LYMPHATIC RESEARCH AND BIOLOGYVolume 4, Number 2, 2006© Mary Ann Liebert, Inc.

Vascular Endothelial Growth Factor–C (VEGF-C)Expression in Normal Human Tissues

K.D. JOORY, 1 J.R. LEVICK2, P.S. MORTIMER3, and D.O. BATES1

ABSTRACT

Objective: To characterize vascular endothelial growth factor-C (VEGF-C) protein expressionin normal human tissues by immunohistochemistry (IHC). VEGF-C is a growth factor for lym-phatic endothelial cells. VEGF-C mRNA and protein are expressed in a variety of canceroustissues, but the localization of VEGF-C protein in many normal human tissues has not beenclearly demonstrated to date. We therefore performed an immunohistochemical survey of thedistribution of intracellular VEGF-C protein in a range of normal human tissue types.

Methods: Five �m sections were cut from archived human tissues. Sections were dewaxed,rehydrated, and subjected to microwave pretreatment. They were incubated with VEGF-C an-tibody before detection with biotinylated secondary antibody using ‘Elite’ avidin-biotin en-zyme complex and diaminobenzidine substrate. The primary antibody recognized the C-ter-minus of the VEGF-C propeptide that is cleaved before secretion and hence only cellularprotein was detected. Negative controls used the same concentration of normal goat IgG.

Results: Staining manifested as small punctate cytoplasmic granules. Strong expression wasobserved in large intestine epithelium, and mammary duct epithelium, skeletal and cardiacmuscle, thyroid, ovary, and the prostate. Weaker expression was also detected in the hepato-cytes close to the terminal hepatic venules of the liver, vascular smooth muscle, and placenta.No expression was consistently detected in spleen or thymus.

Conclusions: Intracellular VEGF-C protein is widely expressed in many normal human adulttissues. Its expression in cancer is not therefore per se indicative of a prolymphangiogenicchange. To demonstrate the latter, a quantitative change in expression level is required.

73

INTRODUCTION

THE PROCESS OF ANGIOGENESIS is the growthand formation of new blood vessels that

sprout from pre-existing blood vessels. Like-wise, lymphangiogenesis is the formation ofnew lymphatics, from pre-existing lymphaticvessels, or from the cardinal vein. These phe-

nomena have received a great deal of attention,particularly in relation to tumor lymphangio-genesis, since tumors often metastasize vialymphatics.1 The most potent mediator of an-giogenesis is the glycoprotein vascular endo-thelial growth factor-A (VEGF-A). Since thediscovery of VEGF-A, a family of related mol-ecules has been described that all bind to and

1Microvascular Research Laboratories, Department of Physiology, Preclinical Veterinary School, University of Bris-tol, Bristol.

Departments of 2Basic Medical Sciences and 3Department of Cardiac and Vascular Sciences, St George’s HospitalMedical School, London.

This work was funded by the Wellcome Trust (Grant Number 62951). Dr. Bates is supported by the British HeartFoundation (BB20000003).

6265_02_p73-82 6/16/06 2:51 PM Page 73

activate endothelial tyrosine kinase receptors.VEGF-C is a member of the family that is a lesspotent angiogenic factor than VEGF-A, but alsohas the ability to stimulate lymphangiogene-sis.2,3 It is produced as a cysteine-rich precur-sor and undergoes several stages of post-trans-lational processing to generate shorter forms ofVEGF-C.4 As the molecule progresses alongthis proteolytic pathway, its affinity for one ofthe VEGF receptors, namely VEGFR-3, is en-hanced.

VEGFR-3 is largely restricted to the lym-phatic endothelium in the adult. Signalingthrough VEGFR-3 is necessary for the mainte-nance and stability of the lymphatic endothe-lial cells in culture. Introduction of exogenoussoluble VEGFR-3, which sequesters VEGF-Cand prevents its action on the endogenous re-ceptor, causes lymphatic endothelial cells toundergo apoptosis, leading to the regression ofdeveloping lymphatics.5,6 VEGFR-3 signalingis required for lymphatic vessel growth anddevelopment in vivo, although not for mainte-nance of the lymphatic vessels.7 The shortest,most proteolytically cleaved form of VEGF-Ccan also bind and activate VEGFR-2, which isfound on blood vascular endothelial cells andstimulates cell proliferation,8 migration,9,10 andincreased vascular permeability.11 It has beensuggested that the importance of the stepwisenature of the proteolytic pathway is to regulateVEGF-C activity in order to prevent unneces-sary angiogenesis whilst maintaining the lym-phatic endothelium. Recombinant VEGF-Cequivalent to the fully processed molecule hasbeen shown to specifically stimulate thegrowth, migration and survival of isolated lym-phatic endothelial cells.6

It is possible that VEGF-C can be held in alatent state bound to the extracellular matrixuntil such time as activity of an unidentifiedprotease may cleave the N-terminal propep-tide, release the protein and hence activate it.4Studies have already shown that VEGF-CmRNA is expressed by various tissues, includ-ing cardiac muscle,12–15 skeletal muscle,13–15

thyroid,16 prostate, ovary,13,15 colon,13,15,17,18

liver,13 placenta, and spleen.13,15 VEGF-C pro-tein has been shown to be expressed in manycancerous tissues including breast carci-noma18–20 and colorectal cancer.18–20 Immuno-

histochemical studies of normal tissue, exam-ining the cellular expression of VEGF-C to in-dicate which cells within a tissue are activelyproducing the protein, ready for secretion andultimately receptor binding, are critical to fur-ther our understanding of the possible rolesand function of the molecule within a physio-logical system. Nevertheless, to date there hasnever been a systematic investigation of the lo-calization of VEGF-C protein in a wide rangeof normal human tissues, and as such, the im-plications of expression of VEGF-C in patho-logical tissues are still unclear. We have there-fore measured VEGF-C expression in a varietyof normal human tissues.

MATERIALS AND METHODS

Histologically normal, formalin-fixed tissuesembedded in paraffin were obtained, in accor-dance with ethics committee approval, fromarchives of tissue removed during surgery,from the Department of Cellular Pathology,Southmead Hospital, Bristol, and from teach-ing archives of the Department of Physiology,University of Bristol. Sections 5 �m thick werecut, mounted, and dried onto slides which hadbeen previously subjected to a double coatingprotocol of immersion in 5 mg/ml gelatine/250�g/ml chromium potassium sulphate solutionin water for 5 minutes, intermediate drying at37°C, and subsequent immersion in 62.5 �g/mlpoly-L lysine dissolved in 0.01 M Tris-HCl atpH 7.6 again for 5 minutes, and dried at 37°C.Sections were dewaxed in Histoclear (Ray-mond A Lamb, Ltd., Eastbourne, UK) rehy-drated by immersion in graded alcohols (100%,90%, 70% ethanol in water for 2 minutes each),and washed twice in phosphate buffered saline(PBS) (137 mM NaCl, 3 mM KCl, 10 mMNa2HPO4, 1.7 mM KH2PO4) for 5 minutes be-fore being subjected to microwave pretreat-ment in 0.1 M Tris/2 mM EDTA buffer (pH 9)for 10 minutes at 800 Watts. The sections wereallowed to cool slowly and washed in distilledwater prior to incubation in 3% hydrogen per-oxide to quench endogenous peroxidase activ-ity. Nonspecific antibody binding was pre-vented by incubating the sections with 1.5%normal horse serum (Vector Laboratories, Ltd.,

JOORY ET AL.74

6265_02_p73-82 6/16/06 2:51 PM Page 74

Peterborough, UK) in PBS for 20 minutes atroom temperature (RT) after a further two 5minute PBS washes. The blocking solution wasblotted off and replaced with the primary an-tibody against VEGF-C (SC-7133, Santa CruzBiotechnology Inc., Santa Cruz, CA) diluted to1.14 �g/ml in blocking solution. Sections werecoverslipped and kept at 4°C overnight for op-timum antibody binding. Following stringentwashing in PBS with 0.05% polyoxyethylenesorbitan monolaurate (Tween 20) (PBS-T) theblocking step was repeated as before but afterblotting was replaced with 2 �g/ml biotinyl-ated anti-goat secondary antibody (Vector) inthe same blocking solution and left to incubatefor 30 minutes at RT. An ‘Elite’ avidin–biotinenzyme complex kit was applied to the sectionsafter two 5 minute PBS-T washes. The proce-dure was completed by performing a furthertwo 5 minute PBS-T washes and treating thesections with 3, 3�-diaminobenzidine substrate(Vector) to yield a brown colored product.Slides were counterstained with hematoxylin,dehydrated, and permanently mounted. Neg-ative controls were performed by taking thesubsequent section and substituting the pri-mary antibody with 1.14 �g/ml of purifiedgoat IgG from a normal nonimmunized animal.Sections were examined using a Nikon E-400microscope and images were captured using aCoolpix 995 digital camera and a DN-100 dig-ital imaging system (Nikon Instruments, Sur-rey, UK). ‘n’ numbers shown represent samplesobtained from different subjects.

RESULTS

Immunostaining for VEGF-C in a multiplesamples of a variety of human tissues re-vealed that the molecule is widely expressed.Punctate cellular granules of brown stainingwere observed in muscle, epithelial, mes-enchymal, and reproductive tissues, butrarely in lymphoid tissue from three differentlymphoid organs.

Muscle

As shown in Figure 1, the protein was con-sistently strongly expressed in all three types

of muscle—a diffuse pattern in cardiac myo-cytes (n � 3), vascular smooth muscle (n � 3),and skeletal muscle (n � 4).

Epithelium

Expression of cellular VEGF-C was strongand consistent in epithelial cells of various tis-sues (Fig. 2). VEGF-C was expressed in ductalepithelial cells of the breast (n � 2), thyroid(n � 4), prostate (n � 4), and by enterocytes inthe colon (n � 3).

Mesenchyme

Expression of VEGF-C in the liver was con-sistently detected at low levels throughout thetissue (Fig. 3C), although the signal was mostconcentrated in the hepatocytes around centralterminal hepatic venules (n � 3).

Reproductive organs

Immunostaining in the ovary was most in-tense in the cortical stroma, although the sur-face epithelium was also positive, as shown inFigure 3A (n � 3). In the placenta, VEGF-Cstaining was most evident in the areas sur-rounding the villus capillaries (n � 4).

Lymphoid tissue

In four out of five spleen samples examined,there was no evidence of VEGF-C expression,whereas in the other one strong staining wasseen in red pulp of spleens (Fig. 4). No clinicaldata was available from these samples. Thethymus (n � 2) and tonsil (n � 5) were consis-tently negative for VEGF-C, although in bothtissues there were occasional leukocytes thatstained positively.

DISCUSSION

We have observed expression of VEGF-Cprotein within the cells of a wide range of normal human tissues. Our results arelargely, but not entirely, consistent with thepattern of VEGF-C mRNA expression de-scribed previously. The results are summa-rized in Table 1.

VEGF-C 75

6265_02_p73-82 6/16/06 2:51 PM Page 75

JOORY ET AL.76

FIG. 1. Expression of VEGF-C protein in vascular smooth muscle (A), cardiac muscle (B), and skeletal muscle (C).Small punctuate granules are visible in many myocytes (arrows). Scale bar 50 �m. Negative controls are shown in theinsert.

6265_02_p73-82 6/16/06 2:51 PM Page 76

VEGF-C expression in muscle

In muscle tissue, VEGF-C mRNA has previ-ously been detected in cardiac tissue by North-ern blot analysis,12–15 although no positive sig-nal was obtained in adult mouse heart by insitu hybridization.21 VEGF-C protein has alsobeen detected in cardiac biopsies from patientswith cardiomyopathy and also those with non-failing hearts by Western blot analysis.12 How-ever, this is the first demonstration of the cel-lular location of VEGF-C protein expression incardiac tissue. Vascular smooth muscle is oneof the few normal tissues that has been specif-ically examined by immunohistochemistry for VEGF-C, though that investigation used different antibodies to the one used in thisstudy.22,23 The findings described above areidentical to those previous studies, confirmingthat the antibody is specific to VEGF-C, at least

in this tissue. In skeletal muscle, an mRNA sig-nal has been described in using in situ hy-bridization,13–15 but this is the first demonstra-tion of VEGF-C protein expression.

VEGF-C expression in epithelial cells

VEGF-C was widely expressed in epithelialcells, particularly ductal cells of secretory organssuch as breast, thyroid, and prostate. A singlestudy described normal human thyroid im-munopositivity in epithelial follicular cells andendothelial vessel lining cells, and the pattern ofstaining demonstrated closely resembled ourown observations, again confirming specificityof the antibody in epithelial cells.24 Moreover,studies examining VEGF-C expression in humanbreast carcinoma by IHC describe, but do not il-lustrate, occasional or weak immunopositivity inductal epithelium of normal breast.25,26

VEGF-C 77

FIG. 2. Expression of VEGF-C was consistently observed (e.g., arrows) in the majority of epithelial cells lining mam-mary ducts (A), colon (B), acini of the thyroid (C), and prostatic glands (D). Scale bar 50 �m. Insert shows negativecontrols.

6265_02_p73-82 6/16/06 2:51 PM Page 77

JOORY ET AL.78

FIG. 3. Expression in the ovary (A) was most intense in areas within the cortical stroma (cs), although the surfaceepithelium (arrow) was also positive. Scale bar 100 �m. Placental expression (B) was most evident in the areas sur-rounding the villus capillaries (vc). Scale bar 50 �m. In the liver (C), the signal was most concentrated in the hepa-tocytes (arrows) around central terminal hepatic venules (star). Scale bar 50 �m. Insert shows negative control.

6265_02_p73-82 6/16/06 2:51 PM Page 78

VEGF-C 79

FIG. 4. No cellular VEGF-C was detected in four of five pathologically normal spleen samples (A–D, scale bar 50�m). No VEGF-C positive cells were observed in any cell type in either thymus (F, scale bar 20 �m) or tonsil (G, scalebar 50 �m).

6265_02_p73-82 6/16/06 2:51 PM Page 79

Hung and colleagues detected expression ofVEGF-C mRNA by real time quantitativeRT–PCR in normal human thyroid tissue frompatients with benign and malignant thyroidtumors.16 We have shown clear epithelial stain-ing in both prostate and ovary tissue, and inthe ovary in stromal cells as well. Althoughsome early articles describe the expression of mRNA in prostate and ovary,13,15 this is thefirst demonstration that there is protein ex-pression in these tissues, and the localizationof that expression.

Strong staining was seen in colonic entero-cytes. Colonic expression of VEGF-C mRNAhas previously been detected by Northern blotanalysis,13,15 RT–PCR/Southern blot analysis17

and in a more recent paper, by quantitative realtime RT–PCR in normal mucosa from patientswith colorectal cancer.18 However, in threestudies investigating VEGF-C expression incolorectal cancer tissues, immunohistochemi-cal staining performed using the same antibodyas the one used in this study did not detectVEGF-C in the adjacent normal colorectal mu-cosa.18–20 It is likely that the reason that proteinexpression was not seen in these studies wasthe lower level of expression relative to carci-noma tissue. Partanen and colleagues have de-scribed strong immunopositivity in normal hu-man colon in nondispersed neuroendocrine

cells that express serotonin23 and chromo-granin A.23

VEGF-C expression in lymphoid tissue

We were unable to detect any VEGF-C pro-tein in the majority of lymphoid tissues fromhistologically normal samples. In three out ofthe five spleen samples examined, VEGF C wasclearly absent. The remaining two spleen sam-ples did give strong expression in the whitematter of the spleen. It is not clear whetherthese samples were from patients with immunechallenges, since no clinical data on these pa-tients was available. Previous studies show thatfaint signals have been observed in this tissueby Northern blot analysis,13,15 but protein ex-pression has not been documented. In thymus,there is concurrence amongst the literature inthat mRNA is not expressed or gives a veryfaint signal by Northern blot.13,15,21 We are notaware of any previous description of VEGF-Cexpression in the tonsil, and so we are unableto compare our finding that VEGF-C protein isabsent from this tissue in normal human sam-ples.

VEGF-C in placenta

We demonstrate clear expression of VEGF-Cprotein in normal placenta. This is in agree-

JOORY ET AL.80

TABLE 1. COMPARISON OF VEGF-C PROTEIN EXPRESSION IN THIS STUDY COMPARED WITH PREVIOUS STUDIES

Proteinexpression

Tissue (other studies) mRNA Expression This study

Skeletal muscle — — YCardiac muscle — Northern—Y12–15 Y

ISH–N12

Vascular smooth muscle Y22,23 Y YBreast 25,26 YProstate — Y13,15 YThyroid 24 Y16 YOvary — Y13,15 YColon — Y13,15,17,18 YSpleen — Weak13,15 NThyroid — N13,15 NTonsil NPlacenta 23 (2nd trimester) Y13,15 Y

N27,28

Liver — Y14,13 YN15,21

6265_02_p73-82 6/16/06 2:51 PM Page 80

ment with VEGF-C mRNA expression, dem-onstrated by Northern blot analysis, in normalplacenta,13,15 and protein has been shown to beexpressed by immunohistochemistry in tro-phoblasts of the second trimester placenta.23

This is in contrast to two studies that failed tosee VEGF-C mRNA expression by in situ hy-bridization at any stage of gestation.27,28

VEGF-C in liver

Expression of VEGF-C protein in normal hu-man liver is described here for the first time.The distribution of the expression appears tocoincide with terminal hepatic venules. mRNAfor VEGF-C in mouse liver has been previouslydescribed,14 and a weak Northern blot signalhas been observed in human liver,13 althoughtwo studies could not find evidence for VEGF-C mRNA expression.15,21

Relation to biological activity andlymphangiogenesis

The results indicate that VEGF-C is a widelyexpressed molecule. We have used an antibodythat detects the cellular form of the protein. Al-though we have not shown secreted protein, itis likely that the protein is indeed readily se-creted. Whether its function is to maintain thelymphatics in and around the tissue, orwhether it has other undescribed functions isnot known. VEGF-C may have a variety ofother actions, because it appears to be ex-pressed in some areas of tissue that do not ex-press VEGFR-3 and/or do not contain lym-phatics. VEGF-C has already been shown to beupregulated in a multitude of tumors, but verylittle work has been done to investigate regu-lation of this molecule in other pathologicalconditions. Further investigation is necessaryto understand more clearly the physiology ofVEGF-C.

ACKNOWLEDGMENTS

The authors would like to thank the Depart-ment of Cellular Pathology, Southmead Hos-pital, Bristol, UK, and the Department of Phys-iology, University of Bristol for provision oftissue samples.

REFERENCES

1. Streit M, Detmar M. Angiogenesis, lymphangiogene-sis, and melanoma metastasis. Oncogene 2003;22:3172–3179.

2. Enholm B, Karpanen T, Jeltsch M, Kubo H, StenbackF, Prevo R, et al. Adenoviral expression of vascularendothelial growth factor-C induces lymphangiogen-esis in the skin. Circ Res 2001;88:623–629.

3. Oh SJ, Jeltsch MM, Birkenhager R, McCarthy JE, Weich HA, Christ B, et al. VEGF and VEGF-C: spe-cific induction of angiogenesis and lymphangiogene-sis in the differentiated avian chorioallantoic mem-brane. Dev Biol 1997;188:96–109.

4. Joukov V, Kaipainen A, Jeltsch M, Pajusola K, Olofs-son B, Kumar V, et al. Vascular endothelial growthfactors VEGF-B and VEGF-C. J Cell Physiol 1997;173:211–215.

5. Kaipainen A, Korhonen J, Mustonen T, van Hins-bergh VW, Fang GH, Dumont D, et al. Expression ofthe fms-like tyrosine kinase 4 gene becomes restrictedto lymphatic endothelium during development. ProcNatl Acad Sci USA 1995;92:3566–3570.

6. Makinen T, Veikkola T, Mustjoki S, Karpanen T,Catimel B, Nice EC, et al. Isolated lymphatic endo-thelial cells transduce growth, survival and migratorysignals via the VEGF-C/D receptor VEGFR-3. EMBOJ 2001;20:4762–4773.

7. Pytowski B, Goldman J, Persaud K, Wu Y, Witte L,Hicklin DJ, et al. Complete and specific inhibition ofadult lymphatic regeneration by a novel VEGFR-3neutralizing antibody. J Natl Cancer Inst 2005;97:14–21.

8. Witzenbichler B, Asahara T, Murohara T, Silver M,Spyridopoulos I, Magner M, et al. Vascular endothe-lial growth factor-C (VEGF-C/VEGF-2) promotes an-giogenesis in the setting of tissue ischemia. Am JPathol 1998;153:381–394.

9. Cao Y, Linden P, Farnebo J, Cao R, Eriksson A, Ku-mar V, et al. Vascular endothelial growth factor C in-duces angiogenesis in vivo. Proc Natl Acad Sci USA1998;95:14389–14394.

10. Joukov V, Kumar V, Sorsa T, Arighi E, Weich H, Sak-sela O, et al. A recombinant mutant vascular endo-thelial growth factor-C that has lost vascular endo-thelial growth factor receptor-2 binding, activation,and vascular permeability activities. J Biol Chem1998;273:6599–6602.

11. Hillman NJ, Whittles CE, Pocock TM, Williams B,Bates DO. Differential effects of vascular endothelialgrowth factor-C and placental growth factor-1 on thehydraulic conductivity of frog mesenteric capillaries.J Vasc Res 2001;38:176–186.

12. Abraham D, Hofbauer R, Schafer R, Blumer R, PaulusP, Miksovsky A, et al. Selective downregulation ofVEGF-A(165), VEGF-R1, and decreased capillary den-sity in patients with dilative but not ischemic car-diomyopathy. Circ Res 2000;87:644–647.

13. Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahti-nen I, Kukk E, et al. A novel vascular endothelial

VEGF-C 81

6265_02_p73-82 6/16/06 2:51 PM Page 81

growth factor, VEGF-C, is a ligand for the Flt4(VEGFR-3) and KDR (VEGFR-2) receptor tyrosine ki-nases. EMBO J 1996;15:290–298.

14. Kukk E, Lymboussaki A, Taira S, Kaipainen A, JeltschM, Joukov V, et al. VEGF-C receptor binding and pat-tern of expression with VEGFR-3 suggests a role inlymphatic vascular development. Development 1996;122:3829–3837.

15. Lee J, Gray A, Yuan J, Luoh SM, Avraham H, WoodWI. Vascular endothelial growth factor-related pro-tein: a ligand and specific activator of the tyrosine ki-nase receptor Flt4. Proc Natl Acad Sci USA1996;93:1988–1992.

16. Hung CJ, Ginzinger DG, Zarnegar R, Kanauchi H,Wong MG, Kebebew E, et al. Expression of vascularendothelial growth factor-C in benign and malignantthyroid tumors. J Clin Endocrinol Metab 2003;88:3694–3699.

17. Andre T, Kotelevets L, Vaillant JC, Coudray AM, We-ber L, Prevot S, et al. Vegf, Vegf-B, Vegf-C and theirreceptors KDR, FLT-1 and FLT-4 during the neoplas-tic progression of human colonic mucosa. Int J Can-cer 2000;86:174–181.

18. Kawakami M, Furuhata T, Kimura Y, Yamaguchi K,Hata F, Sasaki K, et al. Quantification of vascular en-dothelial growth factor-C and its receptor-3 messen-ger RNA with real-time quantitative polymerasechain reaction as a predictor of lymph node metasta-sis in human colorectal cancer. Surgery 2003;133:300–308.

19. Akagi K, Ikeda Y, Miyazaki M, Abe T, Kinoshita J,Maehara Y, et al. Vascular endothelial growth factor-C (VEGF-C) expression in human colorectal cancertissues. Br J Cancer 2000;83:887–891.

20. Maeda K, Yashiro M, Nishihara T, Nishiguchi Y,Sawai M, Uchima K, et al. Correlation between vas-cular endothelial growth factor C expression andlymph node metastasis in T1 carcinoma of the colonand rectum. Surg Today 2003;33:736–739.

21. Lagercrantz J, Farnebo F, Larsson C, Tvrdik T, WeberG, Piehl F. A comparative study of the expression pat-terns for vegf, vegf-b/vrf and vegf-c in the develop-ing and adult mouse. Biochim Biophys Acta 1998;1398:157–163.

22. Paavonen K, Mandelin J, Partanen T, Jussila L, Li TF,Ristimaki A, et al. Vascular endothelial growth fac-tors C and D and their VEGFR-2 and 3 receptors in

blood and lymphatic vessels in healthy and arthriticsynovium. J Rheumatol 2002;29:39–45.

23. Partanen TA, Arola J, Saaristo A, Jussila L, Ora A, Mi-ettinen M, et al. VEGF-C and VEGF-D expression inneuroendocrine cells and their receptor, VEGFR-3, infenestrated blood vessels in human tissues. FASEB J2000;14:2087–2096.

24. Bunone G, Vigneri P, Mariani L, Buto S, Collini P, Pi-lotti S, et al. Expression of angiogenesis stimulatorsand inhibitors in human thyroid tumors and correla-tion with clinical pathological features. Am J Pathol1999;155:1967–1976.

25. Hoar FJ, Chaudhri S, Wadley MS, Stonelake PS. Co-expression of vascular endothelial growth factor C(VEGF-C) and c-erbB2 in human breast carcinoma.Eur J Cancer 2003;39:1698–1703.

26. Valtola R, Salven P, Heikkila P, Taipale J, Joensuu H,Rehn M, et al. VEGFR-3 and its ligand VEGF-C areassociated with angiogenesis in breast cancer. Am JPathol 1999;154:1381–1390.

27. Clark DE, Smith SK, Licence D, Evans AL, Charnock-Jones DS. Comparison of expression patterns for pla-centa growth factor, vascular endothelial growth fac-tor (VEGF), VEGF-B and VEGF-C in the humanplacenta throughout gestation. J Endocrinol 1998;159:459–467.

28. Zhou Y, McMaster M, Woo K, Janatpour M, Perry J,Karpanen T, et al. Vascular endothelial growth factorligands and receptors that regulate human cytotro-phoblast survival are dysregulated in severe pre-eclampsia and hemolysis, elevated liver enzymes, andlow platelets syndrome. Am J Pathol 2002;160:1405–1423.

Address reprint requests to:Dr. David Bates

Microvascular Research LaboratoriesDepartment of Physiology

Preclinical Veterinary School, University of Bristol

Southwell StreetBristol, BS2 8EJ, United Kingdom

E-mail: Dave.Bates@bris.ac.uk

JOORY ET AL.82

6265_02_p73-82 6/16/06 2:51 PM Page 82