Adhesion Transendothelial Migration Without Affecting VCAM-1 Inhibits

An Antibody to the Sixth Ig-like Domain of VCAM-1 Inhibits Leukocyte Transendothelial Migration without Affecting Adhesion This information is current as of September 28, 2021. Sukmook Lee, Il-Hee Yoon, Aerin Yoon, Joan M. Cook-Mills, Chung-Gyu Park and Junho Chung J Immunol published online 1 October 2012 http://www.jimmunol.org/content/early/2012/10/01/jimmun ol.1103803 Downloaded from

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 28, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published October 1, 2012, doi:10.4049/jimmunol.1103803 The Journal of Immunology

An Antibody to the Sixth Ig-like Domain of VCAM-1 Inhibits Leukocyte Transendothelial Migration without Affecting Adhesion

Sukmook Lee,*,1,2 Il-Hee Yoon,*,†,1 Aerin Yoon,‡ Joan M. Cook-Mills,x Chung-Gyu Park,†,3 and Junho Chung‡,3

VCAM-1 plays a key role in leukocyte trafficking during inflammatory responses. However, molecular mechanisms underlying this function have not been clearly elucidated. In this study, using phage display technology, we developed a rabbit/human chimeric VCAM-1 Ab, termed VCAM-1 domain 6 (VCAM-1-D6), which specifically recognizes aa 511–599 within the sixth Ig-like domain. We report that the VCAM-1-D6 Ab blocked U937 cell transmigration across activated HUVECs but did not alter adhesion of

U937 cells to the HUVECs. We also demonstrate that VCAM-1-D6 does not alter TNF-a–stimulated endothelial cell chemokine or Downloaded from cytokine production. Furthermore, through in vivo efficacy testing using a mouse islet allograft model, we demonstrate that VCAM-1-D6 significantly alleviates allograft rejection by blocking leukocyte infiltration to the grafted islets. Taken together, our results suggest that the VCAM-1-D6 Ab may block VCAM-1–mediated inflammation and could be a useful tool in treating inflammatory diseases. The Journal of Immunology, 2012, 189: 000–000.

ascular cell adhesion molecule-1 (also known as CD106) cyte arrest on the endothelium. After the high-affinity adhesion, http://www.jimmunol.org/ is a cell adhesion molecule induced in activated endo- VCAM-1, together with PECAM-1, plays a critical role in leu- V thelial cells. During leukocyte recruitment, VCAM-1 kocyte transendothelial migration (14). binds a4b1 integrin, a leukocyte integrin, in its intermediate- VCAM-1 is a member of the Ig superfamily. The extracellular affinity state and, along with selectins and addressins, plays a region of the full-length form of VCAM-1 contains seven ho- key role in the rolling step of leukocytes on the endothelium (1–3). mologous Ig-like domains. Domains 1 and 4, 2 and 5, and 3 and 6 Selectin binding on leukocytes or activation of chemokine re- are homologous to each other (15–17). There are two human forms ceptors stimulates “outside-in” signals in leukocytes, which and two mouse forms of VCAM-1. Human VCAM-1 (hVCAM-1) increases the affinity of the integrin family of receptors to bind has two splice variants that result in a protein with either seven Ig- endothelial cell adhesion molecules, predominantly ICAM-1and like domains or six Ig-like domains (lacking domain 4) (15, 18). by guest on September 28, 2021 VCAM-1 (4–13). This high-affinity interaction results in leuko- Mouse VCAM-1 (mVCAM-1) also has a full-length form with seven Ig-like domains, as well as a truncated form containing only the first three Ig-like domains. *Cancer Research Institute, Xenotransplantation Research Center, College of Medi- cine, Seoul National University, Seoul 110-799, Korea; †Department of Microbiology All variants of VCAM-1 have been shown to bind to a4b1 and Immunology, Seoul National University, Seoul 110-799, Korea; ‡Department of integrin (19–22). In addition to a4b1 integrin, VCAM-1 can bind Biochemistry and Molecular Biology, Seoul National University, Seoul 110-799, Korea; x other integrins, including a4b7, adb2, aMb2, and a9b1 (23–28). and Allergy-Immunology Division, Northwestern University Feinberg School of Med- icine, Chicago, IL 60611 It is known that a4b1 integrin binds Ig-like domains 1 and 4 (29– 1S.L. and I.-H.Y. contributed equally to this work. 32), and that activated integrins are necessary for firm leukocyte adhesion to the endothelium via Ig-like domains 1 and 4 of VCAM- 2Current address: Scripps Korea Antibody Institute, Hyoja-2-dong, Chuncheon-si, Gangwon-do, Korea. 1 (33). 3C.-G.P. and J.C. codirected this work. Leukocyte binding to adhesion molecules, including VCAM-1, Received for publication January 5, 2012. Accepted for publication August 29, 2012. activates signals within endothelial cells that allow the opening of narrow vascular passageways between endothelial cells or This work was supported by Grant 2011K000737 from the Korea Biotechnology Research and Development Group of Next-Generation Growth Engine Project of through an individual endothelial cell for transendothelial mi- the Ministry of Education, Science and Technology, Republic of Korea, and by a grant gration (34–39). When endothelial cell adhesion molecule sig- from Korea Healthcare Technology Research and Development Project, Ministry of Health and Welfare, Republic of Korea (Project A040004 to C.-G.P.). naling is inhibited, leukocytes bind to the endothelium, but transendothelial migration does not occur (40), and leukocytes are Address correspondence and reprint requests to Prof. Junho Chung or Prof. Chung- Gyu Park, Department of Biochemistry and Molecular Biology, Seoul National Uni- often released from the endothelium to continue in the blood flow versity, Seoul, 110-799, Korea (J.C.) or Department of Microbiology and Immunol- (41, 42). ogy, Seoul National University, Seoul 110-799, Korea (C.-G.P.). E-mail addresses: [email protected] and [email protected] (C.-G.P.) In this study, we developed a chimeric rabbit/human VCAM-1 Ab, specific to Ig-like domain 6, using phage display technology. Abbreviations used in this article: HAEC, human aortic endothelial cell; hTNF-a, human TNF-a; hVCAM-1, human VCAM-1; mTNF-a, mouse TNF-a; mVCAM-1, This Ab, VCAM-1 domain 6 (VCAM-1-D6), inhibits the trans- mouse VCAM-1; MVEC, mouse vascular endothelial cell; rhVCAM-1, recombinant endothelial migration of leukocytes without hindering the leuko- human VCAM-1; rhVCAM-1–Fc, recombinant human VCAM-1 Fc fusion protein; rmVCAM-1, recombinant mouse VCAM-1; rmVCAM-1–Fc, recombinant mouse cyte adhesion process on endothelial cells in vitro. In a mouse VCAM-1 Fc fusion protein; SDF-1a, stromal cell-derived factor-1a; STZ, strepto- islet allograft model, this Ab successfully inhibits transmigration of zotocin; TTBS, 10 mM Tris/HCl, pH 7.5, 150 mM NaCl, and 0.05% (v/v) Tween 20; leukocytes to the graft site and blocks rejection of grafted pan- VCAM-1-D6, VCAM-1 domain 6. creatic b cell islets. Our results suggest that the VCAM-1-D6 Ab Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 may block VCAM-1–mediated inflammation.

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1103803 2 ROLE OF THE SIXTH Ig-LIKE DOMAIN OF VCAM-1

Materials and Methods Immunocytochemistry Construction of the Ab library and selection of binders Immunocytochemistry was performed as described previously (50). In Four New Zealand white rabbits were immunized and boosted four times brief, coverslips were incubated with 1 mg/ml poly-L-lysine for 1 h; then with recombinant hVCAM-1 (rhVCAM-1; R&D Systems, Minneapolis, HUVECs, HAECs, and MVECs were grown on coverslips in the absence MN) with the approval of the Institutional Animal Care and Use Com- or presence of hTNF-a or mTNF-a at the concentration of 20 ng/ml. After mittee of the Seoul National University Hospital. After the final booster rinsing with 1X PBS two times, the cells were fixed with 3.7% (w/v) injection, total RNA was prepared from the spleen and the bone marrow. paraformaldehyde for 30 min at 37˚C. After washing with 1X PBS and cDNA was synthesized using SuperScript First-Strand cDNA synthesis blocking with the 1X PBS containing 5% (w/v) BSA and 0.1% Triton X- system (Invitrogen, Carlsbad, CA) according to manufacturer’s instruc- 100 for 4 h at 4˚C, the cells were incubated overnight at 4˚C with 10 mg/ml tions. A phage-displayed rabbit/human chimeric Fab Ab library was con- VCAM-1-D6 Fab, control Fab, or the mouse anti–hVCAM-1 Ab 51-10C9. structed using pComb3X phagemid vector system as described previously The cells were washed five times with 1X PBS containing 0.05% Triton X- (43). After the library construction, Fab clones were selected from the 100 and then incubated in the same buffer for 1 h with FITC-labeled goat library through six rounds of biopanning as described previously (2). For anti-human Fab or Alexa Fluor 488-labeled goat anti-mouse IgG Ab each round of biopanning, 2.5 mg recombinant mVCAM-1 Fc fusion (1:100), and 2 mg/ml Hoechst (Invitrogen) for 1 h to visualize VCAM-1 protein (rmVCAM-1–Fc; R&D Systems)-coated magnetic beads (Dyna- and the nucleus, respectively. Slides were then examined under a fluores- beads M-270 epoxy; Invitrogen) were used. After the final round of bio- cence microscope (Olympus, Melville, NY). panning, individual phage clones displaying Fab were generated from Real-time interaction analysis colonies grown on output plates and tested for reactivity to hVCAM-1–Fc and mVCAM-1–Fc by phage enzyme immunoassay as described previ- The kinetic parameters of the interactions between either the VCAM-1-D6 ously (4). Fab and VCAM-1, or VCAM-1-D6 IgG and VCAM-1, were determined using the BIAcore system X-100 (Biacore AB, Uppsala, Sweden). In brief, Preparation of VCAM-1-D6 Fab and IgG hVCAM-1–Fc or mVCAM-1–Fc was immobilized on a CM5 dextran Downloaded from sensor chip (Biacore AB) in 10 mM sodium acetate buffer (pH 4.0) at Top10F9 E. coli cells were transformed with phagemid DNA encoding the a flow rate of 5 ml/min using the Amine Coupling Kit (Biacore AB). VCAM-1-D6 Fab. Bacteria were grown in 1 l Luria–Bertani media con- VCAM-1-D6 Fab or IgG in HEPES-buffered saline containing 0.005% taining 50 mg/ml carbenicillin at 37˚C overnight with constant shaking. 3 surfactant P20, 3 mM EDTA, and 0.15 M NaCl were injected over 150 s at The cells were pelleted by centrifugation at 3000 g for 15 min. the flow rate of 30 ml/min at 37˚C. After each analysis, the surface was Supernatants were collected and concentrated 10 times using the Labscale regenerated with 1 M NaCl/50 mM NaOH as described previously (37). TFF System (Millipore, Bedford, MA). The VCAM-1-D6 Fab was purified

Biacore X-100 evaluation software version 1.1 (Biacore AB) was used to http://www.jimmunol.org/ from the concentrated supernatant using hemagglutinin-specific Ab col- calculate K and K constants. umn chromatography as described previously (44, 45). After column on off chromatography, fractions containing Fab were pooled and analyzed by Preparation of VCAM-1 constructs as Fc fusion proteins SDS-PAGE and Coomassie blue staining as described previously (46). The production and purification of VCAM-1-D6 IgG was performed as de- RNA was prepared from HUVECs treated with hTNF-a using TRIzol scribed previously (47). reagent (Invitrogen) following the manufacturer’s instructions. cDNA was synthesized using the SuperScript First-Strand cDNA synthesis ELISA system (Invitrogen) following the manufacturer’s instructions. The extracellular domain of hVCAM-1 (hVCAM-1–WT; aa 1–698) was amplified rhVCAM-1–Fc (R&D Systems) and rmVCAM-1–Fc were dissolved in using the following primers: 59-GGCCCAGGCGGCCATGCCTGGGAA- 50 ml PBS and added to the wells of a microtiter plate. The amount of GATGGTCG-39 and 59-GGCCCAGGCGGCCATGCCTGGGAAGATGG- VCAM-1–Fc added to each well was 5, 10, 20, or 50 ng. After incubation TCG-39. The primers were designed to add SfiI restriction digestion sites at by guest on September 28, 2021 overnight at 4˚C and washing three times with PBST (PBS containing both the 59 and 39 ends. The hVCAM-1 construct hVCAM-C1 (aa 1–599) was 0.05% [v/v] Tween 20), the microtiter plate was incubated for 2 h at 37˚C created using the following primers: 59-GGCCCAGGCGGCCATGCCT- with 3% (w/v) BSA in PBS. After washing with PBST, the plate was in- GGGAAGATGGTCG-39 and 59-GGCCCCACCGGCCCCAGTAACTTG- cubated with 2 nM VCAM-1-D6 Fab or 2 nM of unrelated Fab for 1 h at GATAATTAATTCCAC-39. The hVCAM-C2 (1–510) construct was generated 37˚C, and washed twice with PBST. The plate was incubated with anti- using the following primers: 59-GGCCCAGGCGGCCATGCCTGGGAAGA- human Fab Ab conjugated to HRP (Sigma, St. Louis, MO) diluted 1000- TGGTCG-39 and 59-GGCCCCACCGGCCCC GGCAACATTGACATAAA- fold in 3% (w/v) BSA in PBS. After washing with PBST, 50 ml 3,3,5,5- GTG-39. The hVCAM-C3 (1–407) construct was generated using the tetramethylbenzidine substrate solution (GenDEPOT, Barker, TX) was added following primers: 59-GGCCCAGGCGGCCATGCCTGGGAAGATGG- to each well. OD was measured at 450 nm using a microtiter plate reader TCG-39 and 59-GGCCCCACCGGCCCCTGGATCTCTAGGGAATGA- (Labsystems, Barcelona, Spain). GT-39. The hVCAM-C4 (1–311) construct was generated using the following primers: 59-GGCCCAGGCGGCCATGCCTGGGAAGATGG- Flow cytometry TCG-39 and 59-GGCCCCACCGGCCCC TTTCTCTTGAACAATTAA- 9 HUVECs (Lonza, Baltimore, MD) and human aortic endothelial cells TT-3 . The hVCAM-C5 (1–222) construct was generated using the 9 (HAECs; Lonza, Switzerland) were maintained in endothelial cell growth following primers: 5 -GGCCCAGGCGGCCATGCCTGGGAAGATGGT- CG-39 and 59-GGCCCCACCGGCCCCTGATATGTAGACTTGCAATT-39. medium-2 in a humidified incubator at 37˚C with 5% CO2 according to the manufacturer’s instructions. Mouse vascular endothelial cells (MVECs) The hVCAM-D1 (511–599) construct was generated using the following 9 9 were a kind gift of Dr. Saito (Tsurumi University, Tsurumi, Japan) and primers: 5 -GGCCCAGGCGGCCCCCAGAGATACAACCGTCTT-3 and 9 9 were maintained in Medium 199 supplemented with 5% (v/v) FBS, 10 mg/ml 5 -GGCCGGCCTGGCC GTAACTTGGATAATTAATT-3 .ThePCRfrag- insulin, 2.4 mg/ml hydrocortisone, and 1% (v/v) penicillin/streptomycin. ments were digested with SfiI and cloned into modified pcDNA 3.1 vectors The endothelial cells were stimulated with 20 ng/ml human (hTNF-a)or (Invitrogen) encoding the hinge region and CH2-CH3 domain of human 9 mouse TNF-a (mTNF-a; Millipore) for 24 h. Cells were incubated with IgG1 at the 3 region of the cloning site. Ligated products were transformed into calcium chloride-competent E. coli DH5a cells, and plasmid 10 mg/ml VCAM-1-D6 Fab, an anti-integrin aIIbb3 Fab that was used as 3 a negative control (48), or the mouse anti–hVCAM-1 Ab 51-10C9 (BD DNA was prepared as described previously (38). HEK293F cells (4.0 7 m Bioscience, San Diego, CA) (49), diluted in flow cytometry buffer (1% 10 cells; Invitrogen) were transfected with 50 g of each DNA using 50 m [w/v] BSA in PBS containing 0.05% [w/v] sodium azide) at 37˚C for 1 h. g polyethylenimine (Polysciences, Warrington, PA) as described previ- After centrifugation at 1000 3 g for 10 min, cells were resuspended in the ously (51). The transfected cells were maintained in Freestyle 293 ex- flow cytometry buffer. This washing step was then repeated two more pression media (Invitrogen) supplemented with 0.5% (v/v) penicillin/ times. The cells were then incubated at 37˚C for 1 h with FITC-labeled streptomycin in a humidified incubator at 37˚C with 8% CO2. After 7 d, the anti-human Fab (Jackson Immunoresearch Laboratory, West Grove, PA) or culture media was collected and the fusion proteins were purified using Alexa Fluor 488 labeled goat anti-mouse IgG Ab (Invitrogen) diluted protein A affinity chromatography as described previously (44, 45). 100-fold in flow cytometry buffer. The samples were analyzed by flow Immunoblot analysis cytometry. For all flow cytometry measurements, a FACS Canto II instrument (BD Biosciences, San Jose, CA) equipped with a 488-nm laser SDS-PAGE and immunoblot analysis were performed as described previ- was used. Ten thousand cells were detected per measurement with no ously (5). The purified proteins were dissolved in a Laemmli sample buffer gating, and the results were analyzed using the FlowJo software program and loaded into each well of 4–12% Tris-Glycine Gel (Novex NuPAGE; (Tree Star, Ashland, OR). Invitrogen). After electrophoresis, the proteins were transferred to nitro- The Journal of Immunology 3

cellulose membranes using the wet transfer system (GE Healthcare Life 37˚C inside a humidified incubator with 5% CO2. Wells were washed five Sciences, Pittsburgh, PA). The membrane was incubated with TTBS times with PBS containing 0.2 mM CaCl2 and 0.1 mM MgCl2 to remove (10 mM Tris/HCl, pH 7.5, 150 mM NaCl, and 0.05% [v/v] Tween 20) unbound cells. Bound cells were detached by treating with trypsin/EDTA containing 5% (w/v) skim milk at room temperature for 1 h, followed by (Life Technologies, Toronto, ON, Canada) for 5 min and then washed incubation with 10 mg/ml VCAM-1-D6 Fab in TTBS containing 5% (w/v) two times with DMEM containing 10% FBS. In flow cytometry, cells skim milk at room temperature for 2 h. The membrane was washed with showing fluorescence were counted as CFSE-labeled U937 cells attached TTBS and incubated with anti-human Fab-HRP diluted 1000-fold in TTBS to mock- or hVCAM-1–transfected HEK293 cells, HUVECs, or HAECs. containing 5% (w/v) skim milk at room temperature for 2 h. In a parallel The assay was performed using a FACS CantoII instrument (BD Bio- experiment, the membranes were incubated with HRP-conjugated rabbit sciences) with excitation at 488 nm and the emission filter set to a 530/30 anti-human Fc (Pierce, Rockford, IL) diluted 5000-fold in TTBS con- bandpass. In each measurement, 10,000 cells were counted with no gating. taining 5% (w/v) skim milk at room temperature for 30 min. The mem- Results are expressed as the percentage of CFSE+ cell counts over the total branes were washed with TTBS three times, and protein bands were counted events. visualized using SuperSignal West Pico Chemiluminescent Substrate (Pierce) according to the manufacturer’s instructions. Measurement of HUVEC-secreted cytokines and chemokines 3 5 Human monocyte isolation HUVECs (3.0 10 cells) were added to each well of a 6-well plate and treated with 20 ng/ml hTNF-a for 24 h at 37˚C inside a humidified in- Heparin-treated blood was collected from healthy volunteers after an incubator with 5% CO2. The VCAM-1-D6 IgG or the mouse anti–hVCAM-1 formed consent was signed, and in regulation accordance with the ethical Ab 51-10C9 was added to each well to a final concentration of 10 mg/ml. guidelines of Seoul National University hospital. PBMCs were isolated by The plate was incubated for 24 h at 37˚C inside a humidified incubator Ficoll density gradient centrifugation (Ficoll-Paque; Amersham, Uppsala, with 5% CO2. The media were harvested and subjected to the Bio-Plex Sweden) and resuspended in DMEM supplemented with 10% FBS, 2-ME, Multiplex Chemokine and Cytokine assay (Bio-Rad Laboratories, Hercu-

L-glutamine, gentamicin, essential and nonessential amino acids, and HEPES les, CA) or an IL-8 ELISA (Invitrogen) with a dynamic range between Downloaded from (Life Technologies, Grand Island, NY). CD14+ monocytes were then isolated 15.6 pg/ml and 1 ng/ml, following the manufacturers’ recommendations. by negative selection using the monocyte isolation kit II (Miltenyi Biotec, Auburn, CA) and an AutoMACS instrument (Miltenyi Biotec), according to Mice the manufacturer’s protocol. The purity of the monocytes was evaluated by fluorescent staining with FITC-labeled mouse anti-CD14 Ab (Ancell, Bay- Female C57BL/6 and BALB/c inbred mice, aged 8–10 wk, were purchased port, MN) and flow cytometry analysis. This procedure yielded a monocyte from Charles River Laboratories (Wilmington, MA) and maintained in the fraction containing 83% CD14+ cells. Seoul National University specific pathogen-free animal facility. These

mice were bred at the Biological Services Unit of the Seoul National http://www.jimmunol.org/ Transendothelial cell migration assay University according to the Institutional Animal Care and Use Committee guidelines. HUVECs (5.0 3 104 cells/well) were added to the upper chambers of a 24- transwell plate with polycarbonate membranes of 8.0-mm diameter pores Isolation of mouse pancreatic islets (Corning, Corning, NY) and incubated overnight in a humidiﬁed incubator Murine islets were isolated as described previously (7). In brief, the pan- at 37˚C with 5% CO . Cells were treated with hTNF-a diluted in endo- 2 creas from BALB/c mice was exposed and injected with HBSS (Medi- thelial cell growth medium-2 (Lonza, Baltimore, MD) to a ﬁnal concen- atech, Herndon, VA) containing 0.55 mg/ml collagenase P (Roche, tration of 20 ng/ml for 24 h. VCAM-1-D6 Fab, VCAM-1-D6 IgG, or Indianapolis, IN) via the common bile duct until adequate distension was rhVCAM-1 domain 6 (aa 511–599)-Fc were added to the upper chamber achieved. The distended pancreas was digested at 37˚C for 20 min with at various concentrations. Then U937 human monocytic cells (2.0 3 105

gentle shaking. The digestion was terminated by putting the pancreas on by guest on September 28, 2021 cells/well) were added to the upper chamber. An anti-aIIb3Fab(48) ice. The islets were filtered through a metal mesh and were purified on and an anti-respiratory syncytial virus Ab (palivizumab; MedImmune, Euro-Ficoll gradients (Sigma, St. Louis, MO) by centrifugation at 700 3 g Baltimore, MD) were used as the control Fab and control IgG, respectively. at 8˚C for 15 min. This protocol routinely yielded preparations of .90% The mouse anti–hVCAM-1 Ab 51-10C9 was used as a positive control. purity. Islets were stabilized by culturing overnight in RPMI 1640 media Transmigration assays were also performed using isolated with human supplemented with 11 mM glucose, 2 mM L-glutamine, 10% FCS, 100 U/ml monocytes, following the procedures as described earlier. The final con- penicillin, and 100 mg/ml streptomycin before transplantation. centrations of VCAM-1-D6 Fab, VCAM-1-D6 IgG, mouse anti–hVCAM- 1 Ab 51-10C9, and control Fab were adjusted to 10 mg/ml. The final Induction of diabetes mellitus and islet transplantation concentrations of control IgG and the rhVCAM-1 domain 6-Fc were 20 mg/ ml. RPMI 1640 media containing 50 ng/ml human stromal cell-derived Diabetes mellitus was induced in mice (8–10 wk old) by i.p. administration factor-1a (SDF-1a; R&D Systems) was placed in the lower chamber. After of 250 mg/kg streptozotocin (STZ; Sigma) freshly dissolved in citrate 8 h, cells that migrated to the lower chamber were collected and counted buffer (pH 4.5). Blood glucose in nonfasting conditions was measured under a light microscope. from the snipped tail by a portable glucometer (Lifescan, Milpitas, CA). Two consecutive nonfasting blood glucose readings of .250 mg/dl were Cell adhesion and neutralization assay obtained from whole blood. Seven days after STZ administration, mice were anesthetized with isoflurane. For islet transplantation, the left kidney The full-length cDNA of hVCAM-1 was ligated into the KpnI or XhoI site was exposed and 500 islet equivalents were delivered beneath the kidney of the pcDNA3.1 (+) vector (Invitrogen). After transforming calcium capsule using PE-50 polyethylene tube (Becton Dickinson, Parsippany, chloride-competent E. coli DH5a cells, plasmid DNA was prepared as NJ). Mice were treated with either VCAM-1-D6 IgG or control IgG (0.1 mg, described previously (38). For negative control (mock), pcDNA3.1 (+) i.p. injection on days 21, 0, 1, 2, 3, 4, 5, 6, and 7. Day 0 is defined as the 3 4 vector was also prepared for transfection. HEK293 cells (5.0 10 cells) time of islet transplantation. were transfected with 4 mg of each plasmid DNA using 6 ml TurboFect (Fermentas International, Burlington, Ontario, Canada) following the Immunohistochemistry manufacturer’s recommendations. The transfected cells were cultured in DMEM supplemented with 10% (v/v) FBS, 1% (v/v) penicillin/streptomycin, The kidney with the islet graft was removed and embedded in OCT and 400 mg/ml G418 as a selection marker. Clones stably expressing full- compound (Tissue-Tek, Sakura, Torrance, CA). Frozen sections were ob- length hVCAM-1 or mock-transfected clones were selected and harvested, tained on a cryostat (CM1850; Leica, Bannockburn, IL) at 5-mm thickness respectively, and the expression of hVCAM-1 was determined using flow and fixed in acetone for 10 min at 4˚C. Sections were washed three times cytometry. with PBS, followed by incubation in an endogenous peroxide-blocking Leukocyte adhesion assays were performed as described previously, with solution for 5 min at room temperature. Nonspecific staining was pre- minor modifications (50). In brief, 3.0 3 105 hVCAM-1 or mock- vented by treating the sections with 1% (v/v) FBS in PBS for 30 min at transfected HEK293 cells, HUVECs, or HAECs were plated on six-well room temperature. The primary Abs used were guinea pig anti-insulin dishes. Human endothelial cells were then stimulated with 20 ng/ml hTNF-a (1:20 dilution; 0.7 mg/ml; Dako, Carpinteria, CA), rat anti-CD4 (1:50 for 24 h. The VCAM-1-D6 Fab, control Fab (48), or mouse anti–hVCAM-1 dilution; 0.62 mg/ml, RM4-5; BD Biosciences, San Jose, CA), and rat anti- IgG 51-10C9 was individually added to a final concentration of 10 mg/ml macrophage marker F4/80 Ab (1:50 dilution; 10 mg/ml; eBiosciences, San and incubated for 1 h at 37˚C inside a humidified incubator with 5% CO2. Diego, CA). Primary Abs were applied for 60 min at room temperature. U937 cells were labeled with CFSE (Molecular Probes, Eugene, OR), and HRP-conjugated anti-guinea pig or anti-rat secondary Abs were applied for 6.0 3 105 cells were added to each well. The plate was incubated for 1 h at 40 min at room temperature. Color development was performed using 4 ROLE OF THE SIXTH Ig-LIKE DOMAIN OF VCAM-1

2 a VECTASTAIN Elite ABC kit (PK6101; Vector Laboratories, Burlin- and 7.31 6 0.42 3 10 11, respectively (Table I). These data in- game, CA). The sections were observed and visualized using a light mi- dicate a strong affinity for VCAM-1-D6 Fab and IgG with either croscope (Axiocam; Carl Zeiss, Oberkochen, Germany). hVCAM-1 or mVCAM-1. The reactivity of VCAM-1-D6 Fab to Results endogenous VCAM-1 expressed on human and mouse endothelial cells, and its localization pattern were investigated using flow An Ab isolated from a rabbit/human chimeric Fab library cytometry and immunocytochemistry (Fig. 1C, 1D). VCAM-1- reacts with hVCAM-1 and mVCAM-1 D6 Fab reacted to VCAM-1 on HAECs, HUVECs, and MVECs Rabbits were immunized and boosted four times with purified stimulated by hTNF-a or mTNF-a, but not to unstimulated cells. rhVCAM-1. Total RNA was prepared from spleen and bone In the immunocytochemical study, the localization pattern of marrow, and subjected to cDNA synthesis. Using this cDNA, we VCAM-1 obtained using VCAM-1-D6 Fab was similar to that generated a rabbit/human chimeric Fab library containing rabbit obtained with the mouse anti–hVCAM-1Ab51-10C9thatwas variable regions and human constant regions with a complexity used as a positive control. of 5.7 3 109. After six rounds of biopanning on immobilized mVCAM-1–Fc, clones were randomly selected, rescued by infec- The VCAM-1-D6 Ab specifically recognizes the sixth Ig-like tion of helper phage, and tested for their reactivity to hVCAM-1 and domain of VCAM-1 mVCAM-1 in phage enzyme immunoassay. Ab clones reactive to To identify the epitope region for the VCAM-1-D6 Fab, we both hVCAM-1 and mVCAM-1 were selected for further study. constructed expression vectors for six truncated forms of the ex- After expression in E. coli and subsequent purification, the Fab tracellular domain of hVCAM-1 as Fc fusion proteins (Fig. 2A). purity (.90%) was confirmed by SDS-PAGE and Coomassie blue These constructs were transfected into HEK293F cells, and the Fc Downloaded from staining (data not shown). Using an ELISA, we determined that fusion proteins were purified by protein A column chromatog- 0.02 nM VCAM-1-D6 Fab bound to both hVCAM-1– and mVCAM- raphy. An equal amount of each fusion protein was subjected to 1–coated ELISA plates (Fig. 1A). In a Western blot, the VCAM-1-D6 immunoblot analysis. The VCAM-1-D6 Fab detected hVCAM-1- Fab detected 5–50 ng rhVCAM-1 and rmVCAM-1 (Fig. 1B). WT (1–698)–Fc and hVCAM-1-C1 (1–599)–Fc, whereas it did In real-time interaction analysis using VCAM-1–coated chips not bind to hVCAM-1-C2 (1–510)–Fc, suggesting hVCAM-1 aa

(Table I), the KD constant for VCAM-1-D6 Fab interaction with 511–599 as an epitope for VCAM-1-D6. The HRP-conjugated http://www.jimmunol.org/ hVCAM-1 was 1.35 6 0.02 3 1028. For the interaction with human Fc Ab reacted to all of the Fc fusion proteins (Fig. 2B). 210 mVCAM-1, the KD constant was 4.78 6 0.06 3 10 . When using To further conﬁrm that hVCAM-1 aa sequence 511–599 includes the IgG form of the VCAM-1-D6 Ab, the KD constant for inter- the epitope of the VCAM-1-D6 Fab, we prepared a Fc fusion action with hVCAM-1 and mVCAM-1 was 3.06 6 0.04 3 10210 protein containing these residues (hVCAM-1-D6 [511–599]–Fc; by guest on September 28, 2021

FIGURE 1. VCAM-1-D6 Ab has cross-species reactivity to hVCAM-1 and mVCAM-1. (A) After rhVCAM-1–Fc and rmVCAM-1–Fc were coated onto 96-well plates, ELISA was performed as described with purified control Fab (N) or VCAM-1-D6 Fab (n). These results represent the mean 6 SD obtained from triplicate wells. (B) The indicated amounts of hVCAM-1 and mVCAM-1 recombinant protein were examined by immunoblot analysis with VCAM-1- D6 Fab and anti-human Fab-HRP. (C) HAECs, HUVECs, and MVECs were cultured in the absence (dotted line) or presence (solid line) of hTNF-a or mTNF-a for 24 h, respectively. The cells were then analyzed using flow cytometry with the VCAM-1-D6 Fab Ab, a control Fab, and the mouse anti– hVCAM-1 Ab 51-10C9, followed by FITC-labeled goat anti-human Fab or Alexa Fluor 488-labeled goat anti-mouse IgG, respectively. These results are representative of three independent experiments. (D) HAECs, HUVECs, and MVECs were cultured in the absence or presence of hTNF-a or mTNF-a, and incubated with the VCAM-1-D6 Fab, control Fab, or the mouse anti–hVCAM-1 Ab 51-10C9. Then cells were treated with FITC-labeled goat anti-human Fab Ab or Alexa Fluor 488-labeled goat anti-mouse IgG Ab, respectively (original magnification 3400). The Journal of Immunology 5

Table I. Afﬁnity constants (Ka, Kd) evaluated by BIAcore for VCAM-1-D6 Fab and IgG/hVCAM-1 and mVCAM-1 interactions

21 21 21 Immobilized Molecule Injected Molecule KD (M) Ka (M S ) Kd (S ) hVCAM-1a Anti–VCAM-1 Fab 1.35 6 0.02 3 1028 1.31 6 0.02 3 106 1.76 6 0.03 3 1022 mVCAM-1a Anti–VCAM-1 Fab 4.78 6 0.06 3 10210 5.04 6 0.01 3 105 2.41 6 0.12 3 1024 hVCAM-1a Anti–VCAM-1 lgG 3.06 6 0.04 3 10210 1.03 6 0.002 3 106 3.16 6 0.006 3 1024 mVCAM-1a Anti–VCAM-1 lgG 7.31 6 0.42 3 10211 1.18 6 0.003 3 106 8.64 6 0.84 3 1025 ahVCAM-1 and mVCAM-1 mean Fc fusion proteins. KD, Dissociation equilibrium constant calculated based on Kd/Ka.

Fig. 2A). This purified Fc fusion protein was subjected to immu- migrating U937 cells decreased significantly in a concentration- noblot analysis with VCAM-1-D6 Fab. VCAM-1-D6 Fab suc- dependent manner. We also found that both the VCAM-1-D6 Fab cessfully recognized hVCAM-1-D6 (511–599)–Fc (Fig. 2C). The and IgG almost completely inhibited the migration at 10 mg/ml IgG form of the VCAM-1-D6 Ab also reacted to hVCAM-1-D6 (Fig. 3A, 3B). To check whether rhVCAM-1 domain 6 is directly (511–599)–Fc (data not shown). From these data, we concluded that involved in leukocyte transendothelial migration, we performed the VCAM-1-D6 Fab specifically recognizes a C-terminal region a transmigration assay with rhVCAM-1 domain 6 (aa 511–599)-Fc (aa 511–599) of hVCAM-1, which represents the sixth Ig-like and control IgG. rhVCAM-1 domain 6-Fc significantly inhibited the domain of hVCAM-1 (aa 511–595). transmigration of U937 cells across activated HUVECs in a concentration-dependent manner (Fig. 3C). We also performed the Downloaded from The VCAM-1-D6 Ab significantly inhibits U937 cell and transmigration assay with monocytes isolated from human periph- human monocyte transmigration across activated endothelial eral blood. VCAM-1-D6 Fab or VCAM-1-D6 IgG and rhVCAM-1- cells without inhibiting U937 cell or human monocyte domain 6-Fc inhibited the transmigration of human monocytes adhesion across activated HUVECs (Fig. 3D). To investigate whether the VCAM-1-D6 Ab regulates leukocyte The effect of the VCAM-1 D6 Fab on leukocyte adhesion to transmigration across activated endothelial cells, we performed VCAM-1 was examined using hVCAM-1– or mock-transfected http://www.jimmunol.org/ transendothelial cell migration assays with HUVECs and U937 HEK293 cells, as well as HUVECs and HAECs induced to ex- human monocytic cells. HUVECs were added to the upper chamber press VCAM-1 by treatment with hTNF-a. Cell surface expres- of transwells, allowed to grow to confluence, and then treated with sion in hVCAM-1-transfected HEK293 cells was confirmed by hTNF-a for 24 h to induce VCAM-1 expression. After adding flow cytometric analysis (Fig. 4A). For adhesion assays, the cells U937 cells to the upper chambers, the chemoattractant human were grown to confluence on six-well dishes. U937 cells were SDF-1a was added to the lower chambers. After 24 h, the U937 labeled with CFSE and then added to each well in the absence or cells that migrated through the HUVEC monolayer to the lower presence of 50 mg/ml VCAM-1-D6 Fab, control Fab, or the mouse chamber were collected and counted. When VCAM-1-D6 Fab or anti–hVCAM-1 Ab 51-10C9. After 1 h, U937 cells attached to by guest on September 28, 2021 VCAM-1-D6 IgG was added to the upper chamber, the number of hVCAM-1 overexpressing HEK-293 cells, HUVECs, or HAECs were

FIGURE 2. The VCAM-1-D6 Ab specifically recognizes the sixth Ig-like domain of VCAM-1. (A) The indicated wild-type and C-terminal domain serial deletion mutants were constructed and prepared as Fc-fusion proteins. In addition, a fragment containing aa 511–599 of the hVCAM-1 extracellular domain was fused to the Fc domain. (B) The same amount of puriﬁed hVCAM-1 wild- type and C-terminal domain serial deletion mutant Fc fusion proteins (0.1 mg) was subjected to immunoblot analysis using the VCAM-1-D6 Fab, followed by anti-human Fab-HRP (upper panel) or anti-human Fc- HRP (lower panel). (C) Puriﬁed wild-type hVCAM-1 and the aa 511–599 fragment Fc fusion proteins were analyzed by immunoblot using VCAM-1-D6 Fab, followed by anti-human Fab-HRP (lower panel) or anti- human Fc-HRP (upper panel). In the control lane, recombinant human Fc was loaded as a negative control. These results are representative of three independent experiments. 6 ROLE OF THE SIXTH Ig-LIKE DOMAIN OF VCAM-1 Downloaded from http://www.jimmunol.org/

FIGURE 3. The VCAM-1-D6 Ab specifically inhibits U937 and human monocyte transendothelial migration across TNF-a–activated HUVECs. by guest on September 28, 2021 HUVECs (2.0 3 105 cells) were plated on the upper surface of transwell plates and cultured in the absence or presence of TNF-a for 24 h. After a 1-h preincubation with the indicated concentrations of VCAM-1-D6 Fab (A), control Fab (A), VCAM-1-D6 IgG (B), mouse anti–hVCAM-1 IgG (B), control IgG (B), rhVCAM-1 domain 6-Fc (C), or control IgG (C), U937 cells were added to the upper part of the transwell plate. Recombinant SDF-1a (50 ng/ml) was added to the lower part of the transwell. After 24 h, cells that migrated were counted under a light microscopy. (D) The transendothelial migration assays with human monocytes were performed by preincubating HUVECs with 10 mg/ml of the Fabs or 20 mg/ml rhVCAM-1 domain 6-Fc or control IgG in the same experimental settings described earlier. Results represent the mean 6 SD obtained from experiments performed in triplicate. *p . 0.05. detached by trypsinization and counted using flow cytometry. The VCAM-1-D6 IgG protects against grafted islet rejection by adhesion blocking Ab for domain 1 of VCAM-1, mouse anti–hVCAM- blocking leukocyte infiltration 1 Ab 51-10C9, significantly inhibited U937 adhesion to hVCAM- To investigate the effect of the VCAM-1-D6 Ab on inflammation 1–overexpressing HEK 293 cells, HUVECs, or HAECs, whereas in vivo, we used an MHC-mismatched mouse islet allograft model. VCAM-1-D6 Fab and rhVCAM-1 domain 6-Fc did not affect U937 Diabetes was induced by STZ and the experimental animals re- cell attachment to the VCAM-1–overexpressing cells (Fig. 4B–D). ceived VCAM-1-D6 IgG, whereas control animals received an VCAM-1-D6 IgG does not activate endothelial cells irrelevant IgG. Abs were injected into mice a total of nine times We used a fluorescence multiplex immunoassay to investigate (0.1 mg/injection) before islet transplantation on day 7 (Fig. 6A). whether VCAM-1-D6 IgG activates cells via VCAM-1 cross- Blood glucose levels were monitored twice per week. Median 6 linking. HUVECs were treated with 10 mg/ml VCAM-1-D6 IgG graft survival in the control IgG group was 28.2 8.6 d. In or the mouse anti–hVCAM-1 Ab 51-10C9 (as a positive control) contrast, animals treated with the VCAM-1-D6 Ab showed no (23, 52) for 24 h. In parallel experiments, HUVECs were pre- evidence of graft rejection at the time of their sacrifice, 110 d treated with 20 ng/ml hTNF-a for 24 h before incubation with the posttransplantation (Fig. 6B, 6C). In addition, the islet grafts from VCAM-1 Abs. The endothelial cell culture supernatant was ex- control and VCAM-1-D6 IgG-treated animals were histologically amined for two chemokines (GM-CSF and IL-8), as well as nine examined. The islet tissue from the control animals demonstrated cytokines (IL-1b, IL-2, IL-4, IL-5, IL-10, IL-12, IL-13, IFN-g, clear evidence of immune rejection with the loss of insulin- and TNF-a). The agonistic mouse anti–hVCAM-1 Ab 51-10C9 positive cells (Fig. 7A), as well as significant mononuclear leuko- significantly enhanced TNF-a–stimulated endothelial cell IL-8 cytic infiltrate into the graft (Fig. 7B, 7C). In contrast, the islet production (Fig. 5) but did not affect production of the other grafts from the VCAM-1-D6 Ab-treated mice were intact with cytokines (data not shown). In contrast, the VCAM-1-D6 Ab did many insulin-positive b cells (Fig. 7D). Interestingly, the long-term not enhance IL-8 production (Fig. 5), nor did it enhance any of the functioning grafts (.110 d) from the VCAM-1-D6 IgG-treated mice other cytokines or chemokines tested (data not shown). showed persistent peri-islet mononuclear cellular infiltration of both The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/

FIGURE 4. The VCAM-1-D6 Ab does not alter VCAM-1–mediated binding of U937 cells to TNF-a–activated HUVECs. (A) HEK293 cells overexpressing VCAM-1 or control vector were analyzed by ﬂow cytometry in the absence (dotted line) or presence (solid line) of VCAM-1-D6 Fab. HAECs by guest on September 28, 2021 (B), HUVECs (C), or VCAM-1–transfected HEK293 cells (D) were incubated in the absence or presence of VCAM-1-D6 Fab, control Fab, or the mouse anti–hVCAM-1 IgG 51-10C9 for 1 h. CFSE-labeled U937 cells were then added for 1 h. Bound cells were examined using ﬂow cytometry as described. Results shown represent the mean 6 SD obtained from three separate experiments performed in duplicate. *p . 0.05.

CD4+ T cells (Fig. 7E) and macrophages (Fig. 7F). However, islet allograft survival of .100 d with a survival rate of 75% of the these mononuclear cells were rarely observed within the islets. islet grafts (55). In a mouse model of Crohn’s disease, MK2.7 treatment yielded a 70% resolution of the acute inflammation (56). Discussion Furthermore, MK2.7 also reduced the development of arthritis in In this study, we demonstrate that VCAM-1-D6 binds to the sixth Ig-like domain of VCAM-1, a domain that does not mediate leukocyte binding. However, VCAM-1-D6 binding to this VCAM-1 domain blocks leukocyte transendothelial migration. Furthermore, VCAM-1-D6 IgG does not alter TNF-a–stimulated endothelial cell chemokine or cytokine production. In addition, VCAM-1-D6 IgG inhibited the recruitment of leukocytes to transplanted islets and blocked rejection of islet transplants. Leukocyte a4b1 integrin binding to domains 1 and 4 of VCAM- 1 on the endothelium plays a critical role in recruiting leukocytes during the initiation and progression of inflammation. Abs that block this interaction have anti-inflammatory properties. Natali- zumab, a humanized a4b1 integrin Ab that inhibits the interaction between the integrin and domains 1 and 4 of VCAM1, was approved for treatment of multiple sclerosis (39). In animal experiments, the rat anti-mVCAM-1 Abs MK1.9 and MK2.7 were shown to suc- FIGURE 5. The VCAM-1-D6 Ab does not induce IL-8 secretion of TNF-a–activated HUVECs. HUVECs and TNF-a–activated HUVECs cessfully inhibit the progression of inflammatory processes. MK1.9 (3.0 3 105 cells) were plated in 6-well plates and cultured in the absence or blocks Ramos cell binding to murine endothelioma lines (40) and presence of 10 mg/ml of the agonistic VCAM-1 Ab (mouse anti–hVCAM-1 also increased the graft survival rates of C3H/H3J mice that re- Ab 51-10C9) or VCAM-1-D6 IgG for 24 h. The culture supernatants were ceived tail skin grafts from B10.BR mice (53). In an experimental centrifuged to remove cell debris, and IL-8 secretion was analyzed by autoimmune encephalomyelitis model, MK1.9 reduced disease ELISA. Results shown represent the mean 6 SD obtained from three sep- severity (54). Another VCAM-1 Ab, MK2.7, allowed long-term arate experiments performed in duplicate. 8 ROLE OF THE SIXTH Ig-LIKE DOMAIN OF VCAM-1 Downloaded from

FIGURE 6. The VCAM-1-D6 Ab maintains long-term graft survival in allogeneic islet transplantation assay. (A) Schematic depiction of Ab treatment protocol. (B) Islets harvested from BALB/c mice were transplanted into the subcapsular space of kidney. Blood glucose levels were monitored twice per http://www.jimmunol.org/ week from control IgG (s) or VCAM-1-D6 IgG (N)-treated groups. (C) Treatment with the VCAM-1-D6 Ab induces prolonged graft survival in C57BL/6 recipients (mean survival day . 110 d, n = 5). Control IgG treatment group was hyperglycemic after a brief normoglycemia period (mean survival day = 28.2 d, n = 5). Graft survival is presented as Kaplan–Meier survival curves. a collagen-induced arthritis model (57) and inhibited inflamma- coated with the mouse anti–hVCAM-1 Ab 51-10C9 stimulates tory cell infiltration into the skin in a keratin-14 IL-4 transgenic NADPH oxidase, metalloproteinase-2, and metalloproteinase-9 mouse atopic dermatitis model (58). The mouse anti–hVCAM-1 activity (42). In this study, the agonistic mouse anti–hVCAM-1 Abs 4B9, BBIG-V1, 1G11, and P3H12 block binding of lym- Ab 51-10C9 enhanced TNF-a–stimulated endothelial cell pro- phoma cell lines to VCAM-1–expressing endothelial cells (52, 59, duction of IL-8. However, an anti–VCAM-1 domain 6 Ab did not by guest on September 28, 2021 60). However, these Abs are species specific. A desirable property alter TNF-a–stimulated production of chemokines or cytokines. for a VCAM-1 Ab for therapeutic development is its reactivity to Nevertheless, the VCAM-1-D6 Ab has the unique function of both hVCAM-1 and animal VCAM-1. The VCAM-1-D6 Ab de- blocking leukocyte transendothelial migration without affecting scribed in this study has broad reactivity to hVCAM-1, mVCAM-1, leukocyte adhesion. In vivo intravital microscopy studies dem- and pig VCAM-1 (Fig. 1A and data not shown). This broad onstrate bound leukocytes that do not migrate are often released species reactivity allows evaluation of this Ab in various animal and continue with the blood flow (41). disease models, as well as in human disease interventions. The epitopes of VCAM-1 neutralizing Abs have been reported to be either domain 1 or 4, the domains involved in a4b1 integrin binding (23, 24). The MK2.7 Ab is reactive to VCAM-1 domains 1 and 4 (61). The epitope for the mouse anti–hVCAM-1 Abs, 51- 10C9 and 4B9, is localized to domain 1 (23, 24). In this study, we showed that an Ab reactive to VCAM-1 domain 6 inhibited the transmigration. Because rhVCAM-1 domain 6 (aa 51–599) Fc fusion also inhibited the transmigration of U937 cells and human monocytes across HUVECs, it is more likely that the domain 6 directly interacts with another VCAM-1 molecule or with tetraspanins CD9, CD81, CD151 in tetraspanin-enriched microdomains to initiate downstream signaling (62, 63). This may regulate transendothelial migration because VCAM-1 cross- linking activates downstream signaling molecules in endothelial cells that are required for leukocyte transendothelial migration FIGURE 7. The VCAM-1-D6 Ab significantly promotes islet cell sur- (64). For example, treatment of human and mouse endothelial cells vival by blocking leukocyte migration inside the grafted islets. The kidneys with anti–mVCAM-1–coated beads (clone MV-CAM.A) in- of control animals were isolated at the time of rejection and kidneys fromVCAM-1-D6 Ab-treated mice were isolated 110 d after transplanta- creased the intracellular level of reactive oxygen species, tyrosine tion. Histological comparison of islet grafts from the two groups was phosphatase activity, and protein kinase Ca activity, all of which performed using immunohistochemical staining. (A and D) Islet cell are required for leukocyte migration (49). VCAM-1 cross-linking staining using an insulin Ab. (B and E) CD4 T cell staining using a CD4 with the VCAM-1 Ab (1G11) induced the activation of Rac1, a Ab. (C and F) Macrophage staining using the F4/80 Ab. Images are rep- small GTP binding protein, reactive oxygen species production, resentative of three independent experiments. Arrows indicate the stained and p38 MAPK activation (65). VCAM-1 activation by beads cells. Original magnification 3100 (A–D), 3200 (E, F). The Journal of Immunology 9

In this study, the in vivo efficacy of the VCAM-1-D6 Ab was 13. Weber, C., R. Alon, B. Moser, and T. A. Springer. 1996. Sequential regulation of alpha 4 beta 1 and alpha 5 beta 1 integrin avidity by CC chemokines in demonstrated in a mouse islet allotransplantation model. VCAM-1 monocytes: implications for transendothelial chemotaxis. J. Cell Biol. 134: plays a critical role in recruitment of leukocytes to allografts (66– 1063–1073. 68). VCAM-1 is overexpressed in grafted islets through inflam- 14. Cook-Mills, J. M., Marchese, M., and Abdala-Valencia, H. 2011. VCAM-1 expression and signaling during disease: regulation by reactive oxygen species matory cytokines such as IFN-g, TNF-a, and IL-1 (14, 68). The and antioxidants. Antioxid. Redox Signal. 15: 1607–1638. mVCAM-1–specific Ab MK2.7 blocks adhesion to domain 1 and 15. Cybulsky, M. I., J. W. Fries, A. J. Williams, P. Sultan, V. M. Davis, protects against mouse islet allograft rejection by blocking leu- M. A. Gimbrone, Jr., and T. Collins. 1991. Alternative splicing of human VCAM-1 in activated vascular endothelium. Am. J. Pathol. 138: 815–820. kocyte adhesion to endothelium. In contrast, we found that 16. Hession, C., R. Tizard, C. Vassallo, S. B. Schiffer, D. Goff, P. Moy, G. Chi- VCAM-1-D6 IgG did not block adhesion but did block leukocyte Rosso, S. Luhowskyj, R. Lobb, and L. Osborn. 1991. Cloning of an alternate transendothelial migration. In addition, it successfully inhibited form of vascular cell adhesion molecule-1 (VCAM1). J. Biol. Chem. 266: 6682– 6685. islet graft transplant rejection for at least 110 d after transplan- 17. Polte, T., W. Newman, G. Raghunathan, and T. V. Gopal. 1991. Structural and tation. Immunohistochemical studies also indicated an inhibition functional studies of full-length vascular cell adhesion molecule-1: internal of CD4+ T cell and macrophage infiltration toward the grafted duplication and homology to several adhesion proteins. DNA Cell Biol. 10: 349– 357. islets. This infiltration of CD4 T cells and macrophages in the 18. Cybulsky, M. I., J. W. Fries, A. J. Williams, P. Sultan, R. Eddy, M. Byers, peri-islet space without invasion into the graft site was previously T. Shows, M. A. Gimbrone, Jr., and T. Collins. 1991. Gene structure, chromo- somal location, and basis for alternative mRNA splicing of the human VCAM1 observed when immunological tolerance was achieved (69). Thus, gene. Proc. Natl. Acad. Sci. USA 88: 7859–7863. we have identified a unique function for VCAM-1-D6 Abs in 19. Chan, B. M., M. J. Elices, E. Murphy, and M. E. Hemler. 1992. Adhesion to blocking islet graft transplant rejection and blocking leukocyte vascular cell adhesion molecule 1 and fibronectin. Comparison of alpha 4 beta 1 (VLA-4) and alpha 4 beta 7 on the human B cell line JY. J. Biol. Chem. 267: recruitment. 8366–8370. Downloaded from The unique property of the VCAM-1-D6 Ab to not block leu- 20. Grayson, M. H., M. Van der Vieren, S. A. Sterbinsky, W. Michael Gallatin, kocyte adhesion may provide some advantages. It is reported that P. A. Hoffman, D. E. Staunton, and B. S. Bochner. 1998. alphadbeta2 integrin is expressed on human eosinophils and functions as an alternative ligand for vas- in lymphoid tissues, the interaction of lymphocyte VLA-4 with cular cell adhesion molecule 1 (VCAM-1). J. Exp. Med. 188: 2187–2191. VCAM-1 on stromal cells activates IL-7 production, a pre-B cell 21. Ruëgg, C., A. A. Postigo, E. E. Sikorski, E. C. Butcher, R. Pytela, and D. J. Erle. growth factor (70). The MK1.9 and MK2.7 Abs inhibited B 1992. Role of integrin alpha 4 beta 7/alpha 4 beta P in lymphocyte adherence to fibronectin and VCAM-1 and in homotypic cell clustering. J. Cell Biol. 117: lymphocyte development in bone marrow culture in vitro (40). 179–189. http://www.jimmunol.org/ VCAM-1 is expressed at low levels on interstitial dendritic cells 22. Ross, E. A., M. R. Douglas, S. H. Wong, E. J. Ross, S. J. Curnow, G. B. Nash, E. Rainger, D. Scheel-Toellner, J. M. Lord, M. Salmon, and C. D. Buckley. 2006. and could have a role in Ag presentation for lymphocyte acti- Interaction between integrin alpha9beta1 and vascular cell adhesion molecule-1 vation (68). Abs that block lymphocyte adhesion may inhibit (VCAM-1) inhibits neutrophil apoptosis. Blood 107: 1178–1183. T lymphocyte activation by interfering with the interaction be- 23. Osborn, L., C. Vassallo, and C. D. Benjamin. 1992. Activated endothelium binds lymphocytes through a novel binding site in the alternately spliced domain tween interstitial dendritic cells and lymphocytes. of vascular cell adhesion molecule-1. J. Exp. Med. 176: 99–107. In summary, VCAM-1-D6 Ab, which is reactive to VCAM-1 24. Pepinsky, B., C. Hession, L. L. Chen, P. Moy, L. Burkly, A. Jakubowski, domain 6, inhibits transmigration but not adhesion, suggesting E. P. Chow, C. Benjamin, G. Chi-Rosso, S. Luhowskyj, et al. 1992. Structure/ function studies on vascular cell adhesion molecule-1. J. Biol. Chem. 267: that the sixth IgG-like domain may participate in transmigration 17820–17826. (36, 50). 25. Renz, M. E., H. H. Chiu, S. Jones, J. Fox, K. J. Kim, L. G. Presta, and S. Fong. by guest on September 28, 2021 1994. Structural requirements for adhesion of soluble recombinant murine vascular cell adhesion molecule-1 to alpha 4 beta 1. J. Cell Biol. 125: 1395–1406. Disclosures 26. Vonderheide, R. H., T. F. Tedder, T. A. Springer, and D. E. Staunton. 1994. The authors have no financial conflicts of interest. Residues within a conserved amino acid motif of domains 1 and 4 of VCAM-1 are required for binding to VLA-4. J. Cell Biol. 125: 215–222. 27. Barthel, S. R., D. S. Annis, D. F. Mosher, and M. W. Johansson. 2006. Differ- ential engagement of modules 1 and 4 of vascular cell adhesion molecule-1 References (CD106) by integrins alpha4beta1 (CD49d/29) and alphaMbeta2 (CD11b/18) 1. McEver, R. P. 2002. Selectins: lectins that initiate cell adhesion under flow. Curr. of eosinophils. J. Biol. Chem. 281: 32175–32187. Opin. Cell Biol. 14: 581–586. 28. Barthel, S. R., N. N. Jarjour, D. F. Mosher, and M. W. Johansson. 2006. Dis- 2. Vestweber, D. 2003. Lymphocyte trafficking through blood and lymphatic ves- section of the hyperadhesive phenotype of airway eosinophils in asthma. Am. J. sels: more than just selectins, chemokines and integrins. Eur. J. Immunol. 33: Respir. Cell Mol. Biol. 35: 378–386. 1361–1364. 29. Kilger, G., L. A. Needham, P. J. Nielsen, J. Clements, D. Vestweber, and 3. Johansson, M. W., and D. F. Mosher. 2011. Activation of beta1 integrins on B. Holzmann. 1995. Differential regulation of alpha 4 integrin-dependent blood eosinophils by P-selectin. Am. J. Respir. Cell Mol. Biol. 45: 889–897. binding to domains 1 and 4 of vascular cell adhesion molecule-1. J. Biol. 4. Alon, R., and S. Feigelson. 2002. From rolling to arrest on blood vessels: leu- Chem. 270: 5979–5984. kocyte tap dancing on endothelial integrin ligands and chemokines at sub-second 30. Cook-Mills, J. M., J. S. Gallagher, and T. L. Feldbush. 1996. Isolation and contacts. Semin. Immunol. 14: 93–104. characterization of high endothelial cell lines derived from mouse lymph nodes. 5. Campbell, J. J., S. Qin, K. B. Bacon, C. R. Mackay, and E. C. Butcher. 1996. In Vitro Cell. Dev. Biol. Anim. 32: 167–177. Biology of chemokine and classical chemoattractant receptors: differential 31. Matheny, H. E., T. L. Deem, and J. M. Cook-Mills. 2000. Lymphocyte migration requirements for adhesion-triggering versus chemotactic responses in lymphoid through monolayers of endothelial cell lines involves VCAM-1 signaling via cells. J. Cell Biol. 134: 255–266. endothelial cell NADPH oxidase. J. Immunol. 164: 6550–6559. 6. Shamri, R., V. Grabovsky, S. W. Feigelson, O. Dwir, Y. Van Kooyk, and R. Alon. 32. Muller, W. A. 2009. Mechanisms of transendothelial migration of leukocytes. 2002. Chemokine stimulation of lymphocyte alpha 4 integrin avidity but not of Circ. Res. 105: 223–230. leukocyte function-associated antigen-1 avidity to endothelial ligands under shear 33. Cernuda-Morolloń, E., S. Gharbi, and J. Millań. 2010. Discriminating between flow requires cholesterol membrane rafts. J. Biol. Chem. 277: 40027–40035. the paracellular and transcellular routes of diapedesis. Methods Mol. Biol. 616: 7. Alon, R., V. Grabovsky, and S. Feigelson. 2003. Chemokine induction of integrin 69–82. adhesiveness on rolling and arrested leukocytes local signaling events or global 34. Mamdouh, Z., A. Mikhailov, and W. A. Muller. 2009. Transcellular migration of stepwise activation? Microcirculation 10: 297–311. leukocytes is mediated by the endothelial lateral border recycling compartment. 8. Carr, M. W., R. Alon, and T. A. Springer. 1996. The C-C chemokine MCP-1 dif- J. Exp. Med. 206: 2795–2808. ferentially modulates the avidity of beta 1 and beta 2 integrins on T lymphocytes. 35. Middleton, J., A. M. Patterson, L. Gardner, C. Schmutz, and B. A. Ashton. 2002. Immunity 4: 179–187. Leukocyte extravasation: chemokine transport and presentation by the endo- 9. Chan, J. R., S. J. Hyduk, and M. I. Cybulsky. 2003. Detecting rapid and transient thelium. Blood 100: 3853–3860. upregulation of leukocyte integrin affinity induced by chemokines and chemo- 36. Abdala-Valencia, H., J. Earwood, S. Bansal, M. Jansen, G. Babcock, B. Garvy, attractants. J. Immunol. Methods 273: 43–52. M. Wills-Karp, and J. M. Cook-Mills. 2007. Nonhematopoietic NADPH oxidase 10. Johnston, B., and E. C. Butcher. 2002. Chemokines in rapid leukocyte adhesion regulation of lung eosinophilia and airway hyperresponsiveness in experimen- triggering and migration. Semin. Immunol. 14: 83–92. tally induced asthma. Am. J. Physiol. Lung Cell. Mol. Physiol. 292: L1111– 11. Laudanna, C., and R. Alon. 2006. Right on the spot. Chemokine triggering of L1125. integrin-mediated arrest of rolling leukocytes. Thromb. Haemost. 95: 5–11. 37. Choi, S. H., J. W. Lee, and S. J. Sim. 2005. Enhanced performance of a surface 12. Laudanna, C., J. Y. Kim, G. Constantin, and E. Butcher. 2002. Rapid leukocyte plasmon resonance immunosensor for detecting Ab-GAD antibody based on the integrin activation by chemokines. Immunol. Rev. 186: 37–46. modified self-assembled monolayers. Biosens. Bioelectron. 21: 378–383. 10 ROLE OF THE SIXTH Ig-LIKE DOMAIN OF VCAM-1

38. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecular Cloning: A Lab- 55. Stegall, M. D., P. G. Dean, D. Ninova, A. J. Cohen, G. M. Shepard, C. Gup, and oratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. R. G. Gill. 2001. alpha4 integrin in islet allograft rejection. Transplantation 71: 39. Yednock, T. A., C. Cannon, L. C. Fritz, F. Sanchez-Madrid, L. Steinman, and 1549–1555. N. Karin. 1992. Prevention of experimental autoimmune encephalomyelitis by 56. Burns, R. C., J. Rivera-Nieves, C. A. Moskaluk, S. Matsumoto, F. Cominelli, and antibodies against alpha 4 beta 1 integrin. Nature 356: 63–66. K. Ley. 2001. Antibody blockade of ICAM-1 and VCAM-1 ameliorates in- 40. Miyake, K., K. Medina, K. Ishihara, M. Kimoto, R. Auerbach, and flammation in the SAMP-1/Yit adoptive transfer model of Crohn’s disease in P. W. Kincade. 1991. A VCAM-like adhesion molecule on murine bone marrow mice. Gastroenterology 121: 1428–1436. stromal cells mediates binding of lymphocyte precursors in culture. J. Cell Biol. 57. Carter, R. A., I. K. Campbell, K. L. O’Donnel, and I. P. Wicks. 2002. Vascular 114: 557–565. cell adhesion molecule-1 (VCAM-1) blockade in collagen-induced arthritis 41. Cook-Mills, J. M., M. E. Marchese, and H. Abdala-Valencia. 2011. Vascular cell reduces joint involvement and alters B cell trafficking. Clin. Exp. Immunol. 128: adhesion molecule-1 expression and signaling during disease: regulation by 44–51. reactive oxygen species and antioxidants. Antioxid. Redox Signal. 15: 1607– 58. Chen, L., S. X. Lin, S. Amin, L. Overbergh, G. Maggiolino, and L. S. Chan. 1638. 2010. VCAM-1 blockade delays disease onset, reduces disease severity and 42. Deem, T. L., and J. M. Cook-Mills. 2004. Vascular cell adhesion molecule 1 inflammatory cells in an atopic dermatitis model. Immunol. Cell Biol. 88: 334– (VCAM-1) activation of endothelial cell matrix metalloproteinases: role of re- 342. active oxygen species. Blood 104: 2385–2393. 59. Zohlnho¨fer, D., K. Brand, K. Schipek, G. Pogatsa-Murray, A. Scho¨mig, and 43. Barbas, C. F. 2001. Phage Display: A Laboratory Manual. Cold Spring Harbor F. J. Neumann. 2000. Adhesion of monocyte very late antigen-4 to endothelial Laboratory Press, Cold Spring Harbor, NY. vascular cell adhesion molecule-1 induces interleukin-1beta-dependent expres- 44. Shukla, A. A., B. Hubbard, T. Tressel, S. Guhan, and D. Low. 2007. Downstream sion of interleukin-6 in endothelial cells. Arterioscler. Thromb. Vasc. Biol. 20: processing of monoclonal antibodies—application of platform approaches. J. 353–359. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 848: 28–39. 60. Sung, K. L., Y. Li, M. Elices, J. Gang, P. Sriramarao, and D. H. Broide. 1997. 45. Pitaksajjakul, P., P. Lekcharoensuk, N. Upragarin, C. F. Barbas, III, Granulocyte-macrophage colony-stimulating factor regulates the functional ad- M. S. Ibrahim, K. Ikuta, and P. Ramasoota. 2010. Fab MAbs specific to HA of hesive state of very late antigen-4 expressed by eosinophils. J. Immunol. 158: influenza virus with H5N1 neutralizing activity selected from immunized 919–927. chicken phage library. Biochem. Biophys. Res. Commun. 395: 496–501. 61. Kumar, A. G., X. Y. Dai, C. A. Kozak, M. P. Mims, A. M. Gotto, and 46. Van Meter, K. E., and M. K. Stuart. 2008. A monoclonal antibody that inhibits C. M. Ballantyne. 1994. Murine VCAM-1. Molecular cloning, mapping, and Downloaded from translation in Sf21 cell lysates is specific for glyceraldehyde-3-phosphate de- analysis of a truncated form. J. Immunol. 153: 4088–4098. hydrogenase. Arch. Insect Biochem. Physiol. 69: 107–117. 62. Barreiro, O., M. Yanez-Mo, J. M. Serrador, M. C. Montoya, M. Vicente- 47. Kim, H. Y., S. Tsai, S. C. Lo, D. J. Wear, and M. J. Izadjoo. 2011. Production and Manzanares, R. Tejedor, H. Furthmayr, and F. Sanchez-Madrid. 2002. Dynamic characterization of chimeric monoclonal antibodies against Burkholderia pseu- interaction of VCAM-1 and ICAM-1 with moesin and ezrin in a novel endo- domallei and B. mallei using the DHFR expression system. PLoS ONE 6: thelial docking structure for adherent leukocytes. J. Cell Biol. 157: 1233–1245. e19867. 63. Barreiro, O., M. Zamai, M. Yań˜ez-Mo´, E. Tejera, P. Lo´pez-Romero, P. N. Monk, 48. Chung, J., C. Rader, M. Popkov, Y. M. Hur, H. K. Kim, Y. J. Lee, and E. Gratton, V. R. Caiolfa, and F. Sańchez-Madrid. 2008. Endothelial adhesion

C. F. Barbas, III. 2004. Integrin alphaIIbbeta3-speciﬁc synthetic human receptors are recruited to adherent leukocytes by inclusion in preformed tetra- http://www.jimmunol.org/ monoclonal antibodies and HCDR3 peptides that potently inhibit platelet spanin nanoplatforms. J. Cell Biol. 183: 527–542. aggregation. FASEB J. 18: 361–363. 64. Meager, A. 1999. Cytokine regulation of cellular adhesion molecule expression 49. Deem, T. L., H. Abdala-Valencia, and J. M. Cook-Mills. 2007. VCAM-1 acti- in inﬂammation. Cytokine Growth Factor Rev. 10: 27–39. vation of endothelial cell protein tyrosine phosphatase 1B. J. Immunol. 178: 65. van Wetering, S., N. van den Berk, J. D. van Buul, F. P. Mul, I. Lommerse, 3865–3873. R. Mous, J. P. ten Klooster, J. J. Zwaginga, and P. L. Hordijk. 2003. VCAM-1- 50. Lee, S., J. Chung, I. S. Ha, K. Yi, J. E. Lee, H. G. Kang, I. Choi, K. H. Oh, mediated Rac signaling controls endothelial cell-cell contacts and leukocyte J. Y. Kim, C. D. Surh, and C. Ahn. 2007. Hydrogen peroxide increases human transmigration. Am. J. Physiol. Cell Physiol. 285: C343–C352. leukocyte adhesion to porcine aortic endothelial cells via NFkappaB-dependent 66. Orosz, C. G., R. G. Ohye, R. P. Pelletier, A. M. Van Buskirk, E. Huang, up-regulation of VCAM-1. Int. Immunol. 19: 1349–1359. C. Morgan, P. W. Kincade, and R. M. Ferguson. 1993. Treatment with anti- 51. Baldi, L., N. Muller, S. Picasso, R. Jacquet, P. Girard, H. P. Thanh, E. Derow, and vascular cell adhesion molecule 1 monoclonal antibody induces long-term mu- F. M. Wurm. 2005. Transient gene expression in suspension HEK-293 cells: rine cardiac allograft acceptance. Transplantation 56: 453–460.

application to large-scale protein production. Biotechnol. Prog. 21: 148–153. 67. Isobe, M., J. Suzuki, H. Yagita, K. Okumura, and M. Sekiguchi. 1994. Effect of by guest on September 28, 2021 52. Carlos, T. M., B. R. Schwartz, N. L. Kovach, E. Yee, M. Rosa, L. Osborn, anti-VCAM-1 and anti-VLA-4 monoclonal antibodies on cardiac allograft sur- G. Chi-Rosso, B. Newman, R. Lobb, et al. 1990. Vascular cell adhesion vival and response to soluble antigens in mice. Transplant. Proc. 26: 867–868. molecule-1 mediates lymphocyte adherence to cytokine-activated cultured hu- 68. Stegall, M. D., A. Ostrowska, J. Haynes, F. Karrer, I. Kam, and R. G. Gill. 1995. man endothelial cells. Blood 76: 965–970. Prolongation of islet allograft survival with an antibody to vascular cell adhesion 53. Gorcyznski, R. M., S. Chung, X. M. Fu, G. Levy, B. Sullivan, and Z. Chen. 1995. molecule 1. Surgery 118: 366–369, discussion 369–370. Manipulation of skin graft rejection in alloimmune mice by anti-VCAM-1:VLA- 69. Yoon, I. H., S. E. Choi, Y. H. Kim, S. H. Yang, J. H. Park, C. S. Park, Y. Kim, 4 but not anti-ICAM-1:LFA-1 monoclonal antibodies. Transpl. Immunol. 3: 55– J. S. Kim, S. J. Kim, E. Simpson, and C. G. Park. 2008. Pancreatic islets induce 61. CD4(+) [corrected] CD25(-)Foxp3(+) [corrected] T-cell regulated tolerance to 54. Kanwar, J. R., R. K. Kanwar, D. Wang, and G. W. Krissansen. 2000. Prevention HY-mismatched skin grafts. Transplantation 86: 1352–1360. of a chronic progressive form of experimental autoimmune encephalomyelitis by 70. Sudo, T., M. Ito, Y. Ogawa, M. Iizuka, H. Kodama, T. Kunisada, S. Hayashi, an antibody against mucosal addressin cell adhesion molecule-1, given early in M. Ogawa, K. Sakai, and S. Nishikawa. 1989. Interleukin 7 production and the course of disease progression. Immunol. Cell Biol. 78: 641–645. function in stromal cell-dependent B cell development. J. Exp. Med. 170: 333–338.