Inhibition of Invariant Chain Processing, Antigen-Induced Proliferative Responses, and the Development of -Induced Arthritis and Experimental Autoimmune This information is current as Encephalomyelitis by a Small Molecule of September 30, 2021. Inhibitor Patricia L. Podolin, Brian J. Bolognese, Donald C. Carpenter, T. Gregg Davis, Roy A. Johanson, Josephine H.

Fox, Edward Long III, Xiaoyang Dong, Robert W. Marquis, Downloaded from Stephen M. LoCastro, Gerald J. Terfloth, Edit Kurali, John J. Peterson, Brian R. Smith, Michael S. McQueney, Dennis S. Yamashita and Elizabeth A. Capper-Spudich J Immunol 2008; 180:7989-8003; ;

doi: 10.4049/jimmunol.180.12.7989 http://www.jimmunol.org/ http://www.jimmunol.org/content/180/12/7989

References This article cites 64 articles, 31 of which you can access for free at: http://www.jimmunol.org/content/180/12/7989.full#ref-list-1 by guest on September 30, 2021 Why The JI? Submit online.

• 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

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 © 2008 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Inhibition of Invariant Chain Processing, Antigen-Induced Proliferative Responses, and the Development of Collagen-Induced Arthritis and Experimental Autoimmune Encephalomyelitis by a Small Molecule Inhibitor

Patricia L. Podolin,1* Brian J. Bolognese,* Donald C. Carpenter,* T. Gregg Davis,* Roy A. Johanson,2* Josephine H. Fox,3* Edward Long III,* Xiaoyang Dong,¶ Robert W. Marquis,¶ Stephen M. LoCastro,¶ Gerald J. Terfloth,† Edit Kurali,‡ John J. Peterson,ʈ Brian R. Smith,4§ Michael S. McQueney,# Dennis S. Yamashita,¶ Downloaded from and Elizabeth A. Capper-Spudich5*

Members of the family of cysteine () mediate late stage processing of MHC class II-bound invariant chain (Ii), enabling dissociation of Ii, and binding of antigenic to class II molecules. Recognition of surface class II/Ag -complexes by CD4؉ T cells then leads to T cell activation. Herein, we demonstrate that a pan-active inhibitor, SB http://www.jimmunol.org/ 331750, attenuated the processing of whole cell Ii p10 to CLIP by Raji cells, and DBA/1, SJL/J, and C57BL/6 splenocytes. In Raji cells and C57BL/6 splenocytes, SB-331750 inhibited class II-associated Ii processing and reduced surface class II/CLIP expression, whereas in SB-331750-treated DBA/1 and SJL/J splenocytes, class II-associated Ii processing intermediates were undetectable. Incubation of lymph node cells/splenocytes from collagen-primed DBA/1 mice and myelin basic -primed SJL/J mice with Ag in the presence of SB-331750 resulted in concentration-dependent inhibition of Ag-induced proliferation. In vivo administra- tion of SB-331750 to DBA/1, SJL/J, and C57BL/6 mice inhibited splenocyte processing of whole cell Ii p10 to CLIP. Prophylactic administration of SB-331750 to collagen-immunized/boosted DBA/1 mice delayed the onset and reduced the severity of collagen- induced arthritis (CIA), and reduced paw tissue levels of IL-1␤ and TNF-␣. Similarly, treatment of myelin basic protein-primed SJL/J lymph node cells with SB-331750 delayed the onset and reduced the severity of adoptively transferred experimental au- by guest on September 30, 2021 toimmune encephalomyelitis (EAE). Therapeutic administration of SB-331750 reduced the severity of mild/moderate CIA and EAE. These results indicate that pharmacological inhibition of cathepsins attenuates CIA and EAE, potentially via inhibition of Ii processing, and subsequent Ag-induced T cell activation. The Journal of Immunology, 2008, 180: 7989–8003.

ne of the central occurrences in the generation of a cell- derived from host (autologous) tissue, the subsequent immune re- mediated immune response is the recognition of anti- sponse results in destruction of self-tissue or autoimmunity. Ex- genic peptide by CD4ϩ T cells (1). When the antigenic amples of autoimmune responses include the joint inflammation O ϩ peptide recognized by the CD4 T cell is derived from an invading and bone/cartilage destruction that occur in rheumatoid arthritis (foreign) organism, such as a bacterium, the ensuing immune re- (RA),6 the demyelination of the CNS that is observed in multiple sponse results in elimination of the deleterious organism and, thus, sclerosis, the intestinal lesions that occur in inflammatory bowel is beneficial to the host. In contrast, when the antigenic peptide is disease, and the pancreatic ␤ cell destruction that results in type 1 diabetes. In order for a peripheral CD4ϩ T cell to recognize an antigenic † *Respiratory and Inflammation Center of Excellence for Drug Discovery, Chemical peptide, that peptide must be bound to an MHC class II molecule Development, ‡Discovery Statistics, and §Drug and Pharmacokinetics, GlaxoSmithKline, King of Prussia, PA 19406; and ¶Microbial, Musculoskeletal, and expressed on the cell surface of professional APC (, Proliferative Diseases Center of Excellence for Drug Discovery, ʈResearch Statistics dendritic cells, and B cells). The binding of antigenic peptide to the Unit, and #Screening and Compound Profiling, GlaxoSmithKline, Collegeville, PA 19426 Ag binding groove of an MHC class II molecule occurs intracel- lularly, within the endosomes of APC. Before the intracellular binding Received for publication February 23, 2007. Accepted for publication April 7, 2008. of antigenic peptide to the class II molecule, the class II molecule is The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance bound by a protein called the invariant chain (Ii). This binding occurs with 18 U.S.C. Section 1734 solely to indicate this fact. in the , after which the Ii/class II complex is 1 Address correspondence and reprint requests to Dr. Patricia Podolin, GlaxoSmith- Kline, Mail Code UW2532, 709 Swedeland Road, King of Prussia, PA 19406. E-mail address: [email protected] 6 2 Abbreviations used in this paper: RA, rheumatoid arthritis; cTEC, cortical thymic Current address: Thomas Jefferson University, Philadelphia, PA 19107. epithelial cells; CIA, collagen-induced arthritis; EAE, experimental autoimmune en- 3 Current address: Thomas Jefferson University, Philadelphia, PA 19107. cephalomyelitis; DPBS, Dulbecco’s PBS; MBP, myelin basic protein; LNC, lymph node cells; AUC, area under the curve. 4 Current address: Merck Research Laboratories, West Point, PA 19486. 5 Current address: Drexel University, Philadelphia, PA 19104. Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00 www.jimmunol.org 7990 ATTENUATION OF CIA AND EAE VIA CYSTEINE PROTEASE INHIBITION

transported through the to the endosome. In the Materials and Methods endosome, class II-bound Ii undergoes sequential proteolytic Inhibitors cleavages at the C terminus, generating Ii p22 (LIP, 22 kDa) and 5-(2-Morpholin-4-ylethoxy)benzofuran-2-carboxylic ((S)-3-Methyl- Ii p10 (SLIP, 10 kDa). The final N-terminal proteolytic cleavage of 1-{(S)-3-oxo-1-[2-(3-pyridin-2-ylphenyl)-acetyl]azepan-4-ylcarbamoyl} Ii generates CLIP (3 kDa), which consists of Ii residues 81–104. butyl) (SB-331750) and N-[1S-(2-phenylethyl)-3-phenylsulfonylal- CLIP, which binds to the Ag binding groove of the class II mol- lyl]-4-methyl-2R-piperazinyl carbonylaminovaleramide (APC3328) were ecule, is removed by a class II-like molecule, DM (HLA-DM/H- synthesized at GlaxoSmithKline by the Department of Medicinal Chem- istry, Microbial, Musculoskeletal, and Proliferative Diseases Center of Ex- 2M), enabling antigenic peptide to bind (2–4). If the final proteo- cellence for Drug Discovery, as previously described (24–26). lytic cleavage of Ii does not occur, Ii p10 rather than CLIP remains was purchased from Sigma-Aldrich. bound to the class II molecule. Although HLA-DM/H-2M can in- In vitro assays teract with class II-Ii p10 complexes, its preferred is class II-CLIP (5). Thus, in many cases, inhibition of CLIP generation Cathepsin activity assays were performed in microtiter plates using flu- orogenic substrates at concentrations equal to their Km in assay buffers prevents antigenic peptide from binding to the class II molecule ϩ optimized for each enzyme. The and mouse assays and being presented to CD4 T cells. consisted of 1.25 nM human and mouse cathepsin S, respectively, 30 ␮M Studies using targeted gene deletion and/or cysteine protease Z-Lys-Gln-Lys-Leu-Arg-AMC in 50 mM MES, 10 mM L-cysteine, 5 mM inhibitors have demonstrated a critical role for cathepsin S in the EDTA, and 0.5 mM CHAPS (pH 6.5). The human and V assays consisted of 100 pM human cathepsin L and V, respectively, 8.5 final proteolytic cleavage of Ii in murine B cells and dendritic cells, ␮M Z-Phe-Arg-AMC in 100 mM sodium acetate, 5 mM L-cysteine, and 5 while cathepsins S, L and F may collectively fulfill this function in mM EDTA (pH 5.5). The human assay consisted of 0.4 nM Downloaded from murine macrophages (6–9). In the mouse, cortical thymic epithe- human cathepsin F, 5 ␮M Z-Phe-Arg-AMC in 50 mM potassium phos- lial cells (cTEC), nonprofessional APC that mediate positive se- phate, 2.5 mM 1,4-dithioerythritol, 2.5 mM EDTA, and 0.6 mM CHAPS ϩ (pH 6.5). The human assay consisted of 0.1 nM human ca- lection of CD4 T cells, lack cathepsin S expression and use ca- thepsin K, 30 ␮M Z-Lys-Gln-Lys-Leu-Arg-AMC in 100 mM sodium thepsin L to mediate the final stage of Ii degradation (10), while in acetate, 20 mM L-cysteine, and 1 mM CHAPS (pH 5.5). The human ca- human cTEC, this process has been attributed to (11) thepsin C assay consisted of 0.010 nM human , 5 ␮M (H-Gly- Arg)2-R110 in 50 mM sodium acetate, 30 mM NaCl, 1 mM DTT, 1 mM

and/or cathepsin S (12). In contrast to murine cTEC, intestinal http://www.jimmunol.org/ EDTA, and 0.2 mM CHAPS (pH 5.5). The mouse cathepsin L assay con- epithelial cells in the mouse have been shown to use cathepsin S sisted of 100 pM mouse cathepsin L, 5 ␮M Z-Phe-Arg-AMC in 100 mM for late stage Ii processing (13). Thus, the requirement for a spe- sodium acetate, 20 mM cysteine, and 5 mM EDTA (pH 5.5). The mouse cific lysosomal cysteine protease in order for the final stage of Ii cathepsin K assay consisted of 1 nM mouse cathepsin K, 20 ␮M Z-F-R- processing and subsequent Ag presentation to occur depends, at AFC in 100 mM sodium acetate, 4 mM L-cysteine, 5 mM EDTA, and 1 mM CHAPS (pH 5.5). The mouse cathepsin C assay consisted of 0.015 nM least in part, on the APC type. In addition, the antigenic protein ␮ mouse cathepsin C, 5 M (H-Gly-Arg)2-R110 in 50 mM sodium acetate, and MHC class II allele involved in the interaction influence the 30 mM NaCl, 1 mM DTT, 1 mM EDTA, and 1 mM CHAPS (pH 5.5). All degree to which Ag presentation is dependent on cysteine pro- reactions were initiated by the addition of enzyme-containing assay buffer to the substrate and inhibitor sample and were read after1hofincubation tease-mediated late stage Ii degradation (7). This suggests the po- at 25°C. products of the fluorogenic peptide substrates were by guest on September 30, 2021 tential for inhibitors of cysteine proteases, such as cathepsins S, L, measured using an Analyst AD microtiter plate reader (Molecular Devices) F, and V, to prevent the presentation of self to autoreac- under the following conditions: for AMC substrates, excitation at 360 nm tive CD4ϩ T cells and, thus, inhibit autoimmunity, without result- and fluorescence emission at 460 nm using a 400 nm dichroic filter; for ϩ AFC substrates, excitation at 405 nm and fluorescence emission at 530 nm ing in general unresponsiveness of CD4 T cells to foreign Ags. using a 505 nm dichroic filter; and for R110 substrates, excitation at 485 The importance of the interaction between self peptide/MHC nm and fluorescence emission at 535 nm using a 505 nm dichroic filter. ϩ class II complexes and CD4 T cells in the generation of a number Compound IC50 values were determined from dose-response curves, which ␮ of autoimmune responses is underscored by the requirement for were performed using a starting compound concentration of 25 M that was serially diluted 3-fold over 11 repetitions using the following equation: CD4ϩ T cells to generate and/or transfer disease, and the associ- Ϫ ation between increased disease susceptibility and the expression y ϭ min ϩ ͑max-min͒/͑1 ϩ 10logIC50 x͒ of particular MHC class II alleles. Collagen-induced arthritis In the equation above, x is the log concentration of the test compound, y is (CIA) is a disease characterized by chronic inflammation of the the fluorescence response, min is the minimum response plateau, and max joint, leading to progressive destruction of cartilage and bone, and, is the maximum response plateau. Ki,app were calculated by dividing the IC50 values by two, as the assays were run at substrate concentrations equal thus, is a model that resembles many aspects of RA (14). Like RA, to Km, and the inhibition was assumed to be competitive. the development of CIA in rodents is dependent on the presence of CD4ϩ T cells (15, 16) and the expression of specific MHC class II Mice alleles (17, 18). In a second model of autoimmunity, experimental DBA/1 mice were obtained from Harlan U.K., and SJL/J mice were ob- autoimmune encephalomyelitis (EAE), inflammation of the CNS tained from Harlan France S.A.R.L. or from Taconic. C57BL/6 (B6) mice were obtained from The Jackson Laboratory. Cathepsin KϪ/Ϫ mice and results in demyelination, leading to paralysis. Thus, EAE is a rel- ϩ/ϩ ϩ K littermate controls were bred at the GlaxoSmithKline animal facility. evant model of multiple sclerosis (19, 20). CD4 T cells play a Generation of these mice has been previously described (27). All mice pivotal role in EAE development, as evidenced by the findings that were maintained on a 12 h:12 h light:dark cycle and fed standard rodent CD4-specific mAbs prevent EAE development, or reverse estab- chow and ad libitum. All procedures performed on animals were lished disease, and that CD4ϩ T cells transfer disease to naive reviewed and approved by GlaxoSmithKline Institutional Animal Care and Use Committee. recipients (21). Like CIA, the development of EAE is linked to the expression of particular MHC class II alleles (22, 23). Mouse splenocyte culture Based on the pivotal role of cysteine proteases, such as cathe- Mice were euthanized by CO2 asphyxiation and each spleen placed in 3 ml psins S, L, F and V, in the final stage of Ii degradation and sub- of sterile HBSS (Mediatech) containing 1 mM EGTA. Spleens were me- sequent presentation of antigenic peptide to CD4ϩ T cells, we chanically disrupted with forceps and pipetting before being passed through sterile 70-␮m filters. Cells were centrifuged at 600 ϫ g for 5 min, initiated studies to determine the effect of a pan-active cathepsin ϫ ϩ and RBC were lysed by resuspension of the pellet in 1 M-Lyse buffer inhibitor, SB-331750, in two models of Ag-dependent, CD4 T (R&D Systems) at 2 ml/spleen for 10 min followed by addition of 1ϫ wash cell-mediated autoimmunity: murine CIA and EAE. buffer (R&D Systems) at 8 ml/spleen. Cells were centrifuged as above, The Journal of Immunology 7991 washed in 10 ml of medium (RPMI 1640 containing 2 mM L-glutamine, human CD74 (Ii) mAb, PIN.1 (Santa Cruz Biotechnology) diluted 1/200 in 1.5 g/liter sodium bicarbonate, 4.5 g/liter glucose, 10 mM HEPES, 1 mM Membrane Blocking Buffer. Following incubation with primary mAbs of sodium pyruvate, 90%, and 10% heat inactivated FBS (HyClone), and overnight at 4°C with gentle shaking, the membranes were washed with 1ϫ counted. Cells were cultured at 1 ϫ 106 cells/ml of medium at 37°C in 5% washing buffer (TBS (Bio-Rad) containing 0.05% Tween 20) by quickly

CO2 on a shaking platform for 4–5 h. Where indicated, cultures were rinsing two times followed by three 5 min washes with vigorous shaking. treated with SB-331750, leupeptin, APC3328, or vehicle (0.1% DMSO). Membranes were then incubated with an HRP goat anti-rat IgG (Zymed Following incubation cells were examined for toxicity by trypan blue ex- Laboratories) diluted 1/5000 in Membrane Blocking Buffer, or with an clusion and then processed for immunoprecipitation/Western blot analysis HRP goat anti-mouse IgG (Santa Cruz Biotechnology) diluted 1/3000 in or for flow cytometry. membrane blocking buffer, respectively, for 75 min at room temperature with gentle shaking. The membranes were washed as described following Raji cell culture incubation with the primary Ab and then washed in TBS (Bio-Rad) by quickly rinsing two times followed by one 5-min wash with vigorous shak- The human B cell line, Raji, was purchased from American Type Culture ing. Membranes were treated with Luminol reagent (Santa Cruz Biotech- Collection. Cells were cultured at 5 ϫ 105 cells/ml of medium (RPMI 1640 nology) for 1 min and then exposed to film. containing 2 mM L-glutamine, 1.5 g/liter sodium bicarbonate, 4.5 g/liter glucose, 10 mM HEPES, 1 mM sodium pyruvate, 90%, and 10% heat Assessment of cell surface class II-CLIP and total class II inactivated FBS) at 37°C in 5% CO2 for 24 h. Where indicated, cultures were treated with SB-331750, leupeptin, or vehicle (0.1% DMSO), and expression then processed for immunoprecipitation/Western blot analysis or for flow cytometry. An aliquot of each splenocyte culture described above was centrifuged at 600 ϫ g for 5 min at 4°C, the pellets washed with 3 ml of cold Ab diluent Assessment of intracellular Ii processing intermediates (DPBS, 1% FBS, and 0.1% azide), and each pellet resuspended at 10 ϫ 106 cells/ml in cold Ab diluent containing 10 ␮g/ml of an anti-mouse CD16/ An aliquot of each splenocyte culture described above was centrifuged at CD32 mAb (2.4G2; BD Pharmingen) to block FcR binding. Following Downloaded from 600 ϫ g for 5 min, the pellets washed with 10 ml of cold Dulbecco’s PBS incubation for 5 min at 4°C, 100 ␮l aliquots of each sample (1 ϫ 106 cells) (DPBS), and each pellet resuspended in ice cold lysis buffer (10 mM Tris, were transferred to 96-well round-bottom plates and Abs added at the fol- 150 mM NaCl, 1% Triton X-100, 0.5 mM PMSF, and 1 protease inhibitor lowing amounts/106 cells: 0.5 ␮g of PE anti-mouse CD19 (1D3) or 0.5 ␮g minitab (Roche)) at 5–8.5 ϫ 107 cells/ml, and incubated on a tumbler for of the isotype control PE rat IgG2a (R35–95), purchased from BD Pharm- 30 min at 4°C. The lysates were then clarified at 12,000 ϫ g and the ingen; 20 ␮l(4␮g) of FITC anti-I-Ab/CLIP (15G4) or 20 ␮l of the isotype supernatants collected and stored at Ϫ80°C. A protein assay was performed control FITC mouse IgG1, purchased from Santa Cruz Biotechnology; 1 on each sample using the Bio-Rad protein assay (Bio-Rad). ␮ q b ␮

g of FITC anti-H-2 /H-2 (2G9) or 1 g of the isotype control FITC rat http://www.jimmunol.org/ Alternatively, an aliquot of each Raji cell culture described above was IgG2a (R35–95), purchased from BD Pharmingen; and 0.5 ␮g of Alexa centrifuged at 400 ϫ g for 5 min, the pellets washed with 10 ml of cold Fluor 488 anti-I-Aq/I-As (KH116) or 0.5 ␮g of the isotype control Alexa DPBS, and each pellet resuspended in ice cold lysis buffer at 15–30 ϫ 106 Fluor 488 mouse IgG2b, purchased from BioLegend. Samples were placed cells/ml. Suspensions were incubated on a tumbler for 30 min at 4°C, the on an orbital shaker for 30 min at 4°C, and then 100 ␮l of Ab diluent added lysates clarified at 12,000 ϫ g, and the supernatants collected and stored at and the cells pelleted as described above. Samples were washed twice with Ϫ80°C. A protein assay was performed on each sample using the Bio-Rad 200 ␮l of Ab diluent and then resuspended in 500 ␮l of Ab diluent for flow protein assay. cytometric analysis. Before analysis, 20 ␮l of Via-Probe (BD Pharmingen) For assessment of Ii processing intermediates in whole cell lysates, an was added to allow exclusion of dead cells from the analysis. equal amount of protein from each sample (2 ␮g for C57BL/6, DBA/1, and Alternatively, an aliquot of each Raji cell culture described above was SJL/J samples, and 2–20 ␮g for Raji samples) was diluted in an equal centrifuged at 400 ϫ g for 5 min, the pellets washed with 3 ml of cold Ab volume of loading buffer (62.5 mM Tris-HCl (pH 6.8), 25% glycerol, 4% diluent, and each pellet resuspended at 5–10 ϫ 106 cells/ml in cold Ab by guest on September 30, 2021 SDS, 0.01% bromophenol blue, and 5% 2-ME), and the samples placed in diluent. The 100 ␮l aliquots of each sample (0.5–1 ϫ 106 cells) were boiling water for 5 min. Samples were then loaded onto an 18 or 4–20% transferred to 96-well round-bottom plates and Abs added at the following Tris-HCl minigel (Bio-Rad). To assess class II-associated Ii processing amounts: 10 ␮l of FITC anti-human MHC class II/CLIP (CerCLIP) or 10 intermediates, equal amounts of total protein from each sample (150 ␮g for ␮l of the isotype control FITC mouse IgG1 (MOPC-21), purchased from mouse splenocyte lysates or 50 ␮g for Raji cell lysates) were diluted to a BD Pharmingen; and 10 ␮l of FITC anti-HLA-DR (G46-6) or 10 ␮lofthe final volume of 300 ␮l (splenocytes) or 200 ␮l (Raji cells) in lysis buffer. isotype control FITC mouse IgG2a (G155–178), purchased from BD The samples were precleared by adding isotype control and Protein A/G Pharmingen. Samples were placed on an orbital shaker for 30 min at 4°C, PLUS-Agarose (Santa Cruz Biotechnology) and using the following re- and then 100 ␮l of Ab diluent added and the cells pelleted as described agents: 3 ␮g of normal mouse IgG (Santa Cruz Biotechnology) for DBA/1 above. Samples were washed twice with 200 ␮l of Ab diluent and then and SJL/J samples, 3 ␮g of mouse IgG2a (BioLegend) for C57BL/6 sam- resuspended in 500 ␮l of Ab diluent for flow cytometric analysis. Before ples, and 1 ␮g of normal mouse IgG (Santa Cruz Biotechnology) for Raji analysis propidium iodide (Invitrogen) was added at 20 ␮g/ml, to allow samples. All mouse samples received 30 ␮l of Protein A/G PLUS-Agarose, exclusion of dead cells from the analysis. whereas Raji samples received 20 ␮l. The samples were then tumbled for Events were collected on an LSR or FACSCantos flow cytometer (BD 1 h at 4°C, centrifuged at 1000 ϫ g for 5 min, and supernatants collected. Pharmingen) using CellQuest or FACSDiva software, respectively, with To C57BL/6 supernatants, 3 ␮g of an anti-mouse I-Ab mAb (AF6–120.1; individual unlabeled and one-color samples being prepared to set color BioLegend) or 3 ␮g of the above mouse IgG2a, was added; to DBA/1 and compensation. All data were analyzed using CellQuest software (the SJL/J supernatants, 3 ␮g anti-I-Aq/I-As mAb (KH116; Santa Cruz Bio- FACSDiva-generated data was converted before analysis) gating on live technology) or 3 ␮g of the above normal mouse IgG, was added; and to cells, and using the geometric mean fluorescence 1 content as a direct Raji supernatants, 1 ␮g an anti-HLA-DR mAb (TAL 1B5, Santa Cruz assessment of class II/CLIP or total class II expression. The percent inhi- Biotechnology), or 1 ␮g of the above normal mouse IgG, was added. Sam- bition values of class II/CLIP or total class II expression by SB-331750- or ples were incubated overnight with tumbling at 4°C. Protein A/G PLUS- leupeptin-treated cells, relative to vehicle (0.1% DMSO)-treated cells, were Agarose (30 ␮l to mouse samples and 20 ␮l to Raji samples) was added, calculated following subtraction of the geometric mean fluorescence 1 con- and samples tumbled at 4°C for 2 h. Samples were then pelleted as above tent obtained with relevant isotype control mAb. and the pellets washed three times with 300 ␮l of lysis buffer. The samples were resuspended in 30 ␮l of loading buffer, placed in boiling water for 5 Induction and assessment of CIA min, and then loaded onto an 18 or 4–20% Tris-HCl minigel (Bio-Rad). were resolved in 1ϫ running buffer (25 mM Tris (pH 8.3), 192 On day 0, 12–24-wk-old male DBA/1 mice were immunized intradermally mM glycine, and 0.1% SDS (Bio-Rad)) at 126 V until the dye front reached at the base of the tail with a total of 100 ␮l of IFA (Sigma-Aldrich) con- the bottom of the gel, and were transferred to polyvinylidene difluoride taining 200 ␮g of bovine type II collagen ( Products) and 250 ␮gof membranes (Bio-Rad) in 1ϫ transfer buffer (25 mM Tris (pH 8.3), and 192 Mycobacterium tuberculosis H37Ra (Difco Laboratories). On day 21, mice nM glycine (Bio-Rad) containing 20% methanol) at 100 V for 1 h. The were boosted intradermally with 100 ␮l of PBS containing 200 ␮gof membranes were rinsed with Membrane Blocking Solution (Zymed Labo- bovine type II collagen. For prophylactic studies, SB-331750 (50 mg/kg) or artories) and then blocked with Membrane Blocking Solution for1hat vehicle (200 ␮l PBS) was administered s.c., twice a day (b.i.d.), from days room temperature with gentle shaking. Membranes containing mouse 1 to 40. For therapeutic studies, administration of SB-331750 or vehicle, as splenocyte proteins were incubated with an anti-mouse CD74 (Ii) mAb, described above, was initiated (day 1 of treatment) once an animal exhib- In-1 (BD Pharmingen) diluted 1/1000 in Membrane Blocking Buffer; mem- ited a clinical score of “1” or greater for 2 consecutive days (days 0 and 1 branes containing human Raji cell proteins were incubated with an anti- of treatment). Mice were scored daily for clinical symptoms of disease 7992 ATTENUATION OF CIA AND EAE VIA CYSTEINE PROTEASE INHIBITION using a micrometer caliper to measure paw thickness. Each paw was as- Mice were scored daily for clinical symptoms of disease based on the signed a score ranging from 0 to 4 based on the following criteria: 0, following criteria: 0, asymptomatic; 0.5, distal limp tail; 1.0, complete limp asymptomatic (paw thickness ϭ 1.8–1.9 mm and no swollen digits); 1, tail; 1.5, complete limp tail and unsteady gait; 2.0, paralysis of one hind paw thickness ϭ 1.8–1.9 mm and one or more swollen digits; 2, paw limb; 2.5, partial paralysis of both hind limbs; 3.0, complete bilateral hind thickness ϭ 2.0–2.5 mm and one or more swollen digits; 3, paw thick- limb paralysis; 3.5, complete bilateral hind limb paralysis with forelimb ness ϭ 2.6–3.0 mm and one or more swollen digits; and 4, paw thick- weakness; and 4.0, moribund. ness ϭ 3.0ϩ mm and one or more swollen digits. For each animal, the clinical score for all paws was summed, giving a potential score of 0–16. Ag- or anti-CD3 mAb-induced proliferation by mouse lymph In addition to the mice that were scored throughout the experiment (n ϭ node cells (LNC)/splenocytes 12 per treatment group), on days 28 and 41, mice (n ϭ 6 per treatment For measurement of Ag-induced proliferation, inguinal lymph nodes and group per time point) were removed from the experiment and used to spleens from collagen-immunized/boosted DBA/1 mice, and from MBP- measure paw tissue cytokine levels. These mice were scored on a daily immunized/boosted SJL/J mice (immunizations and boosts performed as basis until their removal from the study and the data integrated into the described above), were collected 24 and 17 days, respectively, postimmu- analysis using the log rank test. nization. Preparation and culture of LNC and splenocytes were performed Quantitative analysis of blood SB-331750 levels under sterile conditions. Lymph nodes and spleens were rinsed in HBSS containing penicillin (100 units/ml)-streptomycin (100 ␮g/ml) (Invitrogen) Analysis of mouse blood samples (25 ␮l of mouse blood diluted with 25 ␮l or gentamicin (50 ␮g/ml) (Sigma-Aldrich), (HBSSϩ), teased apart in 5 ml of water) for SB-331750 was performed using liquid chromatography/tan- of HBSSϩ, and filtered through 50-␮m filters. Samples were centrifuged at dem mass spectrometric detection. Analytical standards were prepared, and 500 ϫ g for 10 min at 4°C, and the resulting LNC pellets resuspended in the compound was isolated from the samples and standards by protein 2 ml HBSSϩ. RBC within splenocyte pellets were lysed, and following precipitation with acetonitrile containing a mass spectral internal standard centrifugation as above, splenocyte pellets were resuspended in 2 ml of ϩ and the resulting mixture was vortex-mixed followed by centrifugation. HBSS . LNC and splenocytes were counted, and cells from three mice Downloaded from The resulting supernatant was subjected to chromatographic separation on combined at a ratio of 80% LNC/20% splenocytes (each group of three a Luna C18 column under isocratic conditions (30/70 - 10 mM ammonium mice considered an n ϭ 1). A total of 1 ϫ 106 cells/ml were cultured in a acetate and acetonitrile mobile phases at a 300 ␮l/min flow rate). The volume of 200 ␮l of RPMI 1640 containing 10% FBS, and penicillin (100 eluent flowed into a Sciex API365 triple-quadrupole mass spectrometer units/ml)-streptomycin (100 ␮g/ml) or gentamicin (50 ␮g/ml), in the ab- (Applied Biosystems) using positive-ion electrospray multiple-reaction sence or presence of the indicated Ag (bovine type II collagen (100 ␮g/ml)) monitoring set to monitor for the (MϩH) plus precursor to product ion or bovine MBP (50 ␮g/ml), and to the Ag-stimulated LNC/splenocytes transition. cultures, SB-331750 was added to a final concentration of 0 (0.1% DMSO Data were reported as quantitative concentrations as determined by stan- vehicle), 0.3, 1, 3, or 10 ␮M. Following incubation for 72 h at 37°C in 5% http://www.jimmunol.org/ ␮ 3 dard calibration curve analysis, using linear fitting of a 1/x-weighted plot CO2,1 Ci [ H]thymidine was added to each well, and the cells cultured of the SB-331750/internal standard peak area ratios vs SB-331750 con- for an additional 24 h. Cell viability was confirmed by trypan blue exclu- centration, with a lower limit of quantification of 10 ng/ml. sion. Cultures were harvested using a Packard Filtermate 196 (Packard Instrument), and radioactivity was quantified using a Packard TopCount Paw tissue cytokine analysis liquid scintillation counter. For measurement of anti-CD3 mAb-induced proliferation, splenocytes Mice removed on days 28 (n ϭ 6) and 41 (n ϭ 6) of the CIA study, as well from naive DBA/1 and SJL/J mice were prepared as described above. A as those scored through day 50 (n ϭ 12), were euthanized on the indicated total of 2 ϫ 106 cells/ml were cultured in a volume of 100 ␮l of RPMI 1640 day by CO2 asphyxiation. The four paws from each mouse were collected containing 10% FBS, penicillin (100 units/ml)-streptomycin (100 ␮g/ml), and paws from two mice combined randomly within each treatment group, and gentamicin (50 ␮g/ml), in the absence or presence of 5 ␮g/ml of by guest on September 30, 2021 resulting in an n ϭ 3, 3, and 6 for days 28, 41, and 50, respectively. The plate-coated anti-mouse CD3 mAb (clone 500A2; Caltag Laboratories), tissue was weighed, flash frozen in liquid , ground using a mortar and to the anti-CD3 mAb-stimulated splenocytes cultures, SB-331750 was and pestle, and then homogenized for 1–2 min with a Polytron homoge- added to a final concentration of 0 (0.1% DMSO vehicle), 0.3, 1, 3, or 10 nizer (model PT 10/35; Brinkmann Instruments) in PBS at 1 ml/gram tis- ␮ ␮ ␮ M. Following incubation for 72 h at 37°C in 5% CO2,10 lof100 M sue. Large debris was removed by centrifugation at 4700 ϫ g. Supernatants BrdU labeling solution (BrdU Cell Proliferation ELISA; Roche Diagnostic were decanted and cleared with an additional centrifugation at 14000 ϫ g, Systems) was added to each well, and, after an additional 24 h of incuba- and the resulting supernatants frozen at Ϫ80°C until use. Cytokine levels tion, cultures were processed according to the manufacturer’s instructions were analyzed using mouse IL-1␤, TNF-␣, and IFN-␥ ELISA kits pur- and chemiluminescence measured using a Wallac Envision 2102 Multila- chased from R&D Systems. Assays were performed according to the man- bel Reader. ufacturer’s instructions. Statistical analysis Induction and assessment of EAE Statistical differences in cytokine levels were determined using the two- For induction of EAE by adoptive transfer, on day 0, 12–16-wk-old female tailed Student’s t test. Analysis of CIA and EAE disease severity was SJL/J mice were immunized s.c. at the base of the tail with a total of 100 performed by calculating, for each animal within a treatment group, the ␮l of IFA (Sigma-Aldrich) containing 400 ␮g of bovine myelin basic pro- area under the curve (AUC) using the trapezoidal rule, and then using tein (MBP) (Sigma-Aldrich) and 50 ␮gofM. tuberculosis H37Ra (Difco either the Student’s t test (for uncensored data), or the log rank test, which Laboratories). On day 7, mice were boosted s.c. in the flank with 100 ␮lof is a nonparametric test that allows for right censored observations (i.e., for IFA containing 400 ␮g of bovine MBP and 50 ␮gofM. tuberculosis any animal removed before the study’s completion, the animal’s complete H37Ra. On day 17 postimmunization, inguinal lymph nodes were collected AUC is considered to be at least as large as the partial AUC exhibited). In and single-cell suspensions prepared as described below. A total of 2 ϫ 106 cases where scores for some days were missing in the interior of the ani- cells/ml were cultured with 50 ␮g/ml of bovine MBP in a volume of 70 ml mal’s time profile, score values were interpolated over time. For CIA and of RPMI 1640 containing 10% FBS and gentamicin (100 ␮g/ml), in the EAE studies, day of disease onset was defined as the first of two consec- presence of vehicle (0.1% DMSO) or SB-331750 (10 ␮M). Following utive days on which an animal exhibited a clinical score Ͼ 0. Analysis of disease onset data was performed using the log rank test. Values of p Ͻ incubation for 72 h at 37°C in 5% CO2, cells were washed twice in PBS and counted, yielding 8.61 ϫ 105 cells/ml (DMSO-treated) and 6.31 ϫ 105 0.05 were considered statistically significant. cells/ml (SB-331750-treated). From each treatment group, 3 ϫ 107 cells in 0.5 ml of PBS were transferred by i.v. injection to naive female SJL/J Results recipients (n ϭ 8–9 per treatment group). Structure and characteristics of SB-331750 For active induction of EAE, on day 0, 12–16-wk-old female SJL/J mice were immunized s.c. in the right flank with a total of 100 ␮l of IFA (Sigma- SB-331750, 5-(2-Morpholin-4-ylethoxy)benzofuran-2-carboxylic Aldrich) containing 400 ␮g of bovine MBP (Sigma-Aldrich) and 50 ␮gof Acid ((S)-3-Methyl-1-{(S)-3-oxo-1-[2-(3-pyridin-2-ylphenyl)- M. tuberculosis H37Ra (Difco Laboratories). On days 0 and 2, mice re- acetyl]azepan-4-ylcarbamoyl}butyl)amide, is a pan-active inhibi- ceived an i.v. injection of 500 ng of perussis toxin (List Biological Labo- tor of the papain family of cysteine proteases, exhibiting potent ratories) in 200 ␮l of PBS. Therapeutic administration of vehicle (PBS) or SB-331750 (50 mg/kg, s.c., b.i.d) was initiated on the first day of exhibition activities for human cathepsins S, L, F, V and K, and weak activity of clinical symptoms (day 1 of treatment) and continued through day 14 for human cathepsin C. Similarly, SB-331750 is a potent inhibitor (n ϭ 4–5 mice per treatment group). of mouse cathepsins S, L, and K, but not mouse cathepsin C (Fig. The Journal of Immunology 7993

FIGURE 1. The chemical structure and selectivity profile of SB-331750, 5-(2-Morpholin-4-ylethoxy)benzofuran-2- ((S)-3-Methyl-1-{(S)-3-oxo-1-[2-(3-pyridin-2-ylphenyl)-acetyl]azepan-4- ylcarbamoyl}butyl)amide. ND ϭ Not determined. FIGURE 2. In vitro treatment with SB-331750 or broad-spectrum cys-

teine protease inhibitors results in inhibition of Ii processing by cathepsin Downloaded from 1). The activity of SB-331750 for mouse cathepsin F is undeter- KϪ/Ϫ splenocytes comparable to that exhibited by cathepsin Kϩ/ϩ spleno- ϩ ϩ Ϫ Ϫ mined. To date, no mouse orthologue of human cathepsin V has cytes. Splenocytes from cathepsin K / (A)orK / (B) mice were cul- been identified (28). SB-331750 does not exhibit activity for tured for 18 h in the absence or presence of vehicle (0.1% DMSO), leu- Ͼ ␮ peptin (200 ␮g/ml), APC3328 (5 ␮M), or SB-331750 (5 ␮M). Splenocytes , aspartyl, or metallo proteases (Ki 10 M), or for mem- bers of the family of cysteine proteases (K Ͼ 10 ␮M). The were lysed, and the resulting lysates subjected to SDS-PAGE on an 18% i Tris-HCl gel, followed by Western blot analysis using the mAb In-1. The representative members of these proteases classes tested were as http://www.jimmunol.org/ results are representative of two independent experiments. follows: , , , and (serine proteases); , , and ␤-secretase (aspartyl pro- teases); matrix -1, -2, -3, -7, -9, angiotensin-con- In vitro treatment with SB-331750 inhibits Ii processing by verting enzyme, and neutral (metallo proteases); DBA/1, SJL/J, and C57BL/6 splenocytes in a and caspase-3 and -4 (caspase family of cysteine proteases). concentration-dependent manner

In vitro treatment with SB-331750 results in comparable Inbred strains of mice, which differ at both MHC-linked and non- inhibition of Ii processing by cathepsin KϪ/Ϫ vs cathepsin Kϩ/ϩ MHC-linked loci, are susceptible to specific experimentally-in- duced autoimmune syndromes. DBA/1 mice, which express the splenocytes by guest on September 30, 2021 H-2q haplotype, exhibit susceptibility to type II collagen-induced Murine cathepsins S, L, and F have been implicated in the final arthritis (14, 17), whereas SJL/J (H-2s) and C57BL/6 (H-2b) mice proteolytic cleavage of Ii (6–10, 13), whereas cathepsin K, to our develop EAE following immunization with the appropriate com- knowledge, has not. To demonstrate the cell-based inhibition of Ii ponent(s) of myelin (29, 30). To determine whether the genetic processing by SB-331750, and to determine whether the activity differences among these strains impact Ii processing, splenocytes exhibited by SB-331750 for mouse cathepsin K (Fig. 1) contrib- from DBA/1, SJL/J, and C57BL/6 mice were treated for 4.5 h with ϩ/ϩ utes to the compound’s inhibition of Ii processing, cathepsin K varying concentrations of SB-331750 and Ii p10 accumulation was Ϫ/Ϫ and K splenocytes were incubated for 18 h in the absence (Un- assessed. SB-331750 induced a concentration-dependent increase treated) or presence of vehicle (0.1% DMSO), SB-331750 (5 ␮M), in Ii p10 accumulation in DBA/1, SJL/J, and C57BL/6 splenocytes or the broad-spectrum cysteine protease inhibitors, leupeptin (200 ␮g/ml) or APC3328 (5 ␮M). Cell lysates were then subjected to SDS-PAGE and Western blot analysis using the anti-mouse Ii mAb, In-1. Untreated cathepsin Kϩ/ϩ splenocytes and those cul- tured in the presence of DMSO expressed both the p41 (data not shown) and p31 isoforms of Ii, but no Ii degradation intermediates. In contrast, cathepsin Kϩ/ϩ splenocytes treated with leupeptin, APC3328, or SB-331750 exhibited prominent accumulation of the CLIP precursor Ii p10 with APC3328 inducing weak accumulation of the Ii p10 precursor Ii p22 as well (Fig. 2A). Under the same experimental conditions, cathepsin KϪ/Ϫ splenocytes (Fig. 2B)ex- hibited Ii processing intermediates comparable to those exhibited by cathepsin Kϩ/ϩ splenocytes. In additional studies, the accumu- lation of Ii processing intermediates was found to occur rapidly, such that treatment of cells for 4–5 h resulted in consistent and quantifiable accumulation of Ii p10 (Fig. 3). In the experiment FIGURE 3. In vitro treatment with SB-331750 induces concentration- depicted in Fig. 2, and in those subsequently described, viability of dependent accumulation of Ii p10 in DBA/1, SJL/J, and C57BL/6 spleno- mouse splenocytes following 18 h or 4–5 h of treatment with ve- cytes. Splenocytes from DBA/1, SJL/J, and C57BL/6 mice were cultured hicle or inhibitor was confirmed by trypan blue exclusion. These for 4.5 h in the presence of SB-331750 at the indicated concentrations. results demonstrate the cell-based inhibition of late stage Ii pro- Following lysis the cells were subjected to SDS-PAGE on an 18% Tris- cessing by SB-331750 and indicate that cathepsin K does not play HCl gel, and then Western blot analysis using the mAb In-1. The results are a role in Ii processing in these cells. representative of four independent experiments. 7994 ATTENUATION OF CIA AND EAE VIA CYSTEINE PROTEASE INHIBITION

FIGURE 4. In vitro treatment with SB-331750 or leupeptin induces ac- cumulation of MHC class II-associ- ated Ii processing intermediates, and decreases cell surface expression of class II/CLIP complexes, by Raji cells. Raji cells were cultured for 24 h in the presence of vehicle (0.1% DMSO), SB-331750 (10 ␮M), or leu- peptin (200 ␮g/ml), and cells lysed for SDS-PAGE/Western blot analy- sis, or intact cells subjected to flow cytometric analysis. Nonimmunopre- cipitated cell lysates (A), and cell ly- sates immunoprecipitated with the anti-HLA-DR mAb, TAL 1B5 (B and C), or with control mouse IgG (C), were subjected to SDS-PAGE on an 18 or 4–20% Tris-HCl gel, and then Downloaded from Western blot analysis using the anti- human Ii mAb PIN.1. The results are representative of two independent ex- periments. Intact Raji cells were stained with the FITC anti-human MHC class II/CLIP mAb CerCLIP or

with control FITC mouse IgG1 (not http://www.jimmunol.org/ shown) (D), or with the FITC anti- HLA-DR mAb G46-6 or with control FITC mouse IgG2a (not shown) (E), and subjected to flow cytometric anal- ysis. Dead cells were excluded using propidium iodide. The results are rep- resentative of three to four indepen- dent experiments. by guest on September 30, 2021

(Fig. 3), whereas untreated and vehicle (0.1% DMSO)-treated peptin-treated Raji cells demonstrated marked accumulation of Ii splenocytes from each strain exhibited no accumulation of Ii p10 p10 and detectable accumulation of Ii p22 (Fig. 4A). Comparable (data not shown). Interestingly, the pattern of Ii p10-fragment ac- results were observed following immunoprecipitation of Raji cumulation differed among the strains, suggesting heterogeneity in whole cell lysates with TAL 1B5 (Fig. 4, B and C), but not with the proteolytic events leading to Ii p10 generation. These data control mouse IgG (Fig. 4C). Ig H and L chains were detected due indicate that SB-331750 inhibits the final stage of Ii processing in to recognition of the mouse immunoprecipitating Abs by the HRP- genetically distinct strains of mice, allowing for investigation of conjugated goat anti-mouse IgG mAb used as the detection re- the effects of inhibition of CLIP generation in multiple models of agent. To determine whether the SB-331750- and leupeptin-medi- autoimmune disease. ated inhibition of class II-associated Ii processing results in modulation of class II/CLIP complexes on the cell surface, In vitro treatment with SB-331750 inhibits processing of MHC DMSO-, SB-331750-, or leupeptin-treated Raji cells were ana- class II-associated Ii, and decreases cell surface expression of lyzed by flow cytometry using the mAb CerCLIP. As shown in class II/CLIP complexes, by Raji cells and C57BL/6 splenocytes Fig. 4D, cells treated with SB-331750 or leupeptin exhibited de- It has been demonstrated that within the cell, Ii is present in large creased class II-associated CLIP expression compared with cells excess relative to class II molecules. In the absence of association treated with DMSO. In three to four experiments, SB-331750 and with class II molecules, this pool of Ii is localized primarily in the leupeptin induced mean decreases in class II-associated CLIP ex- endoplasmic reticulum (31). To determine whether the SB- pression of 65.4% (range, 62.1–71.3%) and 74.4% (range, 72.6– 331750-mediated inhibition of late stage Ii processing observed in 76.1%), respectively. Total class II expression on the cell surface, whole cell lysates is representative of the compound’s activity on detected by the FITC anti-HLA-DR mAb G46-6, was decreased class II-associated Ii, human Raji cells were cultured for 24 h with modestly, but detectably, by SB-331750 (mean reduction of vehicle (0.1% DMSO), SB-331750 (10 ␮M), or leupeptin (200 17.3%; range, 11.2–23.3%) and leupeptin (mean reduction of ␮g/ml), and an aliquot of the cell lysates from each treatment 29.8%; range, 22.1–34.2%) (Fig. 4E). Treatment of Raji cells with group immunoprecipitated using the mouse anti-HLA-DR mAb, SB-331750 at 30 ␮M induced mean decreases of 71.4% (range, TAL 1B5, or control mouse IgG. Nonimmunoprecipitated and im- 70.3–73.1%) and 23.5% (range, 17.9–29.0%) in class II-associated munoprecipitated cell lysates were then subjected to SDS-PAGE CLIP expression and total class II expression, respectively (data and Western blot analysis using the anti-human Ii mAb PIN.1. not shown). Similar to the results observed with nonimmunoprecipitated whole Similar results were obtained following culture of C57BL/6 cell lysates of inhibitor-treated mouse splenocytes (Figs. 2 and 3), splenocytes for 4 h with vehicle (0.1% DMSO), SB-331750 (10 nonimmunoprecipitated whole cell lysates of SB-331750- or leu- ␮M), or leupeptin (200 ␮g/ml). The prominent accumulation of Ii The Journal of Immunology 7995

FIGURE 5. In vitro treatment with SB-331750 or leupeptin induces accu- mulation of MHC class II-associated Ii processing intermediates, and decreases cell surface expression of class II/CLIP complexes, by C57BL/6 splenocytes. Splenocytes from C57BL/6 mice were cultured for4hinthepresence of ve- hicle (0.1% DMSO), SB-331750 (10 ␮M), or leupeptin (200 ␮g/ml), and cells lysed for SDS-PAGE/Western blot analysis, or intact cells subjected to flow cytometric analysis. Nonimmuno- precipitated cell lysates (A and lane 1 of B), and cell lysates immunoprecipitated with the anti-I-Ab mAb, AF6–120.1 (B and C), or with control mouse IgG2a (C), were subjected to SDS-PAGE on an 18% Tris-HCl gel, and then Western Downloaded from blot analysis using the anti-mouse Ii mAb In-1. The results are representa- tive of two pooled mice/experiment in three independent experiments. Intact C57BL/6 splenocytes were stained with the PE anti-mouse CD19 mAb 1D3 or

with control PE rat IgG2a, to allow gat- http://www.jimmunol.org/ ing on CD19ϩ cells. In addition, splenocytes were stained with the FITC anti-I-Ab/CLIP mAb 15G4 or with con- trol FITC mouse IgG1 (not shown) (D), or with the FITC anti-H-2q/H-2b mAb 2G9 or with control FITC rat IgG2a (not shown) (E), and subjected to flow cytometric analysis. Dead cells were excluded using 7-amino-actinomycin D. The results are representative of 1 by guest on September 30, 2021 mouse/experiment in three independent experiments.

p10 and weak accumulation of Ii p22 observed in SB-331750- or Absence of detectable MHC class II-associated Ii processing leupeptin-treated nonimmunoprecipitated whole cell lysates (Fig. intermediates in SB-331750-treated DBA/1 and SJL/J splenocytes 5A) was also seen following immunoprecipitation of whole cell Similar to the results obtained with C57BL/6 splenocytes, follow- lysates with the anti-I-Ab mAb AF6–120.1 (Fig. 5, B and C), but ing in vitro culture of DBA/1 (Fig. 6A) and SJL/J splenocytes (Fig. not with control mouse IgG2a (Fig. 5C). CD19ϩ B cells expressed 7A) for 4 h with SB-331750 (10 ␮M) or leupeptin (200 ␮g/ml), moderate/high cell surface levels of class II/CLIP complexes, (de- nonimmunoprecipitated whole cell lysates exhibited marked accu- tected by the mAb 15G4; Fig. 5D) and total class II molecules mulation of Ii p10 and weak accumulation of Ii p22, compared (detected by the mAb 2G9; Fig. 5E), both of which were decreased with nonimmunoprecipitated whole cell lysates of relevant vehicle following treatment of splenocytes with SB-331750 or leupeptin. (0.1% DMSO)-treated splenocytes. Interestingly, Ii p10 accumu- Class II/CLIP complexes were reduced by an average of 47.1% ϭ lation was not detected following immunoprecipitation of DBA/1 (range, 46.0–47.6%, n 3 mice) by SB-331750, and 31.6% q s (range, 29.6–33.6%, n ϭ 3 mice) by leupeptin (Fig. 5D), while (Fig. 6B) and SJL/J (Fig. 7B) lysates with the anti-I-A /I-A mAb total class II molecules were reduced by an average of 45.9% KH116. In the DBA/1 immunoprecipitates, the identity of the spe- (range, 43.1–47.5%, n ϭ 3 mice) by SB-331750, and 25.1% cies that migrated coincident with Ii p22 is unknown, although its (range, 22.9–26.5%, n ϭ 3 mice) by leupeptin (Fig. 5E). CD19Ϫ appearance following immunoprecipitation with control mouse splenocytes exhibited minimally detectable levels of surface class IgG as well as with KH116, from lysates of both vehicle-treated II/CLIP complexes, and low levels of total class II molecules. To- and inhibitor-treated cells (Fig. 6B), suggests that this band is non- tal class II molecules were decreased by an average of 14.1% specific and does not represent a class II-associated Ii processing (range, 9.2–17.5%, n ϭ 3 mice) and 12.8% (range, 7.8–19.0%, intermediate. To confirm that the lack of detectable class II-asso- n ϭ 3 mice) by SB-331750 and leupeptin, respectively, compared ciated Ii p10 in these strains was not unique to use of KH116 as the with vehicle-treated CD19Ϫ splenocytes (data not shown). immunoprecipitating mAb, DBA/1 whole cell lysates were immu- These data indicate that in Raji cells and C57BL/6 splenocytes, noprecipitated with the anti-H-2q/H-2b mAb 2G9 and similar re- SB-331750 inhibits late stage processing of class II-associated Ii, sults were obtained (data not shown). Levels of surface class II/ and reduces expression of class II/CLIP complexes and total class CLIP on CD19ϩ and CD19Ϫ splenocytes from DBA/1 and SJL/J II molecules on the cell surface. mice, detected by the mAb 15G4, were not elevated above levels 7996 ATTENUATION OF CIA AND EAE VIA CYSTEINE PROTEASE INHIBITION

FIGURE 6. Absence of detectable MHC class II-asso- ciated Ii processing intermediates in SB-331750- or leu- peptin-treated DBA/1 splenocytes. Splenocytes from DBA/1 mice were cultured for4hinthepresence of ve- hicle (0.1% DMSO), SB-331750 (10 ␮M), or leupeptin (200 ␮g/ml), and cells lysed for SDS-PAGE/Western blot analysis, or intact cells subjected to flow cytometric anal- ysis. Nonimmunoprecipitated cell lysates (A and lane 1 of B), and cell lysates immunoprecipitated with the anti-I-Aq/ I-As mAb KH116 or with control mouse IgG (B), were subjected to SDS-PAGE on an 18% Tris-HCl gel, and then Western blot analysis using the anti-mouse Ii mAb In-1. The results are representative of three to four pooled mice/ experiment in four independent experiments. Intact DBA/1 splenocytes were stained with the PE anti-mouse CD19 mAb 1D3 or with control PE rat IgG2a, to allow gating on CD19ϩ cells. In addition, splenocytes were stained with the FITC anti-H-2q/H-2b mAb 2G9 or with control FITC rat IgG2a (not shown) (C), and subjected to

flow cytometric analysis. Dead cells were excluded using Downloaded from 7-amino-actinomycin D. The results are representative of one mouse/experiment in three independent experiments.

detected by an isotype control mAb (data not shown), although (range, 0.43–10.7%, n ϭ 3 mice), on CD19ϩ and CD19Ϫ spleno-

interpretation of this result is confounded by the fact that we can- cytes, respectively, and leupeptin induced mean decreases of 18.0% http://www.jimmunol.org/ not distinguish between absence of surface class II/CLIP expres- (range, 16.1–19.9%, n ϭ 2 mice) and 10.9% (range, 6.0–15.8%, n ϭ sion, and lack of cross-reactivity of 15G4 with the H-2q and H-2s 2 mice), respectively, compared with DMSO-treated CD19ϩ and haplotypes. CD19ϩ B cells from DBA/1 mice expressed high lev- CD19Ϫ splenocytes, respectively. els of total class II molecules on the cell surface (detected by the These data suggest that in two autoimmune-prone strains of mAb 2G9), which were decreased by an average of 46.7% (range, mice, DBA/1 and SJL/J, Ii processing intermediates may readily 44.7–50.5%, n ϭ 3 mice) as a result of treatment with SB-331750, dissociate from MHC class II molecules. Interestingly, similar to and by an average of 33.0% (range, 32.7–33.3%, n ϭ 3 mice) the inhibitor-mediated modulation of total class II expression on following treatment with leupeptin (Fig. 6C). Expression of total the surface of Raji cells and C57BL/6 splenocytes, SB-331750 class II molecules on the surface CD19Ϫ DBA/1 splenocytes was low, treatment resulted in varying degrees of reduction of surface class by guest on September 30, 2021 and was decreased by an average of 14.9% (range, 11.0–18.7%, n ϭ II expression by DBA/1 and SJL/J splenocytes. 3 mice) and 13.5% (range, 3.6–21.2%, n ϭ 3 mice), following treat- ment with SB-331750 and leupeptin, respectively (data not shown). In vitro treatment with SB-331750 inhibits Ag-induced High and low levels of total MHC class II (detected by the mAb proliferation by LNC/splenocytes from collagen-sensitized KH116) were also observed on CD19ϩ (Fig. 7C) and CD19Ϫ (data DBA/1 mice and MBP-sensitized SJL/J mice not shown) SJL/J splenocytes, respectively. SB-331750 induced mean To confirm that inhibition of late stage Ii processing by SB-331750 decreases of 18.6% (range, 17.5–19.4%, n ϭ 3 mice) and 4.9% results in the reduction of Ag-induced cellular responsiveness,

FIGURE 7. Absence of detectable MHC class II-asso- ciated Ii processing intermediates in SB-331750- or leu- peptin-treated SJL/J splenocytes. Splenocytes from SJL/J mice were cultured for4hinthepresence of vehicle (0.1% DMSO), SB-331750 (10 ␮M), or leupeptin (200 ␮g/ml), and cells lysed for SDS-PAGE/Western blot analysis, or intact cells subjected to flow cytometric analysis. Nonim- munoprecipitated cell lysates (A and lane 1 of B), and cell lysates immunoprecipitated with the anti-I-Aq/I-As mAb KH116 or with control mouse IgG (B), were subjected to SDS-PAGE on an 18% Tris-HCl gel, and then Western blot analysis using the anti-mouse Ii mAb In-1. The results are representative of one to four pooled mice/experiment in three independent experiments. Intact SJL/J splenocytes were stained with the PE anti-mouse CD19 mAb 1D3 or with control PE rat IgG2a, to allow gating on CD19ϩ cells. In addition, splenocytes were stained with the AlexaFluor 488 anti-I-As/I-Aq mAb KH116 or with control Alexa- Fluor 488 mouse IgG2b (not shown) (C), and subjected to flow cytometric analysis. Dead cells were excluded using 7-amino-actinomycin D. The results are representative of 1 mouse/experiment in three independent experiments. The Journal of Immunology 7997 Downloaded from http://www.jimmunol.org/

FIGURE 8. In vitro treatment with SB-331750 inhibits Ag-induced proliferation by LNC/splenocytes from collagen-sensitized DBA/1 and MBP-sensitized SJL/J mice in a concentration-dependent manner. For Ag-induced proliferative responses, inguinal lymph nodes and spleens from collagen-immunized/boosted DBA/1 mice (A), and from MBP-immunized/boosted SJL/J mice (B), were collected 24 and 17 days, respectively, postimmunization. Following preparation of single-cell suspensions, LNC and splenocytes were combined at a ratio of 80% LNC/20% splenocytes, giving an n ϭ 3 pooled groups. LNC/splenocytes were cultured in the absence (Unstim.) or presence of type II collagen (100 ␮g/ml) (A) or MBP (50 ␮g/ml) (B). To Ag-stimulated LNC/splenocytes, SB-331750 was added at the concentrations indicated. Cells were cultured for 96 h with the final 24 h of incubation occurring in the presence of [3H]thymidine. Data are expressed by guest on September 30, 2021 as mean Ϯ SEM. The results are representative of three independent experiments. For anti-CD3 mAb-induced proliferative responses, single-cell suspensions were prepared from the spleens of naive DBA/1 (C) and SJL/J (D) mice, and cultured in the absence (Unstim.) or presence of the plate-coated anti-mouse CD3 mAb 500A2 (5 ␮g/ml). To the anti-CD3 mAb-stimulated splenocytes, SB-331750 was added at the concentrations indicated. Cells were cultured for 96 h with the final 24 h of incubation occurring in the presence of BrdU. Data are expressed as mean Ϯ SEM. The results are representative of two independent experiments.

LNC/splenocytes from collagen-immunized/boosted DBA/1 mice, 90 min after a single administration of SB-331750 or vehicle on and from MBP-immunized/boosted SJL/J mice, were collected 24 day 4, spleens were collected and lysed. The resulting whole and 17 days, respectively, postimmunization and stimulated with cell lysates were then subjected to SDS-PAGE and Western blot the relevant immunizing Ag. Culture of collagen-primed DBA/1 analysis using the mAb In-1 to assess Ii p10 accumulation. As cells (Fig. 8A), and of MBP-primed SJL/J cells (Fig. 8B), with shown in Fig. 9, DBA/1, SJL/J, and C57BL/6 mice treated with collagen and MBP, respectively, resulted in robust proliferative SB-331750 exhibited enhanced accumulation of Ii p10 com- responses, each of which was inhibited in a concentration-depen- pared with DBA/1, SJL/J, and C57BL/6 mice, respectively, dent manner by SB-331750. This inhibition was not observed in treated with vehicle, demonstrating that in vivo administration DBA/1 (Fig. 8C) and SJL/J (Fig. 8D) splenocytes induced to pro- of SB-331750 results in the inhibition of processing of Ii p10 liferate via a mechanism not dependent on presentation of Ag, i.e., to CLIP. stimulation with anti-CD3 mAb. A similar lack of inhibition was observed when proliferation was induced by mitogen (data not In vivo prophylactic administration of SB-331750 decreases the shown). Viability of cells following culture with vehicle or SB- severity and delays the onset of CIA, and reduces levels of 331750 was confirmed by trypan blue exclusion. These results tissue proinflammatory cytokines, in DBA/1 mice demonstrate that in conjunction with its ability to inhibit late stage To determine whether the inhibition of Ii processing and col- Ii processing in DBA/1 and SJL/J APC, in vitro treatment with lagen-induced T cell proliferation by SB-331750 is sufficient to SB-331750 prevents T cell proliferation in response to Ags that prevent the development of CIA, collagen-immunized and induce autoimmune syndromes in these strains of mice. boosted DBA/1 mice (24 wk of age at the time of immuniza- tion) were administered SB-331750 (50 mg/kg) or vehicle In vivo administration of SB-331750 inhibits Ii processing by (PBS) s.c., b.i.d. from days 1 to 40. In a previous pharmacoki- DBA/1, SJL/J, and C57BL/6 splenocytes netic study, blood concentrations of SB-331750 were measured To demonstrate the inhibition of Ii processing by SB-331750 in in samples from three DBA/1 mice that were administered SB- vivo, DBA/1, SJL/J, and C57BL/6 mice were administered SB- 331750 (50 mg/kg) s.c., b.i.d. on day 0, and once on day 1. 331750 (50 mg/kg) or vehicle (PBS) s.c., b.i.d. on days 1–3, and Immediately before compound administration on day 1, blood 7998 ATTENUATION OF CIA AND EAE VIA CYSTEINE PROTEASE INHIBITION

10B). All mice in the vehicle-treated group and in the SB-331750- treated group exhibited clinical symptoms of disease during the study, indicating no modulation of cumulative disease incidence by SB-331750. Consistent with the reduced severity of disease, SB-331750- treated mice exhibited decreased paw tissue levels of the proin- flammatory cytokines IL-1␤, TNF-␣, and IFN-␥, compared with vehicle-treated mice. On day 28 (early disease), SB- 331750-treated vs vehicle-treated mice exhibited comparable tissue levels of each cytokine, while on day 41 (mid-disease and immediately following completion of compound administra- tion), tissues collected from SB-331750-treated mice contained significantly reduced IL-1␤ ( p Ͻ 0.001) and TNF-␣ ( p Ͻ 0.05) levels, and a trend toward reduced IFN-␥ ( p ϭ 0.06) levels, compared with tissues collected from vehicle-treated mice (Fig. 10C). On day 50, 10 days following the final administration of inhibitor, paw tissue levels of IL-1␤ ( p Ͻ 0.001) and TNF-␣ FIGURE 9. In vivo administration of SB-331750 induces accumulation ( p Ͻ 0.05), but not IFN-␥, remained significantly reduced. of Ii p10 in DBA/1, SJL/J, and C57BL/6 splenocytes. DBA/1, SJL/J, and These results indicate that prophylactic administration of SB- Downloaded from ϭ C57BL/6 mice were administered vehicle (PBS) (n 2 mice/strain) or 331750 results in the reduction of both the proinflammatory SB-331750 (n ϭ 2 mice/strain) at 50 mg/kg s.c., b.i.d on days 1–3, and 90 cytokine production and clinical manifestations associated with min after a single administration of vehicle or SB-331750 on day 4, spleens were collected. Splenocytes were lysed, and the resulting lysates subjected CIA, potentially via its inhibition of late stage Ii processing and to SDS-PAGE on an 18% Tris-HCl gel, followed by Western blot analysis subsequent Ag (collagen)-induced T cell proliferation. using the mAb In-1. The results are representative of two independent In vivo therapeutic administration of SB-331750 decreases the experiments. http://www.jimmunol.org/ severity of CIA in the DBA/1 mouse To determine whether SB-331750 is capable of modulating the (trough) levels of SB-331750 were 127 Ϯ 48 nM, and at 15, 30, severity of CIA when administered therapeutically, collagen- 60, 120, and 240 min following compound administration on immunized and boosted DBA/1 mice were injected with SB- day 1, blood levels of SB-331750 were 594 Ϯ 52, 619 Ϯ 92, 331750 (50 mg/kg) or vehicle (PBS) s.c., b.i.d. following the 479 Ϯ 61, 764 Ϯ 124, and 184 Ϯ 31 nM, respectively. onset of clinical symptoms. We have observed that mice im- As shown in Fig. 10A, the severity of CIA, represented by mean munized at 12 wk of age exhibit moderate disease severity (Fig. clinical score, was significantly reduced in mice treated with SB- 11A), whereas mice immunized at 24 wk of age develop severe 331750 ( p Ͻ 0.001), compared with mice treated with vehicle disease (Fig. 11B). Under moderate disease conditions, SB- by guest on September 30, 2021 (PBS). In addition, the time of onset of disease was significantly 331750-treated mice exhibited a significantly decreased mean delayed as a result of treatment with SB-331750 ( p Ͻ 0.01) (Fig. clinical score compared with vehicle (PBS)-treated mice

FIGURE 10. In vivo prophylactic admin- istration of SB-331750 decreases the sever- ity and delays the onset of CIA, and reduces paw tissue levels of IL-1␤, TNF-␣, and IFN-␥. The 24-wk-old male DBA/1 mice immunized on day 0 and boosted on day 21 with type II collagen were administered vehicle (PBS) or SB-331750 at 50 mg/kg s.c., b.i.d. on days 1–40, and scored for clinical symptoms of dis- ease. For A, data are expressed as mean Ϯ SEM (initial n ϭ 24 mice, final n ϭ 12 mice). For B, data are expressed as the percentage of mice remaining free of disease (i.e., day of dis- ease onset was defined as the first of two con- secutive days on which an animal exhibited a clinical score Ͼ 0). Thick vertical lines indi- cate censored observations. For C, four paws were collected from each mouse on days 28 (n ϭ 6 mice), 41 (n ϭ 6 mice), and 50 (n ϭ 12 mice). For each time point, paws from two mice were combined randomly within each treatment group, resulting in an n ϭ 3, 3, and 6 for days, 28, 41, and 50, respectively, and supernatants from combined tissue homoge- nates assayed for IL-1␤, TNF-␣, and IFN-␥. p Ͻ ,ء .Data are expressed as mean Ϯ SEM -p Ͻ 0.01, compared with the rel ,ءء and ;0.05 evant vehicle-treated control group. The Journal of Immunology 7999

FIGURE 11. In vivo therapeutic administration of SB-331750 decreases the severity of CIA. Following exhibition of clinical symptoms for 2 consecutive days (days 0 and 1), collagen-immunized/boosted DBA/1 mice that were 12 (A)or24(B) wk of age at the time of immunization were administered vehicle (PBS) or SB-331750 at 50 mg/kg s.c., b.i.d. on days 1–26 (A) or on days 1–24 (B). In B, an additional treatment group was administered SB-331750 at 50 mg/kg s.c., b.i.d. on days 1–5, and then administered vehicle (PBS) s.c., b.i.d on days 6–24. During the course of drug and/or vehicle administration, mice were scored for clinical symptoms of disease. Data are expressed as mean Ϯ SEM (n ϭ 8–13 mice).

( p Ͻ 0.001) (Fig. 11A). Under severe disease conditions, SB- ferred into naive SJL/J recipients. As shown in Fig. 12, A and B, Downloaded from 331750-treated mice exhibited a trend toward decreased disease respectively, the severity of EAE was significantly reduced ( p Ͻ severity, although statistical significance was not achieved (Fig. 0.01), and the onset significantly delayed ( p Ͻ 0.0001), in mice 11B). Under the latter conditions, in addition to administration that received SB-331750-treated LNC compared with mice that of SB-331750 from the onset of symptoms through completion received vehicle-treated LNC. Similar to the findings in CIA, of the study (days 1–24), mice were administered SB-331750 100% of mice in each treatment group exhibited clinical symptoms (50 mg/kg) s.c., b.i.d. on days 1–5, and then administered ve- of disease during the study. These results indicate that the presence http://www.jimmunol.org/ hicle (PBS) for the remainder of the study. As shown in Fig. of SB-331750 during Ag stimulation of MBP-primed LNC reduces 11B, discontinuation of SB-331750 administration resulted in the ability of the cells to adoptively transfer EAE to naive recip- an accelerated return of clinical symptoms to levels exhibited ients, likely through the inhibition of Ag (MBP) presentation to by vehicle-treated mice, although, based on the modest window CD4ϩ T cells. achieved in this study, no statistical difference between the two SB-331750 treatment groups was observed. These results indi- cate potential utility of cathepsin inhibitors in a clinically rel- In vivo therapeutic administration of SB-331750 decreases the evant dosing regimen, and suggest that continued inhibitor ad-

severity of EAE in the SJL/J mouse by guest on September 30, 2021 ministration may be required for extended therapeutic efficacy. To determine whether therapeutic administration of SB-331750 Future studies are planned to determine the effect of discon- modulates the severity of actively-induced EAE, MBP-immu- tinued SB-331750 administration under moderate disease nized SJL/J mice were injected with SB-331750 (50 mg/kg) or conditions. vehicle (PBS) s.c., b.i.d. following the onset of clinical symp- toms, through day 14 of the study. Immunization of SJL/J mice Treatment of MBP-primed LNC with SB-331750 reduces the with bovine MBP produced a mild, transient form of EAE, the severity and delays the onset of adoptively transferred EAE in severity of which was significantly reduced as a result of treat- the SJL/J mouse ment with SB-331750 ( p Ͻ 0.05) (Fig. 13). These results, together To characterize the effects of SB-331750 in a second model of with those obtained in the CIA model, indicate that when admin- Ag-driven autoimmunity, LNC from MBP-immunized/boosted istered in a clinically relevant dosing regimen, SB-331750 pro- SJL/J mice were cultured with MBP (50 ␮g/ml) in the presence of vides therapeutic benefit under mild/moderate autoimmune disease vehicle (0.1% DMSO) or SB-331750 (10 ␮M), and then trans- conditions.

FIGURE 12. Treatment of MBP-primed LNC with SB-331750 reduces the severity and delays the onset of adoptively transferred EAE. Female SJL/J mice were immunized on day 0 and boosted on day 7 with MBP, and inguinal lymph nodes collected on day 17. Following preparation of single-cell suspensions, LNC were cultured for 72 h with MBP (50 ␮g/ml) in the presence of vehicle (0.1% DMSO) or SB-331750 (10 ␮M), and then adoptively transferred to naive female SJL/J mice (n ϭ 8–9 mice). Recipient mice were scored for clinical symptoms of disease. For A, data are expressed as mean Ϯ SEM. For B, data are expressed as the percentage of mice remaining free of disease (i.e., day of disease onset was defined as the first of two consecutive days on which an animal exhibited a clinical score Ͼ 0). 8000 ATTENUATION OF CIA AND EAE VIA CYSTEINE PROTEASE INHIBITION

Similar to the inhibition of presentation of type II collagen to collagen-primed DBA/1 cells by SB-331750, Nakagawa et al. (7) reported a block in the presentation of the type II collagen epitope, 260–270, by macrophages and dendritic cells from ca- thepsin S-deficient DBA/1 mice. Both SB-331750-treated and cathepsin S-deficient DBA/1 mice exhibited delayed onset and reduced severity of CIA compared with relevant control mice. In addition, both cathepsin S deficiency and SB-331750 treat- ment resulted in the absence of detectable class II-associated Ii processing intermediates in DBA/1 (H-2q) splenocytes, but not FIGURE 13. In vivo therapeutic administration of SB-331750 de- in C57BL/6 (H-2b) splenocytes. In contrast to these similarities, creases the severity of EAE. Beginning on the first day of exhibition of cathepsin S deficiency induced minimal whole cell Ii fragment clinical symptoms (day 1 of treatment), MBP-immunized SJL/J mice were administered vehicle (PBS) or SB-331750 at 50 mg/kg s.c., b.i.d. Mice accumulation in DBA/1 splenocytes compared with the accu- received drug or vehicle on days 1–14, and were scored for clinical symp- mulation observed in C57BL/6 splenocytes (7), whereas SB- toms of disease through day 18. Data are expressed as mean Ϯ SEM (n ϭ 331750 treatment induced whole cell Ii fragment accumulation 4–5 mice). in both DBA/1 and C57BL/6 splenocytes. The lack of Ii frag- ment accumulation in the cathepsin S-deficient DBA/1 spleno- cytes, but not in cathepsin S-deficient C57BL/6 splenocytes, Discussion was suggested to result from efficient dissociation of Ii process- Downloaded from The studies described herein demonstrate that a small m.w., ing intermediates from I-Aq but not from I-Ab (7). If this is the pan-active inhibitor of the papain family of cysteine proteases, case, the accumulation of whole cell Ii processing intermediates SB-331750, attenuates late stage Ii processing by DBA/1 and in SB-331750-treated DBA/1 splenocytes, but not in cathepsin SJL/J splenocytes, and Ag-induced T cell proliferation by col- S-deficient DBA/1 splenocytes, may be due to increased sus- lagen-primed DBA/1 and MBP-primed SJL/J cells. In vivo pro- ceptibility of dissociated Ii fragments to degradation by cys- phylactic administration of SB-331750 to collagen immunized/ teine proteases other than cathepsin S, the activities of which http://www.jimmunol.org/ boosted DBA/1 mice delayed the onset and reduced the severity would be inhibited by SB-331750. of CIA, and resulted in decreased paw tissue levels of IL-1␤, In addition to DBA/1 splenocytes, SJL/J splenocytes, but not TNF-␣, and IFN-␥. Similarly, treatment of MBP-primed SJL/J Raji cells or C57BL/6 splenocytes, exhibited an absence of LNC with SB-331750 delayed the onset and reduced the sever- class II-associated Ii processing intermediates following treat- ity of adoptively transferred EAE. When administered thera- ment with SB-331750 or leupeptin. These results offer further peutically, SB-331750 reduced the severity of mild/moderate support for the occurrence of efficient dissociation of Ii degra- CIA and EAE. That SB-331750 exerts its effects on CIA and dation intermediates from selected MHC haplotypes. Class II EAE via inhibition of Ag presentation, and not an alternative

polymorphism has been shown to dramatically affect the affinity by guest on September 30, 2021 mechanism, is suggested by the following: SB-331750 inhibited of the class II-CLIP (39, 40) and class II-Ii p10 (7, 41) inter- processing of whole cell Ii p10 to CLIP, the preferred substrate actions. Whether decreased stability of class II/Ii fragment com- of HLA-DM/H-2M (5), by DBA/1 and SJL/J splenocytes in a plexes contributes to autoimmune susceptibility is unclear. Al- concentration-dependent manner, without affecting cell viabil- ity; SB-331750 inhibited Ag-induced, but not anti-CD3 mAb- though multiple autoimmunity-associated MHC class II or mitogen-induced, proliferation of collagen-primed DBA/1 molecules have been shown to form low stability complexes and MBP-primed SJL/J cells in a concentration-dependent man- with CLIP, including RA- (42), type 1 diabetes- (43, 44), and celiac disease- (45) associated allele products, expression of ner, suggesting that SB-331750 does not exert general anti- b proliferative effects; and SB-331750 exhibited no activity for low affinity I-A /CLIP complexes was reported not to alter the , metalloproteases, or serine proteases whose family peptide repertoire bound by class II molecules or increase the members include the cytokine-activating IL-1␤-con- susceptibility of the mice to autoimmunity (46). verting enzyme (32, 33), TNF-␣ converting enzyme (34), and Consistent with previous reports using leupeptin (47, 48), we proteinase 3 (35), respectively, the inhibition of which could observed reduced cell surface expression of total class II mol- potentially account for the observed reductions in paw tissue ecules as a result of SB-331750 or leupeptin treatment. This IL-1␤ and TNF-␣ levels. This report of attenuated autoimmune decrease may reflect inhibition of the formation of mature class disease development, specifically of murine CIA and EAE, as a II/peptide complexes normally expressed on the cell surface, result of deficient cysteine protease activity is consistent with although measurement of SDS-stable (high-affinity peptide the observations that cathepsin S-deficient mice exhibited re- loaded) class II complexes would be needed to confirm this. duced susceptibility to CIA (7) and myasthenia gravis (36), that Interestingly, the magnitude of the SB-331750- or leupeptin- the irreversible broad-spectrum protease inhibitor morpholin- induced decreases in surface class II levels did not appear to urea-leucyl-homophenylalanyl-vinylsulfone-phenyl (LHVS) de- correlate with the presence of stable class II-associated Ii pro- creases adjuvant-induced arthritis in the rat (37), and that a ca- cessing intermediates. All inhibitor-treated cells exhibited some thepsin S-selective inhibitor abrogated Sjo¨gren syndrome in the degree of reduced surface class II expression, with both Raji mouse (38). This report is, to our knowledge, the first demonstra- cells and SJL/J splenocytes, in which stable class II-associated tion of attenuation of CIA as a result of treatment with a cysteine Ii intermediate complexes were present and absent, respec- protease inhibitor and of modulation of EAE using either a phar- tively, demonstrating small decreases, and C57BL/6 spleno- macologic or genetic approach to inhibit cysteine protease activity. cytes and DBA/1 splenocytes, in which these complexes were Collectively, these observations suggest broad-spectrum and/or se- present and absent, respectively, demonstrating more substan- lective inhibition of members of the papain family of cysteine tial decreases. Consistent with the proposed interpretation of proteases as a potential therapeutic approach to the treatment of these data given above, Villadangos et al. (41) reported that autoimmunity. treatment of APC with the broad-spectrum cysteine protease The Journal of Immunology 8001 inhibitors, leupeptin or morpholinurea-leucyl-homophenylala- To date, evidence supporting a role for the papain family of nyl-vinylsulfone-phenyl (LHVS), prevented Ii degradation and cysteine proteases in the etiology of EAE has been indirect. It formation of SDS-stable class II/peptide complexes in an MHC has been demonstrated that the development of EAE by allele-dependent manner, with Ii p10 dissociating much more C57BL/6 mice in response to the dominant encephalitogenic s b efficiently from I-A than from I-A , and the latter exhibiting peptide, myelin oligodendrocyte glycoprotein (MOG)35–55,re- twice the reduction in class II/peptide complexes as the former. quires Ag processing and presentation in the CNS (58, 59). Based on the results described herein, a potential mechanism , resident CNS APC, express active cathepsins S and by which SB-331750 attenuates CIA and EAE is via the inhi- L (60, 61) and have been shown to function efficiently in the bition of late stage Ii processing, preventing the binding of col- presentation of myelin and other Ags, while the role of astro- lagen/MBP peptide to class II molecules, and the subsequent cytes in Ag presentation and T cell activation in the CNS re- presentation of class II/peptide complexes to CD4ϩ T cells. mains controversial (62–65). The purported lack of efficient However, the absence of stable class II/Ii fragment complexes Ag-presenting capacity by astrocytes has been ascribed to de- in SB-331750-treated DBA/1 and SJL/J splenocytes appears to creased cathepsin S and L activities, and subsequent inhibition argue against this mechanism. It is possible that this observa- of Ii processing (61). As described herein, the results of treat- tion could be explained on the basis of the reduced affinity of Ii ment of MBP-primed LNC with the pan-active cathepsin inhib- intermediates for class II molecules of the H-2q and H-2s hap- itor SB-331750 before transfer of the cells into naive recipients lotypes, coupled with the experimental manipulation necessary directly implicates one or more of these proteases in the devel- for detection of class II/Ii fragment complexes, facilitating en- opment of EAE. In EAE induced by adoptive transfer, Ag- ϩ hanced dissociation of the complexes under in vitro or ex vivo primed donor CD4 T cells collected from the periphery, sub- Downloaded from conditions. The class II-Ii association has been shown to be jected to an in vitro activation step, and transferred to the susceptible to disruption by detergents used for cell solubiliza- periphery of naive recipients migrate across the blood-brain tion, such as Triton X and Nonidet P-40 (49). Alternatively, a barrier, recognize autoantigen in the CNS, and exert their ef- different mechanism may underlie the inhibition of CIA and fector function. The delayed onset and reduced severity of EAE EAE by SB-331750. Based on the lack of Ii fragment accumu- exhibited by recipients of SB-331750-treated cells, compared q lation and the presence of mature SDS-stable I-A dimers ex- with recipients of vehicle-treated cells, indicate that in this http://www.jimmunol.org/ hibited by cathepsin S-deficient DBA/1 splenocytes, Nakagawa model, the compound attenuated effector cell migration and/or et al. (7) proposed that the role of cathepsin S in collagen pre- function, and did so through its action in the periphery. How- sentation and induction of CIA may be one of direct or indirect ever, the efficacy achieved following therapeutic administration involvement in processing of Ag, rather than Ii. Consistent with of SB-331750 in actively-induced EAE may be the result of the this hypothesis, cathepsin S has been shown to play an impor- compound’s action in the periphery and/or the CNS. Thus, our tant function in the generation of a subset of antigenic epitopes observations, together with those of other investigators de- derived from myoglobin, SA85 (50), and hen egg lysozyme scribed above, suggest that cathepsins contribute to the etiology (51), and in cell-free assays, was shown to hydrolyze collagen of EAE by modulating Ag presentation in the periphery and/or (52), and initiate the proteolytic processing of MBP resulting in the CNS. by guest on September 30, 2021 the generation of the epitope MBP111–129 (53). Conversely, the In summary, the studies presented in this paper demonstrate that lack of acetylcholine receptor Ag-induced T cell proliferative a pan-active inhibitor of the papain family of cysteine proteases, and cytokine responses by Ag-primed cathepsin S-deficient SB-331750, attenuates late stage Ii processing by DBA/1 and LNC was demonstrated to be due not to defective processing of SJL/J splenocytes, and subsequent Ag-induced T cell proliferation the acetylcholine receptor protein, but rather to Ii-dependent by collagen-primed DBA/1 and MBP-primed SJL/J cells, respec- restriction of peptide loading, suggesting that inhibition of Ii tively. In conjunction with the inhibition of these processes, SB processing accounts for the reduced susceptibility of cathepsin 331750 induces a delay in onset and decrease in severity in the S-deficient mice to myasthenia gravis (36). Thus, the precise corresponding models of experimentally-induced autoimmune mechanism by which cysteine proteases modulate susceptibility disease, CIA in the DBA/1 mouse and EAE in the SJL/J mouse. to Ag-driven autoimmunity likely depends on multiple factors, These observations suggest that inhibition of Ag presentation including the antigenic epitope being processed and presented, and the resulting activation and expansion of Ag-specific T cells the complement of active proteases present, and the haplotype by inhibitors of the papain family of cysteine proteases offers a of the class II molecule to which the antigenic epitope binds. potential therapeutic approach to the treatment of human In conjunction with the inhibition of the clinical symptoms of autoimmunity. CIA, administration of SB-331750 resulted in reduced paw tissue levels of IL-1␤, TNF-␣, and IFN-␥, proinflammatory cytokines Acknowledgments associated with RA and CIA. It is widely accepted that synovial We thank Liming Qu and Thierry Bernard for their SAS programming macrophages are the principal producers of IL-1␤ and TNF-␣, support. while CD4ϩ Th1 cells, a primary T cell subset in the RA joint, produce IFN-␥ that likely promotes activation, result- Disclosures ing in the abundant secretion of IL-1␤ and TNF-␣ (54, 55). Thus, The authors have no financial conflict of interest. the inhibition of Ag (collagen) presentation resulting from SB- ϩ References 331750 administration may reduce the magnitude of the CD4 T ␥ 1. Schwartz, R. H. 1985. T-lymphocyte recognition of antigen in association with cell-derived IFN- response, thereby decreasing macrophage ac- gene products of the major histocompatibility complex. Annu. Rev. Immunol. 3: tivation and the subsequent production of IL-1␤ and TNF-␣. Fur- 237–261. ther inhibition of cytokine production may occur as the result of 2. Bryant, P. W., A. M. Lennon-Dumenil, E. Fiebiger, C. Lagaudriere-Gesbert, and H. L. Ploegh. 2002. and antigen presentation by MHC class II mol- the attenuation of downstream proinflammatory cytokine cascades. ecules. Adv. Immunol. 80: 71–114. In this regard, IL-1␤ has been shown to stimulate the production of 3. Honey, K., and A. Y. Rudensky. 2003. Lysosomal cysteine proteases regulate ␤ ␣ ␥ ␣ antigen presentation. Nat. Rev. Immunol. 3: 472–482. IL-1 , TNF- , and IFN- (56), and TNF- to induce the produc- 4. Watts, C. 2004. The exogenous pathway for antigen presentation on major his- tion of TNF-␣ and IL-1␤ (57). tocompatibility complex class II and CD1 molecules. Nat. Immunol. 5: 685–692. 8002 ATTENUATION OF CIA AND EAE VIA CYSTEINE PROTEASE INHIBITION

5. Denzin, L. K., C. Hammond, and P. Cresswell. 1996. HLA-DM interactions with 30. Bernard, C. C. A., T. G. Johns, A. Slavin, M. Ichikawa, C. Ewing, J. Liu, and intermediates in HLA-DR maturation and a role for HLA-DM in stabilizing J. Bettadapura. 1997. Myelin oligodendrocyte glycoprotein: a novel candidate empty HLA-DR molecules. J. Exp. Med. 184: 2153–2165. autoantigen in multiple sclerosis. J. Mol. Med. 75: 77–88. 6. Riese, R. J., R. N. Mitchell, J. A. Villadangos, G.-P. Shi, J. T. Palmer, 31. Marks, M. S., J. S. Blum, and P. Cresswell. 1990. Invariant chain trimers are E. R. Karp, G. T. De Sanctis, H. L. Ploegh, and H. A. Chapman. 1998. sequestered in the rough endoplasmic reticulum in the absence of association with Cathepsin S activity regulates antigen presentation and immunity. J. Clin. HLA class II antigens. J. Cell Biol. 111: 839–855. Invest. 101: 2351–2363. 32. Thornberry, N. A., H. G. Bull, J. R. Calaycay, K. T. Chapman, A. D. Howard, 7. Nakagawa, T. Y., W. H. Brissette, P. D. Lira, R. J. Griffiths, N. Petrushova, M. J. Kostura, D. K. Miller, S. M. Molineaux, J. R. Weidner, J. Aunins, et al. J. Stock, J. D. McNeish, S. E. Eastman, E. D. Howard, S. R. M. Clarke, et al. 1992. A novel heterodimeric cysteine protease is required for interleukin-1␤ pro- 1999. Impaired invariant chain degradation and antigen presentation and di- cessing in monocytes. Nature 356: 768–774. minished collagen-induced arthritis in cathepsin S null mice. Immunity 10: 33. Cerretti, D. P., C. J. Kozlosky, B. Mosley, N. Nelson, K. Van Ness, 207–217. T. A. Greenstreet, C. J. March, S. R. Kronheim, T. Druck, L. A. Cannizzaro, et 8. Shi, G.-P., J. A. Villadangos, G. Dranoff, C. Small, L. Gu, K. J. Haley, al. 1992. Molecular cloning of the interleukin-1␤ converting enzyme. Science R. Riese, H. L. Ploegh, and H. A. Chapman. 1999. Cathepsin S required for 256: 97–100. normal MHC class II peptide loading and germinal center development. Im- 34. Black, R. A., C. T. Rauch, C. J. Kozlosky, J. J. Peschon, J. L. Slack, munity 10: 197–206. M. F. Wolfson, K. L. Stocking, P. Reddy, S. Srinivasan, N. Nelson, et al. 1997. ␣ 9. Shi, G.-P., R. A. R. Bryant, R. Riese, S. Verhelst, C. Driessen, Z. Li, D. Bromme, A metalloproteinase disintegrin that releases tumour-necrosis factor- from cells. H. L. Ploegh, and H. A. Chapman. 2000. Role for cathepsin F in invariant chain Nature 385: 729–733. processing and major histocompatibility complex class II peptide loading by mac- 35. Coeshott, C., C. Ohnemus, A. Pilyavskaya, S. Ross, M. Wieczorek, H. Kroona, rophages. J. Exp. Med. 191: 1177–1185. A. H. Leimer, and J. Cheronis. 1999. Converting enzyme-independent release of ␣ ␤ 10. Nakagawa, T., W. Roth, P. Wong, A. Nelson, A. Farr, J. Deussing, - and IL-1 from a stimulated human monocytic cell line J. A. Villadangos, H. Ploegh, C. Peters, and A. Y. Rudensky. 1998. Cathepsin L: in the presence of activated neutrophils or purified proteinase 3. Proc. Natl. Acad. critical role in Ii degradation and CD4 T cell selection in the thymus. Science 280: Sci. USA 96: 6261–6266. 450–453. 36. Yang, H., M. Kala, B. G. Scott, E. Goluszko, H. A. Chapman, and P. Christadoss. 2005. Cathepsin S is required for murine autoimmune myasthenia gravis patho- 11. Tolosa, E., W. Li, Y. Yasuda, W. Wienhold, L. K. Denzin, A. Lautwein, Downloaded from genesis. J. Immunol. 174: 1729–1737. C. Driessen, P. Schnorrer, E. Weber, S. Stevanovic, et al. 2003. Cathepsin V is 37. Biroc, S. L., S. Gay, K. Hummel, C. Magill, J. T. Palmer, D. R. Spencer, S. Sa, involved in the degradation of invariant chain in human thymus and is overex- J. L. Klaus, B. A. Michel, D. Rasnick, and R. E. Gay. 2001. Cysteine protease pressed in myasthenia gravis. J. Clin. Invest. 112: 517–526. activity is up-regulated in inflamed ankle joints of rats with adjuvant-induced 12. Bania, J., E. Gatti, H. Lelouard, A. David, F. Cappello, E. Weber, V. Camosseto, arthritis and decreases with in vivo administration of a vinyl sulfone cysteine and P. Pierre. 2003. Human cathepsin S, but not cathepsin L, degrades efficiently protease inhibitor. Arthritis Rheum. 44: 703–711. MHC class II-associated invariant chain in nonprofessional APCs. Proc. Natl. 38. Saegusa, K., N. Ishimaru, K. Yanagi, R. Arakaki, K. Ogawa, I. Saito, Acad. Sci. USA 100: 6664–6669. N. Katunuma, and Y. Hayashi. 2002. Cathepsin S inhibitor prevents autoantigen

13. Beers, C., A. Burich, M. J. Kleijmeer, J. M. Griffith, P. Wong, and presentation and autoimmunity. J. Clin. Invest. 110: 361–369. http://www.jimmunol.org/ A. Y. Rudensky. 2005. Cathepsin S controls MHC class II-mediated antigen 39. Sette, A., S. Southwood, J. Miller, and E. Appella. 1995. Binding of major his- presentation by epithelial cells in vivo. J. Immunol. 174: 1205–1212. tocompatibility complex class II to the invariant chain-derived peptide, CLIP, is 14. Myers, L. K., E. F. Rosloniec, M. A. Cremer, and A. H. Kang. 1997. Collagen- regulated by allelic polymorphism in class II. J. Exp. Med. 181: 677–683. induced arthritis, an animal model of autoimmunity. Life Sci. 61: 1861–1878. 40. Avva, R. R., and P. Cresswell. 1994. In vivo and in vitro formation and disso- 15. Ranges, G. E., S. Sriram, and S. M. Cooper. 1985. Prevention of type II collagen- ciation of HLA-DR complexes with invariant chain-derived peptides. Immunity induced arthritis by in vivo treatment with anti-L3T4. J. Exp. Med. 162: 1: 763–774. 1105–1110. 41. Villadangos, J. A., R. J. Riese, C. Peters, H. A. Chapman, and H. L. Ploegh. 1997. 16. Latham, K. A., K. B. Whittington, R. Zhou, Z. Qian, and E. F. Rosloniec. 2005. Degradation of mouse invariant chain: roles of cathepsins S and D and the in- Ex vivo characterization of the autoimmune T cell response in the HLA-DR1 fluence of major histocompatibility complex polymorphism. J. Exp. Med. 186: mouse model of collagen-induced arthritis reveals long-term activation of type II 549–560. collagen-specific cells and their presence in arthritic joints. J. Immunol. 174: 42. Patil, N. S., A. Pashine, M. P. Belmares, W. Liu, B. Kaneshiro, J. Rabinowitz, 3978–3985. H. McConnell, and E. D. Mellins. 2001. Rheumatoid arthritis (RA)-associated by guest on September 30, 2021 17. Wooley, P. H., H. S. Luthra, J. M. Stuart, and C. S. David. 1981. Type II col- HLA-DR alleles form less stable complexes with class II-associated invariant lagen-induced arthritis in mice: I. Major histocompatibility complex (I region) chain peptide than non-RA-associated HLA-DR alleles. J. Immunol. 167: linkage and antibody correlates. J. Exp. Med. 154: 688–700. 7157–7168. 18. Wooley, P. H., H. S. Luthra, M. M. Griffiths, J. M. Stuart, A. Huse, and 43. Buckner, J., W. W. Kwok, B. Nepon, and G. T. Nepom. 1996. Modulation of C. S. David. 1985. Type II collagen-induced arthritis in mice: IV. Variations in HLA-DQ binding properties by differences in class II dimer stability and pH- immunogenetic regulation provide evidence for multiple arthritogenic epitopes dependent peptide interactions. J. Immunol. 157: 4940–4945. on the collagen molecule. J. Immunol. 135: 2443–2451. 44. Hausmann, D. H. F., B. Yu, S. Hausmann, and K. W. Wucherpfennig. 1999. 19. Hafler, D. A., and H. L. Weiner. 1995. Immunologic mechanisms and therapy in pH-dependent peptide binding properties of the type 1 diabetes-associated I-Ag7 multiple sclerosis. Immunol. Rev. 144: 75–107. molecule: rapid release of CLIP at an endosomal pH. J. Exp. Med. 189: 20. Swanborg, R. H. 1995. Experimental autoimmune encephalomyelitis in rodents 1723–1733. as a model for human demyelinating disease. Clin. Immunol. Immunopathol. 77: 45. Vartdal, F., B. H. Johansen, T. Friede, C. J. Thorpe, S. Stevanovic, J. E. Eriksen, 4–13. K. Sletten, E. Thorsby, H.-G. Rammensee, and L. M. Sollid. 1996. The peptide ␣ ␤ 21. Kuchroo, V. K., and H. L. Weiner. 1998. Antigen-driven regulation of experi- binding motif of the disease associated HLA-DQ ( 1*0501, 1*0201) molecule. mental autoimmune encephalomyelitis. Res. Immunol. 149: 759–771. Eur. J. Immunol. 26: 2764–2772. 46. Honey, K., K. Forbush, P. E. Jensen, and A. Y. Rudensky. 2004. Effect of de- 22. Fritz, R. B., M. J. Skeen, C.-H. Jen Chou, M. Garcia, and I. K. Egorov. 1985. creasing the affinity of the class II-associated invariant chain peptide on the MHC Major histocompatibility complex-linked control of the murine immune response class II peptide repertoire in the presence or absence of H-2M. J. Immunol. 172: to myelin basic protein. J. Immunol. 134: 2328–2332. 4142–4150. 23. Lamont, A. G., A. Sette, R. Fujinami, S. M. Colon, C. Miles, and H. M. Grey. 47. Neefjes, J. J., and H. L. Ploegh. 1992. Inhibition of endosomal proteolytic activity 1990. Inhibition of experimental autoimmune encephalomyelitis induction in by leupeptin blocks surface expression of MHC class II molecules and their SJL/J mice by using a peptide with high affinity for IAs molecules. J. Immunol. conversion to SDS resistant ␣␤heterodimers in endosomes. EMBO J. 11: 145: 1687–1693. 411–416. 24. Marquis, R. W., Y. Ru, S. M. LoCastro, J. Zeng, D. S. Yamashita, H.-J. Oh, 48. Loss, G. E., and A. J. Sant. 1993. Invariant chain retains MHC class II molecules K. F. Erhard, L. D. Davis, T. A. Tomaszek, D. Tew, et al. 2001. Azepanone-based in the endocytic pathway. J. Immunol. 150: 3187–3197. inhibitors of human and rat cathepsin K. J. Med. Chem. 44: 1380–1395. 49. Castellino, F., R. Han, and R. N. Germain. 2001. The transmembrane segment of 25. Palmer, J. T., D. Rasnick, J. L. Klaus, and D. Bromme. 1995. Vinyl sulfones as invariant chain mediates binding to MHC class II molecules in a CLIP-indepen- mechanism-based cysteine protease inhibitors. J. Med. Chem. 38: 3193–3196. dent manner. Eur. J. Immunol. 31: 841–850. 26. McGrath, M. E., P. A. Sprengeler, C. M. Hill, V. Martichonok, H. Cheung, 50. Hsieh, C.-S., P. deRoos, K. Honey, C. Beers, and A. Y. Rudensky. 2002. A role J. R. Somoza, J. T. Palmer, and J. W. Janc. 2003. Peptide ketobenzoxazole in- for cathepsin L and cathepsin S in peptide generation of MHC class II presen- hibitors bound to cathepsin K. Biochemistry 42: 15018–15028. tation. J. Immunol. 168: 2618–2625. 27. Gowen, M., F. Lazner, R. Dodds, R. Kapadia, J. Feild, M. Tavaria, I. Bertoncello, 51. Plu¨ger, E. B. E., M. Boes, C. Alfonso, C. J. Schro¨ter, H. Kalbacher, H. L. Ploegh, F. Drake, S. Zavarselk, I. Tellis, et al. 1999. Cathepsin K knockout mice develop and C. Driessen. 2002. Specific role for cathepsin S in the generation of antigenic osteopetrosis due to a deficit in matrix degradation but not demineralization. peptides in vivo. Eur. J. Immunol. 32: 467–476. J. Bone Miner. Res. 14: 1654–1663. 52. Kirschke, H., B. Wiederanders, D. Bro¨mme, and A. Rinne. 1989. Cathepsin S 28. Bro¨mme, D., Z. Li, M. Barnes, and E. Mehler. 1999. Human cathepsin V func- from bovine spleen. Biochem. J. 264: 467–473. tional expression, tissue distribution, electrostatic surface potential, enzymatic 53. Beck, H., G. Schwartz, C. J. Schro¨ter, M. Deeg, D. Baier, S. Stevanovic, characterization, and chromosomal localization. Biochemistry 38: 2377–2385. E. Weber, C. Driessen, and H. Kalbacher. 2001. Cathepsin S and an - 29. Pettinelli, C. B., and D. E. McFarlin. 1981. Adoptive transfer of experimental specific endoprotease dominate the proteolytic processing of human myelin basic allergic encephalomyelitis in SJL/J mice after in vitro activation of lymph node protein in vitro. Eur. J. Immunol. 31: 3726–3736. cells by myelin basic protein: requirement for Lyt1ϩ2- T lymphocytes. J. Immu- 54. Feldmann, M., F. M. Brennan, and R. N. Maini. 1996. Role of cytokines in nol. 127: 1420–1423. rheumatoid arthritis. Annu. Rev. Immunol. 14: 397–440. The Journal of Immunology 8003

55. Ma, Y., and R. M. Pope. 2005. The role of macrophages in rheumatoid arthritis. 61. Gresser, O., E. Weber, A. Hellwig, S. Riese, and A. Re´gnier-Vigouroux. 2001. Curr. Pharm. Des. 11: 569–580. Immunocompetent astrocytes and microglia display major differences in the pro- 56. Dinarello, C. A. 1996. Biological basis for interleukin-1 in disease. Blood 87: cessing of the invariant chain and in the expression of active cathepsin L and 2095–2147. cathepsin S. Eur. J. Immunol. 31: 1813–1824. 57. Aggarwal, B. B., A. Samanta, and M. Feldmann. 2001. TNF␣.InCytokine Ref- 62. Matsumoto, Y., K. Ohmori, and M. Fujiwara. 1992. Immune regulation by brain erence. J. J. Oppenheim and M. Feldmann, eds. Academic Press, San Diego, pp. cells in the central nervous system: microglia but not astrocytes present myelin 413–434. basic protein to encephalitogenic T cells under in vivo-mimicking conditions. 58. Slavin, A. J., J. M. Soos, O. Stuve, J. C. Patarroyo, H. L. Weiner, A. Fontana, Immunology 76: 209–216. E. K. Bikoff, and S. S. Zamvil. 2001. Requirement for endocytic antigen pro- 63. Williams, K. C., N. P. Dooley, E. Ulvestad, A. Waage, M. Blain, V. W. Yong, cessing and influence of invariant chain and H-2M deficiencies in CNS autoim- and J. P. Antel. 1995. Antigen presentation by human fetal astrocytes with the munity. J. Clin. Invest. 108: 1133–1139. cooperative effect of microglia or the microglial-derived cytokine IL-1. J. Neu- 59. Tompkins, S. M., J. Padilla, M. C. Dal Canto, J. P. Y. Ting, L. Van Kaer, and rosci. 15: 1869–1878. S. D. Miller. 2002. De novo central nervous system processing of myelin antigen 64. Nikcevich, K. M., K. B. Gordon, L. Tan, S. D. Hurst, J. Kroepfl, M. Gardinier, is required for the initiation of experimental autoimmune encephalomyelitis. T. A. Barrett, and S. D. Miller. 1997. IFN-␥-activated primary murine astrocytes J. Immunol. 168: 4173–4183. express B7 costimulatory molecules and prime naive antigen-specific T cells. 60. Santambrogio, L., S. L. Belyanskaya, F. R. Fischer, B. Cipriani, C. F. Brosnan, J. Immunol. 158: 614–621. P. Ricciardi-Castagnoli, L. J. Stern, J. L. Strominger, and R. Riese. 2001. De- 65. Aloisi, F., F. Ria, G. Penna, and L. Adorini. 1998. Microglia are more efficient velopmental plasticity of CNS microglia. Proc. Natl. Acad. Sci. USA 98: than astrocytes in antigen processing and in Th1 but not Th2 cell activation. 6295–6300. J. Immunol. 160: 4671–4680. Downloaded from http://www.jimmunol.org/ by guest on September 30, 2021