glyc2

J. Biochem. Biophys. Methods 65 (2005) 121–136

www.elsevier.com/locate/jbbm

Two analytical methods to study the interaction of AGEs

with cell surface proteins

Annett Schmitt a,*,1, Ina Meiners a,e,1, Johannes Schmitt b,

Joachim Noller b, Christian Ihling c, Gerald Munch d,

Andrea Sinz c, Karen Nieber e

a Interdisciplinary Centre of Clinical Research (IZKF) at the Faculty of Medicine at the University of Leipzig,

Inselstr. 22, 04103 Leipzig, Germanyb NIMBUS Biotechnologie GmbH, Eilenburger Str. 4, 04317 Leipzig, Germany

c Biotechnological–Biomedical Center, Faculty of Chemistry and Mineralogy, Linnestr. 3, 04103 Leipzig, Germanyd Comparative Genomics Center, James Cook University, Townsville, Australia

e Institute of Pharmacy, Department of Pharmacology for Natural Sciences, University of Leipzig, Germany

Received 5 April 2005; received in revised form 31 October 2005; accepted 31 October 2005

Abstract

Advanced glycation end products (AGEs) are sugar-modified proteins that are known to appear in vivo

and are suspected to be involved in the pathogenesis of several diseases. Although different cellular

responses to AGEs can be measured in cell culture studies, knowledge about the nature of AGE-binding

and their cell surface receptors is poor. In the present paper a method for the purification of AGE-binding

proteins from membrane fractions derived from different rat organs as well as a method for assaying the

binding of fluorescein labelled AGEs to the surface of cells of different cell lines are described. The

presence of more than 10 proteins interacting with AGEs could be shown in membrane fractions obtained

from rat organs. Additionally, binding of AGE-modified BSA to different cells could be shown using

fluorescence-labelled ligands in a flow cytometric approach. The presented methods provide an option to

isolate AGE-interacting proteins which is a precondition for the identification of these proteins.

Furthermore, the measurement of AGE-binding to cell surfaces bears the potential to gain a deeper

0165-022X/$ - see front matter D 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.jbbm.2005.10.006

Abbreviations: AGEs, Advanced glycation end products; BSA, Bovine serum albumin; CEA, Chicken egg albumin

DMEM, Dubelco’s modified eagle’s medium; 2,4-DNPH, 2,4-Dinitrophenyl hydrazine; ESI, Electrospray ionization

HSA, Human serum albumin; FCS, Foetal calf serum; FITC, Fluorescein isothiocyanate; FTICR-MS, Fourier transform

ion cyclotron resonance mass spectrometry; RAGE, Receptor for AGEs; RFU, Relative fluorescence units.

* Corresponding author. Tel.: +49 341 9715894; fax: +49 341 9715979.

E-mail address: [email protected] (A. Schmitt).1 Both authors contributed equally to this work.

;

;

A. Schmitt et al. / J. Biochem. Biophys. Methods 65 (2005) 121–136122

understanding about the nature of AGE-binding to cell surface proteins and might be applied as a

preliminary test before performing cell culture studies about AGE effects.

D 2005 Elsevier B.V. All rights reserved.

Keywords: AGE-receptors; RAGE; AGE-binding; AGE affinity chromatography; Flow cytometry

1. Introduction

Advanced glycation end products (AGEs) arise from the nonenzymatic reaction of reducing

sugars with proteins, namely their amino acid side chains, in a complex multi step reaction. In

vivo, numerous proteins, particularly long-lived proteins are found to be AGE-modified. AGEs

are signals for cellular activation associated with chemotaxis, oxidative stress and cell

proliferation or programmed cell death [1] and are suspected to be involved in the pathogenesis

of several diseases such as chronic clinical complications of diabetes [2], renal failure [2] and

Alzheimer’s disease [3]. Most investigations are dealing with the effects of AGEs mediated by

the receptor for AGEs (RAGE) or the binding of AGEs to RAGE. RAGE is a transmembrane

protein of the immunoglobuline superfamily with a molecular mass of 45 kDa that was

demonstrated to be present in several cell types such as endothelial cells, monocytes, microglia,

and neurons [4]. AGE binding to RAGE was described to cause an activation of p21ras, MAP

kinases and NF-nB [4]. But there are also several other proteins identified that were found to

bind AGEs (e.g. [4–6]). The function of AGE-binding to most of those proteins is largely

unknown but some of them are likely to mediate cellular effects of AGEs (e.g. [1,7]). Therefore,

not only binding to RAGE but also binding to other proteins is likely to be involved in

transducing cellular effects of AGEs. Hence, dealing with biochemical effects of AGE is a very

complex subject since a large number of different AGE structures are potential ligands for a

large number of binding proteins, of which only a small number is identified. Thus, proper

cause-effect relations are nearly impossible to follow when investigating biochemical effects

of AGEs on cells. Due to the fact that the function of many of the AGE-binding proteins is

largely unknown and that there are probably other proteins yet to be identified, researchers are

additionally faced with the problem to find appropriate cellular parameters, which are

activated in the presence of AGEs. To gain a deeper understanding of AGE-effects on cells it

is an essential task to identify the proteins that are able to interact with AGEs and to

investigate their cellular function. When dealing with effects of AGEs on cells, it might be

advantageous to measure AGE-binding to cell surface proteins before starting to search for

their biochemical effects, since cell binding can be assumed to be a precondition for mediating

cellular effects.

In the present study we describe a method for the isolation of AGE-binding proteins from

crude membrane fractions of different tissues, as well as a flow cytometric approach for the

detection of AGE-binding to cell surfaces. To prove the presence of AGE-binding proteins,

membrane fractions from different rat organs were isolated and an affinity column loaded with

AGE-modified albumin was used to isolate AGE-binding proteins. The presence of several

AGE-binding proteins was demonstrated by SDS-PAGE analysis of fractions eluted from the

affinity column. Additionally, we established a flow cytometric assay in order to measure the

binding of modified proteins to cell surface proteins. Binding experiments were performed

with a glucose derived AGE-modified BSA. AGE-BSA was expected to bind to the cell

surfaces, since similar AGEs were found to stimulate cellular effects in different experiments

A. Schmitt et al. / J. Biochem. Biophys. Methods 65 (2005) 121–136 123

(e.g. [8,9]). Our results show that the binding of modified BSA to cell surfaces can be detected

using a flow cytometric approach. Some binding characteristics (e.g. time and concentration

dependence) were studied using the binding of AGE-modified BSA to CaCo-2 cells.

2. Materials and methods

2.1. Chemicals

Chemicals were purchased from the following suppliers: Sigma/Aldrich, Germany: HSA (96–99%,

fraction V, Cat.No.: A 1653), BSA (99%, gamma globulin-free, Cat.No.: A 3059), CEA (grade VII Cat.No.:

A 7641), FITC, EDTA, glyoxal, methyl glyoxal, malonaldehyde bis(dimethylacetal), and Triton X-114;

BIO-RAD, Germany: Affi-Gel 15; Roth, Germany: Roti-Block, sucrose, Tween 20 and components for PBS,

MOPS, and Tris buffer; Pierce, Germany: ECL-kit and BCA reagent 1 and 2; DAKO, Denmark: anti-mouse

antibody-horseradish peroxidase conjugate; Fluka, Germany: 2,4-dinitrophenyl hydrazine (2,4-DNPH),

9,10-phenanthrenequinone, and Thioflavin T; Boehringer Mannheim, Germany: phenylmethylsulfonyl

fluoride (PMSF); Amersham Biosciences, Germany: Rainbow marker 756; Merck, Germany: glucose

monohydrate (Glc), and glyoxylic acid monohydrate; Roche Diagnostics, Germany: Trypsin (sequencing

grade).

Cells were obtained from DSMZ, Germany except for N11 murine microglia, that was a gift from Dr.

Paola Ricciardi-Castagnoli (Centre of Cytopharmacology, Milan, Italy). Dulbecco’s modified Eagle’s

medium (DMEM) and all other components for the cell culture media were purchased from GibcoBRL,

Germany. Accutase was from PAA Laboratories, Austria.

2.2. Isolation of membrane fractions from different rat organs

Animals used in this study were treated in strict accordance to the German law of animal protection § 4/

3 for the care and use of laboratory animals.

Lungs, livers and kidneys were obtained from male Wistar rats. Organs were rapidly removed from

anesthetized and decapitated animals and stored at �20 8C until use.

To disintegrate the cells, organs were mechanically disrupted and homogenized in a Potter homogeniser

in 5 mM Tris/HCl, pH 7.4 containing 300 mM sucrose, 0.1 mM EDTA and 10 AM PMSF on ice.

Membrane fractions were obtained by a three step differential centrifugation. All centrifugation steps were

performed at 4 8C. Nuclei and large fragments were separated by a 15 min centrifugation at 1000 �g.

Mitochondria and small fragments were removed by centrifugating the supernatant of the first step for

30 min at 7000 �g. Finally, the obtained supernatant was centrifuged for 1 h at 20000 �g to separate the

membranes from the soluble fraction. The membrane pellet was resuspended in 5 mM Tris/HCl, pH 7.4

containing 300 mM sucrose, 0.1 mM EDTA and 10 AM PMSF and 1% Triton X-114 [10]. The solution was

centrifuged at 7000 �g for 15 min to remove insoluble components. The supernatant was used for affinity

chromatography. The detection of RAGE (see SDS-PAGE and Western blot paragraphs) in the obtained

fraction was used to prove the presence of membrane proteins.

2.3. Preparation of the AGE-column chromatography

AGEs for the affinity column were prepared by incubating 50 mg/ml HSA or 50 mg/ml CEA in

400 mM phosphate buffer, pH 8.0 using 1 M glucose (Glc), 10 mM methyl glyoxal (MG) or a mixture of

5 mM glyoxal (GO) and 5 mM malonaldehyde bis(dimethylacetal) (MDA) for 8 weeks at 50 8C under

sterile conditions. After the incubation the AGE-solutions were extensively dialysed against 100 mM

MOPS, pH 7.5. Protein concentrations were determined using the BCA assay according to the suppliers

protocol.

The AGE-affinity resin was prepared by incubating 20 ml Affi-Gel 15 with 30 ml AGE solution

containing 5.5 mg/ml of each AGE for 24 h. Coupling and washing procedures were performed at 4 8C


according to the supplier’s protocol. A final protein concentration of 30 mg/ml gel was determined for the

AGE-affinity column.

2.4. AGE-affinity chromatography

The chromatography was performed at 4 8C. 15 ml of the AGE-affinity gel was packed in a column

(17�100 mm) and equilibrated with 50 ml 10 mM Tris/HCl, pH 7.8. The membrane fractions obtained

from differential centrifugation were applied to the column (approximately 20 ml) and unbound proteins

were eluted with the same buffer. Bound proteins were eluted in two steps using 0.1 M NaCl and 0.2 M

NaCl in equilibration buffer. The column was finally washed with 1.5 M NaCl. 1.5 ml fractions were

collected during elution. The elution profile was monitored by detecting the absorbance at 280 nm. As no

pump was used for the chromatography the flow rate was determined by gravity.

2.5. SDS-PAGE and enzymatic proteolysis

SDS-PAGE was performed according to the procedure of Laemmli [11] using 12% acrylamide gels in a

Mini-PROTEAN 3 chamber (BIO-RAD). Rainbow marker 756 was used as molecular weight marker. Gels

were stained with Coomassie brilliant blue or used unstained for Western blot. Selected bands were excised

from the gels and in-gel digested as described previously [12]. Briefly, 5-10 Al of trypsin solution (50 ng/

AL) was added to the desiccated samples (depending on the volume of the gel pieces) and the digests were

incubated at 37 8C for 16 h. The resulting protein fragments were eluted using a mixture of acetonitrile,

water, and formic acid (47.5%, 47.5%, 5%) and the resulting solution was concentrated to a final volume of

5–10 Al.

2.6. Identification of proteins using nano-HPLC/nano-ESI-FTICR mass spectrometry

Peptide mixtures from enzymatic digests were analysed by nano-HPLC/nano-ESI-FTICR (electrospray

ionization Fourier transform ion cyclotron resonance) mass spectrometry as described previously [13].

Database searches were performed using the Mascot software (http://www.matrixscience.com) [14].

2.7. Detection of RAGE using western blot-analysis

Proteins from unstained SDS-PAGE gels were blotted onto a nitrocellulose membrane for 90 min at

150 A in a Semi-dry blotting system (C.B.S. Scientific. Co) using a buffer containing 25 mM Tris and 190

mM glycine, pH 8.3 as transfer buffer. Following a 1 h blocking with 10% Roti-Block the membrane was

incubated for 1 h with a monoclonal mouse-anti-RAGE antibody (kindly provided by Th. Henle, Dresden,

Germany) in Tris buffered saline (TBS) containing 10% Roti-Block. The primary antibody was removed by

washing the membrane four times with TBS. Afterwards the blot was incubated for 1 h with a monoclonal

anti-mouse IgG conjugated with horseradish peroxidase in TBS containing 10% Roti-Block. Unbound

antibody was removed by washing the membranes four times with TBS containing 0.05% Tween 20.

Bound antibody was detected using the ECL-assay (Amersham Biosciences) according to the supplier’s

protocol.

2.8. Preparation of AGE-modified BSA

AGE-modified BSA was prepared by incubating 4 mg/ml BSA in 0.5 M phosphate buffer, pH 8.0

containing 1 mM NaN3 and 1 mM EDTAwith 0.5 M glucose for 6 weeks at 37 8C under sterile conditions.

All glassware and the buffer were autoclaved prior to use to inactivate proteases. BSA incubated under the

same conditions but without glucose was used as control. After the incubation, protein solutions were

extensively dialysed against 10 mM phosphate buffer, pH 7.4 and stored at �20 8C. Protein concentrations

were determined using the BCA assay according to the suppliers protocol.

http://www.matrixscience.com


2.9. Partial characterization of AGE-modified BSA

Characterization procedures were performed as previously described in detail [15]. Briefly, the number

of reacted lysine residues was determined using fluorescamine. Reacted arginine was assayed using 9,10-

phenanthrenequinone as described previously [15,16]. The fibrillar state of modified proteins and the

corresponding controls were determined using Thioflavin T [15,17]. The CML content was determined by

CML-ELISA kit (kindly provided by Roche Diagnosics, Penzberg, Germany) according to the supplier’s

protocol and the carbonyl content was assayed using 2,4-DNPH as previously described [15,18]. AGE-

absorbance was read at 360 nm and AGE-specific fluorescence was detected at excitation/emission

wavelengths of 330/395, 365/440, and 485/530 nm [15]. Additionally, changes in intrinsic protein

fluorescence were detected at excitation/emission wavelength of 280/350 nm [15]. All data are given

normalized to the corresponding control. Freshly prepared BSA solution was used as an additional control

in all cases.

2.10. FITC labelling of samples and determination of the labelling rate

FITC labelling was performed in 100 mM carbonate buffer, pH 9.5 with final protein concentration of

1 mg/ml and final FITC concentration of 0.1 mg/ml. After 1 h incubation at room temperature samples were

extensively dialysed against PBS, aliquoted and stored frozen at �70 8C.The FITC labelling rates were determined by detecting the mass difference between FITC-labelled

proteins and non-labelled proteins using a Bruker Autoflex MALDI-TOF mass spectrometer.

2.11. Cell culture

CaCo-2, SH-SY5Y, N11 and HEK293 cells were used for binding experiments. Cells were grown in

cell culture dishes (145�20 mm, Greiner) in DMEM supplemented with FCS (10% for SH-SY5Y, N11

and HEK293, 20% for CaCo-2), glutamine, non-essential amino acids and penicillin/streptomycin at 37 8C,5% CO2 and 95% humidity until confluence and harvested with Accutase (PAA Laboratories) (5 min for

SH-SY5Y, N11 and HEK293, 20 min for CaCo-2) and counted in a CASY 1 cell counter (Scharfe,

Germany).

2.12. Binding experiments and flow cytometric analysis of surface binding

2�105 cells and 50 Ag of FITC-labelled sample (final concentration 7 AM) were mixed in a total

volume of 100 Al in 0.5 ml tubes and incubated for 30 min at room temperature in the dark. Unlabelled

BSA at a final concentration of 75 mg/ml was used to block non-specific binding. FITC-labelled control for

AGE-BSA and freshly prepared BSAwere used as controls. After binding, samples were transferred to 2 ml

tubes and washed 3 times each with 1.8 ml PBS with intermediate centrifugation (2 min, 2000 rcf). The

washed pellet was resuspended in 1 ml PBS (containing 0.05% BSA), filtered through nylon mesh strainer

caps (BD Biosciences) and analysed in FACSVantage SE (BD Biosciences) with excitation and emission

wavelengths of 488 and 530 nm.

3. Results

3.1. Isolation of AGE-binding proteins from rat organs

Although the receptor for AGEs (RAGE) is the best studied among AGE-binding proteins,

several other proteins were found to specifically bind AGEs [4,5,6]. To prove the presence of

numerous AGE-binding proteins in different rat organs, membrane fractions of rat lung, kidney,

and liver were isolated. Rat was chosen as a model organism because of the extensive

Fig. 1. Image of a SDS-PAGE gel obtained with samples from affinity chromatography of a kidney membrane fraction.

The following samples are shown: lane 1—molecular weight marker, lane 2—kidney membrane fraction as applied to the

column, lane 3—proteins eluted with 0.1 M NaCl, lane 4—proteins eluted with 0.2 M NaCl, and lane 5—wash fraction

obtained from 1.5 M NaCl wash. The arrow in lane 3 marks the band that was later identified to be ezrin.


knowledge about the rat’s genomics and proteomics. Perspectively, this knowledge could allow

the identification of at least some of the isolated proteins. Cultured cells could not be used for the

chromatography because the required amount of biomass was too large to be obtained in

conventional cell culture.

AGE-binding proteins were isolated from membrane fractions of the different rat organs

using their affinity for AGEs. For this purpose an Affi-Gel 15 column was loaded with a

mixture of differently prepared CEA-AGEs and HSA-AGEs. Preparation of the AGEs used

for affinity chromatography was performed for the extended period of 8 weeks at 50 8C to

maximize the AGE-content. A combination of differently prepared AGEs was used to gain

as many differently structured modifications as targets for AGE-binding proteins as

possible.

The AGE-loaded affinity column was used to prove the presence of proteins capable

to bind AGEs. The solubilized membrane proteins from the different rat organs were

Fig. 2. Image of a SDS-PAGE gel obtained with samples from affinity chromatography of liver membrane fraction. The

following samples are shown: lane 1—molecular weight marker, lane 2—liver membrane fraction as applied to the

column, and lane 3—proteins eluted with 0.1 M NaCl.

Fig. 3. Image of a SDS-PAGE gel obtained with samples from affinity chromatography of lung membrane fraction. The

following samples are shown: lane 1—molecular weight marker, lane 2—lung membrane fraction as applied to the

column, lane 3—proteins eluted with 0.1 M NaCl, and lane 4—proteins eluted with 0.2 M NaCl. The arrows in lane 4

mark the bands that were later identified to be chain A of moesin (MW 41,000, upper band)) and lysozyme (MW 17,000

Da, lower band).

Fig. 4. Western blot detection of RAGE in fractions eluted from the AGE affinity column (0.2 M NaCl) from rat kidney

(lane 1) and lung (lane 2). Detection was performed using a mouse anti-RAGE as primary antibody and a HRP-linked

anti-mouse as secondary antibody. Bound antibody was detected using the ECL method.


applied at low ionic strength conditions and unbound proteins were removed by

washing with loading buffer. Proteins that were bound to the column could be eluted

using 0.1 and 0.2 M NaCl. No additional protein was eluted in the following washing

step with 1.5 M NaCl as detected by measuring the absorbance at 280 nm in the wash

fraction.

SDS-PAGE analysis of fractions eluted from the affinity column using NaCl clearly

demonstrates the presence of numerous proteins with at least some affinity to AGEs. In Figs.

1–3 images of Coomassie-stained SDS-PAGE gels are presented, which clearly demonstrate,

that several proteins were bound to the column at low ionic strength and could be eluted

with NaCl. In case of lung and kidney, a portion of bound proteins was eluted using 0.1 M

NaCl while the remaining proteins were detached using 0.2 M NaCl. A different behaviour

was found for the proteins from liver membrane fractions that were completely eluted with

0.1 M NaCl. Only traces of protein have been detected in fractions eluted with 0.2 M NaCl.

The reason for this difference is unknown, but the fact that all AGE-bound proteins were

detached with 0.1 M NaCl suggests that the binding is of no physiological relevance.

Further investigations to clarify the binding behaviour of different proteins are currently in

progress.


However, the results clearly demonstrate the presence of several proteins interacting

with AGEs. For all tissues it is obvious that at least 10 proteins were bound to the

column. The identity of those proteins and the physiological relevance of their interactions

with AGEs are largely unknown at the moment. Identification of some of the protein

bands and more detailed studies on the effects of AGEs towards those proteins are under

investigation.

3.2. Detection of RAGE by western blot

The best known receptor for AGEs (RAGE) can be detected in Western blots using a

specific anti-RAGE antibody. RAGE in blots usually appears in two forms: full length RAGE

with a molecular weight of about 45 kDa and soluble RAGE (proteolytic fragment lacking

the transmembrane and the intracellular domain) with a molecular weight of about 35 kDa

[1,4].

In our experiments the presence of RAGE could be proven in fractions obtained from affinity

chromatography of lung and kidney membranes (Fig. 4). RAGE was bound to the affinity

column and was eluted predominantly with 0.2 M NaCl. With protein samples from kidney

small amounts of RAGE were found to be eluted with 0.1 M NaCl but this appeared to be mainly

the soluble form of RAGE (data not shown). The majority of RAGE protein was eluted with

0.2 M NaCl.

Table 1

Matched peptides for the identification of ezrin from rat lung

Peptide [M+H]+calc. [M+H]+exp.

9–27 2081.00 2080.99

28–35 975.54 975.54

41–53 1659.79 1659.79

54–60 847.46 847.46

72–79 988.53 988.53

73–79 860.44 860.43

84–100 2022.98 2022.98

101–107 893.54 893.54

144–151 897.42 897.42

163–171 1203.56 1203.55

172–180 1174.60 1174.60

185–193 1068.52 1068.51

185–193 Oxid 1084.51 1084.51

194–209 1945.95 1945.94

194–209 Oxid 1961.94 1961.93

213–230 1976.03 1976.03

213–233 2334.17 2334.17

238–246 1103.58 1103.57

255–262 958.59 958.59

255–263 1086.68 1086.68

263–273 1309.68 1309.68

264–273 1181.59 1181.58

296–306 1315.68 1315.69

The protein band was excised from the SDS-PAGE and in gel digested with trypsin. The masses of the fragments

obtained from nano-HPLC/nano-ESI-FTICR mass spectrometry were compared to the theoretical fragments of rat

proteins using Mascot.

Table 2

Matched peptides for the identification of moesin (chain A) from rat lung


9–27 2065.00 2065.00

28–35 975.54 975.54

41–53 1659.79 1659.79

73–79 832.47 832.46

82–100 2280.13 2280.12

84–100 2081.00 2080.99

101–107 893.54 893.53

144–156 1416.77 1416.77

163–171 1216.58 1216.58

172–180 1232.61 1232.60

181–193 1535.8 1535.80

181–193 Oxid 1551.8 1551.79

185–193 1078.55 1078.55

194–209 1889.92 1889.92

194–209 Oxid 1905.92 1905.91

238–246 1103.59 1103.58

255–263 1086.68 1086.68

263–273 1309.69 1309.68

264–273 1181.59 1181.59





RAGE could not be detected in the membrane fraction from rat liver or in any of the fractions

eluted from the column in affinity chromatography (data not shown).

3.3. Identification of proteins by nano-HPLC/nano-ESI-FTICR mass spectrometry

Proteins were identified by peptide mass fingerprint analysis of the fragments resulting

from a trypsin digestion using nano-HPLC/nano-ESI-FTICR-MS. The evaluation of the data

was performed by comparing the fragments molecular weights with theoretical trypsin-

fragments of rat proteins. Among the proteins identified from our experimental setup were

ezrin, moesin (chain A) and lysozyme. While ezrin was found in samples derived from rat

kidney (see Fig. 1), moesin and lysozyme were isolated from lung (see Fig. 3). Result

details are presented in Tables 1–3. Although several more proteins were identified, a more

detailed investigation on these proteins is still in progress and the data will be published in

a separate paper.

Table 3

Matched peptides for the identification of lysozyme from rat lung


62–80 2187.98 2187.98

126–132 915.47 915.47

138–144 822.42 822.42





3.4. Characterization of AGE-modified BSA for binding tests

All data are given normalized to the corresponding control or as percent of the control.

The number of reacted lysine residues was assayed using fluorescamine and the degree of

modified arginine was determined using 9,10-phenanthrenequinone. 69% of the lysine and

30% of the arginine residues were found to be modified. The fibrillar state of the proteins was

determined using the unique feature of Thioflavin T to fluoresce brightly when bound to

cross-h-structures in proteins. Compared to the corresponding control the Thioflavin T

fluorescence increased by the factor 8.5 when incubated with AGE-BSA. Therefore, the

modified protein can be assumed to be significantly conformationally affected by the

modification process.

The CML content of the modified protein was determined by CML-ELISA and was found to

be 1.14 mol/mol BSA. It is a known fact, that CML is formed during incubation of proteins with

glucose [19] and is described to be a ligand for RAGE [20].

Carbonyl groups as a representative for irreversible modifications in proteins were

determined using 2,4-DNPH. AGE-BSA was found to have 3.3 mol carbonyl groups per mol

BSA.

AGE-specific absorbance was read at 360 nm. The absorbance for the AGE sample was

found to be 0.4 when normalized to the absorbance at 280 nm of the corresponding control (see

[15] for details).

AGE-specific fluorescence was detected at the excitation/emission wavelengths of 330/

395, 365/440, and 485/530 nm. The fluorescence of AGE-BSA relative to the control was

found to be 36, 35 and 34 at the indicated wavelengths. Changes in intrinsic protein

fluorescence were detected at the excitation/emission wavelengths of 280/350 nm. The

intrinsic protein fluorescence of AGE-BSA was found to be decreased to 12% of the

corresponding control.

3.5. Binding of AGE-modified BSA to cell surfaces

As described above a large number of proteins interacting with AGEs could be demonstrated

in different rat organs. Hence, biological effects of AGEs are probably not only due to binding

to RAGE but might also be mediated by other cell surface proteins. It has been published for

example, that only a partial reduction of AGE-stimulated NO and chemokine release can be

obtained by blocking RAGE [21]. Therefore, the total binding of AGEs to cell surface proteins

and not only the binding to RAGE might be important for mediating AGE-effects on cells. We

developed a flow cytometric assay to measure the binding of AGEs to the surface of intact cells.

In this approach non-specific binding was blocked using a freshly prepared BSA solution.

Therefore, one can assume that all changes in binding behaviour of incubated samples

compared to fresh BSA should be due to changes in the protein during the incubation procedure

and can thus considered to be specific. To compare the binding of modified proteins to different

cell lines binding data are given in normalized form and corrected for the individual labelling

rates. Data were normalized by dividing the detected fluorescence value with the reference

fluorescence obtained from the binding of freshly prepared BSA.

To determine the binding of AGE-BSA to cells, we measured the binding of labelled

modified BSA, the corresponding control and fresh BSA. Each binding was assayed in the

presence of a 150-fold excess of unlabelled, fresh BSA to block non-specific binding. Cells

without FITC-labelled samples were used as a blank for the flow cytometer.


Some characteristics of the AGE-cell interaction were studied using AGE-binding to CaCo-2

cells as a model system. Those cells were chosen since preliminary experiments have shown

high stability of the cells, good binding properties of AGE-BSA, and a high reproducibility of

the data obtained.

The time dependence of AGE-BSA binding was studied by varying the incubation time

between 15 and 120 min. As shown in Fig. 5 the total amount of bound protein was found to be

time dependent. However, when binding of AGE and its control are normalized to the BSA-

binding, as described above, no increase in relative binding is observed. Therefore, prolonged

incubation (at least 15 min) is advantageous to achieve relatively high fluorescence values in

flow cytometric analysis, but does not influence the relative binding of the samples.

The concentration dependence of AGE-BSA binding was tested by varying the concentration

of sample proteins in the range of 0 to 7.5 AM. Samples were incubated for 60 min to achieve

high fluorescence signals. As shown in Fig. 6 binding was found to be concentration dependent

over the whole investigated range. But as already shown for time dependence of binding, relative

binding was found to be nearly unaffected by the concentration of ligand used in the assay.

However, as the total fluorescence was found to be significantly increased by using high ligand

concentrations and the method should be applicable for AGEs with a much lower extent of

modification, the use of modified protein in a concentration rage of 2–7 AM seems to be

reasonable.

The stability of AGE-BSA binding was tested by performing a binding experiment with

increased sample volume and cell number. After incubation cells were washed with PBS,

resuspended in PBS (containing 0.05% BSA), and aliquoted. One aliquot of cell samples was

analysed for bound ligand immediately, the others after 15, 30, 45, 60, and 90 min. Cell

Fig. 5. Time dependence of AGE-binding to CaCo-2 cells. Binding data were obtained by using the standard assay

containing 7 AM labelled sample with varying incubation times. After incubation, cells were washed and analysed

immediately. Data are given as relative fluorescence units (RFU), all corrected for the individual labelling rates.

Fig. 6. Concentration dependence of AGE-binding to CaCo-2 cells. Binding data were obtained by using the standard

assay containing varying concentrations of labelled sample and an incubation time of 60 min. After incubation, cells were

washed and analysed immediately. Data are given as relative fluorescence units (RFU), that had been corrected for the

individual labelling rates.


suspensions were kept on ice and in the dark until use. The detected fluorescence was found

to gradually decrease to 75–80% of the initial value after 60 min of standing and than

remained stable in all samples (data not shown). However, when relative binding is calculated,

Fig. 7. Binding of BSA-AGE and the corresponding control to different cells. The binding is given relative to the value

detected with FITC-labelled BSA.


the same binding was found at all resting times. Hence, it is suggested to analyse all samples

that should be normalized to a certain sample with bound BSA very quickly. Furthermore cells

should be resuspended directly before analysis to reduce the loss of bound sample to a

minimum.

The binding of AGE-BSA to different cell lines was tested using the optimised conditions

obtained in the described studies. The chosen cell lines are routinely used in AGE-studies in our

laboratory and by other researchers.

As shown in Fig. 7 AGE-BSA was found to bind to SH-SY5Y, HEK293 and CaCo-2 cells.

Strongest AGE-binding was observed with SH-SY5Y cells while approximately equal binding

was detected with HEK293 and CaCo-2 cells. Similar results were found for the corresponding

control BSA. No binding of AGE-modified BSA or the control protein was observed with N11

mouse microglia.

4. Discussion

As AGEs are known to be involved in the pathogenesis of several diseases, AGE-effects on

cells are the objective of numerous studies. However, except for the receptor for AGEs (RAGE)

interactions of AGEs with cell surface proteins are poorly understood. Here we describe a

method to isolate proteins using their ability to interact with AGEs and demonstrate the presence

of several AGE-binding proteins in the membrane fractions of different rat organs. Additionally,

we provide a flow cytometric method for the detection of AGE-binding to cells that should be a

useful tool to gain a deeper understanding about AGE-binding to cell surface proteins.

To investigate the presence of AGE-binding proteins in different rat organs we used their

affinity to immobilized AGEs in a chromatographic system. The relatively crude solubilized

membrane fraction was used as a sample for the affinity purification to make sure that no

potential AGE-binding proteins are removed in additional purification steps. Furthermore,

albumin binding proteins were not removed in an additional purification step with an albumin

coated affinity-column as described by Yang et al. [10]. Since many proteins tend to have more

than one binding site, a protein that is able to bind HSA might also have an AGE-binding site.

Consequently we did not remove HSA-binding proteins before the AGE-affinity chromatog-

raphy and are currently working on an assay that allows to discriminate between HSA-binding

and AGE-binding. This should be applied to assay the affinity of the identified proteins to

different AGEs and HSA.

In our chromatography system NaCl was used to elute bound proteins from the AGE-affinity

column. This was done because only a small number of AGE structures are chemically

characterized and even less are available as modified free amino acids and consequently the

elution of affinity bound protein by using a known ligand is not possible in this application.

Although the use of NaCl is a rather unusual method to detach affinity bound proteins from a

column, it was possible to elute RAGE by this method. Therefore, the use of 0.2 M NaCl seems

to be sufficient to remove AGE-bound proteins from the column. The results of SDS-PAGE

analysis of the eluted protein samples clearly demonstrate the presence of several AGE-binding

proteins in the rat organs studied. Although the applied protein detection assay does not allow

conclusions about the total number of AGE-binding proteins (poor resolution on mini-gels, low

sensitivity of Coomassie stain) it is clear that a minimum of 10 proteins, which are able to

interact with AGEs, are present in the membrane fractions of each of the rat organs studied. The

presence of the known receptor for AGEs (RAGE) could be demonstrated in rat lung and kidney

but not in liver.


Peptide mass fingerprint analysis using nano-HPLC/nano-ESI-FTICR-MS lead to the

identification of ezrin in samples from rat kidney, whereas moesin and lysozyme were found

in lung samples. The N-terminal fragments of ezrin and moesin from rat kidney were previously

found to bind AGEs [22]. The binding of ezrin and moesin to BSA-AGE was shown to inhibit

the ability of LLC-PK1 cells to produce tubules [22]. Lysozyme was also formerly described to

be an AGE-binding protein (e.g. [1,23]). The binding of BSA-AGE to lysozyme was found to

inhibit its enzymatic activity [1,23] and might therefore contribute to the development of

diabetes related infections.

Continuative studies on other eluted proteins shall lead to the identification of the proteins

and more detailed investigations on the nature of their interaction with AGEs and their

physiological relevance.

The second part of the study deals with the development of a flow cytometric method to assay

AGE-binding to intact cells. Some properties that might be important for the application of the

binding assay were studied in more detail using the binding of AGE-BSA to CaCo-2 cells. The

AGE-binding to CaCo-2 cells was shown to be time and concentration dependent. According to

the obtained results an incubation time of 30 min and a sample concentration of 7 AM are

recommended to be used in this assay.

Four different cell lines were assayed for binding of a glucose derived AGE-BSA to their cell

surfaces. AGE-BSA was found to bind to the surface of HEK293, SH-SY5Y and CaCo-2 cells

with the exception that no AGE-binding to mouse microglia (N11) was observed. This is an

unexpected result because AGE-BSA was shown to have carboxymethyl lysine modifications,

that are described to be ligands for RAGE and this receptor was demonstrated to be present in

N11-cells [21]. This effect might be caused by other modifications in AGE-BSA, for instance

fructosamine or carbonyl residues, that might prevent the contact of CML and a potential

receptor due to steric or electrostatic effects.

The binding of FITC-labelled non-incubated, non-modified BSA to the cell surface was

used as a reference in all tests. The fact that even fresh BSA was found to bind to cells to

some extent was previously discussed in the literature [24,25]. This binding is thought to be

due to some conformational changes in BSA proteins that are in contact with any kind of

surfaces [24]. Alternatively the presence of an albumin receptor was suggested, that mediates

the uptake of albumin bound substances into the cell [25]. In our experimental set-up we used

FITC-labelled BSA and an excess of unlabelled BSA to measure the non-specific binding to

the cells. The data from binding tests with FITC-labelled AGE and its control were compared

to the binding of non-incubated BSA. The detected binding difference from AGE-modified

BSA to non-modified BSA should be due to changes in the BSA protein caused by

modifications during the incubation process and is most likely a result of protein–protein

interactions. Binding of AGEs to lipid surfaces was found to be negligible in different

experiments (data to be published). The flow cytometric method described here enables the

measurement of the AGE-binding to the surface proteins of different cells. Therefore it

provides access to a parameter that can be considered to be a precondition for provoking

AGE-mediated cellular effects.

5. Simplified description of the methods and their application

There are two methods described in the current article. The first one allows the isolation of

AGE-binding proteins using their ability to interact with AGEs from a relatively crude

membrane extract in a chromatographic set-up. This method provides a basic tool for the


isolation and identification of AGE-target proteins. The identification of probable AGE-binding

proteins could increase the knowledge about AGE-receptor interactions and the potential

signalling pathways of AGE-induced cellular effects.

The second method was designed to measure the binding of AGE-modified proteins to cell

surfaces using fluorescence labelled AGE-proteins in a flow cytometric approach. This binding

is considered to be a precondition for mediating cellular effects. Hence, the detection of AGE-

cell binding could be used as a preliminary test before searching for cellular effects, that are

often difficult to measure.

Therefore, both methods should be very interesting for researchers working in the field of

AGE-effects on cells. Knowledge about the identity of AGE-binding proteins would most likely

allow to block the AGE-binding by using specific antibodies raised against those proteins, as it

was previously described for RAGE [8]. The use of pharmaceuticals to block binding to specific

proteins would also be possible. Hence, the flow cytometric method described here combined

with results based on the affinity purification of AGE-binding proteins are a promising set-up to

gain a deeper understanding on AGE-receptor interactions.

Acknowledgements

This work was supported by the Interdisziplinares Zentrum fur Klinische Forschung at the

Faculty of Medicine of the University of Leipzig (01KS9504, Project N1), the EU (QLK6-CT-

1999-02112), and the Sachsisches Ministerium fur Forschung und Bildung. The authors thank

Roche Diagnostics GmbH for providing CML-ELISA kits and the group of Dr. Peter Ahnert,

especially Holger Kirsten, at the MBZ, Leipzig for providing the opportunity to use the MALDI-

TOF MS. Additionally we are grateful to Jovana Gasic-Milenkovic for revising the manuscript,

Gabriele Oehme for technical assistance, and Viola Dobel for the FACS operating service.

References

[1] Thornalley PJ. Cell activation by glycated proteins. AGE receptors, receptor recognition factors and functional

classification of AGEs. Cell Mol Biol (Noisy-le-grand) 1998;44:1013–23.

[2] Bucala R, Makita Z, Vega G, Grundy S, Koschinsky T, Cerami A, et al. Modification of low density lipoprotein by

advanced glycation end products contributes to the dyslipidemia of diabetes and renal insufficiency. Proc Natl Acad

Sci U S A 1994;91:9441–5.

[3] Vitek MP, Bhattacharya K, Glendening JM, Stopa E, Vlassara H, Bucala R, et al. Advanced glycation end

products contribute to amyloidosis in Alzheimer disease. Proc Natl Acad Sci U S A 1994; 91:4766–70.

[4] Thornalley PJ. Methods for studying the binding of advanced glycated proteins to receptors for advanced glycation

endproducts (AGE receptors), Methods. Mol Biol 2002;196:49–62.

[5] Stitt AW, He C, Vlassara H. Characterization of the advanced glycation end-product receptor complex in human

vascular endothelial cells. Biochem Biophys Res Commun 1999;256:549–56.

[6] Tamura Y, Adachi H, Osuga J, Ohashi K, Yahagi N, Sekiya M, et al. FEEL-1 and FEEL-2 are endocytic receptors

for advanced glycation end products. J Biol Chem 2003;278:12613–7.

[7] McRobert EA, Gallicchio M, Jerums G, Cooper ME, Bach LA. The amino-terminal domains of the ezrin, radixin,

and moesin (ERM) proteins bind advanced glycation end products, an interaction that may play a role in the

development of diabetic complications. J Biol Chem 2003;278:25783–9.

[8] Collison KS, Parhar RS, Saleh SS, Meyer BF, Kwaasi AA, Hammami MM, et al. RAGE-mediated neutrophil

dysfunction is evoked by advanced glycation end products (AGEs). J Leukoc Biol 2002;71:433–44.

[9] Huttunen HJ, Fages C, Rauvala H. Receptor for advanced glycation end products (RAGE)-mediated neurite

outgrowth and activation of NF-kappaB require the cytoplasmic domain of the receptor but different downstream

signaling pathways. J Biol Chem 1999;274:19919–24.


[10] Yang Z, Makita Z, Horii Y, Brunelle S, Cerami A, Sehajpal P, et al. Two novel rat liver membrane proteins that bind

advanced glycosylation endproducts: relationship to macrophage receptor for glucose-modified proteins. J Exp Med

1991;174:515–24.

[11] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature

1970;227:680–5.

[12] Jensen ON, Shevchenko A, Mann M. In: Creighton TE, editor. Protein Structure, A Practical Approach. Oxford7

Oxford University Press, 1997. p. 48.

[13] Ihling C, Berger K, Hofliger MM, Fuhrer D, Beck-Sickinger AG, Sinz A. Nano-high-performance liquid

chromatography combined with nano-electrospray ionization Fourier transform ion-cyclotron resonance mass

spectrometry for proteome analysis. Rapid Commun Mass Spectrom 2003;17:1240–6.

[14] Perkins DN, Pappin DJ, Creasy DM, Cottrell JS. Probability-based protein identification by searching sequence

databases using mass spectrometry data. Electrophoresis 1999;20:3551–67.

[15] Schmitt A, Schmitt J, Munch G, Gasic-Milencovic J. Characterization of advanced glycation end products for

biochemical studies: side chain modifications and fluorescence characteristics. Anal Biochem 2005;338:201–15.

[16] Smith RE, MacQuarrie R. A sensitive fluorometric method for the determination of arginine using 9,10-

phenanthrenequinone. Anal Biochem 1978;90:246–55.

[17] LeVine III H. Thioflavine T interaction with synthetic Alzheimer’s disease beta-amyloid peptides: detection of

amyloid aggregation in solution. Protein Sci 1993;2:404–10.

[18] Valencia JV, Weldon SC, Quinn D, Kiers GH, DeGroot J, TeKoppele JM, et al. Advanced glycation end product

ligands for the receptor for advanced glycation end products: biochemical characterization and formation kinetics.

Anal Biochem 2004;324:68–78.

[19] Reddy S, Bichler J, Wells-Knecht KJ, Thorpe SR, Baynes JW. N epsilon-(carboxymethyl)lysine is a dominant

advanced glycation end product (AGE) antigen in tissue proteins. Biochemistry 1995;34:10872–8.

[20] Kislinger T, Fu C, Huber B, Qu W, Taguchi A, Du Yan S, et al. N(epsilon)-(carboxymethyl)lysine adducts of

proteins are ligands for receptor for advanced glycation end products that activate cell signaling pathways and

modulate gene expression. J Biol Chem 1999;274:31740–9.

[21] Dukic-Stefanovic S, Gasic-Milenkovic J, Deuther-Conrad W, Munch G. Signal transduction pathways in mouse

microglia N11 cells activated by advanced glycation endproducts (AGEs). J Neurochem 2003;87:44–55.

[22] McRobert EA, Gallicchio M, Jerums G, Cooper ME, Bach LA. The amino-terminal domains of the ezrin, radixin,

and moesin (ERM) proteins bind advanced glycation end products, an interaction that may play a role in the

development of diabetic complications. J Biol Chem 2003;278:25783–9.

[23] Li YM. Glycation ligand binding motif in lactoferrin. Implications in diabetic infection. Adv Exp Med Biol

1998;443:57–63.

[24] Reed RG, Burrington CM. The albumin receptor effect may be due to a surface-induced conformational change in

albumin. Biol Chem 1989;264:9867–72.

[25] Ockner RK, Weisiger RA, Gollan JL. Hepatic uptake of albumin-bound substances: albumin receptor concept. Am J

Physiol 1983;245:G13–8.

Documents

glyc2