  highly 

increased  levels  of  TNFα  were  detected  in  the  skin  of  T2DM  patients.  In  particular,  diabetic 

vasculature (white arrows) and epidermis (white arrowheads) as well as cellular structures in 

between  exhibited  the  strongest  TNFα  expression.  These  data  further  strongly  suggested  a 

possible functional role of TNFα in the skin of type 2 diabetic patients. 

 

 

Figure 15: TNFα expression  in  skin.  Human  diabetic  (upper  panel)  and  non‐diabetic  (lowerpanel)  skin  was  stained  with  anti‐TNFα  antibody.  Increased  staining  could  be  found  invasculature and epidermis of diabetic skin. 

Page 75: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  60

7.1.7 TNFα production by CD68+ macrophages  

Activated macrophages are not only  crucial  regulators of wound healing processes  in diabetic 

skin  but  also  an  important  source  of  TNFα  (Goren  et  al.,  2007).  Immunofluorescent  double 

staining  with  anti‐CD68  and  anti‐TNFα  antibody  reconfirmed  the  increased  macrophage 

infiltration in diabetic (Figure 16 A) compared to normoglycemic skin (Figure 16 B) and showed 

that these macrophages expressed TNFα (magnified inlets). It has been shown that macrophages 

contribute to diabetes‐associated impaired wound healing mainly by producing TNFα (Goren et 

al.,  2007).  Dysfunction  of  cutaneous  macrophages  combined  with  their  reduced  efferocytosis 

(i.e.  their  phagocytosis  capacity)  contributes  to  impaired  resolution of  local  inflammation  and 

complicated wound healing (Khanna et al., 2010). Our data suggested that a pro‐inflammatory 

milieu  in  human  diabetic  skin  is  generated  in  part  by  macrophages  via  production  of  TNFα, 

which in turn could have a detrimental impact on wound healing processes (Eming et al., 2007; 

Goren et al., 2007) as well as migration of immune cells to the afferent LVs (Johnson et al., 2006). 

All  in  all,  enhanced  levels  of  dysfunctional macrophages  seem  to  contribute  to  disturbed  skin 

wound healing and susceptibility to skin infections, which was underlined by our data. 

Figure 16: Production of TNFα by macrophages. Human diabetic (A) and non‐diabetic (B) skin was stained with anti‐CD68 and anti‐TNFα antibody. Increased macrophage infiltration could bedetected  in  diabetic  skin  as  well  as  TNFα  production  by  these macrophages  (A,  lower  panel), whereas no expression was found in macrophages of normoglycemic skin (B, lower panel). 

Page 76: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  61

To  summarize  the  first  part  of  my  thesis,  the  morphological  findings  we  retrieved  from  the 

immunohistochemical  analysis  of  diabetic  and  normoglycemic  skin  showed  increased  lymph 

vessel density, accompanied by a strong macrophage infiltration, which produced on one hand, 

vascular  endothelial  growth  factors,  and,  on  the  other  hand  the  pro‐inflammatory  cytokine 

TNFα. 

However,  no  dramatic  differences  where  found  regarding  deposition  of  extracellular  matrix 

proteins,  including  laminin and collagen  IV. Additionally, on an ultrastructural  level, no BM or 

cellular alterations could be traced for LVs in T2DM skin. Similarly to small LVs, also the bigger 

collecting LVs showed no significant changes when compared  to BVs of T2DM skin, which are 

characterized by an increased staining for smooth muscle actin. 

Altogether,  these  findings  suggest  that,  although  LV  did  not  exhibit  a  form  of  diabetic 

lymphangiopathy,  as  it  was  reported  earlier,  the  dermis  of  T2D  patients  manifests  dramatic 

tissue remodelling processes characterized by enhanced LV density and increased macrophage 

infiltration, supporting the pro‐inflammatory milieu and predisposing for recurrent infections. 

Page 77: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  62

7.2 Ex vivo isolation of LECs from human skin  

Besides  the  fact  that  we  observed  significantly  increased  lymphangiogenesis  associated  with 

enhanced  macrophage  infiltration  in  diabetic  skin,  there  were  no  strong  morphological 

differences evident. However, we hypothesized that dermal LECs might reflect the altered load 

of metabolites and retarded wound healing of T2DM skin on a molecular level. We expected that 

we could answer the question whether complications seen in T2DM skin may be related to LV 

malfunction, because this  leads  to wound healing defects (Saaristo et al., 2006), contributes  to 

local infection (Rockson, 2001) and has an important impact on lipid transport and metabolism 

(Harvey  et  al.,  2005).  This  potential  LV  dysfunction,  together  with  signs  for  enhanced 

lymphangiogenesis,  should  become  evident  by  deregulation  of  respective  gene  expression 

patterns.  

Therefore, the second aim of my thesis was the analysis of the differential transcriptome of LECs 

derived  from T2DM patients and normoglycemic controls.  I  isolated LECs  from human skin of 

four diabetic and four non‐diabetic patients using a combined enzymatic‐mechanical protocol as 

previously established and described  (Wick et  al., 2007, 2008; Kriehuber et  al., 2001,  see also 

Material and Methods). The clinical characteristics of included patients are described in Table 3. 

As shown in Figure 17 A, cells were sorted into LECs (CD31+Podoplanin+, green dots) and BECs 

(CD31+Podoplanin‐, red dots). Reanalysis after sorting demonstrated the enrichment of CD31‐ 

and podoplanin‐positive LECs (in %) in the upper right quadrant, suggesting isolation of highly 

pure populations of CD31/podoplanin double‐positive cells (Figure 17 B). 

7.2.1 Quality control of isolated LECs  

In order to exclude leukocyte contaminations, isolated cells were additionally stained with PC5‐ 

labeled CD45. As shown in Figure 18, the percentage of CD45+ cells after sorting was less than 

one percent in all cell preparations, indicating nearly complete absence of leukocytes. 

To stay as close as possible to the  in vivo situation,  immediately after sorting, cells were lyzed, 

RNA was isolated and transcribed into cDNA. Using RT‐PCR, non‐diabetic (ndLECs) and diabetic 

(dLECs) LECs were  shown  to be positive  for  lymphatic markers  von Willebrand  factor  (vWF), 

podoplanin and PROX­1, but negative for keratin, suggesting exclusion of keratinocytes (Figure 

19 A). Quantitative realtime PCR analysis confirmed similar expression levels of podoplanin in 

dLECs and ndLECs (Figure 19 B). Altogether,  these data confirmed high purity of  isolated LEC 

populations and exclusion of keratinocyte and leukocyte contaminations. 

 

Page 78: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  63

Figure 17: Fluorescent activated cell sorting of ex vivo  isolated LECs and BECs.  LECs andBECs were sorted by fluorescence using antibodies to podoplanin and CD31. BECs were CD31+and podoplanin‐  (A,  red  dots), while  LECs were  double  positive  for  CD31  and  podoplanin  (A,green dots). B: Post‐sort analysis shows accumulation of double positive LECs in the upper rightquadrant.  X‐axis:  CD31  (FITC‐labeled)  intensities,  y‐axis:  podoplanin  (PE‐labeled)  intensities;ndLECs: non‐diabetic LECs, dLECs: diabetic LECs. 

Figure 18: Exclusion of leukocyte contaminations. Staining of cell populations with anti‐CD45staining  was  done  in  order  to  exclude  leukocyte  contaminations.  Post‐sort  analysis  showsexclusion of CD45 positivity in isolated cells. 

Page 79: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  64

  

 

 

 

 

In order to obtain reliable microarray data and due to the low amount of RNA starting material, 

two  RNA  amplification  steps  were  amended  (Wick  et  al.,  2004)  and  checked  by  Agilent 

Bioanalyzer 2100 measurements. Figures 20 and 21 show the Agilent graphs of RNA amounts 

before and after two amplification rounds. This analysis further visualized the high quality of the 

amplified  RNA,  excluded  the  possibility  of  degradation  artefacts  and  guaranteed  sufficient 

amounts of RNA for subsequent array hybridization. 

 

Figure  19:  Quality  control  of  isolated  LECs. A:  Using  RT‐PCR  and  gel  electrophoresis,transcripts  for  actin,  vWF,  keratin,  podoplanin  and  prox1 were  detected;  ndLEC:  non‐diabeticLEC,  dLEC:  diabetic  LEC,  +:  positive  control,  ‐:  negative  control.  B:  Quantitative  realtime  PCRconfirmed  expression  of  podoplanin  in  isolated  cells,  which  showed  no  significant  difference;dLEC: diabetic LEC, ndLEC: non‐diabetic LEC. 

Page 80: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  65

 

Figure  20:  Amplification  check  of  RNA  isolated  from  dLECs.  Left  panel:  RNA  amount  of isolated  dLECs  before  amplification;  right  panel:  RNA  amount  of  isolated  dLECs  after  two amplification  rounds.  The  x‐axis  of  depicted  graphs  represents  the  runtime  (s),  the  y‐axis represents the fluorescence units [FU].  

 

 

 

Page 81: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  66

 

Figure 21: Amplification  check  of RNA  isolated  from ndLECs.  Left  panel:  RNA  amount  of isolated  ndLECs  before  amplification;  right  panel:  RNA  amount  of  isolated  ndLECs  after  two amplification  rounds.  The  x‐axis  of  depicted  graphs  represents  the  runtime  (s),  the  y‐axis represents the fluorescence units [FU].  

 

Page 82: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  67

7.3  Bioinformatical  analysis  of  diabetic  versus  non­diabetic  LEC transcriptomes  

Files  obtained  from  GCOS  (.cel)  were  used  to  analyze  significant  changes  between  gene 

expression profiles of different patients and groups (Wick et al., 2004, 2007, 2008). Only genes 

with postnormalization  values  above  the BioB  control were  included  in  the  calculations. After 

background correction and normalization (Figure 22) using the robust multichip average (RMA) 

method,  datasets were  analyzed  using  two‐sample  t‐test  and  relative  variance method  (RVM, 

Stokić  et  al.,  2006)  (Figure  23).  The  RVM  method  was  developed  to  trace  gene  expression 

alterations  in  small  sample  entities,  and  is  was  successfully  tested  to  identify  important 

endothelial  markers  like  podoplanin  (Stokić  et  al.,  2006).  Further,  a  p‐value  <  0.05  was 

considered to identify genes significantly different between dLECs and LECs.  

 

 

 

Figure 23: Bioinformatical analysis using Student's t­test and RVM. Bioinformatical analysis of diabetic versus non‐diabetic LECs using RVM and t‐test. Both methods in one graph are shown on the right.  

 

Figure  22:  Normalization  of  microarray  chip  data. Normalization  was  done  by  RobustMultichip Average (RMA) using Bioconductor software in R. 

Page 83: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  68

Further,  y‐comparison  of  LEC‐specific marker  genes  and  of  the  genes  differentially  regulated 

between dLECs and dBECs from the same patients was performed. Redundant annotation terms 

were identified and omitted. All data has been deposited in NCBIs Gene Expression Omnibus and 

are  accessible  through  GEO  accession  number  GSE38396 

(http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE38396).  Gehört  schon  hierher: 

Hierarchical clustering was used to analyze expression profiles of the different samples and was 

carried  out  using  Euclidean  distance  metric  and  centroid  linkage  rule  analysis  with  the 

GeneSpringGX  (Ambion)  software.  Data  were  represented  as  a  dendrogram  with  the  closest 

branches of the tree representing arrays with similar gene expression patterns. The expression 

profiles of  the eight different array samples was analyzed, which  illustrated a broad similarity 

among  the  four  dLEC  arrays  or  the  four  LEC  arrays  from  normoglycemic  controls  (ndLECs; 

Figure 24). 

The  t‐test  and  RVM  analysis  retrieved  a  list  of  180  genes  that  were  differentially  expressed 

between dLECs and ndLECs. While 49 genes were upregulated, 131 genes were downregulated 

in  dLECs  compared  to  ndLECs.  This  180  significantly  deregulated dLEC  candidate  genes were 

the basis of a gene network that was assumed to be of functional relevance. The Gene Ontology 

Browser  available  on  the  affymetrix  website  (www.affymetrix.com)  was  used  as  sources  to 

classify  genes  according  to  functionality  context.  Ingenuity  Pathway  Analysis  (IPA)  software 

(Ingenuity  Systems,  Redwood  City,  CA,  USA)  was  used  to  identify  cellular  networks  that 

statistically  fit  to  a  given  input  gene  list  and  according  expression  values.  The  association  of 

genes with particular functions, pathways and diseases was analyzed according to their scores. A 

Canonical pathway analysis  identified pathways which were most significant  to  the  input data 

set. The significance value associated with functions and pathways is a measure of how likely it 

is that genes from the data set file participate in that function. The significance is then expressed 

as a p‐value, which  is calculated by  the right‐tailed Fisher’s exact  test. An extensive NCBI GEO 

and PubMed  ‘in  silico’  research was  amended  to  retrieve  information  on  gene  expression  and 

relevance for LEC biology. 

Page 84: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  69

 

Figure  24:  Hierarchical  cluster  analysis: LECs  derived  from  four  T2DM  patients  exhibit  aunique gene expression pattern. Hierarchical  clustering of  the  four LEC versus  four dLEC geneexpression data (A) and of the 180 differentially expressed genes (FC>1.5) (B). Combined entityand condition trees are shown. On the y‐axis, an entity tree was generated by grouping the probesets  based  on  the  similarity  of  their  expression  profiles.  On  the  x‐axis,  a  condition  tree  wasgenerated showing the relationship between the samples. 

Page 85: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  70

7.3.1  Identification  of  deregulated  pathways  and  gene  functions  using Ingenuity Pathway Analysis  

In order to identify functional gene patterns offering insight into the role of LECs in T2DM, the 

differentially  expressed  genes were  grouped  according  to  their  biological  functions  using  the 

NetAffx  tool  as  well  as  Ingenuity  Pathway  Analysis  (IPA).  IPA  revealed,  which  canonical 

pathways and which gene  functions are mainly altered  in dLECs versus ndLECs. The top 10 of 

the canonical pathways (Figure 25) and gene functions (Figure 26) were visualized using a bar 

chart. Genes associated with these pathways and functions are highlighted in the Tables 9 and 

10. Combining all in silico analyses, this led to the establishment of four overrepresented themes 

in  dLECs,  comprising  (A)  defense  response  and  inflammation,  (B)  tissue  remodeling  and  cell 

motility,  (C)  lymphangiogenesis  and  cell  fate  regulation  and  (D)  lipid  handling  and  small 

molecule biochemistry (Table 11). Furthermore, the dLEC transcripts were highly enriched for 

“plasma  membrane”  (44  transcripts)  and  “cell  periphery  and  extracellular  space”  (71 

transcripts)  compartments  (both  P  <  0.001),  highlighting  an  extensive  modulation  of  dLEC 

molecules at the environmental interface. 

Conclusively, these unbiased, computer‐based methods identified a dLEC transcriptome strongly 

distinct  from  that  of  LECs  under  physiologic  condition.  This  transcriptome  mirrored  the 

phenotypic changes of T2DM skin on a molecular level and highlighted an involvement of LVs in 

T2DM  skin  alterations.  In  the  following  sections,  I  will  present  the  main  candidates  of  the 

deregulated  gene  themes  and  will  discuss  implications  of  their  deregulation  in  LECs. 

Page 86: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  71

Figure  25:  Pathway  analysis  using  IPA. For  the  canonical  pathway  categorization  of  alteredexpressed genes, IPA software was used to calculate a p‐value (set ≤ 0.05, shown as blue bars) byFishers  exact  test,  determining  the  probability  with  which  a  set  of  genes  is  associated  with  aknown  biological  function  or  pathway.  The  ratio  (yellow  squares)  represents  the  number  ofdifferentially  expressed  genes  from  the  dataset  divided  by  the  total  number  of  genes  thatconstitute that canonical pathway. 

Figure 26: Deregulated cellular functions in dLECs. Rankings of the top twenty Molecular andCellular Functions. 

Page 87: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  72

 

Ingenuity Canonical Pathways 

p­value (­log)  Ratio  Molecules 

Fatty Acid Biosynthesis 1,89E+00

 3,92E‐02  ACACB, SLC27A3 

IGF­1 Signaling  1,84E+00  4,9E‐02  JUN, IGFBP3, SFN, CYR61, PRKAR1A 

Complement System  1,84E+00 8,33E‐02 SERPING1, C1S, CD55IL­10 Signaling  1,80E+00 5,71E‐02 IL1R2, SOCS3, JUN, IL1RN

Acute Phase Response Signaling 

1,75E+00  3,93E‐02  SOCS3, SERPING1, HP, JUN, IL1RN, C1S, FGG, MIF 

MIF­Regulation of Innate Immunity 

1,67E+00  6,52E‐02  PLA2G4A, JUN, PTGS2 

Role of Macrophages, Fibroblasts and 

Endothelial Cells in Rheumatoid Arthritis 

1,64E+00  2,79E‐02 IL1R2, SOCS3, VCAM1, HP, JUN, F2RL1, IL1RN, C1S, MMP1, 

WNT5A 

Leukocyte Extravasation Signaling 

1,58E+00  3,57E‐02  VCAM1, CLDN1, EZR, CD44, MMP2, RASSF5, MMP1 

LPS/IL­1 Mediated Inhibition of RXR 

Function 1,57E+00  3,24E‐02 

IL1R2, JUN, SLC27A2, FABP4,SLC27A3, PAPSS2, CHST15 

 VDR/RXR Activation  1,45E+00 5E‐02 CXCL10, IGFBP3, VDR, HSD17B2

Glucocorticoid Receptor Signaling 

1,34E+00  2,84E‐02  IL1R2, VCAM1, JUN, HSPA1A, IL1RN, ANXA1, FGG, MMP1 

IL­8 Signaling  1,30E+00  3,19E‐02  CDH1, VCAM1, NOX4, RND3, MMP2, PTGS2 

Atherosclerosis Signaling  1,19E+00 3,54E‐02 PLA2G4A, VCAM1, IL1RN, MMP1Glycosphingolipid Biosynthesis ­ Neolactoseries 

1,19E+00  2,9E‐02  GLT8D2, GGTA1  

MIF­mediated Glucocorticoid Regulation 

1,10E+00  5,26E‐02  PLA2G4A, PTGS2  

HIF1α Signaling  1,10E+00 3,64E‐02 JUN, SLC2A1, MMP2, MMP1Inhibition of 

Angiogenesis by TSP1 1,08E+00  5,56E‐02  JUN, SDC1 

 Rac Signaling  1,07E+00 3,23E‐02 NOX4, JUN, CD44, IQGAP1

PXR/RXR Activation  1,06E+00 3,3E‐02 SCD, PAPSS2, PRKAR1AO­Glycan Biosynthesis  1,05E+00 4,26E‐02 GLT8D2, GALNTL2 

Table 9: Genes associated with most deregulated pathways (see Figure 25) as evaluated by IPA analysis. 

 

 

 

 

 

 

 

Page 88: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  73

Category  p­value (­log)  Molecules 

Cellular Growth and Proliferation 

1,61E‐13‐4,75E‐03 

 

IL13RA2, SERPINB2, CXCL10, GNA15, EZR, SERPINB7, CLCA2, SLC2A1, MT1X, DCN, RRM2, FGFR2, MMP2, ANLN, KRT10, 

FGFBP1, AREG, CDH1, CTSB, SPINT2, HHEX, CYR61, UNC5A, PHLDA2, CD55, BCL11B, TGM2, DSG3, JUN, HK2, F2RL1, TFAP2A, 

FABP4, KRT2, RASSF5, VDR, CAV2, IGF2BP3, TGFBI, KRT17, MAP7, DEFB1, FOLR1, PLA2G4A, TMEM49, CD44, PTGS2, SFN, 

GATA3, CXADR, NEDD9, ID4, CD81, SOCS3, SLC7A11, HSPA1A, GPNMB, BIRC5, FGFR3, CRIP1, CCL27, RORA, POSTN, SERPINB5, PTX3, KRT16, NOX4, H19, RRAD, LMNA, AURKA, IRF6, STC1, RAPGEF2, IL1RN, 

IGFBP3, LOX, KIAA0101, ANTXR1, RARRES1, APP, GPC1, ANXA1, S100A8, VCAM1, 

TFAP2C, PKP1, TP63, EDNRB, CGA, IL20RB, MST4, TPT1, HOPX, KLRK1, KLF5, TNFRSF19, PRKAR1A, WNT5A 

Cellular Movement 2,54E‐11‐5,13E‐03 

 

CD81, SOCS3, IL13RA2, SERPINB2, CXCL10, CCL27, EZR, CXCL14, POSTN, SERPINB5, 

LYVE1, KRT16, PODXL, SLC2A1, DCN, ANLN, MMP2, AURKA, KRT10, S100A2, AREG, STC1, CDH1, RND3, IL1RN, CTSB, IGFBP3, 

SPINT2, HHEX, CYR61, LY6D, LOX, SERPING1, 

CDC20, UNC5A, CD55, RARRES1, IQGAP1, APP, TGM2, JUN, TFAP2A, F2RL1, ANXA1, TOP2A, FABP4, PTPRZ1, S100A8, VDR, 

RASSF5, MMP1, NUSAP1, TFAP2C, VCAM1, TP63, SDC1, TGFBI, EDNRB, RACGAP1, CD69, DEFB1, LRRC15, LYPD3, TPT1, 

PLA2G4A, HP, KLRK1, KLF5, CD44, PTGS2, GATA3, NEDD9, 

WNT5A, PRKAR1A 

Cell­To­Cell Signaling and Interaction 

1,43E‐09‐5,13E‐03 

 

CD81, MRC1, HSPA1A, MMRN1, FGG, SERPINB2, GP‐NMB, CXCL10, FGFR3, CCL27, CXCL14, POSTN, SERPINB5, KRT1, CLCA2, PTX3, KRT16, PODXL, DCN, FGFR2, MMP2, FGFBP1, AREG, STC1, CDH1, RND3, IL1RN, SLC7A2, IGFBP3, CYR61, ANK3, LY6D, LOX, ZDHHC2, SERPING1, CD55, IQGAP1, APP, TGM2, GPC1, DSG3, F2RL1, ANXA1, KRT2, PTPRZ1, S100A8, RASSF5, MMP1, DSC3, 

VCAM1, TFAP2C, TP63, SDC1, TGFBI, CD69, PERP, LRRC15, CGA, IL20RB, LYPD3, TPT1, CLDN1, LGALS7B, KLRK1, CD44, PTGS2, 

GATA3, NEDD9, CXADR, WNT5A 

Tissue Development 1,43E‐09‐5,13E‐03 

CD81, C4ORF31, MMRN1, FGG, SERPINB2, GPNMB, CXCL10, FGFR3, 

Page 89: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  74

POSTN, SERPINB5, CLCA2, PTX3, KRT16, NOX4, PODXL, DCN, FGFR2, MMP2, AREG, IRF6, STC1, CDH1, RND3, IL1RN, IGFBP3, HHEX, DSC1, CYR61, LY6D, ZDHHC2, 

SERPING1, CD55, IQGAP1, APP, TGM2, DSG3, JUN, TFAP2A, F2RL1, ANXA1, PTPRZ1, S100A8, RASSF5, DSC3, TFAP2C, VCAM1, SDC1, TP63, TGFBI, KRT17, PERP, OVOL1, PLA2G4A, CLDN1, LGALS7B, KLF5, CD44, PTGS2, GATA3, NEDD9, CXADR, PRKAR1A, 

WNT5A 

Cellular Development 1,81E‐09‐5,13E‐03 

 

CD81, SOCS3, CSTA, SLC7A11, HSPA1A, IL13RA2, SERPINB2, BIRC5, CXCL10, FGFR3, GNA15, RORA, EZR, POSTN, NOX4, RRAD, DCN, HBA1, BASP1, LMNA, FGFR2, MMP2, 

AURKA, KRT10, AREG, STC1, IRF6, BHLHE41, CDH1, RND3, IL1RN, IGFBP3, CTSB, HHEX, DSC1, CYR61, SCD, LOX, 

UNC5A, CD55, ANTXR1, BCL11B, SPRR1A, APP, TGM2, GPC1, JUN, F2RL1, TFAP2A, ANXA1, TOP2A, PTPRZ1, FABP4, S100A8, RASSF5, VDR, MMP1, TFAP2C, VCAM1, SDC1, TP63, EDNRB, TGFBI, RACGAP1, 

CD69, MAP7, PERP, OVOL1, FOLR1, MST4, TPT1, PLA2G4A, HOPX, KLRK1, KLF5, 

TMEM49, CD44, PTGS2, SFN, GATA3, CXADR, ID4, WNT5A, PRKAR1A 

Dermatological Diseases and Conditions 

 

6,06E‐09‐5,13E‐03 

 

CD81, SCD, LOX, SOCS3, SERPING1, SRD5A1, CD55, APP, SERPINB2, BIRC5, DSG3, JUN, CCL27, TOP2A, KRT2, S100A8, VDR, KRT1, MMP1, KRT14, KRT16, PKP1, TP63, SDC1, EDNRB, KRT17, C1S, PERP, LMNA, MMP2, KRT10, DSG1, AREG, CDH1, CLDN1, KRT5, 

IL1RN, PTGS2, DSC1 

Hair and Skin Development and  

Function  

1,04E‐08‐5,13E‐03 

 

CSTA, SLC7A11, UNC5A, HSPA1A, ANTXR1, SPRR1A, BIRC5, TGM2, DSG3, ANXA1, KRT2, VDR, KRT1, RIPK4, KRT14, KRT16, TP63, EDNRB, TGFBI, KRT17, OVOL1, KRT10, 

IL20RB, MST4, AREG, IRF6, KRTDAP, CDH1, CLDN1, KRT5, FOXQ1, PTGS2, SFN, CYR61, 

KRT15, TNFRSF19, ID4, WNT5A 

Organ Development 1,04E‐08‐5,13E‐03 

 

LOX, SOCS3, BCL11B, SPRR1A, APP, FGFR3, TGM2, RORA, KRT2, VDR, KRT1, RIPK4, KRT14, KRT16, PODXL, VCAM1, TFAP2C, TP63, EDNRB, KRT17, TLL1, FGFR2, MMP2, CGA, OVOL1, KRT10, AREG, IRF6, KRTDAP, HOPX, CDH1, HP, KRT5, CLDN1, FOXQ1, HHEX, PTGS2, SFN, KRT15, CXADR, TNFRSF19, ID4, PRKAR1A, WNT5A 

Cell Death 2,82E‐08‐5,13E‐03 

 

CD81, CSTA, SOCS3, SLC7A11, HSPA1A, SERPINB2, BIRC5, CXCL10, FGFR3, CCL27, EZR, SERPINB5, CLCA2, NOX4, SLC2A1, DCN, MT1X, HBA1, RRM2, LMNA, FGFR2, ANLN, MMP2, AURKA, TXNL4B, DSG1, AREG, STC1, 

CDH1, RAPGEF2, RND3, IL1RN, CTSB, 

Page 90: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  75

IGFBP3, CYR61, SCD, RAB25, UNC5A, SRPX, CDC20, PHLDA2, CD55, ANTXR1, BCL11B, SPRR1A, APP, TGM2, AQP3, GPC1, DSG3, JUN, HK2, F2RL1, TFAP2A, ANXA1, TOP2A, PTPRZ1, S100A8, VDR, RASSF5, MMP1, VCAM1, TFAP2C, SDC1, TP63, EDNRB, 

TGFBI, CD69, PERP, DEFB1, CGA, SH3BP5, MST4, TPT1, PLA2G4A, CHP2, KLRK1, KLF5, 

CD44, PTGS2, SFN, NEDD9, TNFRSF19, ID4, 

WNT5A, PRKAR1A 

Inflammatory Disease 4,99E‐07‐5,13E‐03 

 

CD81, MRC1, SOCS3, SLC7A11, HSPA1A, IL13RA2, CA12, SNX7, PCDH7, C4ORF31, FGG, SERPINB2, BIRC5, FGFR3, IL1R2, 

CXCL10, RORA, CXCL14, POSTN, SERPINB7, IMPA2, SERPINB5, ADAMTS5, CLCA2, 

PPP1R14C, ERH, C1S, RRM2, TLL1, LMNA, FGFR2, MMP2, C6ORF142, PAPSS2, AREG, IRF6, CDH1, IL1RN, SLC7A2, IGFBP3, CTSB, TAGAP, DSC1, CPS1, MCTP2, SCD, ZDHHC2, LOX, CDC20, SRD5A1, CD55, ANTXR1, PRR5L, APP, TGM2, JUN, F2RL1, ANXA1, TOP2A, PTPRZ1, KRT2, FABP4, S100A8, VDR, MMP1, VCAM1, SDC1, EDNRB, DST, CD69, SH3BP5, HOPX, PLA2G4A, HP, CD44, 

CDH26, PTGS2, GATA3, TN‐ FRSF19, KYNU, WNT5A 

Genetic Disorder 6,26E‐07‐3,71E‐03 

 

MRC1, MALAT1, GLT8D2, SLC27A2, IL13RA2, SNX7, CA12, PCDH7, C4ORF31, GALNTL2, FGG, CXCL10, SERPINB7, 

SLCO2B1, ACACB, SLC2A1, KLHDC10, DCN, MT1X, HBA1, RRM2, FGFR2, TLL1, ANLN, MMP2, KRT10, S100A2, C6ORF142, DSG1, WFDC5, CDH1, FXYD3, CTSB, SPINT2, 

C10ORF58, C13ORF18, CYR61, CPS1, PER2, APOD, ANK3, SH3RF2, SERPING1, SRD5A1, CD55, BCL11B, TGM2, C17ORF67, AQP3, DSG3, HK2, JUN, F2RL1, TFAP2A, TOP2A, FABP4, KRT2, VDR, TRIM29, MMP1, DSC3, TGFBI, KRT17, DST, NLGN4X, SLC39A6, DEFB1, SH3BP5, FOLR1, PLA2G4A, HP, 

KRT5, TMEM49, CD44, PTGS2, GATA3, NEDD9, KYNU, ID4, CD81, SOCS3, SLC7A11, HSPA1A, MT1M, CHST15, BIRC5, FGFR3, IL1R2, CRIP1, CCL27, RORA, CXCL14, POSTN, IMPA2, SERPINB5, KRT1, ADAMTS5, RIPK4, PPP1R14C, KRT16, NOX4, H19, RRAD, C1S, BASP1, LMNA, AURKA, PAPSS2, IRF6, RAPGEF2, IL1RN, SLC7A2, IGFBP3, DSC1, HSD17B2, MCTP2, SCD, CYP46A1, LOX, SRPX, CDC20, KIAA0101, ANTXR1, RARRES1, IQGAP1, APP, RGS17, GPC1, PTPRZ1, S100A8, RNF144B, KRT14, ANXA8L2, PKP1, TP63, SDC1, EDNRB, 

RACGAP1, MIR155HG, PERP, OVOL1, MT1H, 

Page 91: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  76

TSPAN8, TPT1, HOPX, CLDN1, KLF5, TNFRSF19, PRKAR1A, WNT5A 

Hematological System Development and Function 

6,94E‐07‐5,02E‐03 

CD81, SOCS3, SERPING1, HSPA1A, CD55, IL13RA2, BCL11B, MMRN1, APP, BIRC5, CXCL10, TGM2, JUN, CCL27, F2RL1, RORA, EZR, ANXA1, CXCL14, TOP2A, FABP4, 

S100A8, RASSF5, VDR, MMP1, PTX3, VCAM1, EDNRB, CD69, HBA1, PERP, DEFB1, MMP2, 

KRT10, PAPSS2, TSPAN8, TPT1, STC1, HP, IL1RN, SLC7A2, CTSB, CD44, HHEX, PTGS2, DSC1, GATA3, 

NEDD9 

Immune Cell Trafficking 6,94E‐07‐5,13E‐03 

 

CD81, SOCS3, SERPING1, HSPA1A, CD55, APP, TGM2, CXCL10, F2RL1, CCL27, ANXA1, CXCL14, FABP4, S100A8, RASSF5, PTX3, PODXL, VCAM1, SDC1, EDNRB, CD69, DEFB1, MMP2, KRT10, TPT1, STC1, HP, 

IL1RN, SLC7A2, CTSB, CD44, PTGS2, DSC1, NEDD9 

Inflammatory Response 2,33E‐06‐5,13E‐03 

CD81, SOCS3, SERPING1, HSPA1A, CD55, APP, FGG, TGM2, CXCL10, IL1R2, DSG3, F2RL1, CCL27, ANXA1, CXCL14, FABP4, 

S100A8, LYVE1, VDR, PTX3, NOX4, PODXL, VCAM1, EDNRB, CD69, DEFB1, IL20RB, TPT1, STC1, PLA2G4A, HP, IL1RN, KLRK1, SLC7A2, CTSB, CD44, PTGS2, DSC1, GATA3, 

NEDD9 

Organismal Injury and Abnormalities 

1,22E‐05‐5,05E‐03 

 

SOCS3, SCD, SERPING1, SLC7A11, IL13RA2, APP, CXCL10, TGM2, FGFR3, DSG3, F2RL1, ANXA1, POSTN, LYVE1, MMP1, PODXL, 

EDNRB, C1S, FGFR2, LMNA, MMP2, MT1H, KRT10, AREG, STC1, HOPX, PLA2G4A, HP, IL1RN, KLRK1, KLF5, CTSB, CD44, IGFBP3, 

PTGS2 

Cell Morphology 1,41E‐05‐5,13E‐03 

ANK3, HSPA1A, ANTXR1, SPRR1A, IQGAP1, APP, TGM2, FGFR3, CXCL10, JUN, GNA15, 

TFAP2A, EZR, ANXA1, POSTN, PTPRZ1, VDR, SERPINB5, PTX3, KRT16, NOX4, VCAM1, TP63, SDC1, EDNRB, TGFBI, RRAD, DCN, HBA1, BASP1, MAP7, LMNA, MMP2, STC1, IRF6, CDH1, RND3, KLRK1, KLF5, IGFBP3, CD44, PTGS2, CYR61, GATA3, TNFRSF19, 

NEDD9, WNT5A 

Tissue Morphology 1,63E‐05‐4,51E‐03 

 

CD81, ANK3, SOCS3, LOX, SERPING1, APP, BIRC5, CXCL10, FGFR3, JUN, TFAP2A, F2RL1, ANXA1, SERPINB7, S100A8, SERPINB5, VDR, CAV2, MMP1, NOX4, VCAM1, H19, TP63, EDNRB, CD69, DCN, MAP7, FGFR2, MMP2, CGA, TPT1, STC1, IL1RN, CD44, IGFBP3, CTSB, PTGS2, SFN, 

GATA3, CYR61, NEDD9, WNT5A 

Lipid Metabolism 1,73E‐05‐4,41E‐03 

 

PTX3, SCD, CYP46A1, ACACB, SLC27A2,SRD5A1, CGA, APP, IL1R2, GPC1, PLA2G4A, 

F2RL1, IL1RN, ANXA1, CTSB, FABP4, S100A8, PTGS2, VDR, CAV2, SLCO2B1, APOD

Page 92: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  77

Molecular Transport 1,73E‐05‐3,56E‐03 

 

SCD, ACACB, SLC27A2, CGA, APP, IL1R2, GPC1, PLA2G4A, F2RL1, IL1RN, ANXA1, CTSB, FABP4, S100A8, PTGS2, CAV2, 

SLCO2B1, APOD 

Small Molecule Biochemistry 

1,73E‐05‐4,52E‐03 

 

CYP46A1, SCD, SOCS3, SLC27A2, SRD5A1, APP, IL1R2, GPC1, HK2, F2RL1, RORA, ANXA1, FABP4, S100A8, VDR, CAV2, 

SLCO2B1, PTX3, ACACB, SLC2A1, EDNRB, CGA, PLA2G4A, IL1RN, KLRK1, SLC7A2, 

CTSB, PTGS2, APOD 

Table 10: Genes assciated with most deregulated functions (see Figure 26) as evaluated by IPA analysis. 

 

Page 93: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

78

 

AffyID  Accession  Gene  Gene Description  Ratio*  P­value  Molecule / Function† 

Cluster A: Inflammatory Response Innate immune and acute phase response208470_s_at  NM_005143  HP  Haptoglobin  8.43  0.021  Peptidase, acute phase protein 206157_at  NM_02852  PTX3  Pentraxin‐related gene  2.80  0.0004  Acute phase protein, regulation of innate immunity 208747_s_at  NM_001734.3  C1S  Transcribed locus, component 1, s subcomponent  2.76  < 0.05  Complement component, opsonisation 

200838_at  NM_001908  CTSB  Cathepsin B  1.78  0.003  Lysosomal cysteine peptidase, cellular stress and wounding, APP processing 

200986_at  NM_000062  SERPING1  Serpin peptidase inhibitor, clade G (C1 inhibitor), member 1 

1.71  0.004  Peptidase inhibitor, regulates C1S 

200602_at  NM_000484  APP  Amyloid precursor protein  1.54  0.029  Acute phase protein 206785_s_at  NM_2260  KLRC2  Killer cell lectin‐like receptor subfamily C, member 2  0.26  < 0.05  Ca‐dependent lectin, defense response regulation 205821_at  NM_7360  KLRK1  Killer cell lectin‐like receptor subfamily K, member 1  0.23  < 0.05  MHCI receptor, innate immunity 208155_x_at  NM_001468  GAGE1  G antigen 1  0.23  < 0.05  Antigenic peptide, defense response 210397_at  U73945  DEFB1  Defensin, beta 1  0.20  0.0003  Antimicrobial peptide, defense response 

202917_s_at  NM_002964  S100A8  S100 calcium binding protein A8  0.14  0.041 Antimicrobial cytokine, RAGE ligand, inflammatory response 

226926_at  NM_001035516  DMKN  Dermokine  0.13  0.002  Inflammatory protein 218677_at  NM_020672  S100A14  S100 calcium binding protein A14  0.11  0.012  Ca2+ and RAGE ligand 230720_at  NM_001165032  RNF182  Ring finger protein 182  0.10  < 0.05  MHCI associated antigen‐processing 219612_s_at  NM_000509  FGG  Fibrinogen gamma chain  0.08  < 0.05  Fibrin monomer, inflammatory marker 204971_at  NM_005213  CSTA  Cystatin A   0.05  0.012  Cysteine protease inhibitor, regulates CTSB 

Wounding and stress response 236843_at  NM_001143836  NOX4  NADPH oxidase 4  4.81  0.026  ROS production, oxygen sensor, inflammatory response 222748_s_at  AW194729  TXNL4B  Thioredoxin‐like 4B  3.77  0.026  Spliceosomal protein 213425_at  NM_003392  WNT5A  Wingless‐type MMTV integration site family, member 5A  3.64  0.0004  Signaling protein, inflammatory mediator 203060_s_at  NM_004670  PAPSS2  3'‐phosphoadenosine 5'‐phosphosulfate synthase 2  3.27  < 0.05  ATP‐binding enzyme, anti‐inflammatory 200799_at  NM_005345  HSPA1A  Heat shock 70kDa protein 1A  1.88  0.011  Protein folding enzyme, cell stress 

210145_at  NM_024420  PLA2G4A  Phospholipase A2, group IVA (cytosolic, calcium‐dependent) 

0.28  < 0.05  arachidonic acid releasing enzyme, pro‐inflammatory 

206461_x_at  NM_005951  MT1H  Metallothionein 1H  0.27  0.014  Zn‐binding protein, anti‐oxidant 204326_x_at  NM_002450  MT1X  Metallothionein 1X  0.23  0.008  Zn and Ca homeostasis, anti‐oxidant 217546_at  NM_176870  MT1M  Metallothionein 1M  0.19  < 0.05  Cu and Zn binding 205081_at  NM_001311  CRIP1  Cysteine‐rich protein 1 (intestinal)  0.17  0.021  Zn finger protein family, immediate early gene 

204748_at  NM_000963  PTGS2  Prostaglandin G/H synthase and cyclooxygenase  0.17  0.017  Prostaglandin producing enzyme, inflammatory mediator 

Chemotaxis, leukocyte adhesion and extravasation

204533_at  NM_001565  CXCL10  Chemokine (C‐X‐C motif) ligand 10  5.16  0.009  Chemokinge ligand, chemotactic for monocytes and T‐cells 

203868_s_at  NM_001078  VCAM‐1  Vascular cell adhesion molecule 1  4.23  0.001  Integrin receptor, leukocyte adhesion molecule 226017_at  NM_138410  CMTM7  CKLF‐like MARVEL transmembrane domain containing 7  1.52  0.038  Chemokine‐like, unknown function 204438_at  NM_002438  MRC1  Mannose receptor, C type 1  1.47  < 0.05  Endocytosis glycoprotein, pathogen endopinocytosis 205945_at  NM_000565  IL6R  Interleukin 6 receptor  0.30  < 0.05  Subunit of the interleukin 6 (IL6) receptor complex 

Page 94: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

79

205403_at  NM_004633  IL1R2  Interleukin 1 receptor, type II  0.26  < 0.05  Cytokine decoy receptor 206172_at  NM_000640  IL13RA2  Interleukin 13 receptor, alpha 2  0.22  < 0.05  Chemokine receptor, suppresses T‐cell response 

204597_x_at  NM_003155  STC1  Stanniocalcin 1  0.21  < 0.05  Pleiotropic homodimeric glycoprotein, anti‐inflammatory 

213506_at BE965369  

F2RL1  Coagulation factor II (thrombin) receptor‐like 1  0.17  0.010  Receptor for trypsin and trypsin‐like enzymes coupled to G proteins, innate immune response 

222484_s_at  NM_004887  CXCL14  Chemokine (C‐X‐C motif) ligand 14  0.11  0.0005  Chemokine acting on macrophages and DCs 228575_at  NM_144717  IL20RB  interleukin 20 receptor beta  0.11  0.049  Coreceptor for IL20, suppresses T‐cell response 212657_s_at  U65590  IL1RN  Interleukin 1 receptor antagonist  0.08  < 0.05  IL1 antagonist 207955_at  NM_006664  CCL27  Chemokine (C‐C motif) ligand 27  0.06  0.006  T‐cell attracting chemokine 

Cluster B: LEC Adhesion and MigrationGrowth factors and signaling 204379_s_at  NM_000141  FGFR2  Fibroblast growth factor receptor 2  0.30  0.032  FGF receptor signaling, cell growth 203819_s_at  NM_006547  IGF2BP3  Insulin‐like growth factor 2 mRNA binding protein 3  0.28  < 0.05  mRNA binder, translational regulator 

206100_at  NM_001874   CPM  Carboxypeptidase M  0.27  < 0.05 

Membrane peptidase, controls hormone and growth factor activity 

201506_at  NM_000358  TGFBI  Transforming growth factor, beta‐induced, 68kDa  0.23  < 0.05  Cell‐collagen interactions, inhibits cell adhesion 210095_s_at  M31159  IGFBP3  Insulin‐like growth factor binding protein 3  0.22  0.011  Translational repressor, prolongs the half‐life of IGFs 

209795_at  NM_001781.2 CD69 = CLEC2  CD69 molecule  0.20  < 0.05 

Ca dependent lectin superfamily of type II transmembrane receptors 

225316_at  NM_032793  MFSD2A  Major facilitator superfamily domain containing 2  0.19  0.023  multi‐pass membrane protein, supposed tu suppressor 

205239_at  NM_001657   AREG  Amphiregulin  0.17  < 0.05  EGF‐like autocrine growth factor, involved in infection 

and inflammation 204268_at  NM_005978  S100A2  S100 calcium binding protein A2  0.16  0.03  EF hand Ca binding protein, cell growth, pro‐migratory 

203638_s_at  NM_005130   FGFBP1  Fibroblast growth factor binding protein 1  0.11  0.025  Carrier protein that releases FGFs from the 

extracellular matrix 203074_at  NM_001630  ANXA8L2  Annexin A8‐like 2  0.10  0.030  ca‐dependent phospholipid binding, controls migration 

204637_at  NM_000735.3  CGA  Glycoprotein hormones, alpha polypeptide  0.10  < 0.05  Hormone subunit, involved in vascularization, cell death 

204379_s_at  NM_000142  FGFR3  Fibroblast growth factor receptor 3  0.09  0.031  Fibroblast growth factor receptor, signaling pathway 

ECM Proteolysis 

235368_at  NM_007038  ADAMTS5  ADAM metallopeptidase with thrombospondin type 1 motif, 5  3.24  0.002  Metalloprotease, cartilage aggrecan and proteoglycan 

destruction 1555071_at  NM_012464  TLL1  Tolloid‐like 1  2.36  < 0.05  Metalloprotease, regulates ECM formation  201069_at  NM_00453  MMP2  Matrix metallopeptidase 2  2.14  0.009  degrades type IV coll, vascular remodelling 

204475_at  NM_002421  MMP1  Matrix metallopeptidase 1 (interstitial collagenase)  0.26  < 0.05  peptidoglycan metabolism, serum levels reduced in  t2d, 

206421_s_at  NM_003784  SERPINB7  Serpin peptidase inhibitor, clade B (ovalbumin), member 7  0.24  < 0.05  serine‐type endopeptidase inhibitor 

210715_s_at  NM_021102  SPINT2  Serine peptidase inhibitor, Kunitz type, 2  0.15  0.009  inhibits a variety of serine proteases 

204489_s_at  NM_032330  CAPNS2  Calpain, small subunit 2  0.14  0.039 EF‐hand, calcium binding protein, regulatory unit for ca activated proteases 

220051_at  NM_006799.2  PRSS21  Protease, serine, 21 (testisin)  0.14  < 0.05  serine protease 

204855_at  NM_002639  SERPINB5  Serpin peptidase inhibitor, clade B (ovalbumin), member 5  0.13  0.036  serine‐type endopeptidase inhibitor   

Page 95: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

80

204614_at  NM_002575  SERPINB2  Serpin peptidase inhibitor, clade B (ovalbumin), member 2 

0.11  0.038  serine‐type endopeptidase inhibitor , uPA inhibitor 

242204_at  NM_145652  WFDC5  WAP four‐disulfide core domain 5  0.06  0.006  serine‐type endopeptidase inhibitor 

ECM adhesion  / cell migration 

213909_at  NM_001135057  LRRC15  Leucine rich repeat containing 15  7.26  < 0.05 membrane protein, might promote migration and invasion 

236361_at  NM_054110  GALNTL2  UDP‐N‐acetyl‐alpha‐D‐galactosamine:polypeptide N‐acetylgalactosaminyltransferase‐like 2 

3.91 0.008  

catalyzes O‐linked polysaccharide biosynthesis 

1563502_at  NM_016353  ZDHHC2  Palmitoyltransferase ZDHHC2  2.77  < 0.05  palmitoyltransferase, tetraspanin stabilization 220037_s_at  NM_006691  LYVE1  Lymphatic vessel endothelial hyaluronan receptor 1  2.50  0.036  hyaluronic acid binding, cell motion, cell adhesion. 221447_s_at  NM_031302  GLT8D2  Glycosyltransferase 8 domain containing 2  2.13  0.022  glycosyltransferase 

210764_s_at  NM_001554  CYR61  Cysteine rich protein 61  2.09  0.027 Ecm protein, interacts with several integrins and with heparan sulfate 

215446_s_at NM_002317.5   LOX  Lysyl oxidase  2.00  0.007  extracellular copper enzyme that initiates the 

crosslinking of collagens and elastin 205612_at  NM_007351  MMRN1  Multimerin 1  1.75  0.020  elastin microfibril interfacer, integrin ligand 

203066_at  NM_015892  GALNAC4S‐6ST 

Carbohydrate sulfotransferase 14, B cell RAG associated protein 

1.68  0.029  hexose biosynthesis, ECM formation 

210809_s_at  D13665  POSTN  periostin, osteoblast specific factor  0.26  0.022  member of the fasciclin family, heparin binding 

212014_x_at  NM_000610  CD44  CD44 molecule  0.20  0.006 hyaluronic acid binding, vascular integrity, lc extravasation 

201893_x_at  NM_001920  DCN  Decorin  0.19  0.001  proteoglycan, matrix assembly 203824_at  NM_004616  TSPAN8  Tetraspanin 8  0.17  0.029  cell surface glycoprotein, integrin binder, cell mobility 204455_at  NM_001723  DST  Dystonin  0.14  0.008  plakin family protein, hemodesmosomes 201287_s_at  NM_002997  SDC1  Syndecan 1  0.12  0.017  surface proteoglycan, cell spreading and adhesion 

201141_at  NM_002510  GPNMB  Glycoprotein (transmembrane) nmb, osteoactivin  0.10  0.004  type I transmembrane glycoprotein, integrin binding, heparin binding 

Morphogenesis / cytoskeleton organisation 

1569020_at  BC020686  NEDD9  Neural precursor cell expressed, developmentally down‐regulated 9  3.57  0.051  lamellipodia, integrin signaling 

204955_at  NM_006307  SRPX  Sushi‐repeat‐containing protein, X‐linked  2.35  0.045  cytoskeletal association, reg. of cell adhesiveness 224694_at  NM_032208  ANTXR1  Anthrax toxin receptor 1  0.27  < 0.05  type I transmembrane protein, filopodia organization 

202890_at  NM_001198608  MAP7  Microtubule‐associated protein 7  0.26  < 0.05  microtubule‐associated protein, cell polarization and differentiation. 

223687_s_at  NM_017527  LY6K  Lymphocyte antigen 6 complex, locus K  0.23  < 0.05  anchored to membrane 206385_at  NM_001149.3  ANK3  Ankyrin 3, node of Ranvier (ankyrin G)  0.19  < 0.05  cytoskeletal anchoring at plasma membrane 205014_at  NM_031308  EPPK1  Epiplakin 1  0.16  0.028  Intermediate filament cytolinker  

225615_at  NM_001136265  LOC126917=IFFO2  Intermediate filament family orphan 2  0.16  0.0002  intermediate filaments 

201820_at  NM_000424  KRT5  Keratin 5  0.12  0.0002  intermediate filaments 206276_at  NM_003695  LY6D  Lymphocyte antigen 6 complex, locus D  0.09  0.006  leading edges of filopodia, mediates cell adhesion 

Cell­cell adhesion228863_at  AF029343  PCDH17  Protocadherin 17  3.97  < 0.05  Ca‐dependent cell adhesion 

233663_s_at  NM_177980  CDH26  Cadherin‐like 26  3.18  < 0.05 cadherin domain‐containing protein, Ca2+‐dependent cell‐cell adhesion 

Page 96: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

81

205534_at  NM_001173523  PCDH7  Protocadherin 7  0.27  < 0.05  member of cadherin superfamily  1569003_at  NM_030938  TMEM49  Transmembrane protein 49  0.24  < 0.05  Cell‐cell adhesion, induces autophagy of starving cells 226374_at  NM_001338  CXADR  Coxsackie virus and adenovirus receptor  0.21  0.008  homophilic cell adhesion molecule, Lv permeability 

205595_at  NM_001944  DSG3  Desmoglein 3 (pemphigus vulgaris antigen)  0.16  0.033  calcium‐binding transmembrane glycoprotein vascular integrity 

221854_at  AI378979  PKP1  Plakophilin 1  0.15  0.002  member of the arm‐repeat (armadillo) and plakophilin gene, desmosomes 

206032_at  NM_001941  DSC3  Desmocollin 3  0.13  0.003  desmosomal protein, cell‐cell adhesion 222549_at  NM_021101  CLDN1  Claudin 1  0.05  0.002  tight junction protein,  

206642_at  NM_001942  DSG1  Desmoglein 1  0.05  0.001  desmosomal protein, cell‐cell adhesion, vascular integrity 

207324_s_at  NM_004948  DSC1  Desmocollin 1  0.04  0.011  desmosomal protein, cell‐cell adhesion 

Cluster C: LEC Growth / Lymphangiogenesis Response to stimulus / axon guidance236448_at  R37358  UNC5A  Unc‐5 homolog A (C. elegans)  6.23  0.013  netrin‐4 receptor, required for axon‐guidance 

221933_at  NM_020742  NLGN4X  Neuroligin 4, X‐linked  4.27  0.012 neuronal cell surface protein involved in cell junction organization 

213194_at  NM_002941  ROBO1  Roundabout, axon guidance receptor, homolog 1  0.31  < 0.05 immunoglobulin gene superfamily, axon guidance receptor 

204469_at  NM_002851  PTPRZ1 Protein tyrosine phosphatase, receptor‐type, Z polypeptide 1  0.29  < 0.05  receptor protein tyrosine phosphatase, axonogenesis 

202391_at  NM_006317  BASP1  Brain abundant, membrane attached signal protein 1  0.09  0.024  transcription corepression molecule, axon guidance 

G­protein coupled receptor signaling 

203096_s_at  BF439282  RAPGEF2  Rap guanine nucleotide exchange factor (GEF) 2  3.45  0.011  Ras GTPase activating factor, fosters adherence junction formation 

220334_at  NM_012419  RGS17  Regulator of G‐protein signalling 17  2.63  < 0.05  GTPase activating protein, inhibitor of GPCR signaling 206701_x_at  NM_000115  EDNRB  Endothelin receptor type B  1.76  0.023  G protein‐coupled receptor, lymph vessel formation 222077_s_at  NM_001126103  RACGAP1  Rac GTPase activated protein 1  0.30  < 0.05  negatively regulates Rho‐mediated signals 

223322_at  BC004270  RASSF5  Ras association (RalGDS/AF‐6) domain family 5  0.26  0.046  binds to activated Ras proteins, integrin signaling, directional ec movement 

204803_s_at  NM_004165  RRAD  Ras‐related associated with diabetes  0.24  < 0.05  small GTPase associated signal transduction, inhibits vascular lesion formation 

229723_at  NM_054114  TAGAP  T‐cell activation GTPase activating protein  0.21  < 0.05  Rho GTPase‐activating protein   

212724_at  BG054844  RND3  Rho family GTPase 3  0.12  0.023  GTP binding, semaphorin signaling in neurons, stress fiber organization 

218186_at  NM_020387  RAB25  RAB25, member RAS oncogene family  0.11  0.017  small GTPase protein, tumor suppressor 

Other signaling transduction 201811_x_at  NM_004844  SH3BP5  SH3‐domain binding protein 5 (BTK‐associated)  1.57  0.020  protein kinase inhibitor activity 203126_at  NM_014214  IMPA2  Inositol(myo)‐1(or 4)‐monophosphatase 2  0.26  0.003  PI recycling pathway for signaling 221215_s_at  NM_020639  RIPK4  Receptor‐interacting serine‐threonine kinase 4  0.26  0.0004  protein kinase C‐associated kinase 243582_at  NM_152550  SH3RF2  SH3 domain containing ring finger 2  0.25  < 0.05  negative regulation of phosphatase activity 

211003_x_at  NM_0040613  TGM2  Transglutaminase 2  0.22  <0.05  covalent crosslinking of proteins, positive regulation of cell adhesion 

226907_at  NM_030949.2  PPP1R14C  Protein phosphatase 1, regulatory (inhibitor) subunit 14C  0.13  0.004  kinase C‐enhanced PP1 inhibitor 

Transcriptional regulators 

Page 97: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

82

204689_at  NM_002729  HHEX  Hematopoietically expressed homeobox  2.14  0.003  Homeobox TF, regulator of VEGF signaling 201464_x_at  NM_002228  JUN  Jun oncogene  1.65  0.018  Virus derived TF, cell growth, promotes ins resistance 204254_s_at  NM_000376  VDR  Vitamin D (1,25‐ dihydroxyvitamin D3) receptor  0.28  < 0.05  nuclear hormone receptor 222895_s_at  NM_138576  BCL11B  B‐cell CLL/lymphoma 11B  0.25  < 0.05  Zn‐finger TF, Tu‐suppressor 

204653_at  BF343007  TFAP2A  Transcription factor AP‐2 alpha  0.25  0.016  Homo‐ or heterodimeric TF, neg. regulator of proliferation 

221530_s_at  NM_030762  BHLHB3  Basic helix‐loop‐helix domain containing, class B, 3  0.24  < 0.05  Transcriptional repressor 

205286_at  NM_003222  TFAP2C Transcription factor AP‐2 gamma (activating enhancer binding protein 2 gamma) 

0.22  0.016  Homo‐ or heterodimeric TF, developmentally regulated 

202597_at  AU144284  IRF6  Interferon regulatory factor 6  0.21  0.021  Helix‐turn‐helix TF, neg. regulation of proliferation 205251_at  NM_022817  PER2  Period homolog 2 (Drosophila)  0.20  0.002  Circadian clock gene 209603_at  BC003070  GATA3  GATA binding protein 3  0.16  0.016  Zn‐finger TF 

209291_at  NM_001546  ID4  Inhibitor of DNA binding 4, dominant negative helix‐loop‐helix protein 

0.13  < 0.05  Helix‐loop‐helix protein, repressor of cell proliferation 

209212_s_at  AB030824  KLF5  Kruppel‐like factor 5 (intestinal)  0.13  0.039  Zn‐finger TF 227475_at  AI676059  FOXQ1  Forkhead box Q1  0.11  0.015  Member of FOX gene TFs, TGFb mediator 211597_s_at  AB059408  HOPX  HOP homeobox  0.04  0.006  homeobox TF, tu suppressor 

Cell cycle / mitosis 

212086_x_at  NM_005572  LMNA  Lamin A/C  0.53  0.004  nucl. envelope constituent, associated with lipodystrophy 

208079_at  NM_198433  AURKA  Aurora kinase A  0.28  < 0.05  Mitotic spindle organisation 218039_at  NM_016359  NUSAP1  Nucleolar and spindle associated protein 1  0.28  < 0.05  DNA and microtubule binding, cytokinesis 201291_s_at  NM_001067  TOP2A  Topoisomerase II alpha 170 kDa  0.28  < 0.05  fosters cell cycle progression, mitosis, DNA replication. 202870_s_at  NM_001255  CDC20  Cell division cycle 20 homolog  0.27  < 0.05  complexes and activates APC, cell cycle controller 202756_s_at  NM_002081  GPC1  Glypican 1  0.25  0.002  cell surface heparan sulfate proteoglycan  222608_s_at  NM_018685  ANLN  Anillin, actin binding protein     0.22  < 0.05  role during cytokinesis in mitosis  

230104_s_at  NM_015964  TPPP3  Tubulin polymerization‐promoting protein family member 3  0.18  0.043  microtubule bundle formation, centrosome 

amplification 209773_s_at  NM_001165931  RRM2  Ribonucleotide reductase M2 polypeptide  0.14  < 0.05  Deoxyribonucleotide synthesis, cell proliferation,  33323_i_at  X57348  SFN  Stratifin  0.04  0.014  p53 effector, PKC inhibitor 

Cell survival / apoptosis / p53 signaling 

218499_at  NM_016542  RP6‐213H19.1  Serine/threonine protein kinase MST4  0.22  < 0.05  GCK group III family of kinases, cleaved by Casp‐3 

202504_at  NM_012101  TRIM29  Tripartite motif‐containing 29  0.21  0.006  TF, negatively regulates p53222392_x_at  AJ251830  PERP  PERP, TP53 apoptosis effector  0.17  0.028  p53 regulated gene 

209569_x_at  NM_014392  D4S234E=P21 

DNA segment on chromosome 4 (unique) 234 expressed sequence, brain neuron cytoplasmic protein 1  0.14  0.015  transcriptional target of p53, pro‐apoptotic 

206400_at  NM_002307  LGALS7  Lectin, galactoside‐binding, soluble, 7 (galectin 7)  0.05  0.005  p53‐induced, pro‐apoptotic 

Cluster D: Small Molecule BiochemistryLipid transport / metabolic process 1554833_at  NM_018349  MCTP2  Multiple C2 domains, transmembrane 2  5.14  < 0.05  binds Ca in absence of phospholipids, obesity marker! 

220331_at  NM_006668  CYP46A1  Cytochrome P450, family 46, subfamily A, polypeptide 1  4.14  < 0.05  Monooxygenase, cholesterol turnover , steroid degradation 

235978_at  AI766029  FABP4  Fatty acid binding protein 4, adipocyte  3.85  0.006  FFA chaperone, intracell. fa transport 49452_at  NM_001093.3  ACACB  Acetyl‐CoA carboxylase  3.70  0.014  rate‐limiting enzyme in fatty acid oxidation  

Page 98: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

83

201525_at  NM_001647  APOD  Apolipoprotein D  2.64  0.012  lipid transporter glycoprotein 

222217_s_at  NM_024330  SLC27A3 Solute carrier family 27 (fatty acid transporter), member 3 

2.17  0.003  Fa transport and acyl‐CoA activation 

205769_at  NM_003645  SLC27A2  Solute carrier family 27 (fatty acid transporter), member 2  0.28  < 0.05  acyl‐CoA ligase activity 

200832_s_at  NM_005063  SCD‐1  Stearoyl‐CoA desaturase  0.23  0.039  rate‐limiting enzyme of monounsaturated fatty acid biosynthesis 

205220_at  NM_006018  GPR109B  G protein‐coupled receptor 109B  0.07  0.042  Anti‐lipolytic signalling receptor 

Ion transport / metabolic process 203473_at  NM_007256  SLCO2B1 

Solute carrier organic anion transporter family, member 2B1  1.76  0.031  anion and PGE transporter, lipid uptake 

224209_s_at  NM_001242505  GDA  Guanine deaminase  0.26  < 0.05  guanine deamination, 214414_x_at  NM_000558  HBA1  Hemoglobin, alpha 1  0.26  < 0.05  oxygen transport 

209921_at  NM_014331  SLC7A11 Solute carrier family 7, (cationic amino acid transporter, y+ system) member 11  0.25  < 0.05  cysteine:glutamate antiporter activity   

217528_at  NM_006536  CLCA2  Chloride channel, calcium activated, family member 2  0.21  0.042  calcium sensitive chloride conductance protein  204675_at  NM_001047.2  SRD5A1  Steroid‐5‐alpha‐reductase, alpha polypeptide 1  0.20  < 0.05  steroid degradation enzyme, lipid metabolism 39248_at  NM_004925  AQP3  Aquaporin 3  0.17  0.0004  water + glycerol transporter  225516_at  NM_003046  SLC7A2  Solute carrier family 7, member 2  0.12  < 0.05  cationic aa transporter  217564_at  NM_001122633  CPS1  Carbamoyl‐phosphate synthetase 1, mitochondrial  0.10  0.018  removal of excess urea from cells 214164_x_at  NM_001218  CA12  Carbonic anhydrase XII  0.07  0.011  zn metalloenzyme 204437_s_at  NM_000802  FOLR1  Folate receptor 1 (adult)  0.04  0.009  transports 5‐methyltetrahydrofolate into cells 217388_s_at  NM_001032998  KYNU  Kynureninase   0.02  < 0.05  hydrolase , tryptophan metabolism  

Glucose transport / metabolic process 

201250_s_at  NM_006516  SLC2A1  olute carrier family 2 (facilitated glucose transporter), member 1  0.24  0.049  glucose transmembrane transport 

202934_at  NM_000189  HK2  Hexokinase 2  0.17  < 0.05  allosteric enzyme of glycolysis 

Unknown 228977_at  AK090664.1  LOC729680  Uncharacterized LOC729680  2.10  < 0.05  unknown 

228875_at  NM_001085480.2  C6orf189=FAM162B  Family with sequence similarity 162, member B  1.90  0.039  unknown 

217966_s_at  NM_052966.2  FAM129A=GIG39  Family with sequence similarity 129, member A  1.85  0.031  unknown 

229518_at  NM_052943.3  FAM46B  Family with sequence similarity 46, member B  0.29  0.002  unknown 209255_at  NM_014997.3  KIAA0265  Kelch domain containing 10  0.27  < 0.05  unknown 238063_at  NM_152680.2  TMEM154  Transmembrane protein 154  0.24  < 0.05  unknown 

212992_at  NM_138420 C14orf78=AHNAK2  AHNAK nucleoprotein 2  0.23  0.009  unknwon 

Table  11:  Transcript  levels  of  deregulated  candidate  genes  functionally  clustered  in  (A)  Inflammatory  Response,  (B)  LEC  Adhesion  and Migration,  (C) LEC growth and Lymphangiogenesis, and  (D) Small Molecule Biochemistry.  *Shown  are  ratios  of  diabetic  (dLECs)  versus  non‐diabetic (ndLECs) LECs probeset expression. †Statistical significance of that ratio is indicated either as P‐value, or as significant result of RVM analysis, which  integrates a P‐value ≤ 0.05. ‡Differentially expressed genes were grouped according  to  their biological  functions  (GO annotation) and  the  top differentially expressed genes were ranked in order, based on their mean fold change differences. 

Page 99: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  84

7.3.2 Verification of LEC specific genes and comparison with other arrays  

In  order  to  additionally  control  and  confirm  the  purity  of  LEC preparations, we  examined  for 

expression  changes  of  several  LEC‐specific  marker  genes  which  were  shown  to  be  not 

significantly altered between dLECs and ndLECs (Table 12). For example, the transcript level of 

podoplanin, which was used for FACS sorting had a fold change (FC) of 1.08 (dLEC/ndLEC) and 

was considered to be unchanged between the two groups of cells (see also Figure 18 B).  

Expression alterations of established LEC transcripts (e.g. CD44, CCL27, CXADR, CXCL14, LYVE­1, 

PKP1, APP, EDNRB, FABP4, KRT14, MMRN1)  (Podgrabinska  et  al.,  2002; Hirakawa et  al.,  2003; 

Wick  et  al.,  2007;  Wick  et  al.,  2008)  underlined  their  specific  role  in  lymphatic  vessels  and 

validated the quality of our approach. dLECs further revealed some overlapping gene expression 

changes  with  hypoxic  LECs  (Irigoyen  et  al.,  2007),  and  with  LECs  derived  from  skin  of 

lymphedema  patients  (Ogunbiyi  et  al.,  2011)  and  lymph  node  derived  LECs  (Malhotra  et  al., 

2012),  pointing  towards  a  similar  role  of  these  gene  candidates  in  pathological  lymphatic 

dysfunction. In contrast to our results, recent studies detected downregulation of key lymphatic 

markers  PROX­1  and  VEGFR3  in  adipose  tissue  lymphatics  in  familial  human  hyperlipidemia 

(Horra et al., 2009) as well as murine skin inflammation (Vigl et al., 2011).  

Moreover, we  compared  our  data with  data  sets  derived  from LECs  in  pathological  situations 

associated with diabetes, inflammation, wound healing and lymphatic dysfunction, though only 

few  studies  have  investigated  the  gene  expression  changes  of  human  endothelial  cells  in 

pathophysiologic  situations  (St.  Croix  et  al.,  2000;  Seaman  et  al.,  2007;  Clasper  et  al.,  2008; 

Ogunbiyi et al., 2011). Here, we traced some overlaps with gene expression changes in hypoxic 

LECs  (GALNTL2)  (Irigoyen  et  al.,  2007).  However,  compared  to  another  study  using  an 

experimental model  of  acute  post‐surgical  lymphedema,  nearly  no  overlap  could  be  detected 

(Tabibiazar  et  al.,  2006).  Moreover,  only  a  weak  overlap  could  be  detected  with  VEGF­C 

responsive genes in LECs (Yong et al., 2005). This points towards a specific role of deregulated 

genes in LECs in different pathological conditions.  

Overall, our findings indicated that LEC‐specific markers are stable in diabetic skin and that their 

expression  alterations  in  LVs  of  other  tissue  types  might  depend  on  the  specific 

pathophysiological situation and microenvironment. 

Page 100: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  85

 

Gene symbol  Gene name  FC (dLEC/ndLEC)  p­value 

ABCA4 Retinal‐specific  ATP‐binding cassette

transporter  0.66  0.58 

ANGPT2  Angiopoietin‐2 precursor 1.27  0.31KLHL4  Kelch‐like protein 4 0.95  0.89

LYVE­1 Lymphatic vessel endothelial

hyaluronan receptor 1  2.19 0.09 

PDPN  Podoplanin 1.08  0.70THBS4  Thrombospondin 4 precursor 1.88  0.57VEGFR­3  Vascular endothelial cell receptor 3 1.08  0.84

Table 12: Transcript levels of LEC specific genes. No significant differences could be seen for LEC‐specific genes between diabetic (dLECs) and non‐diabetic (ndLECs); FC = fold change. 

 

7.3.3 Diabetic LECs exhibited a distinct gene expression profile compared to diabetic BECs  

The mRNA expression analysis of BECs derived from the same four diabetic and normoglycemic 

patients  revealed  a  completely  diverse  set  of  genes  differentially  regulated  in  diabetic  BECs 

(dBECs).  Out  of  the  180  differentially  regulated  genes  in  dLECs,  only  19 were  also  altered  in 

dBECs (Table 13), which represents an overlap of only 10.5%. This result suggested that a highly 

specific  gene  signature  was  present  in  dLECs  compared  to  dBECs  and  indicated  that  LVs 

undertake completely different tasks than BVs in the skin of T2DM patients. 

Page 101: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  86

 

AffyID  Gene name  Gene description FC 

(dLEC/ndLEC) FC 

(dBEC/ndBEC) Cell adhesion 

210809_s_at  POSTN periostin, osteoblast 

specific factor 0.26  2.52 

206032_at  DSC3  desmocollin 3 0.25 0.22212063_at  CD44  CD44 0.25 0.23

201141_at  GPNMB glycoprotein 

(transmembrane) nmb 0.18  0.17 

212724_at  RND3  Rho family GTPase 3 0.18 0.27206400_at  LGALS7  galectin 7 0.05 0.18

Angiogenesis, wound healing 

210764_s_at  CYR61 cysteine‐rich, angiogenic 

inducer, 61 2.09  2.15 

204855_at  SERPINB5 serpin peptidase inhibitor, clade B, 

member 5 0.22 

0.19  

204614_at  SERPINB2 serpin peptidase inhibitor, clade B, 

member 2 0.19  0.12 

Immune response 

208747_s_at  C1S complement component 1, s subcomponent  2.75  2.86 

222484_s_at  CXCL14 chemokine (C‐X‐C motif) 

ligand 14  0.18  0.13 

Cell growth 

205014_at  FGFBP1 fibroblast growth factor 

binding protein 1 0.11  0.28 

204379_s_at  FGFR3 fibroblast growth factor 

receptor 3  0.09  0.30 

Transport 39248_at  AQP3  aquaporin 3 0.29 0.26

203473_at  SLCO2B1 solute carrier organic anion transporter family, member 2B1 

1.76  0.26 

Transcription 

202597_at  IRF6 interferon regulatory 

factor 6  0.28  0.30 

203074_at  ANXA8  annexin A8 0.09 0.26Other 

226926_at  DMKN  Dermokine 0.13 0.22205251_at  PER2  period homolog 2 0.20 0.34

Table 13: Gene  expression overlap of diabetic LECs  (dLECs) and diabetic BECs  (dBECs) with documented fold change (FC) between diabetic and non­diabetic cell populations. 

 

Page 102: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  87

7.3.4  Genes  linked  to  altered  lipid  transport  and metabolism,  increased oxidative stress and to the pathogenesis of type 2 diabetes mellitus  

First,  we  identified  expression  changes  of  several  genes  that  have  already  been  genetically 

linked  to T2DM in humans (see Table 14),  including haptoglobin  (HP, Quaye et al., 2006),  JUN 

oncogene (Malhotra et al., 2009), tetraspanin 8 (TSPAN8, Zeggini et al., 2008) apolipoprotein d 

(APOD,  Vijayaraghavan  et  al.,  1994),  hematopoietically‐expressed  homeobox  protein  (HHEX, 

Sladek et al., 2007; Saxena et al., 2007),  lamin A/C (LMNA, Wegner et al., 2007) and  fatty acid 

binding  protein  4  (FABP4,  Chan  et  al.,  2010).  Genes  additionally  crucial  for  lipid  metabolic 

processes  included FABP4  and APOD. APOD  is  reported  to  be  a  genetic marker  for  T2DM and 

obesity  (Vijayaraghavan  et  al.,  1994),  to  be  increased  in  T2DM  myotubes  and  gestational 

diabetes (Navarro et al., 2010; Hansen et al., 2004) and to be  involved  in defense mechanisms 

against oxidative stress (Navarro et al., 2010).  

The  results  indicate  that  established  T2D  markers  prove  valid  in  the  lymphatic  vessel 

compartment and confirmed their importance as potential biomarkers.  However, other markers 

very  characteristic  for  T2DM,  like  PPARγ,  did  not  come  up  in  dLECs,  highlighting  the  multi‐

faceted genetic contribution in this disease. Taken together, these alterations presumably confer 

altered adhesive, metabolic and inflammatory properties to type 2 diabetes LECs. 

FABP4  is  highly  and  specifically  expressed  in  LECs  (Ferrell  et  al.,  2008; Wick  et  al.,  2007  and 

unpublished result) and is an important mediator of IR in mice (Uysal et al., 2000). Moreover, its 

expression in human umbilical venous endothelial cells (HUVECs) is regulated by VEGF (Elmasri 

et  al.,  2009).  NOX4  is  a  major  contributor  of  reactive  oxygen  species  (ROS)  production  in 

endothelial cells (Goettsch et al., 2009) and a mediator of angiogenesis (Ushio‐Fukai, 2007). We 

confirmed FABP4, NOX4 and APOD upregulation by quantitative realtime PCR (Figure 27 A) and 

reconfirmed  FABP4  protein  expression  by  immunofluorescence  (Figure  27  B).  Elevated 

expression  levels  of  FABP4,  NOX4  and  APOD  in  dLECs  pointed  towards  a  highly  disturbed 

metabolic status in the skin of T2DM patients. These lipid handling abnormalities, together with 

the downregulation of small molecule transporters (examples) show that LVs are compromised 

in their function in T2DM. 

 

 

Page 103: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  88

 

AffyID Accession 

no. Gene symbol 

Gene name FC 

(dLEC/ndLEC) p­value 

208470_s_at  P00738  HP  Haptoglobin 8.43  0.0209

235978_at  AI766029  FABP4 Fatty acid binding 

protein 4, adipocyte 

3.85  0.006 

201525_at  P05090  APOD  Apolipoprotein D 2.64  0.0115

204689_at  Q03014  HHEX Hematopoietically 

expressed homeobox 

2.14  0.0027 

       201464_x_at  P05412  JUN  Jun oncogene 1.65  0.0182212086_x_at  P02545  LMNA  Lamin A/C 0.53  0.0042203824_at  P19075  TSPAN8  Tetraspanin 8 0.17  0.0285

Table 14: Transcript levels of deregulated genes already associated with type 2 diabetes. Transcript  levels  of  dereguated  genes which were published  to be  genetically  associated with type  2  diabetes.  FC  (dLEC/ndLEC)  =  fold  change  between  diabetic  (dLEC)  and  non‐diabetic (ndLEC) LECs. 

 

Figure  27:  Confirmation  of  altered  expression  of  FABP4, APOD  and NOX4  in  dLECs by realtime PCR  and  immunofluorescence.  A: Differential  expression  of  FABP4,  APOD  and NOX4 was confirmed by quantitative realtime PCR using Taqman ® gene expression assays. B: FABP4 expression was additionally confirmed by double‐labeling of human diabetic and non‐diabetic skin with antibodies to FABP4 and podoplanin; Size bar: 20µm. 

Page 104: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  89

7.3.5 Identification of a gene signature related to wound healing and tissue repair in dLECs  

A huge number of transcripts  involved in ECM remodeling and adhesion processes indicated a 

major  effect  of  T2DM  on  LEC  growth  and  migration,  pointing  at  ongoing  LV  morphogenesis 

and/or  tissue  remodeling  (Table  15).  There  were  prominent  deregulations  of  a  gene  set  of 

proteases  (MMP1, MMP2)  and  protease  inhibitors  (SERPINs),  and  of  a  gene  set  of  transcripts 

involved in formation and binding of ECM components and LEC motility (CYR61, GALNTL2) and 

wound repair (AQP3). 

Two  important  genes  implicated  in  angiogenesis,  tissue  remodeling  and  wound  healing  are 

cysteine rich protein 61 (CYR61) and matrix metalloproteinase 2 (MMP2), which were found to 

be  upregulated  in  dLECs. CYR61  is  an  ECM protein  that mediates wound  healing  (Chen  et  al., 

2001)  and  angiogenesis  via  integrins  (Leu  et  al.,  2002).  Further,  it  is  implicated  in  the 

pathogenesis  of  diabetic  retinopathy  (You  et  al.,  2009)  as  a  downstream  target  of  advanced 

glycation  endproducts  (Hughes  et  al.,  2007).  Additionally,  MMP2  expression  was  shown  to 

stimulate angiogenesis and wound healing (Slyke et al., 2009; Agah et al., 2004). Furthermore, 

we detected downregulation of aquaporin 3 (AQP3), a water and glycerol transporter, which is 

also important for cutaneous wound healing by reducing epidermal proliferation and migration 

(Hara‐Chikuma and Verkman, 2008). Additionally, chemokine  ligand 14 (CXCL14, Maerki et al., 

2009)  and  defensin  beta  1  (DEFB1,  Schittek  et  al.,  2008) were  downregulated  in  dLECs.  Both 

have  antimicrobial  effects,  suggesting  a  function  for  LECs  of  diabetic  skin  in  the  regulation  of 

infections (see Table 15). 

We  confirmed  altered  expression  of  AQP3,  MMP2  and  CYR61  by  quantitative  realtime  PCR 

(Figure  28  A).  Further,  downregulation  of  AQP3  protein  in  diabetic  LVs  was  reconfirmed  by 

immunofluorescent double staining (Figure 28 B), as well as upregulation of CYR61  (Figure 28 

C). MMP2 and CYR61 overexpression in dLECs indicated involvement of LVs in increased tissue 

remodeling  as  observed  during  wound  healing  processes  and  increased  lymphangiogenesis. 

AQP3  expression  was  not  reported  in  LECs  before  but  its  expression  in  diabetic  LECs 

additionally points towards an involvement of LVs in dermal healing processes. Moreover, this 

gene could represent a new water and glycerol transporter in LECs (Ishibashi et al., 2009). 

 

 

 

 

 

 

Page 105: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  90

AffyID Accession 

no. Gene symbol 

Gene name FC 

(dLEC/ndLEC) p­value 

222484_s_at  O95715  CXCL14 Chemokine (C‐X C motif) ligand 14  0.11  0.0005 

210397_at  P60022  DEFB1  Defensin, beta 1 0.20  0.0003

201069_at  P08253  MMP2 Matrix 

metallopeptidase 2 2.14  0.0085 

210764_s_at  O00622  CYR61 Cysteine rich protein 61 

2.09  0.0271 

39248_at  Q92482  AQP3  Aquaporin 3 0.17  0.0004

Table 15: Transcript levels of genes associated with wound healing and tissue repair. FC (dLEC/ndLEC) = fold change between diabetic (dLEC) and non‐diabetic (ndLEC) LECs.  

 

Figure 28: Confirmation of altered AQP3, MMP2 and CYR61 expression in dLECs by realtime PCR  and  immunofluorescence.  A:  Analysis  of  differential  expression  by  quantitative  realtime PCR using Taqman ® gene expression assays. B, C: AQP3 and CYR61 expression was additionally confirmed  by  double‐labeling  of  human  diabetic  and  non‐diabetic  skin  with  antibodies  to respective antigen and podoplanin; Size bar: 20µm.  

Conclusively,  The  gene  alterations  might  indicate  a  stage  of  increased  active  reparative 

mechanism in T2D skin. and show that lvs play an active role in T2D skin disorder. 

Page 106: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  91

7.3.6  Identification  of  a  gene  signature  related  to  increased  adhesion  of inflammatory cells  

A  plethora  of  genes  implicated  in  cell  adhesion  and  inflammatory  processes  was  altered 

significantly  in dLECs  (Table 16). The most upregulated genes  included vascular cell  adhesion 

molecule  (VCAM­1)  and  chemokine  (CXC‐motif)  ligand  10  (CXCL10),  while  coxsackie  and 

adenovirus  receptor  (CXADR)  and  syndecan  1  (SDC1)  were  downregulated.  These  genes  are 

mainly responsible for the regulation of immune cell traffcking. VCAM­1 was previously reported 

to be expressed in LECs upon TNFα stimulation and to be an important regulator of DC adhesion 

and migration in vitro and in vivo (Johnson et al., 2006). CXCL10 seems to be a crucial chemokine 

in T‐cell and macrophage infiltration (Heller et al., 2006; Dufour et al., 2002) and implicated in 

resolution  of  inflammation  (Rahman  et  al.,  2010).  CXADR  interacts  with  junctional  adhesion 

molecule 1 to regulate monocyte, neutrophil and T‐cell migration across junctional connections 

(Luissint et al., 2008; Zen et al., 2005) and was previously reported to be specifically expressed 

on  LECs  (Vigl  et  al.,  2009).  SDC1  has  major  tasks  in  inhibiting  leukocyte  adhesion  to  the 

endothelium (Götte et al., 2002) and is reported to be downregulated by TNFα (Kainulainen et 

al., 1996). 

 

AffyID Accession 

no. Gene symbol 

Gene name  FC (dLEC/ndLEC)  p­value 

204533_at  P02778  CXCL10 Chemokine (C‐X‐C motif) ligand 

10 5.16  0.0093 

203868_s_at  P19320  VCAM­1 Vascular cell adhesion molecule 1 

4.23  0.0013 

207955_at  Q9Y4X3  CCL27 Chemokine (C‐C motif) ligand 27 

0.06  0.0058 

226374_at  P78310  CXADR Coxsackie virus and adenovirus 

receptor 0.21  0.0080 

201287_s_at  P18827  SDC1  Syndecan 1 0.12  0.0165

Table 16: Transcript  levels of genes  linked to  increased adhesion of  inflammatory cells. FC (dLEC/ndLEC) = fold change between diabetic (dLEC) and non‐diabetic (ndLEC) LECs. 

 

Differential  gene  expression  of VCAM­1, CXCL10, CXADR  and  SDC1  was  confirmed  by  realtime 

PCR  and  showed  that  expression  ratios  matched  those  retrieved  from  microarray  analysis 

(Figure  29  A).  Differential  gene  expression  of  CXCL10  and  CXADR  was  further  approved  on 

protein  level  by  double‐immunofluorescence  stainings  of  human  skin  (Figure  29  B  and  C). 

Conclusively, differential expression of VCAM­1, CXCL10, CXADR  and SDC1  suggested  increased 

Page 107: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  92

adhesion  of  inflammatory  cells  to  the  lymphatic  endothelium  in  the  skin  of  T2DM  patients, 

possibly as a result of an increased pro‐inflammatory milieu. 

 

Figure 29: Confirmation of altered VCAM­1, CXCL10, CXADR and SDC1 expression in dLECs by realtime PCR and immunofluorescence. A: Differential expression of VCAM‐1, CXCL10, CXADR and SDC1 was confirmed by quantitative realtime PCR using Taqman ® gene expression assays. B,  C:  Differential  expression  of  CXCL10  and  CXADR  was  reconfirmed  by  double‐labeling  of human diabetic and non‐diabetic skin with antibodies to CXCL10, CXADR and podoplanin; Size bar: 20µm.  

Page 108: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  93

7.3.7 Deregulated genes associated with cellular host defense  

The  prominent  inflammation‐related  gene  cluster  revealed  alterations  of  processes  related  to 

cellular defense, response to infection, chemotaxis/leukocyte adhesion and tissue repair/wound 

healing (Table 17).  

A group of acute phase and antimicrobial defense response factors contained transcripts of well 

described extracellular factors, which previously have been detected as important inflammatory 

markers in plasma, lymph fluid or wound tissue. Of note, we detected upregulation of transcripts 

for  HP  (haptoglobin),  PTX3  (pentraxin),  C1S  (complement  1),  SERPING1  (serpin  peptidase 

inhibitor,  clade  G  (C1  inhibitor),  member  1)  and  APP  (amyloid  precursor  protein),  and 

downregulation of GAGE1 (G antigen 1), DEFB1 (defensin, beta 1), S100A8 (S100 calcium binding 

protein A8) and FGG (fibrinogen gamma chain). Further, innate immune response genes KLRC1 

(killer  cell  lectin‐like  receptor  subfamily  C, member  1), KLRK1  (killer  cell  lectin‐like  receptor 

subfamily  K,  member  1)  and  KYNU  (kynureninase)  were  downregulated.  Transcripts  for 

extracellular  inflammatory  regulators  WNT5A  (wingless‐type  MMTV  integration  site  family, 

member  5A),  AREG  (amphiregulin)  and  IL1RN  (interleukin  1  receptor  antagonist)  were 

deregulated.  

Interestingly,  a  plethora  of  membrane‐associated  receptors  involved  in  proper  inflammatory 

signaling,  like  IL6R  (interleukin  6  receptor),  IL1R2  (interleukin  1  receptor  2),  IL13RA 

(interleukin  13  receptor  A),  and  SLC7A2  (solute  carrier  family  7  (cationic  amino  acid 

transporter,  y+  system), member  2) were  downregulated.  Further,  intracellular  stress  factors 

NOX4 (NADPH oxidase 4), HSPA1A (heat shock 70kDa protein 1A), and CTSB (cathepsin B) were 

upregulated, while components of arachidonic acid uptake and metabolism (SLCO2B1, PLA2G4A, 

PTGS2) were downregulated. These results suggested an extensive involvement of dLECs in skin 

infection and host defense. 

Many  of  the  transcripts  regulated  in  dLECs  have  been  previously  identified  in  whole  tissue 

lysates of chronic human skin  inflammation, venous ulcers and wounds (de  Jongh et al., 2005; 

Charles  et  al.,  2008;  Roy  et  al.,  2008;  Grice  et  al.,  2010),  and  the  emergence  of  these  gene 

products  might  be  a  specific  contribution  of  lymphatic  vessels.  We  identified  overlapping 

deregulation of some transcripts described in atopic dermatitis (AREG, CSTA, DEFB1 and S100A8; 

de Jongh et al., 2005), for psoriasis, (including CXCL10 and S100A8; Guttman‐Yassky et al., 2008), 

nonhealing  venous  ulcers,  (including  AQP3  and  SFN;  Charles  et  al.,  2008),  in  wound 

inflammation,  including CXCL10, CYR61, JUN, PTX3, SOCS3 and VCAM1 (Roy et al., 2008), and in 

the diabetic wound microbiome, including CXCL10, FABP4, CD55, DEFB1 and S100A8 (Grice et al., 

2010). Several of the secreted factors (HB, C1S, FGG, HP, PTX3) have been previously detected as 

inflammatory markers in plasma, lymph fluid or wound tissue. Additionally, some, but generally 

Page 109: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  94

low overlap was detected when comparing our gene list with a murine skin inflammation model, 

including VCAM1, FGFR3, CXCL10 and ADAMTS5 (Vigl et al., 2011). 

Downregulation  of  important  inflammatory  response  genes  in  dLECs,  indicating  a  prolonged 

period  of  inflammatory  disorder  in  the  skin  that  leads  to  hampered  immune  modulatory 

functions of dermal lymphatics. Further, these genes indicate that LVs play an active role in skin 

wounds  and  infections  by  contributing  to  pathogen  modulation.  Altogether,  these  gene 

candidates might highlight functional pathways that are involved in recurrent skin infections. 

 

In  summary,  the  dLEC  transcriptome  highlights  the  processes,  which  are  also  seen  on  a 

morphological  level. The  findings underline  the  importance of LVs and LECs  in  immunological 

processes  and  lipid  metabolism  and  suggest  that  skin  LVs  react  actively  on  the  multiple, 

metabolic as well as inflammatory interstitial burden. 

The results in the ‘inflammatory cluster’ indicate a prolonged period of inflammatory disorder in 

the skin that leads to hampered immune modulatory functions of dermal lymphatics. The large 

number  of  transcripts  involved  in  ECM  remodeling  and  adhesion  processes  indicates  a major 

effect of T2D on LEC growth and migration, pointing at ongoing  lymphatic  vessel  remodeling, 

supporting the finding of increased lymphangiogenesis in T2DM skin. Further, a huge number of 

positive  and  negative  regulators  of  cell  growth,  migration,  proliferation  and  survival  and 

suggested an activation of cellular differentiation and survival processes in LECs in T2D skin.  

Additionally,  LECs might  deal  with  the  lipid  handling  abnormalities,  as  many  small  molecule 

transporters  were  downregulated,  showing  that  LVs  are  compromised  in  their  function, 

regarding lipid, water and electrolyte transport. 

Moreover, the increased expression of TNFα also contributes to dysregulation of genes, mainly 

implicated in the clusters ‘cell adhesion’ and ‘inflammatory response’ (see next section).   

 

 

Page 110: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  95

 

AffyID Accession 

no. Gene symbol 

Gene name FC 

(dLEC/ndLEC) p­value 

206157_at  P26022  PTX3 Pentraxin‐related 

gene 2.80  0.0004 

208747_s_at  P09871  C1S Transcribed locus,component 1, s subcomponent 

2.76  <0.05 

200799_at  P08107  HSPA1A Heat shock protein A1A 

1.88  0.0106 

200838_at  P07858  CTSB  Cathepsin B 1.78  0.0029

203473_at  O94956  SLCO2B1 

Solute carrier organic 

anion transporter family, 

member 2B1 

1.76  0.0313 

200986_at  P05155  SERPING1 

Serpin peptidase inhibitor, clade G (C1 inhibitor), member 1 

1.71  0.0039 

200602_at  P05067  APP Amyloid precursor protein 

1.54  0.0285 

205821_at  P26718  KLRK1 

Killer cell lectin‐like receptor subfamily C, member 1 

0.26  <0.05 

204971_at  P01040  CSTA  Cystatin A 0.05  0.0118236843_at  Q9NPH5  NOX4  NADPH oxidase 4 4.81  0.0255

213425_at  P41221  WNT5A 

Wingless‐type MMTV integration 

site family, member 5A 

3.64  0.0004 

204597_x_at  P52823  STC1  Stanniocalcin 1 0.21  < 0.05

206785_s_at  Q3KQS7  KLRC2 

Killer cell lectin‐like receptor subfamily C, member 1 

0.26  <0.05 

210397_at  P60022  DEFB1  Defensin, beta 1 0.20  0.0003

204748_at  P35354  PTGS2 prostaglandin G/Hsynthase and cyclooxygenase 

0.17  0.0165 

202917_s_at  P05109  S100A8 S100 calcium binding protein 

A8 0.14  0.0713 

212657_s_at  P18510  IL1RN Interleukin 1 receptor antagonist 

0.08  0.0969 

Table 17: Transcript levels of genes associated with cellular host defense. Transcript levels of  dereguated  genes  which  are  implicated  in  the  process  of  cellular  and  host  defense.  FC (dLEC/ndLEC) = fold change between diabetic (dLEC) and non‐diabetic (ndLEC) LECs. 

Page 111: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  96

7.4 TNFα­induced effects on LEC behavior  

In  the  final  part  of  my  thesis,  I  aimed  to  develop  a  mechanistic  link  between  the  observed 

morphological  changes  of  T2DM  skin,  i.e.  the  enhanced  LV  and macrophage  number,  and  the 

specific gene expression profile of dLECs. In order to establish this link between skin lymphatic 

morphology and the dLEC transcriptome, we were looking for a factor possibly bridging the gap 

between these two findings.  

 

7.4.1 TNFα responsiveness of LEC genes in vitro  

Due to the finding that the highly increased macrophage number in diabetic skin was capable of 

producing a local pro‐inflammatory milieu via TNFα, we aimed at investigating whether in vitro 

TNFα  stimulation of  LECs  could  recapitulate  the  differential  gene  expression observed by our 

transcriptomal analysis of dLECs versus ndLECs ex vivo. TNFα treatment  induced upregulation 

of  CXCL10,  VCAM­1  and  CYR61  (Figure  30)  and  downregulation  of  SDC1,  CXADR  and  AQP3 

(Figure  31)  as  it  was  identified  by  the  transcriptomal  analysis  of  ex  vivo  isolated  LECs  from 

T2DM  patients.  Additionally,  treatment  with  an  inhibitory  TNFα  antibody  reversed  TNFα 

induced expression changes of genes and suggested that differential gene expressions are highly 

specific  for  TNFα.  However,  expression  of  FABP4,  GALNTL2  and  APOD  was  downregulated, 

which  is  in  contrast  to  the  array  data  (Figure  32).  Moreover,  no  effect  could  be  found  on 

expression of MMP2 and NOX4 (Figure 33). 

We  conclude  from  these  data  that  increased  TNFα  levels,  mainly  produced  by  infiltrating 

macrophages,  might  directly  influence  the  expression  of  some  genes  in  microvascular  LECs. 

Hence,  we  identified  TNFα‐driven  gene  expression  as  one  signaling  pathway  leading  to 

differential  gene  expression  in  dLECs  versus  ndLECs.  However,  TNFα‐driven  differential 

expression of other genes did not match array data. Therefore, we hypothesize that not only one 

predominant  but  rather  the  interplay  of  several  phenomena,  including  epigenetic, 

transcriptional, or microRNA‐mediated mechanisms, is crucial for the diabetic metabolic status 

and responsible for observed gene expression changes of dLECs versus ndLECs. 

 

Page 112: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  97

 

Figure  30: Upregulation  of  CXCL10,  VCAM­1  and  CYR61 gene  expression  of  cultured  LECs  byTNFα  stimulation.  Primary  dermal  LECs were  cultured  in  EBM‐2/0.5%  FCS  and  stimulated with10ng/ml TNFα for 24 hours (with or without TNFα inhibitor). TNFα treatment of primary culturedLECs  perfectly  recapitulated  upregulation  of CXCL10, VCAM­1  and CYR61  as  found  by microarrayanalysis. 

Figure  31: Downregulation  of  CXADR,  SDC1  and  AQP3  gene  expression  of  cultured  LECs  by TNFα  stimulation.  Primary  dermal  LECs were  cultured  in  EBM‐2/0.5%  FCS  and  stimulated with 10ng/ml TNFα for 24 hours (with or without TNFα inhibitor). TNFα treatment of primary cultured LECs  perfectly  recapitulated  downregulation  of  SDC1,  CXADR  and  AQP3  as  found  by  microarray analysis. 

Page 113: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  98

  

 

Figure 32: Downregulation of FABP4, GALNTL2 and APOD gene expression of cultured LECsby  TNFα  stimulation.  Primary  dermal  LECs were  cultured  in  EBM‐2/0.5%  FCSand  stimulatedwith 10ng/ml TNFα  for 24 hours  (with or without TNFα  inhibitor). Downregulation of FABP4,GALNTL2 and APOD expression by TNFα treatment is contrary to array results. 

Figure  33:  No  change  of  MMP2  and  NOX4  gene  expression of  cultured  LECs  by  TNFα stimulation.  Primary  dermal  LECs  were  cultured  in  EBM‐2/0.5%  FCS  and  stimulated  with 10ng/ml TNFα for 24 hours (with or without TNFα inhibitor). No effect of TNFα could be found on MMP2 and NOX4 expression. 

Page 114: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  99

7.4.1 Increased migration of LECs upon TNFα stimulation  

Interestingly,  expression  of  several  genes  integrated  in  the  clusters  of  ‘cell  adhesion’  and 

‘inflammatory  response’  were  regulated  by  TNF‐α  and  the  in  vitro  stimulation  experiments 

matched  the  array  data.  Therefore,  the  effect  of  TNFα  on  LEC  behavior  was  studied  more 

detailed.  

LVs  are  able  to  react  to  an  inflammatory milieu  by  increasing  lymphangiogenesis  (Cueni  and 

Detmar,  2008).  TNFα  is  a  pro‐inflammatory  cytokine  which  can  stimulate  EC  migration  and 

proliferation (Gao et al., 2002). In order to analyze the effect of TNFα on the behaviour of human 

primary LECs, we performed an  in vitro  scratch wound healing assay with and without TNFα 

stimulation.  Stimulation  of  LECs  seeded  into  24‐well  plates with  10ng/ml  TNFα  for  24  hours 

significantly  increased  LEC  migration  and  induced  faster  wound  closure  in  a  scratch  wound 

assay (Figure 34, p‐value = 0.007). 

This data indicated that TNFα might lead to lymph vessel remodeling or enhanced lymph vessel 

growth (Baluk et al., 2009) and  further corroborated a potential  link between  increased TNFα 

levels and increased LV density in the skin of T2DM patients. 

 

 

7.4.2 Macrophage adhesion to LECs is increased by TNFα stimulation  

Molecular mechanisms which regulate  the entry of  inflammatory cells  from the blood  into  the 

lymphoid  tissues  during  their  circulation  in  the  body  are  well  characterized  (Miyasaka  and 

Tanaka, 2004; von Andrian and Mempel, 2003). In contrast, much less is known about the exit 

mechanisms of these cells, especially of macrophages, from the surrounding tissues into afferent 

lymphatics. However, migration of professional  antigen‐presenting cells  like  lymphocytes, DCs 

and macrophages  is  essential  for  controlling  immune  responses  (Angeli  and Randolph, 2006). 

Increased  TNFα  levels  were  found  in  T2DM  patients  as  well  as  T2DM  mouse  models. 

Additionally, it was shown that endothelial interactions with immune cells are increased during 

hyperglycemia (Algenstaedt et al., 2003). 

 

 

 

Page 115: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  100

 

 

Therefore,  we  tested  whether  TNFα  exerted  effects  on  the  interaction  between  LECs  and 

macrophages.  In  vitro  stimulation  of  primary  LECs  with  TNFα  led  to  a  massively  increased 

adhesion of macrophages (p‐value < 0.001). This was highly specific for TNFα as this effect could 

be  reduced  to  the  amount  of  adherent  macrophages  of  unstimulated  LECs  by  additional 

treatment with an inhibitory anti‐TNFα antibody (Figure 35). 

Conclusively,  we  could  show  increased  macrophage  adhesion  to  LEC  monolayers  after  TNFα 

stimulation,  indicating  massive  upregulation  of  adhesion  molecules  on  LECs  responsible  for 

macrophage  adhesion.  Similar  results  were  shown  for  DCs,  adhering  to  and  transmigrating 

through TNFα‐stimulated LECs, which was dependent  on VCAM­1  expression  (which was  also 

upregulated in dLECs compared to ndLECs) in vitro and in vivo (Johnson et al., 2006). This fact 

indicats that macrophages might undergo enhanced lymphatic traffcking in T2DM. 

 

 

Figure 34: Increased migration of LECs after TNFα stimulation. Primary dermal LECs werecultured  in EBM‐2/0.5% FCS and stimulated with or without 10ng/ml TNFα  for 24 hours. Anartificial wound was made  in each well. Representative  images at  timepoints  t=0h,  t=12h andt=24h and quantification of measured closure of wound area are shown. 

Page 116: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  101

 

 

 

 

 

 

 

Figure 35: Adhesion of macrophages to primary LECs  in vitro. Primary dermal LECs werestarved  overnight  in  EBM‐2/0.5% FCS  and  stimulated with  or without  10ng/ml  TNFα  for  24hours (with or without TNFα inhibitor). Increased macrophage adhesion to LECs was detectedupon TNFα treatment, which was reversible with an anti‐TNFα antibody; Size bar: 100µm. 

Page 117: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  102

7.5  C­X­C motif  chemokine  10  (CXCL10)  expression  and  function  in LECs  

The  array  results  indicated  a  prominent  upregulation  of CXCL10  in  dLECs  (Table  16),  and we 

could  recapitulate  strong  upregulation  of  CXCL10  expression  in  LECs  by  TNFa  stimulation 

(Figure 29). CXCL10 is a small cytokine and is also called interferon gamma‐induced protein 10 

(IP‐10). It has fundamental roles in effector T‐cell generation and function (Dufour et al., 2002). 

Moreover, it is suggested to be involved in the invasion of immune cells into the brain, thereby 

contributing  to  the  pathogenesis  of  multiple  sclerosis  (Salmaggi  et  al.,  2002).  Additionally, 

CXCL10 contributes to airway inflammation in asthma (Medoff et al., 2002) and its upregulation 

is accompanied by a delayed resolution of  local  inflammation (Rahman et al., 2010). CXCL10  is 

angiostatic and was shown to lead to apoptosis and dissociation of newly formed BVs (Bodnar et 

al.,  2009,  2006).  In  diabetes, CXCL10  leads  to  beta  cell  destruction  via  interaction with  TLR4 

(Schulthess  et  al.,  2009).  Here,  we  aimed  to  analyze  the  functional  implications  of  CXCL10 

expression in LECs. 

 

7.5.1 CXCL10 is upregulated and secreted by LECs upon TNFα stimulation  

As  shown  by  realtime  PCR,  CXCL10  was  not  expressed  in  unstimulated  but  dramatically 

upregulated  in  TNFα‐stimulated  LECs  (see  Figure  30).  This  upregulation  could  be  further 

confirmed  by Western  blotting  analysis  on  a  protein  level  (Figure  36 A). Moreover, we  could 

show that LECs secrete  this chemokine upon stimulation with TNFα using ELISA and Western 

blotting of supernatants (Figure 36 B). 

 

Page 118: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  103

 

Figure 36: Expression and  secretion of CXCL10  in LECs upon TNFα  stimulation.  Primary dermal  LECs  were  starved  overnight  in  EBM‐2/0.5%  FCS  and  stimulated  with  or  without 10ng/ml  TNFα  for  24  hours  (with  or  without  TNFα  inhibitor).  Cells  were  lysed  in  Laemmli sample  buffer  and  respective  supernatants were  collected.  A: CXCL10  expression  in  LECs was checked  using western  blotting.  GAPDH was  used  as  loading  control. CXCL10  is  not  visible  in unstimulated but massively upregulated in TNFα‐stimulated LECs. B: Supernatants of stimulated and  unstimulated  LECs  were  checked  for  CXCL10  secretion  by  ELISA  and  Western  blotting showing secretion of this chemokine by TNFα‐stimulated LECs.  

 

7.5.2 CXCL10 mediates macrophage adhesion to LECs  

Due to the fact that CXCL10 was upregulated and secreted upon TNFα treatment, and that TNFα 

significantly  increased adhesion of macrophages  to  the LEC monolayers, we hypothesized that 

CXCL10  could  be  one  factor  responsible  for  the massively  increased macrophage  adhesion  to 

LECs. Therefore, we stimulated primary LECs with 10ng/ml TNFα, with or without an inhibitory 

antibody to CXCL10. A massively increased adhesion of THP‐1 macrophages to LECs was found 

upon  TNFα  treatment,  which  was  almost  completely  abrogated  by  addition  of  an  inhibitory 

CXCL10 antibody (Figure 37, p‐value = 0.04). Conversely, treatment of LECs with TNFα plus an 

inhibitory antibody to VCAM­1 – which was also upregulated by TNFα treatment and shown to 

be  responsible  for  DC  adhesion  in  LECs  (Johnson  et  al.,  2006)  ‐  did  not  affect  macrophage 

adhesion to the lymphatic endothelium (Figure 38, p‐value = 0.65). This suggests that CXCL10 is 

an important mediator of macrophage adhesion to lymphatics, possibly only in an inflammatory 

condition when TNFα levels are increased and, hence, lead to its secretion by LECs. 

 

 

Page 119: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  104

 

 

 

 

 

 

 

Figure  37:  CXCL10  is  responsible  for  macrophage  adhesion  to  lymphatic  endothelialmonolayer. Primary dermal LECs were starved overnight in EBM‐2/0.5% FCS and stimulated withor without 10ng/ml TNFα for 24 hours with or without an inhibitory antibody to CXCL10. Increasedmacrophage adhesion to LECs could be found upon TNFα treatment which was reversed by CXCL10inhibitory antibody; Size bar: 100µm. 

Page 120: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  105

 

 

 

 

 

 

 

 

Figure  38:  VCAM­1  is  dispensable  for  macrophage  adhesion  to  lymphatic  endothelial monolayer.  Primary dermal  LECs were  starved  overnight  in EBM‐2/0.5% FCS  and  stimulatedwith or without 10ng/ml TNFα for 24 hours with or without an inhibitory antibody to VCAM­1.VCAM­1 inhibitory antibody did not change macrophage adhesion to LECs; Size bar: 100µm. 

Page 121: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  106

7.5.3 CXCL10 induces chemotaxis of macrophages  

Having shown that CXCL10 is secreted by LECs upon TNFα stimulation and that CXCL10 is in part 

responsible for macrophage adhesion, we asked whether CXCL10 secreted from stimulated LECs 

had  a  chemotactic  effect  on  macrophages.  Therefore,  we  used  the  supernatants  of  LECs 

stimulated with or without TNFα ‐ with addition of inhibitory antibodies to TNFα and CXCL10 ‐ 

and performed an agarose spot assay  (Wiggins and Rappoport, 2010). As shown  in Figure 39, 

there was increased migration of macrophages into the agarose spots which contained TNFα (C) 

compared to spots lacking TNFα (B). Increased migration induced by TNFα could be significantly 

inhibited by an anti‐TNFα antibody (D). Similarly, when using a specific  inhibitory antibody to 

CXCL10, macrophage migration  into  the agarose spots could be reduced  to  the same extent as 

with the anti‐TNFα antibody at both timepoints (p‐value = 0.008, E). 

Conclusively, CXCL10 was identified as a factor regulated by TNFα in LECs. Our findings suggest 

that  this  chemokine  could  be  responsible  for  the  chemotaxis  of  macrophages  towards  LVs. 

Additionally, it could be possible that the interaction between lymphatic endothelium and other 

inflammatory cells are affected by CXCL10, e.g. T‐cells and DCs, as well as the interaction of LVs 

with tumor cells. However, additional experiments by using a recombinant form of CXCL10 are 

needed to confirm observed effects on adhesion and chemotaxis of macrophages to LECs. 

 

 

Page 122: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  107

Figure 39: Secreted CXCL10 induces chemotaxis of macrophages. Primary dermal LECs werestarved  overnight  in  EBM‐2/0.5%  FCS  and  stimulated with  or without  10ng/ml  TNFα  for  24hours  plus  an  inhibitory  antibody  to  TNFα  and  CXCL10.  Supernatants  were  collected  andconcentrated for use in an agarose spot assay to measure chemotaxis of macrophages. Migrationarea  was  evaluated  after  24  and  48  hours,  respectively.  Representative  pictures  andquantification of migration area  (in µm2)  shows an  increased chemotaxis of macrophages  intospots  containing  TNFα  (C) which was  specifically  inhibited  by  an  inhibitory  antibody  againstTNFα (D) and CXCL10 (E). PBS (A) and supernatant of LECs not stimulated with TNFα (B) wereused as controls. 

Page 123: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  108

7.5.4 CXCL10 enhances LEC­transmigration of macrophages  

Similar  to  CCL27  in  LECs,  which  was  shown  to  be  responsible  for  the  transmigration  of 

CCR10+T‐lymphocytes  (Wick  et  al.,  2008),  we  hypothesized  that  CXCL10  could  also  confer 

transmigration of THP‐1 macrophages. Therefore, we performed a transwell migration assay by 

culturing LECs on the underside of fibronectin‐coated transwell filters. After having grown to   

 

 

Figure  40:  CXCL10  confers  macrophage  transmigration. Dermal  LECs  were  grown  toconfluence on the underside of fibronectin‐coated transwell filters and stimulated with 10ng/mlTNFα  for 24 hours with  or without  an  inhibitory  antibody  to  CXCL10  to  show  involvement ofCXCL10 in macrophage transmigration. After putting the macrophages into the upper chamber,fluorescence of transmigrated macrophages was measured in the lower chamber after one (A),three (B) and six hours (C). As a control for LEC monolayer permeability, measurement of FITC‐dextran  permeability  of  conuent  monolayers  showed  no  difference  when  stimulated  withdifferent  substances  (D).  Conclusively,  TNFα  stimulation  led  to  enhanced  macrophagetransmigration, which could be specifically diminished by an inhibitory CXCL10 antibody. 

Page 124: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  109

confluence,  LEC monolayers were  stimulated with  TNFα with  or without  a  specific  inhibitory 

CXCL10 antibody. TNFα significantly enhanced transmigration of macrophages after one,  three 

and six hours, which was blocked by addition of a specific inhibitory anti‐TNFα antibody. When 

adding  a  blocking  anti‐CXCL10  antibody,  transmigration  of  macrophages  was  significantly 

reduced  at  the  3  hours  stimulation  timepoint,  but  only  slightly  after  one  and  six  hours  (see 

Figure  40 A‐C).  The  increase  of  transmigrated macrophage  number was  not  due  to  increased 

permeability of the confluent LEC monolayers, as permeability for FITC‐dextran was unchanged 

(Figure 40 D). This indicated that CXCL10 is substantially involved in mediating transmigration 

of  macrophages  through  lymphatic  endothelium.  CXCL10  might  regulate  transmigration  via 

binding to CXCR3 expressed on THP‐1 macrophages, which  is  the counterreceptor  for CXCL10 

(Singh et al., 2007). Similar to TNFα, CXCL10 could be an important, yet unrecognized regulator 

of immune cell trafficking in the context of persistent T2DM skin inflammation. 

 

In summary, we could show that CXCL10 was highly upregulated in T2DM skin, possibly due to 

higher levels of TNFα, as it could not be detected in non‐diabetic LECs or LECs not treated with 

TNFα.  Further,  CXCL10  was  shown  to  be  secreted  by  LECs  after  TNFα‐stimulation.  CXCL10 

derived from LECs might be crucial not only for macrophage adhesion but also  for chemotaxis 

and transmigration in a pro‐inflammatory milieu.  

Page 125: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  110

7.6  Characterization  of  Fatty  acid  binding  protein  4  (FABP4) expression and function in LECs  

 

Among  the  metabolically  altered  genes  in  dLECs,  we  detected  strong  upregulation  of  FABP4 

(Table 14).  Fatty acid binding proteins  represent a  family of  lipid  chaperones, which  facilitate 

the  transport  of  fatty  acids  to  different  compartments  of  the  cell,  including  lipid  droplets, 

endoplasmatic  reticulum,  nucleus,  cytoplasm,  peroxisomes  or  mitochondria  (Furuhashi  and 

Hotamisligil,  2008).  Additionally,  they  are  implicated  in  integrating  lipid  signaling  and 

inflammatory pathways. Their expression is highly conserved as they are expressed in different 

species  from  Drosophila  melanogaster  and  Caenorhabditis  elegans  up  to  mice  and  humans 

(Makowski and Hotamisligil, 2004). Fatty acid binding protein 4 (FABP4), also called adipocyte 

protein 2 (aP2) is mainly expressed in adipocytes, macrophages and DCs. Its expression is highly 

regulated  by  fatty  acids,  PPARγ‐agonists  and  insulin  (Furuhashi  and  Hotamisligil,  2008).  Its 

importance  in  the  pathogenesis  of  T2DM  and  IR  was  demonstrated  in  Fabp4  knockout  mice 

(Hotamisligil  et  al.,  1996).  FABP4  expression  has  not  been  reported  in  LECs  before,  although 

hints exist that this gene could be important for LEC behavior, as it was reported as a potential 

candidate gene involved in lymphedema (Ferrell et al., 2008). 

 

Page 126: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  111

7.6.1 FABP4 is specifically expressed in lymphatic endothelial cells  

FABP4 was upregulated in dLECs versus ndLECs. Additionally, we found it exclusively expressed 

in  LECs  compared  to  BECs.  This  differential  expression  could  be  confirmed Western  blotting 

analysis  (Figure  41  A).  Additionally,  FABP4  could  be  detected  by  immunofluorescence  in 

podoplanin‐positive LECs (Figure 41 B). Moreover, FABP4 co‐localized with podoplanin‐positive 

LVs in human skin (Figure 41 C), whereas on a consecutive section, blood vessels stained with 

endoglyx‐1 do not show FABP4 expression (Figure 41 D). 

 

 

Figure 41: FABP4 is specifically expressed in LECs compared to BECs. A: Protein expression analysis  of  FABP4  in  LECs  and  BECs  is  shown.  GAPDH  was  used  as  a  loading  control.  B: Immunofluorescent  stainings  of  LECs  with  antibodies  to  podoplanin  and  FABP4  showed  the strong  FABP4  expression  of  LECs.  C:  Double  staining  of  human  skin  with  antibodies  to podoplanin  and FABP4  showed FABP4  expression by  LVs. D: Double  staining  of  a  consecutive section  using  antibodies  to  podoplanin  and  endoglyx‐1, which  is  highly  specific  for  BVs.  This confirms that FABP4 is not expressed on BVs of human skin. Size bar: 20µm.  

 

Page 127: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  112

7.6.2 FABP4 expression could be specifically knocked down in LECs  

To analyze whether FABP4  knockdown has an  impact on cell behavior, LECs were  transfected 

with FABP4‐specific siRNA and scrambled siRNA as control using RNAiFect. Efficient knockdown 

of  FABP4  expression  was  shown  by  Western  blot  analysis  (Figure  42  A)  as  well  as 

immunofluorescent stainings (Figure 42 B). 

Figure 42: siRNA­mediated knockdown of FABP4 in LECs. LECs were transfected with FABP4‐specific or scrambled siRNA using RNAiFect. Knockdown of FABP4 expression was analyzed byWestern blotting (A) and immunofluorescence stainings (B). 

Page 128: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  113

7.6.3 FABP4 regulates LECs proliferation  

Elmasri et al. (2009) showed that FABP4 expression is induced by VEGF stimulation in HUVECs 

and that it is involved in the regulation of HUVEC proliferation. In order to analyze the specific 

role of FABP4 in LECs, we comparatively determined the proliferation rate of scrambled versus 

FABP4  siRNA‐transfected LECs. As  shown  in Figure 43,  proliferation  of  LECs was  significantly 

reduced in FABP4 siRNA‐transfected LECs after 96 and 144 hours (p‐value = 0.04) compared to 

LECs transfected with scrambled control siRNA. However, in contrast to the findings of Elmasri 

et al. (2009), we did not find induction of FABP4 expression by VEGFs (Figure 44). This suggests 

that  besides  VEGFs  other,  yet  unidentified  factors  are  responsible  for  highly  specific  FABP4 

expression in LECs and LVs compared to BECs and BVs. 

 

Figure 44: FABP4 expression is not upregulated by lymphangiogenic factors. Cell lysates ofLECs  stimulated  with  VEGF‐A,  VEGF‐C  and  VEGF‐D  were  analyzed  for  protein  expression  ofFABP4.  No  differences  of  FABP4  expression  were  observed.  GAPDH  was  used  as  a  loadingcontrol. 

Figure 43: FABP4 regulates LEC proliferation. LECs were transfected with FABP4‐specific or scrambled siRNA using RNAiFect and cells were counted at respective timepoints. 

Page 129: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  114

7.6.4 FABP4 expression increases LEC permeability  

Using these siRNA‐mediated transfection tools, we investigated whether manipulation of FABP4 

expression  has  an  impact  on  LEC‐LEC  adhesion  and  on  the  stability  of  inter‐lymphatic 

endothelial cell junctions by measuring FITC‐dextran permeability and transendothelial electric 

resistance  (TEER)  of  a  LEC monolayer.  Reduced permeability  for  FITC‐dextran was measured 

when  FABP4  expression  was  knocked  down  in  LECs  (Figure  45  A).  In  concordance  with  the 

reduced  permeability,  we  found  increased  TEER  in  FABP4  knockdown  LECs  compared  to 

scrambled siRNA transfected cells (Figure 45 B). These data indicate that FABP4, besides its well 

established  function  as  a  fatty  acid  transporter  in  adipocytes,  seems  to  exert  important  basic 

cellular features in LECs. 

 

 

7.6.5 FABP4 expression regulates PPARγ expression in LECs  

It  has  been  shown  that  peroxisome  proliferator  activated  receptor  gamma  (PPARγ)  regulate 

FABP4 expression in adipocytes and fibroblasts (Furuhashi and Hotamisligil, 2008). Hence, we 

wanted  to  know  if  this  concept  also  applies  to  LECs.  We  transfected  LECs  with  FABP4‐  and 

PPARγ‐specific siRNAs. As shown in Figure 46, knockdown of FABP4 led to a strong decrease of 

Figure  45:  FABP4  regulates  permeability  of  LEC  monolayers. LECs  were  seeded  onfibronectin (FN) coated transwell filters and grown to confluence. A: FITC‐dextran diffusion wasmeasured using an ELISA reader at λ=485‐530nm in transwells with LEC monolayers transfectedwith  FABP4‐specific  and  scrambled  siRNA  and  in  FN  coated  wells  only.  B:  Transendothelialelectric resistance (TEER) was measured using a manual electrode. 

Page 130: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  115

FABP4  expression  and  similarly  also  PPARγ  expression  in  LECs  (A).  Conversely,  PPARγ 

knockdown did not influence FABP4 expression in LECs (B). 

This  could  indicate  that FABP4  expression  in  LECs  is  driven  by  a  rather  solid  differentiation‐

dependent transcriptional regulation, while PPARγ is sensitive to the presence of FABP4.  

 

 

 

 

7.6.6 Possible interactions of FABP4 with PPARγ   

FABPs  serve  as  shuttle  proteins  for  fatty  acids  in  order  to  activate  PPARs  (Wolfrum,  2007), 

which then,  in concert with retinoic acid receptors, activate so‐called peroxisome‐proliferator‐

responsive element (PPREs) in genes. FABP4 was shown to directly interact with PPARγ in COS‐

7  cells  (Tan  et  al.,  2002).  By  functional  analysis  of  PPRE motifs  in  genes  of  fatty  acid‐binding 

proteins,  a  direct  binding  of  PPARγ/RXR  heterodimers  to  PPREs  in  the  FABP4  promoter was 

demonstrated  (Schachtrup  et  al.,  2004).  By  this  way,  FABPs  activate  their  own  transcription, 

regulating  a  positive  feedback‐loop.  Moreover,  it  was  shown  that  treatment  with  the  PPARγ 

agonist rosiglitazone upregulated the expression of FABP4 and PPARγ (Cabré et al., 2007; Allen 

et  al.,  2006)  and  a  nuclear  translocation  of  FABP4  occured  (Ayers  et  al.,  2007;  Gillilan  et  al., 

2007), thereby possibly stimulating the interaction of FABP4 and PPARγ. 

Due to the fact that FABP4 regulated PPARγ expression in LECs (Figure 46), we were interested 

to know whether these two genes were interacting directly with each other in LECs. Therefore, 

we  performed  co‐immunoprecipitations  with  anti‐FABP4  and  anti‐PPARγ  antibodies, 

Figure 46: FABP4 regulates PPARγ expression in LECs but not vice versa. 5x104 LECs wereseeded  into  24  well  plates  and  transfected  with  FABP4‐/PPARγ‐specific  siRNA  as  well  asscrambled  siRNA  using  RNAiFect  transfection  reagent.  PPARγ  (A)  and  FABP4  (B)  proteinexpression  was  checked  usingWestern  blot  analysis.  GAPDH  protein  expression  served  as  aloading control. 

Page 131: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  116

respectively  from whole  LEC  lysates.  As  shown  in  Figure  47,  we  could  not  detect  any  direct 

interaction  between  FABP4  and  PPARγ  in  either  of  the  two  immunoprecipitates.  This  finding 

does not completely exclude an interaction of these two molecules in LECs. Rather, low amounts 

of starting material,  too  low expression of PPARγ in LECs or too weak or too short  interaction 

between these proteins could be the limitations of our practical approach. 

 

 

 

In order to see whether PPARγ protein binds to PPREs in the promotor region of FABP4 gene, 

we performed ChIP assays as previously described (Gal‐Yam et al., 2008). 

 

 

Figure 47: Co­Immunoprecipitations of FABP4 and PPARγ.  Co‐immunoprecipitation,  usinganti‐FABP4  or  anti‐PPARγ  antibodies,  from  the  lysates  of  LECs  stimulated  with  20µM  ofrosiglitazone. The precipitates were  analyzed by  immunoblotting. Western blots  indicate  thatproteins are not interacting with each other. 

Figure 48: Chromatin  immunoprecipitations of PPARγ  to  show potential  interactions withPPREs  in  the  FABP4  promotor  region.  Chromatin  immunoprecipitations  of  LEC  lysatesstimulated with or without 20µM rosiglitatzone  for 24 hours. Binding of PPARγ to PPREs waschecked  using  SybrGreen  quantitative  real‐time  PCR  and  specific  primer  pairs.  MYT1  primerwas used as negative control. 

Page 132: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  117

In  all  of  the  six  potential  PPRE  sites,  except  for  site  6,  no  significant  enrichment  in  site 

amplification  by  quantitative  PCR  could  be  detected  between  unstimulated  and  rosiglitazone‐

stimulated  LECs  (Figure  48).  Site  6  represents  a  PPRE  in  the  core  promotor  region  of  FABP4 

very  near  to  the  5'‐UTR  and  the  start  codon  of  the  FABP4  gene,  suggesting  that  PPARγ 

preferentially  binds  to  this  PPRE  of  endothelial  FABP4.  The  primer  for  myelin  transcription 

factor 1 (MYT1) served as a negative control. 

 

In summary, we could show that FABP4 is significantly increased in diabetic LECs. It could serve 

as  a  new  specific  marker  for  LECs,  as  it  is  not  expressed  in  BECs  and  BVs  of  human  skin. 

Moreover,  it  significantly  influences  proliferation  and  permeability  of  LECs  and  influences 

PPARγ  expression  after  knockdown.  FABP4  could  represent  one  important  connector  gene 

between lipid metabolism and lymphatic vessel functioning, as it was shown that LV dysfunction 

lead to  lymphedema, which  is associated with  increased formation of adipose tissue (Rockson, 

2001) and increased obesity in mice (Harvey et al., 2005).  

 

 

 

Page 133: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  118

  8. Conclusions and Future Perspectives  

A final model for involvement of dermal lymphatic vessels in type 2 diabetes 

 

1. Morphological analysis 

 

During this thesis work, a comprehensive analysis of LVs and LECs in the skin of type 2 diabetic 

patients was performed. A morphological as well as molecular characterization is presented. A 

great merit  is  that  this work boldly analyzed human  tissue specimens, and  it addressed a  less 

studied  vascular  system  in  diabetes.  While  endothelial  dysfunction  of  BVs  is  a  well‐studied 

complication in T2DM, research data on how lymphatics may be altered and involved in disease 

pathogenesis is barely known so far. However, skin complications associated with diabetes are 

not only related to BV but additionally to LV dysfunction. LV malfunctioning may lead to wound 

healing  defects  (Saaristo  et  al.,  2006),  increased  risk  for  local  infections  (Rockson,  2001)  and 

formation of adipose tissue (Harvey et al., 2005).  

Increased  lymphatic  vessel  density  was  observed  in  cholesterol‐rich  atherosclerotic  lesions 

(Kholova et al., 2010) and chronic venous insufficiency ulcers (Fernandez et al., 2011) of human 

patients,  suggesting  causal  relationships  between  these  diseases.  In  contrast  to  newly  grown 

carcinoma‐associated  lymphatic vessels  (Vainionpää et  al., 2007), we could not detect  laminin 

and  type  IV  collagen  expression  alterations,  which  was  consistent  with  unaltered  lymphatic 

vessel  diameter  and  absence  of  hyperplasia.  Although  a  'so‐called'  diabetic  lymphangiopathy 

was  reported  earlier  (Kaufmann et  al.,  1980; Ohkuma,  1979),  no phenotypical  changes of  LVs 

could be found in the skin of T2DM patients, regardless of  their size. We conclude that mostly 

due  to  a  completely  different  assembly  of  the  vessel wall  (no  regular  BM,  no  SMC  coverage), 

observed changes are not similar to BV changes in T2DM. Additionally, intraluminal LV pressure 

is significantly lower and lymph flow rate is approximately 100‐500 times less than blood flow 

rate  (Swartz,  2001).  However,  if  intraluminal  pressure  and  lymph  flow  rate  is  changed  in  a 

similar way than in BVs in type 2 diabetics (Calles‐Escandon and Cipolla, 2001), this would be an 

interesting features to measure, although this is very difficult to conduct in humans. 

Page 134: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  119

Significantly  increased  LV  and  macrophage  density  was  found  in  the  skin  of  T2DM  patients. 

Although controversial  to previously published studies  (Maruyama et al.,  2007; Saaristo et  al., 

2006), we suggest that this could be an intercorrelating feature, where macrophages regulate LV 

density via the expression of VEGFs (Machnik et al., 2009; Maruyama et al., 2005) and possibly 

also via TNFα, which has pro‐migratory and proliferative activity. It needs further investigations 

to  answer  the  question  why  there  is  increased macrophage  infiltration  in  the  skin  of  type  2 

diabetic  patients.  Is  it  related  to  ischemia,  to  hyperglycemia  or  is  it  a  result  of  reduced 

efferocytosis  due  to  a  dysfunction  of  macrophages  which  in  turn  could  sustain  local 

inflammation?  

 

2. The dLEC transcriptome 

 

Extensive bioinformatical and literature analysis of microarray data retrieved a long list of genes 

which  were  dissected  into  different  functional  pathways  using  Ingenuity  Pathway  Analysis 

(IPA).  This  led  to  the  establishment  of  four  overrepresented  themes,  comprising  (Cluster  A) 

defense  response  and  inflammation,  (Cluster  B)  tissue  remodeling  and  cell motility,  (Cluster  C) 

lymphangiogenesis  and  cell  fate  regulation  and  (Cluster  D)  lipid  handling  and  small molecule 

transport.  

We did not trace downregulation of key lymphatic differentiation markers, as was observed in 

obesity and inflammation (Horra et al., 2009; Vigl et al., 2011), and we did not trace deregulation 

of  transcripts  for  lymphatic  valve  and  junction  proteins  (Hämmerling  et  al.,  2006),  though 

lymphatic vessels were postulated to contribute to obesity due to disrupted integrity (Harvey et 

al., 2005). Rather,  the overlaps with hypoxic LECs and LECs derived from skin of  lymphedema 

patients  (Irigoyen et al., 2007; Ogunbiyi et al., 2011) emphasized a  role of  these candidates  in 

pathological  lymphatic  dysfunction.  Further,  overlaps  with  transcriptomes  of  whole  tissue 

lysates  from nonhealing  venous ulcers  (Charles  et  al.,  2008), wound  inflammation  (Roy  et  al., 

2008) and diabetic wound microbiome (Grice et al., 2010) might indicate that these specifically 

derive  from LECs. Therefore, we concluded  that  lymphatic vessels show stable LEC  identity  in 

type  2  diabetic  skin,  while  they  manifest  alterations  that  highlight  the  specific 

pathophysiological microenvironment. 

The  functional  gene  sets  underlined  the  emergence  of  enhanced  lymphatic  vessel  density. 

During  capillary  morphogenesis,  LECs  have  to  be  released  from  their  quiescent  phenotype, 

which  is  triggered  by  dissolution  of  the  basement  membrane  and  complex  changes  of  their 

migratory behaviour. Mature LECs scarcely proliferate and must  re‐enter  the cell  cycle during 

growth  of  new  vessels  from  pre‐existing  ones.  The  huge  number  of  positive  and  negative 

regulators of cell growth, migration, proliferation and survival suggested an activation of cellular 

differentiation and survival processes in LECs in T2DM skin. The transcriptional downregulation 

Page 135: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  120

of GTPase‐activating molecules might indicate relief of a quiescent LEC proliferation state, as it 

was described recently for Ras GTPase‐activating RASA1 (Lapinski et al., 2012). dLECs revealed 

decrease of p53‐mediated downstream effectors, hence extending  their proliferation potential. 

Accordingly, p53‐deficient mice  show enhanced  lymphangiogenesis  (Ruddell  et  al.,  2008),  and 

silencing of p53 signaling improved diabetic wound healing (Nguyen et al., 2010). However, the 

master‐regulators  dictating  the  concerted  gene  suppressions  of  several  gene  clusters  remain 

unknown,  suggesting  involvement  of  epigenetic,  transcriptional,  or  microRNA‐mediated 

mechanisms. Recent  findings point  towards  such  regulatory  traits  in diabetic  endothelial  cells 

(Pirola et al., 2011; Cayrol et al., 2007; Zampetaki et al., 2010).  

 

3. Mechanistic analysis that brings all findings together 

 

In T2DM, the interstitial space of the dermis contains an overload of fatty acids, glucose, proteins 

and adipogenic factors derived from the blood circulation (Tammela et al., 2010). Despite absent 

regulation of adipokines or their cognate receptors in dLECs, enhanced lipid levels per se might 

contribute  to  pathological  alterations  of  lymphatic  vessels.  Dyslipidemia  was  associated  with 

dysfunctional lymphatic vessels that lead to chronic inflammatory disorders of the skin (Lim et 

al.,  2009).  Hence,  incomplete  clearance  of  interstitial  fluid  and  its  contents  might  lead  to 

phenotypic  changes  similar  to  lymphedema  like  increased  collagen  deposition,  fibrotic 

alterations, and dermal swelling (Alitalo, 2011). Type 2 diabetic skin specimens showed obvious 

thickening (own observation), pointing at such alterations. This might be added to the obviously 

damaged  skin  barrier‐immune  axis  and  underlines  the  paramount  importance  of  initial 

lymphatics for maintenance of skin homeostasis.  

Though macrophages are essential for physiological wound healing (Mahdavian Delavary et al., 

2011),  in  type  2  diabetes,  their  increased  influx  has  been  correlated  with  impaired  wound 

healing  (Wetzler  et  al.,  2000),  characterized  by  expression  of  pro‐inflammatory  cytokines, 

especially TNF‐α (Khanna et al., 2010). Here, we show that macrophage infiltration and TNF‐α 

production  is  associated  with  human  dermal  pathology  as  well.  Activated macrophages  have 

been recognized to directly contribute to de novo  lymphangiogenesis (Kerjaschki, 2005) by (1) 

production  of  lymphangiogenic  factors  and  (2)  conversion  into  LECs.  However,  we  detected 

relatively  low  enhancement  of  VEGF‐C  levels,  which  was  not  significantly  different  between 

diabetic and non‐diabetic skin, corresponding to a poor overlap with VEGF‐C responsive genes 

in LECs (Yong et al., 2005), which suggested a minor contribution to the increased lymph vessel 

count. Pathologic lymphangiogenesis might be driven by additional mediators such as PDGF‐BB, 

HIF or FGF‐2 (Alitalo, 2011), though we did not trace altered expression of these factors or their 

cognate  receptors  in  dLECs.  Rather, we  provide  the  first  confirmation  of  TNF‐α‐induced  gene 

deregulation  in  human  LECs  in  vivo  that  is  strongly  correlated  with  a  pro‐migratory  LEC 

Page 136: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  121

phenotype. Accordingly, enhanced TNF‐α  levels were shown to drive  lymph vessel remodeling 

(Baluk et al., 2009).  

Besides indications from murine studies (Tammela & Alitalo, 2010), in humans, enhanced lymph 

vessel density was correlated with accelerated chronic wound healing  (Labanaris et  al., 2009) 

and  restoration  of  gut  homeostasis  in  Crohn’s  disease  (von  der Weid  et  al.,  2011).  However, 

several  reports  indicate  that  inflammatory‐driven  lymphangiogenesis  leads  to  dysfunctional 

vessel  formation  showing  reduced  lymph  drainage  capacity  (Tammela  &  Alitalo,  2010).  The 

active recruitment of macrophages to dLECs via CXCL10 chemotaxis might be indicative of such 

impeded  clearance,  but  could  also  be  the  prerequisite  to  lymph  vessel  integration.  In  a  LPS‐

driven  peritonitis  model,  enhanced  numbers  of  macrophages  were  closely  attached  to  newly 

formed inflammatory lymphatic vessels, then directly incorporating into these (Kim et al., 2009). 

It  is  tempting  to  speculate  that  macrophages  in  human  patients´  skin  are  involved  in  an 

analogous mechanism of lymphatic vessel expansion. This exaggerated macrophage entrapment 

and de novo  lymph vessel formation could lead to decreased tissue fluid, lipid and immune cell 

drainage  and,  finally,  persistent  inflammation  that  altogether  hinder  skin  regeneration.  The 

question  whether  exaggerated  lymph  vessel  formation  is  beneficial  for  the  skin  pathology  in 

type 2 diabetes, has to be explored. 

To summarize, we revealed that dLECs contribute to chronic inflammation, decreased defense of 

infections, leukocyte recruitment, tissue remodeling and severely altered homeostasis of T2DM 

skin. Figure 49 summarizes the effects evoking the altered dLEC properties, including increased 

load  of  metabolites  and macrophage  influx  over  time,  which  lead  to molecular  alterations  in 

dermal  LECs.  LECs  in  T2DM  skin  show  an  unprotected,  activated  phenotype  characterized  by 

increased inflammatory, migratory, lipid handling and lymphangiogenic capacity and apoptosis 

resistance.  A  paracrine  cross‐talk  between  macrophages  and  dLECs  leads  to  increased 

macrophage  recruitment  which  might  be  the  source  of  enhanced  lymph  vessel  expansion. 

Specifically, TNF‐α  is a key mediator of cross talk between proinflammatory macrophages and 

LECs. TNF‐α  

Conclusively,  the  analysis  of  lymphatic  endothelial  cells  from  different  anatomic  sites  and 

pathological situations allows a better understanding of the mechanisms involved in lymphatic 

vascular  growth,  function  and  repair,  depending  on  the  specific  pathophysiological 

microenvironment.  Transcriptomal  comparison  led  to  the  identification  of  a  distinct  gene 

expression profile that characterized lymphatic vascular changes in T2DM on a molecular level, 

suggesting  that LECs are undergoing major changes during T2DM,  including changes  in vessel 

permeability,  lipid  transport, proliferation and vessel  remodelling.  Importantly,  the array data 

underlined our morphological  findings. We could show a correlation between the gene cluster 

‘inflammation’ with  the  increased macrophage  infiltration  and  TNFα  expression.  Additionally, 

Page 137: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  122

mediated  gene  expression  alterations  in  LECs  lead  to  further macrophage  recruitment,  hence 

reinforcing lymph vessel propagation and chronic inflammation. 

 

 

 

Figure  49:  Importance  of  lymphatic  vessels  in  type  2  diabetic  skin  and  the  attempt  of establishing a model  showing  their  contribution,  involvement  and  alterations  in  the  skin  of type 2 diabetic patients. 

 

dLECs  are  characterized  by  increased  tissue  regeneration  efforts,  reflected  by  differentially 

regulated genes asscociated with wound healing and tissue remodelling as well as the fact that 

we found increased lymphangiogenesis in T2DM skin. This finding also underlined the fact that 

we detected an increased number of genes implicated in lymphangiogenesis as well as increased 

Page 138: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  123

survival. Moreover, it is suggested that LECs are directly and actively involved in lipid and small 

molecule transport, highlighted by the differentially regulated genes in the cluster ‘lipid handling 

and small molecule biochemistry’.  

The data suggest that skin lymphatic vessels react actively on the multiple, metabolic as well as 

inflammatory  interstitial  burden.  Inflammatory  reaction  is  a  prominent  event,  as  well  as 

cytoskeletal  regulatory  pathways  that  lead  to  extensive  lymphatic  vessel  expansion  and 

remodelling in T2DM. Overall, dLECs seem to be actively involved in the skin alterations in the 

advanced  stage  of  T2DM  skin  disease.  Moreover,  the  dLEC  transcriptome  could  include 

important marker  genes  for  T2DM,  which will  lead  to  alternative  therapies  as well  as  future 

candidate genes for regulating leukocyte adhesion and migration and potentially be crucial  for 

clearance of tissue inflammation in type 2 diabetic patients. 

Page 139: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  124

  References Abtahian, F., Guerriero, A., Sebzda, E., Lu, M.‐M., Zhou, R., Mocsai, A., Myers, E. E., Huang, B., Jackson, D. G., Ferrari, V. A., Tybulewicz, V., Lowell, C. A., Lepore,  J.  J., Koretzky, G. A., and Kahn, M. L. (2003). Regulation of blood and lymphatic vascular separation by signaling proteins slp‐76 and syk. Science, 299(5604):247‐251.  Agah,  A.,  Kyriakides,  T.  R.,  Letrondo,  N.,  Björkblom,  B.,  and  Bornstein,  P.  (2004).  Thrombospondin  2  levels  are increased in aged mice: consequences for cutaneous wound healing and angiogenesis. Matrix Biol, 22(7):539‐547.  Alberti,  K.  G.  and  Zimmet,  P.  Z.  (1998).  Definition,  diagnosis  and  classification  of  diabetes  mellitus  and  its complications. part 1: diagnosis and classification of diabetes mellitus provisional report of a who consultation. Diabet Med, 15(7):539‐553.  Alexander  JS,  Chaitanya  GV,  Grisham MB,  Boktor  M.  (2010).  Emerging  roles  of  lymphatics  in  inflammatory  bowel disease. Ann N Y Acad Sci, 1207 Suppl 1:E75‐85.  Algenstaedt, P., Schaefer, C., Biermann, T., Hamann, A., Schwarzloh, B., Greten, H., Rther, W., and Hansen‐Algenstaedt, N. (2003). Microvascular alterations in diabetic mice correlate with level of hyperglycemia. Diabetes, 52(2):542‐549.  Alitalo K. (2011). The lymphatic vasculature in disease. Nat Med, 17(11):1371‐1380.  Alitalo, K., Tammela, T.,  and Petrova, T. V.  (2005).  Lymphangiogenesis  in development  and human disease. Nature, 438(7070):946‐953.  Alizadeh, A. A., Eisen, M. B., Davis, R. E., Ma, C.,  Lossos,  I.  S., Rosenwald, A., Boldrick,  J.  C.,  Sabet, H., Tran, T., Yu, X., Powell,  J.  I.,  Yang,  L., Marti,  G.  E., Moore,  T.,  Hudson,  J.,  Lu,  L.,  Lewis, D.  B.,  Tibshirani,  R.,  Sherlock,  G.,  Chan, W.  C., Greiner, T. C., Weisenburger, D. D., Armitage, J. O., Warnke, R., Levy, R., Wilson, W., Grever, M. R., Byrd, J. C., Botstein, D., Brown, P. O., and Staudt, L. M. (2000). Distinct types of diffuse  large b‐cell  lymphoma identified by gene expression profiling. Nature, 403(6769):503‐511.  Allen, T., Zhang, F., Moodie, S. A., Clemens, L. E., Smith, A., Gregoire, F., Bell, A., Muscat, G. E. O., and Gustafson, T. A. (2006).  Halofenate  is  a  selective  peroxisome  proliferator‐activated  receptor  gamma  modulator  with  antidiabetic activity. Diabetes, 55(9):2523‐2533.  Altshuler, D., Hirschhorn, J. N., Klannemark, M., Lindgren, C. M., Vohl, M. C., Nemesh, J., Lane, C. R., Schaffner, S. F., Bolk, S.,  Brewer,  C.,  Tuomi,  T.,  Gaudet,  D.,  Hudson,  T.  J.,  Daly,  M.,  Groop,  L.,  and  Lander,  E.  S.  (2000).  The  common ppargamma pro12ala polymorphism is associated with decreased risk of type 2 diabetes. Nat Genet, 26(1):76‐80.  Amatschek, S., Kriehuber, E., Bauer, W., Reininger, B., Meraner, P., Wolpl, A., Schweifer, N., Haslinger, C., Stingl, G., and Maurer, D. (2007). Blood and lymphatic endothelial cell specific differentiation programs are stringently controlled by the tissue environment. Blood, 109(11):4777‐4785.  Angeli, V. and Randolph, G. J. (2006). Inflammation, lymphatic function, and dendritic cell migration. Lymphat Res Biol, 4(4):217‐228.  Aselli,  G.  (1627).  De  lacteibus  sive  lacteis  venis,  quarto  vasorum  mesarai  corum  genere  novo  invento  mediolani, milano.  Asselot‐Chapel, C., Borchiellini, C., Labat‐Robert, J., and Kern, P. (1996). Expression of laminin and type iv collagen by basement  membrane‐producing  ehs  tumors  in  streptozotocin‐induced  diabetic  mice:  in  vivo  modulation  by  low‐molecular‐weight heparin fragments. Biochem Pharmacol, 52(11):1695‐1701.  Association, A. D. (2007). Standards of medical care in diabetes. Diabetes Care, 30 Suppl 1:S4‐S41.  Association, A. D. (2005). Diagnosis and classification of diabetes mellitus. Diabetes Care, 28 Suppl 1:S37{‐42.  Ayers,  S.  D.,  Nedrow,  K.  L.,  Gillilan,  R.  E.,  and  Noy,  N.  (2007).  Continuous  nucleocytoplasmic  shuttling  underlies transcriptional activation of ppargamma by fabp4. Biochemistry, 46(23):6744‐6752.  

Page 140: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  125

Bäckhed, F., Crawford, P. A., O'Donnell, D., and Gordon,  J.  I.  (2007). Postnatal  lymphatic partitioning  from the blood vasculature in the small intestine requires fasting‐induced adipose factor. Proc Natl Acad Sci U S A, 104(2):606‐611.  Baldwin, M. E., Halford, M. M., Roufail, S., Williams, R. A., Hibbs, M. L., Grail, D., Kubo, H., Stacker, S. A., and Achen, M. G. (2005). Vascular endothelial growth  factor d  is dispensable  for development of  the  lymphatic system. Mol Cell Biol, 25(6):2441‐2449.  Balfour, J. A. and Plosker, G. L. (1999). Rosiglitazone. Drugs, 57(6):921‐930; Discussion 931‐932.  Baluk,  P.,  Tammela,  T.,  Ator,  E.,  Lyubynska,  N.,  Achen,  M.  G.,  Hicklin,  D.  J.,  Jeltsch,  M.,  Petrova,  T.  V.,  Pytowski,  B., Stacker,  S.  A.,  Ylä‐Herttuala,  S.,  Jackson,  D.  G.,  Alitalo,  K.,  and  McDonald,  D.  M.  (2005).  Pathogenesis  of  persistent lymphatic vessel hyperplasia in chronic airway inflammation. J Clin Invest, 115(2):247‐257.  Baluk, P., Yao, L.‐C., Feng, J., Romano, T., Jung, S. S., Schreiter, J. L., Yan, L., Shealy, D. J., and McDonald, D. M. (2009). Tnf‐alpha  drives  remodeling  of  blood  vessels  and  lymphatics  in  sustained  airway  inflammation  in  mice.  J  Clin  Invest, 119(10):2954‐2964.  Banerji, S., Ni, J., Wang, S. X., Clasper, S., Su, J., Tammi, R., Jones, M., and Jackson, D. G. (1999). Lyve‐1, a new homologue of the cd44 glycoprotein, is a lymph‐specific receptor for hyaluronan. J Cell Biol, 144(4):789‐801.  Bazigou, E., Xie, S., Chen, C., Weston, A., Miura, N.,  Sorokin, L., Adams, R., Muro, A. F.,  Sheppard, D.,  and Makinen, T. (2009). Integrin‐alpha9 is required for fibronectin matrix assembly during lymphatic valve morphogenesis. Dev Cell, 17(2):175‐186.  Belperio  JA,  Keane MP,  Arenberg  DA,  Addison  CL,  Ehlert  JE,  Burdick MD,  Strieter  RM.  (2000).  CXC  chemokines  in angiogenesis. J Leukoc Biol, 68(1):1‐8.  Bergman, R. N. and Ader, M. (2000). Free fatty acids and pathogenesis of type 2 diabetes mellitus. Trends Endocrinol Metab, 11(9):351‐356.  Bergqvist, L., Strand, S. E., and Persson, B. R. (1983). Particle sizing and biokinetics of interstitial lymphoscintigraphic agents. Semin Nucl Med, 13(1):9‐19.  Bertozzi, C. C., Schmaier, A. A., Mericko, P., Hess, P. R., Zou, Z., Chen, M., Chen, C.‐Y., Xu, B., min Lu, M., Zhou, D., Sebzda, E., Santore, M. T., Merianos, D. J., Stadtfeld, M., Flake, A. W., Graf, T., Skoda, R., Maltzman, J. S., Koretzky, G. A., and Kahn, M. L.  (2010). Platelets regulate  lymphatic vascular development  through clec‐2‐slp‐76 signaling. Blood, 116(4):661‐670.  Bittner, M., Meltzer,  P.,  Chen, Y.,  Jiang,  Y.,  Seftor,  E., Hendrix, M.,  Radmacher, M.,  Simon, R.,  Yakhini,  Z., Ben‐Dor, A., Sampas,  N.,  Dougherty,  E.,  Wang,  E.,  Marincola,  F.,  Gooden,  C.,  Lueders,  J.,  Glatfelter,  A.,  Pollock,  P.,  Carpten,  J., Gillanders, E., Leja, D., Dietrich, K., Beaudry, C., Berens, M., Alberts, D., and Sondak, V. (2000). Molecular classification of cutaneous malignant melanoma by gene expression profiling. Nature, 406(6795):536‐540.  Bodnar, R. J., Yates, C. C., Rodgers, M. E., Du, X., andWells, A. (2009). Ip‐10 induces dissociation of newly formed blood vessels. J Cell Sci, 122(Pt 12):2064‐2077.  Bodnar, R. J., Yates, C. C., and Wells, A. (2006). Ip‐10 blocks vascular endothelial growth factor‐induced endothelial cell motility and tube formation via inhibition of calpain. Circ Res, 98(5):617‐625.  Bonifati, C. and Ameglio, F. (1999). Cytokines in psoriasis. Int J Dermatol, 38(4):241‐251.   Breiteneder‐Geleff,  S., Matsui, K.,  Soleiman, A., Meraner, P.,  Poczewski, H., Kalt, R.,  Schachner, G.,  and Kerjaschki, D. (1997).  Podoplanin,  novel  43‐kd membrane  protein  of  glomerular  epithelial  cells,  is  down‐regulated  in  puromycin nephrosis. Am J Pathol, 151(4):1141‐1152.  Breiteneder‐Geleff,  S.,  Soleiman,  A.,  Horvat,  R.,  Amann,  G.,  Kowalski,  H.,  and  Kerjaschki,  D.  (1999).  Podoplanin  –a  specific marker for lymphatic endothelium expressed in angiosarcoma. Verh Dtsch Ges Pathol, 83:270‐275.  Bridenbaugh,  E.  A.,  Gashev,  A.  A.,  and  Zawieja,  D.  C.  (2003).  Lymphatic  muscle:  a  review  of  contractile  function. Lymphat Res Biol, 1(2):147‐158.  Brown, P. O. and Botstein, D. (1999). Exploring the new world of the genome with dna microarrays. Nat Genet, 21(1 Suppl):33‐37.  Brownlee, M. (2001). Biochemistry and molecular cell biology of diabetic complications. Nature, 414(6865):813‐820.  

Page 141: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  126

Cabré, A., Lázaro,  I., Girona,  J., Manzanares,  J. M., Marimón, F., Plana, N., Heras, M., and Masana, L. (2007). Fatty acid binding  protein  4  is  increased  in  metabolic  syndrome  and  with  thiazolidinedione  treatment  in  diabetic  patients. Atherosclerosis, 195(1):e150‐e158.  Calcutt, N.  A.,  Cooper, M.  E.,  Kern,  T.  S.,  and  Schmidt,  A. M.  (2009).  Therapies  for  hyperglycaemia‐induced  diabetic complications: from animal models to clinical trials. Nat Rev Drug Discov, 8(5):417‐429.  Calles‐Escandon,  J.  and Cipolla, M.  (2001). Diabetes  and  endothelial  dysfunction:  a  clinical perspective. Endocr Rev, 22(1):36‐52.  Cayrol C, Lacroix C, Mathe C, Ecochard V, Ceribelli M, Loreau E, Lazar V, Dessen P, Mantovani R, Aguilar L, Girard JP. (2007). The THAP‐zinc finger protein THAP1 regulates endothelial cell proliferation through modulation of pRB/E2F cell‐cycle target genes. Blood, 109(2):584‐594.  Chan, K.‐H. K., Song, Y., Hsu, Y.‐H., You, N.‐C. Y., Tinker, L. F., and Liu, S. (2010). Common genetic variants in fatty acid‐binding  protein‐4  (fabp4)  and  clinical  diabetes  risk  in  the  women's  health  initiative  observational  study.  Obesity (Silver Spring). Charles CA, Tomic‐Canic M, Vincek V, Nassiri M, Stojadinovic O, Eaglstein WH, Kirsner RS. (2008). A gene signature of nonhealing venous ulcers: potential diagnostic markers. J Am Acad Dermatol, 59(5):758‐771.  Chen, C. C., Mo, F. E., and Lau, L. F. (2001). The angiogenic factor cyr61 activates a genetic program for wound healing in human skin fibroblasts. J Biol Chem, 276(50):47329‐47337.  Chia,  S.,  Qadan, M.,  Newton,  R.,  Ludlam,  C.  A.,  Fox,  K.  A.  A.,  and Newby,  D.  E.  (2003).  Intra‐arterial  tumor  necrosis factor‐alpha impairs endothelium‐dependent vasodilatation and stimulates local tissue plasminogen activator release in humans. Arterioscler Thromb Vasc Biol, 23(4):695‐701.  Chiasson,  J.  L.,  Gomis,  R.,  Hanefeld,  M.,  Josse,  R.  G.,  Karasik,  A.,  and  Laakso,  M.  (1998).  The  STOP‐NIDDM  trial:  an international study on the efficacy of an alpha glucosidase inhibitor to prevent type 2 diabetes in a population with impaired  glucose  tolerance:  rationale,  design,  and  preliminary  screening  data.  study  to  prevent  non‐insulin‐dependent diabetes mellitus. Diabetes Care, 21(10):1720‐1725.  Christian, S., Ahorn, H., Novatchkova, M., Garin‐Chesa, P., Park, J. E., Weber, G., Eisenhaber, F., Rettig, W. J., and Lenter, M.  C.  (2001).  Molecular  cloning  and  characterization  of  endoglyx‐1,  an  emilin‐like  multisubunit  glycoprotein  of vascular endothelium. J Biol Chem, 276(51):48588‐48595.  Clasper  S,  Royston D, Baban D,  Cao Y,  Ewers  S,  Butz  S,  Vestweber D,  Jackson DG.  (2008). A novel  gene  expression profile in lymphatics associated with tumor growth and nodal metastasis. Cancer Res, 15;68(18):7293‐7303.  Cole,  K.  A.,  Krizman,  D.  B.,  and  Emmert‐Buck,  M.  R.  (1999).  The  genetics  of  cancer  ‐  a  3d model. Nat Genet,  21(1 Suppl):38‐41.  Correia,  S.,  Carvalho,  C.,  Santos, M.  S.,  Sei ca,  R.,  Oliveira,  C.  R.,  and Moreira,  P.  I.  (2008). Mechanisms  of  action  of metformin in type 2 diabetes and associated complications: an overview. Mini Rev Med Chem, 8(13):1343‐1354.  Cueni,  L.  N.,  Chen,  L.,  Zhang,  H.,  Marino,  D.,  Huggenberger,  R.,  Alitalo,  A.,  Bianchi,  R.,  and  Detmar,  M.  (2010a). Podoplanin‐fc  reduces  lymphatic  vessel  formation  in  vitro  and  in  vivo  and  causes  disseminated  intravascular coagulation when transgenically expressed in the skin. Blood 18, 116(20):4376‐4384.  Cueni, L. N. and Detmar, M. (2008). The lymphatic system in health and disease. Lymphat Res Biol, 6(3‐4):109‐122.  Cueni, L. N., Hegyi, I., Shin, J. W., Albinger‐Hegyi, A., Gruber, S., Kunstfeld, R., Moch, H., and Detmar, M. (2010b). Tumor lymphangiogenesis  and metastasis  to  lymph  nodes  induced  by  cancer  cell  expression  of  podoplanin. Am  J  Pathol, 177(2):1004‐1016.  Danussi,  C.,  Spessotto,  P.,  Petrucco,  A.,  Wassermann,  B.,  Sabatelli,  P.,  Montesi,  M.,  Doliana,  R.,  Bressan,  G.  M.,  and Colombatti, A. (2008). Emilin1 deficiency causes structural and functional defects of  lymphatic vasculature. Mol Cell Biol, 28(12):4026‐4039.  Deeb,  S.  S.,  Fajas, L., Nemoto, M., Pihlajamäki,  J., Mykkänen, L., Kuusisto,  J., Laakso, M.,  Fujimoto, W.,  and Auwerx,  J. (1998). A pro12ala substitution in ppargamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity. Nat Genet, 20(3):284‐287.  de Jongh GJ, Zeeuwen PL, Kucharekova M, Pfundt R, van der Valk PG, Blokx W, Dogan A, Hiemstra PS, van de Kerkhof PC,  Schalkwijk  J.  (2005).  High  expression  levels  of  keratinocyte  antimicrobial  proteins  in  psoriasis  compared with atopic dermatitis. J Invest Dermatol, 125(6):1163‐1173.  

Page 142: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  127

 Dixon,  J.  B.,  Raghunathan,  S.,  and  Swartz,  M.  A.  (2009).  A  tissue‐engineered  model  of  the  intestinal  lacteal  for evaluating lipid transport by lymphatics. Biotechnol Bioeng, 103(6):1224‐1235.  Draghici, S., Khatri, P., Eklund, A. C., and Szallasi, Z.  (2006). Reliability and reproducibility  issues  in dna microarray measurements. Trends Genet, 22(2):101‐109.  Du,  X.  L.,  Edelstein,  D.,  Rossetti,  L.,  Fantus,  I.  G.,  Goldberg,  H.,  Ziyadeh,  F.,  Wu,  J.,  and  Brownlee,  M.  (2000). Hyperglycemia‐induced  mitochondrial  superoxide  overproduction  activates  the  hexosamine  pathway  and  induces plasminogen activator inhibitor‐1 expression by increasing sp1 glycosylation. Proc Natl Acad Sci U S A, 97(22):12222‐12226.  Duby,  J.  J., Campbell, R. K., Setter, S. M., White,  J. R., and Rasmussen, K. A.  (2004). Diabetic neuropathy: an  intensive review. Am J Health Syst Pharm, 61(2):160‐173.  Dufour, J. H., Dziejman, M., Liu, M. T., Leung, J. H., Lane, T. E., and Luster, A. D. (2002). Ifn‐gamma‐inducible protein 10 (ip‐10;  cxcl10)‐deficient  mice  reveal  a  role  for  ip‐10  in  effector  T  cell  generation  and  trafficking.  J  Immunol, 168(7):3195‐3204.  Dumont, D.  J.,  Jussila, L., Taipale,  J., Lymboussaki, A., Mustonen, T., Pajusola, K., Breitman, M., and Alitalo, K.  (1998). Cardiovascular failure in mouse embryos deficient in VEGF receptor‐3. Science, 282(5390):946‐949.  Dunworth, W. P., Fritz‐Six, K. L., and Caron, K. M. (2008). Adrenomedullin stabilizes the lymphatic endothelial barrier in vitro and in vivo. Peptides, 29(12):2243‐2249.  Eberl, G., Marmon, S., Sunshine, M.‐J., Rennert, P. D., Choi, Y., and Littman, D. R. (2004). An essential function for the nuclear receptor rorgamma(t) in the generation of fetal lymphoid tissue inducer cells. Nat Immunol, 5(1):64‐73.  Elmasri, H., Karaaslan, C., Teper, Y., Ghelfi, E., Weng, M., Ince, T. A., Kozakewich, H., Bischoff,J., and Cataltepe, S. (2009). Fatty  acid  binding  protein  4  is  a  target  of  vegf  and  a  regulator  of  cell  proliferation  in  endothelial  cells.  FASEB  J, 23(11):3865‐3873.  Eming, S. A., Krieg, T., and Davidson, J. M. (2007). Inflammation in wound repair: molecular and cellular mechanisms. J Invest Dermatol, 127(3):514‐525.  Enholm B, Paavonen K, Ristimäki A, Kumar V, Gunji Y, Klefstrom J, Kivinen L, Laiho M, Olofsson B, Joukov V, Eriksson U,  Alitalo  K.  (1997).  Comparison  of  VEGF,  VEGF‐B,  VEGF‐C  and  Ang‐1 mRNA  regulation  by  serum,  growth  factors, oncoproteins and hypoxia. Oncogene, 14(20):2475‐2483.  Fan,  J. B., Chen, X., Halushka, M. K., Berno, A., Huang, X., Ryder, T., Lipshutz, R.  J., Lockhart, D.  J., and Chakravarti, A. (2000).  Parallel  genotyping  of  human  snps  using  generic  high‐density  oligonucleotide  tag  arrays.  Genome  Res, 10(6):853‐860.  Fernandez  AP, Miteva M,  Roberts  B,  Ricotti  C,  Rouhani  P,  Romanelli  P.  (2011).  Histopathologic  analysis  of  dermal lymphatic alterations in chronic venous insufficiency ulcers using D2‐40. J Am Acad Dermatol, 64(6):1123.e1‐12.  Ferrell,  R.  E.,  Kimak,  M.  A.,  Lawrence,  E.  C.,  and  Finegold,  D.  N.  (2008).  Candidate  gene  analysis  in  primary lymphedema. Lymphat Res Biol, 6(2):69‐76.  Förster, R., Davalos‐Misslitz, A. C., and Rot, A. (2008). Ccr7 and its ligands: balancing immunity and tolerance. Nat Rev Immunol, 8(5):362‐371.  Fox, C. S., Coady, S., Sorlie, P. D., Levy, D., Meigs, J. B., D'Agostino, R. B., Wilson, P. W. F., and Savage, P. J. (2004). Trends in cardiovascular complications of diabetes. JAMA, 292(20):2495‐2499.  Franois,  M.,  Caprini,  A.,  Hosking,  B.,  Orsenigo,  F.,  Wilhelm,  D.,  Browne,  C.,  Paavonen,  K.,  Karnezis,  T.,  Shayan,  R., Downes, M., Davidson, T., Tutt, D., Cheah, K. S. E., Stacker, S. A., Muscat, G. E. O., Achen, M. G., Dejana, E., and Koopman, P. (2008). Sox18 induces development of the lymphatic vasculature in mice. Nature, 456(7222):643‐647.  Fritsch, H. and K uhnel, W. (2001). Taschenatlas der Anatomie in 3 Bänden, Teil 2: Innere Organe. Thieme.  Fritz‐Six, K.  L., Dunworth, W. P.,  Li, M.,  and Caron, K. M.  (2008). Adrenomedullin  signaling  is  necessary  for murine lymphatic vascular development. J Clin Invest, 118(1):40‐50.  Fu, J., Gerhardt, H., McDaniel, J. M., Xia, B., Liu, X., Ivanciu, L., Ny, A., Hermans, K., Silasi‐Mansat, R., McGee, S., Nye, E., Ju, T., Ramirez, M. I., Carmeliet, P., Cummings, R. D., Lupu, F., and Xia, L. (2008). Endothelial cell o‐glycan deficiency causes blood/lymphatic misconnections and consequent fatty liver disease in mice. J Clin Invest, 118(11):3725‐3737. 

Page 143: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  128

Fujita N, Takagi S. (2012). The impact of Aggrus/podoplanin on platelet aggregation and tumor metastasis. J Biochem, [Epub ahead of print]  Furuhashi, M. and Hotamisligil, G. S. (2008). Fatty acid‐binding proteins: role  in metabolic diseases and potential as drug targets. Nat Rev Drug Discov, 7(6):489‐503.  Gal‐Yam, E. N., Egger, G., Iniguez, L., Holster, H., Einarsson, S., Zhang, X., Lin, J. C., Liang, G., Jones, P. A., and Tanay, A. (2008). Frequent switching of polycomb repressive marks and DNA hypermethylation in the pc3 prostate cancer cell line. Proc Natl Acad Sci U S A, 105(35):12979‐12984.  Gale, N. W., Prevo, R., Espinosa, J., Ferguson, D. J., Dominguez, M. G., Yancopoulos, G. D., Thurston, G., and Jackson, D. G. (2007). Normal lymphatic development and function in mice deficient for the lymphatic hyaluronan receptor lyve‐1. Mol Cell Biol, 27(2):595‐604.  Gale, N. W., Thurston, G., Hackett, S. F., Renard, R., Wang, Q., McClain, J., Martin, C., Witte, C., Witte, M. H., Jackson, D., Suri,  C.,  Campochiaro,  P.  A., Wiegand,  S.  J.,  and Yancopoulos,  G. D.  (2002).  Angiopoietin‐2  is  required  for  postnatal angiogenesis and lymphatic patterning, and only the latter role is rescued by angiopoietin‐1. Dev Cell, 3(3):411‐423.  Galiano RD, Tepper OM,  Pelo CR, Bhatt KA,  Callaghan M, Bastidas N, Bunting  S,  Steinmetz HG, Gurtner GC.  (2004). Topical vascular endothelial growth factor accelerates diabetic wound healing through increased angiogenesis and by mobilizing and recruiting bone marrow‐derived cells. Am J Pathol, 164(6):1935‐1947.  Gao,  B.,  Saba,  T.  M.,  and  Tsan,  M.‐F.  (2002).  Role  of  alpha(v)beta(3)‐integrin  in  tnf‐alphainduced  endothelial  cell migration. Am J Physiol Cell Physiol, 283(4):C1196‐C1205.  Gerli,  R.,  Solito,  R.,  Weber,  E.,  and  Aglianó,  M.  (2000).  Specific  adhesion  molecules  bind  anchoring  filaments  and endothelial cells in human skin initial lymphatics. Lymphology, 33(4):148‐157.  Gillilan, R. E., Ayers, S. D., and Noy, N. (2007). Structural basis for activation of fatty acid‐binding protein 4. J Mol Biol, 372(5):1246‐1260.  Gnepp,  D.  R.  and  Green,  F.  H.  (1979).  Scanning  electron  microscopy  of  collecting  lymphatic  vessels  and  their comparison to arteries and veins. Scan Electron Microsc, 3(3):756‐762.  Goettsch,  C.,  Goettsch,  W.,  Muller,  G.,  Seebach,  J.,  Schnittler,  H.‐J.,  and  Morawietz,  H.  (2009).  Nox4  overexpression activates  reactive  oxygen  species  and  p38  MAPK  in  human  endothelial  cells.  Biochem  Biophys  Res  Commun, 380(2):355‐360.  Goldberg, R. B. (2003). Cardiovascular disease in patients who have diabetes. Cardiol Clin, 21(3):399‐413, vii.  Goova MT, Li J, Kislinger T, Qu W, Lu Y, Bucciarelli LG, Nowygrod S, Wolf BM, Caliste X, Yan SF, Stern DM, Schmidt AM. (2001). Blockade of receptor for advanced glycation end‐products restores effective wound healing in diabetic mice. Am J Pathol, 159(2):513‐525.  Goren, I., Müller, E., Schiefelbein, D., Christen, U., Pfeilschifter, J., Mühl, H., and Frank, S. (2007). Systemic anti‐tnfalpha treatment  restores diabetes‐impaired  skin  repair  in  ob/ob mice by  inactivation of macrophages.  J  Invest Dermatol, 127(9):2259‐2267.  Götte, M., Joussen, A. M., Klein, C., Andre, P., Wagner, D. D., Hinkes, M. T., Kirchhof, B., Adamis, A. P., and Bernfield, M. (2002). Role of syndecan‐1 in leukocyte‐endothelial interactions in the ocular vasculature. Invest Ophthalmol Vis Sci, 43(4):1135‐1141.  Grainger, R. M.  (1992). Embryonic  lens  induction:  shedding  light on vertebrate  tissue determination. Trends Genet, 8(10):349‐355.  Grice EA, Snitkin ES, Yockey LJ, Bermudez DM; NISC Comparative Sequencing Program, Liechty KW, Segre JA. (2010). Longitudinal shift in diabetic wound microbiota correlates with prolonged skin defense response. Proc Natl Acad Sci USA, 107(33):14799‐14804.  Gröger, M., Niederleithner, H., Kerjaschki, D., and Petzelbauer, P. (2007). A previously unknown dermal blood vessel phenotype in skin inflammation. J Invest Dermatol, 127(12):2893‐2900.   Group, U. (1998a). Effect of intensive blood‐glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). UK prospective diabetes study (UKPDS) group. Lancet, 352(9131):854‐865.  

Page 144: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  129

Group,  U.  (1998b).  Intensive  blood‐glucose  control  with  sulphonylureas  or  insulin  compared  with  conventional treatment  and  risk  of  complications  in  patients  with  type  2  diabetes  (UKPDS  33).  UK  prospective  diabetes  study (UKPDS) group. Lancet, 352(9131):837‐853.  Group,  U.  P.  D.  S.  (1995).  United  Kingdom  prospective  diabetes  study  (UKPDS).  13:  Relative  efficacy  of  randomly allocated diet, sulphonylurea, insulin, or metformin in patients with newly diagnosed non‐insulin dependent diabetes followed for three years. BMJ, 310(6972):83‐88.  Guttman‐Yassky  E,  Lowes MA,  Fuentes‐Duculan  J,  Zaba  LC,  Cardinale  I,  Nograles  KE,  Khatcherian  A,  Novitskaya  I, Carucci JA, Bergman R, Krueger JG. (2008). Low expression of the IL‐23/Th17 pathway in atopic dermatitis compared to psoriasis. J Immunol, 181(10):7420‐7427.  Haddy, N., Sass, C., Droesch, S., Zaiou, M., Siest, G., Ponthieux, A., Lambert, D., and Visvikis, S. (2003). Il‐6, TNF‐alpha and atherosclerosis  risk  indicators  in a healthy  family population:  the stanislas cohort. Atherosclerosis, 170(2):277‐283.  Hämmerling B, Grund C, Boda‐Heggemann  J, Moll R, Franke WW.  (2006). The  complexus adhaerens of mammalian lymphatic endothelia revisited: a junction even more complex than hitherto thought. Cell Tissue Res, 324(1):55‐67.  Hagiwara, A., Takahashi, T., and Oku, N. (1989). Cancer chemotherapy administered by activated carbon particles and liposomes. Crit Rev Oncol Hematol, 9(4):319‐350.  Halgren, R. G.,  Fielden, M. R.,  Fong, C.  J.,  and Zacharewski, T. R.  (2001). Assessment of  clone  identity  and  sequence fidelity for 1189 image cdna clones. Nucleic Acids Res, 29(2):582‐588.   Halin,  C.  and Detmar, M.  (2008).  Chapter  1.  inflammation,  angiogenesis,  and  lymphangiogenesis. Methods Enzymol, 445:1‐25.  Hansen, L., Gaster, M., Oakeley, E. J., Brusgaard, K., Nielsen, E.‐M. D., Beck‐Nielsen, H., Pedersen, O., and Hemmings, B. A. (2004).  Expression  profiling  of  insulin  action  in  human myotubes:  induction  of  inflammatory  and  pro‐angiogenic pathways in relationship with glycogen synthesis and type 2 diabetes. Biochem Biophys Res Commun, 323(2):685‐695.  Hara‐Chikuma,  M.  and  Verkman,  A.  S.  (2008).  Aquaporin‐3  facilitates  epidermal  cell  migration  and  proliferation during wound healing. J Mol Med, 86(2):221‐231.  Harvey, N. L. (2008). The link between lymphatic function and adipose biology. Ann N Y Acad Sci, 1131:82‐88.  Harvey, N. L., Srinivasan, R. S., Dillard, M. E., Johnson, N. C., Witte, M. H., Boyd, K., Sleeman, M. W., and Oliver, G. (2005). Lymphatic vascular defects promoted by prox1 haploinsufficiency cause adult‐onset obesity. Nat Genet, 37(10):1072‐1081.  He, Y., Kozaki, K.‐I., Karpanen, T., Koshikawa, K., Yla‐Herttuala, S., Takahashi, T., and Alitalo, K. (2002). Suppression of tumor  lymphangiogenesis  and  lymph  node  metastasis  by  blocking  vascular  endothelial  growth  factor  receptor  3 signaling. J Natl Cancer Inst, 94(11):819‐825.  Heller, E. A., Liu, E., Tager, A. M., Yuan, Q., Lin, A. Y., Ahluwalia, N., Jones, K., Koehn, S. L., Lok, V. M., Aikawa, E., Moore, K. J.,  Luster,  A.  D.,  and  Gerszten,  R.  E.  (2006).  Chemokine  CXCL10  promotes  atherogenesis  by  modulating  the  local balance of effector and regulatory t cells. Circulation, 113(19):2301‐2312.  Hirakawa, S., Hong, Y.‐K., Harvey, N., Schacht, V., Matsuda, K., Libermann, T., and Detmar, M. (2003). Identification of vascular lineage‐specific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am J Pathol, 162(2):575‐586.  Hong, Y.‐K., Harvey, N., Noh, Y.‐H., Schacht, V., Hirakawa, S., Detmar, M., and Oliver, G. (2002). Prox1 is a master control gene in the program specifying lymphatic endothelial cell fate. Dev Dyn, 225(3):351‐357.  Horra  A,  Salazar  J,  Ferré  R,  Vallvé  JC,  Guardiola M,  Rosales  R, Masana  L,  Ribalta  J.  (2009)  Prox‐1  and  FOXC2  gene expression  in  adipose  tissue:  A  potential  contributory  role  of  the  lymphatic  system  to  familial  combined hyperlipidaemia. Atherosclerosis, 206(2):343‐345.   Hotamisligil,  G.  S.  (1999).  Mechanisms  of  tnf‐alpha‐induced  insulin  resistance.  Exp  Clin  Endocrinol  Diabetes, 107(2):119‐125.  Hotamisligil, G. S. (2000). Molecular mechanisms of insulin resistance and the role of the adipocyte. Int J Obes Relat Metab Disord, 24 Suppl 4:S23‐S27.  Hotamisligil, G. S. (2006). Inflammation and metabolic disorders. Nature, 444(7121):860‐867. 

Page 145: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  130

 Hotamisligil,  G.  S.  (2010).  Endoplasmic  reticulum  stress  and  the  inflammatory  basis  of  metabolic  disease.  Cell, 140(6):900‐917.  Hotamisligil,  G.  S.,  Arner,  P.,  Caro,  J.  F.,  Atkinson,  R.  L.,  and  Spiegelman,  B.  M.  (1995).  Increased  adipose  tissue expression of tumor necrosis factor‐alpha in human obesity and insulin resistance. J Clin Invest, 95(5):2409‐2415.  Hotamisligil, G. S., Johnson, R. S., Distel, R. J., Ellis, R., Papaioannou, V. E., and Spiegelman, B. M. (1996). Uncoupling of obesity from insulin resistance through a targeted mutation in aP2, the adipocyte fatty acid binding protein. Science, 274(5291):1377‐1379.  Hotamisligil,  G.  S.,  Murray,  D.  L.,  Choy,  L.  N.,  and  Spiegelman,  B.  M.  (1994).  Tumor  necrosis  factor  alpha  inhibits signaling from the insulin receptor. Proc Natl Acad Sci U S A, 91(11):4854‐4858.  Hotta, K., Funahashi, T., Arita, Y., Takahashi, M., Matsuda, M., Okamoto, Y., Iwahashi, H., Kuriyama, H., Ouchi, N., Maeda, K., Nishida, M., Kihara, S., Sakai, N., Nakajima, T., Hasegawa, K., Muraguchi, M., Ohmoto, Y., Nakamura, T., Yamashita, S., Hanafusa, T.,  and Matsuzawa, Y.  (2000). Plasma  concentrations of  a novel,  adipose‐specific  protein,  adiponectin,  in type 2 diabetic patients. Arterioscler Thromb Vasc Biol, 20(6):1595‐1599.  Hu, E., Liang, P., and Spiegelman, B. M. (1996). Adipoq is a novel adipose‐specific gene dysregulated in obesity. J Biol Chem, 271(18):10697‐10703.  Huang, D. W., Sherman, B. T., and Lempicki, R. A. (2009). Systematic and integrative analysis of large gene lists using david bioinformatics resources. Nat Protoc, 4(1):44‐57.  Huang,  X.  Z.,  Wu,  J.  F.,  Ferrando,  R.,  Lee,  J.  H.,  Wang,  Y.  L.,  Farese,  R.  V.,  and  Sheppard,  D.  (2000).  Fatal  bilateral chylothorax in mice lacking the integrin alpha9beta1. Mol Cell Biol, 20(14):5208‐5215.  Huggenberger R, Siddiqui SS, Brander D, Ullmann S, Zimmermann K, Antsiferova M, Werner S, Alitalo K, Detmar M. (2011). An important role of lymphatic vessel activation in limiting acute inflammation. Blood, 117(17):4667‐4678.   Hughes,  J. M., Kuiper, E.  J., Klaassen,  I., Canning, P.,  Stitt, A. W., Bezu,  J. V., Schalkwijk, C. G., Noorden, C.  J. F. V.,  and Schlingemann,  R.  O.  (2007).  Advanced  glycation  end  products  cause  increased  ccn  family  and  extracellular matrix gene expression in the diabetic rodent retina. Diabetologia, 50(5):1089‐1098.  Hundal, R. S. and Inzucchi, S. E. (2003). Metformin: new understandings, new uses. Drugs, 63(18):1879‐1894.  Huntington, G. and McClure, C.  (1910). The anatomy and development of  the  jugular  lymph sac  in  the domestic cat (felis domestica). Am J Anat, 10:177‐311.  Ichise, H., Ichise, T., Ohtani, O., and Yoshida, N. (2009). Phospholipase cgamma2 is necessary for separation of blood and lymphatic vasculature in mice. Development, 136(2):191‐195.  Irigoyen  M,  Ansó  E,  Martínez  E,  Garayoa  M,  Martínez‐Irujo  JJ,  Rouzaut  A.  (2007)  Hypoxia  alters  the  adhesive properties of lymphatic endothelial cells. A transcriptional and functional study. Biochim Biophys Acta, 1773(6):880‐890.   Irjala, H.,  Johansson,  E.  L.,  Grenman,  R.,  Alanen,  K.,  Salmi, M.,  and  Jalkanen,  S.  (2001). Mannose  receptor  is  a  novel ligand for l‐selectin and mediates lymphocyte binding to lymphatic endothelium. J Exp Med, 194(8):1033‐1042.  Ishibashi, K., Hara, S., and Kondo, S. (2009). Aquaporin water channels in mammals. Clin Exp Nephrol, 13(2):107‐117.  Itani, S. I., Ruderman, N. B., Schmieder, F., and Boden, G. (2002). Lipid‐induced insulin resistance in human muscle is associated with changes in diacylglycerol, protein kinase c, and I kappa b alpha. Diabetes, 51(7):2005‐2011.  Jackson,  A.  J.  (1981).  Intramuscular  absorption  and  regional  lymphatic  uptake  of  liposomeentrapped  inulin. Drug Metab Dispos, 9(6):535‐540.  Jackson,  D.  G.  (2009).  Immunological  functions  of  hyaluronan  and  its  receptors  in  the  lymphatics.  Immunol  Rev, 230(1):216‐231.  Jelinek, J. E. (1994). Cutaneous manifestations of diabetes mellitus. Int J Dermatol, 33(9):605‐617.   Jeltsch, M., Kaipainen, A., Joukov, V., Meng, X., Lakso, M., Rauvala, H., Swartz, M., Fukumura, D., Jain, R. K., and Alitalo, K. (1997). Hyperplasia of lymphatic vessels in VEGF‐C transgenic mice. Science, 276(5317):1423‐1425.  

Page 146: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  131

Jeltsch, M., Tammela, T., Alitalo, K., and Wilting, J. (2003). Genesis and pathogenesis of  lymphatic vessels. Cell Tissue Res, 314(1):69‐84.  Ji,  R.‐C.,  Kurihara, K.,  and Kato,  S.  (2006).  Lymphatic  vascular  endothelial  hyaluronan  receptor  (lyve)‐1‐  and  ccl21‐positive lymphatic compartments in the diabetic thymus. Anat Sci Int, 81(4):201‐209.  Johnson,  L.  A.,  Clasper,  S.,  Holt,  A.  P.,  Lalor,  P.  F.,  Baban,  D.,  and  Jackson,  D.  G.  (2006).  An  inflammation‐induced mechanism for leukocyte transmigration across lymphatic vessel endothelium. J Exp Med, 203(12):2763‐2777.  Johnson, L. A., Prevo, R., Clasper, S., and Jackson, D. G.  (2007).  Inflammation‐induced uptake and degradation of  the lymphatic endothelial hyaluronan receptor LYVE‐1. J Biol Chem, 282(46):33671‐33680.  Johnson, N. C., Dillard, M. E., Baluk, P., McDonald, D. M., Harvey, N. L.,  Frase,  S.  L.,  and Oliver, G.  (2008). Lymphatic endothelial cell identity is reversible and its maintenance requires PROX1 activity. Genes Dev, 22(23):3282‐3291. Joshi, N., Caputo, G. M., Weitekamp, M. R., and Karchmer, A. W. (1999). Infections in patients with diabetes mellitus. N Engl J Med, 341(25):1906‐1912.  Jurisic G, Maby‐El Hajjami H, Karaman S, Ochsenbein AM, Alitalo A, Siddiqui SS, Ochoa Pereira C, Petrova TV, Detmar M.  (2012).  An  unexpected  role  of  semaphorin3a‐neuropilin‐1  signaling  in  lymphatic  vessel  maturation  and  valve formation. Circ Res, 111(4):426‐436.     Jurisic, G. and Detmar, M. (2009). Lymphatic endothelium in health and disease. Cell Tissue Res, 335(1):97‐108.  Jussila, L. and Alitalo, K. (2002). Vascular growth factors and lymphangiogenesis. Physiol Rev, 82(3):673‐700.  Kainulainen,  V.,  Nelimarkka,  L.,  Järveläinen,  H.,  Laato,  M.,  Jalkanen,  M.,  and  Elenius,  K.  (1996).  Suppression  of syndecan‐1 expression in endothelial cells by tumor necrosis factor‐alpha. J Biol Chem, 271(31):18759‐18766.  Kaipainen, A., Korhonen,  J., Mustonen, T., van Hinsbergh, V. W., Fang, G. H., Dumont, D., Breitman, M., and Alitalo, K. (1995).  Expression  of  the  fms‐like  tyrosine  kinase  4  gene  becomes  restricted  to  lymphatic  endothelium  during development. Proc Natl Acad Sci U S A, 92(8):3566‐3570.  Kang, S., Lee, S.‐P., Kim, K. E., Kim, H.‐Z., Mémet, S., and Koh, G. Y. (2009). Toll‐like receptor 4 in lymphatic endothelial cells  contributes  to  lps‐induced  lymphangiogenesis  by  chemotactic  recruitment  of  macrophages.  Blood, 113(11):2605‐2613.  Kapoor S. (2012). Podoplanin expression and its emerging role as a prognostic indicator in systemic tumors besides head and neck malignancies. J Surg Oncol, [Epub ahead of print].  Kaprio,  J.,  Tuomilehto,  J.,  Koskenvuo, M.,  Romanov,  K.,  Reunanen,  A.,  Eriksson,  J.,  Stengård,  J.,  and  Kesäniemi,  Y.  A. (1992).  Concordance  for  type  1  (insulin‐dependent)  and  type  2  (non  insulin‐dependent)  diabetes  mellitus  in  a population‐based cohort of twins in Finland. Diabetologia, 35(11):1060‐1067.  Karikoski,  M.,  Irjala,  H.,  Maksimow,  M.,  Miiluniemi,  M.,  Granfors,  K.,  Hernesniemi,  S.,  Elima,  K.,  Moldenhauer,  G., Schledzewski,  K.,  Kzhyshkowska,  J.,  Goerdt,  S.,  Salmi,  M.,  and  Jalkanen,  S.  (2009).  Clever‐1/stabilin‐1  regulates lymphocyte  migration  within  lymphatics  and  leukocyte  entrance  to  sites  of  inflammation.  Eur  J  Immunol, 39(12):3477‐3487.  Karkkainen, M.  J.,  Haiko,  P.,  Sainio,  K.,  Partanen,  J.,  Taipale,  J.,  Petrova,  T.  V.,  Jeltsch, M.,  Jackson,  D.  G.,  Talikka, M., Rauvala, H., Betsholtz, C., and Alitalo, K. (2004). Vascular endothelial growth factor c is required for sprouting of the first lymphatic vessels from embryonic veins. Nat Immunol, 5(1):74‐80.  Karkkainen, M. J., Saaristo, A., Jussila, L., Karila, K. A., Lawrence, E. C., Pajusola, K., Bueler, H., Eichmann, A., Kauppinen, R., Kettunen, M. I., Yla‐Herttuala, S., Finegold, D. N., Ferrell, R. E., and Alitalo, K. (2001). A model for gene therapy of human hereditary lymphedema. Proc Natl Acad Sci U S A, 98(22):12677‐12682.  Kaufmann,  A.,  Molnár,  B.,  Craciun,  C.,  and  Itcus,  A.  (1980).  Diabetic  lymphangiopathy.  An  optical  and  electron microscopic study. Lymphology, 13(4):202‐206.  Kerjaschki, D. (2006). Lymphatic neoangiogenesis in renal transplants: a driving force of chronic rejection? J Nephrol, 19(4):403‐406.  Kerjaschki D. (2005). The crucial role of macrophages in lymphangiogenesis. J Clin Invest, 115(9):2316‐2319.  Kerjaschki D, Regele HM, Moosberger I, Nagy‐Bojarski K, Watschinger B, Soleiman A, Birner P, Krieger S, Hovorka A, Silberhumer  G,  Laakkonen  P,  Petrova  T,  Langer  B,  Raab  I.  (2004).  Lymphatic  neoangiogenesis  in  human  kidney transplants is associated with immunologically active lymphocytic infiltrates. J Am Soc Nephrol, 15(3):603‐612. 

Page 147: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  132

Khanna, S., Biswas, S., Shang, Y., Collard, E., Azad, A., Kauh, C., Bhasker, V., Gordillo, G. M., Sen, C. K., and Roy, S. (2010). Macrophage dysfunction impairs resolution of inflammation in the wounds of diabetic mice. PLoS One, 5(3):e9539.  Kholová  I,  Dragneva  G,  Cermáková  P,  Laidinen  S,  Kaskenpää  N,  Hazes  T,  Cermáková  E,  Steiner  I,  Ylä‐Herttuala  S. (2011). Lymphatic vasculature is increased in heart valves, ischaemic and inflamed hearts and in cholesterol‐rich and calcified atherosclerotic lesions. Eur J Clin Invest,41(5):487‐497.   Kikkawa, R., Koya, D., and Haneda, M. (2003). Progression of diabetic nephropathy. Am J Kidney Dis, 41(3 Suppl 1):S19‐S21.  Kim  KE,  Koh  YJ,  Jeon  BH,  Jang  C,  Han  J,  Kataru  RP,  Schwendener  RA,  Kim  JM,  Koh  GY.  (2009).  Role  of  CD11b+ macrophages  in  intraperitoneal  lipopolysaccharide‐induced  aberrant  lymphangiogenesis  and  lymphatic  function  in the diaphragm. Am J Pathol,175(4):1733‐1745.   Kirpichnikov, D., McFarlane, S. I., and Sowers, J. R. (2002). Metformin: an update. Ann Intern Med, 137(1):25‐33.  Kivelä, R., Silvennoinen, M., Lehti, M., Kainulainen, H., and Vihko, V. (2007). Effects of acute exercise, exercise training, and  diabetes  on  the  expression  of  lymphangiogenic  growth  factors  and  lymphatic  vessels  in  skeletal muscle. Am  J Physiol Heart Circ Physiol, 293(4):H2573‐H2579.  Koch S, Tugues S, Li X, Gualandi L, Claesson‐Welsh L. (2011). Signal transduction by vascular endothelial growth factor receptors. Biochem J, 437(2):169‐183.  Krebs, D. L. and Hilton, D. J. (2003). A new role for SOCS in insulin action. Suppressor of cytokine signaling. Sci STKE, (169):PE6.  Kriederman,  B. M., Myloyde,  T.  L., Witte, M. H.,  Dagenais,  S.  L., Witte,  C.  L.,  Rennels, M.,  Bernas, M.  J.,  Lynch, M.  T., Erickson, R. P., Caulder, M. S., Miura, N.,  Jackson, D., Brooks, B. P., and Glover, T. W.  (2003). Foxc2 haploinsufficient mice are a model for human autosomal dominant lymphedema‐distichiasis syndrome. Hum Mol Genet, 12(10):1179‐1185.  Kriehuber, E., Breiteneder‐Geleff, S., Groeger, M., Soleiman, A., Schoppmann, S. F., Stingl, G., Kerjaschki, D., and Maurer, D.  (2001).  Isolation  and  characterization  of  dermal  lymphatic  and  blood  endothelial  cells  reveal  stable  and functionally specialized cell lineages. J Exp Med, 194(6):797‐808.  Kukk, E., Lymboussaki, A., Taira, S., Kaipainen, A., Jeltsch, M., Joukov, V., and Alitalo, K. (1996). Vegf‐c receptor binding and  pattern  of  expression  with  Vegfr‐3  suggests  a  role  in  lymphatic  vascular  development.  Development, 122(12):3829‐3837.  Kulesh, D. A., Clive, D. R., Zarlenga, D. S., and Greene, J. J. (1987). Identification of interferon modulated proliferation‐related cDNA sequences. Proc Natl Acad Sci U S A, 84(23):8453‐8457.  Kume  T.  (2010).  Specification  of  arterial,  venous,  and  lymphatic  endothelial  cells  during  embryonic  development. Histol Histopathol, 25 (5): 637 – 646.  Labanaris AP, Polykandriotis E, Horch RE. (2009). The effect of vacuum‐assisted closure on lymph vessels in chronic wounds. J Plast Reconstr Aesthet Surg, 62(8):1068‐1075.   Lander, E. S. (1999). Array of hope. Nat Genet, 21(1 Suppl):3‐4.  Lapinski  PE,  Kwon  S,  Lubeck  BA,  Wilkinson  JE,  Srinivasan  RS,  Sevick‐Muraca  E,  King  PD.  RASA1  maintains  the lymphatic vasculature in a quiescent functional state in mice. J Clin Invest, 122(2):733‐747.  Leak, L. V. (1970). Electron microscopic observations on lymphatic capillaries and the structural components of the connective tissue‐lymph interface. Microvasc Res, 2(4):361‐391.  Leak, L. V. and Burke, J. F. (1966). Fine structure of the lymphatic capillary and the adjoining connective tissue area. Am J Anat, 118(3):785‐809.  Leak, L. V. and Burke, J. F. (1968). Ultrastructural studies on the lymphatic anchoring filaments. J Cell Biol, 36(1):129‐149.  Leu, S.‐J., Lam, S. C.‐T., and Lau, L. F.  (2002). Pro‐angiogenic activities of CYR61 (CCN1) mediated  through  integrins alphaVbeta3 and alpha6beta1 in human umbilical vein endothelial cells. J Biol Chem, 277(48):46248‐46255.  Lim,  H.  Y.,  Rutkowski,  J.  M.,  Helft,  J.,  Reddy,  S.  T.,  Swartz,  M.  A.,  Randolph,  G.  J.,  and  Angeli,  V.  (2009). Hypercholesterolemic mice exhibit lymphatic vessel dysfunction and degeneration. Am J Pathol, 175(3):1328‐1337. 

Page 148: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  133

Liu FY, Tang XC, Deng M, Chen P, Ji W, Zhang X, Gong L, Woodward Z, Liu J, Zhang L, Sun S, Liu JP, Wu K, Wu MX, Liu XL, Yu MB, Liu Y, Li DW. (2012). The Tumor Suppressor p53 Regulates c‐Maf and Prox‐1 to Control Lens Differentiation. Curr Mol Med, 12(8):917‐928.  Luissint,  A.‐C.,  Lutz,  P.  G.,  Calderwood,  D.  A.,  Couraud,  P.‐O.,  and  Bourdoulous,  S.  (2008).  Jam‐l‐mediated  leukocyte adhesion to endothelial cells is regulated in cis by alpha4beta1 integrin activation. J Cell Biol, 183(6):1159‐1173.  Luong, M. X., Tam, J., Lin, Q., Hagendoorn, J., Moore, K. J., Padera, T. P., Seed, B., Fukumura, D., Kucherlapati, R., and Jain, R. K. (2009). Lack of lymphatic vessel phenotype in lyve‐1/cd44 double knockout mice. J Cell Physiol, 219(2):430‐437.  Lyman, G. H., Giuliano, A. E., Somerfield, M. R., Benson, A. B., Bodurka, D. C., Burstein, H. J., Cochran, A. J., Cody, H. S., Edge, S. B., Galper, S., Hayman, J. A., Kim, T. Y., Perkins, C. L., Podoloff, D. A., Sivasubramaniam, V. H., Turner, R. R., Wahl, R., Weaver, D. L., Wolff, A. C., Winer, E. P., and of Clinical Oncology, A. S. (2005). American society of clinical oncology guideline  recommendations  for  sentinel  lymph node biopsy  in early‐stage breast  cancer.  J Clin Oncol,  23(30):7703‐7720.  Lymboussaki, A.,  Partanen,  T. A., Olofsson, B.,  Thomas‐Crusells,  J.,  Fletcher,  C. D.,  de Waal, R. M.,  Kaipainen, A.,  and Alitalo, K. (1998). Expression of the vascular endothelial growth factor c receptor VEGFR‐3 in lymphatic endothelium of the skin and in vascular tumors. Am J Pathol, 153(2):395‐403.  Maby‐El  Hajjami  H  and  Petrova  TV.  (2008).  Developmental  and  pathological  lymphangiogenesis:  from  models  to human disease. Histochem Cell Biol, 130(6):1063‐1078.   Machnik, A., Neuhofer, W.,  Jantsch,  J., Dahlmann, A., Tammela, T., Machura, K., Park,  J.‐K., Beck, F.‐X., M uller, D. N., Derer, W., Goss, J., Ziomber, A., Dietsch, P., Wagner, H., van Rooijen, N., Kurtz, A., Hilgers, K. F., Alitalo, K., Eckardt, K.‐U., Luft, F. C., Kerjaschki, D., and Titze, J. (2009). Macrophages regulate salt‐dependent volume and blood pressure by a vascular endothelial growth factor‐c‐dependent buffering mechanism. Nat Med, 15(5):545‐552.  Maerki, C., Meuter, S., Liebi, M., M uhlemann, K., Frederick, M. J., Yawalkar, N., Moser, B., and Wolf, M. (2009). Potent and  broad‐spectrum  antimicrobial  activity  of  CXCL14  suggests  an  immediate  role  in  skin  infections.  J  Immunol, 182(1):507‐514.  Mäkinen,  T.,  Adams,  R.  H.,  Bailey,  J.,  Lu,  Q.,  Ziemiecki,  A.,  Alitalo,  K.,  Klein,  R.,  and  Wilkinson,  G.  A.  (2005).  Pdz interaction site in ephrinb2 is required for the remodeling of lymphatic vasculature. Genes Dev, 19(3):397‐410.  Mahdavian Delavary B, van der Veer WM, van Egmond M, Niessen FB, Beelen RH. (2011). Macrophages in skin injury and repair. Immunobiology,216(7):753‐762.  Makowski, L. and Hotamisligil, G. S. (2004). Fatty acid binding proteins‐the evolutionary crossroads of inflammatory and metabolic responses. J Nutr, 134(9):2464S‐2468S.  Malhotra D, Fletcher AL, Astarita J, Lukacs‐Kornek V, Tayalia P, Gonzalez SF, Elpek KG, Chang SK, Knoblich K, Hemler ME,  Brenner  MB,  Carroll  MC,  Mooney  DJ,  Turley  SJ;  Immunological  Genome  Project  Consortium.  (2012). Transcriptional  profiling  of  stroma  from  inflamed  and  resting  lymph  nodes  defines  immunological  hallmarks. Nat Immunol,13(5):499‐510.  Malhotra, A., Igo, R. P., Thameem, F., Kao, W. H. L., Abboud, H. E., Adler, S. G., Arar, N. H., Bowden, D. W., Duggirala, R., Freedman, B. I., Goddard, K. A. B., Ipp, E., Iyengar, S. K., Kimmel, P. L., Knowler, W. C., Kohn, O., Leehey, D., Meoni, L. A., Nelson, R. G., Nicholas, S. B., Parekh, R. S., Rich, S. S., Chen, Y.‐D. I., Saad, M. F., Scavini, M., Schelling, J. R., Sedor, J. R., Shah, V. O., Taylor, K. D., Thornley‐Brown, D., Zager, P. G., Horvath, A., Hanson, R. L., of Nephropathy, F. I., and Group, D. R. (2009). Genome‐wide linkage scans  for type 2 diabetes mellitus  in four ethnically diverse populations‐significant evidence  for  linkage on chromosome 4q  in african americans:  the  family  investigation of nephropathy and diabetes research group. Diabetes Metab Res Rev, 25(8):740‐747.  Maresh, J. G. and Shohet, R. V. (2008). In vivo endothelial gene regulation in diabetes. Cardiovasc Diabetol, 7:8.  Martin P, Leibovich SJ.  (2005).  Inflammatory cells during wound repair:  the good,  the bad and the ugly. Trends Cell Biol, 15(11):599‐607.  Martín‐Villar,  E.,  Scholl,  F.  G.,  Gamallo,  C.,  Yurrita,  M.  M.,  noz  Guerra,  M.  M.,  Cruces,  J.,  and  Quintanilla,  M.  (2005). Characterization  of  human  pa2.26  antigen  (t1alpha‐2,  podoplanin),  a  small  membrane  mucin  induced  in  oral squamous cell carcinomas. Int J Cancer, 113(6):899‐910.  Marttila‐Ichihara, F., Turja, R., Miiluniemi, M., Karikoski, M., Maksimow, M., Niemelä, J., Martinez‐Pomares, L., Salmi, M., and Jalkanen, S. (2008). Macrophage mannose receptor on lymphatics controls cell trafficking. Blood, 112(1):64‐72.  

Page 149: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  134

Maruyama, K., Asai, J., Ii, M., Thorne, T., Losordo, D. W., and D'Amore, P. A. (2007). Decreased macrophage number and activation lead to reduced lymphatic vessel formation and contribute to impaired diabetic wound healing. Am J Pathol, 170(4):1178‐1191.  Maruyama, K.,  Ii, M.,  Cursiefen,  C.,  Jackson, D. G.,  Keino, H.,  Tomita, M.,  Rooijen, N.  V.,  Takenaka, H., D'Amore,  P. A., Stein‐Streilein, J., Losordo, D. W., and Streilein, J. W. (2005). Inflammation‐induced lymphangiogenesis in the cornea arises from cd11b‐positive macrophages. J Clin Invest, 115(9):2363‐2372.  Matsui,  K.,  Breitender‐Geleff,  S.,  Soleiman,  A.,  Kowalski,  H.,  and  Kerjaschki,  D.  (1999).  Podoplanin,  a  novel  43‐kda membrane protein, controls the shape of podocytes. Nephrol Dial Transplant, 14 Suppl 1:9‐11.  Matsui,  K.,  Breiteneder‐Geleff,  S.,  and  Kerjaschki,  D.  (1998).  Epitope‐specific  antibodies  to  the  43‐kD  glomerular membrane protein podoplanin cause proteinuria and rapid attening of podocytes. J Am Soc Nephrol, 9(11):2013‐2026.  Mattacks,  C.  A.,  Sadler,  D.,  and Pond,  C. M.  (2003).  The  cellular  structure  and  lipid/protein  composition  of  adipose tissue surrounding chronically stimulated lymph nodes in rats. J Anat, 202(6):551‐561.  Mebius, R. E. (2003). Organogenesis of lymphoid tissues. Nat Rev Immunol, 3(4):292‐303.  Medoff, B. D., Sauty, A., Tager, A. M., Maclean, J. A., Smith, R. N., Mathew, A., Dufour, J. H., and Luster, A. D. (2002). Ifn‐gamma‐inducible  protein  10  (cxcl10)  contributes  to  airway  hyperreactivity  and  airway  inflammation  in  a  mouse model of asthma. J Immunol, 168(10):5278‐5286.  Meigs,  J.  B.,  Hu,  F.  B.,  Rifai,  N.,  and Manson,  J.  E.  (2004).  Biomarkers  of  endothelial  dysfunction  and  risk  of  type  2 diabetes mellitus. JAMA, 291(16):1978‐1986.  Meriden,  T.  (2004).  Progress  with  thiazolidinediones  in  the  management  of  type  2  diabetes  mellitus.  Clin  Ther, 26(2):177‐190.  Miklos, G. L. G. and Maleszka, R. (2004). Microarray reality checks in the context of a complex disease. Nat Biotechnol, 22(5):615‐621.  Miyasaka, M.  and Tanaka, T.  (2004).  Lymphocyte  trafficking  across high  endothelial  venules:  dogmas and enigmas. Nat Rev Immunol, 4(5):360‐370.  Moczulski, D. K., Fojcik, H., Wielgorecki, A., Trautsolt, W., Gawlik, B., Kosiorz‐Gorczynska, S., Oczko‐Wojciechowska, M., Wiench,  M.,  Strojek,  K.,  Zukowska‐Szczechowska,  E.,  and  Grzeszczak,  W.  (2007).  Expression  pattern  of  genes  in peripheral blood mononuclear cells in diabetic nephropathy. Diabet Med, 24(3):266‐271.  Morigi, M., Angioletti,  S.,  Imberti, B., Donadelli, R., Micheletti, G.,  Figliuzzi, M., Remuzzi, A.,  Zoja,  C.,  and Remuzzi, G. (1998). Leukocyte‐endothelial interaction is augmented by high glucose concentrations and hyperglycemia in a nf‐kb‐dependent fashion. J Clin Invest, 101(9):1905‐1915.  Moriguchi, P., Sannomiya, P., Lara, P. F., Oliveira‐Filho, R. M., Greco, K. V., and Sudo‐Hayashi, L. S. (2005). Lymphatic system changes in diabetes mellitus: role of insulin and hyperglycemia. Diabetes Metab Res Rev, 21(2):150‐157.  Mortimer, P. S., Simmonds, R., Rezvani, M., Robbins, M., Hopewell,  J. W., and Ryan, T.  J.  (1990). The measurement of skin lymph ow by isotope clearance‐reliability, reproducibility, injection dynamics, and the effect of massage. J Invest Dermatol, 95(6):677‐682.  Muller, L. M. A. J., Gorter, K. J., Hak, E., Goudzwaard, W. L., Schellevis, F. G., Hoepelman, A. I. M., and Rutten, G. E. H. M. (2005).  Increased  risk  of  common  infections  in  patients  with  type  1  and  type  2  diabetes mellitus.  Clin  Infect Dis, 41(3):281‐288.  Nadler,  S.  T.,  Stoehr,  J.  P.,  Schueler,  K.  L.,  Tanimoto,  G.,  Yandell,  B.  S.,  and  Attie,  A.  D.  (2000).  The  expression  of adipogenic genes is decreased in obesity and diabetes mellitus. Proc Natl Acad Sci U S A, 97(21):11371‐11376.  Nathanson, S. D. (2003). Insights into the mechanisms of lymph node metastasis. Cancer, 98(2):413‐423.  Navarro, A., Alonso, A., Garrido, P., González, C., Rey, C. G. D., nez, C. O.,  and Tolivia,  J.  (2010).  Increase  in placental apolipoprotein D as an adaptation to human gestational diabetes. Placenta, 31(1):25‐31.  Nesto, R. W., Bell, D., Bonow, R. O., Fonseca, V., Grundy, S. M., Horton, E. S., Winter, M. L., Porte, D., Semenkovich, C. F., Smith,  S.,  Young,  L.  H.,  and  Kahn,  R.  (2004).  Thiazolidinedione  use,  fluid  retention,  and  congestive  heart  failure:  a consensus  statement  from  the  american  heart  association  and  american  diabetes  association.  Diabetes  Care, 27(1):256‐263.  

Page 150: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  135

Newman, B., Selby, J. V., King, M. C., Slemenda, C., Fabsitz, R., and Friedman, G. D. (1987). Concordance for type 2 (non‐insulin‐dependent) diabetes mellitus in male twins. Diabetologia, 30(10):763‐768.  Nguyen PD, Tutela JP, Thanik VD, Knobel D, Allen RJ Jr, Chang CC, Levine JP, Warren SM, Saadeh PB. (2010). Improved diabetic wound healing through topical silencing of p53 is associated with augmented vasculogenic mediators. Wound Repair Regen, 18(6):553‐559.   Nightingale,  T.  D.,  Frayne,  M.  E.  F.,  Clasper,  S.,  Banerji,  S.,  and  Jackson,  D.  G.  (2009).  A  mechanism  of  sialylation functionally silences the hyaluronan receptor lyve‐1 in lymphatic endothelium. J Biol Chem, 284(6):3935‐3945.  Nishikawa,  T.,  Edelstein,  D.,  Du,  X.  L.,  Yamagishi,  S., Matsumura,  T.,  Kaneda,  Y.,  Yorek, M.  A.,  Beebe,  D.,  Oates,  P.  J., Hammes, H. P., Giardino, I., and Brownlee, M. (2000). Normalizing mitochondrial superoxide production blocks three pathways of hyperglycaemic damage. Nature, 404(6779):787‐790.  Nissen,  S.  E.  and  Wolski,  K.  (2007).  Effect  of  rosiglitazone  on  the  risk  of  myocardial  infarction  and  death  from cardiovascular causes. N Engl J Med, 356(24):2457‐2471.  Norrmén, C., Ivanov, K. I., Cheng, J., Zangger, N., Delorenzi, M., Jaquet, M., Miura, N., Puolakkainen, P., Horsley, V., Hu, J., Augustin, H. G.,  Ylä‐Herttuala,  S.,  Alitalo, K.,  and Petrova,  T.  V.  (2009).  Foxc2  controls  formation  and maturation of lymphatic collecting vessels through cooperation with NFATc1. J Cell Biol, 185(3):439‐457.  Nougues, J., Reyne, Y., and Dulor, J. P. (1988). Differentiation of rabbit adipocyte precursors in primary culture. Int J Obes, 12(4):321‐333.  Ny, A., Koch, M., Vandevelde, W., Schneider, M., Fischer, C., Diez‐Juan, A., Neven, E., Geudens,  I., Maity, S., Moons, L., Plaisance, S., Lambrechts, D., Carmeliet, P., and Dewerchin, M. (2008). Role of VEGF‐D and VEGFR‐3 in developmental lymphangiogenesis, a chemicogenetic study in xenopus tadpoles. Blood, 112(5):1740‐1749.  Ogunbiyi S, Chinien G, Field D, Humphries J, Burand K, Sawyer B, Jeffrey S, Mortimer P, Clasper S, Jackson D, Smith A; London  Lymphedema  Consortium.  (2011)  Molecular  characterization  of  dermal  lymphatic  endothelial  cells  from primary lymphedema skin. Lymphat Res Biol, 9(1):19‐30.  Ohkuma,  M.  (1979).  Histochemical  change  of  the  endothelial  basal  lamina  of  the  diabetic  lymphatic  vessel. Lymphology, 12(1):37‐39.  Ohl,  L.,  Mohaupt,  M.,  Czeloth,  N.,  Hintzen,  G.,  Kiafard,  Z.,  Zwirner,  J.,  Blankenstein,  T.,  Henning,  G.,  and  Förster,  R. (2004).  Ccr7  governs  skin  dendritic  cell  migration  under  inflammatory  and  steady‐state  conditions.  Immunity, 21(2):279‐288.  Oliver, G. (2004). Lymphatic vasculature development. Nat Rev Immunol, 4(1):35‐45.  Oliver, G. and Harvey, N.  (2002). A stepwise model of  the development of  lymphatic vasculature. Ann N Y Acad Sci, 979:159‐65; discussion 188‐96.  Oliver,  G.,  Sosa‐Pineda,  B.,  Geisendorf,  S.,  Spana,  E.  P.,  Doe,  C.  Q.,  and  Gruss,  P.  (1993).  Prox  1,  a  prospero‐related homeobox gene expressed during mouse development. Mech Dev, 44(1):3‐16.  Oliver, G. and Srinivasan, R. S. (2010). Endothelial cell plasticity: how to become and remain a lymphatic endothelial cell. Development, 137(3):363‐372.  Olsson, A.‐K., Dimberg, A., Kreuger, J., and Claesson‐Welsh, L. (2006). Vegf receptor signaling ‐  in control of vascular function. Nat Rev Mol Cell Biol, 7(5):359‐371.  Olszewski, W., Engeset, A., Jaeger, P. M., Sokolowski, J., and Theodorsen, L. (1977). Flow and composition of leg lymph in normal men during venous stasis, muscular activity and local hyperthermia. Acta Physiol Scand, 99(2):149‐155.  Ordonez, N. G. (2006). Podoplanin: a novel diagnostic immunohistochemical marker. Adv Anat Pathol, 13(2):83‐88.  Ostergård,  T.,  Jessen, N.,  Schmitz,  O.,  and Mandarino,  L.  J.  (2007).  The  effect  of  exercise,  training,  and  inactivity  on insulin sensitivity in diabetics and their relatives: what is new? Appl Physiol Nutr Metab, 32(3):541‐548.  Parsons, R.  J. and McMaster, P. D.  (1938). The effect of  the pulse upon the  formation and  flow of  lymph.  J Exp Med, 68(3):353‐376.  Pennisi, D., Gardner, J., Chambers, D., Hosking, B., Peters, J., Muscat, G., Abbott, C., and Koopman, P. (2000). Mutations in SOX18 underlie cardiovascular and hair follicle defects in ragged mice. Nat Genet, 24(4):434‐437.  

Page 151: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  136

Perou, C. M., Jeffrey, S. S., van de Rijn, M., Rees, C. A., Eisen, M. B., Ross, D. T., Pergamenschikov, A., Williams, C. F., Zhu, S. X.,  Lee,  J.  C.,  Lashkari,  D.,  Shalon,  D.,  Brown,  P.  O.,  and  Botstein,  D.  (1999).  Distinctive  gene  expression  patterns  in human mammary epithelial cells and breast cancers. Proc Natl Acad Sci U S A, 96(16):9212‐9217.  Perou, C. M., Sorlie, T., Eisen, M. B., van de Rijn, M., Jeffrey, S. S., Rees, C. A., Pollack, J. R., Ross, D. T., Johnsen, H., Akslen, L.  A.,  Fluge,  O.,  Pergamenschikov,  A., Williams,  C.,  Zhu,  S.  X.,  Lonning,  P.  E.,  Borresen‐Dale,  A.  L.,  Brown,  P.  O.,  and Botstein, D. (2000). Molecular portraits of human breast tumours. Nature, 406(6797):747‐752.  Pessin, J. E. and Saltiel, A. R. (2000). Signaling pathways in insulin action: molecular targets of insulin resistance. J Clin Invest, 106(2):165‐169. Petrova, T. V., Karpanen, T., Norrmén, C., Mellor, R., Tamakoshi, T., Finegold, D., Ferrell, R., Kerjaschki, D., Mortimer, P., Ylä‐Herttuala,  S.,  Miura,  N.,  and  Alitalo,  K.  (2004).  Defective  valves  and  abnormal  mural  cell  recruitment  underlie lymphatic vascular failure in lymphedema distichiasis. Nat Med, 10(9):974‐981.  Petrova,  T.  V.,  Mäkinen,  T.,  Mäkelä,  T.  P.,  Saarela,  J.,  Virtanen,  I.,  Ferrell,  R.  E.,  Finegold,  D.  N.,  Kerjaschki,  D.,  Ylä‐Herttuala, S., and Alitalo, K. (2002). Lymphatic endothelial reprogramming of vascular endothelial cells by the prox‐1 homeobox transcription factor. EMBO J, 21(17):4593‐4599.  Pfaffl  M. W.  (2001).  A  new mathematical model  for  relative  quantification  in  real‐time  RT‐PCR. Nucleic Acids Res, 29(9):e45.  Pierce  A,  Smith  DL,  Jakobsen  LV, Whetton  AD,  Spooncer  E.  (2001).  The  specific  enhancement  of  interferon  alpha induced growth inhibition by BCR/ABL only occurs in multipotent cells. Hematol J, 2(4):257‐64.  Pierce,  M.,  Keen,  H.,  and  Bradley,  C.  (1995).  Risk  of  diabetes  in  offspring  of  parents  with  noninsulin‐  dependent diabetes. Diabet Med, 12(1):6‐13.  Pinhas‐Hamiel, O.  and Zeitler, P.  (2007). Clinical presentation and  treatment of  type 2 diabetes  in  children. Pediatr Diabetes, 8 Suppl 9:16‐27.  Pirola  L,  Balcerczyk  A,  Tothill  RW,  Haviv  I,  Kaspi  A,  Lunke  S,  Ziemann  M,  Karagiannis  T,  Tonna  S,  Kowalczyk  A, Beresford‐Smith B, Macintyre G, Kelong M, Hongyu Z, Zhu  J, El‐Osta A.  (2011). Genome‐wide analysis distinguishes hyperglycemia regulated epigenetic signatures of primary vascular cells. Genome Res, 21(10):1601‐1615.   Podgrabinska, S., Braun, P., Velasco, P., Kloos, B., Pepper, M. S.,  and Skobe, M.  (2002). Molecular characterization of lymphatic endothelial cells. Proc Natl Acad Sci U S A, 99(25):16069‐16074.  Pollock,  J.  D.  (2002).  Gene  expression  profiling:  methodological  challenges,  results,  and  prospects  for  addiction research. Chem Phys Lipids, 121(1‐2):241‐256.  Pond, C. M. and Mattacks, C. A. (1995). Interactions between adipose tissue around lymph nodes and lymphoid cells in vitro. J Lipid Res, 36(10):2219‐2231.  Pond, C. M. and Mattacks, C. A. (1998). In vivo evidence for the involvement of the adipose tissue surrounding lymph nodes in immune responses. Immunol Lett, 63(3):159‐167.  Porta, M. and Bandello, F. (2002). Diabetic retinopathya clinical update. Diabetologia, 45(12):1617‐1634.  Pradhan,  A.  D.,  Rifai,  N.,  Buring,  J.  E.,  and  Ridker,  P.  M.  (2007).  Hemoglobin  a1c  predicts  diabetes  but  not cardiovascular disease in nondiabetic women. Am J Med, 120(8):720‐727.  Pradhan, A. D., Manson, J. E., Rifai, N., Buring, J. E., and Ridker, P. M. (2001). C‐reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA, 286(3):327‐334.  Qi, W., Chen, X., Gilbert, R. E., Zhang, Y., Waltham, M., Schache, M., Kelly, D. J., and Pollock, C. A. (2007). High glucose‐induced thioredoxin‐interacting protein in renal proximal tubule cells is independent of transforming growth factor‐beta1. Am J Pathol, 171(3):744‐754.  Qu, P., Ji, R. C., and Kato, S. (2003). Histochemical analysis of lymphatic endothelial cells in the pancreas of non‐obese diabetic mice. J Anat, 203(5):523‐530.  Qu, P., Ji, R.‐C., Shimoda, H., Miura, M., and Kato, S. (2004). Study on pancreatic lymphatics in nonobese diabetic mouse with  prevention  of  insulitis  and  diabetes  by  adjuvant  immunotherapy.  Anat  Rec  A  Discov  Mol  Cell  Evol  Biol, 281(2):1326‐1336.  Quaye,  I. K., Ababio, G.,  and Amoah, A. G.  (2006). Haptoglobin 2‐2 phenotype  is  a  risk  factor  for  type 2 diabetes  in ghana. J Atheroscler Thromb, 13(2):90‐94. 

Page 152: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  137

Rahman, F. Z., Smith, A. M., Hayee, B., Marks, D. J. B., Bloom, S. L., and Segal, A. W. (2010). Delayed resolution of acute inflammation  in  ulcerative  colitis  is  associated  with  elevated  cytokine  release  downstream  of  TLR4.  PLoS  One, 5(3):e9891.  Rajala,  M.  W.  and  Scherer,  P.  E.  (2003).  Minireview:  The  adipocyte‐at  the  crossroads  of  energy  homeostasis, inflammation, and atherosclerosis. Endocrinology, 144(9):3765‐3773.  Ramirez,  M.  I.,  Millien,  G.,  Hinds,  A.,  Cao,  Y.,  Seldin,  D.  C.,  and  Williams,  M.  C.  (2003).  T1alpha,  a  lung  type  I  cell differentiation gene, is required for normal lung cell proliferation and alveolus formation at birth. Dev Biol, 256(1):61‐72.  Randolph,  G.  J.,  Angeli,  V.,  and  Swartz,  M.  A.  (2005).  Dendritic‐cell  trafficking  to  lymph  nodes  through  lymphatic vessels. Nat Rev Immunol, 5(8):617‐628.  Ratzinger, G., Stoitzner, P., Ebner, S., Lutz, M. B., Layton, G. T., Rainer, C., Senior, R. M., Shipley, J. M., Fritsch, P., Schuler, G., and Romani, N. (2002). Matrix metalloproteinases 9 and 2 are necessary for the migration of Langerhans cells and dermal dendritic cells from human and murine skin. J Immunol, 168(9):4361‐4371.  Ravussin, E. and Smith, S. R.  (2002).  Increased fat  intake,  impaired  fat oxidation, and failure of  fat cell proliferation result in ectopic fat storage, insulin resistance, and type 2 diabetes mellitus. Ann N Y Acad Sci, 967:363‐378.  Reinehr, T. (2005). Clinical presentation of type 2 diabetes mellitus in children and adolescents. Int J Obes (Lond), 29 Suppl 2:S105‐S110.  Rishi, A. K.,  Joyce‐Brady, M., Fisher,  J., Dobbs, L. G., Floros,  J., VanderSpek,  J., Brody,  J. S., and Williams, M. C.  (1995). Cloning, characterization, and development expression of a rat lung alveolar type I cell gene in embryonic endodermal and neural derivatives. Dev Biol, 167(1):294‐306.  Rockson, S. G. (2001). Lymphedema. Am J Med, 110(4):288‐295.  Rockson, S. G. (2008). Diagnosis and management of lymphatic vascular disease. J Am Coll Cardiol, 52(10):799‐806.  Roesli,  C.,  Mumprecht,  V.,  Neri,  D.,  and  Detmar,  M.  (2008).  Identification  of  the  surface‐accessible,  lineage‐specific vascular proteome by two‐dimensional peptide mapping. FASEB J, 22(6):1933‐1944.  Rosen, E. D.  (2002). The molecular control of adipogenesis, with special  reference  to  lymphatic pathology. Ann N Y Acad Sci, 979:143‐58.  Rovenská E, Rovenský J. (2011). Lymphatic vessels: structure and function. Isr Med Assoc J, 13(12):762‐768.  Roy S, Khanna S, Rink C, Biswas S, Sen CK. (2008). Characterization of the acute temporal changes in excisional murine cutaneous wound inflammation by screening of the wound‐edge transcriptome. Physiol Genomics,34(2):162‐184.   Roy,  S.,  Cagliero,  E.,  and  Lorenzi,  M.  (1996).  Fibronectin  overexpression  in  retinal  microvessels  of  patients  with diabetes. Invest Ophthalmol Vis Sci, 37(2):258‐266.  Roy, S., Maiello, M., and Lorenzi, M. (1994). Increased expression of basement membrane collagen in human diabetic retinopathy. J Clin Invest, 93(1):438‐442.  Ruddell  A,  Kelly‐Spratt  KS,  Furuya M,  Parghi  SS,  Kemp CJ.  (2008).  p19/Arf  and  p53  suppress  sentinel  lymph node lymphangiogenesis and carcinoma metastasis. Oncogene, 27(22):3145‐3155.   Rydén, L., Standl, E., Bartnik, M., den Berghe, G. V., Betteridge, J., de Boer, M.‐J., Cosentino, F.,  Jönsson, B., Laakso, M., Malmberg, K., Priori, S., Ostergren, J., Tuomilehto, J., Thrainsdottir, I., Vanhorebeek, I., Stramba‐Badiale, M., Lindgren, P., Qiao, Q., Priori, S. G., Blanc, J.‐J., Budaj, A., Camm, J., Dean, V., Deckers, J., Dickstein, K., Lekakis, J., McGregor, K., Metra, M., ao Morais,  J., Osterspey, A., Tamargo, J., Zamorano, J. L., Deckers, J. W., Bertrand, M., Charbonnel, B., Erdmann, E., Ferrannini, E., Flyvbjerg, A., Gohlke, H.,  Juanatey,  J. R. G., Graham,  I., Monteiro, P. F., Parhofer, K., Pyörälä, K., Raz,  I., Schernthaner, G., Volpe, M., Wood, D., on Diabetes, T. F., of the European Society of Cardiology (ESC), C. D., and for the Study of Diabetes (EASD), E. A. (2007). Guidelines on diabetes, pre‐diabetes, and cardiovascular diseases: executive summary. the task force on diabetes and cardiovascular diseases of the european society of cardiology (esc) and of the european association for the study of diabetes (EASD). Eur Heart J, 28(1):88‐136.  Saaristo, A., Tammela, T., Farkkilā, A., Kärkkäinen, M., Suominen, E., Ylä‐Herttuala, S., and Alitalo, K. (2006). Vascular endothelial growth factor‐c accelerates diabetic wound healing. Am J Pathol, 169(3):1080‐1087.  

Page 153: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  138

Saaristo,  A.,  Veikkola,  T.,  Tammela,  T.,  Enholm,  B.,  Karkkainen,  M.  J.,  Pajusola,  K.,  Bueler,  H.,  Ylä‐Herttuala,  S.,  and Alitalo, K.  (2002). Lymphangiogenic gene  therapy with minimal blood vascular  side effects.  J Exp Med,  196(6):719‐730.  Sabin, F. R. (1902). On the origin of the lymphatic system from the veins, and the development of the lymph hearts and thoracic duct in the pig. Am J Anat, 1:367‐389.  Salmaggi, A., Gelati, M., Dufour, A., Corsini, E., Pagano, S., Baccalini, R., Ferrero, E., Scabini, S., Silei, V., Ciusani, E., and Rossi, M. D. (2002). Expression and modulation of IFN‐gammainducible chemokines (IP‐10, MIG, and I‐TAC) in human brain endothelium and astrocytes: possible relevance for the immune invasion of the central nervous system and the pathogenesis of multiple sclerosis. J Interferon Cytokine Res, 22(6):631‐640.  Salmi, M., Koskinen, K., Henttinen, T., Elima, K., and Jalkanen, S. (2004). Clever‐1 mediates lymphocyte transmigration through vascular and lymphatic endothelium. Blood, 104(13):3849‐3857.  Saltiel,  A.  R.  and Kahn,  C.  R.  (2001).  Insulin  signalling  and  the  regulation  of  glucose  and  lipid metabolism. Nature, 414(6865):799‐806.  Saxena, R., Voight, B. F., Lyssenko, V., Burtt, N. P., de Bakker, P. I. W., Chen, H., Roix, J. J., Kathiresan, S., Hirschhorn, J. N., Daly, M. J., Hughes, T. E., Groop, L., Altshuler, D., Almgren, P., Florez, J. C., Meyer, J., Ardlie, K., Bostrm, K. B., Isomaa, B., Lettre, G., Lindblad, U., Lyon, H. N., Melander, O., Newton‐Cheh, C., Nilsson, P., Orho‐Melander, M., Råstam, L., Speliotes, E. K., Taskinen, M.‐R., Tuomi, T., Guiducci, C., Berglund, A., Carlson,  J., Gianniny, L., Hackett, R., Hall, L., Holmkvist,  J., Laurila,  E.,  Sjögren,  M.,  Sterner,  M.,  Surti,  A.,  Svensson,  M.,  Svensson,  M.,  Tewhey,  R.,  Blumenstiel,  B.,  Parkin,  M., Defelice, M.,  Barry, R.,  Brodeur, W.,  Camarata,  J.,  Chia, N.,  Fava, M.,  Gibbons,  J., Handsaker,  B., Healy,  C., Nguyen, K., Gates, C., Sougnez, C., Gage, D., Nizzari, M., Gabriel, S. B., Chirn, G.‐W., Ma, Q., Parikh, H., Richardson, D., Ricke, D., and Purcell, S. (2007). Genome‐wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science, 316(5829):1331‐1336.  Schacht,  V.,  Dadras,  S.  S.,  Johnson,  L.  A.,  Jackson,  D.  G.,  Hong,  Y.‐K.,  and  Detmar,  M.  (2005).  Up‐regulation  of  the lymphatic marker podoplanin, a mucin‐type transmembrane glycoprotein,  in human squamous cell carcinomas and germ cell tumors. Am J Pathol, 166(3):913‐921.  Schacht, V., Ramirez, M.  I., Hong, Y.‐K., Hirakawa,  S.,  Feng, D., Harvey, N., Williams, M., Dvorak, A. M., Dvorak, H.  F., Oliver, G.,  and Detmar, M.  (2003). T1alpha/podoplanin deficiency disrupts normal  lymphatic vasculature  formation and causes lymphedema. EMBO J, 22(14):3546‐3556.  Schachtrup,  C.,  Emmler,  T.,  Bleck,  B.,  Sandqvist,  A.,  and  Spener,  F.  (2004).  Functional  analysis  of  peroxisome‐proliferator‐responsive element motifs in genes of fatty acid‐binding proteins. Biochem J, 382(Pt 1):239‐245. Schad, H., Flowaczny, H., Brechtelsbauer, H., and Birkenfeld, G. (1978). The significance of respiration for thoracic duct ow in relation to other driving forces of lymph flow. Pugers Arch, 378(2):121‐125.  Schalkwijk, C. G. and Stehouwer, C. D. A. (2005). Vascular complications  in diabetes mellitus: the role of endothelial dysfunction. Clin Sci (Lond), 109(2):143‐159.  Schena, M. (1996). Genome analysis with gene expression microarrays. Bioessays, 18(5):427‐431.  Schena, M., Shalon, D., Davis, R. W., and Brown, P. O. (1995). Quantitative monitoring of gene expression patterns with a complementary dna microarray. Science, 270(5235):467‐470.  Schittek, B., Paulmann, M., Senyürek, I., and Steffen, H. (2008). The role of antimicrobial peptides in human skin and in skin infectious diseases. Infect Disord Drug Targets, 8(3):135‐143.  Schmid‐Schönbein, G. W. (1990). Microlymphatics and lymph flow. Physiol Rev, 70(4):987‐1028.  Schneider, M., Conway, E. M., and Carmeliet, P. (2005). Lymph makes you fat. Nat Genet, 37(10):1023‐1024.  Schneider, M., Othman‐Hassan, K., Christ, B., and Wilting, J. (1999). Lymphangioblasts in the avian wing bud. Dev Dyn, 216(4‐5):311‐319.  Schulthess, F. T., Paroni, F., Sauter, N. S., Shu, L., Ribaux, P., Haataja, L., Strieter, R. M., Oberholzer,  J., King, C. C., and Maedler, K.  (2009). CXCL10  impairs beta  cell  function and viability  in diabetes  through TLR4 signaling. Cell Metab, 9(2):125‐139.  Schwarzenberger, K. and Udey, M. C.  (1996). Contact allergens and epidermal proinflammatory cytokines modulate langerhans cell e‐cadherin expression in situ. J Invest Dermatol, 106(3):553‐558.  

Page 154: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  139

Seaman S, Stevens J, Yang MY, Logsdon D, Graff‐Cherry C, St Croix B. (2007). Genes that distinguish physiological and pathological angiogenesis. Cancer Cell, 11(6):539‐554.  Seifarth,  C.  C.,  Hinkmann,  C.,  Hahn,  E.‐G.,  Lohmann,  T.,  and  Harsch,  I.  A.  (2008).  Reduced  frequency  of  peripheral dendritic cells in type 2 diabetes. Exp Clin Endocrinol Diabetes, 116(3):162‐166.  Senn,  J.  J.,  Klover,  P.  J.,  Nowak,  I.  A.,  Zimmers,  T.  A.,  Koniaris,  L.  G.,  Furlanetto,  R.  W.,  and  Mooney,  R.  A.  (2003). Suppressor  of  cytokine  signaling‐3  (SOCS‐3),  a  potential mediator  of  interleukin‐6‐dependent  insulin  resistance  in hepatocytes. J Biol Chem, 278(16):13740‐13746.  Seyama K,  Kumasaka T,  Kurihara M, Mitani K,  Sato T.  (2010).  Lymphangioleiomyomatosis:  a  disease  involving  the lymphatic system. Lymphat Res Biol, 8(1):21‐31.  Shah, B. R. and Hux,  J. E.  (2003). Quantifying the risk of  infectious diseases  for people with diabetes. Diabetes Care, 26(2):510‐513.  Shields, J. D., Fleury, M. E., Yong, C., Tomei, A. A., Randolph, G. J., and Swartz, M. A. (2007). Autologous chemotaxis as a mechanism  of  tumor  cell  homing  to  lymphatics  via  interstitial  flow  and  autocrine  CCR7  signaling.  Cancer  Cell, 11(6):526‐538.  Shields, J. D., Kourtis, I. C., Tomei, A. A., Roberts, J. M., and Swartz, M. A. (2010). Induction of lymphoidlike stroma and immune escape by tumors that express the chemokine CCL21. Science, 328(5979):749‐752.  Simonson, G. D.  and Kendall, D. M.  (2005). Diagnosis of  insulin  resistance and associated  syndromes:  the spectrum from the metabolic syndrome to type 2 diabetes mellitus. Coron Artery Dis, 16(8):465‐472.  Singh, U. P., Venkataraman, C., Singh, R., and Lillard, J. W. (2007). CXCR3 axis: role in inflammatory bowel disease and its therapeutic implication. Endocr Metab Immune Disord Drug Targets, 7(2):111‐123.  Sladek, R., Rocheleau, G., Rung, J., Dina, C., Shen, L., Serre, D., Boutin, P., Vincent, D., Belisle, A., Hadjadj, S., Balkau, B., Heude, B., Charpentier, G., Hudson, T. J., Montpetit, A., Pshezhetsky, A. V., Prentki, M., Posner, B. I., Balding, D. J., Meyre, D.,  Polychronakos, C.,  and Froguel,  P.  (2007). A  genome‐wide association  study  identifies novel  risk  loci  for  type 2 diabetes. Nature, 445(7130):881‐885.  Slyke, P. V., Alami,  J., Martin, D., Kuliszewski, M., Leong‐Poi, H., Sefton, M. V., and Dumont, D. (2009). Acceleration of diabetic wound healing by an angiopoietin peptide mimetic. Tissue Eng Part A, 15(6):1269‐1280.  Sosa‐Pineda, B., Wigle, J. T., and Oliver, G. (2000). Hepatocyte migration during liver development requires PROX1. Nat Genet, 25(3):254‐255.  Soukas,  A.,  Cohen,  P.,  Socci,  N.  D.,  and  Friedman,  J. M.  (2000).  Leptin‐specific  patterns  of  gene  expression  in white adipose tissue. Genes Dev, 14(8):963‐980. Stacker, S. A., Achen, M. G., Jussila, L., Baldwin, M. E., and Alitalo, K. (2002). Lymphangiogenesis and cancer metastasis. Nat Rev Cancer, 2(8):573‐583.  Stacker, S. A., Caesar, C., Baldwin, M. E., Thornton, G. E., Williams, R. A., Prevo, R., Jackson, D. G., Nishikawa, S., Kubo, H., and Achen, M. G. (2001). VEGF‐D promotes the metastatic spread of tumor cells via the lymphatics. Nat Med, 7(2):186‐191.  Staiger, H., Machicao, F., Fritsche, A., and Häring, H.‐U. (2009). Pathomechanisms of type 2 diabetes genes. Endocr Rev, 30(6):557‐585.  St Croix B, Rago C, Velculescu V, Traverso G, Romans KE, Montgomery E, Lal A, Riggins GJ, Lengauer C, Vogelstein B, Kinzler KW. (2000) Genes expressed in human tumor endothelium. Science, 18;289(5482):1197‐1202.  Stehouwer, C. D., Lambert, J., Donker, A. J., and van Hinsbergh, V. W. (1997). Endothelial dysfunction and pathogenesis of diabetic angiopathy. Cardiovasc Res, 34(1):55‐68.  Stehouwer, C. D. and Schaper, N. C. (1996). The pathogenesis of vascular complications of diabetes mellitus: one voice or many? Eur J Clin Invest, 26(7):535‐543.  Stokić,  D.,  Wick,  N.,  Biely,  C.,  Gurnhofer,  E.,  and  Thurner,  S.  (2006).  Statistically  consistent  identification  of differentially  expressed  genes  in  DNA  chip  data  over  the  whole  expression  range:  relative  variance method. Appl Bioinformatics, 5(4):277‐284.  Stumvoll, M., Goldstein, B. J., and van Haeften, T. W. (2005). Type 2 diabetes: principles of pathogenesis and therapy. Lancet, 365(9467):1333‐1346. 

Page 155: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  140

Suzuki‐Inoue,  K.  (2009).  CLEC‐2,  the  novel  platelet  activation  receptor  and  its  internal  ligand,  podoplanin. Rinsho Ketsueki, 50(5):389‐398.  Suzuki‐Inoue, K., Fuller, G. L. J., García, A., Eble, J. A., Pöhlmann, S., Inoue, O., Gartner, T. K., Hughan, S. C., Pearce, A. C., Laing, G. D., Theakston, R. D. G., Schweighoffer, E., Zitzmann, N., Morita, T., Tybulewicz, V. L. J., Ozaki, Y., and Watson, S. P.  (2006).  A  novel  SYK‐dependent  mechanism  of  platelet  activation  by  the  c‐type  lectin  receptor  CLEC‐2.  Blood, 107(2):542‐549.  Suzuki‐Inoue, K., Kato, Y.,  Inoue, O., Kaneko, M. K., Mishima, K., Yatomi, Y., Yamazaki, Y., Narimatsu, H., and Ozaki, Y. (2007). Involvement of the snake toxin receptor CLEC‐2, in podoplanin‐mediated platelet activation, by cancer cells. J Biol Chem, 282(36):25993‐26001. Swartz  MA,  Lund  AW.  (2012).  Lymphatic  and  interstitial  flow  in  the  tumour  microenvironment:  linking mechanobiology with immunity. Nat Rev Cancer. 12(3):210‐219.  Swartz, M. A. (2001). The physiology of the lymphatic system. Adv Drug Deliv Rev, 50(1‐2):3‐20.  Swartz, M. A. and Skobe, M. (2001). Lymphatic function, lymphangiogenesis, and cancer metastasis. Microsc Res Tech, 55(2):92‐99.  Tabibiazar R,  Cheung L, Han  J,  Swanson  J, Beilhack A, An A, Dadras  SS, Rockson N,  Joshi  S, Wagner R, Rockson SG. (2006). Inflammatory manifestations of experimental lymphatic insufficiency. PLoS Med,3(7):e254.  Tammela, T. and Alitalo, K. (2010). Lymphangiogenesis: Molecular mechanisms and future promise. Cell, 140(4):460‐476.  Tammela,  T.,  Enholm,  B.,  Alitalo,  K.,  and  Paavonen,  K.  (2005a).  The  biology  of  vascular  endothelial  growth  factors. Cardiovasc Res, 65(3):550‐563.  Tammela,  T.,  Petrova,  T.  V.,  and  Alitalo,  K.  (2005b).  Molecular  lymphangiogenesis:  new  players.  Trends  Cell  Biol, 15(8):434‐441.  Tammela, T., Saaristo, A., Lohela, M., Morisada, T., Tornberg, J., Norrmén, C., Oike, Y., Pajusola, K., Thurston, G., Suda, T., Ylä‐Herttuala,  S.,  and  Alitalo,  K.  (2005c).  Angiopoietin‐1  promotes  lymphatic  sprouting  and  hyperplasia.  Blood, 105(12):4642‐4648.  Tan, N.‐S.,  Shaw, N.  S.,  Vinckenbosch, N.,  Liu,  P.,  Yasmin, R., Desvergne, B., Wahli, W.,  and Noy, N.  (2002).  Selective cooperation  between  fatty  acid  binding  proteins  and  peroxisome  proliferator‐activated  receptors  in  regulating transcription. Mol Cell Biol, 22(14):5114‐5127.  Tattersal, R. B. and Fajans, S. S. (1975). Prevalence of diabetes and glucose intolerance in 199 offspring of thirty‐seven conjugal diabetic parents. Diabetes, 24(5):452‐462.  Thangarajah H, Vial  IN, Chang E, El‐Ftesi S,  Januszyk M, Chang EI, Paterno  J, Neofytou E, Longaker MT, Gurtner GC. (2009). IFATS collection: Adipose stromal cells adopt a proangiogenic phenotype under the influence of hypoxia. Stem Cells,27(1):266‐274.  Tilg, H. and Moschen, A. R. (2008). Inflammatory mechanisms in the regulation of insulin resistance. Mol Med, 14(3‐4):222‐231.  Trzewik, J., Mallipattu, S. K., Artmann, G. M., Delano, F. A., and Schmid‐Schönbein, G. W. (2001). Evidence for a second valve system in lymphatics: endothelial microvalves. FASEB J, 15(10):1711‐1717.  Tuomilehto, J., Lindström, J., Eriksson, J. G., Valle, T. T., Hämäläinen, H., Ilanne‐Parikka, P., Keinänen‐Kiukaanniemi, S., Laakso, M., Louheranta, A., Rastas, M., Salminen, V., Uusitupa, M., and Group, F. D. P. S.  (2001). Prevention of  type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med, 344(18):1343‐1350.  Uhrin, P., Zaujec, J., Breuss, J. M., Olcaydu, D., Chrenek, P., Stockinger, H., Fuertbauer, E., Moser, M., Haiko, P., Fässler, R., Alitalo, K., Binder, B. R., and Kerjaschki, D. (2010). Novel function for blood platelets and podoplanin in developmental separation of blood and lymphatic circulation. Blood, 115(19):3997‐4005.  Ushio‐Fukai, M. (2007). Vegf signaling through nadph oxidase‐derived ros. Antioxid Redox Signal, 9(6):731‐739.  Uysal, K. T.,  Scheja, L., Wiesbrock,  S. M., Bonner‐Weir,  S.,  and Hotamisligil, G.  S.  (2000).  Improved glucose and  lipid metabolism in genetically obese mice lacking aP2. Endocrinology, 141(9):3388‐3396.  

Page 156: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  141

Vainionpää, N., Bützow, R., Hukkanen, M., Jackson, D. G., Pihlajaniemi, T., Sakai, L. Y., and Virtanen, I. (2007). Basement membrane  protein  distribution  in  LYVE‐1‐immunoreactive  lymphatic  vessels  of  normal  tissues  and  ovarian carcinomas. Cell Tissue Res, 328(2):317‐328.  van der Putte, S. C.  (1975). The early development of  the  lymphatic  system  in mouse embryos. Acta Morphol Neerl Scand, 13(4):245‐286.  Van Hattem S, Bootsma AH, Thio HB. (2008). Skin manifestations of diabetes. Cleve Clin J Med, 75(11):772‐787.   van Tilburg,  J., van Haeften, T. W., Pearson, P., and Wijmenga, C.  (2001). Defining the genetic contribution of  type 2 diabetes mellitus. J Med Genet, 38(9):569‐578.  Vigl B, Aebischer D, Nitschké M, Iolyeva M, Röthlin T, Antsiferova O, Halin C. (2011). Tissue inflammation modulates gene expression of  lymphatic endothelial  cells and dendritic  cell migration  in a  stimulus‐dependent manner. Blood, 7;118(1):205‐215.   Vigl, B., Zgraggen, C., Rehman, N., Banziger‐Tobler, N. E., Detmar, M., and Halin, C. (2009). Coxsackie‐ and adenovirus receptor (car) is expressed in lymphatic vessels in human skin and affects lymphatic endothelial cell function in vitro. Exp Cell Res, 315(2):336‐347.  Vijayaraghavan, S., Hitman, G. A., and Kopelman, P. G. (1994). Apolipoprotein‐D polymorphism: a genetic marker for obesity and hyperinsulinemia. J Clin Endocrinol Metab, 79(2):568‐570.  von Andrian, U. H. and Mempel, T. R. (2003). Homing and cellular traffic in lymph nodes. Nat Rev Immunol, 3(11):867‐878.  von der Weid PY, Rehal S, Ferraz JG. (2011) Role of the lymphatic system in the pathogenesis of Crohn's disease. Curr Opin Gastroenterol, 27(4):335‐341.  Wegner, L., Andersen, G., Sparso, T., Grarup, N., Glümer, C., Borch‐Johnsen, K., Jorgensen, T., Hansen, T., and Pedersen, O. (2007). Common variation in LMNA increases susceptibility to type 2 diabetes and associates with elevated fasting glycemia  and  estimates  of  body  fat  and  height  in  the  general  population:  studies  of  7,495 danish whites. Diabetes, 56(3):694‐698.  Weisberg, S. P., McCann, D., Desai, M., Rosenbaum, M., Leibel, R. L., and Ferrante, A. W. (2003). Obesity is associated with macrophage accumulation in adipose tissue. J Clin Invest, 112(12):1796‐1808.  Wetzler  C,  Kämpfer  H,  Stallmeyer  B,  Pfeilschifter  J,  Frank  S.  (2000).  Large  and  sustained  induction  of  chemokines during  impaired  wound  healing  in  the  genetically  diabetic  mouse:  prolonged  persistence  of  neutrophils  and macrophages during the late phase of repair. J Invest Dermatol, 115(2):245‐253.  Wick,  N.,  Bruck,  J.,  Gurnhofer,  E.,  Steiner,  C.‐W.,  Giovanoli,  P.,  Kerjaschki,  D.,  and  Thurner,  S.  (2004).  Nonuniform hybridization: a potential source of error in oligonucleotide‐chip experiments with low amounts of starting material. Diagn Mol Pathol, 13(3):151‐159.  Wick, N., Haluza, D., Gurnhofer, E., Raab, I., Kasimir, M.‐T., Prinz, M., Steiner, C.‐W., Reinisch, C., Howorka, A., Giovanoli, P.,  Buchsbaum,  S.,  Krieger,  S.,  Tschachler,  E.,  Petzelbauer,  P.,  and  Kerjaschki,  D.  (2008).  Lymphatic  precollectors contain  a  novel,  specialized  subpopulation  of  podoplanin  low,  CCL27‐expressing  lymphatic  endothelial  cells. Am  J Pathol, 173(4):1202‐1209.  Wick,  N.,  Saharinen,  P.,  Saharinen,  J.,  Gurnhofer,  E.,  Steiner,  C. W.,  Raab,  I.,  Stokic,  D.,  Giovanoli,  P.,  Buchsbaum,  S., Burchard,  A.,  Thurner,  S.,  Alitalo,  K.,  and  Kerjaschki,  D.  (2007).  Transcriptomal  comparison  of  human  dermal lymphatic endothelial cells ex vivo and in vitro. Physiol Genomics, 28(2):179‐192.  Wicki, A. and Christofori, G. (2007). The potential role of podoplanin in tumour invasion. Br J Cancer, 96(1):1‐5.  Wicki, A., Lehembre, F., Wick, N., Hantusch, B., Kerjaschki, D., and Christofori, G. (2006). Tumor invasion in the absence of  epithelial‐mesenchymal  transition:  podoplanin‐mediated  remodeling  of  the  actin  cytoskeleton.  Cancer  Cell, 9(4):261‐272.  Wiggins, H. and Rappoport, J. (2010). An agarose spot assay for chemotactic invasion. Biotechniques, 48(2):121‐124.  Wigle, J. T., Chowdhury, K., Gruss, P., and Oliver, G. (1999). PROX1 function is crucial for mouse lens‐fibre elongation. Nat Genet, 21(3):318‐322.  Wigle,  J.  T.,  Harvey,  N.,  Detmar, M.,  Lagutina,  I.,  Grosveld,  G.,  Gunn, M.  D.,  Jackson,  D.  G.,  and  Oliver,  G.  (2002).  An essential role for prox1 in the induction of the lymphatic endothelial cell phenotype. EMBO J, 21(7):1505‐1513. 

Page 157: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  142

Wigle, J. T. and Oliver, G. (1999). PROX1 function is required for the development of the murine lymphatic system. Cell, 98(6):769‐778.  Wild, S., Roglic, G., Green, A., Sicree, R., and King, H. (2004). Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care, 27(5):1047‐1053.  Wilting, J., Papoutsi, M., Christ, B., Nicolaides, K. H., von Kaisenberg, C. S., Borges, J., Stark, G. B., Alitalo, K., Tomarev, S. I., Niemeyer, C., and Rössler, J. (2002). The transcription factor PROX1 is a marker for lymphatic endothelial cells  in normal and diseased human tissues. FASEB J, 16(10):1271‐1273.  Wolfrum, C. (2007). Cytoplasmic fatty acid binding protein sensing fatty acids for peroxisome proliferator activated receptor activation. Cell Mol Life Sci, 64(19‐20):2465‐2476.  Xu, H., Barnes, G. T., Yang, Q., Tan, G., Yang, D., Chou, C.  J., Sole,  J., Nichols, A., Ross, J. S., Tartaglia, L. A., and Chen, H. (2003). Chronic inflammation in fat plays a crucial role in the development of obesity‐related insulin resistance. J Clin Invest, 112(12):1821‐1830.  Xu, Y., Yuan, L., Mak, J., Pardanaud, L., Caunt, M., Kasman, I., Larrivée, B., Toro, R. D., Suchting, S., Medvinsky, A., Silva, J., Yang,  J.,  Thomas,  J.‐L.,  Koch,  A.  W.,  Alitalo,  K.,  Eichmann,  A.,  and  Bagri,  A.  (2010).  Neuropilin‐2  mediates  VEGF‐C‐induced lymphatic sprouting together with VEGFR3. J Cell Biol, 188(1):115‐130.  Yang, C.‐X.,  Shi,  J.‐F.,  and Gu, G.‐H.  (2007).  Single nucleotide polymorphism analysis  applied  to  research  susceptible genes of type 2 diabetes. Journal of Clinical Rehabilitative Tissue Engineering Research, 11:2559‐2562.  Ye,  J.‐M.,  Dzamko,  N.,  Cleasby,  M.  E.,  Hegarty,  B.  D.,  Furler,  S.  M.,  Cooney,  G.  J.,  and  Kraegen,  E.  W.  (2004).  Direct demonstration  of  lipid  sequestration  as  a  mechanism  by  which  rosiglitazone  prevents  fatty‐acid‐induced  insulin resistance in the rat: comparison with metformin. Diabetologia, 47(7):1306‐1313.  Yin N, Zhang N, Lal G, Xu  J, Yan M, Ding Y, Bromberg  JS.  (2011) Lymphangiogenesis  is  required  for pancreatic  islet inflammation and diabetes. PLoS One, 6(11):e28023.  Yki‐Järvinen, H. (2004). Thiazolidinediones. N Engl J Med, 351(11):1106‐1118.  Yong C, Bridenbaugh EA, Zawieja DC, Swartz MA. (2007). Microarray analysis of VEGF‐C responsive genes in human lymphatic endothelial cells. Lymphat Res Biol,3(4):183‐207.  Yoshizumi, M., Perrella, M. A., Burnett, J. C., and Lee, M. E. (1993). Tumor necrosis factor downregulates an endothelial nitric oxide synthase mRNA by shortening its half‐life. Circ Res, 73(1):205‐209.  You,  J.‐J., Yang, C.‐H., Chen, M.‐S.,  and Yang, C.‐M.  (2009). Cysteine‐rich 61, a member of  the CCN  family, as a  factor involved in the pathogenesis of proliferative diabetic retinopathy. Invest Ophthalmol Vis Sci, 50(7):3447‐3455.  You, L.‐R., Lin, F.‐J., Lee, C. T., DeMayo, F.  J., Tsai, M.‐J., and Tsai, S. Y. (2005). Suppression of Notch signalling by the COUP‐TFII transcription factor regulates vein identity. Nature, 435(7038):98‐104.  Yuan,  L.,  Moyon,  D.,  Pardanaud,  L.,  Bréant,  C.,  Karkkainen,  M.  J.,  Alitalo,  K.,  and  Eichmann,  A.  (2002).  Abnormal lymphatic vessel development in neuropilin 2 mutant mice. Development, 129(20):4797‐4806.  Zampell  JC, Aschen S, Weitman ES, Yan A, Elhadad S, De Brot M, Mehrara BJ.  (2012). Regulation of adipogenesis by lymphatic fluid stasis: part I. Adipogenesis, fibrosis, and inflammation. Plast Reconstr Surg, 129(4):825‐834.  Zampetaki A, Kiechl S, Drozdov I, Willeit P, Mayr U, Prokopi M, Mayr A, Weger S, Oberhollenzer F, Bonora E, Shah A, Willeit J, Mayr M. (2010). Plasma microRNA profiling reveals loss of endothelial miR‐126 and other microRNAs in type 2 diabetes. Circ Res, 107(6):810‐817.  Zeggini, E., Scott, L.  J., Saxena, R., Voight, B. F., Marchini,  J. L., Hu, T., de Bakker, P.  I. W., Abecasis, G. R., Almgren, P., Andersen, G., Ardlie, K., Bostrm, K. B., Bergman, R. N., Bonnycastle, L. L., Borch‐Johnsen, K., Burtt, N. P., Chen, H., Chines, P. S., Daly, M. J., Deodhar, P., Ding, C.‐J., Doney, A. S. F., Duren, W. L., Elliott, K. S., Erdos, M. R., Frayling, T. M., Freathy, R. M., Gianniny, L., Grallert, H., Grarup, N., Groves, C.  J., Guiducci, C., Hansen, T., Herder, C., Hitman, G. A., Hughes, T. E., Isomaa, B.,  Jackson, A. U.,  Jrgensen, T., Kong, A., Kubalanza, K., Kuruvilla, F. G., Kuusisto, J., Langenberg, C., Lango, H., Lauritzen, T., Li, Y., Lindgren, C. M., Lyssenko, V., Marvelle, A. F., Meisinger, C., Midthjell, K., Mohlke, K. L., Morken, M. A., Morris,  A.  D.,  Narisu,  N.,  Nilsson,  P.,  Owen,  K.  R.,  Palmer,  C.  N.  A.,  Payne,  F.,  Perry,  J.  R.  B.,  Pettersen,  E.,  Platou,  C., Prokopenko, I., Qi, L., Qin, L., Rayner, N. W., Rees, M., Roix, J. J., Sandbaek, A., Shields, B., Sjgren, M., Steinthorsdottir, V., Stringham, H. M., Swift, A. J., Thorleifsson, G., Thorsteinsdottir, U., Timpson, N. J., Tuomi, T., Tuomilehto, J., Walker, M., Watanabe, R. M., Weedon, M. N., Willer, C. J., Consortium, W. T. C. C., Illig, T., Hveem, K., Hu, F. B., Laakso, M., Stefansson, K., Pedersen, O., Wareham, N. J., Barroso, I., Hattersley, A. T., Collins, F. S., Groop, L., McCarthy, M. I., Boehnke, M., and 

Page 158: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  143

Altshuler, D. (2008). Meta‐analysis of genome‐wide association data and large‐scale replication identifies additional susceptibility loci for type 2 diabetes. Nat Genet, 40(5):638‐645.  Zen, K., Liu, Y., McCall,  I. C., Wu, T., Lee, W., Babbin, B. A., Nusrat, A., and Parkos, C. A. (2005). Neutrophil migration across tight junctions is mediated by adhesive interactions between epithelial coxsackie and adenovirus receptor and a junctional adhesion molecule‐like protein on neutrophils. Mol Biol Cell, 16(6):2694‐2703.  Zeyda, M., Farmer, D., Todoric, J., Aszmann, O., Speiser, M., Györi, G., Zlabinger, G. J., and Stulnig, T. M. (2007). Human adipose  tissue  macrophages  are  of  an  anti‐inflammatory  phenotype  but  capable  of  excessive  pro‐inflammatory mediator production. Int J Obes (Lond), 31(9):1420‐1428.  Zeyda, M. and Stulnig, T. M. (2007). Adipose tissue macrophages. Immunol Lett, 112(2):61‐67.  Zeyda,  M.  and  Stulnig,  T.  M.  (2009).  Obesity,  inflammation,  and  insulin  resistance‐a  mini‐review.  Gerontology, 55(4):379‐386.  Zhang,  Q.,  Lu,  Y.,  Proulx,  S.  T.,  Guo,  R.,  Yao,  Z.,  Schwarz,  E.  M.,  Boyce,  B.  F.,  and  Xing,  L.  (2007).  Increased lymphangiogenesis in joints of mice with inflammatory arthritis. Arthritis Res Ther, 9(6):R118.  

 

 

 

 

 

Page 159: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  144

   Curriculum vitae Contact details 

 Private address: Blumenthalstraße 49 69120 Heidelberg, Deutschland 

  Working address: University of Heidelberg        German Cancer Research Center (DKFZ) Institute of Pathology   Helmholtz‐University‐Group "Molecular 

RNA Biology & Cancer" Im Neuenheimer Feld 224        Im Neuenheimer Feld 280 69120 Heidelberg          69120 Heidelberg  Email: [email protected][email protected]‐heidelberg.de Mobil: +49‐1577‐8408901   Personal Information 

Date and Place of Birth:  22 June 1983, Lustenau/Vorarlberg, Austria Nationality:       Austrian Academic Degree:     Medical Doctor (MD)   Education 

2010­present:    Resident pathologist  at  the  Institute  of  Pathology, University Hospital & University of Heidelberg 

2010­present:    Postdoctoral  research  fellow  at  the  Research  Group  "Molecular  RNA Biology  &  Cancer"  –  Dr.  Sven  Diederichs  (www.diederichslab.org)  Main  focus  on:  Identification  and  functional  analysis  of  long  non­coding RNAs in hepatocellular carcinoma 

 2007­present:     PhD  student  at  the  Clinical  Institute  of  Pathology,  General 

Hospital and Medical University Vienna, Austria:     “Molecular and  functional  characterization of  lymphatic  vessels and 

lymphatic endothelial cells in type 2 diabetes mellitus”  

Page 160: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  145

July – Dec 2011:   Gerok stipend within the SFB TRR77 ‘Liver cancer’: Subproject B3 (Non‐coding RNAs in Hepatocellular Carcinoma), Dr. Sven Diederichs  

  September 2007:  Research  internship  at  the  Institute  of  Physiology,  Katowice  University, 

Poland     “Matrix Metalloproteinases in child tumors” 

2006­2007:    MD  thesis  at  the  Neurological  Research  Laboratory  Innsbruck: ”Matrix Metalloproteinases and their  Inhibitors  in Multiple Sclerosis: Correlation with Inflammation and Clinical Parameters”  

2001­2007:     Studies of Human Medicine at the Medical University of Innsbruck, Austria (Graduation: 08/2007) 

 

Scholarships 

2005:   Scholarship  of  the  Federal  State  of  Vorarlberg  for  a  study  visit  at  the UniversityHospital Monterrey, Mexico 

2006:  Merit scholarship of the Medical University of Innsbruck     

 

Language skills  German:   Mother tongue English:   fluent in speaking and writing French:  basic Swedish:   basic   Software skills  Statistical software:  SPSS, R   Office:     MS Office (Word, Excel, PowerPoint), Open Office, Adobe Creative Suite 

(Acrobat, Illustrator, Photoshop), Latex, Jabref, Endnote Operating systems:   Windows, Linux       Other activities 

  Project manager: Molecular mechanisms of diabetic microangiopathy  Responsible Organizer:  

‐ BRIDGING THE GAP ‐ International workshop on cell communication in health and disease  (February 2009 and 2010) at the Medical University of Vienna (www.phd‐cchd.at/cchd‐workshop) 

Member of the Young Scientist Association (YSA) of the Medical University (www.ysa‐muv.org)  and  organisation  of  the  PhD‐Symposium  2010  at  the  Medical  University  of Vienna 

  

Page 161: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  146

Publications   M. Hämmerle, T.M. Keller, B. Hantusch, D. Kerjaschki. Transcriptome  analysis  reveals  inflammatory,  growth  and  lipid metabolic  changes  of  lymphatic vessels in human type 2 diabetes. In revision.  T.  Gutschner, M.  Hämmerle,  M.  Eißmann,  J.  Hsu,  Y.  Kim,  G.  Hung,  A.  Revenko,  G.  Arun,  M. Stentrup, M. Groß, M. Zörnig, A. Robert MacLeod, D. L. Spector, S. Diederichs The non­coding RNA MALAT1  is a  critical  regulator of  the metastasis phenotype of  lung  cancer cells. In revision.  M. Polycarpou‐Schwarz, T. Gutschner, M. Hämmerle, S. Grund, A. Roth, C. Hildenbrand, A. Warth, T. Longerich, S. Aulmann, J. Rom, M. Meister, T. Muley, H. Zabeck, S. Schmidt, T. Ivacevic, V. Benes, K. Breuhahn, P. Schnabel, P. Sinn, H. Hoffmann, P. Schirmacher, S. Diederichs Defining  the  non­coding  RNA  landscape  of  lung,  liver  and  breast  cancer  reveals  concerted regulation, novel tumor­associated long ncRNAs and high tissue specificity. Submitted.  M.  Eißmann*,  T.  Gutschner*,  M.  Hämmerle,  S.  Günther,  M.  Caudron‐Herger,  M.  Groß,  P. Schirmacher, K. Rippe, T. Braun, M. Zörnig* and S. Diederichs*.  Loss of the abundant nuclear non­coding RNA MALAT1 is compatible with life and development. RNA Biology 2012, Aug 1;9(8).  M. Zeyda, B. Wernly, S. Demyanets, C. Kaun, M. Hämmerle, B. Hantusch, M. Schranz, A. Neuhofer, B.K. Itariu, M. Keck, G. Prager, J. Wojta, T.M. Stulnig. Severe obesity increases adipose tissue expression of interleukin‐33 and its receptor ST2, both predominantly detectable in endothelial cells of human adipose tissue. Int J Obes (Lond). 2012 Jul 17. [Epub ahead of print]  C.  Vonach,  K.  Viola,  B.  Giessrigl,  N.  Huttary,  I.  Raab,  R.  Kalt,  S.  Krieger,  TP.  Vo,  S. Madlener,  S. Bauer, B. Marian, M. Hämmerle, N. Kretschy, M. Teichmann, B. Hantusch, S. Stary, C. Unger, M. Seelinger, A. Eger, R. Mader, W. Jäger, W. Schmidt, M. Grusch, N. Dolznig, W. Mikulits, G. Krupitza. NF­κB  mediates  the  12(S)­HETE­induced  endothelial  to  mesenchymal  transition  of lymphendothelial cells during the intravasation of breast carcinoma cells. British Journal of Cancer 2011; 105(2):263‐71.  D. Kerjaschki, M. Rudas, G. Bartel, Z. Bago‐Horvath, V. Sexl, S. Wolbank, C. Schneckenleithner, H. Dolznig,  S. Krieger, B. Hantusch, K. Nagy‐Bojarszky, N. Huttary,  I.  Raab, R. Kalt, K.  Lackner, M. Hämmerle, T. Keller, K. Viola, M. Schreiber, A. Nader, W. Mikulits, M. Gnant, K. Krautgasser, H. Schachner, K. Kaserer, S. Rezar, S. Madlener, C. Vonach, A. Davidovits, H. Nosaka, S. Hirakawa, M. Detmar, K. Alitalo, S. Nijman, F. Offner, T. J. Maier, D. Steinhilber and G. Krupitza. Lipoxygenase  driven  tumor  cell  invasion  of  intrametastatic  lymphatic  vessels  propagates  lymph node metastasis of human carcinomas. Journal of Clinical Investigation 2011; 121(5):2000‐12.  M. Hämmerle, T.M. Keller, B. Hantusch, D. Stokic, C.W. Steiner, D. Kerjaschki. Transcriptomal comparison of human dermal diabetic versus non­diabetic  lymphatic endothelial cells ex vivo. Wiener Klinische Wochenschrift 2008; Vol. 120 (Suppl. 1): 117.  B.  Kuenz,  A.  Lutterotti,  R.  Ehling,  C.  Gneiss, M.  Hämmerle,  C.  Rainer,  F.  Deisenhammer,  M. Schocke, T. Berger, M. Reindl. Cerebrospinal  fluid  B  cells  correlate  with  early  brain  inflammation  in  multiple  sclerosis.  PLoS ONE. 2008 Jul 2;3(7):e2559.  

Page 162: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  147

M.  Hämmerle,  T.M.  Keller,  B.  Hantusch,  D.  Stokic,  N.  Wick,  E.  Gurnhofer,  S.  Thurner,  D. Kerjaschki. Transcriptomal comparison of human dermal diabetic versus non­diabetic lymphatic endothelial cells ex vivo. Der Pathologe 2008, Suppl. 1: 26 (Do‐091).  B. Kuenz, A. Lutterotti, R. Ehling, C. Gneiss, M. Hämmerle, M. Schocke, T. Berger, M. Reindl. Cerebrospinal fluid B cells correlate with early brain inflammation in multiple sclerosis.  Multiple Sclerosis 2008, 14 (Suppl. 1): S133‐134 (P 356).  M. Hämmerle. Matrix  metalloproteinases  and  their  inhibitors  in  multiple  sclerosis  –  Correlation  with  inflammation and clinical parameters. Medizinische Dissertation, Medizinische Universität Innsbruck.   Scientific presentations 

Talks:  M. Hämmerle Molecular and functional analysis of long non­coding RNAs in hepatocellular carcinoma. German Society of Pathology (DGP), Berlin, Germany, May 2012  M Hämmerle. Non­coding RNA in malignant tumors: A new world of tumor biomarkers and target structures in cancer cells. German Society for Endocrinology (DGE), Mannheim, Germany, March 2012.  M. Hämmerle. Molecular and Functional Analysis of long non­coding RNAs in Liver Cancer. SFB TRR77 Annual Meeting, Fulda, Germany, November 2011.  M Hämmerle Non­coding RNA profiling in hepatocellular carcinoma. German Society of Pathology (DGP), Leipzig, Germany, June 2011.  M. Hämmerle. Morphological and molecular characterization of lymphatic vessels and lymphatic endothelial cells in type 2 diabetes mellitus. Invited talk, German Cancer Research Center, Heidelberg, May 2010.   M. Hämmerle. Morphological and molecular characterization of lymphatic vessels and lymphatic endothelial cells in type 2 diabetes mellitus. Invited talk, Institute of Pathology, Heidelberg, April 2010.   M. Hämmerle. Characterization of lymphatic endothelial cells in type 2 diabetes mellitus. Hearing of the Austrian Science Fund (FWF), Vienna, May 2009.  M. Hämmerle. Characterization of diabetic versus non­diabetic lymphatic endothelial cells ex vivo. '92. Jahrestagung der Deutschen Gesellschaft für Pathologie', Berlin, May 2008.  

Page 163: CHARACTERIZATION OF LYMPHATIC VESSELS AND LYMPHATIC ... · vessels in type 2 diabetes in comparison to blood endothelial cells (BECs) and blood vessels. FABP4 was shown to regulate

  148

Posters:  M. Hämmerle, T. Gutschner, S. Diederichs. Molecular and Functional Analysis of long non­coding RNAs in Liver Cancer. SFB TRR77 Annual Meeting, Fulda, Germany, November 2011.  M. Hämmerle, T. Gutschner, H. Uckelmann, M. Baas, K. Breuhahn, P. Schirmacher, S. Diederichs. Molecular and Functional Analysis of long non­coding RNAs in Liver Cancer. 12th Young Scientist Meeting of the German Society of Cell Biology (DGZ): "RNA & Disease", Jena, Germany, September 2011.  M. Hämmerle, T.M. Keller, C.W. Steiner, D. Stokic, B. Hantusch D. Kerjaschki. Lymphatic endothelial cells and lymphatic vessels are altered by the diabetic condition. 6th PhD Symposium, Medical University Vienna, June 2010.  M. Hämmerle, T.M. Keller, B. Hantusch, C.W. Steiner, D. Haluza, D. Kerjaschki. Characterization of lymphatic endothelial cells in type 2 diabetes. 5th PhD Symposium, Medical University Vienna, June 2009.  M. Hämmerle, T.M. Keller, B. Hantusch, C.W. Steiner, D. Haluza, D. Kerjaschki. Characterization of lymphatic endothelial cells in type 2 diabetes. ESH: Interdisciplinary Conference on Angiogenesis, Helsinki, June 2009.  M. Hämmerle, T.M. Keller, B. Hantusch, C.W. Steiner, D. Haluza, D. Kerjaschki. Characterization of lymphatic endothelial cells in type 2 diabetes. EMBO Workshop: Lymphatic & blood vasculature ‐ from models to human disease, Biomedicum Helsinki, June 2009.  M. Hämmerle, T.M. Keller, B. Hantusch, C.W. Steiner, D. Haluza, D. Kerjaschki. Transcriptomal comparison of diabetic versus non­diabetic lymphatic endothelial cells ex vivo.  Joint Annual Meeting of Immunology, Vienna General Hospital, September 2008.  M. Hämmerle, T.M. Keller, B. Hantusch, C.W. Steiner, D. Haluza, D. Kerjaschki. Transcriptomal comparison of diabetic versus non­diabetic lymphatic endothelial cells ex vivo.  4th PhD Symposium, Medical University Vienna, May 2008.  M. Hämmerle, T.M. Keller, B. Hantusch, C.W. Steiner, D. Haluza, D. Kerjaschki. Transcriptomal comparison of diabetic versus non­diabetic lymphatic endothelial cells ex vivo.  Gordon Research Conference, Ventura, March 2008.  B. Künz, M. Hämmerle, R. Ehling, A. Lutterotti, C. Gnei, K. Egger, M. Schocke, T. Berger, M. Reindl. Characterization of Cell Populations in Cerebrospinal Fluid of Patients with Multiple Sclerosis.  6. Jahrestagung der Österreichischen Gesellschaft für Neurologie, Innsbruck, February 2008.  B. Künz, M. Hämmerle, R. Ehling, C. Gneiss, A. Millonig, F. Deisenhammer, T. Berger, M. Reindl. Characterization of Cell Populations in Cerebrospinal Fluid of Patients with Multiple Sclerosis.  3rd Meeting of Doctoral students from Innsbruck's Medical University, November 2007.