Supplementary methods: Reagents.Recombinant mouse and human Gremlin-1, mouse Gremlin-2, mouse and human MIF, and human MCP-1 were from R&D Systems, MN, U.S.A. INF-γ was from Pepro Tech, NJ, U.S.A. TNF-alpha was purchased from tebu-bio GmbH, Germany. Antibodies to Gremlin-1 were from Abnova, Heidelberg, Germany (polyclonal) and Abgent, San Diego, CA (clone RB2060). Antibodies to CD68 (polyclonal) and CD14 (polyclonal) were from Abbiotec, San Diego, CA, U.S.A, to Mac-3 from BD Biosciences, CA, U.S.A. (clone M3/84), to MIF (polyclonal) from R&D Systems, Wiesbaden, Germany, to TNF-alpha (polyclonal) from Abcam, Cambridge, UK, and to human IgG (polyclonal) from Jackson Immuno Research, West Grove, U.S.A.
Cell isolation and cell lines.Human monocytes were isolated from peripheral venous blood samples by adherence after Ficoll-Paque purification of peripheral blood mononuclear cells as described 1. Human umbilical vein endothelial cells (HUVEC) were cultivated until confluency in complete endothelial cell basal medium (PAA) containing growth factors and 10% FCS. Human aortic endothelial cells (HAECs) were grown in complete endothelial cell basal medium (PAA) containing growth factors, 10% human serum and 10% FCS and grown to confluency. Flp-InTM-CHO cells were grown following the recommendations of the manufacturer (Invitrogen). ECV 304 cells (human urinary bladder carcinoma cell line) were cultivated in complete EAGLE medium with 10% FCS and growth factors.
MIF-induced TNF-α-secretion from macrophages.Macrophages were differentiated from human monocytes isolated from buffy coats by density gradient centrifugation and differentiated as described 2. Human macrophages were stimulated with MIF 0.25 µg/mL, Gremlin-1 0.5 µg/mL (or as indicated in the figure), or MIF and Gremlin-1 for 12 hours in RPMI 1640 with 2% FCS and 0.5 µg/ml polymyxin B. When MIF and Gremlin-1 were added together, they were pre-incubated with each other for 10 min before added to the culture. TNF-α concentration was determined in cell supernatants using ELISA (R&D Systems Inc., MN, U.S.A.).
Cloning and PCR materials.MMLV Reverse Transcriptase (Invitrogen, U.S.A.); ImPromII reverse transcription system (Promega, Madison, U.S.A.); restriction enzymes (Promega, WI, U.S.A., and New England BioLabs, MA, U.S.A.); DNA ligase, DNA taq polymerase, RNAsin, T3 RNA polymerase and T7 RNA polymerase (Promega, WI, U.S.A.); DNA ladder, dNTPs and proteinase K (Roche, Basel, Switzerland);RNase-Free DNase Set, RNAesy Kit, MiniPrep Kit and MidiPrep Kit (Qiagen, Hilden, Germany); Gremlin-1 coding sequence and 3´utr pBluescript SK II (+) vector (StrataGene, La Jolla, USA); control pcDNA5/FRT/T0 vector (Invitrogen GmbH, Karlsruhe, Germany); competent Escherichia coli DH5α (Stratagene, Heidelberg, Germany);primers for cloning and PCR (Eurofin MWG Operon, Ebersberg, Germany).
Cloning, expression and purification of fusion protein mGremlin-1-Fc and Fc. A murine Gremlin-1-Fc fusion protein consisting of the leader sequence of IgG kappa (IgK), the fragment crystallisable region of human IgG2 (Fc) with the hinge region, and the extracellular domain of murine Gremlin-1 (mGremlin-1-Fc) and a corresponding IgG2 Fc control protein without the mGremlin-1 were designed (Supplementary Fig. 4) and cloned into the pcDNA5/FRT vector (Invitrogen) for expression under control of the CMV promoter. For amplification, competent Escherichia coli DH5α (Stratagene) were transformed with these expression vectors, and the amplified plasmids were isolated using Midi Prep Kit (Qiagen). The identity of the constructs was verified by DNA sequencing (Eurofin MWG Operon, Germany).
Stable cell lines for the expression of mGremlin-1-Fc fusion protein and Fc control protein, respectively, were generated with the Flp-InTM system and Flp-InTM-CHO cells, respectively, following the recommendations of the manufacturer (Invitrogen GmbH, Germany). mGremlin-1-Fc stable expressing cell lines were cultivated in HAM’s-F12 medium (Biochrom AG, Germany) containing 10% FCS, 1% P/S and 250 μg/ml Hygromycin B at 37°C and 5% CO2. For lab scale expression of mGremlin-1-Fc and Fc, 3x106 cells were seeded on T-160 cell culture flasks in serum free CHO (A) medium (Gibco, Scotland). After 8 days of incubation, cell culture supernatants of the stable Flp-InTM CHO expression cells producing mGremlin-1-Fc or Fc, respectively, were collected. The supernatants were purified using Protein G agarose beads (Pierce, USA) following the recommendations of the manufacturer. Purified protein were pooled, dialysed in PBS overnight at 4°C, and frozen at –80°C until used. Protein concentrations were determined with the human IgG ELISA (Helvetica Healthcare, Switzerland) following the instructions of the manufacturer. Purity and function of mGremlin-1-Fc was verified by SDS-PAGE and SPR as indicated.
ELISA and Arrays. The following proteins were quantified by ELISA: human MIF (Ray Biotech Inc., GA, U.S.A.), murine MIF (Uscn Life Science Inc., China), TNF-α (R&D Systems Inc., MN, U.S.A.), Gremlin-1 (Uscn Life Science Inc., China), and human IgG (Helvetica Healthcare, Geneva, Switzerland). Cell culture supernatants were analyzed for various chemokines/cytokines using capture antibody based array kit (Human Cytokine Array Kit, Panel A (R&D Systems, MN, U.S.A.).
Isolation of human and murine monocytes. Human monocytes were isolated from peripheral venous blood of healthy donors by differential centrifugation through Ficoll gradient and adherence as described previously 1;3.
Murine monocytes were isolated from spleens of wild type (C57Bl/6J) and MIF-/- mice. MIF-/- mice were obtained from Prof. Jürgen Bernhagen, Institute of Biochemistry & Molecular Cell Biology, RWTH Aachen University, Germany.
The spleens were extruded trough a 40 µm filter (cell strainer, BD Falcon) on the top of a 50 ml Falcon tube. The filter was rinsed once with 10 ml PBS. Then the falcon tube was centrifugated for 10 min at 2100 rpm. Afterwards the supernatant was discarded. To remove erythrocytes the pellet was lysed using Ammonium-Cl solution (5min, 4°C). After a further centrifugation for 10 min at 1600 rpm the pellet was seeded out into petri dishes (BD, Falcon) in RPMI medium (RPMI 1640, + 10%FCS, +1% L-Glut, 1%P/S, PAA) . At the next day the supernatant was discarded and the petri dishes were washed once with PBS. To remove the attached monocytes were trypsinated. After centrifugation for 10min at 1600 rpm, the cells were counted and seeded out in 96-well plates (5 x 104/well).
Secretion of Gremlin-1 and MIF from activated monocytes.Human monocytes (3 x 105/well) were stimulated for 24h with the following cytokines: oxLDL (50µg/mL, Kalen Biomedical), LPS (1µg/mL, Sigma-Aldrich), TNF (100 ng/mL,tebu-bio), and medium control. After incubation time as indicated cell supernatants were collected and cell lysates were generated. For the generation of cell lysates, the cells were once washed with PBS. The cells were lysed afterwards for 45 min on ice using RIPA buffer (150 mM NaCl, 50 mM tris, 0.1% tris, 0.5% sodium-desoxycholat, 1% Triton X-100, protease inhibitors (1:25); pH 8.0). The lysate was centrifugated (15 min, 4°C, 10,500 x g) and the supernatant was collected and frozen at -20°C. Cell lysates and culture supernatants were analysed by immunoblotting using antibodies to MIF (polyclonal, R&D Systems, Wiesbaden, Germany), to Gremlin-1 (Abgent, San Diego, CA, polyclonal, clone RB2060) and as a loading control beta-Actin (Abcam, polyclonal, Ab1801) and by ELISA (MIF, Ray Biotech Inc., GA, U.S.A.; Gremlin-1, Uscn Life Science Inc., China).
In vitro model of macrophage and foam cell generation. The effects of Gremlin-1 on formation of macrophages/foam cells were studied in vitro as described 3;4. The formation of macrophages and foam cells were analyzed in the absence or presence of Gremlin-1 or Gremlin-2 (0.5µg/ml) by morphological assessment and CD68-staining at time points as indicated 3;4. For detection of chemokines/cytokines supernatants were analyzed using a human cytokine array kit (R&D, panel A) and by MIF ELISA.
In a second approach human monocytes were seeded out (5x104 cells/well) in foam cell medium (RPMI 1640, + 1% Sodium Pyruvate, +1% L-Glut., 1%P/S, 1%NEAA, 20% human serum) to generate foam cells. At day 1, 5 and 10 the cells were stimulated with different cytokines as following: Gremlin-1 (0.5 µg/ml, R&D Systems, Wiesbaden, Germany, 5190-GR), MIF (10 ng/ml, R&D Systems, Wiesbaden, Germany, 298-MF-002), Gremlin-2 (1 µg/ml, R&D Systems, Wiesbaden, Germany, 2069-PR), Gremlin-1 (0.5 µg/ml) + MIF (10 ng/ml) (10 min preincubated), blocking MIF antibody (25 µg/ml, provided by Prof. Jürgen Bernhagen, RWTH Aachen University, Germany, clone IIID9), isotype control: normal mouse IgG (25 µg/ml) (Santa Cruz Biotechnologies, sc-2025) or medium alone. At day 4 in every setting the cells were feeded with acLDL (80 µg/ml, Kalen Biomedical). The cell culture supernatant of day 10 was analyzed using the human TNF-α ELISA (Kit R&D Systems Inc., MN, U.S.A.). The cells were photographed and counted at day 10 to get the number of foam cells per visual field.
To generate murine foam cells the monocytes were seeded out (5x104 cells/well) in RPMI 1640, + 10 FCS, +1% L-Glut., 1%P/S. At day 1 and 4 the cells were stimulated with different cytokines as follows: Gremlin-1 (0,5 µg, R&D Systems, Wiesbaden, Germany, 956-GR/CF), MIF (10 ng, 50 ng/ml, 50 ng/ml, R&D Systems, Wiesbaden, Germany, 1978-MF/CF), Gremlin-1 (0.5 µg/ml) + MIF (10 ng/ml) (10 min preincubated), blocking MIF antibody (25 µg/ml, provided by Prof. Jürgen Bernhagen, RWTH Aachen University, Germany, clone IIID9), isotype control: normal mouse IgG (25 µg/ml) (Santa Cruz Biotechnologies, sc-2025) or medium alone. The cell culture supernatant of day 8 was analyzed using the mouse TNF-α ELISA Kit (R&D Systems Inc., MN, U.S.A.). The cells were photographed and counted at day 7 to get the number of foam cells per visual field.
Sudan red staining. Cells were washed with PBS before each incubation step, fixed with 2% formaldehyde solution (20 min), and incubated with 0.5% Sudan red (Sigma; 20 min). Nuclei were
counterstained with hematoxyline solution (Sigma) for 5 min and analyzed by standard microscopy.
Macrophage cholesterol loading and efflux.
Experiments were performed according to a previously published protocol 5. Briefly, primary mouse peritoneal macrophages, elicited with thioglycollate, were allowed to adhere in 48-well plates in RPMI 1640 medium supplemented with 1% FBS and 100 U/ml penicillin/ 100 µg/ml streptomycin for 4h. Following a wash with HBSS to remove non-adherent cells, loading was carried out using RPMI 1640 containing 50 µg/ml acetylated LDL and 3 µCi/ml 3H-cholesterol (NEN Life Science Products) for 20h. For loading experiments the respective chemokines were added 1h before loading as indicated and these macrophages were harvested to determine uptake of radioactivity following the loading period as detailed below. To determine efflux, after an additional wash with HBSS the macrophages were equilibrated in RPMI 1640 containing 2% BSA for 18h. Then HDL (50 µg protein/ml in RPMI 1640) isolated by ultracentrifugation (1.063 < d < 1.21) from a pool of healthy, normolipidemic donors was added for 5h in the presence of the respective chemokines as indicated, which were added 1h before the efflux experiment and kept present throughout. After this period, radioactivity in the medium was determined by liquid scintillation counting (Beckman LS6500, Beckman Instruments, Palo Alto, CA) after tabletop centrifugation to pellet cellular debris. To determine radioactivity still present in the cells, plates were incubated for 30 min at room temperature with 0.1M NaOH followed by liquid scintillation counting. Efflux is expressed as the percentage of counts recovered from the medium relative to the initially present total dose of radioactivity (counts recovered within the medium added to the counts recovered from the cells). To correct for unspecific non-HDL-mediated efflux, values from control cells without added HDL were subtracted from all respective experimental values. All experiments were performed in triplicates. Statistics on macrophage cholesterol loading and efflux were performed using Kruskal Wallis - Conover test.
Immunoprecipitation.Human aortic endothelial cells (HAEC) were grown in complete endothelial cell basal medium containing growth factors, 10% human serum and 10% FCS, until confluency. For generation of cell lysate, the cells were once washed with PBS. Then the cells were lysed for 45 min on ice using IP buffer (15 mM tris-hydrochlorid, 155 mM NaCl, 1 mM EDTA, 0.08 mM sodium azide, proteinase inhibitor (1:25)). The lysate was centrifugated (5 min, 4°C, 2700 x g) and the supernatant was collected and frozen at -20°C until further processing. The protein concentration in the cell lysate was determinated using Biorad Protein Assay with protein standard BSA (Sigma Aldrich) and measurement of light absorption in the ELISA Reader (Biorad Model 550) at 405 nm. For precipitation, samples were incubated over night at 4°C under rotation with sepharose beads (GE Health Care) and 5 µg of the following antibodies (Santa Cruz Biotechnology): human MIF IgG goat antibody (polyclonal), human Gremlin-1 IgG rabbit antibody (polyclonal), isotype control IgG goat (polyclonal) and isotype control IgG rabbit (polyclonal). Before precipitation, sepharose was washed three times with IP buffer. Samples were centrifugated (20,000 x g, 5 min, 4°C), washed three times, and pre-heated for 10 min to 95°C after dilution with buffer (2x Lämmli buffer + 5% Mercaptoethanol). 1.25 mg/ml protein per sample was separated on a 15% SDS-polyacrylamide gel electrophoresis (PAGE) gel for MIF detection or on a 10-20% Novex gradient gel (Invitrogen) to determine gremlin-1. Blotting of proteins onto a polyvinylidene difluoride membrane (Immibilon, Millipore) was performed using a SemiDry Tranfer Cell System (Peqlab). Membranes were stained with rabbit polyclonal anti-human Gremlin-1 (Abgent, San Diego, CA) and goat polyclonal anti-human MIF antibody (R&D Systems, Wiesbaden, Germany).
SDS-PAGE and immunoblotting. For immunoblot analysis lysates of human monocytes, macrophages, foam cells, HUVECs, HAECs and supernatant from mGremlin-1-Fc producing CHO cells were separated on SDS-PAGE according to Lämmli 7. Protein concentration was determinated using Biorad Protein Assay with protein standard BSA (Sigma) and measurement of absorption at 405 nm. The samples were diluted with Lämmli buffer (1x, +5% mercaptoethanol) and heated for 10 min up to 95°C. Thirty micrograms of total protein were seperated on a 15% SDS-polyacrylamide gel electrophoresis (PAGE) (Invitrogen). Blotting of the protein onto a polyvinylidene difluoride membrane (PVDF, Immibilon, Millipore) was performed using Semi Dry Tranfer Cell System (Peqlab). We used rabbit polyclonal anti-human Gremlin-1 antibody (clone RB2060, Abgent, San Diego, CA) and goat polyclonal anti-human MIF antibody (R&D Systems, Wiesbaden, Germany) for detection of Gremlin-1 and MIF. Rabbit polyclonal mouse anti-Gremlin-1 antibody (Abcam, Cambridge, UK) and anti-human IgG-HRP (polyclonal) (Jackson Immuno Research, PA, USA) were used to detect the mGremlin-1-Fc fusion protein. ß-actin antibody (Sigma-Aldrich, Steinheim, Germany) and α-actin antibody (polyclonal) (Abcam, Cambridge, UK) was used as internal loading control. For detection of antibody binding, corresponding secondary fluorescence labeled antibodies and the Odyssey infrared imaging system (LI-COR, Bad Homburg, Germany) were used. Bands were quantified using ImageJ software (National Institutes of Health, USA).
Isolation of RNA. For total RNA extraction from arteries, the tissue was fixed and lysed directly after mouse sacrification using TRI reagent (Sigma) according to the user´s manual provided by the manufactorer. RNA was further treated with RNase-free DNase for 30 min at 37°C to remove genomic DNA. One microgram of total RNA was then reverse transcribed to cDNA using the MMLV reverse transciptase (Invitrogen). cDNA was used as template for PCR analysis of gremlin-1 and gremlin-2 expression. For total RNA extraction from human monocytes (buffy coat derived) cells were pelleted and mRNA was isolated using the RNeasy Kit (Qiagen).
RT-PCR. 200 nanogram or 1 µg of total RNA was then reverse transcribed to cDNA using the c DNA synthesis Kit (IMIIProm Reverse Transkriptase synthesis Kit, Promega). cDNA was used as template for PCR analysis of MIF and Gremlin-1 expression. As a loading control Aldolase was used.
RNA was reverse transcribed into cDNA from monocytes, macrophages and foam cells as follows: 200 ng or 1 µg of RNA was mixed with 100 ng of oligo(dT)15 and incubated for 5 min at 65°C. One mM dNTPs, 60 mM KCl, 15 mM Tris-Cl, pH 8.4, 3 mM MgCl2, 0.3% Tween 20, 10 mM mercaptoethanol, 10 U RNAsin (Promega), and 100 U Mo-MLVRT (Invitrogen) were added, and the mix was incubated for 55 min at 37°C followed by enzyme inactivation for 5 min at 95°C. Expression of Gremlin-1, MIF and Aldolase was examined by semi-quantitative PCR. PCR was performed with 20 ng of cDNA using TaqDNA polymerase (Promega) for 24 to 35 cycles (1 min at 95°C, 1 min at 60 to 65°C, and 1 min at 72°C).
Primer sets were designed to cross intron/exon bounderies to exclude the possibility of amplification of contaminating genomic DNA. Primer sets used were as follows: mouse Gremlin-1, forward primer, 5’- cgggatccacagcgaagaacctgaggacc-3’, reverse primer, 5’- gctctagacaggctgaatgtgcccgctttg-3’; mouse aldolase: forward primer, 5’- agctgtctgacatcgctcaccg -3’, reverse primer, 5’- cacatactggcagcgcttcaag -3’.
Primer sets used to clone mouse Gremlin-1 into pBluescript SK II (+) (StrataGene) for in situ hybridization probe synthesis, were as follows: forward primer, 5’- ggaattcccagtccatgctcttgccaagatg-3’, reverse primer, 5’- cgggatccctagcttaacctttcctggtg -3’.
Primer sets for human MIF were as follows: forward primer, 5´-CTCTCCGAGC TCACCCAGCAG-3´, reverse primer, 5´-CGCGTTCATGTCGTAATAGTT-3´; human Gremlin-1: forward primer, 5´-GTATGAGCCGCACAGCCTACA-3´, reverse primer, 5´-CTCGCTTCAGGTAT TTG CGCT-3´; human Aldolase-b: forward primer, 5´-GCCACTTCTCAACCTCAATGC-3´, reverse primer, 5´-TCTCCTTCCCAACCTACCAC-3´. In situ hybridization (ISH). Aortic tissue from 8 and 16-weeks-old ApoE-/- and wild type mice, fed for 4 and 12 weeks with a high cholesterol diet, was snap frozen. ISH was performed on cryostat sections, 8-10 µm in diameter each, according to the method decribed 8;9, with the following modifications: Gremlin-1 antisense digoxigenin-labeled RNA probes were synthesized after cloning mouse gremlin-1 coding region and 3´utr into pBluescript SK II (+) vector (StrataGene) via BamHI and XbaI restriction sites, linearization of the vector using BamHI, and amplification of RNA probes using T3 RNA polymerase and DIG RNA labeling Kit (Roche). Gremlin-1 sense digoxigenin-labeled RNA probes served as controls. Cryo sections were dried, fixed in 4% paraformaldehyde 1´ PBS (phosphate-buffered saline), brought stepwise to pure ethanol, and frozen at -20°C. Then the sections were rehydrated stepwise in ethanol/PBS and finally put back in 100% PBS. Tissue sections were then treated with proteinase K. The reaction was stopped by rinsing sections with PBS. The tissue was postfixed in 4% paraformaldehyde 1´ PBS for 20 min and rinsed in PBS 5 times for 5 min each afterwards. The embryos were prehybridized at least 1 hour at 70°C in hybridization buffer (50% formamide, 5´ SSC, 50 mg/ml heparin, 500 mg/ml tRNA, 0.1% Tween, 20.9 mM citric acid). The hybridization was done in the same buffer containing 50 ng to 100 ng of probe per section overnight at 70°C. The tissue sections were washed at 70°C for 10 min in (75% hybridization buffer, 25% 2´ SSC), 10 min in (50% hybridization buffer, 50% 2´ SSC), 10 min in (25% hybridization mix, 75% 2´ SSC), 10 min in 2´ SSC, 2 times for 30 min in 0.2´ SSC. Further washes were performed at room temperature for 5 min in (75% 0.2´ SSC, 25% PBS), 5 min in (50% 0.2´ SSC, 50% PBS), 5 min in (25% 0.2´ SSC, 75% PBS), 5 min in PBS, and then 1 hour in (PBS with 2 mg/ml, BSA (bovine serum albumin), 2% sheep serum). Then the sections were incubated overnight at 4°C with the pre-absorbed alkaline phosphatase-coupled anti-digoxigenin antibody (Roche) at a 1/5000 dilution in a PBS buffer containing 2 mg/ml BSA, 2% sheep serum. Finally the slides were rinsed 6 times for 15 min each in PBS at room temperature, followed by 4 to 6 washes with NTMT buffer for 15 min each. Staining was performed using BCIP/NBT in NTMT buffer as decribed 9. When the color was developed, the reaction was stopped in 1´ with PBS. The tissue sections were then refixed in 4% paraformaldehyde for 20 min, clarified in PBS, dried and mounted in Permount Mounting Medium (Dako Inc., USA).
Molecular modeling. Molecular models of Gremlin-1 and mGremlin-1-Fc were generated by template LOMETS homology modeling with structure fragmentation and reassembling by replica-exchange Monte Carlo simulations using the I-TASSER standalone package (version 1.1) 10;11;12. The structure of MIF was obtained from Protein Data Bank (1MIF, residues 1-115) 13. Flexible blind docking of either Gremlin-1 or mGremlin-1-Fc to MIF was performed using Hex 6.3 software 14;15. The original parameters of blind docking were used in combination with an evaluation algorithm based on binding free energy (G) 16. Three-dimensional models of MIF-Gremlin-1 and MIF/mGremlin-1-Fc complexes with surface charge distribution were created with Molegro Virtual Docker 5 17.
Analysis of protein-protein interactions. For surface plasmon resonance analysis MIF was covalently bound to a linear polycarboxylated three dimensional hydrogel sensor SD C80 m by using the amine coupling kit EDC/NHS (both XanTec Bioanalytics, Münster, Germany). The sensor chips were coated with an 800 nM MIF solution. For the determination of the interaction between Gremlin-1 and MIF proteins were dissolved in binding buffer (10 mM MOPS, 0.3 mM EDTA, 1.43 mM β-mercaptoethanol, 150 mM NaCl, 0.001 % (v/v) NP40 (Igepal CA-630), adjusted to pH 7.0, and applied to the detector cell (IBIS dual channel optical measuring system, XanTec Bioanalytics) at 20°C. Interaction was analyzed for 300 sec; sensor surfaces were regenerated with a solution containing 4 mM HCl and 1 M NaCl. Nonlinear regression of the primary data collected during the association phase was used to calculate the kinetic parameters. Apparent rate constants for the single class interaction sites were determined, and the apparent KD were calculated.
The interaction between mGremlin-1-Fc and MIF was analysed by a dynamic light scattering method. The average hydrodynamic radii (ravg) of the molecules were used as measurable parameters, because they represent the increase of total molar concentration of the self-associating system 18. By measuring ravg as a function of molar concentration and fitting the experimental data to equilibrium relations, we determined the KD of the protein-protein complex18. The average hydrodynamic radii of either Gremlin or mGremlin-1-Fc and MIF mixtures, where the Gremlin proteins remained fixed at 10 nM and MIF concentrations varyied (1-500 nM), were measured (Zetasizer Nano ZS, Malvern Instruments, UK). The KD of the Gremlin and MIF protein complex was calculated by fitting the data to a one binding site model (A + B ↔ AB). Due to dimerization of the Fc fusion proteins, for the KD of the mGremlin-1-Fc and MIF protein complexes was calculated according to a two binding sites model (A + A + B ↔ ABA).
Immunohistochemistry and Histology. For immunostaining and confocal microscopy human cells were fixed with 2% formaldehyde and permeabilized with 0.2% Triton X-100. Human cells were stained with rabbit polyclonal anti-Gremlin-1 (clone RB2060, Abnova), rabbit polyclonal CD68 antibody (Abbiotec), polyclonal anti-MIF antibody (R&D Systems, Wiesbaden, Germany), or isotype control antibodies for 1 hour. For immunohistochemical analysis, aortic tissue samples of ApoE-/- mice were embedded in paraffin and cut into 5µm sections. The sections were deparaffinized, heated in citrate buffer (0.1M 8.2% tri sodium citrate dehydrate (Applichem, Darmstadt, Germany), 1.8% 0.1M citric acid (Sigma-Aldrich, Steinheim, Germany); pH6) in a microwave oven at 360 Watt for a total of 15 minutes. They were immunostained either with avidin-biotin complex (ABC) method (LSAB + System HRP, Dako, Belgium) or with immunofluorescence using primary antibodies and secondary Alexa Fluor labeled antibodies (Molecular Probes). Mouse aortic tissue was stained using the following primary antibodies: rabbit polyclonal anti-Gremlin-1 (clone RB2060, Abnova, Germany), rabbit polyclonal mouse anti-CD68 (Abbiotec, U.S.A.), rat monoclonal macrophage antibody (clone M3/84, Mac-3, BD Biosciences), goat polyclonal anti-MIF (R&D Systems), anti-TNF-alpha (rabbit anti mouse, polyclonal) from Abcam, Cambridge, UK, and isotype control antibodies according to standard protocol. Corresponding biotinylated secondary antibodies (DAKO) were used. Adjacent sections were stained with haematoxylin and eosin to visualize the corresponding structures of the plaques. To detect binding of the recombinant fusion protein mGremlin-1-Fc or control Fc, anti-human IgG, Fcγ fragment specific antibody (Jackson ImmunoResearch Laboratories, U.S.A.) was used. Unspecific binding was prevented by protein block (serum free, DAKO) or bovine serum albumin (3%, 30 min). Samples were covered with mounting medium (DAKO) and analyzed by compound microscopy (Axiovert 200, Zeiss, and Nikon Digital Sight DS-U1, Nikon, Japan) or confocal microscopy (LSM510, Zeiss and Leica TCSSP, Leica Microsystems).
H&E staining was performed according to standard protocol.
Immunofluorescence staining of mouse tissue followed by confocal microscopy. The paraffin sections were deparaffinized and heated in citrate buffer. After cooling and washing with PBS-T (containing 0.05% Tween; Merck, Hohenbrunn, Germany) the sections were separated using a lipid pen (Dako, Glostrup, Denmark) and inkubated with serum-free protein block (Dako, Carpinteria, USA) for 30 minutes at room temperature in a humid chamber. The Gremlin-1 antibody (Santza Cruz Biotechnology, Sanra Cruz, USA) was diluted 1/10, IgG control (Dako, Glostrup, Denmark) was used at the same concentration for an incubation over night at 4°C in a humid chamber. After washing with PBS-T the secondary antibody (dilution 1/100) Alexa 647 goat-anti-rabbit (Invitrogen, Darmstadt, Germany) was incubated for 2h at room temperature in a humid chamber. Followed by another washing-step, the sections were incubated for 30 minutes using serum-free protein block (Dako, Carpinteria, USA) in a humid chamber at room temperature. The CD68 antibody (abbiotec, San Diego, USA; dilution 1/10) and the IgG control (Dako, Glostrup, Denmark) at the same concentration were incubated over night at 4°C in a humid chamber. After washing with PBS-T 80 µl of the secondary antibody Alexa 488 goat-anti-rabbit (Invitrogen, Darmstadt, Germany; dilution 1/100) were incubated for 2h at room temperature in a humid chamber. The sections were washed in PBS-T and stained for 15 minutes with Draq5 (biostatus limited, Shepshed, UK; dilution 1/2000). After washing with PBS the sections were covered with mounting medium (Dako, Carpinteria, USA) and cover slides (R Langenbrinck, Emmendingen, Germany). Confocal microscopic analysis and capture of images were carried out with Zeiss LSM5 EXCITER Confocal Laser Scanning Microscope (Carl Zeiss Micro Imaging, Jena, Germany) with an A-Plan 40x and 63x ocular.
Mouse model of atherosclerosis and disease progression. Male ApoE-/- mice (B6.129P2-ApoetmUnc1) were purchased from Jackson laboratories. Starting at the age of 4 weeks the mice were fed a cholesterol-rich diet (1.25% cholesterol, 0.2 % cholate, Provimi Kliba AG, Kaiseraugst, Switzerland,) throughout the experiments. C57BL/6J wild-type mice (Charles River Laboratories) served as controls. To assess Gremlin-1 expression in atherosclerotic arteries, mice received this diet for 4 or 12 weeks. At the end of this period mice were sacrificed and the main arteries (aortic arch, aorta thoracica and aorta abdominalis) were collected, conserved in trizol, Sigma, Germany, and frozen at -80° Celsius for later total RNA extraction or embedded in paraffin for immunohistochemistry. To study the effect of mGremlin-1-Fc, 4-weeks old ApoE-/- mice were fed with a cholesterol-rich diet. At the age of 10 weeks mGremlin-1-Fc (1μg/g body weight; n = 8) or control Fc (equimolar, n = 8) was administered i.p. thrice weekly for further 4 weeks. Further inactivated (heat inactivated over 10 min at 95oC) mGremlin-1-Fc (1μg/g body weight; n = 3) was used and compared to active mGremlin-1-Fc (1μg/g body weight; n = 3) in a second independent experiment following the same protocol as described above. Thereafter, mice were sacrificed in general anesthesia. All animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996) and the German law for the welfare of animals. Animal studies were approved by the local authorities (regional board Tübingen).
After euthanasia of the animals, the vessels were perfused with saline in situ followed by perfusion 4% paraformaldehyde through the left ventricle. Subsequently, the vessels were transfered into 4% PFA for fixation. Arteries from mice with long time mGremlin-1-Fc treatment were immediately stained with oil red O. Afterwards the vessels were photographed with a Zeiss Axiovert 200 and the Axiocam MRc5 (Zeiss) using the Axio Vision software. Plaque areas and the total vessel area were determined and the relative plaque extension was expressed as percentage of the total vessel area. Afterwards, vessels were embedded in paraffin, cut in 5μm sections and stained with HE or immunostained for the specific visualization of macrophages, foam cells and T-lymphocytes using anti-CD68, anti-Gremlin-1, anti-MIF, anti-Mac3, anti-TNF-alpha and anti-Fc.After staining and immunostaining, sections were viewed under a Nikon Compound Microscope (Digital Sight DS-U1, Nikon) and digital images were taken with a Nikon camera (Digital Sight DS-5M, Nikon Japan) at a resolution of 2592 x 1944 pixels. The imaging software NIS Elements Basic Research (Nikon, Germany) was used. Macrophage and T-cell numbers were determined by counting blue nuclei in positively stained cells within lesions. On tissue sections of the aortic arch 9 plaques on random sections were analysed per mouse at 10x magnification regarding macrophage/foam cell number (CD68 positive cells), total cell number, macrophage/foam cell – cell area and total plaque area.
Pharmacokinetics. C57BL/6Jmice received mgremlin-1-Fc (1 µg/g body weight) by intraperitoneal injection. At indicated time intervals (before, 6, 24 and 48 hrs after administration), 100 µL blood was collected in a tube containing 10% (v/v) 20u/ml heparin taken from the retro-orbital plexus under isofluran anaesthesia. mGremlin-1-Fc levels were determined in the plasma using a human IgG ELISA (IgG ELISA Kit, Immunotek, ZMC) according to manufacturer’s instruction.
Statistical Analysis. Data are presented as mean ± SEM. Statistical analyses on continuous data were performed using unpaired two-tailed Student’s t-test. Correlations were analyzed using Spearman's rank correlation coefficient. A p-value of 0.05 or less was regarded as significant.
Supplementary Figure Legends Supplements to Figure 1: Gremlin-1 is expressed in early and advanced atherosclerotic lesions in ApoE-/- mice. Aortic tissue was obtained from ApoE-/- mice at the age of 10 to 20 weeks. Expression of Gremlin-1 in aortic tissue on RNA-level was assessed using in situ hybridization. Expression of Gremlin-1 on protein level was assessed by immunostaining using mouse anti-Gremlin-1. Macrophages and foam cells were detected using anti-CD68 mAb. Expression was analyzed by compound and confocal microscopy.
Supplements to Figure 2: Characterization of macrophages, differentiated from human CD34+ cells and monocytes, influence of different concentrations of Grem1 on foam cell formation, impact of MIF and Grem1 on murine monocyte differentiation to macrophages/foam cells, and the effects of Grem1 on cholesterol uptake and efflux. (a) Representative phase contrast image (DIC) of CD34+ cells and foam cell generation. Sudan red III marks large granular and lipid-rich cells. May-Gruenwald staining revealed a nonsegmented nucleus surrounded by a large cytoplasm with enhanced granularity. Naphtyl acetate esterase (NSE) and CD68 immunostaining indicates macrophage/monocyte lineage; TEM image of foam cells, showing a typical morphology and multiple vacuoles. (b) Representative DIC images of a co-culture of monocytes and thrombocytes, in which monocytes had differentiated into macrophages and foam cells (larger round cells) over 15 days. The upper three panels show an inhibited differentiation into macrophages and foam cells under the influence of Grem1 in the concentration of 0.5 µg/mL, 1 µg/mL and 2 µg/mL. The effects of these three concentrations of Grem1 were comparable regarding their effect on the inhibition of foam cell generation. Medium, Grem1 protein diluent (HCl 0.4 mM) and a protein control (IgG) served as negative controls. (c) + (d) Murine monocytes were isolated from wild type mice and from MIF-/- mice. Macrophages/foam cells were differentiated from these isolated mouse monocytes. Formation of macrophages/foam cells was evaluated over time through cell counting in the presence of Grem1 (0.5 µg/mL), MIF (10 ng/mL), Grem1 (0.5 µg/mL) + MIF (10 ng/mL). Representative photomicrographs are shown. Data represent the mean ± SEM (** = p<0.01). (e) + (f) Macrophage uptake and efflux of cholesterol under the influence of Gremlin-1 (2 µg/mL), MIF (250 ng/mL), and Gremlin-1 + MIF. All experiments were performed in triplicates. Data represent the mean ± SEM (* = p<0.05).
Supplements to Figure 5: Effects of mGremlin-1-Fc or heat-inactivated mGremlin-1-Fc onatherosclerosis in ApoE-/- mice. (a) mGremlin-1-Fc reduces atherosclerotic plaque extension compared to inactivated mGremlin-1-Fc. Representative photographs of Sudan red III–stained aortic arches and thoracic aortas are shown. The mean percentages of plaque areas in relation to total vessel areas of all mice are shown with mean ± SEM (n=3 for each group). * = p<0.05, ** = p<0.01, statistical significance. (b) Representative immunostainings together with their isotype- and PBS-control (only secondary antibody) stainings of atherosclerotic aortic tissue of mGremlin-1-Fc and Fc-treated ApoE-/- mice are shown.
Schmidt R, Bultmann A, Ungerer M, Joghetaei N, Bulbul O, Thieme S, Chavakis T, Toole BP, Gawaz M, Schomig A, May AE. Extracellular matrix metalloproteinase inducer regulates matrix metalloproteinase activity in cardiovascular cells: implications in acute myocardial infarction. Circulation. 2006;113:834-841.
Colognato R, Slupsky JR, Jendrach M, Burysek L, Syrovets T, Simmet T. Differential expression and regulation of protease-activated receptors in human peripheral monocytes and monocyte-derived antigen-presenting cells. Blood. 2003;102:2645-2652.
Daub K, Siegel-Axel D, Schonberger T, Leder C, Seizer P, Muller K, Schaller M, Penz S, Menzel D, Buchele B, Bultmann A, Munch G, Lindemann S, Simmet T, Gawaz M. Inhibition of foam cell formation using a soluble CD68-Fc fusion protein. J Mol Med (Berl). 2010;88:909-920.
4. Daub K, Langer H, Seizer P, Stellos K, May AE, Goyal P, Bigalke B, Schonberger T, Geisler T, Siegel-Axel D, Oostendorp RA, Lindemann S, Gawaz M. Platelets induce differentiation of human CD34+ progenitor cells into foam cells and endothelial cells. FASEB J. 2006;20:2559-2561.
5. Nijstad N, de Boer JF, Lagor WR, Toelle M, Usher D, Annema W, van der Giet M, Rader DJ, Tietge UJ. Overexpression of apolipoprotein O does not impact on plasma HDL levels or functionality in human apolipoprotein A-I transgenic mice. Biochim Biophys Acta. 2011, 1811:294-299.
6. Stellos K, Langer H, Daub K, Schoenberger T, Gauss A, Geisler T, Bigalke B, Mueller I, Schumm M, Schaefer I, Seizer P, Kraemer BF, Siegel-Axel D, May AE, Lindemann S, Gawaz M. Platelet-derived stromal cell-derived factor-1 regulates adhesion and promotes differentiation of human CD34+ cells to endothelial progenitor cells. Circulation. 2008;117:206-215.
7. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680-685.
8. Beck H, Acker T, Puschel AW, Fujisawa H, Carmeliet P, Plate KH. Cell type-specific expression of neuropilins in an MCA-occlusion model in mice suggests a potential role in post-ischemic brain remodeling. J Neuropathol Exp Neurol. 2002;61:339-350.
9. Thisse C, Thisse B, Schilling TF, Postlethwait JH. Structure of the zebrafish snail1 gene and its expression in wild-type, spadetail and no tail mutant embryos. Development. 1993;119:1203-1215.
10. Roy A, Kucukural A, Zhang Y. I-TASSER: a unified platform for automated protein structure and function prediction. Nat Protoc. 2010;5:725-738.
11. Zhang Y. I-TASSER server for protein 3D structure prediction. BMC Bioinformatics. 2008;9:40.
12. Zhang Y. I-TASSER: fully automated protein structure prediction in CASP8. Proteins. 2009;77 Suppl 9:100-113.
13. Sun HW, Bernhagen J, Bucala R, Lolis E. Crystal structure at 2.6-A resolution of human macrophage migration inhibitory factor. Proc Natl Acad Sci U S A. 1996;93:5191-5196.
14. Ritchie DW, Kozakov D, Vajda S. Accelerating and focusing protein-protein docking correlations using multi-dimensional rotational FFT generating functions. Bioinformatics. 2008;24:1865-1873.
15. Ritchie DW, Venkatraman V. Ultra-fast FFT protein docking on graphics processors. Bioinformatics. 2010;26:2398-2405.
16. Hetenyi C, van der SD. Efficient docking of peptides to proteins without prior knowledge of the binding site. Protein Sci. 2002;11:1729-1737.
17. Thomsen R, Christensen MH. MolDock: a new technique for high-accuracy molecular docking. J Med Chem. 2006;49:3315-3321.
18. Hanlon AD, Larkin MI, Reddick RM. Free-solution, label-free protein-protein interactions characterized by dynamic light scattering. Biophys J. 2010;98:297-304.
19. Massberg S, Konrad I, Bultmann A, Schulz C, Munch G, Peluso M, Lorenz M, Schneider S, Besta F, Muller I, Hu B, Langer H, Kremmer E, Rudelius M, Heinzmann U, Ungerer M, Gawaz M. Soluble glycoprotein VI dimer inhibits platelet adhesion and aggregation to the injured vessel wall in vivo. FASEB J. 2004;18:397-399.
20. Massberg S, Gawaz M, Gruner S, Schulte V, Konrad I, Zohlnhofer D, Heinzmann U, Nieswandt B. A crucial role of glycoprotein VI for platelet recruitment to the injured arterial wall in vivo. J Exp Med. 2003;197:41-49.
21. Massberg S, Konrad I, Schurzinger K, Lorenz M, Schneider S, Zohlnhoefer D, Hoppe K, Schiemann M, Kennerknecht E, Sauer S, Schulz C, Kerstan S, Rudelius M, Seidl S, Sorge F, Langer H, Peluso M, Goyal P, Vestweber D, Emambokus NR, Busch DH, Frampton J, Gawaz M. Platelets secrete stromal cell-derived factor 1alpha and recruit bone marrow-derived progenitor cells to arterial thrombi in vivo. J Exp Med. 2006;203:1221-1233.