Arthritis and Collagen-Induced Arthritis Receptor in Myeloid Cells

Published April 19, 2013, doi:10.4049/jimmunol.1201675 The Journal of Immunology

The Novel Role of IL-7 Ligation to IL-7 Receptor in Myeloid Cells of Rheumatoid Arthritis and Collagen-Induced Arthritis

Zhenlong Chen,*,1 Seung-jae Kim,*,1 Nathan D. Chamberlain,* Sarah R. Pickens,* Michael V. Volin,† Suncica Volkov,* Shiva Arami,* John W. Christman,‡ Bellur S. Prabhakar,x William Swedler,* Anjali Mehta,* Nadera Sweiss,* and Shiva Shahrara*

Although the role of IL-7 and IL-7R has been implicated in the pathogenesis of rheumatoid arthritis (RA), the majority of the studies have focused on the effect of IL-7/IL-7R in T cell development and function. Our novel data, however, document that patients with RA and greater disease activity have higher levels of IL-7, IL-7R, and TNF-a in RA monocytes, suggesting a feedback regulation between IL-7/IL-7R and TNF-a cascades in myeloid cells that is linked to chronic disease progression. Investigations into the involved mechanism showed that IL-7 is a novel and potent chemoattractant that attracts IL-7R+ monocytes through activation of the PI3K/AKT1 and ERK pathways at similar concentrations of IL-7 detected in RA synovial fluid. To determine whether ligation of IL-7 to IL-7R is a potential target for RA treatment and to identify their mechanism of action, collagen-induced arthritis (CIA) was therapeutically treated with anti–IL-7 Ab or IgG control. Anti–IL-7 Ab treatment significantly reduces CIA monocyte recruitment and osteoclast differentiation as well as potent joint monocyte chemoattractants and bone erosion markers, suggesting that both direct and indirect pathways might contribute to the observed effect. We also demonstrate that reduction in joint MIP-2 levels is responsible for suppressed vascularization detected in mice treated with anti–IL-7 Ab compared with the control group. To our knowledge, we show for the first time that expression of IL-7/IL-7R in myeloid cells is strongly correlated with RA disease activity and that ligation of IL-7 to IL-7R contributes to monocyte homing, differentiation of osteoclasts, and vascularization in the CIA effector phase. The Journal of Immunology, 2013, 190: 000–000.

heumatoid arthritis (RA) is a chronic autoimmune dis- Consistent with our findings, RA macrophages were determined order in which the numbers of monocyte-derived mac- to be the main source of IL-7 production, because the expression R rophages are greater than normal joints; it is well correlated of IL-7 in the lining and sublining closely correlated with the with radiologic damage, joint pain, and inflammation (1, 2). IL-7 number of CD68+ cells (6). However, others have shown that IL- is a member of the IL-2/IL-15 family of cytokines that signals 7R is expressed on T and B cells in addition to macrophages in RA through IL-7R ligation (3, 4). We have shown recently that IL-7 synovium (7). and IL-7R are coexpressed in RA synovial tissue (ST) lining and The role of IL-7 and IL-7R has been implicated in several au- sublining macrophages as well as sublining endothelial cells (5). toimmune diseases including multiple sclerosis, psoriasis, Sjo¨gren syndrome, juvenile idiopathic arthritis, and RA (8, 9). Interest- ingly, most of the previous studies have focused on determining *Division of Rheumatology, Department of Medicine, University of Illinois at Chi- the role of IL-7/IL-7R in T cell function because it has been cago, Chicago, IL 60612; †Department of Microbiology and Immunology, Midwest- demonstrated that IL-7 is responsible for maintaining T cell ho- ern University, Chicago College of Osteopathic Medicine, Downers Grove, IL 60515; meostasis by expanding TH-1 and TH-17 cells through inhibition ‡Division of Pulmonary/Critical Care, Department of Medicine, University of Illinois at Chicago, Chicago, IL 60612; and xDepartment of Microbiology and Immunology, of T cell apoptosis via upregulation of Bcl-2 (10, 11). IL-7R li- College of Medicine, University of Illinois at Chicago, Chicago, IL 60612 gation by IL-7 can also contribute to T cell proliferation, positive 1Z.C. and S.-j.K. contributed equally to this work. and negative selection, activation, and cytokine production (12). Received for publication June 18, 2012. Accepted for publication March 19, 2013. However, IL-7–activated T cells were unable to secrete TNF-a This work was supported in part by awards from the National Institutes of Health and required cell to cell contact with monocytes for this function (AR056099), the Arthritis National Research Foundation, a grant from Within Our (6, 13). Conversely, when human peripheral blood monocytes Reach from the American College of Rheumatology, the U.S. Department of De- fense (PR093477), an Arthritis Foundation Innovative Research Grant (to S.S.), were stimulated with IL-7, significant levels of proinflammatory National Institutes of Health Grants NIH R01 HL 075557, RO1 HL 103643, and cytokines such as IL-6, IL-8 TNF-a, IL-1a, IL-1b (14, 15) were T32 HL 082547, and the U.S. Department of Veterans Affairs Merit Review Grant produced, suggesting that IL-7R ligation to IL-7 could also have 5I01BX000108 (to J.W.C.). an important role in myeloid cell function. Furthermore, recent Address correspondence and reprint requests to Dr. Shiva Shahrara, Division of Rheumatology, Department of Medicine, University of Illinois at Chicago, Medical data show that TNF-a is the common factor that modulates ex- Sciences Building, 835 South Wolcott Avenue, E807-E809, Chicago, IL 60612. pression of IL-7 and IL-7R in the synovial lining (RA macro- E-mail address: [email protected] phages and fibroblasts) and the endothelial cells, suggesting a cross- Abbreviations used in this article: CIA, collagen-induced arthritis; Ctl, control; DAS28, regulation between these two cascades (5). disease activity score based on 28 defined joints; DMARD, disease-modifying antirheumatic drug; DN, dominant negative; NL, normal; OA, osteoarthritis, PB, peripheral Among a panel of 16 factors, IL-7 was the most potent inducer + blood; RA, rheumatoid arthritis; RANKL, receptor activator for NF-kB ligand; SF, in differentiating CD14 RA synovial fluid macrophages to mul- synovial fluid; ST, synovial tissue; TRAP, tartrate-resistant acid phosphatase; VWF, tinucleated osteoclasts (16, 17). Although IL-7–mediated bone von Willebrand factor. erosion has also been demonstrated to be due to T cell production Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 of receptor activator for NF-kB ligand (RANKL) (18, 19), other

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1201675 2 NOVEL ROLE OF IL-7 LIGATION IN RA studies suggest that IL-7/IL-7R–mediated osteoclastogenesis in To show that RA synovial fluid–mediated monocyte chemotaxis is in part RA could extend beyond their role in T cells and have other due to IL-7 function, 12 synovial fluids were diluted (1:20) and neutralized critical implications in myeloid cells (16, 17). with anti–IL-7 Ab (10 mg/ml; R&D Systems) or control IgG. To demonstrate that RA synovial fluid monocyte trafficking is mediated through IL-7 ligation Based on the significant elevation of IL-7 and IL-7R in RA ST to myeloid IL-7R, cells were incubated with Ab to IL-7R (10 mg/ml; R&D and fluid macrophages (5), the ability of IL-7 to induce potent Systems) or IgG control for 1 h prior to performing monocyte chemotaxis in proinflammatory cytokines from myeloid cells (15), and the role response to 6 RA synovial fluids (1:20 dilution) for 2 h. of IL-7 in modulating differentiation of RA synovial fluid myeloid To validate that IL-7–mediated monocyte chemotaxis is promoted through activation of the AKT pathway, NL monocytes were transfected cells to mature osteoclasts (16, 17), we examined the significance with control (Ctl) or dominant negative (DN)-AKT plasmid (a gift from of IL-7 ligation to IL-7R on myeloid cells. We also investigated Dr. B.S. Prabhakar’s laboratory) (23) at 2.5 mg for 48 h. Cells were either whether expression of IL-7/IL-7R in RA blood myeloid cells is untreated or stimulated with 100 ng/ml of IL-7 for 30 and 60 min prior to linked to TNF-a and disease activity levels. Western blotting. After demonstrating that DN-AKT significantly sup- In this study, we found that in RA blood monocytes, concen- presses IL-7–mediated AKT1 phosphorylation, chemotaxis of NL monocyte transfected with Ctl or DN-AKT was examined in response to 10 ng/ trations of IL-7, IL-7R, and TNF-a are closely correlated with each ml IL-7. other and disease activity, suggesting that activation of IL-7/IL-7R To confirm that ERK activation contributes to IL-7–mediated myeloid cascade plays a crucial role in myeloid cell–mediated RA patho- cell infiltration, THP-1 cells (American Type Culture Collection, Manassas, genesis. The pathogenic role of IL-7R ligation was demonstrated VA) were transfected with 100 nM scrambled or ERK siRNA (Dharmacon, Thermo Scientific, Waltham, MA) for 48 h according to the manufacturer’s by showing that IL-7 is a novel and potent monocyte chemo- instructions. Thereafter, transfected THP-1 cells were probed for ERK and + + – attractant that attracts IL-7R blood CD14 CD16 monocytes via actin. Next, chemotaxis of control or ERK knockdown THP-1 cells was PI3K/AKT1 and ERK pathways. In collagen-induced arthritis examined in response to 10 ng/ml IL-7. (CIA), a chronic model of RA, we demonstrate that blockade of RA patient population endogenous IL-7 relieves arthritis by markedly reducing joint monocyte homing, bone erosion, and vascularization either di- RA specimens were obtained from patients with RA, diagnosed according rectly or indirectly through decreasing potent monocyte chemo- to the 1987 revised criteria of the American College of Rheumatology (24). PB was obtained from 76 patients, 71 women, and 5 men (mean age, 48.2 6 attractants (TNF-a and CCL2/MCP-1), markers of bone erosion 15.3 y). At the time of evaluation, patients received no treatment (n =7,all (RANKL and Cathepsin K), and proangiogenic factor (MIP-2/ women; mean age, 53.1 6 19.5 y); treatment with nonbiological disease- CCL8). To our knowledge, these results suggest for the first modifying antirheumatic drugs (DMARDs; methotrexate, leflunomide, time that ligation of IL-7 to IL-7R can affect myeloid cell mi- sulfasalazine azathioprine, hydroxychloroquine, or minocycline) alone (n = 29; 26 women and 3 men; mean age, 52.0 6 16.0 y; of which two gration and function, as well as have a positive feedback regula- received hydroxychloroquine only [both women, 62 y old]); treatment with tion with TNF-a production from these cells, leading to it being DMARDs plus prednisone (n = 7, all women; mean age, 54.7 6 12.7 y), a potential therapeutic target in RA. DMARDs plus rituximab (n = 1 woman, age 52 y); TNF-a inhibitor alone (n =7;6women,1man;meanage,38.36 8.9 y), with prednisone (n =1 woman, age 48 y), with a DMARD (n = 17, all women; mean age, 45.6 6 Materials and Methods 12.6 y), with a DMARD plus prednisone (n =5,4women,and1man, Monocyte chemotaxis average age 37.2 6 18.1; or anti–T cell therapy (Abatacept) with a DMARD (n = 1 woman, age 41 y) or with DMARD and prednisone (n = 1 women, age Experiments were performed to determine the direct effect of IL-7 on 28 y). monocyte chemotaxis. Mononuclear cells were isolated by Histopaque These studies were approved by the University of Illinois at Chicago (Sigma-Aldrich) gradient centrifugation. Subsequently, monocytes were Institutional Ethics Review Board, and all donors gave informed written isolated from normal (NL) or RA peripheral blood (PB) using negative consent. The number of patients was 76, but refer to the figure legends for selection kit for CD14+CD16+CD16- (StemCell Technology, Vancouver, the exact number of patients in each experiment. Canada; cat #19058; used for all monocyte chemotaxis experiments) and + – CD14 CD16 (StemCell Technology; cat #19019 used only in Fig. 2C) Cytokine quantification according to the manufacturer’s instructions (5, 20). Chemotaxis was performed in a Boyden chamber (Neuroprobe; Gaithersburg, MD) using HumanIL-7(R&DSystems)wasquantifiedbyELISAinSTsfrom NL monocytes for 2 h with IL-7 (R&D Systems, Minneapolis, MN) patients with RA or osteoarthritis (OA) or from NL samples, as well as concentrations varying from 0.0001 to 100 ng/ml; fMLF (1 mM; Sigma synovial fluids (SFs) from patients with RA or OA according to the manu- Aldrich) was used as positive control, and PBS was used as negative con- facturers’ instructions. Mouse IL-6, IL-1b,IL-17,TNF-a, CCL2/MCP-1, trol (21, 22). CCL5/RANTES, CXCL1, Ang-1, bFGF, VEGF, and MIP-2 ELISA Kit To determine whether blockade of IL-7R on monocytes would inhibit IL- (R&D Systems) was used according to the manufacturers’ instructions, and 7–induced monocyte chemotaxis, monocytes were blocked with anti–IL- the expression level of each factor was normalized to the ankle protein 7R Ab (10 mg/ml; R&D Systems), or isotype control for 1 h prior to concentration. performing the monocyte chemotaxis in response to 10 and 50 ng/ml IL-7 for 2 h. IL-7 signaling in human peripheral blood monocytes To determine the contribution of monocyte subtypes (CD16+ versus Monocytes were untreated or treated with IL-7 (10 ng/ml) for 15–65 min. CD16–) to IL-7 (1 and 50 ng/ml for 2 h)–mediated myeloid cell infiltration, Cell lysates were examined by Western blot analysis as described previ- in vitro chemotaxis was compared between CD14+CD16+CD16– and ously (22, 25). Blots were probed with p-ERK, p-AKT1, p-STAT1, p- CD14+CD16– cells. STAT3, or p-STAT5 (Cell Signaling; 1:1000 dilution) overnight and after To determine signaling pathways associated with IL-7–induced mono- stripping, were probed with ERK, AKT, STAT3, or actin (Cell Signaling or cyte chemotaxis, monocytes were treated with 1 and 5 mM inhibitors to Sigma-Aldrich; 1:3000 dilution). ERK (U0126) and PI3K/AKT (LY294002) and STAT3 (WP1066) or 10 and 50 mM for STAT5 (573108 STAT5 inhibitor; EMD Millipore; Bill- Study protocol for CIA and anti–IL-7 treatment erica, MA) for 1 h. Subsequently, monocyte chemotaxis was performed in response to 10 ng/ml of IL-7 for 2 h. The animal studies were approved by the Institutional Animal Care and Use To demonstrate that inhibition of PI3K/AKT1 is specific to monocyte Committee. DBA/1J mice (7–8 wk old; Jackson Laboratories, Bar Harbor, migration mediated by IL-7, extravasation of monocytes pretreated with ME) were immunized with collagen on days 0 and 21. Bovine collagen DMSO or PI3K inhibitor (LY294002; 5 mM) was examined in response to type II (2 mg/ml; Chondrex, Redmond, WA) was emulsified in equal potent monocyte chemoattractant such as fMLF (1 mM) as well as CCL2 volumes of CFA (2 mg/ml; Chondrex). The DBA/1J mice were immunized (0.9 nM; R&D Systems), CCL5 (1.01 nM; R&D Systems), IL-17 (0.667 s.c. in the tail with 100 ml emulsion. On day 21, mice were injected in- nM; R&D Systems) or IL-7 (0.58 nM). To validate that activation of p38 tradermally with 100 ml collagen type II (2 mg/ml) emulsified in equal MAPK promotes CCL2, CCL5, and IL-17 but not IL-7–induced myeloid volumes of IFA (25). CIA mice were treated with IgG or anti–IL-7 (100 cell infiltration, an inhibitor to p38 (SB203580; 5 mM) was included in mg/injection; R&D Systems) Ab i.p. on days 26, 29, 33, 36, 40, and 42. these experiments. Changes in ankle circumference (in millimeters) were recorded (n = 10–11 The Journal of Immunology 3 mice). Mice were sacrificed on day 43, and ankles were harvested for proteinase K digestion buffer (0.05 M/l; Dako) for 5 min at room tem- protein and mRNA extraction or histologic studies. Serum was saved for perature. Endogenous peroxidase activity was blocked by incubation with laboratory tests. In a separate study, CIA ankle joints were harvested on 3% H2O2 for 5 min. Nonspecific binding of avidin and biotin was blocked day 45 and were compared with control mice that were not induced by CIA using an avidin/biotin blocking kit (Dako). Nonspecific binding of Abs for IL-7R immunostaining and quantifying joint IL-7 by ELISA. to the tissues was blocked by pretreatment of tissues with Protein Block (Dako). Sham injection or CIA tissues were immunostained with IL-7R Clinical assessments (1:100 dilution; Santa Cruz Biotechnology). Additional CIA mice treated Ankle circumferences were determined by measuring two perpendicular with IgG or anti–IL-7 were stained with F480 (1:100 dilution; Serotec), von diameters, including the laterolateral diameter and the anteroposterior di- Willebrand factor (1:1000 dilution; Dako), CD3 (1:100 dilution; Abcam), or control IgG Ab (Beckman Coulter). Positive immunostaining was scored on ameter, as measured with a caliper (Lange Caliper; Cambridge Scientific a 0–5 scale; score data were pooled, and the mean 6 SEM was calculated in Industries). Circumference was determined using the following for- rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi! each data group. Each slide was evaluated by two masked observers. a2 þ b2 mula: C ¼ 2p 3 where C is the circumference and a and To exclude the chance of ambiguous cells migrating in response to IL-7, 2 cells attracted to chemotaxis polycarbonate membrane were stained with b represent the diameters (25). Ankle circumference evaluations were anti-CD68 Ab. For this purpose, membranes harvested from NL monocytes performed on days 3, 7, 10, 14, 17, 20, 23, 26, 28, 30, 33, 35, 37, 40, 41, migrated in response to PBS, fMLF (1 mM) and IL-7 (10 ng/ml) were fixed, and 42. stained with mouse anti-human CD68 (1:100 dilution; Dako), and visualized with goat anti-mouse Ab labeled with Alexa 594 (1:500 dilution; Abs and immunohistochemistry Molecular Probes, Eugene, OR). The number of cells migrated was determined in 15 high-powered fields based on double-positive DAPI (blue The ankles of mice were fixed in formalin, decalcified, embedded in nuclear staining) and CD68 (red) staining. paraffin, and sectioned. Inflammation, synovial lining, and bone erosion (based on a score of 0–5) were determined using H&E-stained sections by Quantification of hemoglobin in the ankles of mice two masked observers. Mouse ankles were stained with immunoperoxidase using Vector Elite ABC Kits (Vector Laboratories), with diaminobenzidine Using methemoglobin (Sigma-Aldrich), serial dilutions were prepared to (Vector Laboratories) as a chromogen. Slides were deparaffinized in xylene generate a standard curve from 70 to 1.1 g/dl (25–28). Fifty microliters of for 15 min at room temperature, followed by rehydration by transfer the ankles of mice or standard were added to a 96-well plate in duplicate, through graded alcohols. Ags were unmasked by incubating slides in and 50 ml tetramethylbenzidine (Sigma-Aldrich) was added to each sam-

FIGURE 1. IL-7 and IL-7R expression correlates with DAS28 and TNF-a levels in RA monocytes, and ligation of RA synovial fluid IL-7 to myeloid IL- 7R promotes chemotaxis. Linear regression analysis was used to compare expression of TNF-a with IL-7 (n = 76) (A), IL-7R (n = 76) (B), mRNA levels, and DAS28 (n = 76) (C) in RA monocytes. Correlation was also calculated for DAS28 score (n = 76) and expression levels of IL-7 (D) and IL-7R (E)inRA monocytes. The mRNA expression levels in RA monocytes are shown as fold increase above NL PB monocytes and are normalized to GAPDH. (F) Concentration of IL-7 was quantified by ELISA in RA, OA, and NL STs (n = 10) as well as in (G), which shows SFs from RA and OA patients (n = 18). (H) Twelve RA SFs were preincubated with anti–IL-7 Ab (10 mg/ml) or control IgG for 1 h prior to performing monocyte chemotaxis in response to the RA SFs. (I) Monocytes were incubated with Ab to IL-7R (10 mg/ml) and IgG control for 1 h prior to performing monocyte chemotaxis in response to six RA SFs. Values demonstrate mean 6 SE. *p , 0.05. 4 NOVEL ROLE OF IL-7 LIGATION IN RA ple. The plate was allowed to develop at room temperature for 15–20 min Tartrate-resistant acid phosphatase staining with gentle shaking, and the reaction was terminated with 50 ml2N Tartrate-resistant acid phosphatase (TRAP) staining was performed using an H2SO4 for 3–5 min. Absorbance was read with an ELISA plate reader at Acid Phosphatase Leukocyte Kit (Sigma-Aldrich) in paraffin-embedded 450 nm. To calculate hemoglobin concentrations in the mice ankles, the + values (grams per deciliter) were normalized to the weights of the ankles mice ankles according to manufacturer’s instructions. Next, TRAP cells (milligrams per milliliter). stained in CIA mouse ankles were scored on a scale of 0–5: 0 = no staining, 1 = few or rare positive cells, 2 = scattered staining, 3 = multiple foci of Flow cytometry analysis of CD3, CD4, TH-1, and TH-17 cells positive cells, 4 = clusters of positive cells, and 5 = diffuse staining (29). in CIA splenocytes Statistical analysis 1 1 To quantify the percent CD3 and CD4 cells in CIA splenocytes, cells The data were analyzed using one-way ANOVA followed by a post hoc two- were washed, blocked using anti-mouse CD16/CD32, and stained with tailed Student t test for paired and unpaired samples using GraphPad and allophycocyanin-conjugated anti-CD3 or anti-CD4 Abs (eBioscience, San Microsoft Excel. In RA monocytes, expression levels of IL-7, IL-7R, and Diego, CA). To determine whether the percentage of TH-1 or TH-17 cells TNF-a were correlated with each other and disease activity score based on were effected by anti-IL-7 Ab treatment in CIA mice, splenocytes were 28 defined joints (DAS28) score using linear regression analysis in Excel. stimulated with PMA (5 ng/ml) and ionomycin (500 ng/ml; Sigma- The p values , 0.05 were considered significant. Aldrich) in the presence of Brefeldin A (3 mg/ml; Sigma-Aldrich) for 4 h. Cells were then washed, blocked using anti-mouse CD16/CD32, and Results stained with APC-conjugated anti-CD4 Abs (eBioscience, San Diego, CA). Following CD4 staining, cells were blocked, fixed, and permeabilized In RA monocytes, IL-7 and IL-7R correlate with DAS28 score using an IC-Flow Kit according to the manufacturer’s instructions and TNF-a levels (Imgenex, San Diego, CA). The percentages of TH-1– and TH-17–pro- Because expression of IL-7 and IL-7R is elevated in RA compared ducing cells were determined by staining the splenocytes with FITC- conjugated anti–IFN-g or PE-labeled anti–IL-17 Abs (eBioscience), and with NL monocytes (5), and because IL-7–activated monocytes the data were analyzed by flow cytometry (Beckman Coulter Cyan ADP). can produce TNF-a (15), we asked whether there is a relationship between IL-7/IL-7R cascade with TNF-a and DAS28 score in RA Real-time RT-PCR monocytes. We found that the levels of IL-7 (R2 = 0.51) and IL-7R 2 Total cellular RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA) (R = 0.85) are closely correlated with TNF-a in RA monocytes from the ankle homogenates. Subsequently, reverse transcription and real- (Fig. 1A, 1B). Furthermore, patients with greater levels of DAS28 time RT-PCR were performed to determine the expression of IL-7, IL-7R, had increased expression of IL-7 (R2 = 0.55), IL-7R (R2 = 0.56; and TNF-a levels in RA monocytes as well as RANKL and cathepsin K Fig. 1D, 1E), and TNF-a (R2 = 0.57; Fig. 1C) in RA monocytes. expression levels in CIA ankle joints as described previously (21, 22, 25). We noticed that there were two groups of RA patients, in regard to Relative gene expression was determined using the ΔΔCt method, and results were expressed as fold increase above conditions indicated in the myeloid IL-7 expression, that included groups below and above figure legends. the 50-fold range. Therefore, in addition to linear regression

FIGURE 2. IL-7 induces monocyte migration through ligation to IL-7R and activation of AKT1/PI3K and ERK pathways. (A) IL-7 monocyte chemotaxis was performed in a Boyden chemotaxis chamber with varying concentrations of IL-7 (n = 3). (B) Monocytes were incubated with anti–IL-7R Ab (10 mg/ml) or control IgG for 1 h; thereafter, chemotaxis was performed in response to IL-7 (10 and 50 ng/ml; n = 3). (C) Chemotaxis was performed using monocytes negatively selected for CD14+CD16+/2 and CD14+CD162 in response to IL-7 (1 and 50 ng/ml; n = 3). (D) To determine the mechanism of IL-7 in monocytes, cells were stimulated with IL-7 (100 ng/ml) for 0–65 min, and the cell lysates were probed for p-ERK, p-AKT1 and p-STAT1, and p-STAT3 and STAT5 (n = 3). (E) To examine signaling pathways associated with IL-7 monocyte migration, cells were preincubated with 1 and 5 mM of the identified chemical inhibitors for ERK (U0126), PI3K (LY294002; LY), and STAT3 (WP1066; WP) or 10 and 50 mM of STAT5 inhibitor (573108 STAT5 inhibitor; S5i) prior to performing chemotaxis in the Boyden chamber (2 h; n = 3). (F) Monocytes were treated with DMSO or inhibitors for PI3K/AKT (LY294002; 5 mM) and p38 (SB203580; 5 mM) to examine the impact of these pathways on fMLF (1 mM) as well as of CCL2 (0.9 nM), CCL5 (1.01 nM), IL-17 (0.667 nM), or IL-7 (0.58 nM)–induced monocyte migration (n = 3–5). Values demonstrate mean 6 SE.*p , 0.05. The Journal of Immunology 5 analysis, the data were reevaluated using nonlinear curve analysis. Ligation of IL-7 to IL-7R contributes to monocyte trafficking Interestingly the nonlinear regression analyses for TNF and IL-7 through activation of PI3K/AKT1 and ERK pathways 2 3 216 2 (R = 0.60; p = 1.53 10 ), DAS28 and IL-7 (R = 0.60; p = IL-7 contributes to myeloid cell trafficking into the RA joints; 3 216 2 3 216 1.61 10 ) and DAS28 and IL-7R (R = 0.61; p =1.01 10 ) therefore, the involved mechanism was examined next. IL-7 was were still able to demonstrate a strong correlation between these chemotactic for monocytes at concentration as low as 0.1 ng/ml factors. These results suggest that elevated levels of IL-7/IL-7R can (n = 3; Fig. 2A). The mean concentration of IL-7 in the 18 RA predict higher RA disease activity and that there is a positive feed- synovial fluids analyzed was 138 6 19 pg/ml (up to 414 pg/ml; Fig. back regulation between IL-7/IL-7R and TNF-a pathways by pro- 1G), a value that was highly chemotactic (Fig. 2A). Furthermore, ducing and responding to TNF-a. blockade of IL-7R on monocytes suppressed IL-7–mediated chemotaxis (Fig. 2B), suggesting that monocyte recruitment occurs IL-7 and IL-7R play an important role in RA synovial through direct ligation of IL-7 to IL-7R on these cells. To further fluid–mediated monocyte trafficking demonstrate which subtypes of monocyte participate in IL-7–me- Because RA ST and fluid expressed significantly higher levels of diated myeloid cell recruitment, chemotaxis of CD14+CD16+/2 IL-7 compared with OA or NL ST and fluid (Fig. 1F, 1G), ex- monocytes was compared with CD14+CD162 cells in response periments were performed to determine whether the IL-7 identified to IL-7. We show that the mean number of CD14+CD16+/2 and in RA synovial fluid was chemotactic for monocytes. Neutralization CD14+CD162 cells migrating in response to IL-7 was similar, of IL-7 significantly reduced (40%; p , 0.05) monocyte chemotaxis suggesting that CD14+CD16+ does not have an imperative role in compared with control IgG-treated RA synovial fluids (Fig. 1H). In this process, whereas CD14+CD162 cells are the main responders addition, blockade of IL-7R on monocytes was effective in sup- to IL-7–mediated cell migration (Fig. 2C). Experiments were pressing RA synovial fluid–mediated monocyte migration (Fig. 1I). performed to determine signaling pathways involved in IL-7–me- These results indicate that IL-7 present in RA synovial fluid attracts diated monocyte chemotaxis. We found that ERK, AKT1, STAT3, circulating IL-7R+ monocytes into the joints. and STAT5 but not STAT1 were phosphorylated by IL-7 activation

FIGURE 3. Inhibition of AKT and ERK cascades suppresses IL-7–mediated monocyte homing. As with RA, IL-7R is expressed in CIA ST lining and sublining macrophages and sublining endothelial cells. (A) NL monocytes were transfected with Ctl or DN-AKT plasmid at 2.5 mg for 48 h. Cells were either untreated or stimulated with 100 ng/ml of IL-7 for 30 and 60 min and probed for p-AKT, AKT, and actin (n = 3). (B) Migration of Ctl or DN-AKT– transfected NL monocytes was examined in response to 10 ng/ml IL-7 or PBS, and response to fMLF was tested only in Ctl-transfected monocytes (n = 3). (C) THP-1 cells were transfected with 100 nM scrambled (ctl si) or ERK siRNA (ERK si) for 48 h. Subsequently, transfected cells were probed for ERK and actin (n = 3). (D) Migration of THP-1 Ctl or ERK knockdown cells was examined in response to 10 ng/ml IL-7 or PBS and the ability of fMLF to attract cells was examined only in control knockdown cells (n = 3). (E) Polycarbonate membranes harvested from NL monocytes that migrated in response to PBS, fMLF (1 mM), or IL-7 (10 ng/ml) were fixed, stained with anti-CD68 Ab (1:100 dilution), and visualized with a secondary Ab labeled with Alexa 594 (1:500 dilution; original magnification 3400). (F) The total number of migrated cells was quantified in 15 high-powered fields based on double-positive DAPI (blue) and CD68 (red) staining. (G) Ankles harvested from PBS Ctl or CIA-induced mice were stained with anti–IL-7R Ab (original magnification 3200). (H) IL-7R positive immunostaining was scored on a 0–5 scale on ST lining and sublining macrophages (mac) as well as sublining endothelial cells (Endo; n = 5–7; original magnification 3200). (I) IL-7 levels were determined in ankles harvested from PBS-injected control or CIA mice by ELISA (n = 5). Values demonstrate mean 6 SE. *p , 0.05. 6 NOVEL ROLE OF IL-7 LIGATION IN RA in monocytes (Fig. 2D). To determine which of these pathways can strongly attract THP-1 cells, a human monocytic cell line might contribute to IL-7–mediated chemotaxis, monocytes were (Fig. 3D). We also show that similar to fMLF, cells attracted to IL-7 preincubated with inhibitors of the ERK, PI3K/AKT1, STAT3, and on the chemotaxis membrane are 99.02% positive for CD68 (based STAT5 before performing the chemotaxis. Although inhibition of on colocalization of CD68+ cells with DAPI nucleus staining; Fig. STAT3 and STAT5 pathways was ineffective, inhibition of the 3E, 3F). Collectively, our results demonstrate that IL-7 can strongly PI3K/AKT1 and ERK cascades significantly reduced IL-7–induced recruit monocytes through ligation of IL-7R and activation of PI3K/ monocyte recruitment (Fig. 2E). We also demonstrate that inhibi- AKT1 and ERK but not STAT pathways. tion of PI3K/AKT1 is specific to monocyte migration mediated by IL-7, because suppression of this pathway did not affect extrava- IL-7 and IL-7R expression levels are greatly elevated in CIA sation induced by other potent monocyte chemoattractants includ- ankle joints and anti–IL-7 therapy reduces joint inflammation ing CCL2/MCP-1, CCL5/RANTES, IL-17, or fMLF (Fig. 2F). We and bone destruction also show that when monocyte chemoattractants are used at To evaluate the mechanism by which IL-7/IL-7R induces RA the similar concentration, IL-7 is at least as potent as CCL2/MCP-1 pathogenesis, an experimental arthritis model was used. We show in attracting monocytes (Fig. 2F). Although CCL2-, CCL5-, and IL- that, as in RA, IL-7R is significantly elevated in the lining and 17–induced (21) myeloid cell migration is promoted through p38 sublining macrophages as well as sublining endothelial cells in CIA (Fig. 2F), this pathway is not activated by IL-7 in myeloid cells, compared with PBS-treated ankles (Fig. 3G, 3H). In addition, CIA suggesting that the IL-7 mechanism of action does not overlap with mice produced 3-fold higher joint IL-7 levels compared with the other monocyte chemoattractants. control group (Fig. 3I). Therefore, to examine the role of IL-7/IL- To eliminate the possible nonspecific effect of kinase inhibitors, 7R in CIA pathology, CIA mice were treated therapeutically with significance of AKT and ERK activation was assessed on IL-7 anti–IL-7 Ab or IgG control (Fig. 4A, 4B) starting on day 26 after monocyte recruitment using DN-AKT vector or ERK siRNA. CIA induction. These studies demonstrated that anti–IL-7 Ab We demonstrate that expression of DN-AKT in normal myeloid treatment significantly reduced joint inflammation on days 33, 37, cells was capable of markedly reducing AKT1 phosphorylation and 40, 41, and 42 after CIA induction compared with the control monocyte infiltration mediated by IL-7, whereas the control had no group; however, there were no differences detected between the effect on these two functions (Fig. 3A, 3B). Next, we confirm that two treatment groups on days 28 and 30 (Fig. 4A). We document IL-7–induced THP-1 chemotaxis is significantly suppressed by ERK that blockade of IL-7 was also capable of reducing CIA synovial compared with the control knockdown (Fig. 3C, 3D). inflammation (40%), joint lining thickness (45%), and erosion To document that cells migrating in response to IL-7 are mono- (40%) compared with the control treated mice (Fig. 4B, 4C). cytes rather than other ambiguous cells, we demonstrate that IL-7 Consistent with the histologic studies, we demonstrate that TRAP+

FIGURE 4. Therapeutic treatment of anti-IL-7 Ab ameliorates CIA pathology and bone erosion. (A) Changes in joint circumference were recorded for CIA mice that were treated with IgG or anti–IL-7 Ab (100 mg) i.p. on days 26, 29, 33, 36, 40, and 42 (n = 10–11 mice). (B) Effect of anti–IL-7 Ab treatment on inflammation, lining thickness, and bone erosion was scored on a 0–5 scale (n = 8). (C) Representative ankle H&E staining (original magnification 3 200). (D) TRAP staining of CIA mice treated with IgG or anti–IL-7 Ab was scored on a 0–5 scale (n = 8). (E) Representative ankle TRAP staining (original magnification 3 200); arrows demonstrate TRAP+ cells. (F) mRNA concentration of RANKL and cathepsin K was determined in CIA ankles treated with IgG or anti–IL-7 Ab using real-time RT-PCR (n = 5). The data are shown as fold increase above IgG-treated CIA ankles and are normalized to GAPDH. Values are mean 6 SE. *p , 0.05. The Journal of Immunology 7 cells are markedly higher in control mice (45%) compared with suggest that like in RA, IL-7 plays a key role in myeloid cell traf- anti–IL-7 Ab treatment group (Fig. 4D, 4E). We next show that ficking and function in CIA joint. bone loss detected in anti–IL-7 Ab-treated CIA mice is due to Anti–IL-7 treatment reduced CIA vascularization, joint MIP-2, significant decrease in key bone erosion markers including and hemoglobin levels compared with control treatment RANKL (3-fold) and cathepsin K (10-fold) expression (Fig. 4F) compared with control mice. These results indicated that IL-7 has Because we have demonstrated previously that macrophages and a crucial role in CIA disease progression and bone erosion; endothelial cells stimulated with IL-7 could produce a number of therefore, subsequent studies were performed to identify the cell proangiogenic factors (5), the effect of anti–IL-7 therapy was types and the mechanism by which IL-7 promotes disease. examined on CIA vascularization and joint proangiogenic factors. We show that anti–IL-7 Ab treatment in CIA ankles specifically a Anti–IL-7 Ab therapy reduces joint TNF- and CCL2/MCP-1 reduces MIP-2 levels while having no effect on CXCL1, Ang-1, levels and monocyte trafficking in CIA mice bFGF, and VEGF concentration compared with the control treat- Based on the IL-7/IL-7R feedback regulation with TNF-a shown ment (Fig. 6A–E). Interestingly, anti–IL-7 Ab treatment was ca- by us (5) and others (13) and the ability of IL-7 to induce monocyte pable of reducing CIA ankle hemoglobin levels by 6-fold (Fig. 6F) migration, we asked whether joint TNF-a or other potent monocyte and joint vascularization (Fig. 6G, 6H), suggesting that IL-7 can chemoattractants were affected by anti–IL-7 therapy in CIA. We contribute to angiogenesis in CIA. found that joint TNF-a (Fig. 5A) as well as ankle and serum levels of CCL2/MCP-1 (Fig. 5B) were 2-fold higher in the IgG group Joint TH-17 promoting cytokines, TH-17 polarization, and compared with anti–IL-7 Ab-treated CIA mice; however, concen- T cell trafficking was unaffected by anti–IL-7 Ab treatment in tration of joint CCL5/RANTES (Fig. 5C) was not significantly CIA mice different in the two treatment groups. Because of these findings, we Previous studies had demonstrated the importance of IL-7 in TH-1 next examined the role of IL-7 in CIA-mediated monocyte migra- (30) and TH-17 differentiation (11); therefore, we asked whether tion. We found that anti–IL-7 Ab treatment significantly reduced blockade of IL-7 could affect the percentage of CD3, CD4, TH-1, monocyte joint recruitment compared with control group (Fig. 5D, and TH-17–positive cells in splenocytes. We show that the per- 5E). This finding may be due to disruption in joint IL-7 ligation to centage of CD3, CD4, TH-1, and TH-17–positive cells was similar IL-7R on monocytes and their homing into the inflammatory site or in anti–IL-7 and IgG treatment groups (Fig. 7A). Consistent with indirectly because of reduced potent monocyte chemoattractant these findings, joint IL-6 levels (Fig. 7B) were unaffected by anti– such as TNF-a and CCL2. It is also possible that both mechanisms IL-7 therapy, but there was an insignificant trend toward lower of action contribute to this detected effect. Collectively, these results levels of joint IL-1b and IL-17 (Fig. 7C, 7D). Therefore, the effect

FIGURE 5. Neutralization of IL-7 reduces potent monocyte chemoattractants and CIA monocyte homing. Changes in TNF-a (A), CCL2 (B), and CCL5 (C) expression levels in ankle homogenates (n = 7) and sera (n = 10) from CIA mice treated with IgG control or anti–IL-7 Ab were determined by ELISA. (D) STs from CIA mice treated with IgG or anti–IL-7 Ab were harvested on day 43 and immunostained with anti-F480 Ab (original magniﬁcation 3 200). Arrows demonstrate F480+ cells. (E) Macrophage staining was quantiﬁed on a 0–5 scale (n = 5–7). Values are mean 6 SE. *p , 0.05. 8 NOVEL ROLE OF IL-7 LIGATION IN RA

FIGURE 6. MIP-2 and joint vascularization is reduced in CIA mice treated with anti–IL-7 Ab compared with the control group. Changes in MIP-2 (A), Ang-1 (B), CXCL1 (C), basic fibroblast growth factor (D), and vascular endothelial growth factor (E) expression levels in ankle homogenates (n = 7) from CIA mice treated with IgG control or anti–IL-7 Ab were determined by ELISA. (F) Hemoglobin levels are quantified in CIA ankles harvested from control and anti–IL-7 Ab treatment groups on day 43. Results are demonstrated as hemoglobin (grams per deciliter) per joint weight (milligrams per milliliter). (G) STs from CIA mice treated with IgG or anti–IL-7 Ab were harvested and immunostained with anti-VWF Ab (original magnification 3 200; small insets, original magnification 3400). Cells positive for VWF staining are shown by arrows. (H) Endothelial immunostaining was quantified on a 0–5 scale (n = 5–7). Values are mean 6 SE. *p , 0.05. of anti–IL-7 Ab treatment was determined on joint T cell migra- levels of TNF-a and number of CD68+ cells in RA tissue, but tion. We found that although joint T cells were not significantly there was no correlation found between IL-7 levels and numbers reduced, there was a trend toward a lower number of T cells in the of RA ST CD3+, CD8+, or CD8+ T cells, suggesting that myeloid anti–IL-7 Ab-treated CIA mice compared with the control group cells are the main source of IL-7 production (13). Furthermore, it (Fig. 7E, 7F). has been shown that TNF-a is the common factor that induces expression of IL-7 and IL-7R in RA myeloid cells and endothelial Discussion cells (5), suggesting that there is a positive feedback regulation This study identifies a novel mechanism for IL-7/IL-7R in CD14+ between TNF-a and the IL-7/IL-7R cascade. As such, anti–TNF-a CD162 cell trafficking through activation of the AKT/PI3K and responders (13) show significantly reduced circulating IL-7 levels, ERK cascades that is distinct from the effect of IL-7 on T cells, reflecting our results that demonstrate a strong association be- which is mediated through the JAK/STAT pathway (31). In ad- tween IL-7/IL-7R and TNF-a and RA disease severity. dition, we show that expression of IL-7 and IL-7R in RA mono- Macrophages in RA synovial fluid expressed 45 fold higher IL- cytes is linked to increased TNF-a and DAS28 levels, suggesting 7R compared with control cells (5); therefore, we postulated that the importance of IL-7/IL-7R function in myeloid cells to RA IL-7 expressed in RA synovial fluid could be important for progression. To demonstrate the effect of IL-7R ligation to IL-7 attracting myeloid cells from blood into the joint. We show that in disease pathogenesis, we document that therapeutic anti–IL-7 blockade of IL-7R and neutralization of IL-7 can significantly treatment in CIA relieves arthritis by reducing monocyte extrav- reduce synovial fluid–mediated monocyte migration, justifying asation, osteoclast differentiation, and joint vascularization (Fig. elevated levels of IL-7R on RA joint macrophages. Interestingly, 7G). Although the role of IL-7 and IL-7R in T cell differentiation the mechanism by which IL-7 mediates monocyte recruitment is is well described, their significance in the effector phase of RA not shared by other potent monocyte chemoattractant and is dis- and myeloid cell function is unknown. tinct from signaling pathways associated with IL-7–induced T cell Based on our recent studies demonstrating that IL-7 and IL-7R differentiation (31). are highly elevated in RA synovial fluid, tissue, and blood myeloid Expression of IL-7R in CIA lining and sublining myeloid cells cells (5), we asked whether expression of these factors has a sig- as well as upregulation of IL-7 in CIA ankle joints justified the use nificant role in myeloid cell function. Previous studies had shown of this experimental arthritis model to examine whether the IL-7/ that cell-to-cell contact of macrophages with T cells was required IL-7R cascade could be used as a target for RA therapy. We show for IL-7–mediated TNF-a production (6); however, IL-7–activated that therapeutic treatment of CIA mice markedly reduces joint PB monocytes did not require T cell interaction and were capable inflammation, lining thickness, and bone degradation. We dem- of producing TNF-a (15). Concentration of IL-7 correlated with onstrate that downregulation of TRAP+ cells in anti–IL-7 treated The Journal of Immunology 9

FIGURE 7. Spleen TH-1 cells, TH-17 cells, and TH-17 promoting cytokines and joint CD3 immunostaining were unaffected by anti–IL-7 Ab treatment and schematic representation of the mechanism that contributes to IL-7–mediated pathogenesis in RA and CIA. (A) Percent spleen CD3+, CD4+, TH-1, and TH-17–positive cells were determined by flow cytometry analysis in control IgG and anti–IL-7 Ab treatment in CIA. Changes in IL-6 (B), IL-1b (C), and IL-17 (D) expression levels in ankle homogenates (n = 7) from CIA mice treated with IgG control or anti–IL-7 Ab were quantified by ELISA. (E) STs from CIA mice treated with IgG or anti–IL-7 Ab were immunostained with anti-CD3 Ab (original magnification 3 200) and cells positive for CD3 staining are demonstrated by arrows. (F) T cell immunostaining was quantified on a 0–5 scale (n = 5–7). (G) IL-7 is capable of modulating monocyte homing in an RA joint as well as contributing to CIA myeloid cell trafficking and function, bone erosion, and vascularization.

CIA mice is due to decreased RANKL and cathepsin K levels. duction levels, despite their study showing a lack of an effect in Interestingly, others have shown that osteoblasts activated with number of spleen myeloid cells and that joint macrophage im- TNF-a and IL-1b produce IL-7, which can contribute to RANKL- munostaining was not performed (36). Our in vivo results in CIA dependent or independent osteoclastogenesis (32). In RA, syno- correlate with findings of in vitro studies demonstrating that in vial tissue fibroblasts or synovial fluid T cells (33) are responsible myeloid cells, concentrations of TNF-a (5) and IL-7 (15) are in for RANKL secretion. Because anti–IL-7 treatment was unable to cross regulation, which is consistent with the strong relation noted markedly reduce joint T cell migration, our results might suggest between CD68+ cells and levels of IL-7 and TNF-a in RA ST as that IL-7 mainly affects RANKL transcription from joint fibro- well as reduction in circulating IL-7 levels detected in TNF-a blasts. Consistent with this conclusion, we (5) and others (34) responders compared with nonresponders (13). However, we were demonstrate that RA fibroblasts express both IL-7 and IL-7R and unable to demonstrate that IL-7 is capable of inducing CCL2 from that levels of this ligand and receptor are modulated by TNF-a, myeloid cells, RA fibroblasts, and endothelial cells; therefore, re- indicating the responsiveness of fibroblasts to IL-7 stimulation. duction of CCL2 levels in anti–IL-7 Ab therapy may be due to Our results also highlight that fold reduction of cathepsin K (10- reduced joint monocyte differentiated macrophages rather than a fold decrease), a collagen degrading protease, present in myeloid direct effect of IL-7 on the joint. In addition, the lack of effect cells and mature osteoclasts (35) is greater than RANKL (3-fold observed following anti–IL-7 Ab treatment on joint CCL5 levels decrease), which is in agreement with the effect of anti–IL-7 may be the result of the inefficient reduction of joint T cells, because treatment on CIA myeloid cell homing. CCL5 is primarily secreted from T cells and myeloid cells (37). We also show that potent monocyte chemoattractants, such as One of the mechanisms by which IL-7 induces pathogenesis is its TNF-a and CCL2 but not CCL5, were significantly reduced in the ability to induce potent proangiogenic factors such as IL-8 and ankles of mice with CIA that were anti–IL-Ab treated compared Ang-1 from macrophages, endothelial cells, or both (5). In CIA, with controls. This finding suggests that reduction in joint myeloid neutralization of IL-7 was capable of reducing joint MIP-2 (ho- cell extravasation following anti–IL-7 Ab treatment could be due molog of human IL-8); however, regarding Ang-1, bFGF, CXCL1, to the disrupted ligation of IL-7 to IL-7R on monocytes or indi- and VEGF, although lower, their concentration was not signifi- rectly due to the lower levels of joint TNF-a and CCL2. Our cantly reduced compared with the control group. Consistently, results corroborate earlier studies demonstrating that preventative others have shown that despite the inefficiency of anti–IL-7R anti–IL-7R treatment could reduce CIA front paw TNF-a pro- Ab treatment in reducing CIA joint bFGF concentrations, von 10 NOVEL ROLE OF IL-7 LIGATION IN RA

Willebrand factor (VWF) expression levels were significantly re- 8. Hartgring, S. A., J. W. Bijlsma, F. P. Lafeber, and J. A. van Roon. 2006. duced compared with the control group (36), suggesting that IL-7/ Interleukin-7 induced immunopathology in arthritis. Ann. Rheum. Dis. 65(Suppl 3): iii69–iii74. IL-7R–mediated angiogenesis in CIA is distinct from VEGF or 9. Haas, J., M. Korporal, A. Schwarz, B. Balint, and B. Wildemann. 2011. The bFGF pathways. Although IL-7R expression is elevated in endo- interleukin-7 receptor a chain contributes to altered homeostasis of regulatory T cells in multiple sclerosis. Eur. J. Immunol. 41: 845–853. thelial cells both in RA (5) and CIA ankles, we were unable to 10. Churchman, S. M., and F. Ponchel. 2008. Interleukin-7 in rheumatoid arthritis. demonstrate that ligation of IL-7 to endothelial IL-7R could Rheumatology (Oxford) 47: 753–759. promote endothelial cell migration, suggesting that reduced vas- 11. Liu, X., S. Leung, C. Wang, Z. Tan, J. Wang, T. B. Guo, L. Fang, Y. Zhao, B. Wan, X. Qin, et al. 2010. Crucial role of interleukin-7 in T helper type 17 cularization in CIA is indirectly due to modulation of joint MIP-2 survival and expansion in autoimmune disease. Nat. Med. 16: 191–197. levels. 12. Fry, T. J., and C. L. Mackall. 2002. Interleukin-7: from bench to clinic. Blood 99: We show that anti–IL-7 Ab treatment in CIA mice had no effect 3892–3904. 13. van Roon, J. A., S. A. Hartgring, M. Wenting-van Wijk, K. M. Jacobs, P. P. Tak, on the percentage of spleen CD3, CD4, TH1, and TH-17–positive J. W. Bijlsma, and F. P. Lafeber. 2007. Persistence of interleukin 7 activity and cells compared with control treatment. In agreement with these levels on tumour necrosis factor alpha blockade in patients with rheumatoid results, although there was a trend toward lower joint CD3 im- arthritis. Ann. Rheum. Dis. 66: 664–669. + 14. Standiford, T. J., R. M. Strieter, R. M. Allen, M. D. Burdick, and S. L. Kunkel. munostaining in the anti–IL-7 Ab treatment group, CD3 cells 1992. IL-7 up-regulates the expression of IL-8 from resting and stimulated hu- were not markedly reduced compared with the control group. man blood monocytes. J. Immunol. 149: 2035–2039. Similarly, levels of joint IL-6, IL-1b, and IL-17 were unaffected 15. Alderson, M. R., T. W. Tough, S. F. Ziegler, and K. H. Grabstein. 1991. Inter- leukin 7 induces cytokine secretion and tumoricidal activity by human peripheral by the treatment. In contrast with our results, preventative treat- blood monocytes. J. Exp. Med. 173: 923–930. ment of CIA with anti–IL-7R Ab demonstrated a significant re- 16. Toyosaki-Maeda, T., H. Takano, T. Tomita, Y. Tsuruta, M. Maeda-Tanimura, duction in front paw IL-17 and IL-1b levels as well as the Y. Shimaoka, T. Takahashi, T. Itoh, R. Suzuki, and T. Ochi. 2001. Differentiation + + of monocytes into multinucleated giant bone-resorbing cells: two-step differ- percentage of CD4 , CD8 , naı¨ve, and memory T cells (36). The entiation induced by nurse-like cells and cytokines. Arthritis Res. 3: 306–310. contrasting data may be due to treatment time point, indicating 17. Takano, H., T. Tomita, T. Toyosaki-Maeda, M. Maeda-Tanimura, H. Tsuboi, that ligation of IL-7 to IL-7R before disease onset could be im- E. Takeuchi, M. Kaneko, K. Shi, K. Takahi, A. Myoui, et al. 2004. Comparison of the activities of multinucleated bone-resorbing giant cells derived from CD14- portant in maintaining and restoring T cell homeostasis, whereas positive cells in the synovial fluids of rheumatoid arthritis and osteoarthritis subsequent to disease onset these factors could have a role in patients. Rheumatology (Oxford) 43: 435–441. myeloid cell migration and function. Previous studies demonstrate 18. Lee, S. K., and C. D. Surh. 2005. Role of interleukin-7 in bone and T-cell homeostasis. Immunol. Rev. 208: 169–180. that IL-7 contributes to TH-17 cell differentiation by increasing 19. Toraldo, G., C. Roggia, W. P. Qian, R. Pacifici, and M. N. Weitzmann. 2003. IL- IL-1R1 expression on CD4+ cells, thereby making these CD4+IL- 7 induces bone loss in vivo by induction of receptor activator of nuclear factor 1R1+ cells more responsive to IL-1b stimulation (38). Therefore, kappa B ligand and tumor necrosis factor alpha from T cells. Proc. Natl. Acad. Sci. USA 100: 125–130. anti–IL-7 Ab treatment may have failed to reduce the percentage 20. Pickens, S. R., N. D. Chamberlain, M. V. Volin, R. M. Pope, A. M. Mandelin, II, of TH-17 cell development because of its lack of effect on IL-1R1 and S. Shahrara. 2011. Characterization of CCL19 and CCL21 in rheumatoid + arthritis. Arthritis Rheum. 63: 914–922. expression in CD4 cells owing to treatment after onset. Similar to 21. Shahrara, S., S. R. Pickens, A. Dorfleutner, and R. M. Pope. 2009. IL-17 induces our findings in CIA, others have shown that TH-1 cell population monocyte migration in rheumatoid arthritis. J. Immunol. 182: 3884–3891. is unaffected by IL-7R antagonist treatment in autoimmune en- 22. Shahrara, S., S. R. Pickens, A. M. Mandelin, II, W. J. Karpus, Q. Huang, J. K. Kolls, and R. M. Pope. 2010. IL-17-mediated monocyte migration occurs cephalomyelitis (11). partially through CC chemokine ligand 2/monocyte chemoattractant protein-1 Our findings in the effector phase of CIA suggest that ligation induction. J. Immunol. 184: 4479–4487. of IL-7 to IL-7R contributes to increased monocyte homing and 23. Gingras, A. C., S. G. Kennedy, M. A. O’Leary, N. Sonenberg, and N. Hay. 1998. 4E-BP1, a repressor of mRNA translation, is phosphorylated and inactivated by angiogenesis, which reflects the expression pattern of IL-7 and IL- the Akt(PKB) signaling pathway. Genes Dev. 12: 502–513. 7R in RA ST (5). Therefore, association of IL-7 and IL-7R with 24. Arnett, F. C., S. M. Edworthy, D. A. Bloch, D. J. McShane, J. F. Fries, the two aforementioned functions, as well as its close relation with N. S. Cooper, L. A. Healey, S. R. Kaplan, M. H. Liang, H. S. Luthra, et al. 1988. The American Rheumatism Association 1987 revised criteria for the classifi- TNF-a in myeloid cells, suggests that IL-7 and IL-7R can be used cation of rheumatoid arthritis. Arthritis Rheum. 31: 315–324. as targets for RA treatment. 25.Pickens,S.R.,N.D.Chamberlain,M.V.Volin,A.M.Mandelin,II, H. Agrawal, M. Matsui, T. Yoshimoto, and S. Shahrara. 2011. Local expression of interleukin-27 ameliorates collagen-induced arthritis. Arthritis Disclosures Rheum. 63: 2289–2298. The authors have no financial conflicts of interest. 26. Haas, C. S., M. A. Amin, J. H. Ruth, B. L. Allen, S. Ahmed, A. Pakozdi, J. M. Woods, S. Shahrara, and A. E. Koch. 2007. In vivo inhibition of angiogenesis by interleukin-13 gene therapy in a rat model of rheumatoid arthritis. Arthritis Rheum. 56: 2535–2548. References 27. Pickens, S. R., M. V. Volin, A. M. Mandelin, II, J. K. Kolls, R. M. Pope, and 1. Brentano, F., D. Kyburz, O. Schorr, R. Gay, and S. Gay. 2005. The role of Toll-like S. Shahrara. 2010. IL-17 contributes to angiogenesis in rheumatoid arthritis. J. receptor signalling in the pathogenesis of arthritis. Cell. Immunol. 233: 90–96. Immunol. 184: 3233–3241. 2. Drexler, S. K., S. M. Sacre, and B. M. Foxwell. 2006. Toll-like receptors: a new 28. Pickens, S. R., N. D. Chamberlain, M. V. Volin, R. M. Pope, N. E. Talarico, target in rheumatoid arthritis? Expert Rev. Clin. Immunol. 2: 585–599. A. M. Mandelin, 2nd, and S. Shahrara. 2012. Role of the CCL21 and CCR7 3. Kondrack, R. M., J. Harbertson, J. T. Tan, M. E. McBreen, C. D. Surh, and pathways in rheumatoid arthritis angiogenesis. Arthritis Rheum. 64: 2471–2481. L. M. Bradley. 2003. Interleukin 7 regulates the survival and generation of 29. Scatizzi, J. C., E. Bickel, J. Hutcheson, G. K. Haines, III, and H. Perlman. 2006. memory CD4 cells. J. Exp. Med. 198: 1797–1806. Bim deficiency leads to exacerbation and prolongation of joint inflammation in 4. Seddon, B., P. Tomlinson, and R. Zamoyska. 2003. Interleukin 7 and T cell experimental arthritis. Arthritis Rheum. 54: 3182–3193. receptor signals regulate homeostasis of CD4 memory cells. Nat. Immunol. 4: 30. Bikker, A., F. M. Moret, A. A. Kruize, J. W. Bijlsma, F. P. Lafeber, and J. A. van 680–686. Roon. 2012. IL-7 drives Th1 and Th17 cytokine production in patients with 5. Pickens, S. R., N. D. Chamberlain, M. V. Volin, R. M. Pope, N. E. Talarico, primary SS despite an increase in CD4 T cells lacking the IL-7Ra. Rheuma- A. M. Mandelin, II, and S. Shahrara. 2011. Characterization of interleukin-7 and tology (Oxford) 51: 996–1005. 31. Hofmeister, R., A. R. Khaled, N. Benbernou, E. Rajnavolgyi, K. Muegge, and interleukin-7 receptor in the pathogenesis of rheumatoid arthritis. Arthritis S. K. Durum. 1999. Interleukin-7: physiological roles and mechanisms of action. Rheum. 63: 2884–2893. Cytokine Growth Factor Rev. 10: 41–60. 6. van Roon, J. A., M. C. Verweij, M. W. Wijk, K. M. Jacobs, J. W. Bijlsma, and 32. Weitzmann, M. N., S. Cenci, L. Rifas, C. Brown, and R. Pacifici. 2000. F. P. Lafeber. 2005. Increased intraarticular interleukin-7 in rheumatoid arthritis Interleukin-7 stimulates osteoclast formation by up-regulating the T-cell pro- patients stimulates cell contact-dependent activation of CD4(+) T cells and duction of soluble osteoclastogenic cytokines. Blood 96: 1873–1878. macrophages. Arthritis Rheum. 52: 1700–1710. 33. Schett, G., S. Hayer, J. Zwerina, K. Redlich, and J. S. Smolen. 2005. Mecha- 7. Hartgring, S. A., J. A. van Roon, M. Wenting-van Wijk, K. M. Jacobs, nisms of Disease: the link between RANKL and arthritic bone disease. Nat. Clin. Z. N. Jahangier, C. R. Willis, J. W. Bijlsma, and F. P. Lafeber. 2009. Elevated ex- Pract. Rheumatol. 1: 47–54. pression of interleukin-7 receptor in inflamed joints mediates interleukin-7-induced 34. Harada, S., M. Yamamura, H. Okamoto, Y. Morita, M. Kawashima, T. Aita, and immune activation in rheumatoid arthritis. Arthritis Rheum. 60: 2595–2605. H. Makino. 1999. Production of interleukin-7 and interleukin-15 by fibroblast- The Journal of Immunology 11

like synoviocytes from patients with rheumatoid arthritis. Arthritis Rheum. 42: T cell activity and proinﬂammatory mediators. Arthritis Rheum. 62: 2716– 1508–1516. 2725. 35. Kneissel, M., N. H. Luong-Nguyen, M. Baptist, R. Cortesi, S. Zumstein-Mecker, 37. Miyagishi, R., S. Kikuchi, C. Takayama, Y. Inoue, and K. Tashiro. 1997. Iden- S. Kossida, T. O’Reilly, H. Lane, and M. Susa. 2004. Everolimus suppresses tiﬁcation of cell types producing RANTES, MIP-1 alpha and MIP-1 beta in rat cancellous bone loss, bone resorption, and cathepsin K expression by osteoclasts. experimental autoimmune encephalomyelitis by in situ hybridization. J. Neu- Bone 35: 1144–1156. roimmunol. 77: 17–26. 36. Hartgring, S. A., C. R. Willis, D. Alcorn, L. J. Nelson, J. W. Bijlsma, 38. Lee, W. W., S. W. Kang, J. Choi, S. H. Lee, K. Shah, E. E. Eynon, R. A. Flavell, F. P. Lafeber, and J. A. van Roon. 2010. Blockade of the interleukin-7 re- and I. Kang. 2010. Regulating human Th17 cells via differential expression of ceptor inhibits collagen-induced arthritis and is associated with reduction of IL-1 receptor. Blood 115: 530–540.