endothelial cell adhesion molecule and pmnl response to inflammatory stimuli and age-modified...

Endothelial cell adhesion molecule and PMNL response toinflammatory stimuli and AGE-modified fibronectin

GURKAN SENGOELGE, MANUELA FODINGER, SONJA SKOUPY, ILSE FERRARA, CHRISTA ZANGERLE,MICHAEL ROGY, WALTER H. HORL, GERE SUNDER-PLASSMANN, and JOHANNES MENZEL

Department of Internal Medicine III, Division of Nephrology and Dialysis; Department of Laboratory Medicine, Division of MolecularBiology; and Department of Surgery, Institute of Immunology, University of Vienna, Vienna, Austria

Endothelial cell adhesion molecule and PMNL response toinflammatory stimuli and AGE-modified fibronectin.

Background. Atherosclerotic vascular disease is the leadingcause of death in patients with diabetes mellitus and end-stagerenal disease. Advanced glycation end products (AGEs) arestrongly suggested to be involved in the pathogenesis of athero-sclerosis in these patients who also frequently experience infec-tious complications. We hypothesized that the interaction ofAGEs and inflammatory mediators contributes to the up-regula-tion of endothelial cell activation.

Methods. We investigated the effect of advanced glycatedfibronectin in the presence or absence of inflammatory stimuli onthe endothelial cell surface and mRNA expression of cell adhe-sion molecules. Furthermore, the influence of advanced glycatedfibronectin on the transendothelial migration pattern of polymor-phonuclear cells was analyzed.

Results. Exposure to advanced glycated fibronectin togetherwith inflammatory stimuli such as interleukin (IL)-1a, tumornecrosis factor-a (TNF-a) or lipopolysaccharide (LPS) led to asignificant increase in the surface expression of the cell adhesionmolecules E-selectin, ICAM-1, VCAM-1 and PECAM-1 on en-dothelial cells. Soluble AGEs in combination with advancedglycated fibronectin significantly enhanced the endothelial cellsurface expression of ICAM-1, VCAM-1 and PECAM-1, whereasthis was not the case for E-selectin. At the transcriptional levelshort-time exposure of endothelial cells to advanced glycatedfibronectin and inflammatory mediators resulted in an increasedexpression of E-selectin, ICAM-1 and VCAM-1 mRNA levels,whereas PECAM-1 repeatedly showed a significant decrease ofgene transcript levels. An increase of mRNA levels was alsoobserved for E-selectin, ICAM-1, VCAM-1 and PECAM-1 fol-lowing incubation with a combination of advanced glycated fi-bronectin and soluble advanced glycation end-products. Further-more, polymorphonuclear cells responded with a sevenfoldincrease in transendothelial migration following exposure ofendothelial cells to advanced glycated fibronectin and inflamma-tory mediators.

Conclusions. These results suggest that the combination ofmatrix glycation and inflammation up-regulates the activation ofthe endothelial cell adhesion cascade, a mechanism that mightcontribute to the increased burden of atherosclerotic morbidityand mortality in patients suffering from diabetes mellitus orchronic renal failure.

The incidence of vascular disease morbidity and mortal-ity is manifold increased in patients with diabetes mellitusand in patients with end-stage renal failure [1–3]. One keyevent in the development of atherosclerosis is the adhesionof leukocytes to the vascular endothelium [4]. Cytokineinducible cellular adhesion molecules (CAM) expressed onthe surface of endothelial cells including E-selectin, inter-cellular CAM-1 (ICAM-1), vascular CAM-1 (VCAM-1),and platelet-endothelial CAM (PECAM) have been shownto participate in the atherosclerotic plaque formation[5–13]. Additionally, advanced glycation end products(AGEs) have been suggested to be involved in the patho-genesis of atherosclerosis, since both circulating and con-nective tissue AGEs have been identified in patients suf-fering from diabetes mellitus or end-stage renal failure [14,15].

AGEs represent a heterogeneous group of metabolicallyaltered substrates that accumulate during long-term hyper-glycemia. Advanced glycation starts with a non-enzymaticcondensation reaction between circulating blood glucoseand amino groups, forming slowly reversible Amadoriproducts. Upon prolonged exposure to elevated plasmaglucose levels, these early glycated products are trans-formed into AGEs that are now irreversibly bound toproteins, lipids and nucleic acids [16]. Thus, AGE-forma-tion alters long-lived molecules in particular, such as extra-cellular matrix components, and increases with age.

One candidate extracellular matrix protein that might bealtered by advanced glycation is fibronectin, a componentof the basal membrane of the blood vessel wall that iscovered by endothelial cells. Endothelial cells are able tointeract with AGEs via specific AGE receptors resulting in

Key words: advanced glycation end products, inflammatory mediators,diabetes, uremia, endothelial cells, polymorphonuclear leukocytes, E-selectin, ICAM-1, VCAM-1, PECAM.

Received for publication March 27, 1998and in revised form May 20, 1998Accepted for publication June 5, 1998

© 1998 by the International Society of Nephrology

Kidney International, Vol. 54 (1998), pp. 1637–1651

1637

an increased vascular permeability and the up-regulation ofendothelial cell surface procoagulants [17]. Furthermore,AGEs quench nitric oxide and thus interfere with endothe-lium-dependent vascular relaxation [18].

Since AGEs are strongly suggested to be involved in thepathogenesis of atherosclerosis in patients with diabetesmellitus and renal failure who also frequently experienceinfectious complications [19, 20], the hypothesis was raisedthat the interaction of AGEs and inflammatory mediatorsenhances the up-regulation of endothelial cell adhesionmolecules and the transendothelial migration of leuko-cytes. This interaction could contribute to the multifactorialacceleration of atherosclerosis in these patients who fre-quently experience episodes of infections.

METHODS

Reagents and antibodies

Fibronectin and endothelial cell growth supplement werepurchased from Collaborative Biomedical Products (Bed-ford, MA, USA). Fetal calf serum, collagenase type I,Hank’s solution, Dulbecco’s phosphate-buffered saline(PBS) and RPMI 1640 medium were obtained from GibcoLaboratories (Grand Island, NY, USA). Ficoll-Paque wasobtained from Pharmacia (Uppsala, Sweden). Bovine se-rum albumin, b2-microglobulin, complete Freund’s adju-vant, fluorescein isothiocyanate (FITC), RNAse and endo-toxin of E. coli serotype 026:B6 were purchased from SigmaChemical Co. (St. Louis, MO, USA). Human IgG was fromBiochemie GmbH (Vienna, Austria). Glucose-6-phosphateand 2,29-Azinobis-3-ethylbenzthiazoline-sulfonic acid(ABTS) were from Boehringer Mannheim (Mannheim,Germany). RNazol-B was obtained from Biotech Lab. Inc.(Houston, TX, USA). Calcein AM was purchased fromMolecular Probes (Eugene, OR, USA). The oligonucleo-tide primers were purchased from Pharmacia (Vienna,Austria) and are summarized in Table 1. Ampli-Taq DNA-

Polymerase was from Perkin Elmer Cetus (Norwalk, CT,USA). The 6% polyacrylamide mini gels were from Novex(San Diego, CA, USA). The limulus assay (Coatest Endo-toxin) was obtained from Boehringer Ingelheim (In-gelheim, Germany).

The antibodies 5.6E (PECAM), 84H10 (ICAM-1), 1.2B6(E-selectin), 1G2 (endoglin), 1G11 (VCAM-1) and 4F9(von Willebrand factor) were obtained from Immunotech(Marseille Cedex, France). Interleukin-1a (IL-1a), tumornecrosis factor-a (TNF-a) and blocking antibodies BBIG-I1 (ICAM-1), BBIG-E4 (E-selectin) and BBIG-V1(VCAM-1) were supplied by R&D Systems (Minneapolis,MN, USA). FITC labeled goat F(ab)2 anti-mouseIgG1IgM and non-specific anti-IgG11IgG2a antibodieswere from An der Grub (Vienna, Austria).

Preparation of advanced glycated fibronectin

Advanced glycated fibronectin (AGE-fibronectin) wasprepared using fibronectin (1 mg/ml) in a reaction mixturecontaining 5 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.5 M

glucose-6-phosphate (G-6-P) and proteinase inhibitors (1mM phenyl-methyl-sulfonyl-fluoride, 5 mM EDTA (pH 8.0),trypsin inhibitor 50 mg/ml, aprotinin 2 mg/ml, 1 mM amin-ocaproic acid and 0.02% sodium-azide). Native fibronectin,which served as a control, was incubated under identicalconditions without G-6-P. Using Limulus assay, the lipo-polysaccharide (LPS) level in the AGE-fibronectin stocksolution was determined to be , 0.6 ng/ml, which wasfurther diluted (1:40). Only AGE-fibronectin without de-tectable LPS was used in all experiments. Quality controlsto assess formation of AGE-fibronectin included the mea-surement of the fluorescence intensity (at 370 nm and 440nm, respectively) using a spectrofluorometer (ShimadzuRF-551; the specific fluorescence intensity of AGE-fi-bronectin used in all experiments was 15000 RFUs, nativefibronectin showed no fluorescence) and the analysis of the

Table 1. Relative surface and mRNA expression of E-selectin, ICAM-1, VCAM-1 and PECAM cultured on native fibronectin (N-FN) or AGE-fibronectin (AGE-FN) and stimulated with IL-1a, TNF-a, LPS or AGE-albumin (sAGE) following different times of incubation (in comparison to

culture of cells on native fibronectin in medium; n.d., not done).

Expression of Location hours N-FN1IL-1a N-FN1TNFa N-FN1LPS N-FN1sAGE AGE-FN-Co AGE-FN1IL-1a

E-selectin surface, 4 hr 380 6 27 229 6 18 366 6 29 112 6 6 130 6 15 1260 6 113surface, 18 hr 184 6 30 124 6 17 123 6 25 80 6 5 91 6 10 302 6 30mRNA, 2.5 hr 292 6 9 n.d. n.d. 317 6 69 237 6 62 570 6 121mRNA, 10 hr 483 6 42 n.d. n.d. 106 6 25 120 6 23 684 6 155

ICAM-1 surface, 4 hr 185 6 14 162 6 14 187 6 17 104 6 2 141 6 5 259 6 29surface, 18 hr 368 6 23 288 6 19 341 6 18 100 6 6 112 6 6 468 6 42mRNA, 2.5 hr 269 6 29 n.d. n.d. 144 6 29 132 6 22 421 6 52mRNA, 10 hr 400 6 24 n.d. n.d. 109 6 15 83 6 11 368 6 18

VCAM-1 surface, 4 hr 556 6 46 536 6 41 611 6 46 124 6 8 167 6 14 1071 6 96surface, 18 hr 963 6 106 757 6 74 776 6 84 119 6 11 130 6 13 1416 6 148mRNA, 2.5 hr 80 6 21 n.d. n.d. 193 6 48 169 6 30 299 6 61mRNA, 10 hr 537 6 34 n.d. n.d. 171 6 22 137 6 21 435 6 49

PECAM surface, 4 hr 102 6 4 101 6 3 94 6 4 102 6 2 122 6 6 119 6 3surface, 18 hr 107 6 5 106 6 5 103 6 5 91 6 3 117 6 3 132 6 10mRNA, 2.5 hr 60 6 10 n.d. n.d. 91 6 19 101 6 15 59 6 13mRNA, 10 hr 46 6 6 n.d. n.d. 91 6 5 91 6 5 40 6 5

Abbreviations are in the Appendix.

Sengoelge et al: AGE-modified fibronectin and inflammation1638

inhibition properties of AGE antibody binding in the solidphase immunoassay (as described below).

Generation of anti-AGE antibodies using AGE-RNAse

AGE-RNAse was prepared by incubating RNAse (25mg/ml) in PBS (pH 7.4) containing 0.5 M G-6-P, 0.1%sodium-azide, 100 U/ml penicillin and 100 mg/ml strepto-mycin for five weeks at 37°C under sterile conditions. Priorto incubation the reaction mixture was passed through aDetoxigel column (Pierce, Rockford, IL, USA) to eliminatecontaminating LPS. At the end of the incubation, unboundG-6-P was removed by dialysis against PBS.

For the preparation of anti-AGE antibodies, white fe-male New Zealand rabbits were immunized with 10 mg/mlor 25 mg/ml of AGE-RNAse emulsified in an equal volumeof complete Freund’s adjuvant. Two hundred and fiftymicroliters were injected subcutaneously every two weeksat two different sites of the back of the animals. A total ofsix injections were given. After the second injection bloodwas collected for the determination of antibody titers priorto each injection. High titers were observed ten weeks afterthe immunization. AGE-antibody titers were determinedby a solid-phase assay as described below.

Advanced glycated end product immunoassay

Advanced glycated albumin (AGE-albumin) was pre-pared using the same conditions as described above forAGE-RNAse, but with a bovine serum albumin concentra-tion of 10% (RFU 2300). Ninety-six-well microtiter plateswere coated with AGE-albumin (50 mg/ml) for one hour at37°C followed by an incubation period of three days at 4°C.After washing with PBS, the plates were incubated with0.4% BSA in PBS for two hours at room temperature toprevent nonspecific binding. The AGE containing sampleswere diluted in 0.4% BSA/PBS (1:10) and were incubated

in glass vials with the rabbit anti-AGE-RNAse antiserum(1:1000) for one hour at 37°C and 16 hours at 4°C. Thesemixtures (0.2 ml) were added to the AGE-albumin coatedwells and incubated for one hour at 37°C followed by a finalincubation step of 30 minutes at 4°C. After thoroughwashing with PBS peroxidase labeled anti-rabbit antibody(1:2000) was added for one hour at 37°C and 30 minutes at4°C. Binding of the second antibody was determined by acolor reaction induced by the addition of ABTS (0.1%solution in diluent buffer) and the measurement of theoptical density at 405 nm. The degree of inhibition ofanti-AGE antibody binding to the solid phase was propor-tional to the AGE concentration of the sample. AGE-modified b2-microglobulin (AGE-b2m, 25 mg/dl), whichwas prepared under the same conditions as described forAGE-RNAse, was used as a reference. Results were ex-pressed in molar equivalents of AGE-b2m in relativefluorescence units (RFU). The molar equivalent of AGE-fibronectin was 25 to 34 mg/ml AGE-b2m, and 17 mg/ml forAGE-albumin. Neither native albumin nor native fibronec-tin showed any inhibition properties of AGE antibodybinding in the solid phase immunoassay.

Endothelial cell culture

Human umbilical cord vein endothelial cells (HUVEC)were cultured according to Jaffe et al [21]. Briefly, cellswere pooled after collagenase treatment and seeded onsix-well cell culture plates coated with native fibronectin(2.5 mg/cm2) or AGE-fibronectin (2.5 mg/cm2). Cells weregrown in RPMI 1640 medium supplemented with 20%FCS, 1% L-glutamine, 100 U/ml penicillin G, 100 mg/mlstreptomycin and 1 mg/ml fungizone and after the first daywith 5% FCS under standard cell culture conditions (hu-midified atmosphere, 5% CO2, 37°C). Reaching confluencewithin three to four days, primary endothelial cell cultureswere used for all flow cytometric analyses and preparationof total RNA. All cells showed cobblestone morphologythat did not differ between cells grown on native fibronectinor AGE-fibronectin and showed uptake of acetylated LDLand expression of von Willebrand factor and angiotensinconverting enzyme. The presence of contaminating CD14positive cells (monocytes) was excluded by flow cytometry.

Immunofluorescence analysis of cellular adhesionmolecule expression

Endothelial cells were incubated in native fibronectin orAGE-fibronectin coated six-well plates in RPMI 1640medium containing either IL-1a (100 U/ml), TNF-a (500U/ml), LPS (1 mg/ml), or AGE-albumin (500 mg/ml) ormedium alone (control) for four hours or 18 hours. Afterincubation confluent HUVEC-monolayers were detachedby scraping with a rubber policeman in ice-cold PBS. Cellswere then washed twice and centrifuged at 300 g for 10minutes. Fifty microliters of the cell suspension (5 3

AGE-FN1TNFa AGE-FN1LPS AGE-FN1sAGE

743 6 73 1218 6 129 124 6 12113 6 15 156 6 18 81 6 8

n.d. n.d. 337 6 45n.d. n.d. 169 6 18

238 6 20 266 6 29 136 6 4384 6 36 418 6 23 143 6 14

n.d. n.d. 131 6 19n.d. n.d. 147 6 29

854 6 86 1140 6 116 166 6 191170 6 139 1129 6 83 197 6 31

n.d. n.d. 226 6 36n.d. n.d. 280 6 56

119 6 6 115 6 4 117 6 6140 6 10 129 6 9 113 6 5

n.d. n.d. 93 6 14n.d. n.d. 99 6 11

Sengoelge et al: AGE-modified fibronectin and inflammation 1639

105/ml) were incubated with 50 ml of the respective mono-clonal antibody (mAb; 5 mg/ml) for 30 minutes at 4°C.After washing, cells were stained with FITC-conjugatedgoat F(ab9)2 anti-mouse IgG1IgM antibodies for 30 min-utes at 4°C. Following two additional washing steps endo-thelial cells were analyzed using a FACScan (Becton Dick-inson, Mountain View, CA, USA) and the LYSIS IIsoftware package (Becton Dickinson). The cell viability ofHUVEC prior to staining procedures was typically . 95%as determined by trypan blue exclusion. Cytofluorimetricanalysis of cellular adhesion molecule cell surface expres-sion was performed in a total of 10 independent cell cultureexperiments. Data are given as mean fluorescence intensityin percent of control expriments.

Semiquantitative reverse transcriptase-polymerase chainreaction analysis of cell adhesion molecules

Human umbilical cord vein endothelial cells were cul-tured on 175 cm2 cell culture flasks coated with nativefibronectin (2.5 mg/cm2) or AGE-fibronectin (2.5 mg/cm2)in the presence of IL-1a (100 U/ml), AGE-albumin (500mg/ml) or medium alone for 2.5 hours or 10 hours. Afterincubation, cells were washed twice in diethylpyrocarbon-ate-treated PBS and 1 3 107 cell aliquots were lysed byaddition of 1.8 ml RNazol-B. Total RNA was extracted asdescribed [22]. The quality of isolated RNA was controlled byelectrophoresis through formaldehyde agarose gels. Highquality RNA was quantitated by measuring absorption at

Fig. 1. Expression of E-selectin. (A) Surfaceexpression (mean fluorescence intensity, mfi) ofE-selectin on HUVEC cultured on nativefibronectin or on AGE-fibronectin after fourhours or 18 hours of incubation with IL-1a,TNF-a, LPS, AGE-albumin (sAGE), ormedium as determined by flow cytometricanalysis. The results are given as percentages ofexpression on non-stimulated HUVEC grownon native fibronectin 6 SEM. P values (**P ,0.01) refer to comparison of equally treatedcells grown on native fibronectin or AGE-fibronectin. (B) Semiquantitative RT-PCRanalysis of E-selectin transcript levels inendothelial cells. Cells were cultured on nativefibronectin or AGE-fibronectin for 2.5 hours or10 hours with IL-1a, AGE-albumin (sAGE) ormedium alone. Results were corrected by theABL values of the respective cDNA. Theresults are given as percentages of expression innon-stimulated HUVEC grown on nativefibronectin 6 SEM. P values (*P , 0.05) referto comparison of equally treated cells grown onnative fibronectin or AGE-fibronectin.


260 nm and 1 mg of total RNA was subjected to cDNAsynthesis as described elsewhere [23].

For semiquantitative analysis of PECAM, ICAM-1, E-selectin and VCAM-1 mRNA, a RT-PCR protocol, whichallows measurement of relative transcript levels was ap-plied [24]. The oligonucleotide primer sequences used forPCR amplification of the respective mRNAs have beenpreviously described [25, 26]. Polymerase chain reactionamplification of ABL transcripts was used as a reference toassess variations of total RNA or cDNA between samplesas described previously [27]. The linear ranges of PCRamplifications were established as a function of cyclenumber and cDNA dilutions for each primer pair [24].Reaction conditions included 3 ml cDNA, 10 to 50 pmol ofeach primer, 2.5 to 3.0 mM MgCl2, 200 mM of each dNTP,

1.25 U Ampli-Taq DNA polymerase and a (32P)-dCTP(150,000 cpm) in a 50 ml reaction volume. The thermalcycling conditions were denaturation at 94°C (1 min),annealing at 57°C 265°C (1 min) and extension at 72°C (1min), preceded by an initial denaturation step at 94°C forfive minutes and followed by a terminal extension of 10minutes at 72°C. The PCR amplification products weresubjected to 6% polyacrylamide gels and dried gels wereexposed to Kodak XAR-5 films at 270°C for 12 hours.

For quantification of PCR products incorporated a(32P)-dCTP was measured on autoradiograms using an imagingdensitometer (model 670; BioRad Laboratories, Hercules,CA, USA) and the system9s volume integration program(BioRad Gel DOC 1000 system, Molecular Analyst/PC soft-ware). Potential differences of total cellular RNA/cDNA in

Fig. 1. Continued.


PCR analyses were corrected by dividing PECAM,ICAM-1, E-selectin and VCAM-1 values by the ABL valueobtained from the respective cDNA. Each relative tran-script level was measured in a total of 12 PCR analysesusing six cDNAs that were synthesized in duplicates fromRNAs of three independent cell culture experiments.

Transendothelial polymorphonuclear leukocyte migration

Isolation of polymorphonuclear leukocyte (PMNL) wasperformed as described previously [28]. Briefly, heparin-ized peripheral blood obtained from healthy volunteers waslayered over Ficoll-Paque (1.077 g/ml). After an initialincubation step of 45 minutes at room temperature thesupernatants were layered on 63% Percoll underlayed with72% Percoll. Following centrifugation at 500 g for 25

minutes at room temperature, cells were washed twice inCa21- and Mg21- free Hank’s solution. The cell pelletswere resuspended in RPMI 1640 medium at a final con-centration of 2.5 3 106 cells/ml. The cell viability in allexperiments was . 95% as determined by trypan blueexclusion.

Human umbilical vein endothelial cells were grown onnative fibronectin or AGE-fibronectin coated permeablemembrane inserts (diameter of 9 mm, pore size of 3.0 mm)of a 24 multiwell double-chamber system and were subse-quently incubated with IL-1a (100 U/ml) or medium (con-trol) for four hours and 18 hours. Reaching confluence,HUVEC monolayers of passages 2 to 4 were incubated withPMNL (5 3 105) in the upper chambers of the permeableinserts for two hours. At the end of the incubation period,

Fig. 2. Expression of ICAM-1. (A) Surfaceexpression (mean fluorescence intensity, mfi) ofICAM-1 on HUVEC cultured on nativefibronectin or on AGE-fibronectin after fourhours or 18 hours of incubation with IL-1a,TNFa, LPS, AGE-albumin (sAGE), or mediumas determined by flow cytometric analysis. Theresults are given as percentages of expressionon non-stimulated HUVEC grown on nativefibronectin 6 SEM. P values (**P , 0.01) referto comparison of equally treated cells grown onnative fibronectin or AGE-fibronectin. (B)Semiquantitative RT-PCR analysis of ICAM-1transcript levels in endothelial cells. Cells werecultured on native fibronectin or AGE-fibronectin for 2.5 hours or 10 hours with IL-1a, AGE-albumin (sAGE) or medium alone.Results were corrected by the ABL values ofthe respective cDNA. The results are given aspercentages of expression in non-stimulatedHUVEC grown on native fibronectin 6 SEM.P values (*P , 0.05) refer to comparison ofequally treated cells grown on native fibronectinor AGE-fibronectin.


the membrane inserts containing the non-migrated leuko-cytes were discarded. Without further washing or transfersteps, migrated leukocytes in the lower chamber wereimmediately exposed to 2 mM of the fluorescent dyecalcein-acetoxymethylester (calcein-AM). Following incu-bation in the dark at room temperature for 30 minutes(mild shaking), the relative fluorescence intensity was mea-sured using the cytofluor 2350 fluorescence plate reader

(Millipore, Bedford, MA, USA). Absolute cell counts weredetermined by comparison with dilution series of cal-cein-AM labeled PMNLs cultured in RPMI 1640 mediumas described [29].

For blocking experiments, HUVEC were pre-incubatedfor one hour with 300 ml of a mixture containing blockingantibodies against PECAM (50 mg/ml), E-selectin (50mg/ml), ICAM-1 (50 mg/ml) and VCAM-1 (60 mg/ml).

Fig. 2. Continued.


Polymorphonuclear leukocytes were pre-treated with hu-man IgG (10 mg/ml) at 37°C for one hour, to saturateFc-gamma receptors and to prevent nonspecific binding toantibodies. Following preincubation, PMNL were coincu-bated with HUVEC on the membrane inserts for twohours. The number of transmigrated PMNL in individualwells was determined as described above. Each experimentwas performed in triplicates for at least three times.

Statistical analyses

All data are presented as means 6 SEM. Statisticalanalysis included ANOVA and Student’s t-test, which wereapplied when appropriate.

RESULTS

AGE-fibronectin up-regulates cytokine inducedexpression of endothelial cell adhesion molecules

The detailed results of flow cytometric analysis of cellsurface expression and of mRNA expression of endothelialcell adhesion molecules are indicated in Table 1. Cellsurface expression of E-selectin, ICAM-1, and VCAM-1was remarkably higher on HUVEC grown on AGE-fi-bronectin after four hours of stimulation with IL-1a,TNF-a or LPS, compared to HUVEC cultured on nativefibronectin as determined by flow cytometric analysis (Figs.1, 2, and 3). Additionally, the presence of IL-1a led to a

Fig. 3. Expression of VCAM-1. (A) Surfaceexpression (mean fluorescence intensity; mfi) ofVCAM-1 on HUVEC cultured on nativefibronectin or on AGE-fibronectin after fourhours or 18 hours of incubation with IL-1a,TNFa, LPS, AGE-albumin (sAGE), or mediumas determined by flow cytometric analysis. Theresults are given as percentages of expressionon non-stimulated HUVEC grown on nativefibronectin 6 SEM. P values (*P , 0.05; **P ,0.01) refer to comparison of equally treatedcells grown on fibronectin or AGE-fibronectin.(B) Semiquantitative RT-PCR analysis ofVCAM-1 transcript levels in endothelial cells.Cells were cultured on native fibronectin orAGE-fibronectin for 2.5 hours or 10 hours withIL-1a, AGE-albumin (sAGE) or medium alone.Results were corrected by the ABL values ofthe respective cDNA. The results are given aspercentages of expression in non-stimulatedHUVEC grown on native fibronectin 6 SEM. Pvalues (*P , 0.05) refer to comparison ofequally treated cells grown on fibronectin orAGE-fibronectin.


prolonged elevated E-selectin expression on HUVECgrown on AGE-fibronectin compared to standard cultureconditions (Fig. 1). IL-1a, TNFa and LPS enhancedICAM-1 and VCAM-1 expression on HUVEC after 18hours of culture on AGE-fibronectin (Figs. 2, 3). AGE-fibronectin significantly up-regulated the IL-1a inducedexpression of E-selectin mRNA at both timepoints,whereas ICAM-1, and VCAM-1 transcripts were increasedafter 2.5 hours (Figs. 1 to 3).

Neither AGE-fibronectin nor AGE-albumin nor a com-bination of both enhanced the cell surface expression ofE-selectin in the absence of inflammatory stimuli (Fig. 1).On the other hand, AGE-fibronectin alone increased thesurface expression of ICAM-1 or VCAM-1 on HUVECafter four hours of culture. An additive effect of AGE-albumin on AGE-fibronectin-induced ICAM-1 andVCAM-1 surface expression and transcript levels was ob-served after long-term incubation (Figs. 2 and 3), whereas

Fig. 3. Continued.


the combination of AGE-albumin and AGE-fibronectinand even AGE-fibronectin alone increased E-selectin tran-script levels already after 2.5 hours. No significant addi-tional effect of AGE-albumin on the expression of E-selectin mRNA in HUVEC was observed at the latertimepoint (Fig. 1).

PECAM expression on HUVEC was slightly, but signif-icantly enhanced by culture on AGE-fibronectin for fourhours and 18 hours with or without stimulating agents (Fig.4). AGE-albumin showed a weak, but significant stimula-tory effect on the PECAM transcript levels of endothelialcells grown on AGE-fibronectin at either timepoint (Fig.4). Most strikingly, IL-1a significantly reduced PECAM

mRNA expression independent from the presence ofAGE-fibronectin (Fig. 4).

Effect of AGE-fibronectin on transendothelial migrationof PMNLs

Confluent HUVEC monolayers grown on native fi-bronectin or AGE-fibronectin were incubated for fourhours or 18 hours with IL-1a or with medium (control)prior to co-incubation with PMNL. The spontaneous mi-gration of unstimulated PMNL through HUVEC-monolay-ers was increased about threefold if endothelial cells weregrown on AGE-fibronectin for four hours and 18 hourscompared to the migration pattern obtained on native

Fig. 4. Expression of PECAM. (A) Surfaceexpression (mean fluorescence intensity; mfi) ofPECAM on HUVEC cultured on nativefibronectin or on AGE-fibronectin after fourhours or 18 hours of incubation with IL-1a,TNFa, LPS, AGE-albumin (sAGE), or mediumas determined by flow cytometric analysis. Theresults are given as percentages of expressionon non-stimulated HUVEC grown on nativefibronectin 6 SEM. P values (*P , 0.05; **P ,0.01) refer to comparison of equally treatedcells grown on native fibronectin or AGE-fibronectin. (B) Semiquantitative RT-PCRanalysis of PECAM transcript levels inendothelial cells. Cells were cultured on nativefibronectin or AGE-fibronectin for 2.5 hours or10 hours with IL-1a, AGE-albumin (sAGE) ormedium alone. Results were corrected by theABL values of the respective cDNA. Theresults are given as percentages of expression innon-stimulated HUVEC grown on nativefibronectin 6 SEM. P values (*P , 0.05) referto comparison of equally treated cells grown onnative fibronectin or AGE-fibronectin.


fibronectin (Fig. 5A). There was a fivefold increase ofPMNL migration through IL-1a activated endotheliumcultured on native fibronectin that was more than sevenfoldincreased if endothelial cells were grown on AGE-fibronec-tin (Fig. 5A). To demonstrate that the enhanced migrationof PMNLs is due to the AGE-fibronectin induced expres-sion of cellular adhesion molecules, we performed addi-tional migration experiments in the presence of a cocktail

containing blocking antibodies against E-selectin, ICAM-1,VCAM-1, and PECAM. Blocking antibodies suppressedPMNL migration through IL-1a stimulated endothelialcells cultured on AGE-fibronectin to the levels of migrationthrough resting endothelial cells on AGE-fibronectin (Fig.5B), whereas antibodies against endoglin (antibody 1G2)and von Willebrand factor had no effect on PMNL migra-tion (data not shown).

Fig. 4. Continued.


DISCUSSION

The present study provides evidence that the interactionof inflammatory mediators and glycated matrix proteinsresults in a significant up-regulation of endothelial celladhesion structures and in a tremendous increase of thetransendothelial migration of PMNL.

It is well established that expression of cellular adhesionmolecules on endothelial cells is enhanced by variouscytokines during inflammation [9]. Furthermore, it hasbeen described that cellular adhesion molecules are presenton endothelial cells in atherosclerotic lesions [5, 7, 12].Based on the observation that AGEs can be found indiabetics and patients with uremia [15], the hypothesis wasraised that AGEs (matrix bound and soluble AGEs) couldmodulate the cell surface expression of cellular adhesionmolecules in these patients. In particular, we were inter-ested in the effect of AGEs on endothelial cell surface andmRNA expression of cellular adhesion molecules in the

presence of inflammatory stimuli, since patients with dia-betes mellitus and uremia frequently experience inflamma-tory episodes.

In the present study, we observed that AGE-matrixtogether with inflammatory stimuli significantly increasedthe cell surface expression of E-selectin, ICAM-1 andVCAM-1. Furthermore, a combination of AGE-albuminand matrix bound AGEs enhanced the cell surface expres-sion of ICAM-1 and VCAM-1. This effect of AGE-albuminwas not observed for E-selectin. If endothelial cells wereexposed to AGE-fibronectin alone, a significant up-regula-tion of ICAM-1 and VCAM-1 was observed (although to amuch lesser extent), but not of E-selectin. At the molecularlevel, we could demonstrate that mRNA levels of E-selectin, ICAM-1 and VCAM-1 significantly increased fol-lowing short-time stimulation of endothelial cells withmatrix bound AGEs and inflammatory mediators.Furthermore, the combination of AGE-albumin and

Fig. 5. Effect of AGE-fibronectin on transendothelial migration of PMNLs. Confluent HUVEC monolayers grown on native fibronectin (M) orAGE-fibronectin (o) were incubated for four hours or 18 hours with IL-1a or with medium (Control) prior to co-incubation with 5 3 105 PMNL ina double chamber system. Results are shown as number of migrated PMNL 6 SEM. (A) Migration of PMNL through HUVEC monolayers. *P , 0.05refers to comparison between the numbers of migrated PMNL through equally treated HUVEC grown on native fibronectin or AGE-fibronectin. (B)Blocking experiments. Incubation of HUVEC monolayers with a cocktail of mAbs against PECAM, ICAM-1, E-selectin and VCAM-1 for one hour afterincubation with medium (Control) or IL-1a and prior to co-incubation with PMNL. #P , 0.05 refers to comparison between the numbers of migratedPMNL through equally treated HUVEC grown on the same type of fibronectin with blocking mAbs (B) versus incubation without blocking mAbs (A).


AGE-fibronectin also resulted in an increase of ICAM-1,VCAM-1 and E-selectin transcript levels.

Several studies demonstrated that AGE-albumin en-hances cell surface expression of endothelial ICAM-1 andVCAM-1 [30–32]. In our study, AGE-albumin enhancedthe expression of ICAM-1 and VCAM-1 on endothelialcells grown on AGE-fibronectin, but not on endothelialcells cultured on native fibronectin. This discrepancy mightbe due to the different degrees of long-term glycation usedin these studies. The concentration of albumin (500 mg/ml)in our study was higher than that used by Schmidt et al [32].Though our albumin concentration was high, fluorescenceintensity was low compared to that of AGE-fibronectin(2300 RFU vs. 15000 RFU, respectively). The fluorescenceintensity of the AGE-albumin used in the present study,however, resembles the in vivo situation, where proteinswith longer half life (such as, fibronectin) may be prone tomore intense glycation compared to proteins with shorterhalf life (for example, serum albumin). Furthermore, weused G-6-P instead of glucose for in vitro glycation. Sinceproteins in diabetic patients are chronically exposed tohyperglycemic conditions for many years, we took highlyconcentrated (0.5 M) G-6-P as used by others [33] toachieve comparable in vitro glycation in a relatively shorttime.

In addition to E-selectin, ICAM-1 and VCAM-1, weshow that matrix bound AGEs also enhanced the endothe-lial cell surface expression of PECAM in the presence ofinflammatory mediators. This up-regulatory effect was alsoobserved on endothelial cells cultured on AGE-fibronectinin the presence of AGE-albumin. At the gene transcriptlevel we demonstrate that an inflammatory stimulus, suchas IL-1a, led to a significant reduction of PECAM mRNAlevels, probably due to enhanced translational and cellsurface expression activities at the time points investigatedin this study.

An interesting finding of the present study is the timedependent modulatory effect of AGE-fibronectin on thesurface expression of cellular adhesion molecules. Celladhesion molecules are responsible for different sequentialsteps in the leukocyte diapedesis process. E-selectin isinvolved in “rolling,” ICAM-1 and VCAM-1 in “firmadhesion” and PECAM, ICAM-1 and VCAM-1 in “trans-endothelial migration” of leukocytes [34, 35]. Our resultsindicate that the cooperation of inflammatory cytokinesand AGE-fibronectin leads to a prolonged expression of“early” endothelial adhesion molecules and to a prematureexpression of “late” adhesion molecules. The co-stimula-tion via AGE-fibronectin and IL-1a, which is supposed tofrequently occur in diabetic or renal failure patients, sug-gests a mechanism that enhances leukocyte recruitment tothe blood vessel wall and finally may be involved in theacceleration of atherogenesis [20, 21].

Migration of leukocytes across the endothelial barrier isone of the key events in atherosclerotic plaque formation

[4, 36]. Therefore, we studied transendothelial migration ofPMNLs following activation of endothelial cells with IL-1aand observed a more than sevenfold increase in PMNLmigration if endothelial cells were grown on AGE-fi-bronectin. These results demonstrate that the exposure ofendothelial cells to AGE-fibronectin enhances the cyto-kine-induced transendothelial migration of PMNL.

From previous studies with mononuclear phagocytes andAGE-albumin it seems likely that monocytes are able toadhere to and migrate through AGE modified endothelium[37]. They are retained in the subendothelial space due toa chemotactic effect of AGEs and the expression of recep-tors for AGE and CD49d/29. In the present study, we haveused PMNL since they lack AGE-receptors and fibronec-tin-receptors (CD49d/29, VLA-4). Therefore, they couldnot be captured by fibronectin or AGE in our migrationassay. Since AGE-compounds were described to havechemotactic effects on leukocytes [37], we performedblocking experiments, which clearly confirmed the involve-ment of IL-1a and AGE-fibronectin induced up-regulationof cellular adhesion molecule expression on HUVEC intransendothelial migration. Furthermore, in our study,growing endothelial cells on AGE-fibronectin did not resultin enhanced production of IL-8 (data not shown), which isa potent chemoattractant for PMNL [38].

Studies from other groups demonstrated that advancedglycation of fibronectin does not alter the adhesion prop-erties of endothelial cells on this growth surface and it doesnot modify spreading of HUVEC in cell culture experi-ments [33]. Therefore, the augmented transendothelialmigration of PMNL cannot be explained by increasedporosity of the endothelial cell monolayers. We thereforeconclude that the increased transendothelial migration ofPMNL through AGE-fibronectin versus native fibronectinin our experiments is due to cellular adhesion molecule-related interactions and was not due to chemotactic factorsor structural alterations of the endothelial cell monolayer.

Adhesion of PMNL to endothelial cells and their matrixproteins as well as transendothelial migration of PMNL hasrecently received considerable interest in the pathogenesisof atherosclerotic lesions [4, 39, 40]. PMNL can releasemediators including oxygen-derived free radicals, resultingin an impairment of endothelial cell function [41]. Thedamaged endothelium then exposes receptors for immuno-globulins and complement, which further promote adher-ence of PMNL to endothelial cells [42]. The terminal C5b-9complement complex has been localized in atheroscleroticplaques, indicating that complement activation had oc-curred. This process was mediated by activation of PMNL[42]. Furthermore, it has been shown that PMNL causeendothelial dysfunction during ischemia and are involved inreperfusion injury [4]. Thus, our data support the conceptthat the interaction of leukocytes with endothelial cellsexposed to glycated matrix proteins and inflammatorymediators is putatively involved in the development of


vascular pathology. These findings suggest some clinicalimportance of AGEs and inflammatory mediators, espe-cially in patients with recurrent acute infectious episodes.

It was not the purpose of the present study to clarify theregulatory mechanisms of AGE modulated mRNA andprotein expression of endothelial cellular adhesion mole-cules. According to Stern and colleagues [43], however, theNF-kB/IkBa system has been suggested to participate inthe regulation of the expression of these molecules. It hasbeen shown that IkBa is capable of displacing NF-kBp50/p65 from E-selectin- and VCAM-1 NF-kB bindingelements, thereby negatively regulating the expression ofthese molecules. It might be possible that signaling viaAGE proteins/AGE receptors influences this negative reg-ulatory process that is supposed to prevent the inappropri-ate expression of adhesion molecules. By offsetting thephysiologic and dynamic balance of the NF-kB/IkBa sys-tem, AGEs might promote the onset and the perpetuationof vascular disease. Another potential candidate for medi-ating AGE effects is the cytokine-induced enhancer com-plex, which provides an attractive model for superinductionof endothelial adhesion structures by the complex interac-tion of multiple transcription regulatory elements [44].

In conclusion, this is the first study to our knowledgedemonstrating that the effect of an AGE-modified matrixprotein, AGE-fibronectin, in combination with inflamma-tory stimuli, enhances the endothelial cellular adhesionmolecule expression and augments transmigration ofPMNL through cytokine activated endothelial cells. Weprovide evidence that irreversibly glycated matrix proteinfibronectin modulates the endothelial expression of E-selectin, ICAM-1, VCAM-1 and PECAM both on the cellsurface expressed protein and the mRNA-level, an effectthat is enhanced by the presence of inflammatory stimuli.The irreversibly glycated matrix compound fibronectincaused a dramatic increase of transendothelial migration ofPMNL. These findings represent a potentially importantlink between glycated matrix proteins, inflammation, andactivation of the endothelial adhesion cascade that may beinvolved in vascular pathology contributing to the develop-ment of the excess atherosclerotic burden in patients withdiabetes mellitus or chronic renal failure.

ACKNOWLEDGMENTS

This work was supported by the “Medizinisch-Wissenschaftlicher Fondsdes Burgermeisters der Stadt Wien” of the City of Vienna, by the“Hochschuljubilaumsfond” of the University of Vienna and by the “Fondszur Forderung der Wissenschaftlichen Forschung,” project number10918MED. We wish to thank Raute Sunder-Plassmann for helpfulcomments and careful reading of the manuscript.

Reprint requests to Dr. Gere Sunder-Plassmann, Department of InternalMedicine III, Division of Nephrology and Dialysis, Wahringer Gurtel 18-20,A-1090, Wien, Austria.E-mail: [email protected]

APPENDIX

Abbreviations used in this article are: ABTS, 2,29-Azinobis-3-ethylben-zthiazoline-sulfonic acid; AGE, advanced glycation end products; AGE-albumin, advanced glycated albumin; AGE-fibronectin, advanced glycatedfibronectin; B2m, B2-microglobulin; BSA, bovine serum albumin; calcein-AM, calcein-acetoxymethylester; FCS, fetal calf serum; FITC, fluoresceinisothiocyanate; G-6-P, glucose-6-phosphate; HUVEC, human umbilicalchord vein endothelial cells; ICAM, intercellular cellular adhesion mole-cule; IL, interleukin; LPS, lipopolysaccharide; mAb, monoclonal antibod-ies; NF-kB, nuclear factor-kB; PBS, phosphate buffered saline; PECAM,platelet-endothelial cellular adhesion molecule; PMNL, polymorphonu-clear leukocyte; RFU, relative fluorescence units; RT-PCR, reversetranscription-polymerase chain reaction; TNF-a, tumor necrosis factor-a;VCAM, vascular cellular adhesion molecule.

REFERENCES

1. RUDERMAN NB, HAUDENSCHILD C: Diabetes as an atherogenic factor.Progr Cardiovasc Dis 26:373–412, 1984

2. MANSKE CL, WILSON RF, WANG Y, THOMAS W: Prevalence of, andrisk factors for, angiographically determined coronary artery diseasein type I diabetic patients with nephropathy. Arch Intern Med 152:2450–2455, 1992

3. WHEELER DC: Cardiovascular disease in patients with chronic renalfailure. Lancet 348:1673–1674, 1996

4. BELCH JJ: The relationship between white blood cells and arterialdisease. Curr Opin Lipidol 5:440–446, 1994

5. CYBULSKY MI, GIMBRONE, MA JR: Endothelial expression of amononuclear leukocyte adhesion molecule during atherogenesis. Sci-ence 251:788–791, 1991

6. LI H, CYBULSKY MI, GIMBRONE MA, LIBBY P: An atherogenic dietrapidly induces VCAM-1, a cytokine-regulatable mononuclear leuko-cyte adhesion molecule, in rabbit aortic endothelium. ArteriosclerThromb 13:197–204, 1993

7. VAN DER WAL AC, DAS PK, TIGGES AJ, BECKER AE: Adhesionmolecules on the endothelium and mononuclear cells in humanatherosclerotic lesions. Am J Pathol 141:1427–1433, 1992

8. O’BRIEN KD, ALLEN MD, MCDONALD TO, CHAIT A, HARLAN JM,FISHBEIN D, MCCARTY J, FERGUSON M, HUDKINS K, BENJAMIN CD,LOBB R, ALPERS CE: Vascular cell adhesion molecule-1 is expressedin human coronary atherosclerotic plaques. Implications for the modeof progression of advanced coronary atherosclerosis. J Clin Invest92:945951, 1993

9. BEVILACQUA MP, NELSON RM, MANNORI G, CECCONI O: Endothelial-leukocyte adhesion molecules in human disease. Annu Rev Med45:361–378, 1994

10. POSTON RN, HASKARD DO, COUCHER JR, GALL NP, JOHNSON-TIDEYRR: Expression of intercellular adhesion molecule-1 in atheroscle-rotic plaques. Am J Pathol 140:665–673, 1992

11. RICHARDSON M, HADCOCK SJ, DERESKE M, CYBULSKY MI: Increasedexpression in vivo of VCAM-1 and E-selectin by the aortic endothe-lium of normolipemic and hyperlipemic diabetic rabbits. ArteriosclerThromb 14:760–769, 1994

12. HAKKERT BC, KUIJPERS TW, LEEUWENBERG JFM, VAN MOURIK JA,ROOS D: Neutrophil and monocyte adherence to and migration acrossmonolayers of cytokine-activated endothelial cells: The contributionof CD18, ELAM-1 and VLA-4. Blood 78:2721–2726, 1991

13. POSTON RN, JOHNSON-TIDEY RR: Localized adhesion of monocytes tohuman atherosclerotic plaques demonstrated in vitro: Implications foratherogenesis. Am J Pathol 149:73–80, 1996

14. MAKITA Z, RADOFF S, RAYFIELD EJ, YANG Z, SKOLNIK E, DELANEYV, FRIEDMAN EA, CERAMI A, VLASSARA H: Advanced glycosylationend products in patients with diabetic nephropathy. N Engl J Med325:836–842, 1991

15. NAKAMURA Y, HORII Y, NISHINO T, SHIIKI H, SAKAGUCHI Y, KA-GOSHIMA T, DOHI K, MAKITA Z, VLASSARA H, BUCALA R: Immuno-histochemical localization of advanced glycosylation endproducts incoronary atheroma and cardiac tissue in diabetes mellitus. Am J Pathol143:1649–1656, 1993

16. VLASSARA H, BUCALA R, STRIKER L: Biology of disease: Pathogeniceffects of advanced glycosylation: Biochemical, biologic, and clinicalimplications for diabetes and aging. Lab Invest 70:138–151, 1994


17. ESPOSITO C, GERLACH H, BRETT J, STERN D, VLASSARA H: Endothe-lial receptor-mediated binding of glucose-modified albumin is associ-ated with increased monolayer permeability and modulation of cellsurface coagulant properties. J Exp Med 170:1387–1407, 1989

18. BUCALA R, TRACEY KJ, CERAMI A: Advanced glycosylation productsquench nitric oxide and mediate defective endothelium-dependentvasodilatation in experimental diabetes. J Clin Invest 87:432–438, 1991

19. HAAG-WEBER M, HORL WH: Are granulocyte inhibitory proteinscontributing to enhanced susceptibility to infections in uremia? Neph-rol Dial Transplant 11(Suppl 2):S98–S100, 1996

20. HAAG-WEBER M, HORL WH: Uremia and infection: Mechanisms ofimpaired cellular host defense. Nephron 63:125–131, 1993

21. JAFFE EA, NACHMAN RL, BECKER CG, MINICK CR: Culture of humanendothelial cells derived from umbilical veins. Identification by mor-phologic and immunologic criteria. J Clin Invest 52:2745–2756, 1973

22. CHOMCZYNSKI P, SACCHI N: Single-step method of RNA isolation byacid guanidinium thiocyanate-phenol-chloroform extraction. AnalBiochem 162:156–169, 1987

23. MAURER J, JANSSEN JW, THIEL E, VAN DENDEREN J, LUDWIG WD,AYDEMIR U, HEINZE B, FONATSCH C, HARBOTT J, REITER A, RIEHMH, HOELZER D, BARTRAM CR: Detection of chimeric BCR-ABLgenes in acute lymphoblastic leukaemia by the polymerase chainreaction. Lancet 337:1055–1058, 1991

24. RUSSELL ME, ADAMS DH, WYNER LR, YAMASHITA Y, HALNON NJ,KARNOVSKY MJ: Early and persistent induction of monocyte chemoat-tractant protein in rat cardiac allografts. Proc Natl Acad Sci USA90:6086–6090, 1993

25. KIRSCHBAUM NE, GUMINA RJ, NEWMAN PJ: Organization of the genefor human platelet/endothelial cell adhesion molecule-1 shows alter-natively spliced isoforms and a functionally complex cytoplasmicdomain. Blood 84:4028–4037, 1994

26. SAITO I, TERAUCHI K, SHIMUTA M, NISHIIMURA S, YOSHINO K,TAKEUCHI T, TSUBOTA K, MIYASAKA N: Expression of cell adhesionmolecules in the salivary and lacrimal glands of Sjogren’s syndrome.J Clin Lab Anal 7:180–187, 1993

27. GEISSLER K, OHLER L, FODINGER M, VIRGOLINI I, LEIMER M,KABRNA E, KOLLARS M, SKOUPY S, BOHLE B, ROGY M, LECHNER K:Interleukin-10 inhibits growth and granulocyte/macrophage colony-stimulating factor production in chronic myelomonocytic leukemiacells. J Exp Med 184:1377–1384, 1996

28. METCALF JA, GALLIN JI, NAUSEEF WM, ROOT RK: LaboratoryManual of Neutrophil Function. New York, Raven Press, 1986, pp 3–17

29. SUNDER-PLASSMANN G, HOFBAUER R, SENGOELGE G, HORL WH:Quantification of leukocyte migration: Improvement of a method.Immunol Invest 25:49–63, 1996

30. SCHMIDT AM, ZHANG JH, CRANDALL J, CAO R, YAN SD, BRETT J,STERN D: Interaction of advanced glycation endproducts with theirendothelial cell receptor leads to enhanced expression of VCAM-1: Amechanism for augmented monocyte vessel wall interactions in dia-betes. (abstract) FASEB J 8(Part II):662, 1994

31. VLASSARA H, FUH H, DONNELLY T, CYBULSKY M: Advanced glycation

endproducts promote adhesion molecule (VCAM-1, ICAM-1) expres-sion and atheroma formation in normal rabbits. Mol Med 1:447–456,1995

32. SCHMIDT AM, HORI O, CHEN JX, LI JF, CRANDALL J, ZHANG J, CAO

R, YAN SD, BRETT J, STERN D: Advanced glycation endproductsinteracting with their endothelial receptor induce expression of vas-cular adhesion molecule-1 (VCAM-1) in cultured human endothelialcells and in mice. A potential mechanism for the accelerated vascu-lopathy of diabetes. J Clin Invest 96:1395–1403, 1995

33. BOBBINK IWG, DE BOER HC, TEKELENBURG WLH, BANGA JD, DE

GROOT PG: Effect of extracellular matrix glycation on endothelial celladhesion and spreading: Involvement of vitronectin. Diabetes 46:87–93, 1997

34. CARLOS TM, HARLAN JM: Leukocyte-endothelial adhesion molecules.Blood 84:2068–2101, 1994

35. MULLER WA, WEIGL SA, DENG X, PHILLIPS DM: PECAM-1 isrequired for transendothelial migration of leukocytes. J Exp Med178:449–460, 1993

36. ROSS R: The pathogenesis of atherosclerosis: A perspective for the1990s. Nature 362:801–809, 1993

37. KIRSTEIN M, BRETT J, RADOFF S, OGAWA S, STERN D, VLASSARA H:Advanced protein glycosylation induces transendothelial humanmonocyte chemotaxis and secretion of platelet-derived growth factor:Role in vascular disease of diabetes and aging. Proc Natl Acad Sci USA87:9010–9014, 1990

38. ROT A: Neutrophil attractant/activation protein-1 (interleukin-8) in-duces in vitro neutrophil migration by haptotactic mechanism. EurJ Immunol 23:303–306, 1993

39. MAYROVITZ HN, WIEDMAN MP, TUMA RF: Factors influencingleukocyte adherence in microvessels. Thromb Haemost 38:823–830,1977

40. WELBOURN R, GOLDMAN G, KOBZIK L, PATERSON I, VALERI CR,SHEPRO D, HECHTMAN HB: Neutrophil adherence receptors (CD18)in ischemia. Dissociation between quantitative cell surface expressionand diapedesis mediated by leukotriene B4. J Immunol 145:1906–1911, 1990

41. ZIMMERMANN GA, PRESCOTT SM, MCINTYRE TM: Endothelial inter-actions with granulocytes: Tethering and signaling molecules. Immu-nol Today 13:93–100, 1992

42. NICULESCU F, RUS HG, VLAICU R: Activation of the terminalcomplement pathway in atherosclerosis. Clin Immunol Immunopathol45:147–155, 1987

43. SCHMIDT AM, HORI O, BRETT J, YAN SD, WAUTIER JL, STERN D:Cellular receptors for advanced glycation end products. Implicationsfor induction of oxidant stress and cellular dysfunction in the patho-genesis of vascular lesions. Arterioscler Thromb 14:1521–1528, 1994

44. COLLINS T, READ MA, NEISH AS, WHITLEY MZ, THANOS D, MANIA-TIS T: Transcriptional regulation of endothelial cell adhesion mole-cules: NF-kB and cytokine-inducible enhancers. FASEB J 9:899–909,1995


endothelial cell adhesion molecule and pmnl response to inflammatory stimuli and age-modified...

Documents