Platelet-Derived Growth Factor Receptor Activation Promotes The

Home , TGF alpha

Platelet-Derived Growth Factor Receptor Activation Promotes the Prodestructive Invadosome-Forming Phenotype of Synoviocytes from Patients with Rheumatoid This information is current as Arthritis of October 2, 2021. Martine Charbonneau, Roxane R. Lavoie, Annie Lauzier, Kelly Harper, Patrick P. McDonald and Claire M. Dubois J Immunol 2016; 196:3264-3275; Prepublished online 14

March 2016; Downloaded from doi: 10.4049/jimmunol.1500502 http://www.jimmunol.org/content/196/8/3264

References This article cites 93 articles, 21 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/196/8/3264.full#ref-list-1

Why The JI? Submit online.

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

• No Triage! Every submission reviewed by practicing scientists by guest on October 2, 2021

• 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 © 2016 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Platelet-Derived Growth Factor Receptor Activation Promotes the Prodestructive Invadosome-Forming Phenotype of Synoviocytes from Patients with Rheumatoid Arthritis

Martine Charbonneau,*,1 Roxane R. Lavoie,*,1 Annie Lauzier,* Kelly Harper,* Patrick P. McDonald,† and Claire M. Dubois*

Fibroblast-like synoviocytes (FLS) play a major role in invasive joint destruction in rheumatoid arthritis (RA). This prodestructive phenotype has been shown to involve autocrine TGF-b that triggers formation of matrix-degrading invadosomes through molecular mechanisms that are not fully elucidated. The platelet-derived growth factor (PDGF) receptor (PDGFR) family of receptor tyrosine kinases (RTK) has been shown to cooperate with TGF-b in various pathological conditions. We therefore sought to determine whether RTK activity played a role in invadosome biogenesis. We demonstrated that, among the common RTKs, PDGFR-ab was specifically phosphorylated in FLS from RA patients. Phosphorylation of PDGFR-ab was also elevated in RA Downloaded from synovial tissues. Interference with PDGFR activation or PDGF neutralization inhibited invadosome formation in RA synoviocytes, indicating the presence of an autocrine PDGFR activation loop that involved endogenous PDGF. Among the PDGF-A–D isoforms, only PDGF-B was found both significantly elevated in FLS lines from RA patients, and related to high-invadosome forming cells. Addition of TGF-b upregulated invadosome formation, PDGF-B mRNA expression, and phosphorylation of PDGFR. All of these functions were efficiently suppressed by TGF-b neutralization or interference with the Smad/TbR1or PI3K/Akt pathway. Among the class 1 PI3K family proteins known to be expressed in RA synoviocytes, PI3Ka was selectively involved in PDGF-B expression, http://www.jimmunol.org/ whereas both PI3Ka and PI3Kd participated in invadosome formation. Our findings demonstrate that PDGFR is a critical RTK required for the prodestructive phenotype of RA synovial cells. They also provide evidence for an association between autocrine TGF-b and PDGFR-mediated invadosome formation in RA synoviocytes that involves the production of PDGF-B induced by TGF-b. The Journal of Immunology, 2016, 196: 3264–3275.

heumatoid arthritis (RA) is a systemic autoimmune dis- inﬂammation and joint destruction (2–4). Arthritic FLS resemble ease that mainly affects the joints, leading to joint in- transformed mesenchymal cells that are highly invasive in vitro

R flammation and erosive structural damages. Although and in vivo. This property correlates with elevated production of by guest on October 2, 2021 important progress has been made in managing the pain and in- inflammatory cytokines and proteolytic enzymes that sustain inflammation associated with the disease, strategies to directly inflammation and joint matrix degradation. We have reported that terfere with the process of erosion are lacking. The onset of RA the ability of arthritic FLS to degrade the extracellular matrix causes important morphological changes in joint lining, including depends on the formation of plasma membrane structures that formation of an aggressive tumor-like synovial tissue that invades resembled invadopodia in tumor cells (5, 6). These structures were and erodes cartilage and bone (1). A large body of evidence from detected in fibroblast-like cells strategically located at the carti- patients and experimental animal models indicated that fibroblast- lage–synovial membrane interface. They were shown to contain like synoviocytes (FLS) are the main cell type that actively drives actin components, signaling molecules, such as Src, and high levels of proteolytic enzymes known to be particularly efficient at *Immunology Division, Department of Pediatrics, Faculty of Medicine, University of inducing cartilage damage. Importantly, interference with the Sherbrooke, Sherbrooke, Quebec J1H 5N4, Canada; and †Pneumology Division, Department of Medicine, Faculty of Medicine, University of Sherbrooke, Sherbrooke formation of invadosomes in arthritic FLS strongly inhibited Quebec J1H 5N4, Canada matrix degradation in vitro and ex vivo as well as cartilage deg- 1M.C. and R.R.L. are cofirst authors. radation in a rat model of arthritis (5, 6). These observations ORCIDs: 0000-0001-6253-8090 (R.R.L.); 0000-0002-0751-2821 (C.M.D.). suggested that invadosomes were directly involved in joint deg- Received for publication March 4, 2015. Accepted for publication February 15, 2016. radation, leading to the conclusion that an in-depth understanding This work was supported by Canadian Institutes for Health Research (CIHR) Grants of the mechanism of invadosome formation is of importance for MOP-86634 and MOP-286621 (to C.M.D.). C.M.D. is a member of the Fonds de la development of joint protection strategies for the clinical man- Recherche en Sante´ du Que´bec–funded Centre de Recherche du Centre Hospitalier agement of RA. Universitaire de Sherbrooke. K.H. is recipient of a scholarship from CIHR. The mechanisms involved in invadosome formation in synovial Address correspondence and reprint requests to Dr. Claire M. Dubois, Immunology Division, Department of Pediatrics, Faculty of Medicine, Universite´ de Sherbrooke, cells are not fully known. Invadosome formation and in vivo 3001 12th Avenue North, Sherbrooke, QC J1H 5N4, Canada. E-mail address: Claire. cartilage degradation capability of synovial cells of collagen- [email protected] induced arthritis rats were shown to depend on an autocrine ac- Abbreviations used in this article: ACR, American College of Rheumatology; tivation loop that involved TGF-b (6). Analysis of the protein and ECM, extracellular matrix; FLS, fibroblast-like synoviocyte; LPA, lysophosphatidic acid; NA, nonarthritis; OA, osteoarthritis; PDGF, platelet-derived growth factor; mRNA in RA synovial tissues revealed that TGF-b was highly PDGFR, PDGF receptor; PLC, phospholipase C; pY, phosphotyrosine; RA, rheumatoid expressed in RA patients (7–9). However, few studies have arthritis; RTK, receptor tyrosine kinase. addressed the role of TGF-b in the functions of synovial fibro- Copyright Ó 2016 by The American Association of Immunologists, Inc. 0022-1767/16/$30.00 blasts derived from these patients. TGF-b was shown to increase www.jimmunol.org/cgi/doi/10.4049/jimmunol.1500502 The Journal of Immunology 3265 the expression of proinflammatory cytokines and metalloproteinases agnosed using the 1986 ACR clinical criteria (34). The protocol was ap- in RA synoviocytes (9), an effect that was found to be dramati- proved by the Centre Hospitalier Universitaire de Sherbrooke Ethics cally potentiated by receptor tyrosine kinase (RTK)–dependent Committee, and written consent was obtained from all participants. Synoviocytes were isolated using standard procedures (35), and the signaling (10). These studies suggested a potential association culture was maintained in DMEM-F12 medium supplemented with 10% between TGF-b and RTK signaling to promote proarthritic FBS and 40 mg/ml gentamicin. Cells were used between passages 3 and 8. functions of synovial fibroblasts. Synoviocyte cultures exhibited a classic spindle-shape fibroblastic mor- RTKs comprise a large family of cell surface receptors that are phology that formed parallel clusters at confluency. The cell surface phenotypic marker analysis showed that they were consistently positive for essential components of signal transduction pathways that medi- the stromal mesenchymal marker fibronectin (.99%) and negative (,1%) ate cell survival, proliferation, differentiation, and motility, and for the macrophage marker CD68. modulate cell metabolism (11). These transmembrane proteins bind polypeptide ligands, mainly growth factors. Among the 58 Plasmids and transfections RTK family members, epidermal growth factor receptor, platelet- pLKO.1-puro short hairpin RNA targeting PI3Ka, PI3Kb, PI3Kd, and derived growth factor (PDGF) receptor (PDGFR), fibroblast control (scrambled) short hairpin RNA plasmids were from Sigma-Aldrich growth factor receptor, vascular endothelial growth factor recep- (Oakville, ON, Canada). Viral particles were generated by transient trans- fection of 293T cells with the ViraPower lentiviral expression system tor, hepatocyte growth factor receptor (c-Met), and stem cell (Invitrogen Thermo Fisher Scientific, Burlington, ON, Canada). Experi- growth factor receptor (c-kit) have been shown to be expressed in ments were conducted 48 h following lentiviral infection with Polybrene rheumatoid synovial tissues (11–19). On the basis of their roles as (5 mg/ml; EMD Millipore). Transfected cells were selected by treatment growth factor receptors, several RTKs are the driving force for with puromycin for 72 h. onset or progression of various diseases such as malignancy and Immunofluorescence and confocal microscopy arthritis and may represent key targets in therapeutic treatments. Downloaded from For instance, in the case of arthritis management, immatinib Synoviocytesculturedoncoverslipsfor4hwerefixedin2%paraformaldehyde, permeabilized with 0.05% saponin, and blocked with 2% mesylate, a tyrosine kinase inhibitor initially used for treatment of BSA. The following Abs or reagents were used to stain for actin chronic myeloid leukemia, has been shown to reduce activation of (phalloidin conjugated to Alexa647), cortactin (EMD Millipore), phos- RA synoviocytes by interfering with PDGFR signaling (10, 20). In photyrosine (pY) (EMD Millipore), and p-Src Tyr418 (Novus Biologicals, experimental arthritis, imatinib has also been reported to markedly Littleton, CO). Negative control slides were treated with isotype-

matched primary Abs, followed by secondary Abs. Confocal images http://www.jimmunol.org/ reduce joint erosion when administered before onset of the disease were acquired using a Fluoview 1000 scanning confocal microscope or during its progression (20–22). In RA patients, several reports (Olympus, Richmond Hill, ON, Canada) in line with an inverted have confirmed the efficacy of imatinib, and this has lead, in some Olympus microscope equipped with a 340 oil immersion objective. Color cases, to complete clinical remission (23–26). Despite these en- channels were scanned sequentially to avoid overlapping signals. A set of z m couraging results, the toxicity of imatinib was found to be high -stack images was collected at 0.25- m intervals and reconstructed using the FluoView software (Olympus). (27), possibly due to its pan-effect on multiple RTKs. The bulk of these observations suggested that identification of RTKs specif- Invadosome assays ically involved in joint erosion would be invaluable to design Coverslips were prepared using Oregon Green488–conjugated gelatin (Life specific strategies for development of targeted therapies to mini- Technologies, Burlington, ON, Canada) at a final concentration of 1%, as by guest on October 2, 2021 mize or prevent joint damage in RA. described (36). Thirty thousand cells were seeded on each coverslip, cul- The goal of the current study was to investigate whether RTKs tured for 40–48 h, and fixed with 1% paraformaldehyde. Nuclei were stained with DAPI, and F-actin was stained with Alexa647-conjugated played a role in invadosome formation by synovial cells and to phalloidin (Life Technologies; dilution 1:50). To determine the percentage identify the nature of these RTKs. RTKs, such as PDGFR, he- of cells forming invadosomes, stained cells were visualized using a Zeiss patocyte growth factor receptor, and epidermal growth factor Axioskop fluorescence microscope and invadosomes were identified by receptor, have been shown to direct matrix degradation and cell F-actin–enriched areas of matrix degradation. Three fields of 100 cells 3 invasion, a situation linked to invadosome formation in cancer (original magnification 40) were counted per coverslip. In selected experiments, we also measured the capacity of the cells to form actin/ cells (28–31). In this study, we show that activation (phos- cortactin-rich invadosomal structures. For this, cells were also stained phorylation) of PDGFR is specifically upregulated in RA syn- with anti–p-cortactin Abs, and clusters of p-cortactin/actin were calculated oviocytes and synovial tissues. We also show that PDGFR for 25 cells per slide. The capacity of invadosomes to degrade gelatin was activation involves TGF-b–induced PDGF-B upregulation me- also quantitated by measuring the areas of degradation per cell. Pictures of fluorescent matrix were analyzed using the ImagePro software, and deg- diated by TbR1/Smad and PI3K/AKT pathways. These findings radation areas associated with invadosome-positive cells were calculated in suggested the involvement of an overreactive TGF-b/PDGF-B/ pixels. A minimum of 25 cells was counted per coverslip. When specified PDGFR pathway in synoviocyte-driven extracellular matrix degra- under the figure legends, RA synovial cells or control synoviocytes were dation in RA. stimulated for 48 h with PDGF-BB or TGF-b1 (Peprotech, Montreal, QC, Canada), in the presence or absence of imatinib mesylate (Cayman Chemical, Cedarlane Laboratories, Burlington, ON, Canada), PDGFR ty- Materials and Methods rosine kinase inhibitor V (EMD Millipore, Billerica, MA), PDGF or FLS cell lines TGFb1-3 neutralizing Abs (R&D Systems, Minneapolis, MN), PI3K inhibitor LY294002 (Cayman), AKT inhibitor XI (EMD Millipore), phos- Human synovial cells derived from joint tissues (mostly knee joints) of pholipase C (PLC) inhibitor U73122, or the control compound U73343 patients diagnosed with osteoarthritis (OA) and RA, or from control (inactive analog of U73122). nonarthritis (NA) individuals were purchased from Asterand (Detroit, MI). The 1987 American College of Rheumatology (ACR) or 2010 ACR/ Western blotting European League Against Rheumatism classification criteria were used for diagnosing RA (32, 33), whereas the OA patients were classified using the Whole-cell extracts were prepared by lysis of overnight serum-starved cells 1986 ACR clinical criteria for OA (34). The NA cell lines were derived in radioimmunoprecipitation assay buffer. When specified in the figure from synovial tissues of cadavers who died of cardiovascular diseases. All legends, cells were incubated with imatinib mesylate (3 mM), PDGF-BB synovial tissues were obtained under approval of the local Institutional (10 ng/ml), or TGF-b (5 ng/ml). Proteins were immunoblotted, as de- Review Board and appropriate signed informed consent. Human synovial scribed (37), using anti-pPDGFRa (Tyr849)/pPDGFRb (Tyr857)(Cell tissue samples used for immunohistochemistry were obtained from OA or Signaling Technology; dilution 1:2000), anti-PDGFRb (Cell Signaling RA patients undergoing total knee joint replacement surgery. The 1987 Technology; dilution 1:1000), anti–p-AKT (Ser473) (Cell Signaling Tech- ACR or 2010 ACR/European League Against Rheumatism classification nology; dilution 1:2000), anti-AKT (Cell Signaling Technology; dilution criteria were used for diagnosing RA (32, 33), and OA patients were di- 1:1000), or anti–a-tubulin (Sigma-Aldrich, Oakville, ON, Canada; dilution 3266 PDGFR ACTIVATION PROMOTES INVADOSOME FORMATION IN RA-FLS

1:1000) Abs. Band intensities were analyzed using the Quantity One Results software (Bio-Rad Laboratories, Mississauga, ON, Canada). Invadosome generation and function are increased in RTK arrays synoviocytes from RA patients and require RTK activity Synovial cells were incubated in the presence or absence of imatinib We have recently shown that synovial cells from collagen-induced (3 mM) for 16 h in serum-free medium. A mixture of phosphatase arthritic rats spontaneously formed f-actin–enriched degradative inhibitors (10 mM sodium fluoride, 1 mM sodium ortho-vanadate, complexes that colocalized with zones enriched for the invado- 20 mM b-glycerophosphate) was added 20 min prior to cell lysis in the cell lysis buffer. Extracted proteins (0.3 mg/ml) were analyzed with the some markers cortactin, an actin assembly protein; Src, a tyrosine human PathScan RTK Signaling Antibody Array kit (Cell Signaling kinase involved in phosphorylation of actin-regulatory proteins; Technology), according to the manufacturer’s instructions. A fluores- and pY (5). In this study, we investigated whether similar obser- cent image of the slide was captured using an Odyssey Infrared Im- vations could be made in the case of human synovial cells of aging System, and fluorescence intensities were measured using the rheumatoid arthritic patients. Cells were plated on a cross-linked ImageJ software. fluorochrome-labeled gelatin matrix and left for 40 h at 37˚C. Real-time PCR Synoviocytes were incubated in serum-free condition in the presence or absence of TGF-b and in the presence or absence of PI3K inhibitor LY294002 (Cayman Chemical), TbR1 inhibitors LY364947 (Tocris Bio- science), or SB431542 (Sigma-Aldrich). Total RNA was isolated using the TRI-Reagent (Life Technologies) protocol, as previously described (38), and quantitative real-time PCR was performed on a Rotor-Gene 3000

(Corbett Research). Primer sequences were as follows: PDGF-A forward Downloaded from (59-CGTAGGGAGTGAGGATTCTTTG-39), PDGF-A reverse (59-GCTTC- CTCGATGCTTCTCTT-39); PDGF-B forward (59-CCATTCCCGAGGA- GCTTTATC-39), PDGF-B reverse (59-GGTCATGTTCAGGTCCAACTC-39); PDGF-C forward (59-GTCAATGTGTCCCAAGCAAAG-39), PDGF-C reverse (59-CCACGTCGGTGAGTGATTT-39); PDGF-D forward (59-GAAATTGTGGCTGTGGAACTG-39), PDGF-D reverse (59-GGCCA- GGCTCAAACTGTAATA-39); PAI-1 forward (59-AATCAGACGGCAG- CACTGTCT-39), PAI-1 reverse (59-GGCAGTTCCAGGATGTCGTAGT-39); http://www.jimmunol.org/ PI3K-a forward (59-GTATCCCGAGAAGCAGGATTTAG-39), PI3K-a reverse (59- CAGAGAGAGGATCTCGTGTAGAA-39); PI3K-b forward (59-GCACTTGGTAATCGGAGGATAG-39), PI3K-b reverse (59-TTGTA- CTGAGACAGCAGGAATG-39); PI3K-d forward (59-CCCACAGGTGA- TCCTAACATATC-39), PI3K-d reverse (59-ACTTCTGGCTCTGTTGAGTTT- 39); RPLPO forward (59-GATTACACCTTCCCACTTGC-39), RPLO reverse (59-CCAAATCCCATATCCTCGTCCG-39). Each reaction was runinduplicate,andvalueswerenormalized against the RPLPO housekeeping gene. by guest on October 2, 2021 Immunohistochemistry Paraffin-embedded synovial tissue sections were freed of paraffin and rehydrated, and immunohistochemical staining was performed according to the standard avidin–biotin immunoperoxidase complex technique by using the following Abs: anti-pPDGFRa(Tyr849)/b(Tyr857) (Cell Signaling Technology; dilution 1:75), anti-PDGFRb (Cell Signaling Technology; dilution 1:100), or rabbit isotype IgG. Diaminobenzidine was used for the detection of the labeled proteins, and the sections were counterstained with Harris hematoxylin. Slides were scanned with a Hamamatsu Nanozoomer 2.0-RS scanner. For each patient, six random fields (at original magnification 320) were captured with the NPD viewing software, and intensity of labeling in the synovial membrane was analyzed using the immunohistochemistry quantification technique described by Pham et al. (39). Images were converted to CMYK with the FIJI software. Next, gray image of the yellow (Y) channel was extracted, and chromogen intensities were analyzed using the Image Pro software (Media Cybernetics, Bethesda, MD). Results are expressed as the sum of labeling intensity (density) relative to total area. Measurement of intracellular calcium Synovial cells were incubated with 3 mM Fura-2/AM (Molecular Probes) for 30 min at room temperature in 1 mL HBSS containing 0.35 g/L FIGURE 1. RA-FLS generate ECM-degrading invadosomes. RA-FLS 488 NaHCO3 and 10 mM HEPES (pH 7.0). The dye-loaded cells were then were cultured on Oregon Green –conjugated gelatin for 4 h and stained washed and resuspended in HBSS containing 1.5 mM CaCl2, and baseline for invadopodia markers. Confocal microscopy images of the basal surface fluorescence was measured at 0 s. Fluorescence released after lysophos- of the cells showing colocalization of (A) F-actin (red) and cortactin (blue), phatidic acid (LPA) stimulation (10 mM) was recorded every 2 s for 4 min. (B) F-actin and tyrosine-phosphorylated proteins (pY) (blue), and (C) Where indicated, cells were pretreated with the PLC inhibitor U73122 488 (0.1 and 1 mM) before stimulation with LPA. Analysis was performed by F-actin and p-Src (blue) to areas of Oregon Green –conjugated gelatin spectrophotometry using the Hitachi F2500 fluorescence spectrophotom- (green) degradation. Nuclei were stained with DAPI (blue). Boxed areas eter and fluorescence solutions software. outlining invadosomes are enlarged in the corresponding right panels. The enlarged merged images also show fluorescence intensity profiles for Statistical analysis F-actin, Oregon Green488–conjugated gelatin, and the specified invadosome Comparison between two groups was analyzed using unpaired Student t marker at the cell–matrix interface. Representative images of three inde- test. One-way ANOVA was used for comparison among three or more pendent experiments generated with two independent RA-FLS lines are groups. The p values ,0.05 were considered significant. shown. Scale bars, 10 mM. Original magnification 360 (insets). The Journal of Immunology 3267

Confocal analysis of the cells indicated that they extended dot-like Increased phosphorylation of PDGFR is a distinctive feature of actin-rich degradative structures from the ventral side of the plasma RA synovial cells and tissues membrane in contact with the matrix (Fig. 1A). To confirm that the We next investigated the nature of RTKs responsible for the ag- observed structures were indeed invadosomes, we stained the cells gressive invadosome-forming phenotype. A phospho-specific RTK for invadosome markers cortactin, pY, and p-src (40, 41). Stacking Ab array was used to screen synovial cells for 28 common RTKs along the z-axis of the cell–matrix interface clearly showed that all that included EGFR, FGFR, INSR, NGFR, HGFR, PDGFR (c-kit, these components were located within the matrix in a configuration PDGFR, FLT3, CSF-1R), EphR, and Axl family members and Tie similar to that of actin, suggesting that they were located within and VEGFR. High levels of phosphorylated PDGFR were observed the cell membrane protrusions (Fig. 1). Because invadosomes are in RA synovial cells compared with cells of OA and NA indi- transient dynamic structures, some areas of gelatin degradation viduals (Fig. 3A). In contrast, there were no significant changes in were observed that were no longer associated with invadosome their levels of phosphorylation in the case of other RTKs. Western markers at the time of fixation. Together, these results indicate that blot analysis of synovial cell lysates using a phospho-specific synovial cells of arthritis patients produce typical invadosome mAb that recognized active (phosphorylated) PDGFR-ab con- structures similar to those produced by cancer cells or synovial cells firmed the RTK array results with a 2.6-fold increase in PDGFR from collagen-induced arthritis rats (5, 42, 43). phosphorylation in RA synoviocytes compared with control NA We next investigated the invadosome formation potential of cells, whereas levels of PDGFR did not change (Fig. 3B–D). synovial cell lines from two types of arthritic patients, namely OA PDGFR phosphorylation in RA cells was significantly inhibited by and RA, and compared the results with cells from NA. Results imatinib mesylate both in the phospho-RTK array and Western showed that synoviocytes of RA patients had a 3.4- and 2.8-fold

blotting (Fig. 3A–D), suggesting that the observed inhibitory ef- Downloaded from increase in the percentage of cell-forming invadosomes as com- fect of imatinib mesylate on invadosome formation by RA syno- pared with cells from NA and OA patients, respectively (Fig. 2A). vial cells was most likely due to inhibition of PDGFR kinase Quantification of the extent of extracellular matrix (ECM) deg- activity. We also examined pPDGFR expression levels in synovial radation (a measure of invadosome function), as well as the tissues. Results showed that pPDGFR is strongly expressed in the number of invadosomal structures formed per cell, identified by rheumatoid synovium with prominent staining in the synovial the colocalization of actin and phospho-cortactin, two known intimal lining (Fig. 3E). Quantification of staining intensity indi- http://www.jimmunol.org/ markers of invadosomes (44), showed a significant increase in cated that pPDGFR expression is significantly elevated (p , 0.05) both invadosome formation and function in RA synoviocytes in RA compared with OA (Fig. 3F). Together, these results indi- compared with NA or OA cells (Fig. 2B, 2C). Of note, there were cate that PDGFR could be part of a dominant RTK signaling no statistically significant differences between the OA and NA pathway involved in ECM degradation by RA synoviocytes. groups of synovial cells, indicating that, in contrast to RA cells, OA synovial cells displayed a nonaggressive invadosome-forming Signaling through PDGFR is required for the enhanced phenotype. invadosome biogenesis observed in RA synovial cells The involvement of RTK in invadosome formation of RA syn- To investigate whether endogenous PDGF and PDGFR activity

oviocytes was investigated by culturing cells in the presence or ab- were required for invadosome formation in RA synovial cells, we by guest on October 2, 2021 sence of imatinib mesylate, a RTK inhibitor that possesses strong ﬁrst tested whether PDGF promotes invadosome formation in con- activity against PDGFR and c-Kit and weaker activity against trol nonarthritic cells. Results showed that addition of PDGF-B, FLT3 and CSF-1R. Results showed that inhibition of RTK activ- which binds PDGFR a and b receptors, induced a concentration- ity resulted in a concentration-dependent decrease in invadosome dependent increase of dot-like actin-rich invasive structures formation by RA synoviocytes, reaching basal levels of OA and NA (Fig. 4A, associated micrograph panel) similar to those observed control cells (Fig. 2D). These results led to the conclusion that RTK in RA synoviocytes (Fig. 1A–C). In contrast, PDGF-B had no ef- were involved in invadosome formation by RA synovial cells. fect on the capacity of OA cells to form invadosomes and did not

FIGURE 2. RA-FLS display increased invadosome-forming ability. (A and B) Synoviocytes of three NA, four OA, and four RA individuals were plated on Oregon Green488–conjugated gelatin- coated coverslips and incubated for 48 h. (A) Per- centage of invadosome-forming cells was calculated for 300 cells in four independent experiments. Each column represents the results of individual cell lines. (B) The mean area of degradation was calculated for 20 cells in three independent experiments. (C) Synoviocytes were incubated on Oregon Green488 gelatin-coated coverslips for 6 h and stained for phospho-cortactin and actin. The number of phospho-cortactin/actin clusters was counted for 25 cells in three independent experiments. (D) Synovial cells were plated on Oregon Green488 gelatin-coated coverslips and incubated in the presence or absence of imatinib mesylate, used at the indicated concentrations, for 48 h (n = 4 independent experiments). *p , 0.05, **p , 0.01, ***p , 0.001. 3268 PDGFR ACTIVATION PROMOTES INVADOSOME FORMATION IN RA-FLS Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 3. Phosphorylation of PDGFR is increased in RA synovial cells and tissue sections. Synovial cells of five controls (Ctl) (three NA and two OA) and five RA individuals were incubated in serum-free medium for 16 h. Four RA-FLS lines were incubated in the presence or absence of imatinib mesylate (3 mM). (A) Representative images of cell lysates hybridized to a human phospho-RTK array. Hybridization signals at the corners and within the middle of the strip served as internal controls. Dots within squared boxes represent phospho-PDGFR. The associated graph shows mean phospho-PDGFR fluorescence intensities calculated for each cell lysate. (B) Cell lysates were analyzed by immunoblotting with the indicated Abs. The associated graphs show (C) pPDGFRab/PDGFRb ratio or (D) PDGFRb/a tubulin ratio of four independent experiments. (E) Representative images (original magnification 340) of PDGFRb, pPDGFRab immunostaining, and control rabbit IgG of synovial tissues of OA and RA patients. (F) The associated graph shows relative labeling intensities in arbitrary units (au) for 11 OA and 5 RA patients. *p , 0.05, **p , 0.01. increase further the number of invadosomes in RA cells (Fig. 4A). Consistent with previous studies (47, 48), the action of PDGF-B We then blocked either PDGFR activity using PDGFR tyrosine on nonarthritic synovial cells results in a rapid phosphorylation kinase inhibitor V, which inhibits ligand-induced PDGFR phos- of AKT(S476), further suggesting that this pathway is directly phorylation and kinase activity (45), or endogenous PDGFR li- triggered by PDGFR activation in these cells (Fig. 5A). In gands using PDGF-B and pan-PDGF neutralizing Abs. Results addition, treatment of RA synovial cells or PDGF-stimulated showed that invadosome formation induced by PDGF-B in control nonarthritic cells with inhibitors of PI3K (LY294002) or AKT synovial cells or already elevated in RA cells was abolished by the (AKT inhibitor XI) resulted in a concentration-dependent in- addition of PDGF neutralizing Abs or PDGFR tyrosine kinase hibition of invadosome formation. In contrast, inhibition of inhibitor V (Fig. 4B, 4C). In contrast, an isotype-matched negative PLCg (U73122), a downstream mediator of PDGFR signaling control Ab has no effect. We therefore concluded that activation of (49), had no effect (Fig. 5B). The efficiency of the PLC in- PDGFR in RA synoviocytes triggers signals that result in inhibitor in synoviocytes was confirmed by its ability to block creased invadopodia formation. calcium mobilization induced by LPA (Fig. 5C). Furthermore, knockdown of the class I PI3K family catalytic subunits known PI3K/AKT pathway is an effector of PDGF-driven invadosome to be expressed in human synoviocytes (47) indicated that formation in RA synovial cells PI3Ka and PI3Kd wereselectivelyinvolvedininvadosome PDGFR activation triggers the recruitment of signaling proteins formation in RA cells (Fig. 5D, 5E). These results suggest that that contain Src homology 2, Src homology 3, and pTyr-binding activation of the PI3K-AKT pathway is required for PDGFR- domains and that lead to the activation of several pathways with induced invadosome biogenesis in RA synoviocytes and fur- the PI3K-AKT pathway being a major downstream effector ther support the role of this pathway in synoviocyte invasive of both PDGFR signaling and invadosome formation (43, 46). capabilities (48). The Journal of Immunology 3269 Downloaded from http://www.jimmunol.org/ FIGURE 4. PDGF activity regulates invadosome formation by RA synoviocytes. (A–C) Synovial cells of three NA, two OA, and four RA individuals were plated on Oregon Green488–conjugated gelatin-coated coverslips and incubated for 48 h in the presence or absence of (A) PDGF-BB used at the indicated concentrations (n = 3 independent experiments). Associated micrographs show representative invadosome formation (stained for F-actin [red] and DAPI [blue]) of control and PDGF-BB–treated cells. (B) PDGF-pan–specific neutralizing Ab, PDGF-BB–specific neutralizing Ab, or an isotype-matched IgG used at saturating concentrations (20 mg/ml); NA synoviocytes were also incubated in the presence or absence of PDGF-BB (n = 4 independent experiments). Scale bars, 10 mM. (C) PDGFR tyrosine kinase inhibitor V used at the indicated concentrations (n = 4 independent experiments). Graphs represent mean + SEM of the percentage of invadosome-forming cells. **p , 0.01, ***p , 0.001.

Relationship between TGF-b and PDGF in invadosome significantly increased in RA synoviocytes compared with non- by guest on October 2, 2021 formation in RA synovial cells arthritic cells (Fig. 6C). Of interest, PDGF-B mRNA expression We have recently reported that autocrine TGF-b production was was significantly increased in high-invadosome–producing RA part of the mechanism that explained the aggressive invadosome- cell lines, but not in RA cell lines, with low production of inva- forming phenotype of rat synoviocytes derived from collagen- dosomes. In contrast, there was an absence of discriminatory induced arthritis joints (6). TGF-b has also been shown to promote effect on PDGF-C mRNA expression (Fig. 6D). Furthermore, proliferation of smooth muscle cells and to induce epithelial– incubation of RA synoviocytes with TGF-b1–3 neutralizing Abs mesenchymal transition in mammary epithelial cells through a resulted in a strong decrease of PDGF-B mRNA expression, PDGF-dependent mechanism (50, 51). These observations led whereas mRNA expression of PDGF-A, -C, and -D was not af- us to investigate the relationship between TGF-b and PDGFR- fected (Fig. 6E). These data demonstrate that TGF-b contributes mediated invadosome formation in RA cells. As observed in to PDGFR-induced invadosome formation in RA synoviocytes, the case of rat synovial cells (6), addition of TGF-b1 to human through upregulation of PDGFR ligand, PDGF-B. synoviocytes increased invadosome formation in a dose- Previous reports have shown that TGF-b can modulate mRNA dependent fashion (Fig. 6A). Addition of a neutralizing anti– expression of PDGF ligands in macrophages and endothelial and TGF-b1–3 Ab significantly decreased the high incidence of glioma cells (52–54). To obtain further evidence that TGF-b in- invadosome formation in RA cells (Fig. 6A), indicating that duced PDGF-B ligand expression in synovial cells, we analyzed autocrine TGF-b production also drives invadosome formation in mRNA expression of PDGF-B in nonarthritic cells stimulated with these cells. To examine the contribution of PDGF to TGF-b– TGF-b. PDGF-A, -C, and -D mRNA expression were assessed for induced invadosome formation, TGF-b1–stimulated control syno- comparison. Addition of TGF-b led to a strong and significant viocytes were incubated in the presence or absence of anti–PDGF-BB upregulation of PDGF-B, whereas only moderate changes in or pan-PDGF neutralizing Abs. Invadosome formation was simi- PDGF-A, PDGF-C, and PDGF-D were observed (Fig. 7A). This larly diminished upon PDGF neutralization by either of the two was associated with the ability of TGF-b to increase the phos- Abs, supporting a role for autocrine PDGF-B in invadosome for- phorylation of the PDGFR (Fig. 7B), suggesting production of mation triggered by TGF-b (Fig. 6). functional PDGF ligands. Previous reports have shown that Smads Data described above suggested that TGF-b was part of an bind and transactivate the PDGF-B promoter in macrophages, activation loop in RA synoviocytes that contributed to invadosome endothelial cells, and gliomas (52–54), and PI3K/Akt pathways formation through regulation of expression of PDGFR ligands. were shown to participate in invadosome production by synovial This possibility was first tested by analysis of mRNA expression cells (Fig. 5A, 5B) (6). To determine whether the Smad and/or of PDGF-A–D in RA cells compared with nonarthritic cells. Re- PI3K/Akt pathways were involved in the expression of PDGF-B in sults showed that only mRNA levels of PDGF-B and C were synoviocytes, we treated the cells with the TbR1/ALK5 inhibitors, 3270 PDGFR ACTIVATION PROMOTES INVADOSOME FORMATION IN RA-FLS Downloaded from http://www.jimmunol.org/

FIGURE 5. PI3K/AKT as effectors of invadosome formation in RA synovial cells. (A) NA synovial cells were serum starved for 16 h and incubated with PDGF-BB (10 ng/ml) for 0–30 min. Total proteins were extracted and analyzed by Western blot analysis for pPDGFRab (a[Tyr849]/b[Tyr857]), PDGFRb, pAkt(S476), and AKT. A representative blot of three independent experiments using three independent NA cell lines is shown. (B) Nonarthritic synovial cells 488

stimulated with PDGF-BB (25 ng/ml) and RA synovial cells were plated on Oregon Green –conjugated gelatin-coated coverslips and incubated for 48 h by guest on October 2, 2021 in the presence or absence of PI3K inhibitor LY294002, the AKT inhibitor XI, the PLC inhibitor U73122, or the inactive U73343 analog used at the indicated concentrations (n = 3–4 independent experiments using 7 independent NA cell lines and 5 independent RA cell lines). Graphs represent mean + SEM of the percentage of invadosome-forming cell. (C) Graph showing the ability of U73122 to block LPA-induced calcium mobilization in NA synoviocytes. LPA was used at 10 mM. (D and E) RA synoviocytes were transduced with lentivirus-expressing short hairpin RNAs targeting individual PI3Ka (shp110a), PI3Kb (shp110b), and PI3Kd (shp110d), and (D) expression of the PI3K isoforms was determined by RT-PCR or (E) cells were submitted to the invadosome assay (n = 4 independent experiments with 4 independent RA cell lines). *p , 0.05, **p , 0.01, ***p , 0.001.

LY364947 (55), or SB-431542, a TbR1 inhibitor that was also tyrosine kinase c-Src and the autocrine transglutaminase-2/TGF-b shown to block the activation and nuclear translocation of the loop were previously identified as part of the mechanism that Smads (56–58) or the PI3K inhibitor LY294002 and assessed triggered invadosome formation in arthritic FLS. In this work, we the induction of PDGF-B mRNA by TGFb. PAI-1, a Smad2/3- show that, among the main RTKs known to be expressed in RA, regulated gene, was used as an internal control. Results indi- the RTK PDGFR, through activation of the PI3K/AKT pathway, cated that pharmacological suppression of the TbR1/Smad or the also participates in promoting invadosome formation and matrix PI3K pathway impeded the induction of PDGF-B mRNA ex- degradation in synovial cells from RA patients. Our data reveal an pression by TGF-b, whereas, as expected, PAI-1 induction was association between activation of the RTK PDGFR and TGF-b for selectively blocked by the TbR1/Smad inhibitors (Fig. 7C, 7D). In invadosome formation in RA-FLS that involves induction of addition, knockdown of p110a, p110b, or p110d indicates that PDGF-B by TGF-b, an event under the control of both TbR1/ P13Ka isoform was selectively involved in PDGF-B production Smad and PI3K pathways. These findings identify PDGFR as an by RA synoviocytes (Fig. 7E). These results suggest that PDGF-B important RTK required for the aggressive invadosome-forming induction in synoviocytes was dependent on both TbR1/Smad and phenotype of RA synovial cells and suggest the participation of an PI3Ka activities. overactive TGF-b/PDGFR pathway during cartilage degradation processes. Discussion One central finding of the current studies is the potent role of Among the cells involved in RA pathogenesis, FLS are recognized PDGFR in invadosome formation in RA synoviocytes. PDGFRa as the primary effectors of cartilage degradation. These cells pos- and PDGFRb, and their ligands PDGFs A–D, have previously sess strong invasive properties and produce extensive quantities of been implicated in RA-FLS aggressiveness. Increased levels of ECM-degrading enzymes. We have previously shown that this PDGFRa and PDGFR ligands were detected in FLS cultures of aggressive behavior is largely dependent on their ability to form RA patients (59, 60), and PDGFRab was found to be expressed in the actin-rich structures invadosomes, which concentrate matrix stromal cells in the synovial lining, and in smooth muscle cells metalloproteinases to focal sites of matrix degradation (5, 6). The and capillary cells in RA synovium (14). Furthermore, PDGFs The Journal of Immunology 3271

FIGURE 6. Involvement of autocrine TGF-b in PDGFR-induced invadosome formation in RA synovial cells. (A and B) Synovial cells of NA and RA individuals were plated on Oregon Green488–conjugated gelatin-coated coverslips and incubated for 48 h in the presence or absence of the following reagents: NA synoviocytes were incubated in the presence or absence of TGF-b used at the indicated concentrations, TGF-b1–3–neutralizing Ab, PDGF pan-specific neutralizing Ab, PDGF- BB–specific neutralizing Ab, or an isotype- matched IgG. RA synoviocytes were incubated in the presence or absence of TGF-b1–3– neutralizing Ab or an isotype-matched IgG (n = 4 independent experiments using 4 NA and 4 RA cell lines). Graph represents mean + SEM of the percentage of invadosome-forming cells. (C–E) RT-PCR quantification of PDGF Downloaded from mRNA expression in synovial cells of four NA and/or six RA individuals. (C)PDGF-A– D mRNA expression levels. (D)PDGF-Band PDGF-C expression in RA synoviocytes expressing low (,2-fold increase compared with control synoviocytes, n = 3 independent http://www.jimmunol.org/ experiments) and high (.2-fold increase compared with control synoviocytes, n =4 independent experiments) invadosome-forming capability. (E) PDGF-A–D expression in three independent RA FLS lines previously exposed to TGF-b1–3–neutralizing Ab or matching IgG used at 20 mg/mlfor16h(n = 3 independent experiments). Graphs show the mean + SEM. *p , 0.05, **p , 0.01, ***p , 0.001. by guest on October 2, 2021

have been found to regulate synoviocyte proliferation, invasion, overexpressed PDGF-B and that the invadosome-forming pheno- anchorage-independent growth, and collagenase transcription (61– type of RA cells was inhibited using PDGF-BB–specific or PAN- 63). Our finding that PDGFR activation/phosphorylation was a PDGF– neutralizing Abs suggest direct phosphorylation of main trigger of signaling events that led to invadosome formation PDGFR as a result of ligation by autocrine PDGF-B. Such in RA-FLS adds an original facet to the role of PDGF/PDGFR in autoregulatory loop most likely involves TGF-b expression and synovial aggressiveness that is directly related to the key effector activity because PDGF-B production was blocked by TGF-b functions of these cells in cartilage degradation. neutralization in RA cells. Furthermore, addition of TGF-b to Results described in this work showed that PDGFRab phos- synovial cells increased PDGF-B expression as well as PDGFR phorylation was specifically increased in FLS of RA patients and phosphorylation/activation and thereby recapitulated the RA strongly elevated in the lining layer of RA synovium. It can be phenotype. Second, we have reported that the active form of Src argued that mechanisms responsible for PDGFR phosphorylation was elevated in synovial tissues and FLS of collagen-induced can be direct, that is PDGF driven or indirect, which would not arthritis rats and that c-Src activity mediated invadosome forma- involve PDGFs. Direct activation involves PDGF ligand-induced tion in arthritic cells (5). Data indicating that Src inhibition effi- PDGFR dimerization, leading to a robust activation of RTK ac- ciently blocked invadosome formation in arthritic FLS (5), along tivity, and appears to be a main mechanism responsible for with the known association of PDGFR and Src family kinases and PDGFR phosphorylation. In contrast, reactive oxygen species, its site-specific phosphorylation by c-Src (67, 68), raised the which could be induced by growth factors outside of the PDGF possibility that activation of Src kinases in arthritic FLS could also family, can activate Src family kinases that have been reported to contribute to PDGFR phosphorylation. act as intracellular mediators of PDGFR phosphorylation (49). In TGF-b and PDGFs are prime cytokines involved in the induc- arthritic synoviocytes, evidence suggests that at least two path- tion of an aggressive behavior by synovial cells (10). Based on ways may be involved in PDGFR phosphorylation and activation. a TGF-b neutralization approach, autocrine TGF-b production First, RA or growth factor–activated synoviocytes were shown to contributed to PDGFR-induced invadosome formation in RA pa- secrete PDGF ligands and to express the intracellular (e.g., furin) tients through PDGF-B production. In accordance with previous or extracellular (e.g., tPA, uPA) proteases required for their acti- reports that TGF-b induced PDGF ligands in various cell types vation (59, 60, 64–66). In this study, the findings that RA FLS (69–71), including synovial cells (72), PDGF-B and PDGF-C 3272 PDGFR ACTIVATION PROMOTES INVADOSOME FORMATION IN RA-FLS

FIGURE 7. Implication of the TbR1/Smad pathway in PDGF-B expression induced by TGF-b.(A and B) Four NA synoviocyte lines were serum starved and incubated in the presence or absence of TGF-b (5 ng/ml) in serum- free condition for 6 or 24 h. (A)RT-PCR quantiﬁcation of PDGF-A–D mRNA expression (n =3).(B) Western blot analysis of pPDGFRab (a[Tyr849]/b[Tyr857]), PDGFRb, pAkt (S476), and AKT. Representative blot of three independent experiments. (C and D)RT-PCR quantiﬁcation of (C) PDGF-B and (D) PAI-1 mRNA expression in NA cells incubated for 6 h with TGF-b (5 mg/ml) in the presence or absence of PI3K inhibitor LY294002 (10 mM), TbR1 inhibitors LY364947 (1 mM), or SB431542

(1 mM) (n = 3 independent experiments). Downloaded from (E) RA synoviocyte lines from four patients were transduced with lentivirus- expressing short hairpin RNAs targeting individual PI3Ka (shp110a), PI3Kb (shp110b), and PI3Kd (shp110d), and expression of PDGF-B was deter- minedbyRT-PCR(n = 3 indepen- http://www.jimmunol.org/ dent experiments). Graphs represent the mean + SEM. *p , 0.05, **p , 0.01, ***p , 0.001.

mRNA were found overexpressed in RA synoviocytes, whereas know which isoforms contributed to invadosome formation. PDGF-B was strongly induced by TGF-b in nonarthritic cells. Our Knockdown of class I PI3Ka, PI3Kb, and PI3Kd isoforms, which results are consistent with a recent report that PDGF-B and TGF-b are known to be expressed in human synoviocytes (47), indicated by guest on October 2, 2021 synergized to strongly potentiate the response of FLS to cytokines, that both PI3Ka and PI3Kd were selectively involved in invado- leading to the development of an aggressive proinflammatory some formation in RA cells. In contrast, only PI3Ka participated phenotype (10). in autocrine PDGF-B production. These observations are consis- It is of note that OA synoviocytes were not able to form inva- tent with previous findings indicating a role for PI3Kd in RA dosomes upon PDGF stimulation, whereas an efficient response synoviocyte invasiveness and joint destruction (47). They also was observed in synoviocytes from nonarthritic patients (Fig. 4A). describe a new function for PI3Ka in growth factor gene ex- The exact reason for the lack of PDGF response in OA cells is pression. PI3Ka is an oncogene frequently mutated in human unknown but may involve, among others, alterations in the TGFb/ cancer and has been shown to play a role in tumor progression PDGF pathway in OA. Among the candidate genes that have been (83). New findings that the PI3Ka isoform is also involved in reported as harboring risk alleles for OA are GDF5, a growth factor PDGF-mediated prodestructive functions of RA synovial cells that is part of the TGF-b superfamily, and SMAD3, which also suggest that dual inhibition PI3Ka and PI3Kd may protect from operates in the TGF-b pathway (73, 74). Both GDF5 and SMAD3 cartilage damage by modulating the overactive prodestructive are involved in synovial joint development, maintenance, and repair, functions of synoviocytes. and may thereby be involved in invadosome formation (75). Another interesting finding of our investigation was the fact that Formation of invadopodia in cancer cells or podosomes in Src- both PI3K and Smad activity were required in the regulation of transformed fibroblasts requires activation of Smads and PI3K PDGF-B expression by TGF-b. Although TGF-b is known to (76–79). In this work, we showed that pharmacological inhibition activate PI3K in a number of cell types, including synovial cells of PI3K blocked invadosome-mediated ECM degradation in RA (Fig. 7B), the PI3K/AKT pathway was shown to antagonize synoviocytes or in cells stimulated with PDGF. We also showed Smad-mediated effects in most TGF-b–induced responses. This that PI3K and Smad activation was required for PDGF-B mRNA effect was due to the ability of AKT to interfere with phosphor- expression. PI3Ks comprise a family of signaling molecules that ylation and nuclear localization of Smad3 (84, 85) or to counteract have been involved in a wide range of cellular functions, including the activity of the Smad3/FoxO nuclear complexes (86, 87). In this cell growth, apoptosis, survival, and migration (80). Among them, work, findings that inhibition of PI3K or TbR1/Smad pathway in class I PI3K isoforms (PI3Ka, PI3Kb, PI3Kϒ, and PI3Kd) are the synoviocytes blocked PDGF-B mRNA expression induced by best studied and are emerging drug targets in the treatment of TGF-b suggest positive cooperation between both signaling arthritis (23297340). Mice lacking PI3Kd were shown to be re- pathways. Cross-talk between Smad and PI3K pathways remains sistant to the development of experimental arthritis (47, 81), and largely unexplored. In the late phase of tumor progression, en- PI3Kϒ blockade by both genetic and pharmacological approaches hancement of the PI3K/AKT pathway has been shown to switch suppressed joint inflammation and cartilage damage in various TGF-b into a tumor-promoting signal through inactivation of mice models of RA (47, 81, 82). It was therefore of interest to glycogen synthase kinase GSK3b that promotes Smad3 stabili- The Journal of Immunology 3273 zation (88, 89). Overexpressed Smad3 can in turn transactivate the transforming growth factor-beta s and transforming growth factor-beta type II receptor in the synovial tissues of patients with rheumatoid arthritis. Lab. Invest. PDGF-B promoter (53), suggesting that one possible way PI3K 70: 620-630. and Smads positively cooperate for PDGF-B expression in syno- 9. Cheon, H., S. J. Yu, D. H. Yoo, I. J. Chae, G. G. Song, and J. Sohn. 2002. In- viocytes is through an increase in Smad3 protein stability. Inter- creased expression of pro-inflammatory cytokines and metalloproteinase-1 by TGF-beta1 in synovial fibroblasts from rheumatoid arthritis and normal indi- estingly, Smad3, but not Smad2, seems to be the primary target of viduals. Clin. Exp. Immunol. 127: 547–552. regulation by PI3K/Akt. This observation was consistent with our 10. Rosengren, S., M. Corr, and D. L. Boyle. 2010. Platelet-derived growth factor previous findings of the selective role of Smad3 in invadopodia and transforming growth factor beta synergistically potentiate inflammatory mediator synthesis by fibroblast-like synoviocytes. Arthritis Res. Ther. 12: R65. formation mediated by TGF-b (79). Although further studies are 11. Lemmon, M. A., and J. Schlessinger. 2010. Cell signaling by receptor tyrosine needed to delineate the exact downstream targets of PI3K and kinases. Cell 141: 1117–1134. TbR1/Smad activities and how these pathways interact with each 12. Yamane, S., S. Ishida, Y. Hanamoto, K. Kumagai, R. Masuda, K. Tanaka, N. Shiobara, N. Yamane, T. Mori, T. Juji, et al. 2008. Proinflammatory role of other, our results clearly suggest that both pathways play an es- amphiregulin, an epidermal growth factor family member whose expression is sential role in the development of the prodestructive function of augmented in rheumatoid arthritis patients. J. Inflamm. 5: 5. 13. Hallbeck, A. L., T. M. Walz, K. Briheim, and A. Wasteson. 2005. TGF-alpha and RA synoviocytes. ErbB2 production in synovial joint tissue: increased expression in arthritic joints. Results of clinical trials have shown that current therapies for the Scand. J. Rheumatol. 34: 204–211. treatment of RA have only a moderate effect on joint destruction 14. Reuterdahl, C., A. Tingstrom, L. Terracio, K. Funa, C. H. Heldin, and K. Rubin. 1991. Characterization of platelet-derived growth factor beta-receptor expressing (90–93). Although encouraging results have been reported using a cells in the vasculature of human rheumatoid synovium. Lab. Invest. 64: 321– combination of the classical drug, methotrexate, with new bio- 329. logical treatments, such as TNF inhibitors, these new therapies are 15. Remmers, E. F., H. Sano, and R. L. Wilder. 1991. Platelet-derived growth factors and heparin-binding (fibroblast) growth factors in the synovial tissue pathology only effective in a subset of patients, indicating a need for alter- of rheumatoid arthritis. Semin. Arthritis Rheum. 21: 191–199. Downloaded from native treatments (94). Interestingly, TNF resistance has recently 16. Paavonen, K., J. Mandelin, T. Partanen, L. Jussila, T. F. Li, A. Ristimaki, been associated with a distinct synovial fibroid phenotype K. Alitalo, and Y. T. Konttinen. 2002. Vascular endothelial growth factors C and D and their VEGFR-2 and 3 receptors in blood and lymphatic vessels in healthy enriched for gene sets associated with TGF-b/SMAD signaling, as and arthritic synovium. J. Rheumatol. 29: 39–45. well as cell projection processes (95). In this work, we show that 17.Feuerherm,A.J.,M.Borset,C.Seidel,A.Sundan,L.Leistad,M.Ostensen, blockade of the TbR1/PDGFR axis or downstream PI3K-AKT and A. Faxvaag. 2001. Elevated levels of osteoprotegerin (OPG) and hepatocyte growth factor (HGF) in rheumatoid arthritis. Scand. J. Rheumatol. 30: pathway impaired invadosome formation, a process previously 229–234. linked to cartilage damage (5, 6). Various small-molecule inhibi- 18. Byrd, V., X. M. Zhao, W. L. McKeehan, G. G. Miller, and J. W. Thomas. 1996. http://www.jimmunol.org/ Expression and functional expansion of fibroblast growth factor receptor T cells tors of TGF-b are being evaluated in preclinical and clinical trials in rheumatoid synovium and peripheral blood of patients with rheumatoid ar- (96). Furthermore, several inhibitors that target PI3Ks are cur- thritis. Arthritis Rheum. 39: 914–922. rently being tested in clinical trials for treatment of human cancers 19. Ceponis, A., Y. T. Konttinen, M. Takagi, J. W. Xu, T. Sorsa, M. Matucci-Cerinic, S. Santavirta, H. C. Bankl, and P. Valent. 1998. Expression of stem cell factor (97, 98) and are under consideration for the treatment of RA (99). (SCF) and SCF receptor (c-kit) in synovial membrane in arthritis: correlation Although further investigation of the interplay between the TGF-b with synovial mast cell hyperplasia and inflammation. J. Rheumatol. 25: 2304– and PDGF pathways and downstream PI3K signaling pathway is 2314. 20. Paniagua, R. T., O. Sharpe, P. P. Ho, S. M. Chan, A. Chang, J. P. Higgins, needed to define the most promising target(s), the potential action B. H. Tomooka, F. M. Thomas, J. J. Song, S. B. Goodman, et al. 2006. Selective of these inhibitors on both TGF-b pathways and PDGF-B in- tyrosine kinase inhibition by imatinib mesylate for the treatment of autoimmune by guest on October 2, 2021 duction highlights their therapeutic potential for treatment of joint arthritis. J. Clin. Invest. 116: 2633–2642. 21. Ando, W., J. Hashimoto, A. Nampei, H. Tsuboi, K. Tateishi, T. Ono, destruction in RA. N. Nakamura, T. Ochi, and H. Yoshikawa. 2006. Imatinib mesylate inhibits osteoclastogenesis and joint destruction in rats with collagen-induced arthritis (CIA). J. Bone Miner. Metab. 24: 274–282. Acknowledgments 22. Koyama, K., K. Hatsushika, T. Ando, M. Sakuma, M. Wako, R. Kato, H. Haro, We thank Dr. Gilles Dupuis for critical reading of the manuscript and H. Sugiyama, Y. Hamada, H. Ogawa, and A. Nakao. 2007. Imatinib mesylate helpful comments. We also thank Dr. Leónid Volkov for expert assistance both prevents and treats the arthritis induced by type II collagen antibody in mice. Mod. Rheumatol. 17: 306–310. with confocal microscopy. 23. Eklund, K. K., K. Lindstedt, C. Sandler, P. T. Kovanen, L. Laasonen, A. Juurikivi, H. Wolff, M. Mykka¨nen, and H. Joensuu. 2008. Maintained efficacy of the tyrosine kinase inhibitor imatinib mesylate in a patient with rheumatoid Disclosures arthritis. J. Clin. Rheumatol. 14: 294–296. The authors have no financial conflicts of interest. 24. Eklund, K. K., and H. Joensuu. 2003. Treatment of rheumatoid arthritis with imatinib mesylate: clinical improvement in three refractory cases. Ann. Med. 35: 362–367. References 25. Miyachi, K., A. Ihara, R. W. Hankins, R. Murai, S. Maehiro, and H. Miyashita. 2003. Efficacy of imatinib mesylate (STI571) treatment for a patient with 1. Hamilton, J. A. 1983. Hypothesis: in vitro evidence for the invasive and tumor- rheumatoid arthritis developing chronic myelogenous leukemia. Clin. Rheuma- like properties of the rheumatoid pannus. J. Rheumatol. 10: 845–851. tol. 22: 329–332. 2. Bartok, B., and G. S. Firestein. 2010. Fibroblast-like synoviocytes: key effector 26. Pereira, I., S. Fialho, G. Castro, and A. Zimmermann. 2010. Imatinib mesylate cells in rheumatoid arthritis. Immunol. Rev. 233: 233–255. induces clinical remission in rheumatoid arthritis. Joint Bone Spine 77: 372–373. 3. Noss, E. H., and M. B. Brenner. 2008. The role and therapeutic implications of 27. Druker, B. J., F. Guilhot, S. G. O’Brien, I. Gathmann, H. Kantarjian, fibroblast-like synoviocytes in inflammation and cartilage erosion in rheumatoid N. Gattermann, M. W. Deininger, R. T. Silver, J. M. Goldman, R. M. Stone, et al; arthritis. Immunol. Rev. 223: 252–270. IRIS Investigators. 2006. Five-year follow-up of patients receiving imatinib for 4. Muller-Ladner,€ U., J. Kriegsmann, B. N. Franklin, S. Matsumoto, T. Geiler, chronic myeloid leukemia. N. Engl. J. Med. 355: 2408–2417. R. E. Gay, and S. Gay. 1996. Synovial fibroblasts of patients with rheumatoid 28. Eckert, M. A., T. M. Lwin, A. T. Chang, J. Kim, E. Danis, L. Ohno-Machado, arthritis attach to and invade normal human cartilage when engrafted into SCID and J. Yang. 2011. Twist1-induced invadopodia formation promotes tumor me- mice. Am. J. Pathol. 149: 1607–1615. tastasis. Cancer Cell 19: 372–386. 5. Lauzier, A., M. Charbonneau, K. Harper, M. Jilaveanu-Pelmus, and 29. Rajadurai, C. V., S. Havrylov, K. Zaoui, R. Vaillancourt, M. Stuible, C. M. Dubois. 2011. Formation of invadopodia-like structures by synovial cells M. Naujokas, D. Zuo, M. L. Tremblay, and M. Park. 2012. Met receptor tyrosine promotes cartilage breakdown in collagen-induced arthritis: involvement of the kinase signals through a cortactin-Gab1 scaffold complex, to mediate invado- protein tyrosine kinase Src. Arthritis Rheum. 63: 1591–1602. podia. J. Cell Sci. 125: 2940–2953. 6. Lauzier, A., M. Charbonneau, M. Paquette, K. Harper, and C. M. Dubois. 2012. 30. Kimura, F., K. Iwaya, T. Kawaguchi, H. Kaise, K. Yamada, K. Mukai, Transglutaminase 2 cross-linking activity is linked to invadopodia formation and O. Matsubara, N. Ikeda, and N. Kohno. 2010. Epidermal growth factor- cartilage breakdown in arthritis. Arthritis Res. Ther. 14: R159. dependent enhancement of invasiveness of squamous cell carcinoma of the 7. Fava, R., N. Olsen, J. Keski-Oja, H. Moses, and T. Pincus. 1989. Active and breast. Cancer Sci. 101: 1133–1140. latent forms of transforming growth factor beta activity in synovial effusions. 31. Mader, C. C., M. Oser, M. A. Magalhaes, J. J. Bravo-Cordero, J. Condeelis, J. Exp. Med. 169: 291–296. A. J. Koleske, and H. Gil-Henn. 2011. An EGFR-Src-Arg-cortactin pathway 8. Taketazu, F., M. Kato, A. Gobl, H. Ichijo, P. ten Dijke, J. Itoh, M. Kyogoku, J. mediates functional maturation of invadopodia and breast cancer cell invasion. Ronnelid, K. Miyazono, C. H. Heldin, et al. 1994. Enhanced expression of Cancer Res. 71: 1730–1741. 3274 PDGFR ACTIVATION PROMOTES INVADOSOME FORMATION IN RA-FLS

32. Arnett, F. C., S. M. Edworthy, D. A. Bloch, D. J. McShane, J. F. Fries, 57. Hjelmeland, M. D., A. B. Hjelmeland, S. Sathornsumetee, E. D. Reese, N. S. Cooper, L. A. Healey, S. R. Kaplan, M. H. Liang, H. S. Luthra, et al. 1988. M. H. Herbstreith, N. J. Laping, H. S. Friedman, D. D. Bigner, X. F. Wang, and The American Rheumatism Association 1987 revised criteria for the classifi- J. N. Rich. 2004. SB-431542, a small molecule transforming growth factor-beta- cation of rheumatoid arthritis. Arthritis Rheum. 31: 315–324. receptor antagonist, inhibits human glioma cell line proliferation and motility. 33. Aletaha, D., T. Neogi, A. J. Silman, J. Funovits, D. T. Felson, C. O. Bingham, III, Mol. Cancer Ther. 3: 737–745. N. S. Birnbaum, G. R. Burmester, V. P. Bykerk, M. D. Cohen, et al. 2010. 2010 58. Laping, N. J., E. Grygielko, A. Mathur, S. Butter, J. Bomberger, C. Tweed, Rheumatoid arthritis classification criteria: an American College of W. Martin, J. Fornwald, R. Lehr, J. Harling, et al. 2002. Inhibition of trans- Rheumatology/European League Against Rheumatism collaborative initiative. forming growth factor (TGF)-beta1-induced extracellular matrix with a novel Arthritis Rheum. 62: 2569–2581. inhibitor of the TGF-beta type I receptor kinase activity: SB-431542. Mol. 34. Altman, R., E. Asch, D. Bloch, G. Bole, D. Borenstein, K. Brandt, W. Christy, Pharmacol. 62: 58–64. T. D. Cooke, R. Greenwald, M. Hochberg, et al; Diagnostic and Therapeutic 59. Pohlers, D., R. Huber, B. Ukena, and R. W. Kinne. 2006. Expression of platelet- Criteria Committee of the American Rheumatism Association. 1986. Develop- derived growth factors C and D in the synovial membrane of patients with ment of criteria for the classification and reporting of osteoarthritis: classification rheumatoid arthritis and osteoarthritis. Arthritis Rheum. 54: 788–794. of osteoarthritis of the knee. Arthritis Rheum. 29: 1039–1049. 60. Watanabe, N., K. Ando, S. Yoshida, S. Inuzuka, M. Kobayashi, N. Matsui, and 35. Baker, D. G., J. M. Dayer, M. Roelke, H. R. Schumacher, and S. M. Krane. 1983. T. Okamoto. 2002. Gene expression profile analysis of rheumatoid synovial fi- Rheumatoid synovial cell morphologic changes induced by a mononuclear cell broblast cultures revealing the overexpression of genes responsible for tumor- factor in culture. Arthritis Rheum. 26: 8–14. like growth of rheumatoid synovium. Biochem. Biophys. Res. Commun. 294: 36. Baldassarre, M., I. Ayala, G. Beznoussenko, G. Giacchetti, L. M. Machesky, 1121–1129. A. Luini, and R. Buccione. 2006. Actin dynamics at sites of extracellular matrix 61. Butler, D. M., T. Leizer, and J. A. Hamilton. 1989. Stimulation of human degradation. Eur. J. Cell Biol. 85: 1217–1231. synovial fibroblast DNA synthesis by platelet-derived growth factor and fi- 37. Charbonneau, M., K. Harper, F. Grondin, M. Pelmus, P. P. McDonald, and broblast growth factor: differences to the activation by IL-1. J. Immunol. 142: C. M. Dubois. 2007. Hypoxia-inducible factor mediates hypoxic and tumor 3098–3103. necrosis factor alpha-induced increases in tumor necrosis factor-alpha converting 62. Kumkumian, G. K., R. Lafyatis, E. F. Remmers, J. P. Case, S. J. Kim, and enzyme/ADAM17 expression by synovial cells. J. Biol. Chem. 282: 33714– R. L. Wilder. 1989. Platelet-derived growth factor and IL-1 interactions in 33724. rheumatoid arthritis: regulation of synoviocyte proliferation, prostaglandin pro- 38. Chomczynski, P. 1993. A reagent for the single-step simultaneous isolation of duction, and collagenase transcription. J. Immunol. 143: 833–837. Downloaded from RNA, DNA and proteins from cell and tissue samples. Biotechniques 15: 532– 63. Lafyatis, R., E. F. Remmers, A. B. Roberts, D. E. Yocum, M. B. Sporn, and 534, 536–537. R. L. Wilder. 1989. Anchorage-independent growth of synoviocytes from ar- 39. Pham, N. A., A. Morrison, J. Schwock, S. Aviel-Ronen, V. Iakovlev, M. S. Tsao, thritic and normal joints: stimulation by exogenous platelet-derived growth J. Ho, and D. W. Hedley. 2007. Quantitative image analysis of immunohisto- factor and inhibition by transforming growth factor-beta and retinoids. J. Clin. chemical stains using a CMYK color model. Diagn. Pathol. 2: 8. Invest. 83: 1267–1276. 40. Buschman, M. D., P. A. Bromann, P. Cejudo-Martin, F. Wen, I. Pass, and 64. Blanchette, F., R. Day, W. Dong, M. H. Laprise, and C. M. Dubois. 1997. S. A. Courtneidge. 2009. The novel adaptor protein Tks4 (SH3PXD2B) is re- TGFbeta1 regulates gene expression of its own converting enzyme furin. J. Clin. quired for functional podosome formation. Mol. Biol. Cell 20: 1302–1311. Invest. 99: 1974–1983. 41. Diaz, B., G. Shani, I. Pass, D. Anderson, M. Quintavalle, and S. A. Courtneidge. 65. Blanchette, F., P. Rudd, F. Grondin, L. Attisano, and C. M. Dubois. 2001. In- http://www.jimmunol.org/ 2009. Tks5-dependent, nox-mediated generation of reactive oxygen species is volvement of Smads in TGFbeta1-induced furin (fur) transcription. J. Cell. necessary for invadopodia formation. Sci. Signal. 2: ra53. Physiol. 188: 264–273. 42. Beaty, B. T., and J. Condeelis. 2014. Digging a little deeper: the stages of 66. Cerinic, M. M., S. Generini, G. Partsch, A. Pignone, G. Dini, Y. T. Konttinen, invadopodium formation and maturation. Eur. J. Cell Biol. 93: 438–444. and M. Del Rosso. 1998. Synoviocytes from osteoarthritis and rheumatoid ar- 43. Hoshino, D., K. M. Branch, and A. M. Weaver. 2013. Signaling inputs to thritis produce plasminogen activators and plasminogen activator inhibitor-1 and invadopodia and podosomes. J. Cell Sci. 126: 2979–2989. display u-PA receptors on their surface. Life Sci. 63: 441–453. 44. Bowden,E.T.,E.Onikoyi,R.Slack,A.Myoui,T.Yoneda,K.M.Yamada,and 67. Kypta, R. M., Y. Goldberg, E. T. Ulug, and S. A. Courtneidge. 1990. Association S. C. Mueller. 2006. Co-localization of cortactin and phosphotyrosine iden- between the PDGF receptor and members of the src family of tyrosine kinases. tifies active invadopodia in human breast cancer cells. Exp. Cell Res. 312: Cell 62: 481–492. 1240–1253. 68. Hansen, K., M. Johnell, A. Siegbahn, C. Rorsman, U. Engstro¨m, C. Wernstedt, 45. Furuta, T., T. Sakai, T. Senga, T. Osawa, K. Kubo, T. Shimizu, R. Suzuki, C. H. Heldin, and L. Ro¨nnstrand. 1996. Mutation of a Src phosphorylation site in T. Yoshino, M. Endo, and A. Miwa. 2006. Identification of potent and selective the PDGF beta-receptor leads to increased PDGF-stimulated chemotaxis but by guest on October 2, 2021 inhibitors of PDGF receptor autophosphorylation. J. Med. Chem. 49: 2186–2192. decreased mitogenesis. EMBO J. 15: 5299–5313. 46. Choudhury, G. G., L. Mahimainathan, F. Das, B. Venkatesan, and N. Ghosh- 69. Paulsson, Y., M. P. Beckmann, B. Westermark, and C. H. Heldin. 1988. Density- Choudhury. 2006. c-Src couples PI 3 kinase/Akt and MAPK signaling to PDGF- dependent inhibition of cell growth by transforming growth factor-beta 1 in induced DNA synthesis in mesangial cells. Cell. Signal. 18: 1854–1864. normal human fibroblasts. Growth Factors 1: 19–27. 47. Bartok, B., D. L. Boyle, Y. Liu, P. Ren, S. T. Ball, W. D. Bugbee, C. Rommel, 70. Soma, Y., and G. R. Grotendorst. 1989. TGF-beta stimulates primary human skin and G. S. Firestein. 2012. PI3 kinase d is a key regulator of synoviocyte function fibroblast DNA synthesis via an autocrine production of PDGF-related peptides. in rheumatoid arthritis. Am. J. Pathol. 180: 1906–1916. J. Cell. Physiol. 140: 246–253. 48. Bartok, B., D. Hammaker, and G. S. Firestein. 2014. Phosphoinositide 3-kinase d 71. Gotzmann, J., A. N. Fischer, M. Zojer, M. Mikula, V. Proell, H. Huber, regulates migration and invasion of synoviocytes in rheumatoid arthritis. M. Jechlinger, T. Waerner, A. Weith, H. Beug, and W. Mikulits. 2006. A crucial J. Immunol. 192: 2063–2070. function of PDGF in TGF-beta-mediated cancer progression of hepatocytes. 49. Valius, M., and A. Kazlauskas. 1993. Phospholipase C-gamma 1 and phospha- Oncogene 25: 3170–3185. tidylinositol 3 kinase are the downstream mediators of the PDGF receptor’s 72. Sakuma, M., K. Hatsushika, K. Koyama, R. Katoh, T. Ando, Y. Watanabe, mitogenic signal. Cell 73: 321–334. M. Wako, M. Kanzaki, S. Takano, H. Sugiyama, et al. 2007. TGF-beta type I 50. Jechlinger, M., A. Sommer, R. Moriggl, P. Seither, N. Kraut, P. Capodiecci, receptor kinase inhibitor down-regulates rheumatoid synoviocytes and prevents M. Donovan, C. Cordon-Cardo, H. Beug, and S. Grunert.€ 2006. Autocrine the arthritis induced by type II collagen antibody. Int. Immunol. 19: 117–126. PDGFR signaling promotes mammary cancer metastasis. J. Clin. Invest. 116: 73. Loughlin, J. 2011. Genetics of osteoarthritis. Curr. Opin. Rheumatol. 23: 479– 1561–1570. 483. 51. Majack, R. A., M. W. Majesky, and L. V. Goodman. 1990. Role of PDGF-A 74. Valdes, A. M., T. D. Spector, A. Tamm, K. Kisand, S. A. Doherty, expression in the control of vascular smooth muscle cell growth by transforming E. M. Dennison, M. Mangino, A. Tamm, I. Kerna, D. J. Hart, et al. 2010. Genetic growth factor-beta. J. Cell Biol. 111: 239–247. variation in the SMAD3 gene is associated with hip and knee osteoarthritis. 52. Chow, E. K., R. M. O’connell, S. Schilling, X. F. Wang, X. Y. Fu, and G. Cheng. Arthritis Rheum. 62: 2347–2352. 2005. TLR agonists regulate PDGF-B production and cell proliferation through 75. Tateishi, K., C. Higuchi, W. Ando, K. Nakata, J. Hashimoto, D. A. Hart, TGF-beta/type I IFN crosstalk. EMBO J. 24: 4071–4081. H. Yoshikawa, and N. Nakamura. 2007. The immunosuppressant FK506 pro- 53. Taylor, L. M., and L. M. Khachigian. 2000. Induction of platelet-derived growth motes development of the chondrogenic phenotype in human synovial stromal factor B-chain expression by transforming growth factor-beta involves trans- cells via modulation of the Smad signaling pathway. Osteoarthritis Cartilage 15: activation by Smads. J. Biol. Chem. 275: 16709–16716. 709–718. 54. Bruna, A., R. S. Darken, F. Rojo, A. Ocanã, S. Pen˜uelas, A. Arias, R. Paris, 76. Nakahara, H., T. Otani, T. Sasaki, Y. Miura, Y. Takai, and M. Kogo. 2003. In- A. Tortosa, J. Mora, J. Baselga, and J. Seoane. 2007. High TGFbeta-Smad ac- volvement of Cdc42 and Rac small G proteins in invadopodia formation of tivity confers poor prognosis in glioma patients and promotes cell proliferation RPMI7951 cells. Genes Cells 8: 1019–1027. depending on the methylation of the PDGF-B gene. Cancer Cell 11: 147–160. 77. Yamaguchi, H., S. Yoshida, E. Muroi, N. Yoshida, M. Kawamura, Z. Kouchi, 55. Sawyer, J. S., B. D. Anderson, D. W. Beight, R. M. Campbell, M. L. Jones, Y. Nakamura, R. Sakai, and K. Fukami. 2011. Phosphoinositide 3-kinase sig- D. K. Herron, J. W. Lampe, J. R. McCowan, W. T. McMillen, N. Mort, et al. naling pathway mediated by p110a regulates invadopodia formation. J. Cell 2003. Synthesis and activity of new aryl- and heteroaryl-substituted pyrazole Biol. 193: 1275–1288. inhibitors of the transforming growth factor-beta type I receptor kinase domain. 78. Hoshino, D., J. Jourquin, S. W. Emmons, T. Miller, M. Goldgof, K. Costello, J. Med. Chem. 46: 3953–3956. D. R. Tyson, B. Brown, Y. Lu, N. K. Prasad, et al. 2012. Network analysis of the 56. Inman, G. J., F. J. Nicola´s, J. F. Callahan, J. D. Harling, L. M. Gaster, focal adhesion to invadopodia transition identifies a PI3K-PKCa invasive sig- A. D. Reith, N. J. Laping, and C. S. Hill. 2002. SB-431542 is a potent and naling axis. Sci. Signal. 5: ra66. specific inhibitor of transforming growth factor-beta superfamily type I activin 79. Arsenault, D., K. Brochu-Gaudreau, M. Charbonneau, and C. M. Dubois. 2013. receptor-like kinase (ALK) receptors ALK4, ALK5, and ALK7. Mol. Pharma- HDAC6 deacetylase activity is required for hypoxia-induced invadopodia for- col. 62: 65–74. mation and cell invasion. PLoS One 8: e55529. The Journal of Immunology 3275

80. Vanhaesebroeck, B., J. Guillermet-Guibert, M. Graupera, and B. Bilanges. 2010. 90. van der Heijde, D. M., P. L. van Riel, I. H. Nuver-Zwart, F. W. Gribnau, and The emerging mechanisms of isoform-specific PI3K signalling. Nat. Rev. Mol. L. B. vad de Putte. 1989. Effects of hydroxychloroquine and sulphasalazine on Cell Biol. 11: 329–341. progression of joint damage in rheumatoid arthritis. Lancet 1: 1036–1038. 81. Randis, T. M., K. D. Puri, H. Zhou, and T. G. Diacovo. 2008. Role of PI3Kdelta 91. Wolfe, F., and J. T. Sharp. 1998. Radiographic outcome of recent-onset rheu- and PI3Kgamma in inflammatory arthritis and tissue localization of neutrophils. matoid arthritis: a 19-year study of radiographic progression. Arthritis Rheum. Eur. J. Immunol. 38: 1215–1224. 41: 1571–1582. 82. Winkler, D. G., K. L. Faia, J. P. DiNitto, J. A. Ali, K. F. White, E. E. Brophy, 92. van Aken, J., L. R. Lard, S. le Cessie, J. M. Hazes, F. C. Breedveld, and M. M. Pink, J. L. Proctor, J. Lussier, C. M. Martin, et al. 2013. PI3K-d and PI3K- T. W. Huizinga. 2004. Radiological outcome after four years of early versus g inhibition by IPI-145 abrogates immune responses and suppresses activity in delayed treatment strategy in patients with recent onset rheumatoid arthritis. autoimmune and inflammatory disease models. Chem. Biol. 20: 1364–1374. Ann. Rheum. Dis. 63: 274–279. 83. Gabelli, S. B., I. Echeverria, M. Alexander, K. C. Duong-Ly, D. Chaves-Moreira, 93. Smolen, J. S., and P. Emery. 2000. Efficacy and safety of leflunomide in active E. T. Brower, B. Vogelstein, and L. M. Amzel. 2014. Activation of PI3Ka by rheumatoid arthritis. Rheumatology 39(Suppl. 1): 48–56. physiological effectors and by oncogenic mutations: structural and dynamic 94. Lipsky, P. E., D. M. van der Heijde, E. W. St Clair, D. E. Furst, F. C. Breedveld, effects. Biophys. Rev. 6: 89–95. J. R. Kalden, J. S. Smolen, M. Weisman, P. Emery, M. Feldmann, et al; Anti- 84. Conery, A. R., Y. Cao, E. A. Thompson, C. M. Townsend, Jr., T. C. Ko, and Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy K. Luo. 2004. Akt interacts directly with Smad3 to regulate the sensitivity to Study Group. 2000. Infliximab and methotrexate in the treatment of rheumatoid TGF-beta induced apoptosis. Nat. Cell Biol. 6: 366–372. arthritis. N. Engl. J. Med. 343: 1594–1602. 85. Remy, I., A. Montmarquette, and S. W. Michnick. 2004. PKB/Akt modulates 95. Dennis, G., Jr., C. T. Holweg, S. K. Kummerfeld, D. F. Choy, A. F. Setiadi, TGF-beta signalling through a direct interaction with Smad3. Nat. Cell Biol. 6: J. A. Hackney, P. M. Haverty, H. Gilbert, W. Y. Lin, L. Diehl, et al. 2014. Sy- 358–365. novial phenotypes in rheumatoid arthritis correlate with response to biologic 86. Seoane, J., H. V. Le, L. Shen, S. A. Anderson, and J. Massague´. 2004. Integration therapeutics. Arthritis Res. Ther. 16: R90. of Smad and forkhead pathways in the control of neuroepithelial and glioblas- 96. Sheen, Y. Y., M. J. Kim, S. A. Park, S. Y. Park, and J. S. Nam. 2013. Targeting toma cell proliferation. Cell 117: 211–223. the transforming growth factor-b signaling in cancer therapy. Biomol. Ther. 21: 87. Gomis, R. R., C. Alarcoń, W. He, Q. Wang, J. Seoane, A. Lash, and J. Massague´. 323–331. 2006. A FoxO-Smad synexpression group in human keratinocytes. Proc. Natl. 97. Engelman, J. A. 2009. Targeting PI3K signalling in cancer: opportunities, Acad. Sci. USA 103: 12747–12752. challenges and limitations. Nat. Rev. Cancer 9: 550–562. 88. Lim, J. H., H. Jono, K. Komatsu, C. H. Woo, J. Lee, M. Miyata, T. Matsuno, 98. Rodon, J., R. Dienstmann, V. Serra, and J. Tabernero. 2013. Development of Downloaded from X. Xu, Y. Huang, W. Zhang, et al. 2012. CYLD negatively regulates transforming PI3K inhibitors: lessons learned from early clinical trials. Nat. Rev. Clin. Oncol. growth factor-b-signalling via deubiquitinating Akt. Nat. Commun. 3: 771. 10: 143–153. 89. Zhou, B. P., J. Deng, W. Xia, J. Xu, Y. M. Li, M. Gunduz, and M. C. Hung. 2004. 99. Banham-Hall, E., M. R. Clatworthy, and K. Okkenhaug. 2012. The therapeutic Dual regulation of Snail by GSK-3beta-mediated phosphorylation in control of potential for PI3K inhibitors in autoimmune rheumatic diseases. Open Rheu- epithelial-mesenchymal transition. Nat. Cell Biol. 6: 931–940. matol. J. 6: 245–258. http://www.jimmunol.org/ by guest on October 2, 2021