Donate Help Contact The AHA Sign In Home
American Heart Association
Arteriosclerosis, Thrombosis, and Vascular Biology
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents | Next Article »
Arteriosclerosis, Thrombosis, and Vascular Biology. 1995;15:1456-1465

This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Camejo, E. H.
Right arrow Articles by Bondjers, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Camejo, E. H.
Right arrow Articles by Bondjers, G.
(Arteriosclerosis, Thrombosis, and Vascular Biology. 1995;15:1456-1465.)
© 1995 American Heart Association, Inc.

Articles

Interferon Gamma Binds to Extracellular Matrix Chondroitin-Sulfate Proteoglycans, Thus Enhancing Its Cellular Response

Eva Hurt Camejo; Birgitta Rosengren; Germán Camejo; Peter Sartipy; Gunnar Fager; Göran Bondjers

From the Wallenberg Laboratory for Cardiovascular Research, University of Gothenburg, Sahlgrenska University Hospital, and the Biochemistry Department Preclinical Research Laboratories (G.C.), Astra Hässle, Mölndal, Sweden.

Correspondence to Eva Hurt-Camejo, Wallenberg Laboratory, Sahlgrenska University Hospital, Göteborg University, Gothenburg S-41 345, Sweden. E-mail walevah@wlab.wall.gu.se.


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Abstract The amino acid sequence of interferon gamma (IFN-{gamma}) has basic amino acid clusters similar to the heparin-binding consensus sequences found in other proteins that bind to proteoglycans (PGs). We investigated whether recombinant human IFN-{gamma} could bind to extracellular matrix (ECM) PGs secreted by human arterial smooth muscle cells (HASMCs) in vitro and whether the interaction affected the cellular response to IFN-{gamma}. As an in vitro model of ECM we used the basement membrane from HASMCs in culture. The binding of 125I-IFN-{gamma} to ECM was reduced significantly by pretreatment of ECM with chondroitinase ABC, an enzyme that degrades chondroitin-sulfate glycosaminoglycans. IFN-{gamma} binding to ECM was reduced by increasing concentrations of chondroitin-6-sulfate. 125I-IFN-{gamma} (0.05 to 2 ng/mL) binding data indicated an apparent Kd of 2x10-11 mol/L and a maximum binding of 1.6x106 IFN-{gamma} molecules bound per square millimeter of ECM. Experiments with synthetic peptides suggested that residues 127 through 135 (AKTGKRKRS) are involved in the binding. The binding to chondroitin-sulfate PGs was confirmed by affinity chromatography of isolated [35S]chondroitin-sulfate PGs from ECM and cell-culture medium on immobilized IFN-{gamma}. The binding was abolished by treatment with chondroitinase ABC. ECM-bound IFN-{gamma} was more effective in inducing the expression of class II major histocompatibility antigens such as HLA-DR in HASMCs and human arterial endothelial cells than soluble IFN-{gamma}. These results suggest a role for chondroitin-sulfate PGs in immobilizing IFN-{gamma} in the ECM compartment and enhancing the cellular response to IFN-{gamma}.


Key Words: arteriosclerosis • interferon gamma • inflammation • proteoglycans


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMethods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
The cytokine IFN-{gamma} is produced by activated T lymphocytes and natural killer cells and released during the immune response and in inflammatory conditions. IFN-{gamma} exerts multiple effects on many different types of cells.1 2 3 SMCs respond to IFN-{gamma} by expressing class II MHC genes such as HLA-DR antigen.4 These genes are activated by SMCs in the vicinity of T cell infiltrates in experimentally injured arteries and human atherosclerotic plaques. This is probably due to a paracrine secretion of IFN-{gamma}.5 6 IFN-{gamma} is also a potent inhibitor of cell proliferation of SMCs in culture and ECM production6 7 8 ; thus, it may function as a negative growth regulator in vascular lesions.6 9 In vivo and in vitro data imply that IFN-{gamma} is involved in adhesion and transmigration processes between endothelium and leukocytes by activation of ECs.10 11 12 13 The expression of intracellular adhesion molecule–1, vascular cell adhesion molecule–1, class II MHC antigens, and interleukin-1 on cultured human ECs increases after treatment with IFN-{gamma}.14 IFN-{gamma} also inhibits EC proliferation.14 In macrophages, IFN-{gamma} inhibits lipoprotein lipase, scavenger receptor expression, and foam cell formation.15 16 IFN-{gamma} also regulates 15-lipoxygenase expression and suppresses platelet-derived growth factor expression in these cells.17 18

Arterial SMCs, monocytes, macrophages, ECs, and T lymphocytes are cells involved in atherosclerosis and restenosis.19 20 The multiple effects that IFN-{gamma} exerts on these cells suggest that local production of IFN-{gamma} in the arterial wall may either stimulate or suppress several cellular events involved in these processes. Although in vitro studies indicate that IFN-{gamma} may exist and exert its effects as a soluble factor, it is difficult to imagine how IFN-{gamma} may act to induce activation of cells and adhesion and transendothelial migration of leukocytes in vivo unless it is retained close to the site of secretion. Soluble cytokine molecules would be rapidly washed away from the area of production by extracellular fluid flow. Several cytokines and growth factors, such as granulocyte macrophage colony-stimulating factor,21 basic fibroblast growth factor,22 monocyte-colony stimulating factor,23 interleukin-1ß,24 interleukin-3,25 platelet-derived growth factor,26 27 and IFN-{gamma}28 bind to GAGs. In addition, LDL, lipoprotein lipase, and thrombomodulin bind to a specific type of GAG.29 This binding may increase the residence time of these macromolecules in the extracellular environment and may modulate their functional activity.29 30

GAGs are linear polymers of repeating disaccharides that contain one hexosamine and carboxyl and sulfate groups. GAG chains are covalently bound to a core protein that forms a PG; PGs are found inside cells, at the cell surface as a complete pericellular envelope, and in the ECM. PGs in blood vessels are synthesized and secreted by ECs and SMCs.31 The GAG chains of PGs are long, hydrophilic, and highly negatively charged. These physical characteristics are consistent with their role in binding proteins with positively charged peptide regions rich in lysine and arginine.32 33 Molecular modeling of protein-GAG interactions has uncovered consensus sequences of basic amino acids required for the interaction with GAGs.32 IFN-{gamma} has three stretches of basic amino acids similar to the GAG-binding sequences. Crystallographic analysis of human IFN-{gamma} indicates that these sequences of basic amino acids are exposed on the surface of the protein,34 a location that may facilitate their interaction with GAGs. Thus, in the present study we investigated whether human IFN-{gamma} could bind to ECM PGs synthesized by HASMCs and whether the interaction may modulate the biological activity of IFN-{gamma}. As an experimental model of ECM we used basement membrane secreted by HASMCs in culture. We also used cell-synthesized [35S]PGs and [3H]PGs isolated from ECM and cell-culture medium. Our results show that IFN-{gamma} binds to ECM CSPGs secreted by HASMCs. The ECM-bound IFN-{gamma} remains functionally active, inducing higher expression of HLA-DR antigen in HASMCs and HAECs than does soluble (ie, not bound to ECM) IFN-{gamma}.


*    Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Methods
down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Materials
Guanidine HCl (grade I), HEPES, Triton X-100, N-ethylmaleimide, {varepsilon}–aminocaproic acid, benzamidine HCl, phenylmethylsulfonylfluoride, and ethylaminohexanoic acid were purchased from Sigma Chemical Co. Chondroitin-4-sulfate, C6S, HS, chondroitinase ABC (protease free) and heparinase I were purchased from Seikagaku Kogyo Co and Sigma Chemical Co. Cyanogen bromide–activated Sepharose and Hi-Trap Q columns were bought from Pharmacia Fine Chemicals. Collagen I was purchased from Collaborative Biomedical Products, Becton Dickinson. Cell-culture media, antibiotics, nonessential amino acids, FCS, and culture vessels were from Flow Laboratories. Dulbecco's PBS with and without calcium and magnesium were purchased from JRH Biosciences, Sera-Lab. Cell culture–tested BSA and trypsin-EDTA were purchased from Sigma Chemical Co. Sodium [35S]sulfate (100 mCi/mmol), L-[4,5-3H]leucine (120 to 190 Ci/mmol), unlabeled recombinant human IFN-{gamma}, recombinant human 125I-IFN-{gamma} (1000 Ci/mmol; 17 000 MW; 10 U/ng), and hyperfilms for autoradiography were from Amersham International. Liquid scintillation cocktail (Ready Safe) for aqueous samples was from Beckman Instruments Inc. Monoclonal mouse anti-human HLA-DR antigen, CRI/43, negative control mouse IgG1, biotinylated rabbit anti-mouse immunoglobulins, normal rabbit serum, and ABC complex/alkaline phosphatase for determination of HDL-DR antigen expression in human cells were purchased from Dako (Dakopatts AB). Proliferation kit II (XTT) for nonradioactive quantification of cell proliferation and viability was purchased from Boehringer Mannheim. Salts, buffer substances, and detergents used in this work were of analytical grade and were purchased from Merck.

Cell Culture
Primary cultures of HASMCs from the inner media of human uterine arteries were established by using an explantation technique.35 The experiments were carried out with cells between passages 3 and 8. The cells were harvested by trypsinization and cultured at a cell density of 5x103 cells/cm2 in 6-, 12-, 24-, and 96-well plates for binding experiments to ECM and in 80-cm2 dishes for the isolation of PGs synthesized by the cells. HASMCs were cultured in dishes coated with a film of collagen I.36 Briefly, wells were coated with 2 to 5 µg/cm2 collagen I dissolved in 0.5 mol/L acetic acid. The wells were covered with the solution of collagen I and allowed to air dry. Coated wells were washed three times with PBS and three times with Waymouth medium before adding the cells. The cells were allowed to proliferate in Waymouth medium plus pooled human serum (10%, vol/vol), FCS (10%, vol/vol), 100 U/mL penicillin, 100 µg/mL streptomycin, 1 mmol/L sodium pyruvate, 4 mmol/L glutamine, and nonessential amino acids until they were confluent (about 8 days). Cell-culture medium was changed and fresh medium was added every 2 to 3 days. When the HASMCs were incubated with IFN-{gamma} the cell-culture medium was supplemented with 10% FCS (no human serum was added).

Normal HAECs were purchased from Clonetics Corp and Cytotech. The cells were harvested by trypsinization and cultured at a cell density of 5x103 cells/cm2 according to the manufacturer's specifications. The experiments were carried out with cells between passage 3 and 12. HASMCs and HAECs were tested for mycoplasma contamination during each passage by using a Mycoplasma test kit from Gen-Probe Inc. Endotoxin levels were regularly tested in cell-culture media and cell-culture reagents with Coatest/endotoxin (Chromogenix AB). Levels detected were <=0.01 EU/mL.

Preparation of Dishes Coated With ECM From HASMCs
Once HASMCs were confluent, the sub-SMC layer of ECM was exposed by dissolving the cell layer first with a solution of 0.5% Triton X-100 and then with 25 mmol/L NH4OH in PBS for 3 minutes each, followed by four washes with PBS. This ECM from HASMCs in culture remained intact, firmly attached to the entire area of the tissue-culture dish and free of nuclear or cellular debris.37

Before the binding assays with 125I-IFN-{gamma} some wells were pretreated with GAG hydrolyzing enzymes. ECM-coated wells were incubated with chondroitinase ABC (0.16 U/mL) overnight in 10 mmol/L HEPES buffer, pH 7.4, containing 140 mmol/L NaCl and 0.4% BSA. Treatment with heparinase I 0.4 U/mL overnight was done in 10 mmol/L HEPES buffer, pH 7.4, containing 140 mmol/L NaCl, 10 µmol/L CaCl2, and 0.4% BSA. After treatment the buffer with the enzyme was removed, and the ECM was washed three times with the same type of buffer. Different concentrations and incubation times were tested for the enzymes. The activities of the enzymes chondroitinase ABC and heparinase I were tested in parallel with the experiments by following the increase in absorbance at 232 nm caused by formation of the unsaturated disaccharides liberated by the action of the enzymes on C6S and HS. The conditions described above for incubation with the enzymes gave the highest removal of each type of GAG in ECM. Similar conditions were used when treating isolated PGs from ECM with the enzymes.

ECM Binding Studies
The ability of 125I-IFN-{gamma} to bind to ECM in vitro was measured in binding assays carried out in 12-, 24-, or 96-well plates coated with ECM. Before each experiment the ECM from confluent HASMCs, prepared as described above, was incubated with binding buffer (5 mmol/L HEPES, pH 7.4, 140 mmol/L NaCl, 5 mmol/L CaCl2, 2 mmol/L MgCl2, and 0.4% BSA) for 1 hour at 37°C to saturate nonspecific protein binding sites. The ECM-coated wells were then washed three times with the binding buffer and incubated with human recombinant 125I-IFN-{gamma} (0.02 to 2 ng/mL; 20 000 to 90 000 cpm/ng) at 37°C for 4 hours with and without a 200-fold excess of unlabeled IFN-{gamma}. After incubation, buffer with 125I-IFN-{gamma} was removed, and the wells were washed three times each with binding buffer and binding buffer without BSA. The amount of iodinated IFN-{gamma} bound was determined by dissolving the ECM with 0.2 mol/L NaOH. The nonspecific binding of the iodinated IFN-{gamma} in the presence of a 200-fold excess of unlabeled IFN-{gamma} represented 68±10% of the total amount of labeled IFN-{gamma} bound to the ECM. This nonspecific binding was subtracted from the total binding. The amount of radioactivity was counted in a Compugamma counter (LKB), and aliquots were used for protein determination.38 All the experiments were performed in triplicate or quadruplicate and repeated at least once. The data from the binding experiments were analyzed according to the Scatchard method39 by using the program GRAFIT, version 3.0 (R.J. Leatherbarrow, Erithacus Software Ltd, 1992).

Competition for Binding With GAGs
Competition for 125I-IFN-{gamma} binding to ECM by unlabeled GAGs was carried out in 96-well plates coated with ECM and prepared as described above. The plates were incubated with 125I-IFN-{gamma} (90 000 cpm/ng) in the presence of increasing concentrations of C6S or HS. After a 4-hour incubation at 37°C the plates were washed, and the amount of bound IFN-{gamma} was measured as described above.

Isolation of PGs from ECM and Cell-Culture Medium
For the isolation of PGs from cell-culture medium and ECM the cells were cultured as described.40 Briefly, 80-cm2 dishes with proliferative HASMCs (5000 cells/cm2) were maintained in Waymouth medium plus human serum (10%, vol/vol) and FCS (10%, vol/vol) for 3 days. The medium was then changed to basal minimal Eagle's diploid medium (sulfate free) plus human serum (10%, vol/vol) and FCS (10%, vol/vol). After 3 days in this medium, the cells were placed in fresh medium with the same composition plus 25 µCi/mL [35S]sulfate (100 mCi/mmol) and 10 µCi/mL [3H]leucine (120 Ci/mmol). After 48 hours the cell-culture medium was harvested and centrifuged to remove cells and cell debris, and guanidine HCl was added (final concentration, 4 mol/L). The ECM was prepared as described above and dissolved with 4 mol/L guanidine HCl, pH 5.8. A mixture of protease inhibitors with the following final concentrations was immediately added to both the ECM and cell-culture medium extractions: 5 mmol/L {varepsilon}–aminocaproic acid, 0.1 mmol/L phenylmethylsulfonylfluoride, 5 mmol/L N-ethylmaleimide, and 5 mmol/L benzamidine HCl. The extractions were incubated overnight at 4°C with gentle agitation and then dialyzed against an ion-exchange chromatography buffer (8 mol/L urea, 2 mmol/L EDTA, 0.5% Triton X-100, and 20 mmol/L Tris-HCl, pH 7.5) containing protease inhibitors for 2 days with two changes per day (3500 cut-off). After dialysis the samples were passed through Hi-Trap Q ion-exchange columns equilibrated with binding buffer. After passing the samples the columns were washed with binding buffer containing 0.20 mol/L NaCl to remove glycoproteins.41 The bound material was finally eluted with a gradient from 0.20 to 3 mol/L NaCl in binding buffer. Total counts in each fraction were determined by liquid scintillation counting. The fractions rich in [35S]PGs eluted around 1.5 mol NaCl. The fraction-containing PGs were pooled, dialyzed against water, and lyophilized. These samples were used for IFN-{gamma} affinity chromatography as described below. GAGs were isolated from [35S]PGs after papain digestion,42 and GAG composition was analyzed by agarose-gel electrophoresis.43 After electrophoresis the gel was dried and stored with an autoradiography film at -70°C.

Affinity Chromatography on a Sepharose–IFN-{gamma} Column
A Sepharose–IFN-{gamma} column (5x1-cm diameter) was prepared from IFN-{gamma} bound to cyanogen bromide–activated Sepharose according to the manufacturer's procedure. The column was equilibrated in binding buffer (5 mmol/L HEPES, pH 7.4, 20 mmol/L NaCl, 5 mmol/L CaCl2, and 2 mmol/L MgCl2). A similar column containing no IFN-{gamma} and blocked with ethanolamine served as a control column for nonspecific binding. The cell-culture medium and ECM fractions containing the [35S]sulfate- and [3H]leucine-labeled PGs synthesized by HASMCs were equilibrated in binding buffer and passed half through a IFN-{gamma} column and half through a control column. The columns were washed with 25 mL of the same buffer, the bound material was eluted with a gradient from 20 to 500 mmol/L NaCl, and fractions of 1 mL were collected. The peaks containing the ECM PGs and medium PGs that were bound to the immobilized IFN-{gamma} were dialyzed and lyophilized. Thereafter the samples were dissolved in 1.5 mL of 20 mmol/L Tris, pH 7.5, and divided in three equal parts. The first part was treated with chondroitinase ABC (20 U/mL), the second with heparinase I (20 U/mL), and the third part was used as control without enzyme treatment. The samples were incubated overnight at 37°C. The reaction was stopped by rapidly chilling the samples to -20°C. The samples were equilibrated in binding buffer and passed through the IFN-{gamma} affinity column as described above.

Blocking the Binding of IFN-{gamma} to ECM With Synthetic Peptide
The domains in IFN-{gamma} responsible for the binding of IFN-{gamma} to CSPGs present in ECM were investigated with synthetic peptide. The binding of IFN-{gamma} to ECM was blocked by preincubating ECM with synthetic peptide. Synthetic peptide corresponding to the basic amino acid–rich domains of the IFN-{gamma} sequence were synthesized in a Milligen 9050 peptide synthesizer. After synthesis the peptides were precipitated with ether and purified by high-performance liquid chromatography.44 After lyophilization, the peptides were stored in a desiccation chamber. The sequences of the peptides were confirmed by automated protein sequencing (Edman degradation) by using an Applied Biosystems 477 A protein sequencer. The wells were incubated with 100 nmol/L of each type of synthetic peptide in binding buffer. After a 2-hour incubation at 37°C, the binding buffer with the excess peptide was removed. The wells were washed three times with binding buffer and incubated with125I-IFN-{gamma} (200 pg/mL). After 4 hours' incubation at 37°C the plates were washed three times each with binding buffer and binding buffer without BSA. The amount of 125I-IFN-{gamma} bound to ECM was determined by dissolving the ECM with 0.2 mol/L NaOH and counting the radioactivity. The amount of protein per well was also measured. The values are expressed in terms of inhibition of IFN-{gamma} binding compared with 100% binding in the control (without competitors). The control bound 438±113 pg IFN-{gamma} per milligram of protein (n=6).

IFN-{gamma} Bioassay: HLA-DR Antigen Expression in HASMCs and HAECs
The ability of ECM-bound compared with unbound IFN-{gamma} (added as aliquots to the culture medium) in inducing the expression of HLA-DR antigen in cells was assayed in ECM-coated 96-well plates prepared as described above. The cells were incubated with similar amounts of both types of IFN-{gamma}. To calculate the amount of IFN-{gamma} bound to ECM, 96-well plates were washed and preincubated with binding buffer as described above. The plates were then incubated with increasing concentrations of 125I-IFN-{gamma} (0, 1, 10, 100, and 1000 U/mL [10 U/ng, 1000 cpm/U]) in 100 µL binding buffer in quadruplicate. After a 4-hour incubation at 37°C the plates were washed three times with binding buffer, and the amount of125I-IFN-{gamma} bound to the ECM was determined by dissolving the ECM with 0.2N NaOH and counting the radioactivity. To measure the expression of HLA-DR antigen in cells, ECM-coated 96-well plates were incubated as described above with increasing concentrations of unlabeled IFN-{gamma} in binding buffer. ECM-coated control plates were incubated in parallel with binding buffer alone. After a 4-hour incubation, the plates were washed three times with binding buffer and immediately 100 µL cell medium containing HAECs or HASMCs (both 5x104 cells/mL) was added to each well. The cells were allowed to adhere for 15 minutes at 37°C, after which two sets of ECM-coated control plates not preincubated with IFN-{gamma} received 10-µL aliquots of unlabeled IFN-{gamma}. One set received in the medium similar amounts of IFN-{gamma} as the ones bound to the ECM (calculated with radiolabeled IFN-{gamma} ). The second set received 0, 0.1, 1, 10, or 100 U added to the medium. This last control set was run to check for the activity and cellular response to IFN-{gamma} in concentrations reported to induce HLA-DR antigen in similar cell-culture systems.4 6 11 After 6 days of incubation, the expression of HDL-DR antigen was measured by enzyme-linked immunoassay with monoclonal antibody HLA-DR antigen, CRI/43, or mouse IgG1 negative control antibody following the manufacture's specifications and a modification of a previously described procedure.6 The optimal dilutions of the antibodies and ABC complex for each type of cell were determined by checkerboard titration. The absorbance values obtained with the negative control antibody were subtracted from the absorbance values obtained with the antibody against HLA-DR antigen. Net absorbance values obtained were expressed per number of cells counted or per XTT colorimetric values. The number of cells or XTT assay for quantification of cell proliferation and viability was determined in parallel plates.

The statistical significance of the pertinent data was evaluated by Student's t test by using the statistical program WINSTAT FOR WINDOWS. (R.K Fitch, Kalmia Co, Inc).


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 *Results
down arrowDiscussion
 down arrowReferences
 
ECM GAGs and Their Effect on IFN-{gamma} Binding to ECM
The possibility that IFN-{gamma} binds to GAGs in the ECM was addressed by using different degradative enzymes that cleave specific GAGs while leaving others intact. For this purpose, ECM was treated with either heparinase I or chondroitinase ABC, washed extensively, and tested for the ability to bind IFN-{gamma}. The treatment with chondroitinase ABC resulted in a significant (P<=.006, n=6) reduction of interferon binding to ECM, while the treatment with heparinase I was ineffective (Fig 1Down). These results were supported by the results obtained with the competition experiments. Fig 2Down shows the competition for 125I-IFN-{gamma} binding to ECM by excess unlabeled C6S and HS. At all concentrations used CS was more effective in inhibiting the binding of IFN-{gamma} to ECM than was HS. These results indicate that the C6S type of GAGs are involved in the binding of IFN-{gamma} to PGs present in the ECM from HASMCs.



View larger version (11K):
[in this window]
[in a new window]
  		Figure 1. Bar graph showing effect of GAGs present in the ECM on IFN-{gamma} binding. ECM-coated 24-well plates were treated with the GAG-degrading enzymes heparinase I (0.4 U/mL) and chondroitinase ABC (ChABC; 0.16 U/mL) overnight at 37°C. Control wells were incubated in parallel without enzymes. The ECM-coated wells were then washed six times with binding buffer and tested for their ability to bind 125I-IFN-{gamma} (0.2 ng/mL). After a 4-hour incubation at 37°C the amount of 125I-IFN-{gamma} bound to ECM was determined by dissolving the wells' contents with 0.2 mol/L NaOH. Nonspecific binding was measured in the presence of a 200-fold excess of unlabeled IFN-{gamma}. Values represent mean±SD of two experiments; n=6-8. **P<=.006.



View larger version (12K):
[in this window]
[in a new window]
  		Figure 2. Line graph showing competition for 125I-IFN-{gamma} by unlabeled GAGs. ECM-coated 96-well plates were incubated in binding buffer with 125I-IFN-{gamma} (20 ng/mL [90 000 cpm/well]) with increasing concentrations of C6S (Ch-6-Sulfate) or HS (Heparan S) in equimolar concentrations. After a 4-hour incubation the amount of 125I-IFN-{gamma} bound to ECM was determined. Values represent mean±SD of triplicate wells.

125I-IFN-{gamma} Binding to ECM From HASMCs in Culture
The binding of 125I-IFN-{gamma} to ECM-coated plates was analyzed as a function of IFN-{gamma} concentration as described in "Methods." 125I-IFN-{gamma} binding to ECM showed maximum binding at 4 hours at 37°C (data not shown). Fig 3Down shows the binding curve and Scatchard plot analysis of the binding. An apparent Kd of 2x10-11 mol/L was obtained; maximum binding capacity was 124 fmol/mg ECM protein, which represents 1.6x106 molecules/mm2. 125I-IFN-{gamma} did not bind covalently to the ECM, as more than 80% of the ECM-bound 125I-IFN-{gamma} was released upon a mild treatment with 200 mmol/L NaCl, 2 mmol/L EDTA, 10 mmol/L Tris, and 0.1% Tween 20, pH 7.4, for 30 minutes at 37°C (data not shown).



View larger version (16K):
[in this window]
[in a new window]
  		Figure 3. Concentration-dependent binding curve of 125I-IFN-{gamma} to ECM (top) and corresponding Scatchard plot (bottom). ECM-coated 12-well plates were incubated with increasing concentrations (0.05-1 ng/mL [2.94-58.8 pmol/L]) 125I-IFN-{gamma} at 37°C. Nonspecific binding was measured in the presence of a 200-fold excess of unlabeled IFN-{gamma}. After a 4-hour incubation at 37°C the 125I-IFN-{gamma} bound to ECM was measured. Values represent mean±SD of triplicate determinations.

ECM Binding Domain of IFN-{gamma}
To characterize the ECM binding domain of IFN-{gamma}, competition experiments were performed by preincubating ECM with 100 nmol/L of three synthetic peptides with analogy to sequences in the carboxyl-terminal region of IFN-{gamma}. Peptide I contained residues 87 through 95 (SNKKKRDDF), net charge +2; peptide II, residues 127 through 135 (AKTGKRKRS), net charge +5; peptide III, residues 138 through 146 (LFRGRRASE), net charge +2; and an irrelevant peptide, Hp-7 (EDYLILRVIGNMGQTMEQLTPELKS), net charge -2. This control or irrelevant peptide does not interact with PGs.45 Inhibition of interferon binding was highest when ECM wells were preincubated with peptide II (Fig 4Down). The subsequent binding of125I-IFN-{gamma} to ECM was inhibited by 50% (P<=.096, n=4). Preincubation with peptide I inhibited the binding of125I-IFN-{gamma} by 38%. It is interesting to observe that peptide III was not as effective as peptide I in inhibiting the binding of IFN-{gamma} even though both peptides have the same net positive charge (+2). Peptide III inhibited the binding of IFN-{gamma} by only 10%. These results suggest that residues Ala-Lys-Thr-Gly-Lys-Arg-Lys-Arg-Ser (127 through 135) in the carboxyl-terminal region of IFN-{gamma} constitute the domain that contributes most to the interaction of IFN-{gamma} with CSPGs in the ECM.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 4. Bar graph showing blockage of 125I-IFN-{gamma} binding to ECM by preincubation of ECM-coated wells with synthetic IFN-{gamma} fragment analogues: peptide I (SNKKKRDDF), peptide II (AKTGKRKRS), peptide III (LFRGRRASE), and peptide HP-7 (EDYLILRVIGNMGQTMEQLTPELKS). ECM-coated 24-well plates were preincubated with 100 nmol/L of each type of the other competitors in binding buffer at 37°C. After 2 hours the wells were washed and incubated with 125I-IFN-{gamma} (0.2 ng/mL). After a 4-hour incubation wells were washed and processed with 0.2 mol/L NaOH. Values are expressed in terms of percent IFN-{gamma} binding compared with 100% binding in the absence of competitors (438±113 pg IFN-{gamma} · mg protein-1 · well-1; n=4) and represent mean±SD of four determinations. P<=.096 for peptide II.

GAG Composition of PGs Synthesized by HASMCs
The results presented above indicate that CSPGs are involved in the binding of IFN-{gamma} in ECM. The isolation and characterization of these CSPGs synthesized by proliferating HASMCs in vitro has been reported.40 The monomer of this CSPG has an Mr of 1.1x106. More than 90% of the GAG chains consist of C6S and chondroitin-4-sulfate in a 6:4 ratio.40 PGs isolated from both cell-culture medium and ECM contain mainly CS-type GAGs and some HS GAGs (Fig 5Down). For this reason, these CS-rich PGs are named CSPGs. Hyaluronic acid was not detected. Similar GAG composition of PGs of the versican family synthesized by arterial SMCs in vitro has been reported by others.41



View larger version (70K):
[in this window]
[in a new window]
  		Figure 5. Autoradiographs showing GAG composition of the isolated PGs synthesized by HASMCs. Standards included chondroitin-4-sulfate (CsA; 2.2 µg), dermatan sulfate (CsB; 10 µg), C6S (CsC; 5 µg), HS (HepS; 10 µg), hyaluronic acid (HyalA; 20 µg), and a standard mixture containing 2 µg of each GAG standard. [35S]PGs and [3H]leucine PGs isolated from cell-culture medium and ECM were treated with papain to remove protein, precipitated, and run in an agarose gel. GAGs were from ECM (1.228 cpm) and cell-culture medium (2.021 cpm). Standards were visualized by staining with alcian blue and toluidine blue.

IFN-{gamma} Affinity Chromatography of Cell-Synthesized PGs
To better characterize the interaction of IFN-{gamma} with PGs from ECM we used affinity chromatography of labeled PGs isolated from ECM and cell-culture medium from HASMCs on immobilized IFN-{gamma}. After binding to Sepharose–IFN-{gamma} the total [35S]- and [3H]leucine-labeled PGs secreted by the cells and isolated from medium and ECM were eluted with a linear NaCl gradient. Fig 6Down (top) shows that [35S]CSPGs and [3H]CSPGs from cell-culture medium bound to the IFN-{gamma} column and eluted between 100 and 150 mmol/L of NaCl. This material represents 52% of the total amount of the [35S]- and [3H]PG-rich fraction from the cell-culture medium passed through the column and 82% of the material initially bound to the column. Fig 6Down (bottom) shows the results obtained with PGs isolated from ECM. The [35S]- and [3H]CSPGs isolated from ECM and bound to the IFN-{gamma} eluted between 120 and 220 mmol/L NaCl concentration. This material represents 52% of the total amount of the PG-rich fraction from ECM passed through the column and 62% of the material initially bound to the IFN-{gamma} column. The difference in elution patterns suggests that PGs from ECM have a higher affinity for IFN-{gamma} compared with CSPGs secreted into the medium that remain in solution. Fractions 5 through 13 from both elution profiles, ECM and medium, were pooled, dialyzed against water, and lyophilized. After the incubation with chondroitinase ABC or heparinase I the samples were passed through a Sepharose–IFN-{gamma} affinity column (Fig 7Down). Elution patterns similar to the ones shown in Fig 6Down were obtained with control samples. The difference in the elution pattern between PGs from ECM and cell-culture medium was again observed after additional chromatographic analysis. These results confirmed that PGs from ECM have a higher affinity for IFN-{gamma} compared with CSPGs that are found in the cell-culture medium. Treatment with chondroitinase ABC, which degrades CS GAGs, completely abolished the binding of [35S]CSPGs and [3H]CSPGs to IFN-{gamma} compared with the control samples. Heparinase I had no effect on the binding of cell-medium [35S]- and [3H]CSPGs to IFN-{gamma}. The elution profile of [35S]- and [3H]CSPGs from ECM changed after treatment with heparinase I. However, the total number of counts was the same in the heparinase I–treated sample and in the control sample; thus the removal of the small amounts of HS by heparinase I did not substantially alter the affinity of the PGs for IFN-{gamma}.



View larger version (24K):
[in this window]
[in a new window]
  		Figure 6. Line graphs showing IFN-{gamma} affinity chromatography of [ 35S]- ({bullet}) and [3H]leucine- ({triangleup}) labeled CSPGs isolated from cell medium and ECM from proliferating HASMCs. Cell-synthesized and labeled CSPGs isolated by ion-exchange chromatography were loaded and eluted from a column (5x1 cm) to which IFN-{gamma} was covalently attached to Sepharose and from a control (ctrl; {circ}) column for nonspecific binding. The bound CSPGs were eluted by a linear gradient between 20 and 500 mmol/L NaCl (dotted line); flow was 0.5 mL/min. Fractions (1 mL) were collected, and the amount of radioactivity in each fraction was determined.



View larger version (25K):
[in this window]
[in a new window]
  		Figure 7. Line graphs showing affinity chromatography of [35S] CSPGs after treatment with chondroitinase ABC (Ch ABC) and heparinase I (Hep I). The labeled CSPGs eluted between fractions 5 and 13 in Fig 6Up were pooled, divided in three equal parts, and treated with chondroitinase ABC ({circ}) or heparinase I ({triangleup}) or control samples not treated with enzymes (no enz; {square}). The samples were then passed through the Sepharose–IFN-{gamma} column and eluted by using a linear NaCl gradient (Conc.; dotted line).

Effect of Soluble and ECM-Bound IFN-{gamma} on the Expression of HLA-DR Antigen by HAECs and HASMCs
IFN-{gamma} induces the expression of class II MHC HLA-DR antigens in cells. This is a unique biological property of IFN-{gamma}. For this reason, we measured the expression of HLA-DR antigen in human macrophages, ie, HASMCs and HAECs cultured in the presence of either ECM-bound or soluble (unbound) IFN-{gamma} for 6 days. Figs 8Down and 9Down show the results obtained with HASMCs and HAECs, respectively. Both HASMCs and HAECs cultured in the presence of ECM-bound IFN-{gamma} expressed higher levels of HLA-DR antigen than the same type of cells cultured on ECM-coated plates in the presence of similar amounts of soluble IFN-{gamma}. In the case of HASMCs (Fig 8Down), 2.1 U/well of ECM-bound IFN-{gamma} induced approximately the same levels of HLA-DR antigen as 10 U/well of soluble IFN-{gamma}. In the case of HAECs (Fig 9Down), 1.2 U/well of ECM-bound IFN-{gamma} induced approximately similar levels of HLA-DR antigen as 10 U/well of soluble IFN-{gamma}. We used low concentrations of IFN-{gamma} in these experiments that may be close to those present in vivo. Therefore, the levels of HLA-DR antigen measured with soluble IFN-{gamma} are somewhat lower than those reported by other laboratories.4 6 11 Figs 8Down and 9Down also show that HASMCs and HAECs, as expected, expressed HLA-DR antigen in a dose-dependent fashion when incubated with soluble IFN-{gamma} in a range from 0 to 100 U/well, a dose usually used to induce activation of cells in vitro by IFN-{gamma}.4 6 7 8 11 From these results we concluded that the binding of IFN-{gamma} to CSPGs in ECM enhances the cellular response to IFN-{gamma} by HASMCs and HAECs.



View larger version (15K):
[in this window]
[in a new window]
  		Figure 8. Bar graph showing effect of ECM-bound or soluble IFN-{gamma} in inducing HLA-DR antigen expression in HASMCs. Three sets of cells were cultured with 100 µL medium in ECM-coated 96-well plates containing ECM-bound or soluble IFN-{gamma} as shown. The number of cells was counted in parallel dishes. Values represent mean±SD of triplicate determinations.



View larger version (15K):
[in this window]
[in a new window]
  		Figure 9. Bar graph showing effect of ECM-bound or soluble IFN-{gamma} in inducing HLA-DR antigen expression in HAECs. Cells were cultured with 100 µL medium in ECM-coated 96-well plates containing ECM-bound or soluble IFN-{gamma} as shown. XTT levels were measured in parallel dishes. Values represent mean±SD of triplicate determinations.


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 *Discussion
down arrowReferences
 
IFN-{gamma} is a pleiotropic protein that has antiviral, growth regulatory, and various immune modulatory activities that specifically regulate the expression of class II MHC antigens.1 2 3 4 6 7 8 9 10 11 12 13 14 Close examination of the human IFN-{gamma} amino acid sequence has revealed three stretches of basic residues on the carboxyl-terminal region similar to the consensus sequences present in heparin-binding proteins.32 33 Furthermore, these basic segments include amino acids such as serine and threonine, with hydrophilic side chains, and therefore they should be exposed on the surface of the protein.33 We investigated whether human recombinant IFN-{gamma} could bind to the GAG moiety of PGs and if the interaction could affect the biological function of IFN-{gamma}. CSPGs are the main components of the ECM in arteries.46 47 Arterial SMCs and ECs are the main cells involved in the synthesis and maintenance of the ECM in arterial tissues and also synthesize ECM proteins and PGs in vitro.31 41 As an in vitro model of ECM we used the basement membrane secreted by HASMCs in vitro. This basement membrane, which occurs as a thin sheet-like structure at the basal side of the cells, could mimic the in vivo distribution and structural arrangement of the PGs present in ECM.48 A similar basement-membrane model of ECM synthesized by ECs in vitro has been used to study the binding of thrombin, lipoprotein lipase, and LDLs to ECM PGs.37 49

Our results indicated that IFN-{gamma} binds to CSPGs present in ECM from HASMCs. By using a radiolabeled IFN-{gamma} with high specific activity it was possible to study the binding of IFN-{gamma} to ECM components in the physiological range of concentrations (30 to 300 pmol/L) expected to be found at sites of IFN-{gamma} production.50 Binding-data analysis revealed a ligand-binding mechanism with an apparent Kd of 2.2x10-11 mol/L and a maximum binding capacity of approximately 1.6x106 IFN-{gamma} molecules/mm2 of ECM. This high maximum binding suggests the possibility that a GAG chain may bind several molecules of IFN-{gamma}.

Experiments with different GAG-degrading enzymes suggest that CS is the component through which IFN-{gamma} binds to ECM. This result was confirmed by affinity chromatography experiments on immobilized IFN-{gamma} with PGs isolated from ECM and cell-culture medium. In these experiments, the degradation of CSPGs completely abolished the binding to IFN-{gamma}. Our results indicate that the binding of IFN-{gamma} to ECM in vitro is essentially irreversible at physiological concentrations of NaCl. IFN-{gamma} binds to the HS present in basement membrane from Engelbreth-Holm-Swarm tumor.28 The reported binding affinity of IFN-{gamma} for HS was lower (Kd=1.5x10-9 mol/L) than the affinity for CS obtained in the present work (Kd=2.2x10-11 mol/L). The GAG composition of ECM PGs from HASMCs showed small amounts of HS. Therefore, the low levels of inhibition observed with HS in the competition experiments may be due to the presence of small amounts of HS in ECM and not merely to a nonspecific binding effect. It is possible, then, that in vivo IFN-{gamma} may interact with different affinities to each type of GAG present in ECM systems.

PGs are a large family of macromolecules and major determinants of the physical and physiological properties of connective tissues. Because of their high sulfate- and carboxyl-group content in their GAG moiety, PGs are the most negatively charged polymers in living tissues. This allows them to interact with proteins with clusters of positively charged amino acids.28 29 32 33 46 There are three stretches with consensus sequences of basic amino acids in the carboxyl-terminal part of human recombinant IFN-{gamma}: peptide I, Ser87->Phe95 (SNKKKRDDF), net charge +2; peptide II, Ala127->Ser135 (AKTGKRKRS), net charge +5; and peptide III, Leu138->Gln146 (LFRGRRASE), net charge +2. These three sequences represent 19% of the total human IFN-{gamma} protein sequence. These are the only three sequences along the IFN-{gamma} protein with clusters of basic amino acids that resemble the heparin-binding consensus sequences found in other proteins.32 33 The contribution of additional sequences from other regions of IFN-{gamma} protein to ECM binding was not investigated here, but it is possible that patches of noncontinuous basic amino acid residues could also contribute to GAG binding.51 We investigated which of these sequences contribute most to the binding to ECM CSPGs. In competition experiments in which a constant amount of 125I-IFN-{gamma} was mixed with increasing concentrations of unlabeled synthetic peptide within the sequences of interest, higher binding of 125I-IFN-{gamma} to ECM was obtained in the presence of the synthetic peptide than in its absence. This unexpected result may be caused by association of the peptide with the 125I-IFN-{gamma} molecules, which consequently would increase the IFN-{gamma} binding capacity to the ECM CSPGs. For this reason we performed blocking experiments by preincubating ECM-coated wells with the synthetic peptide before incubating with 125I-IFN-{gamma}. The highest inhibition of IFN-{gamma} binding (50%) was obtained with peptide II. Although peptides I and III have the same net positive charge (+2), they inhibited the binding of IFN-{gamma} by only 38% and 10%, respectively, indicating that the presence of a net positive charge on the sequence is not sufficient to induce binding to ECM. This suggests that the binding of IFN-{gamma} to ECM is not only due to electrostatic interaction but may involve specific sequences with defined spatial distribution of the basic residues.

Although human IFN-{gamma} binds to a specific cell-surface receptor,2 the structures in IFN-{gamma} that induce the diverse biological functions on cells are not known. Several groups have reported the importance of the stretches of basic amino acids in the carboxyl terminal of IFN-{gamma} for expressing biological activity and receptor binding.52 53 54 55 56 However, the involvement of the last carboxyl-terminal 21 amino acids in conferring activity to human IFN-{gamma} is still under investigation.57 The consensus sequences of basic amino acids in the carboxyl-terminal region of IFN-{gamma} appear to be required for efficient intracellular translocation of intact IFN-{gamma} into the nucleus of target cells.58 59 This may be an additional functional role of the polycationic carboxyl-terminal region of human IFN-{gamma}.

The immobilization of IFN-{gamma} in the ECM through PG binding could be biologically relevant. We found that the binding of IFN-{gamma} to ECM PGs enhanced the cellular response to the cytokine in terms of HLA-DR antigen expression. The reasons behind this potentiation of the cellular response toward ECM-bound IFN-{gamma} were partially investigated. We observed that up to 80% of the ECM-bound IFN-{gamma} remained associated with the ECM for up to 6 days at 37°C. Furthermore, the binding of IFN-{gamma} to ECM PGs prevented the cytokine from extensive proteolytic degradation (data not shown), and HS protects the carboxyl-terminal region of IFN-{gamma} against proteolytic degradation.60 These results suggest that the binding of IFN-{gamma} to ECM PGs may prevent degradation of IFN-{gamma} and consequently prolong the presence of an active form of IFN-{gamma} around the cells. Another possibility could be that the binding to ECM PGs may increase the local concentration of IFN-{gamma} and consequently increase the number of cell-surface receptors that will be occupied by IFN-{gamma} molecules per cell. However, more experiments are needed to explore this hypothesis. The expression of MHC class II antigens, different adhesion molecules, and interleukin-1 is increased after treatment of ECs with IFN-{gamma}.13 These molecules appear to be involved in the adhesion of leukocytes to ECs.11 12 13 Bound IFN-{gamma} induced higher levels of class II MHC HLA-DR antigen expression in HASMCs and HAECs than did soluble IFN-{gamma}. In addition, human leukocyte adhesion was observed on HAECs cultured in ECM-coated wells containing bound IFN-{gamma} (data not shown). This suggests that the sequestration of IFN-{gamma} in an ECM system by interaction with PGs could potentiate lymphocyte and monocyte adhesion and transmigration of these cells from blood through the endothelium in vivo.

The ECM is assembled locally by cells into an organized three-dimensional network of glycoproteins, collagen, and PGs. The ECM not only provides a mechanical support and divides tissues into compartments, but it is also the space through which essential molecules are transported to and between the cells. In addition, the ECM plays a central role in the control of cell proliferation, differentiation, and migration by mediating cell adhesion and communication.49 61 ECM may also function as a reservoir for growth factors, cytokines, and other proteins with a capacity to bind to PGs.29 30 Results reported by other groups and those from the present study indicate that different biologically active molecules can be sequestered and stabilized by their interaction with ECM components, allowing more persistent and localized effects compared with the same molecules in a fluid system.22 25 28 36 This could be of particular importance for cells that are adhered to and surrounded by ECM-like arterial SMCs and ECs. Our results suggest that the interaction of IFN-{gamma} to the ECM PGs secreted by arterial SMCs could modulate its biological activity and availability to cells under physiological conditions. During SMC proliferation and increased production of ECM, such as occur in restenosis and atherogenesis, the described phenomenon may be of significance in triggering and maintaining local inflammatory processes.


*    Selected Abbreviations and Acronyms
 
BSA = bovine serum albumin
CS = chondroitin sulfate
CSPG = chondroitin sulfate–rich proteoglycan
C6S = chondroitin-6-sulfate
EC = endothelial cell
ECM = extracellular matrix
FCS = fetal calf serum
GAG = glycosaminoglycan
HAEC = human arterial endothelial cell
HASMC = human arterial smooth muscle cell
HS = heparan sulfate
IFN-{gamma} = interferon gamma
MHC = major histocompatibility
PBS = phosphate-buffered saline
PG = proteoglycan
SMC = smooth muscle cell


*    Acknowledgments
 
This work was supported by grants from the Swedish Medical Research Council (project No. 4531), the Heart and Lung Foundation (project No. 41027), the Swedish Medical Society (project No. 437.0), and Astra Hässle AB, Mölndal, Sweden. We thank Professor Göran K. Hansson and Olle Wiklund for useful comments during the preparation of this manuscript.

Received February 12, 1995; accepted June 14, 1995.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
 up arrowResults
 up arrowDiscussion
 *References
 

  1. Trinchieri G, Perussia B. Immune interferon: a pleiotropic lymphokine with multiple effects. Immunol Today. 1985;6:131-136.
  2. Farrar MA, Schreiber RD. The molecular cell biology of interferon-{gamma} and its receptor. Annu Rev Immunol. 1993;11:571-611. [Medline] [Order article via Infotrieve]
  3. Dalton KD, Pitts-Meek S, Keshav S, Figar IS, Bradley A, Stewart A. Multiple defects of immune cell functions in mice with disrupted interferon-{gamma} genes. Science. 1993;259:1739-1745. [Abstract/Free Full Text]
  4. Stemme S, Fager G, Hansson GK. MHC class II antigen expression in human vascular smooth muscle cells is induced by interferon-gamma and modulated by tumour necrosis factor and lymphotoxin. Immunology. 1990;69:243-249. [Medline] [Order article via Infotrieve]
  5. Hansson GK, Holm J, Jonasson L. Detection of activated T-lymphocytes in the human atherosclerotic plaque. Am J Pathol. 1989;135:169-175. [Abstract]
  6. Hansson GK, Jonasson L, Holm J, Clowes MM, Clowes AW. {gamma}-Interferon regulates vascular smooth muscle cell proliferation and Ia antigen expression in vivo and in vitro. Circ Res. 1988;63:712-719. [Abstract/Free Full Text]
  7. Warner SJC, Friedman GB, Libby P. Immune interferon inhibits proliferation and induces 2'5'-oligoadenylate synthetase gene expression in human vascular smooth muscle cells. J Clin Invest. 1989;83:1174-1182.
  8. Amento EP, Ehsani N, Palmer H, Libby P. Cytokines and growth factors positively and negatively regulate interstitial collagen gene expression in human vascular smooth muscle cells. Arterioscler Thromb. 1991;11:1223-1232. [Abstract/Free Full Text]
  9. Hansson GK, Holm J, Holm S, Foteu Z, Hedrich H-J, Fingerle J. T lymphocytes inhibit the vascular response to injury. Proc Natl Acad Sci U S A. 1991;88:10530-10534. [Abstract/Free Full Text]
  10. Oppenheimer-Marks N, Ziff M. Migration of lymphocytes through endothelial cell monolayer: augmentation by interferon-gamma. Cell Immunol. 1988;114:307-323. [Medline] [Order article via Infotrieve]
  11. Masuyama J-I, Minato N, Kano S. Mechanisms of lymphocyte adhesion to human vascular endothelial cells in culture: T lymphocyte adhesion to endothelial cells through endothelial HLA-DR antigens induced by gamma interferon. J Clin Invest. 1986;77:1596-1605.
  12. Munro JM, Pober JS, Cotran RS. Tumor necrosis factor and interferon-{gamma} induce distinct patterns of endothelial activation and associated leukocyte accumulation in skin of Papio anubis. Am J Pathol. 1989;135:121-133. [Abstract]
  13. Pober JM, Cotran RS. Cytokines and endothelial cell biology. Physiol Rev. 1990;70:427-451. [Free Full Text]
  14. Friesel R, Komoriya A, Maciag T. Inhibition of endothelial cell proliferation by gamma-interferon. J Cell Biol. 1987;104:689-696. [Abstract/Free Full Text]
  15. Jonasson L, Hansson GK, Bondjers G, Noe L, Etienne J. Interferon-gamma inhibits lipoprotein lipase in human monocyte-derived macrophages. Biochim Biophys Acta. 1990;1053:43-48. [Medline] [Order article via Infotrieve]
  16. Geng Y-J, Hansson GK. Interferon-{gamma} inhibits scavenger receptor expression and foam cell formation in human monocyte-derived macrophages. J Clin Invest. 1992;89:1-9.
  17. Conrad DJ, Kuhn H, Mulkins M, Highland E, Sigal E. Specific inflammatory cytokines regulate the expression of human monocyte 15-lipoxygenase. Proc Natl Acad Sci U S A. 1992;89:217-222. [Abstract/Free Full Text]
  18. Kosaka C, Masuda J, Shimokado K, Zen K, Yokota T, Sasaguri T, Ogata J. Interferon-{gamma} suppresses PDGF production from THP-1 cells and blood monocyte-derived macrophages. Atherosclerosis. 1992;97:75-87. [Medline] [Order article via Infotrieve]
  19. Ross R. The pathogenesis of atherosclerosis: a prospective for the 1990s. Nature. 1993;362:801-809. [Medline] [Order article via Infotrieve]
  20. Tanaka H, Sukhova GK, Swanson SJ, Clinton SK, Ganz P, Cybulski MI, Libby P. Sustained activation of vascular cells and leukocytes in the rabbit aorta after balloon injury. Circulation. 1993;88:1788-1803. [Abstract/Free Full Text]
  21. Gordon MY, Riley GP, Watt SM, Greaves MF. Compartmentalization of a haematopoietic growth factor (GM-CSF) by glycosaminoglycans in the bone marrow microenvironment. Nature. 1987;326:403-405. [Medline] [Order article via Infotrieve]
  22. Bashkin P, Doctrow M, Klagsbrun M, Svahn CM, Folkman J, Vlodavsky I. Basic fibroblast growth factor binds to subendothelial matrix and is released by heparinase and heparin-like molecules. Biochemistry. 1989;28:1737-1743. [Medline] [Order article via Infotrieve]
  23. Suzu S, Ohtsuki T, Makishima M, Yanai N, Kawashima K, Nagata N, Motoyoshi K. Biological activity of a proteoglycan form of macrophage colony-stimulating factor and its binding to type V collagen. J Biol Chem. 1992;267:16812-16815. [Abstract/Free Full Text]
  24. Ramsden L, Rider CHC. Selective and differential binding of interleukin (IL)-1{alpha}, IL-1ß, IL-2 and IL-6 to glycosaminoglycans. Eur J Immunol. 1992;22:3027-3031. [Medline] [Order article via Infotrieve]
  25. Roberts R, Gallagher J, Sooncer E, Allen TD, Bloomfield F, Dexter TM. Heparan sulphate bound growth factors: a mecha- nism for stromal cell mediated haemopoiesis. Nature. 1988;332:376-378. [Medline] [Order article via Infotrieve]
  26. Fager G, Camejo G, Olsson U, Östergre-Lunden G, Bondjers G. Heparin-like glycosaminoglycans influence growth and phenotype of human arterial smooth muscle cells in vitro, II: the platelet-derived growth factor A-chain contains a sequence that specifically binds heparin. In Vitro Cell Dev Biol. 1992;28:176-180.
  27. Raines EW, Ross R. Compartmentalization of PDGF on extracellular binding sites dependent on exon-6-encoded sequences. J Cell Biol. 1992;116:533-543. [Abstract/Free Full Text]
  28. Lortat-Jacob H, Kleinman HK, Grimaud JA. High-affinity binding of interferon-{gamma} to a basement membrane complex (Matrigel). J Clin Invest. 1991;87:878-883.
  29. Camejo G, Hurt-Camejo E, Olsson U, Bondjers G. Proteoglycans and lipoproteins in atherosclerosis. Curr Opin Lipidol. 1993;4:385-391.
  30. Ruoslahti E, Yamaguchi Y. Proteoglycans as modulators of growth factor activities. Cell. 1991;64:867-869. [Medline] [Order article via Infotrieve]
  31. Wight TN, Heinegård DK, Hascall VC. Proteoglycans: structure and function. In: Hay ED, ed. Cell Biology of Extracellular Matrix. 2nd ed. New York, NY: Plenum Press; 1991:45-71.
  32. Cardin AD, Weintraub HJR. Molecular modeling of protein-glycosaminoglycan interactions. Arteriosclerosis. 1989;9:21-32. [Abstract/Free Full Text]
  33. Jackson RL, Bush SJ, Cardin AD. Glycosaminoglycans: molecular properties, protein interactions and role in physiological processes. Physiol Rev. 1991;71:481-539. [Free Full Text]
  34. Ealick SE, Cook WJ, Vijay-Kummar S, Carson M, Nagabhushan TL, Trotta PP, Bugg CHE. Three-dimensional structure of recombinant human interferon-{gamma}. Science. 1991;252:698-702. [Abstract/Free Full Text]
  35. Fager G, Hansson G, Ottosson P, Dahllof B, Bondjers G. Human arterial smooth muscle cells in culture: inverse relationships between proliferation and expression of contractile proteins. In Vitro Cell Dev Biol. 1988;25:511-519.
  36. Thyberg J, Hedin U, Bottger BA. Attachment substrate for smooth muscle cells. In: Piper HM, ed. Cell Culture Techniques in Heart and Vessel Research. Berlin, FRG: Springer; 1991:315-333.
  37. Bar-Shavit R, Eldor A, Vlodavsky I. Binding of thrombin to subendothelial extracellular matrix: protection and expression of functional properties. J Clin Invest. 1989;84:1096-1104.
  38. Markwell MAK, Hass SN, Bieber LL, Tolbert NE. A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal Biochem. 1978;87:206-210. [Medline] [Order article via Infotrieve]
  39. Hulme EC, Birdsall NJM. Strategy and tactics in receptor-binding studies. In: Receptor-Ligand Interactions: A Practical Approach. Hulme EC, ed. New York, NY: Oxford University Press; 1992.
  40. Camejo G, Fager G, Rosengren B, Hurt-Camejo E, Bondjers G. Binding of low density lipoproteins by proteoglycans synthesized by proliferating and quiescent human arterial smooth muscle cells. J Biol Chem. 1993;268:14131-14137. [Abstract/Free Full Text]
  41. Wight TN, Hascall VC. Proteoglycans in primate arteries, III: characterization of the proteoglycans synthesized by arterial smooth muscle cells in culture. J Cell Biol. 1993;96:167-176. [Abstract/Free Full Text]
  42. Alavi MZ, Moore S. Glycosaminoglycan composition and biosynthesis in the endothelium covered neointima of deendothelialized rabbit aorta. Exp Mol Pathol. 1985;42:389-400. [Medline] [Order article via Infotrieve]
  43. Alves CS, Mourao AS. Interaction of high molecular weight chondroitin sulfate from human aorta with plasma low density lipoproteins. Atherosclerosis. 1988;73:113-124. [Medline] [Order article via Infotrieve]
  44. Olsson U, Camejo G, Bondjers G. Binding of a synthetic apolipoprotein B-100 peptide and peptide analogues to chondroitin 6-sulphate: effects of the lipid environment. Biochemistry. 1993;7:1858-1865.
  45. Camejo G, Olofsson S-O, López F, Carlsson P, Bondjers G. Identification of apoB-100 segments mediating the interaction of low-density lipoproteins with arterial proteoglycans. Arteriosclerosis. 1988;8:368-377. [Abstract/Free Full Text]
  46. Wight TN. Arterial proteoglycans. Arteriosclerosis. 1988;9:1-20. [Abstract/Free Full Text]
  47. Volker W, Schmidt A, Buddecke E. Mapping of proteoglycans in human arterial tissue. Eur J Cell Biol. 1987;45:72-79. [Medline] [Order article via Infotrieve]
  48. Paulsson M. Basement membrane proteins: structure, assembly and cellular interactions. Crit Rev Biochem Mol Biol. 1992;27:93-127. [Medline] [Order article via Infotrieve]
  49. Tabas I, Li Y, Brocia RW, Xu SW, Swanson TL, Williams KJ. Lipoprotein lipase and sphingomyelinase synergistically enhance the association of atherogenic lipoproteins with smooth muscle cells and extracellular matrix: a possible mechanism for low density lipoprotein and lipoprotein (a) retention and macrophage foam cell formation. J Biol Chem. 1993;268:20419-20432. [Abstract/Free Full Text]
  50. Nathan C, Yoshida R. Cytokines: interferon-{gamma}. In: Gallin JI, Goldstein IM, Snyderman R, eds. Inflammation: Basic Principles and Clinical Correlates. New York, NY: Raven Press Ltd; 1988:229-251.
  51. Margalit H, Fischer N, Ben-Sasson SA. Comparative analysis of structurally defined heparin binding sequences reveals a distinct spatial distribution of basic residues. J Biol Chem. 1993;268:19228-19231. [Abstract/Free Full Text]
  52. Lundell D, Lunn CH, Dalgarno D, Fossetta J, Greenberg R, Reim R,