PIM-1 kinase is a novel regulator of proinflammatory cytokine-mediated responses in rheumatoid arthritis fibroblast-like synoviocytes

You-Jung Ha 1, Yong Seok Choi2, Dong Woo Han3, Eun Ha Kang1, In Seol Yoo4, Jin Hyun Kim4, Seong Wook Kang4, Eun Young Lee5, Yeong Wook Song5,6 and Yun Jong Lee1,3,5

Abstract

Objectives. This study investigated the expression of proviral-integration site for Moloney murine leukaemia virus (PIM) -1 kinase in RA synovium and RA fibroblast-like synoviocytes (FLSs) along with its impact on RA-FLS aggressiveness.
Methods. The expression of PIM kinases was assessed in synovial tissues by immunohistochemistry and double IF. After PIM-1 inhibition using either small-interfering RNA or the chemical inhibitor AZD1208, we performed proliferation and migration assays and measured the levels of MMPs and IL-6 released from RA-FLSs under stimulation with proinflammatory cytokines (TNF-a, S100A4 and IL-6/soluble IL-6 receptor). Additionally, PIM-1-associated downstream signalling pathways were analysed by immunoblotting.
Results. Three isoforms of PIM kinases were immunodetected in the synovial tissues from patients with RA or OA. Specifically, PIM-1 and PIM-3 were upregulated in RA synovium and PIM-1 was expressed in T cells, macrophages and FLSs. Additionally, upon stimulation of RA-FLSs with TNF-a, S100A4 and IL-6/sIL-6R, PIM-1 and PIM-3, but not PIM-2, were significantly inducible. Moreover, PIM-1 knockdown or AZD1208 treatment significantly suppressed basal or cytokine-induced proliferation and migration of RA-FLS and the secretion of MMPs from stimulated RA-FLSs. PIM-1 knockdown significantly affected the phosphorylation levels of extracellular signal-regulated kinase and cAMP responsive element binding protein in RA-FLSs.
Conclusion. PIM-1 was upregulated in RA synovial tissues and RA-FLSs and its inhibition significantly reduced the proliferation, migration and MMP production of RA-FLSs in vitro. These findings suggest PIM-1 as a novel regulator of the aggressive and invasive behaviour of RA-FLSs and indicate its potential as a target for RA treatment.

Key words: RA, fibroblast-like synoviocytes, synovium, PIM kinase.

Rheumatology key messages

. PIM-1 is upregulated in RA synovial tissue and RA fibroblast-like synoviocytes.
. PIM-1 inhibition suppresses the cancer-like features of RA fibroblast-like synoviocytes.
. PIM-1 might be a potential therapeutic target for RA treatment.

Introduction

RA is a chronic inflammatory systemic autoimmune disease affecting diarthrodial joints and leading to progressive joint damage and functional disability. Among professional immune and non-immune cells involved in RA development or perturbation, RA fibroblast-like synoviocytes (FLSs) are key components of invasive RA synovial tissue and exhibit unique aggressive features that contribute to destructive joint inflammation [1]. RA-FLSs produce proinflammatory cytokines and matrix-degrading enzymes and exhibit cancer-like proliferation and migration [2]. These activities enable RA-FLSs to potentially promote immune-cell recruitment and activation, neoangiogenesis, cartilage degradation and bone erosion.
The proviral-integration site for Moloney murine leukaemia virus (PIM) family includes serine/threonine kinases that cooperate with c-Myc [3]. The three PIM kinase members (PIM-1, -2 and -3) regulate multiple cellular functions, including cell cycle progression and growth. Because of their potential oncogenic activity and higher expression in cancer cells, PIM kinases represent attractive therapeutic targets for some haematological and solid malignancies. Unlike other protein kinases, PIM kinases do not have a regulatory domain for activation and are constitutively active soon after gene expression [4]. Because the expression of PIM kinases is regulated by the Janus kinase/signal transducers and activators of transcription (STAT) pathway, PIM kinases are considered a downstream effector of many cytokine signalling pathways.
Considering that RA-FLSs display characteristics similar to transformed cells and their abnormal proliferation is associated with c-Myc or STAT3 [5, 6], it is possible that PIM kinases are key factors related to the aggressive and invasive traits associated with RA-FLSs. Previously, PIM-1 kinase was reported to be associated with Th1 differentiation and enhance RANK-ligand-induced osteoclastogenesis [7, 8]. Additionally, PIM-1 kinase inhibition improved experimental murine colitis through downregulation of Th1 and Th17 responses [9].
However, the studies regarding the expression of PIM kinases in RA synovial tissues and their role in the biological responses of RA-FLSs are limited. Here, for the first time, we found that the expression of PIM-1 kinase was upregulated in RA synovial tissues and induced by proinflammatory cytokines in RA-FLSs. Additionally, we found that inhibition of PIM-1 kinase significantly decreased the proliferation and migration of FLSs, as well as MMP production, through the suppression of extracellular signalregulated kinase (ERK)/cAMP responsive element binding protein (CREB) activity.

Methods

Study subjects and reagents

Thirteen patients with RA (five for immunohistochemical analysis and eight for RA-FLS isolation), who had undergone synovectomy or joint-replacement surgery, were enrolled in this study. RA was diagnosed according to the 1987 ACR criteria [10]. Synovial tissues from four patients with knee OA undergoing joint replacement served as controls for the immunohistochemical study. Demographic and clinical features of the patients are summarized in Supplementary Table S1, available at Rheumatology online. This study was approved by the Seoul National University Bundang Hospital Institutional Review Board (B-0905/075-013) and was performed according to the recommendations of the Declaration of Helsinki. Informed consent was obtained from all patients. Reagents used in our experiments are given in the Supplementary Data, available at Rheumatology online.

Isolation and culture of RA-FLS

Briefly, synovial tissue was minced and incubated with 1 mg/ml collagenase for 1 h at 37C. After filtering through a 100-mm filter, the cells were transferred to culture dishes and allowed to adhere overnight. Adherent FLSs were grown in DMEM supplemented with 10% fetal bovine serum. RA-FLSs at passages 39 were used for the experiments.

Immunohistochemistry and double IF staining

Formalin-fixed paraffin-embedded tissue sections were deparaffinized and rehydrated, and the slides were blocked with 5% normal donkey serum. For immunohistochemical analysis, the sections were incubated with anti-PIM-1 (1:200), anti-PIM-2 (1:100) and anti-PIM-3 (1:1000) antibodies. Also, to check for non-specific binding, isotype control antibodies were added on the tissue sections of colon cancer (for PIM-1), lymphoma (for PIM-2), or hepatocellular carcinoma (for PIM-3). Supplementary Fig. S1, available at Rheumatology online, shows no positive staining in these negative controls. The UltraVision LP Detection System HRP Polymer (Thermo Fisher Scientific, Waltham, MA, USA) was used as a secondary antibody. All sections were counterstained with haematoxylin. To provide a semi-quantitative analysis, we manually counted the number of immunostained cells among synovial lining and sublining stromal cells. A minimum of 1000 cells per slide were analysed in at least three high-power fields, which were randomly selected areas and excluded the area of lymphoid follicles or aggregates.
To perform double IF staining, the slides were incubated with the anti-PIM-1 antibody (1:200) overnight at 4C. After repeated washing, the slides were stained with anti-CD20 (1:400), anti-CD68 (1:100), anti-CD3 (1:50) or anti-vimentin (1:100) antibodies overnight at 4C. The slides were then incubated with fluorescent dye-conjugated detection antibodies (Alexa Fluor 488 for PIM-1 and Texas Red-X for cell-specific markers) for 1 h at room temperature in the dark. After counterstaining with 40,6-diamidino-2-phenylindole (DAPI), the slides were mounted with mounting solution and observed under a confocal microscope (LSM 710; Carl Zeiss MicroImaging, Jena, Germany).

RT-PCR and quantitative real-time PCR

To semi-quantitatively evaluate the expression of PIM mRNA, RT-PCR was performed with primers specific for PIM-1 (50-GCTCGGTCTACTCAGGCATC-30 and 50-CGGG CATCTGACAAGAGAGG-30; accession number: NM_002648), PIM-2 (50-GCCTCACAGATCGACTCCAG-30 and 50-GAAGCAGGGCACCAGAACC-30; NM_006875) and PIM-3 (50-ACCGCGACATTAAGGACGAAA-30 and 50-ACACACCATATCGTAGAGA-AGCA-30; NM_001001852). Quantitative real-time PCR was performed using an ABI7500 system and commercially available TaqMan probes for each PIM kinase (Applied Biosystems, Foster City, CA, USA). The fold difference in the expression of target mRNA was calculated using the comparative Ct method (2Ct) and normalized to RPLP0 (large ribosomal protein).

Immunoblotting

Total cell lysates were separated on 10% polyacrylamide gels, and signals on polyvinylidene difluoride membranes were developed using an enhanced chemiluminescence system (Amersham Biosciences, Little Chalfont, UK). b-Actin was used as an internal control. The relative expression of each protein was determined by densitometric analysis using ImageJ software (NIH, Bethesda, MD, USA).

Small-interfering RNA transfection

RA-FLSs were seeded in six-well plates (3 105 cells/well) and transfected with PIM-1 or control small-interfering (si) RNA using siRNA transfection reagent (Santa Cruz Biotechnology, Dallas, TX, USA). The cells were incubated in 2 complete medium overnight and then transferred to 24-well plates (3 104 cells/well) or 48-well plates (1 104 cells/well). Transfected cells were used for further experiments 48 h after transfection.

Cell viability, proliferation and migration assays

For cell viability assays, PIM-1 or control siRNA-transfected RA-FLSs were seeded in 48-well plates. After 24 h, the cells were stimulated with TNF-a (05 ng/ml) for a maximum of 96 h. The cells were then incubated with medium containing 2,5-diphenyltetrazolium bromide (MTT) for 30 min and lysed in dimethyl sulphoxide. To measure the cell cytotoxicity of AZD1208, RA-FLSs were treated with AZD1208 (050 mM) for 24 h in serum-free medium. To evaluate the effect of PIM-1 on cell proliferation, bromodeoxyuridine (BrdU) incorporation was measured using a Cytoselect BrdU cell proliferation ELISA kit (Cell Biolabs, San Diego, CA, USA). PIM-1 or control siRNA-transfected RA-FLSs (3 105 cells/well) were stimulated with TNF-a (05 ng/ml) for up to 96 h. After 24, 48 or 96 h, cells were pulsed with BrdU for 2 h. ELISAs were performed using anti-BrdU antibody and the absorbance was measured at 450 nm. Relative BrdU incorporation was calculated as the absorbance with respect to the unstimulated and control siRNA-transfected cells at 24 h.
A scratch assay was performed to quantify cell migration. RA-FLSs were seeded in six-well plates and scraped using a sterile cell scraper (3 mm width). Cells in complete medium were stimulated with TNF-a (05 ng/ml) for 24, 48 and 96 h, and migrated cells in the scratch area were counted under the microscope at 24, 48 and 72 h. For the cell migration assay in the presence of AZD1208, RAFLSs were pretreated with AZD1208 (020 mM) for 2 h, followed by stimulation with TNF-a. For the scratch assay in the presence of AZD1208, cells were scraped using a sterile cell scraper (3 mm width) and subsequently washed twice with complete medium. Cells were then pretreated with AZD1208 (0, 2, or 20 mM) for 2 h, followed by stimulation with TNF-a (5 ng/ml). Migrating cells within the scratch were counted at 24, 48 and 72 h.

ELISA

Conditioned medium was collected after RA-FLSs in serum-free medium were stimulated with TNF-a or S100A4 for 24 h. Levels of MMP-1, MMP-3, MMP-13 and IL-6 were measured using commercial ELISA kits according to the manufacturers’ recommendations.

Flow cytometric analysis

Forty-eight hours after the transfection with control or PIM-1 siRNA, RA-FLSs were incubated for 24h in the presence of TNF-a (5ng/ml) or S100A4 (1mg/ml). They were detached, washed twice with cold PBS, and then resuspended in the cell staining buffer (1.0 105 cells/100 ml). Cells were labelled with PE-anti-CXCR4 or IgG2a isotype control antibodies at 4C for 20min. At least 5 104 cells from each sample were analysed in a FACScan flow cytometer (BD Biosciences, San Jose, CA, USA).

Statistical analysis

All experiments were repeated at least three times, and data are expressed as the mean ± S.E.M. Quantitative realtime PCR was performed in triplicate, and ELISA determination was performed in duplicate. The MannWhitney U-test or KruskalWallis test was used to compare continuous variables. In cases of multiple comparisons, a Bonferroni correction was applied. Statistical analyses were performed using SPSS Statistics for Windows version 20 (IBM Corp., Armonk, NY, USA), and a P or corrected P (Pc) < 0.05 was considered statistically significant.

Results

The expression of three PIM kinase members in RA synovial tissues

Three PIM kinases were detected in all synovial tissue samples from patients with OA and RA. In some areas showing infiltrated inflammatory cells, PIM kinases showed weak to moderate immunostaining in the OA synovial membrane but the intensity of PIM kinases was stronger in RA synovial tissues (Fig. 1A). PIM-1-positive cells were distributed in both the lining and sublining of the RA synovial membrane, whereas PIM-3-positive cells tended to be found more frequently in the sublining layer. In the lymphoid follicles of RA synovium, PIM-1- or PIM-3positive cells were frequently observed, whereas PIM-2positive cells were scarcely observed. When comparing the numbers of immunopositive cells between RA and OA synovium, the fraction of PIM-1 (P = 0.038) or PIM-3 (P = 0.010) immunostained cells was significantly higher in RA synovium (Fig. 1B).

Cellular expression of PIM-1 kinase in RA synovial tissues

To examine which cells express PIM-1 protein in RA synovium, we performed double IF staining for PIM-1 and cellspecific markers (CD3 for T cells, CD20 for B cells, CD68 for macrophages and vimentin for FLSs). As shown in Fig. 2, PIM-1 was immunostained in CD3-, CD68- or

Immunohistochemical staining for three PIM kinases

(A) Representative light-microscopy photographs showing that all PIM kinase expression was elevated in synovial tissues from three patients with RA as compared with that observed in OA synovium. Immunostaining for PIM-1 or PIM-3 was more pronounced than that for PIM-2 in RA synovium. PIM-1-positive cells were detected in the lining layer, sublining layer and lymphoid follicles of RA synovium, whereas PIM-3-positive cells were observed mainly in the sublining layer and lymphoid follicles. White boxes in the left panel (original magnification, 100) indicate the magnified regions shown in the right panel (original magnification, 400). (B) Semi-quantitative analysis of the expression of PIM kinases in RA and OA synovial tissues. When comparing the number of PIM kinase-immunostained cells, the fraction of PIM-1- or PIM-3positive cells was significantly higher among synovial lining and stromal cells of RA synovial membrane (RA, n = 6; OA, n = 4). *P < 0.05. PIM: proviral-integration site for Moloney murine leukaemia virus. Scale bars = 200 mm. vimentin-positive cells in the RA synovium, whereas CD20-positive cells did not express PIM-1.

Induction of PIM-1 kinase by proinflammatory cytokines

We then investigated whether PIM-1 expression could be induced by proinflammatory cytokines, such as TNF-a, IL-6 and extracellular S100A4. S100A4 is a metastasisassociated protein previously reported to promote the proliferation and invasion of RA-FLSs [11]. The expression of PIM-1 and PIM-3 transcripts was significantly upregulated (both P = 0.001 by KruskalWallis test; Fig. 3A, upper panel) when RA-FLSs were stimulated with TNF-a (510 ng/ml), IL-6/sIL-6 (50 ng/ml) or S100A4 (1 mg/ml) for 8 h; however, PIM-2 mRNA levels were unchanged. Additionally, in RA-FLSs, PIM-1 exhibited a greater increase in transcript level relative to PIM-3 (average 1.63vs 1.26-fold following stimulation with 5 ng/ml TNF-a; 2.40- vs 1.66-fold following S100A4 stimulation; and 3.27- vs 1.82-fold following IL-6/sIL-6R stimulation). In the analysis based on the Ct (Ct of RPLP0 Ct of a target gene) method, only PIM-1 mRNA levels were significantly changed in cytokine-stimulated conditions (P = 0.004 by KruskalWallis test; Fig. 3A, lower panel). Moreover, corresponding immunoblots confirmed that PIM-1 protein levels were also upregulated after 24 h by TNF-a, IL-6/sIL-6 or S100A4 by 2-fold (P = 0.011; Fig. 3B), which agreed with our immunohistochemical results (Fig. 1).

The effect of PIM-1 knockdown on RA-FLS proliferation and migration

To silence PIM-1 expression, we employed the siRNA approach in RA-FLSs. After transfection, siRNA efficiently knocked down PIM-1 at the mRNA and protein levels without affecting the expression of PIM-2 and PIM-3 (Fig. 4A). In both the presence and absence of TNF-a stimulation, PIM-1 knockdown significantly decreased cell viability measured by MTT assay at 48h (both P = 0.001) and 96h (P = 4.94 104 without TNF-a and P =2.84 105 with TNF-a, by MannWhitney U-test; Fig. 4B). PIM-1 knockdown also significantly reduced basal and TNF-a stimulated proliferation assessed by BrdU incorporation at 48h (P = 0.009 without TNF-a and P= 0.021 with TNF-a) and 96h (P =0.003 without TNF-a and P =0.0002 with TNF-a; Fig. 4C). Furthermore, the cell migration activity of RA-FLSs was significantly inhibited following PIM-1 knockdown (Fig. 4D). At both 48 and 72h, horizontal migration on culture plates was significantly reduced in PIM-1 siRNA-treated RA-FLSs, in both conditions without (P =0.037 at 48h and P = 0.016 at 72h) and with TNF-a stimulation (P =0.010 at 48h and P= 0.006 at 72h).

The effect of PIM-1 knockdown on the expression of MMPs and IL-6 from RA-FLSs

Double IF for PIM-1 and cell-specific markers

Representative confocal images showing PIM-1 (green) expression in vimentin-positive (synovial fibroblasts), CD68positive (macrophages) and CD3-positive (T cells) cells. However, CD20-positive cells did not colocalize with PIM-1positive cells. White boxes in the haematoxylin-and-eosin-stained section indicate areas corresponding to the confocal images (original magnification, 400). PIM: proviral-integration site for Moloney murine leukaemia virus. Scale bars, 500 mm (haematoxylin-and-eosin stained image) and 100 mm (confocal images).
Because increased RA-FLS invasiveness can be conferred by matrix-degrading enzymes, the transcription of which is regulated by the nuclear factor (NF)-kB-signalling pathway [12], and given that PIM-1 can control NF-kB signalling [13], we investigated changes in MMPs release from PIM-1-knockdown RA-FLSs. We observed that PIM1 knockdown did not affect basal levels of MMPs released from RA-FLSs (Fig. 5); however, TNF-a- or S100A4induced secretion of MMP-1 and MMP-3 was significantly reduced in PIM-1-siRNA-transfected RA-FLSs (TNF-a stimulation: P = 0.009 and P = 0.023, respectively; S100A4 stimulation: P = 0.043 and P = 0.023, respectively, by MannWhitney U-test) as compared with that observed in RA-FLSs treated with control siRNAs. Additionally, TNF-a-induced MMP-13 secretion was significantly suppressed in PIM-1-knockdown FLSs (P = 4.87 104). However, in contrast with previous observations showing that PIM-1 blockade downregulated IL-6 production [13, 14], IL-6 levels did not differ significantly between PIM-1knockdown and control-siRNA-treated RA-FLSs.

The effect of PIM-1 inhibition by a chemical inhibitor, AZD1208

To validate our results obtained following PIM-1 knockdown, we treated RA-FLSs with AZD1208, a highly selective and potent PIM kinase inhibitor. AZD1208 (020 mM) significantly decreased the migration activity and production of MMP-1 (P = 0.008 by KruskalWallis test), MMP-3 (P = 4.29 104) and MMP-13 (P = 0.006) in RA-FLSs in a dose-dependent manner (Supplementary Fig. S2, available at Rheumatology online).

Associations between PIM-1 and the ERK/CREB signalling pathway in RA-FLSs

Because PIM-1 affects the activation status of NF-kB [8, 13] and mitogen-activated protein (MAP) kinases [1518], we analysed levels of active MAP kinases and NF-kB by immunoblotting. In RA-FLSs, PIM-1 knockdown did not change the phosphorylation levels of p65, p38 and c-Jun N-terminal kinase (JNK) following TNF-a or S100A4 stimulation; however, phospho-ERK levels were significantly attenuated in PIM-1-knockdown RA-FLSs (Fig. 6). Because ERK activation results in CREB phosphorylation, and MMP expression is regulated by CREB [19], we performed phospho-CREB immunoblotting, revealing that PIM-1 knockdown suppressed CREB activity (Fig. 6).
Although PIM-1 could regulate c-Myc and the mammalian

Induction of PIM-1 by TNF-a, IL-6, and S100A4

(A) Real-time RT-PCR analysis showing significant increases in fold change (upper panel) of PIM-1 and PIM-3 mRNA levels in RA fibroblast-like synoviocytes 8 h after stimulation with TNF-a (510 ng/ml), IL-6/sIL-6R (50 ng/ml) or S100A4 (1 mg/ml) in serum-free media (n = 5). When Ct values were compared among basal and cytokine-stimulated conditions (lower panel), only PIM-1 transcript levels were significantly changed. (B) The expression of PIM-1 protein was significantly induced after 24 h by TNF-a, IL-6/sIL-6R or S100A stimulation, as assessed by immunoblotting (n = 4). Data are expressed as the mean ± S.E.M. *Pc < 0.05. PIM: proviral-integration site for Moloney murine leukaemia virus. target of rapamycin (mTOR) pathway and CXCR4 surface expression in haematopoietic cells [3, 20, 21], PIM-1 knockdown did not significantly affect the levels of cMyc and mTOR phosphorylation and the cell surface expression of CXCR4 in RA-FLSs (data not shown).

Discussion

Although PIM-1 is highly expressed in haematopoietic progenitors of the human fetal liver and spleen and all haematopoietic organs of the developing mouse embryo [22, 23], PIM-1 mutant mice exhibit only minor phenotypic changes [24]. More interestingly, PIM-1, PIM-2, and PIM-3 triple knockout mice are postnatally viable and fertile and showed decreased response to growth factors in haematopoietic stem cells and IL-2 in peripheral T lymphocytes [25]. These findings indicate that PIM-1 might contribute to cellular response to cytokines and could be involved in autoimmune diseases in the pathogenesis of which T cells have a role, such as RA. However, most previous studies have focused on the role of PIM-1 in cancer cells of haematological and epithelial origin, because PIM-1 acts as a strong synergistic partner of c-Myc in the cell cycle pathway and tumorigenesis. Our results are the first to clearly demonstrate that PIM-1 expression was upregulated in RA synovium and could be inducible by TNF-a, IL-6 or S100A4 in RA-FLSs.
In non-malignant cells, it was previously reported that PIM-1 is expressed in activated monocytes/macrophages stimulated with GM-CSF or lipopolysaccharides [9, 26] and activated ab-T cells [27, 28]. These findings were in line with our double IF data showing co-localization of CD3- or CD68-positive cells with PIM-1. Therefore, in RA synovium, PIM-1 might function as a survival factor for activated macrophages and T cells; however, no consistent results have been obtained concerning PIM-1 expression in B cell lineages. Some reports showed that PIM-1 was not inducible in peripheral B cells [27] or had no effect on early B cell progenitor pools [29], whereas others reported that PIM-1 is induced in murine pro-B cells [30] and in IL-27-stimulated human naı¨ve B cells [31]. In the present study, CD20-positive cells did not exhibit PIM-1-specific signals.
Apart from professional immune cells, FLSs are also considered major players in RA pathogenesis.

PIM-1 knockdown affects RA-FLSs proliferation and migration

(A) Suppression of PIM-1 expression in RA-FLSs by PIM-1-specific siRNA without affecting PIM-2 and PIM-3 expression. (B) Cell viability by MTT assay was significantly attenuated in PIM-1-knockdown RA-FLSs at 48 h and 96 h, regardless of TNF-a stimulation (n = 5). (C) Cell proliferation, assessed by BrdU incorporation, was also significantly decreased by PIM1 siRNA transfection at 48 and 96 h in both absence and presence of TNF-a stimulation (n = 4). (D) Relative migration into the scratch area was significantly decreased in PIM-1-knockdown RA-FLSs at 48 and 72 h, regardless of TNF-a stimulation (n = 6). Rates were relative to those of control-siRNA-treated cells without TNF-a stimulation at 24 h. Data are expressed as the mean ± S.E.M. P-values were calculated by using MannWhitney test for comparing between control and PIM-1 siRNA-treated RA-FLSs. *P < 0.05, yP 4 0.01, zP < 0.001. BrdU: bromodeoxyuridine; FLS: fibroblast-like synoviocyte; MTT: 2,5-diphenyltetrazolium bromide; PIM: proviral-integration site for Moloney murine leukaemia virus; siRNA:
Specifically, they are actively involved in bone and cartilage damage through the direct invasion of proliferative and invasive synovial pannus, the secretion of MMPs or proinflammatory cytokines, and the induction of boneturnover imbalance [2]. The present study showed that PIM-1 was expressed in RA-FLSs, and that its expression was elevated when stimulated with TNF-a, IL-6, or S100A4. Considering that PIM-1 is a regulator of small-interfering RNA. proliferation and invasion in some non-haematological cancer cells [32, 33], it could be expected that increased PIM-1 expression might be related to the aggressive behaviour of RA-FLSs. Here, PIM-1 regulated the proliferation of unstimulated RA-FLSs. Moreover, PIM-1 knockdown by a specific siRNA or use of a chemical inhibitor of PIM kinase revealed that PIM-1 regulated proinflammatory cytokine-induced proliferation, migration and MMP secretion in RA-FLSs. Neutralizing antibodies against TNF-a and IL-6 are well-established therapeutic agents. Janus kinase inhibitors, another agent approved for RA treatment, directly block phosphorylation of STAT3, which primarily controls the expression of PIM kinases [33]. Considering our results together, the effect of TNF-a, IL-6 or Janus kinase/STAT blockade therapy might result from the modulation of PIM-1 expression.
MMPs are implicated in the breakdown of extracellular matrix, which contributes to cartilage and bone destruction and loss of joint function in chronic inflammatory arthritis such as RA. MMP-1, -3, -9 and -13 are actively transcribed in RA-FLSs [34], and circulating levels of MMP-1 or MMP-3 are related to RA disease activity and the progression of RA-related joint damage [35, 36]. Recently, Lim et al. [14] reported that PIM-1 knockdown decreases MMP-9 levels in human amnion cells. In the present study and for the first time, we observed that PIM-1 might represent a novel regulator capable of controlling the TNF-a-induced expression of MMPs in RAFLSs.
S100A4 is known to be associated with metastatic progression in various cancers and was reported to be upregulated in RA synovial tissue, especially at sites of bone invasion [37]. Additionally, extracellular S100A4 levels are elevated and its levels are correlated with disease activity and radiographic damage in RA patients [11, 38]. Moreover, secreted S100A4 can increase MMP levels in RA-FLSs [37]. We showed for the first time that S100A4 increased PIM-1 expression, and that S100A4-induced MMP production was dependent upon PIM-1.
Many papers have provided an excellent overview of the downstream effectors involved in PIM-1-mediated responses [3, 4]. Previous studies also reported that PIM-1 is capable of activating NF-kB signalling after TNF-a stimulation in HeLa cells [13] and is involved in the

Effect of PIM-1 knockdown on TNF-a- or S100A4-induced MMP secretion from RA-FLSs

PIM-1 knockdown significantly reduced TNF-a- or S100A4-induced secretion of MMP-1 (A) and MMP-3 (B) from RAFLSs (n = 5). The release of MMP-13 from TNF-a-stimulated RA-FLSs was also decreased by PIM-1 knockdown (C); however, IL-6 levels were unaffected by PIM-1 knockdown under both unstimulated and stimulated conditions (D). Data are expressed as the mean ± S.E.M. *P < 0.05, yP 4 0.01, zP < 0.001. FLS: fibroblast-like synoviocyte; PIM: proviral-integration site for Moloney murine leukaemia virus; siRNA: small-interfering RNA. activation of p38 in basophils [15], JNK in colon cancer cells [16], and ERK in chronic lymphocytic leukaemia and prostate cancer cells [17, 18]. However, in the present study, we did not observe a significant effect of PIM-1 on NF-kB and p38/JNK signalling in RA-FLSs; however, we found that PIM-1 inhibition most significantly affected the ERK/CREB pathway. Previous studies showed that CREB is upregulated in RA synovial AZD1208 tissues, and that its nuclear translocation is inducible by IL-1b and TNF-a in RA-FLSs [39, 40]. CREB is also involved in the hyperfunction of RA-FLSs through cell proliferation and the production of proinflammatory cytokines and MMPs [40]. These findings support out observation that PIM-1 could act as a regulator of abnormal RA-FLS hyperfunction via the ERK/ CREB pathway.
The endogenous levels of c-Myc are increased by PIM1/PIM-2 in cancer cells and c-Myc is potentially involved in synovial hyperplasia and aggressiveness of RA-FLSs [41, 42]. Also, mTOR has been reported to be a regulator of invasiveness of RA-FLSs, and PIM-1 indirectly activates mTOR in haematopoietic FDC-P1 cells [21, 43]. It was recently reported that PIM-2 indirectly suppressed 4-hydroxynonenal-induced NF-kB activation through mTORC1 activation in an immortalized MH7A RA-FLS cell line [44]. However, in the present study, PIM-1 knockdown did not change the levels of c-Myc and phosphomTOR in primary RA-FLSs. Chemokine receptors mediate migration and proliferation of RA-FLSs and are involved in MMP production [45]. Additionally, PIM-1 is essential for CXCR4 surface expression in haematopoietic cells [20]. But PIM-1 did not have any influence on the surface expression of CXCR4 in our study.
In summary, this study showed that PIM-1 was upregulated in RA synovial membranes, and that the cancer-like phenotype of RA-FLSs was affected by PIM-1 expression. Our results suggest that PIM-1 might represent a potential therapeutic target for RA.

References

1 Bottini N, Firestein GS. Duality of fibroblast-like synoviocytes in RA: passive responders and imprinted aggressors. Nat Rev Rheumatol 2013;9:2433.
2 Bustamante MF, Garcia-Carbonell R, Whisenant KD, Guma M. Fibroblast-like synoviocyte metabolism in the pathogenesis of rheumatoid arthritis. Arthritis Res Ther 2017;19:110.
3 Tursynbay Y, Zhang J, Li Z et al. Pim-1 kinase as cancer drug target: an update. Biomed Rep 2016;4:1406.
4 Le BT, Kumarasiri M, Adams JR et al. Targeting Pim kinases for cancer treatment: opportunities and challenges. Future Med Chem 2015;7:3553.
5 Qu Z, Garcia CH, O’Rourke LM et al. Local proliferation of fibroblast-like synoviocytes contributes to synovial hyperplasia. Results of proliferating cell nuclear antigen/ cyclin, c-myc, and nucleolar organizer region staining. Arthritis Rheum 1994;37:21220.
6 Krause A, Scaletta N, Ji JD, Ivashkiv LB. Rheumatoid arthritis synoviocyte survival is dependent on Stat3. J Immunol 2002;169:66106.
7 Aho TL, Lund RJ, Ylikoski EK et al. Expression of human pim family genes is selectively up-regulated by cytokines promoting T helper type 1, but not T helper type 2, cell differentiation. Immunology 2005;116:828.
8 Kim K, Kim JH, Youn BU, Jin HM, Kim N. Pim-1 regulates RANKL-induced osteoclastogenesis via NF-kappaB activation and NFATc1 induction. J Immunol 2010;185:74606.
9 Shen YM, Zhao Y, Zeng Y et al. Inhibition of Pim-1 kinase ameliorates dextran sodium sulfate-induced colitis in mice. Dig Dis Sci 2012;57:182231.
10 Arnett FC, Edworthy SM, Bloch DA et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:31524.
11 Oslejskova L, Grigorian M, Gay S, Neidhart M, Senolt L. The metastasis associated protein S100A4: a potential novel link to inflammation and consequent aggressive behaviour of rheumatoid arthritis synovial fibroblasts. Ann Rheum Dis 2008;67:1499504.
12 Vincenti MP, Brinckerhoff CE. Transcriptional regulation of collagenase (MMP-1, MMP-13) genes in arthritis: integration of complex signaling pathways for the recruitment of gene-specific transcription factors. Arthritis Res 2002;4:15764.
13 Nihira K, Ando Y, Yamaguchi T et al. Pim-1 controls NFkappaB signalling by stabilizing RelA/p65. Cell Death Differ 2010;17:68998.
14 Lim R, Barker G, Lappas M. Inhibition of PIM1 kinase attenuates inflammation-induced pro-labour mediators in human foetal membranes in vitro. Mol Hum Reprod 2017;23:42840.
15 Didichenko SA, Spiegl N, Brunner T, Dahinden CA. IL-3 induces a Pim1-dependent antiapoptotic pathway in primary human basophils. Blood 2008;112:394958.
16 Weirauch U, Beckmann N, Thomas M et al. Functional role and therapeutic potential of the pim-1 kinase in colon carcinoma. Neoplasia 2013;15:78394.
17 Wang J, Anderson PD, Luo W et al. Pim1 kinase is required to maintain tumorigenicity in MYC-expressing prostate cancer cells. Oncogene 2012;31:1794803.
18 Decker S, Finter J, Forde AJ et al. PIM kinases are essential for chronic lymphocytic leukemia cell survival (PIM2/3) and CXCR4-mediated microenvironmental interactions (PIM1). Mol Cancer Ther 2014;13:123145.
19 Ha YJ, Choi YS, Kang EH et al. SOCS1 suppresses IL1beta-induced C/EBPbeta expression via transcriptional regulation in human chondrocytes. Exp Mol Med 2016;48:e241.
20 Grundler R, Brault L, Gasser C et al. Dissection of PIM serine/threonine kinases in FLT3-ITD-induced leukemogenesis reveals PIM1 as regulator of CXCL12-CXCR4mediated homing and migration. J Exp Med 2009;206:195770.
21 Zhang F, Beharry ZM, Harris TE et al. PIM1 protein kinase regulates PRAS40 phosphorylation and mTOR activity in FDCP1 cells. Cancer Biol Ther 2009;8:84653.
22 Amson R, Sigaux F, Przedborski S et al. The human protooncogene product p33pim is expressed during fetal hematopoiesis and in diverse leukemias. Proc Natl Acad Sci U S A 1989;86:885761.
23 Eichmann A, Yuan L, Breant C, Alitalo K, Koskinen PJ. Developmental expression of pim kinases suggests functions also outside of the hematopoietic system. Oncogene 2000;19:121524.
24 Laird PW, van der Lugt NM, Clarke A et al. In vivo analysis of Pim-1 deficiency. Nucleic Acids Res 1993;21:47505.
25 Mikkers H, Nawijn M, Allen J et al. Mice deficient for all PIM kinases display reduced body size and impaired responses to hematopoietic growth factors. Mol Cell Biol 2004;24:610415.
26 Lehtonen A, Matikainen S, Miettinen M, Julkunen I. Granulocyte-macrophage colony-stimulating factor (GMCSF)-induced STAT5 activation and target-gene expression during human monocyte/macrophage differentiation. J Leukoc Biol 2002;71:5119.
27 Wingett D, Stone D, Davis WC, Magnuson NS. Expression of the pim-1 protooncogene: differential inducibility between alpha/beta- and gamma/delta-T cells and B cells.
28 Tahvanainen J, Kylaniemi MK, Kanduri K et al. Proviral integration site for Moloney murine leukemia virus (PIM) kinases promote human T helper 1 cell differentiation. J Biol Chem 2013;288:304858.
29 Xu Z, Gwin KA, Li Y, Medina KL. Developmental stagespecific effects of Pim-1 dysregulation on murine bone marrow B cell development. BMC Immunol 2016;17:16.
30 Zhu N, Ramirez LM, Lee RL et al. CD40 signaling in B cells regulates the expression of the Pim-1 kinase via the NF-kappa B pathway. J Immunol 2002;168:74454.
31 Charlot-Rabiega P, Bardel E, Dietrich C, Kastelein R, Devergne O. Signaling events involved in interleukin 27 (IL-27)-induced proliferation of human naive CD4+ T cells and B cells. J Biol Chem 2011;286:2735062.
32 Xu D, Allsop SA, Witherspoon SM et al. The oncogenic kinase Pim-1 is modulated by K-Ras signaling and mediates transformed growth and radioresistance in human pancreatic ductal adenocarcinoma cells. Carcinogenesis 2011;32:48895.
33 Liu Z, He W, Gao J et al. Computational prediction and experimental validation of a novel synthesized pan-PIM inhibitor PI003 and its apoptosis-inducing mechanisms in cervical cancer. Oncotarget 2015;6:801935.
34 Araki Y, Tsuzuki Wada T, Aizaki Y et al. Histone methylation and STAT-3 differentially regulate interleukin-6induced matrix metalloproteinase gene activation in rheumatoid arthritis synovial fibroblasts. Arthritis Rheumatol 2016;68:111123.
35 Green MJ, Gough AK, Devlin J et al. Serum MMP-3 and MMP-1 and progression of joint damage in early rheumatoid arthritis. Rheumatology (Oxford) 2003;42:838.
36 Shinozaki M, Inoue E, Nakajima A et al. Elevation of serum matrix metalloproteinase-3 as a predictive marker for the long-term disability of rheumatoid arthritis patients in a prospective observational cohort IORRA. Mod Rheumatol 2007;17:4038.
37 Senolt L, Grigorian M, Lukanidin E et al. S100A4 is expressed at site of invasion in rheumatoid arthritis synovium and modulates production of matrix metalloproteinases. Ann Rheum Dis 2006;65:16458.
38 Erlandsson MC, Forslind K, Andersson SE, Lund A, Bokarewa MI. Metastasin S100A4 is increased in proportion to radiographic damage in patients with RA. Rheumatology (Oxford) 2012;51:93240.
39 Wakisaka S, Suzuki N, Takeno M et al. Involvement of simultaneous multiple transcription factor expression, including cAMP responsive element binding protein and OCT-1, for synovial cell outgrowth in patients with rheumatoid arthritis. Ann Rheum Dis 1998;57:48794.
40 Takeba Y, Suzuki N, Wakisaka S et al. Involvement of cAMP responsive element binding protein (CREB) in the synovial cell hyperfunction in patients with rheumatoid arthritis. Clin Exp Rheumatol 2000;18:4755.
41 Zhang Y, Wang Z, Li X, Magnuson NS. Pim kinase-dependent inhibition of c-Myc degradation. Oncogene 2008;27:480919.
42 Pap T, Nawrath M, Heinrich J et al. Cooperation of Rasand c-Myc-dependent pathways in regulating the growth and invasiveness of synovial fibroblasts in rheumatoid arthritis. Arthritis Rheum 2004;50:2794802.
43 Laragione T, Gulko PS. mTOR regulates the invasive properties of synovial fibroblasts in rheumatoid arthritis. Mol Med 2010;16:3528.
44 Yin G, Li Y, Yang M, Cen XM, Xie QB. Pim-2/mTORC1 pathway shapes inflammatory capacity in rheumatoid arthritis synovial cells exposed to lipid peroxidations. Biomed Res Int 2015;2015:240210.
45 Garcia-Vicuna R, Gomez-Gaviro MV, Dominguez-Luis MJ et al. CC and CXC chemokine receptors mediate migration, proliferation, and matrix metalloproteinase production by fibroblast-like synoviocytes from rheumatoid arthritis patients. Arthritis Rheum 2004;50:386677.