- Email: [email protected]
A pro-inflammatory role for A20 and ABIN family proteins in human fibroblast-like synoviocytes in rheumatoid arthritis
Immunology Letters 141 (2011) 246–253
Contents lists available at SciVerse ScienceDirect
Immunology Letters journal homepage: www.elsevier.com/locate/immlet
A pro-inflammatory role for A20 and ABIN family proteins in human fibroblast-like synoviocytes in rheumatoid arthritis Hideya Igarashi a , Ayano Yahagi a , Taro Saika a , Jun Hashimoto b , Tetsuya Tomita b , Hideki Yoshikawa b , Katsuhiko Ishihara a,∗ a b
Department of Immunology and Molecular Genetics, Kawasaki Medical School, 577 Matsushima, Kurashiki City, Okayama 701-0192, Japan Department of Orthopaedics, Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita City, Osaka 565-0871, Japan
a r t i c l e
i n f o
Article history: Received 22 June 2011 Received in revised form 26 October 2011 Accepted 31 October 2011 Available online 9 November 2011 Keywords: Human fibroblast-like synoviocytes TNF␣ NF-B A20 ABINs
a b s t r a c t Circuit of chronic inflammation in the joints of rheumatoid arthritis (RA) starts from the production of inflammatory cytokines by fibroblast-like synoviocytes (FLS) stimulated by TNF␣ produced by inflammatory cells mainly composed of macrophages. In this context, TNF␣/NF-B pathway plays an essential role for the transcription of pro-inflammatory cytokines. Here we show that the kinetics of pro-inflammatory cytokine genes induced by TNF␣ in FLS from RA was synchronized with that of A20, ABIN1, and ABIN3 that have been thought as negative regulators for NF-B activation. Furthermore, based on this finding, we could tentatively categorize the RA-FLS into two groups; TNF␣ low-responder and high-responder FLS. The high responders that have abundant mRNA levels of NF-B inhibitory molecules were also accompanied with the marked induction of the pro-inflammatory cytokines by the stimulation with TNF␣. The low responders RA-FLS did not show this property, nor did FLS from osteoarthritis. Phosphorylation dependent degradation of IB␣ as well as NF-B activation upon stimulation with TNF␣ was significantly enhanced in the high-responder FLS lines. Surprisingly, single transfection of each NF-B inhibitor was enough to facilitate the transcription of pro-inflammatory cytokines, suggesting that there is an unknown pro-inflammatory function for A20 and ABIN family proteins in RA-FLS. © 2011 Elsevier B.V. All rights reserved.
1. Introduction TNF␣ is considered as most appropriate target for the treatment of rheumatoid arthritis (RA), because TNF␣ is located at the up-stream of all other pro-inflammatory cytokines [1]. Therefore, many biological modifiers against TNF␣ such as infliximab, adalimumab or etanercept have been developed and exert dramatic improvement for the prognosis of RA patients [2]. On the other hand, there are some patients who show the resistance to these biological modifiers. One considerable reason for this is the predominance of other inflammatory cytokine than TNF␣, such as IL-6 or IL-1, in the RA patients receiving anti-TNF therapy. The other is the different reactivity of fibroblast-like synoviocytes (FLS) to TNF␣ existing in the microenvironment of the inflammatory lesion in each patient. Our previous analysis suggested that transformed RA-FLS lines have heterogeneity in the expression levels of pro-inflammatory cytokines and metalloproteinases (manuscript in preparation). Then, we reported another example of heterogeneity in RA-FLS characterized by the ectopic expression of activation
∗ Corresponding author. Tel.: +81 86 462 1111; fax: +81 86 464 1187. E-mail address: [email protected] (K. Ishihara). 0165-2478/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.imlet.2011.10.011
induced cytidine deaminase (AID) together with mutations in p53, which could cause generation of a subtype of RA-FLS with tumorlike phenotype [3]. Here we show that reactivity to TNF␣ could be added as an attribute depicting the variations among RA-FLS lines. TNF␣ signal is transduced through the canonical pathway in which NF-B, a latent transcription factor, plays an essential role. In RA, NF-B regulates production of pro-inflammatory cytokines such as TNF␣, IL-1 and IL-6, cell cycle progression, cell survival, adhesion, and inhibition of apoptosis of FLS [4–6]. Constitutive activation of NF-B in some autoimmune disease including RA has been reported [7]. One possible mechanism in this case is a defect or insufficient activity of physiological NF-B inhibitory pathway such as A20 or ABIN family protein involved in negative feedback loop. A20 (TNFAIP3) is originally identified as inducible zinc finger protein in human umbilical vein endothelial cell lines by the stimulation with TNF␣ [8]. ABIN (A20 binding inhibitor of NF-B activation, also called as TNIP; TNFAIP3 interacting protein) family members including ABIN-1, -2 and -3 are identified as A20 associated molecules. Both A20 and ABIN proteins were reported to suppress NF-B activation evidenced by their overexpression and/or knock-down in the human carcinoma cells, A431 and DU145 over-expressing EGFR [9]. A20 may exert this suppressive effect for NF-B through the regulation of ubiquitin status
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
of signaling molecules with its dual enzymatic activities; ubiquitination and deubiquitination [10]. On the other hand, ABINs do not have enzymatic activity but common structures, ABIN homology domain (AHD) and ubiquitin binding domains [11]. AHD is essential to inhibit NF-B activation by recruiting A20 to signaling modules, suggesting that ABINs have function as an adaptor. ABIN1 knockout embryo died due to the hypersensitivity to TNF␣ [12]. Enforced expression of human ABIN-3 in human monocytic cells suppressed the cytoplasmic degradation of IB␣, the activation of NF-B, and the induction of pro-inflammatory genes [13]. In addition, it has been reported that the polymorphisms in A20 gene and its nearest downstream gene at 6q23 are associated with the susceptibility of RA [14,15]. However, still little is known about the precise mechanism how they negatively regulate NF-B activation. In this study, we found that primary RA-FLS lines could be divided into two groups with a novel criterion. The RA-FLS in one group produce the abundant mRNA of pro-inflammatory cytokines by responding to TNF␣, whereas those in another not. The kinetics of transcription of A20 and ABINs were parallel to those of pro-inflammatory cytokines in the former group with enhanced activation of NF-B. Furthermore, single transfection of each A20, ABIN-1 or ABIN-3 into RA-FLS showed the augmented expression of pro-inflammatory cytokine genes without affecting the suppressive effect on the canonical NF-B pathway. On the contrary, knocking-down of each A20, ABIN-1 or ABIN-3 in RA-FLS weakened the induction of pro-inflammatory cytokines by the stimulation with TNF␣. These data demonstrate a certain subtype of RA-FLS that is responsible for enhanced production of pro-inflammatory cytokines dependent on TNF␣ through unknown positive regulatory function of A20/ABINs. 2. Materials and methods 2.1. Cells and cell culture Primary human FLS cell lines were established from the synovial tissues of RA and osteoarthritis (OA) patients described as previously [3]. The tissues were obtained with informed consent of the patients following the procedure approved by the institutional review boards for ethics at the Faculty of Medicine, Osaka University (#340-1). MH7A, a RA-FLS line transformed with SV40TAg, was obtained from Riken cell bank (Ibaraki, Japan). The transformed and original non-transformed primary FLS cells were maintained in DMEM (Sigma–Aldrich, USA) supplemented with 10% heatinactivated FCS with antibiotics. For the cell stimulation, FLS were cultured in DMEM complete medium with various concentrations of TNF␣ (Pepro Tech Inc., USA) for various times as indicated in the text. This study protocol was approved by the institutional review boards for ethics at the Faculty of Medicine, Osaka University (#340-1) and the Kawasaki Medical School (#291).
247
2.3. Western blot analysis The cell lysates were prepared in 1% TNE lysis buffer and protein concentration was measured by Bradford method. The 25 g equivalent whole cell crudes were resuspended in SDS-PAGE sample buffer, boiled for 5 min, resolved on SDS-PAGE, transferred electrically to PVDF membrane (Immun-Blot, Bio-Rad, USA) and subjected to western blot analysis with signal enhancing kit provided from SantaCruz Inc. Membrane was probed with anti-IB␣, antiphospho-IB␣ (CST, USA), and anti--actin Ab (Santa Cruz, USA). 2.4. EMSA assay One million primary FLS were stimulated with TNF␣ at the concentration of 10 ng/ml in 12 well culture plate for 3 h, then nuclear protein was extracted with NE-PER Nuclear and Cytoplasmic Extraction Kit (Pierce, USA). EMSA was performed with purified nuclear extracts using LightShift Chemiluminescent EMSA Kit (Pierce, USA). 3 -Biotinylated NF-B probe is as follows: AGTTGAGGGGACTTTCCCAGGC. Oct-1 was used for loading control and normalization. 3 -Biotinylated Oct-1 probe is as follows: TGTCGAATGCAAATCACTAGAA. 2.5. Construction of GFP fusion proteins, transfection, and cell sorting The expression vectors for GFP fusion protein of A20 and ABINs was prepared as follows. The total RNAs form RA-FLS were extracted with TRIzol (Invitrogen, USA). Random hexamer-primed cDNAs were prepared with ReverTra Ace (TOYOBO, Japan) and then amplified by PCR with the following primer pairs containing appropriate restriction enzyme sites at 5 -end and 3 -end: A20 (5 ) (5 -CGGAATTCTATGGCTGAACAAGTCCTTCC-3 ) and (3 ) (5 -ACGCGTCGACTTAGCCATACATCTGCTTGA-3 ) primers, ABIN-1 (5 ) (5 -CCCAAGCTTCGATGGAAGGGAGAGGACCG-3 ) and (3 ) (5 -CGGGATCCTCACTGAGGCCCCTCACGGTCA-3 ) primers, ABIN-3 (5 ) (5 -GAAGATCTATGGCACATTTTGTACAAGG-3 ) and (3 ) (5 ACGCGTCGACCTGAGGATAAACCATTTG-3 ) primers. The cloning product was once inserted into pBluescript, and the sequence of the inserted fragment was confirmed by sequencing with T7 or T3 primer. Thereafter, the insert was introduced into pEGFP-C1 expression vector (Clontech, USA) to add EGFP at the N-termini of the proteins. Then, these vectors were transferred into RA-FLS by nucleofector device (Amaxa, Germany) under epidermal fibroblast conditions. Twenty four hours later, the localization of GFP-fusion proteins was observed under the fluorescence microscope Power IX 81 (Olympus, Japan). For the study of the effect from over-doses of A20, ABIN-1 and ABIN-3, 24 h after gene transfer, GFP− or GFP+ cells were purified by sorting with FACSAria (BD, Australia) and the transcription of inflammatory cytokines in each fraction was examined by real-time PCR. The sorted cells were plated with fresh medium and cultured for 24 h before analyses for morphology, biochemistry, and gene expression.
2.2. Quantitative real-time RT-PCR 2.6. Quantitative analysis of IL-6 The total RNAs were extracted with TRIzol (Invitrogen, USA), according to the manufacturer’s instructions. Random hexamerprimed cDNAs were prepared using ReverTra Ace qPCR RT Kit (TOYOBO, Japan). The specific primers for each gene were designed using Universal Primer Design Tools available on the web site of Roche Company. The PCR in real time with the SYBR Green 1 dye was carried out by the 7300 Real-Time PCR System (PE Applied Biosystems, USA). The expression levels of target cDNAs were normalized to the endogenous transcription levels of the human ˇ-actin.
The twenty thousand cells sorted as GFP− or GFP+ fraction described above were cultured in 96 well plate and the amount of IL-6 secreted into the culture supernatant was measured by Instant ELISA kit (eBioscience, USA) at the time of 24 h post-sorting. 2.7. Gene silencing by siRNA transfection The siRNA (MISSION® siRNA) for each gene were purchased from Sigma. The siRNA equivalent to 133 pmol was transfected into
248
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
MH7A with X-tremeGENE siRNA Transfection Reagent according to manufacture’s protocol (Roche, Swiss). The efficiency of gene silencing was evaluated by real-time PCR.
3. Results
of the transcription of TNF˛ and IL-6 genes, and transient suppression followed in the transcription of three pro-inflammatory cytokines. Such kinetics is compatible with the prevailing notion that A20/ABINs have negative regulator for TNF␣/NF-B pathway. Among ABIN families, the extent of ABIN-2 induction was relatively modest (less than 2.2 fold) in comparison with other ABIN members, thus we focused on ABIN-1 and -3 at following experiments.
3.1. A20, ABIN-1, ABIN-3 but not ABIN-2 is induced by the stimulation with TNF˛ in primary FLS cell lines derived from RA 3.2. RA-FLS cell lines differentially respond to TNF˛ In order to know the biological response of FLS to TNF␣, inducible expression of TNF˛, IL-1ˇ, and IL-6 genes as the output of TNF␣ effect was evaluated by quantitative-PCR in three different RA-FLS lines tentatively designated as RA-FLS1, RA-FLS2, and RAFLS3. The kinetics study revealed bi-phasic responses, early (3–6 h) and late (24–48 h later), although the time point and magnitude of peak response are variable depending on the lines (Fig. S1A). RA-FLS1 and RA-FLS2 have peaks of cytokine gene induction in the late phases, whereas RA-FLS3 has peaks of TNF˛ and IL-6 gene induction in the early phases. The cytokine with the highest fold induction was IL-6 in RA-FLS1, while that is IL-1ˇ in RA-FLS2 and RA-FLS3. These data indicated that there are some variations in the responses to TNF␣ among primary RA-FLS lines. However, the bi-phasic responses in the cytokine gene induction, which are common features among three RA-FLS lines, indicate that negative feedback regulation to the initial activation of NF-B by TNF␣ is invariable among distinct RA-FLS lines. To evaluate the suppressive pathway for NF-B activation at the downstream of TNF␣ signaling, we also assessed the transcription of A20 and ABIN family (Fig. S1B). The kinetics of A20/ABINs transcription also exhibited biphasic responses. The early peak of A20/ABINs transcription was synchronized with that
Since our previous analyses suggested the variations of TNF␣ responses among RA-FLS cell lines, we examined the panels of RA-FLS cell lines for their induction levels of pro-inflammatory cytokines and A20/ABINs at 24 h after TNF␣ stimulation. We chose the late phase, because we speculated that cytokine gene induction overcoming the negative feedback might be involved in the perpetuation of inflammation in RA. The results indicated that RA-FLS lines can be obviously categorized into two types based on the responsiveness to TNF␣; One group consists of five FLS lines, aligned from the far left in the figure, which synthesize the negligible induction (less than 1.5 fold) of mRNA for NF-B inhibitory molecules upon TNF␣ stimulation compared with unstimulated samples. In contrast, another group consists of six FLS lines showing much higher induction levels (more than 4 fold of A20; 5 fold of ABIN1; 9 fold of ABIN3) (Fig. 1A). Curiously the group with abundant mRNA levels of NF-B inhibitory molecules was also accompanied with the marked induction of the pro-inflammatory cytokines by the stimulation with TNF␣ (more than 2.0 fold of TNF˛; 30 fold of IL-1ˇ; 20 fold of IL-6) (Fig. 1B and Fig. S1). Based on this observation, we tentatively categorize the RA-FLS into two groups; TNF␣ low-responder (designated as L1 to L5) and high-responder (designated as H1 to
Fig. 1. Inducible expression of NF-B inhibitory molecules and pro-inflammatory cytokines in non-transformed primary FLS stimulated with 10 ng/ml of TNF␣ for 24 h. (A) The mRNA of NF-B inhibitory molecules were quantified (relative quantification) by real-time PCR and normalized to house-keeping gene, ˇ-actin. The fold expression over that of unstimulated sample in each line was calculated by CT method. (B) The mRNA of inflammatory cytokines were quantified (relative quantification) by real-time PCR and normalized to housekeeping gene, ˇ-actin. The fold induction over the expression level of unstimulated sample in each FLS line was calculated by CT method.
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
249
Fig. 2. Comparison of NF-B activities between TNF␣ high-responder and low-responder primary FLS. (A) High- or low-responder primary FLS lines were stimulated with 10 ng/ml of TNF␣ for 3 h. TNF␣-induced IB␣ phosphorylation and degradation was examined by western blot analysis with phospho-IB␣ and IB␣ specific mAb respectively. (B) TNF␣-induced NF-B activation was confirmed by EMSA with biotin labeled NF-B or Oct-1 binding probe. The intensities of the corresponding bands were measured on the densitometer and normalized to that of -actin (A) or Oct-1 (B), and shown as the mean ± SEM with P values calculated by Student’s t-test at the right.
H6) FLS. Of note, we found super-responder cells, typically as H2 and H3, showing the hyper induction of those cytokines (more than 10 fold of TNF˛; 150 fold of IL-1ˇ; 400 fold of IL-6) among highresponders. No difference was found between low-responder and high-responder regarding the expression level of both TNFR1 and TNFR2, and also clinical characteristics, such as medication and data of laboratory examination containing complete blood count, and serum levels of CRP, MMP-3 and RF (data not shown). In addition, OA-FLS lines showed the similar pattern to low-responder of RAFLS, suggesting that the high-responder of RA-FLS is unique in RA and reflecting some cell-intrinsic abnormalities regarding the TNF␣ signaling pathways (Fig. S2). 3.3. TNF˛ high-responder RA-FLS lines showed the facilitated activation of NF-B To clarify the molecular mechanisms for the enhanced gene expression in high-responder RA-FLS, we examined the TNF␣ signaling. Phosphorylation and degradation of IB␣ upon stimulation with TNF␣ was significantly enhanced in the high-responder FLS lines, which could cause the increased activation of NF-B leading to inductive production of pro-inflammatory cytokines (relative intensity of phosphorylated bands compared with -actin bands; average ± SD low-responders 0.29 ± 0.11 vs. highresponders 1.01 ± 0.28, p < 0.01) (Fig. 2A). Actually, the significant elevation of NF-B activity in TNF␣ high-responder FLS lines than that in low-responders was confirmed by EMSA (Fig. 2B and Fig. S3). These data indicated that high-responder group has characteristics of the enhanced activation of NF-B in the early signaling events induced by TNF␣. 3.4. Positive effect of A20/ABINs on pro-inflammatory cytokine induction in spite of suppression of NF-B However, it seemed difficult to explain why induced A20 and ABIN family could not affect to NF-B activity even at 24 h after stimulation of TNF␣. In order to know the direct effect of A20/ABINs for the activation of NF-B and the induction of pro-inflammatory cytokines, GFP-tagged protein of each was over-expressed in RAFLS whether they can suppress the intrinsic NF-B activity and
basal expression level of pro-inflammatory cytokines. As shown in Fig. 3A, these transfection experiments revealed distinct intracellular distribution of A20/ABINs. In contrast to the diffuse distribution in both nucleus and cytoplasm by the introduction of vector alone, A20 resides not only in the nucleus but also in cytoplasm along with plasma membrane as dotted pattern. ABIN-1 was preferentially localized at peri-nuclear region. ABIN-3 accumulated in the nucleus. Similar distribution patterns were also observed in two other lines of RA-FLS (Fig. S4). These unique distributions suggest that A20 and ABIN family proteins play their functions at distinct portion in the cell. Then, GFP+ cells were enriched by cell sorting and compared to GFP− cells with respect to the transcription level of TNF␣, IL1 and IL-6, and the phosphorylation status of IB␣. The extent of gene induction by each transfection is shown in Supplementary Fig. 5A and B. Surprisingly, the single transfection of each A20/ABINs rather enhanced the transcription of pro-inflammatory cytokines that are positively regulated usually by NF-B, clearly indicating that these “inhibitors” exert positive effects on transcription of pro-inflammatory genes in RA-FLS (Fig. 3B). This unexpected effect of A20/ABINs for pro-inflammatory cytokines was finally confirmed by the enhanced production of IL-6 in the supernatant from the culture of transfectants (Fig. 3C). On the other hand, in the same cells, the phosphorylation of IB␣ reflecting the NF-B activation was significantly reduced (Fig. 3D). Furthermore, EMSA assay revealed that forced expression of A20/ABINs significantly suppressed the canonical NF-B activation pathway even after the stimulation with TNF␣, although some comparison did not reach the statistic significance, indicating that the known A20/ABINmediated inhibitory pathway, at least in the earliest step, is intact (Fig. 3E). To ensure the positive function of A20/ABINs for proinflammatory cytokine production, we tried to suppress the expression of A20/ABINs by siRNA technology in primary RA-FLS lines. However, it was not successful due to the difficulties for the efficient delivery of siRNA to primary FLSs. Thus, we used a transformed RA-FLS line, MH7A, which is more competent for this purpose. Importantly, MH7A showed the very similar phenotype with high responder RA-FLS, indicating the compatibility of this cell line as a representative of high responder RA-FLSs (Fig. S6). As expected,
250
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
Fig. 3. GFP-tagged A20 and ABIN family proteins were expressed in three different lines of RA-FLS. (A) The localization of each GFP-tagged A20 and ABIN family proteins in H4 RA-FLS observed under fluorescence microscope at 24 h post transfection. Scale bar indicates 100 m. (B) Increased expression levels of TNF˛, IL-1ˇ and IL-6 in the test fraction (GFP+ ) over-expressing A20 and ABIN family proteins compared with control fraction (GFP− ). Data of three high-responder RA-FLS lines are shown; H2 (left), H3 (middle) and H4 (right). Dotted line indicates the standard expression level “1”. (C) The amount of IL-6 in the culture supernatant at the time of 24 h post-sorting between control fraction (GFP− ) and test fraction (GFP+ ) over-expressing A20 and ABIN family proteins in H2, H3 and H4 RA-FLS was assessed by measuring the production of IL-6 on ELISA in triplicate, and shown as the mean ± SEM with P values calculated by Student’s t-test. *p < 0.05, **p < 0.01. Two independent experiments (Exp. 1 and Exp. 2) are exhibited. (D) Endogenous IB␣ phosphorylation was examined by western blot analysis with phospho-IB␣ specific mAb either in control fraction (GFP− ) or test fraction (GFP+ ) over-expressing A20, ABIN-1 and ABIN-3 in H2, H3 and H4 RA-FLS. (E) Each transfectant was unstimulated or stimulated with 10 ng/ml of TNF␣ for 3 h prior to cell harvest, then sorted into control fraction (GFP− ) and test fraction (GFP+ ). After sorting, cells were immediately lysed and nuclear protein was extracted for EMSA. The intensity of the bands corresponding to NF-B were measured on the densitometer and normalized to that of Oct-1. The NF-B/Oct-1 ratio among three high-responder (H2, H3 and H4) are shown as the mean ± SEM with P values calculated by Student’s t-test at the right.
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
251
Fig. 3. (Continued )
siRNA significantly reduced the expression level of each target genes due to the more efficient delivery of reagent in comparison with the case of primary FLS (Fig. S7). The suppressed expression of each A20, ABIN-1 and ABIN-3 resulted in the insufficient
induction of pro-inflammatory cytokines by the stimulation with TNF␣ (Fig. 4), clearly demonstrating that A20, ABIN-1, and ABIN3 have positive roles in TNF␣-dependent cytokine induction in RA-FLS.
252
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
Fig. 4. Knock-down of A20/ABINs reduced expression levels of pro-inflammatory cytokine genes in FLS stimulated with TNF␣. MH7A cells were transfected with siRNA for each A20/ABINs, then 24 h later, cells were stimulated with 10 ng/ml of TNF␣ for further 24 h. The mRNA of inflammatory cytokines were quantified (relative quantification) in triplicate by real-time PCR and normalized to housekeeping gene, ˇ-actin. The fold induction over the expression level of unmanipulated and unstimulated sample was calculated by CT method. The results are shown as the mean ± SEM with P values calculated by Student’s t-test. The statistic significance of the changes in the expression levels of cytokine genes were evaluated between MH7A cells untreated or treated with siRNA. *p < 0.01. Dotted line indicates the standard expression level “1”. The results from two independent experiments are exhibited.
4. Discussion In this study, we demonstrated that the TNF␣ mediated induction of inflammatory cytokines that are regulated by NF-B activity in RA-FLS showed the same kinetics with NF-B inhibitory molecules, A20 and ABINs not only at early phase (3H) but also at late phase (24H). It is no wonder that A20 is induced by TNF␣ stimulation, because it was originally identified as TNF␣ inducible gene in endothelial cell [8]. Likewise, there is a NF-B binding site at the promoter region of both ABIN-1 and ABIN-3 gene, and TNF␣ signals induce the expression of these genes [16,17]. Thus, at early time point after stimulation with TNF␣, it is reasonable that the expression of A20/ABINs were correlated with the induction of pro-inflammatory cytokines by NF-B. However, at later time point such as 24 h post stimulation, it does not make sense that both NF-B and its inhibitors were up-regulated simultaneously. This unexpected result drove us to think that A20 and ABINs do not work as inhibitors in RA-FLS rather may augment the production of pro-inflammatory cytokines via enhancement of NF-B activity. Importantly, single transfection of each A20, ABIN-1 and ABIN-3 was sufficient to induce the transcription of pro-inflammatory cytokine genes in RA-FLS, whereas knock-down of each molecule conversely reduced it (Figs. 3B and 4), definitively demonstrating the novel, positive regulatory roles for A20 and ABINs in terms of cytokine gene induction in the TNF␣ signaling. In spite of the positive end effect to induce cytokine gene expression, the biochemical data prevented us from clear understanding of the mechanism, because TNF␣ stimulation under the over-expression of A20/ABINs induced the simultaneous reduction of phospho-IB␣ and increase of IB␣ protein, which are usually thought to lead inhibition of NF-B activity. One possible reason for this discrepancy is the alteration of signaling pathway to induce pro-inflammatory cytokines other than the conventional
NF-B activation. For example, the enhanced recruitment of other transcription factors, such as c-Fos, c-Jun/AP-1, p53 or NF-B2, which is related to non-canonical NF-B pathway, to the promoter region of pro-inflammatory cytokines might occur by the function of A20/ABINs as an adapter protein. Indeed, these transcription factors are known to be ubiquitinated [18–21], and A20/ABINs may bind with them through its ubiquitin-binding domain, leading to successfull recruitment of those transcription factors to the promoter region of pro-inflammatory cytokines. Elsby et al. reported that the expression level of A20 in RA-FLS is relatively weaker than that in OA-FLS and the relationship to the polymorphisms at 6q23, although its difference did not reach to statistical significance [22]. Based on the known function of A20, this finding seems reasonable to account for the enhanced NF-B activation leading to production of pro-inflammatory cytokines in RA-FLS, but down-regulation of A20 does not necessarily mean NF-B activation in the synovium if the heterogeneity of FLS is considered. As we are showing here, some subtypes of RA-FLS have potential to express A20 as well as ABINs by the secondary or continuous stimulation with TNF␣ and they alter the intracellular signaling cascade, which leads to pro-inflammatory cytokine production without affecting the conventional inhibitory pathway for NF-B. One may think that the clinical backgrounds of individual patients, especially the kind of medicines used for the treatment, might affect to the response of FLS to TNF␣. However, the patients examined in this study were medicated with similar drugs, such as NSAIDS, DMARDS and corticosteroids, and no bias of the usage for these drugs was found in patients with either low-responder or high-responder FLSs. It will be informative if we could obtain and analyze the FLS from the naïve patients. Although precise mechanism is currently unknown and deserves further investigation, the results demonstrating proinflammatory roles for A20 and ABINs have brought us one of
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
possible mechanisms for chronic inflammation in RA. Furthermore, our approach to categorize the RA-FLS subset is beneficial to reveal the unexpected abnormality of RA-FLS that could cause the pathophysiology in the synovium of RA. Acknowledgments
[9]
[10]
We thank Ms. Reina Tanaka and Ms. Yuka Kenmotsu for their technical assistance. This work was supported by Research Project Grants (20-418I to H.I. and 19-406M to K.I.) from Kawasaki Medical School and Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, Culture and Technology of Japan (19590390 to. K.I.).
[11]
Appendix A. Supplementary data
[14]
[12]
[13]
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.imlet.2011.10.011.
[15]
References
[16]
[1] Taylor PC, Feldmann M. Anti-TNF biologic agents: still the therapy of choice for rheumatoid arthritis. Nat Rev Rheumatol United States 2009:578–82. [2] van Vollenhoven RF. Treatment of rheumatoid arthritis: state of the art 2009. Nat Rev Rheumatol United States 2009:531–41. [3] Igarashi H, Hashimoto J, Tomita T, Yoshikawa H, Ishihara K. TP53 mutations coincide with the ectopic expression of activation-induced cytidine deaminase in the fibroblast-like synoviocytes derived from a fraction of patients with rheumatoid arthritis. Clin Exp Immunol England 2010:71–80. [4] Handel ML, McMorrow LB, Gravallese EM. Nuclear factor-kappa B in rheumatoid synovium. Localization of p50 and p65. Arthritis Rheum United States 1995:1762–70. [5] Marok R, Winyard PG, Coumbe A, Kus ML, Gaffney K, Blades S, et al. Activation of the transcription factor nuclear factor-kappaB in human inflamed synovial tissue. Arthritis Rheum United States 1996:583–91. [6] Brown KD, Claudio E, Siebenlist U. The roles of the classical and alternative nuclear factor-kappaB pathways: potential implications for autoimmunity and rheumatoid arthritis. Arthritis Res Ther England 2008:212. [7] Okamoto T. NF-kappaB rheumatic F diseases. Endocr Metab Immune Disord Drug Targets United Arab Emirates 2006:359–72. [8] Dixit VM, Green S, Sarma V, Holzman LB, Wolf FW, O’Rourke K, et al. Tumor necrosis factor-alpha induction of novel gene products in human endothelial
[17]
[18] [19]
[20]
[21] [22]
253
cells including a macrophage-specific chemotaxin. J Biol Chem United States 1990:2973–8. Huang L, Verstrepen L, Heyninck K, Wullaert A, Revets H, De Baetselier P, et al. ABINs inhibit EGF receptor-mediated NF-kappaB activation and growth of EGF receptor-overexpressing tumour cells. Oncogene England 2008:6131– 40. Vereecke L, Beyaert R, van Loo G. The ubiquitin-editing enzyme A20 (TNFAIP3) is a central regulator of immunopathology. Trends Immunol England 2009:383–91. Wagner S, Carpentier I, Rogov V, Kreike M, Ikeda F, Lohr F, et al. Ubiquitin binding mediates the NF-kappaB inhibitory potential of ABIN proteins. Oncogene England 2008:3739–45. Oshima S, Turer EE, Callahan JA, Chai S, Advincula R, Barrera J, et al. ABIN-1 is a ubiquitin sensor that restricts cell death and sustains embryonic development. Nature England 2009:906–9. Weaver BK, Bohn E, Judd BA, Gil MP, Schreiber RD. ABIN-3, a molecular basis for species divergence in interleukin-10-induced anti-inflammatory actions. Mol Cell Biol United States 2007:4603–16. Dieguez-Gonzalez R, Calaza M, Perez-Pampin E, Balsa A, Blanco FJ, Canete JD, et al. Analysis of TNFAIP3, a feedback inhibitor of nuclear factor-kappaB and the neighbor intergenic 6q23 region in rheumatoid arthritis susceptibility. Arthritis Res Ther England 2009:R42. Orozco G, Hinks A, Eyre S, Ke X, Gibbons LJ, Bowes J, et al. Combined effects of three independent SNPs greatly increase the risk estimate for RA at 6q23. Hum Mol Genet England 2009:2693–9. Gallagher J, Howlin J, McCarthy C, Murphy EP, Bresnihan B, FitzGerald O, et al. Identification of Naf1/ABIN-1 among TNF-alpha-induced expressed genes in human synoviocytes using oligonucleotide microarrays. FEBS Lett Netherlands 2003:8–12. Verstrepen L, Adib-Conquy M, Kreike M, Carpentier I, Adrie C, Cavaillon JM, et al. Expression of the NF-kappaB inhibitor ABIN-3 in response to TNF and toll-like receptor 4 stimulation is itself regulated by NF-kappaB. J Cell Mol Med Romania 2008:316–29. Petroski MD, Deshaies RJ. Function and regulation of cullin-RING ubiquitin ligases. Nat Rev Mol Cell Biol England 2005:9–20. Vasilescu J, Zweitzig DR, Denis NJ, Smith JC, Ethier M, Haines DS, et al. The proteomic reactor facilitates the analysis of affinity-purified proteins by mass spectrometry, application for identifying ubiquitinated proteins in human cells. J Proteome Res United States 2007:298–305. Gu Q, Tan M, Sun Y. SAG/ROC2/Rbx2 is a novel activator protein-1 target that promotes c-Jun degradation and inhibits 12-O-tetradecanoylphorbol13-acetate-induced neoplastic transformation. Cancer Res United States 2007:3616–25. Lee JT, Gu W. The multiple levels of regulation by p53 ubiquitination. Cell Death Differ England 2010:86–92. Elsby LM, Orozco G, Denton J, Worthington J, Ray DW, Donn RP. Functional evaluation of TNFAIP3 (A20) in rheumatoid arthritis. Clin Exp Rheumatol Italy 2010:708–14.
Contents lists available at SciVerse ScienceDirect
Immunology Letters journal homepage: www.elsevier.com/locate/immlet
A pro-inflammatory role for A20 and ABIN family proteins in human fibroblast-like synoviocytes in rheumatoid arthritis Hideya Igarashi a , Ayano Yahagi a , Taro Saika a , Jun Hashimoto b , Tetsuya Tomita b , Hideki Yoshikawa b , Katsuhiko Ishihara a,∗ a b
Department of Immunology and Molecular Genetics, Kawasaki Medical School, 577 Matsushima, Kurashiki City, Okayama 701-0192, Japan Department of Orthopaedics, Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita City, Osaka 565-0871, Japan
a r t i c l e
i n f o
Article history: Received 22 June 2011 Received in revised form 26 October 2011 Accepted 31 October 2011 Available online 9 November 2011 Keywords: Human fibroblast-like synoviocytes TNF␣ NF-B A20 ABINs
a b s t r a c t Circuit of chronic inflammation in the joints of rheumatoid arthritis (RA) starts from the production of inflammatory cytokines by fibroblast-like synoviocytes (FLS) stimulated by TNF␣ produced by inflammatory cells mainly composed of macrophages. In this context, TNF␣/NF-B pathway plays an essential role for the transcription of pro-inflammatory cytokines. Here we show that the kinetics of pro-inflammatory cytokine genes induced by TNF␣ in FLS from RA was synchronized with that of A20, ABIN1, and ABIN3 that have been thought as negative regulators for NF-B activation. Furthermore, based on this finding, we could tentatively categorize the RA-FLS into two groups; TNF␣ low-responder and high-responder FLS. The high responders that have abundant mRNA levels of NF-B inhibitory molecules were also accompanied with the marked induction of the pro-inflammatory cytokines by the stimulation with TNF␣. The low responders RA-FLS did not show this property, nor did FLS from osteoarthritis. Phosphorylation dependent degradation of IB␣ as well as NF-B activation upon stimulation with TNF␣ was significantly enhanced in the high-responder FLS lines. Surprisingly, single transfection of each NF-B inhibitor was enough to facilitate the transcription of pro-inflammatory cytokines, suggesting that there is an unknown pro-inflammatory function for A20 and ABIN family proteins in RA-FLS. © 2011 Elsevier B.V. All rights reserved.
1. Introduction TNF␣ is considered as most appropriate target for the treatment of rheumatoid arthritis (RA), because TNF␣ is located at the up-stream of all other pro-inflammatory cytokines [1]. Therefore, many biological modifiers against TNF␣ such as infliximab, adalimumab or etanercept have been developed and exert dramatic improvement for the prognosis of RA patients [2]. On the other hand, there are some patients who show the resistance to these biological modifiers. One considerable reason for this is the predominance of other inflammatory cytokine than TNF␣, such as IL-6 or IL-1, in the RA patients receiving anti-TNF therapy. The other is the different reactivity of fibroblast-like synoviocytes (FLS) to TNF␣ existing in the microenvironment of the inflammatory lesion in each patient. Our previous analysis suggested that transformed RA-FLS lines have heterogeneity in the expression levels of pro-inflammatory cytokines and metalloproteinases (manuscript in preparation). Then, we reported another example of heterogeneity in RA-FLS characterized by the ectopic expression of activation
∗ Corresponding author. Tel.: +81 86 462 1111; fax: +81 86 464 1187. E-mail address: [email protected] (K. Ishihara). 0165-2478/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.imlet.2011.10.011
induced cytidine deaminase (AID) together with mutations in p53, which could cause generation of a subtype of RA-FLS with tumorlike phenotype [3]. Here we show that reactivity to TNF␣ could be added as an attribute depicting the variations among RA-FLS lines. TNF␣ signal is transduced through the canonical pathway in which NF-B, a latent transcription factor, plays an essential role. In RA, NF-B regulates production of pro-inflammatory cytokines such as TNF␣, IL-1 and IL-6, cell cycle progression, cell survival, adhesion, and inhibition of apoptosis of FLS [4–6]. Constitutive activation of NF-B in some autoimmune disease including RA has been reported [7]. One possible mechanism in this case is a defect or insufficient activity of physiological NF-B inhibitory pathway such as A20 or ABIN family protein involved in negative feedback loop. A20 (TNFAIP3) is originally identified as inducible zinc finger protein in human umbilical vein endothelial cell lines by the stimulation with TNF␣ [8]. ABIN (A20 binding inhibitor of NF-B activation, also called as TNIP; TNFAIP3 interacting protein) family members including ABIN-1, -2 and -3 are identified as A20 associated molecules. Both A20 and ABIN proteins were reported to suppress NF-B activation evidenced by their overexpression and/or knock-down in the human carcinoma cells, A431 and DU145 over-expressing EGFR [9]. A20 may exert this suppressive effect for NF-B through the regulation of ubiquitin status
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
of signaling molecules with its dual enzymatic activities; ubiquitination and deubiquitination [10]. On the other hand, ABINs do not have enzymatic activity but common structures, ABIN homology domain (AHD) and ubiquitin binding domains [11]. AHD is essential to inhibit NF-B activation by recruiting A20 to signaling modules, suggesting that ABINs have function as an adaptor. ABIN1 knockout embryo died due to the hypersensitivity to TNF␣ [12]. Enforced expression of human ABIN-3 in human monocytic cells suppressed the cytoplasmic degradation of IB␣, the activation of NF-B, and the induction of pro-inflammatory genes [13]. In addition, it has been reported that the polymorphisms in A20 gene and its nearest downstream gene at 6q23 are associated with the susceptibility of RA [14,15]. However, still little is known about the precise mechanism how they negatively regulate NF-B activation. In this study, we found that primary RA-FLS lines could be divided into two groups with a novel criterion. The RA-FLS in one group produce the abundant mRNA of pro-inflammatory cytokines by responding to TNF␣, whereas those in another not. The kinetics of transcription of A20 and ABINs were parallel to those of pro-inflammatory cytokines in the former group with enhanced activation of NF-B. Furthermore, single transfection of each A20, ABIN-1 or ABIN-3 into RA-FLS showed the augmented expression of pro-inflammatory cytokine genes without affecting the suppressive effect on the canonical NF-B pathway. On the contrary, knocking-down of each A20, ABIN-1 or ABIN-3 in RA-FLS weakened the induction of pro-inflammatory cytokines by the stimulation with TNF␣. These data demonstrate a certain subtype of RA-FLS that is responsible for enhanced production of pro-inflammatory cytokines dependent on TNF␣ through unknown positive regulatory function of A20/ABINs. 2. Materials and methods 2.1. Cells and cell culture Primary human FLS cell lines were established from the synovial tissues of RA and osteoarthritis (OA) patients described as previously [3]. The tissues were obtained with informed consent of the patients following the procedure approved by the institutional review boards for ethics at the Faculty of Medicine, Osaka University (#340-1). MH7A, a RA-FLS line transformed with SV40TAg, was obtained from Riken cell bank (Ibaraki, Japan). The transformed and original non-transformed primary FLS cells were maintained in DMEM (Sigma–Aldrich, USA) supplemented with 10% heatinactivated FCS with antibiotics. For the cell stimulation, FLS were cultured in DMEM complete medium with various concentrations of TNF␣ (Pepro Tech Inc., USA) for various times as indicated in the text. This study protocol was approved by the institutional review boards for ethics at the Faculty of Medicine, Osaka University (#340-1) and the Kawasaki Medical School (#291).
247
2.3. Western blot analysis The cell lysates were prepared in 1% TNE lysis buffer and protein concentration was measured by Bradford method. The 25 g equivalent whole cell crudes were resuspended in SDS-PAGE sample buffer, boiled for 5 min, resolved on SDS-PAGE, transferred electrically to PVDF membrane (Immun-Blot, Bio-Rad, USA) and subjected to western blot analysis with signal enhancing kit provided from SantaCruz Inc. Membrane was probed with anti-IB␣, antiphospho-IB␣ (CST, USA), and anti--actin Ab (Santa Cruz, USA). 2.4. EMSA assay One million primary FLS were stimulated with TNF␣ at the concentration of 10 ng/ml in 12 well culture plate for 3 h, then nuclear protein was extracted with NE-PER Nuclear and Cytoplasmic Extraction Kit (Pierce, USA). EMSA was performed with purified nuclear extracts using LightShift Chemiluminescent EMSA Kit (Pierce, USA). 3 -Biotinylated NF-B probe is as follows: AGTTGAGGGGACTTTCCCAGGC. Oct-1 was used for loading control and normalization. 3 -Biotinylated Oct-1 probe is as follows: TGTCGAATGCAAATCACTAGAA. 2.5. Construction of GFP fusion proteins, transfection, and cell sorting The expression vectors for GFP fusion protein of A20 and ABINs was prepared as follows. The total RNAs form RA-FLS were extracted with TRIzol (Invitrogen, USA). Random hexamer-primed cDNAs were prepared with ReverTra Ace (TOYOBO, Japan) and then amplified by PCR with the following primer pairs containing appropriate restriction enzyme sites at 5 -end and 3 -end: A20 (5 ) (5 -CGGAATTCTATGGCTGAACAAGTCCTTCC-3 ) and (3 ) (5 -ACGCGTCGACTTAGCCATACATCTGCTTGA-3 ) primers, ABIN-1 (5 ) (5 -CCCAAGCTTCGATGGAAGGGAGAGGACCG-3 ) and (3 ) (5 -CGGGATCCTCACTGAGGCCCCTCACGGTCA-3 ) primers, ABIN-3 (5 ) (5 -GAAGATCTATGGCACATTTTGTACAAGG-3 ) and (3 ) (5 ACGCGTCGACCTGAGGATAAACCATTTG-3 ) primers. The cloning product was once inserted into pBluescript, and the sequence of the inserted fragment was confirmed by sequencing with T7 or T3 primer. Thereafter, the insert was introduced into pEGFP-C1 expression vector (Clontech, USA) to add EGFP at the N-termini of the proteins. Then, these vectors were transferred into RA-FLS by nucleofector device (Amaxa, Germany) under epidermal fibroblast conditions. Twenty four hours later, the localization of GFP-fusion proteins was observed under the fluorescence microscope Power IX 81 (Olympus, Japan). For the study of the effect from over-doses of A20, ABIN-1 and ABIN-3, 24 h after gene transfer, GFP− or GFP+ cells were purified by sorting with FACSAria (BD, Australia) and the transcription of inflammatory cytokines in each fraction was examined by real-time PCR. The sorted cells were plated with fresh medium and cultured for 24 h before analyses for morphology, biochemistry, and gene expression.
2.2. Quantitative real-time RT-PCR 2.6. Quantitative analysis of IL-6 The total RNAs were extracted with TRIzol (Invitrogen, USA), according to the manufacturer’s instructions. Random hexamerprimed cDNAs were prepared using ReverTra Ace qPCR RT Kit (TOYOBO, Japan). The specific primers for each gene were designed using Universal Primer Design Tools available on the web site of Roche Company. The PCR in real time with the SYBR Green 1 dye was carried out by the 7300 Real-Time PCR System (PE Applied Biosystems, USA). The expression levels of target cDNAs were normalized to the endogenous transcription levels of the human ˇ-actin.
The twenty thousand cells sorted as GFP− or GFP+ fraction described above were cultured in 96 well plate and the amount of IL-6 secreted into the culture supernatant was measured by Instant ELISA kit (eBioscience, USA) at the time of 24 h post-sorting. 2.7. Gene silencing by siRNA transfection The siRNA (MISSION® siRNA) for each gene were purchased from Sigma. The siRNA equivalent to 133 pmol was transfected into
248
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
MH7A with X-tremeGENE siRNA Transfection Reagent according to manufacture’s protocol (Roche, Swiss). The efficiency of gene silencing was evaluated by real-time PCR.
3. Results
of the transcription of TNF˛ and IL-6 genes, and transient suppression followed in the transcription of three pro-inflammatory cytokines. Such kinetics is compatible with the prevailing notion that A20/ABINs have negative regulator for TNF␣/NF-B pathway. Among ABIN families, the extent of ABIN-2 induction was relatively modest (less than 2.2 fold) in comparison with other ABIN members, thus we focused on ABIN-1 and -3 at following experiments.
3.1. A20, ABIN-1, ABIN-3 but not ABIN-2 is induced by the stimulation with TNF˛ in primary FLS cell lines derived from RA 3.2. RA-FLS cell lines differentially respond to TNF˛ In order to know the biological response of FLS to TNF␣, inducible expression of TNF˛, IL-1ˇ, and IL-6 genes as the output of TNF␣ effect was evaluated by quantitative-PCR in three different RA-FLS lines tentatively designated as RA-FLS1, RA-FLS2, and RAFLS3. The kinetics study revealed bi-phasic responses, early (3–6 h) and late (24–48 h later), although the time point and magnitude of peak response are variable depending on the lines (Fig. S1A). RA-FLS1 and RA-FLS2 have peaks of cytokine gene induction in the late phases, whereas RA-FLS3 has peaks of TNF˛ and IL-6 gene induction in the early phases. The cytokine with the highest fold induction was IL-6 in RA-FLS1, while that is IL-1ˇ in RA-FLS2 and RA-FLS3. These data indicated that there are some variations in the responses to TNF␣ among primary RA-FLS lines. However, the bi-phasic responses in the cytokine gene induction, which are common features among three RA-FLS lines, indicate that negative feedback regulation to the initial activation of NF-B by TNF␣ is invariable among distinct RA-FLS lines. To evaluate the suppressive pathway for NF-B activation at the downstream of TNF␣ signaling, we also assessed the transcription of A20 and ABIN family (Fig. S1B). The kinetics of A20/ABINs transcription also exhibited biphasic responses. The early peak of A20/ABINs transcription was synchronized with that
Since our previous analyses suggested the variations of TNF␣ responses among RA-FLS cell lines, we examined the panels of RA-FLS cell lines for their induction levels of pro-inflammatory cytokines and A20/ABINs at 24 h after TNF␣ stimulation. We chose the late phase, because we speculated that cytokine gene induction overcoming the negative feedback might be involved in the perpetuation of inflammation in RA. The results indicated that RA-FLS lines can be obviously categorized into two types based on the responsiveness to TNF␣; One group consists of five FLS lines, aligned from the far left in the figure, which synthesize the negligible induction (less than 1.5 fold) of mRNA for NF-B inhibitory molecules upon TNF␣ stimulation compared with unstimulated samples. In contrast, another group consists of six FLS lines showing much higher induction levels (more than 4 fold of A20; 5 fold of ABIN1; 9 fold of ABIN3) (Fig. 1A). Curiously the group with abundant mRNA levels of NF-B inhibitory molecules was also accompanied with the marked induction of the pro-inflammatory cytokines by the stimulation with TNF␣ (more than 2.0 fold of TNF˛; 30 fold of IL-1ˇ; 20 fold of IL-6) (Fig. 1B and Fig. S1). Based on this observation, we tentatively categorize the RA-FLS into two groups; TNF␣ low-responder (designated as L1 to L5) and high-responder (designated as H1 to
Fig. 1. Inducible expression of NF-B inhibitory molecules and pro-inflammatory cytokines in non-transformed primary FLS stimulated with 10 ng/ml of TNF␣ for 24 h. (A) The mRNA of NF-B inhibitory molecules were quantified (relative quantification) by real-time PCR and normalized to house-keeping gene, ˇ-actin. The fold expression over that of unstimulated sample in each line was calculated by CT method. (B) The mRNA of inflammatory cytokines were quantified (relative quantification) by real-time PCR and normalized to housekeeping gene, ˇ-actin. The fold induction over the expression level of unstimulated sample in each FLS line was calculated by CT method.
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
249
Fig. 2. Comparison of NF-B activities between TNF␣ high-responder and low-responder primary FLS. (A) High- or low-responder primary FLS lines were stimulated with 10 ng/ml of TNF␣ for 3 h. TNF␣-induced IB␣ phosphorylation and degradation was examined by western blot analysis with phospho-IB␣ and IB␣ specific mAb respectively. (B) TNF␣-induced NF-B activation was confirmed by EMSA with biotin labeled NF-B or Oct-1 binding probe. The intensities of the corresponding bands were measured on the densitometer and normalized to that of -actin (A) or Oct-1 (B), and shown as the mean ± SEM with P values calculated by Student’s t-test at the right.
H6) FLS. Of note, we found super-responder cells, typically as H2 and H3, showing the hyper induction of those cytokines (more than 10 fold of TNF˛; 150 fold of IL-1ˇ; 400 fold of IL-6) among highresponders. No difference was found between low-responder and high-responder regarding the expression level of both TNFR1 and TNFR2, and also clinical characteristics, such as medication and data of laboratory examination containing complete blood count, and serum levels of CRP, MMP-3 and RF (data not shown). In addition, OA-FLS lines showed the similar pattern to low-responder of RAFLS, suggesting that the high-responder of RA-FLS is unique in RA and reflecting some cell-intrinsic abnormalities regarding the TNF␣ signaling pathways (Fig. S2). 3.3. TNF˛ high-responder RA-FLS lines showed the facilitated activation of NF-B To clarify the molecular mechanisms for the enhanced gene expression in high-responder RA-FLS, we examined the TNF␣ signaling. Phosphorylation and degradation of IB␣ upon stimulation with TNF␣ was significantly enhanced in the high-responder FLS lines, which could cause the increased activation of NF-B leading to inductive production of pro-inflammatory cytokines (relative intensity of phosphorylated bands compared with -actin bands; average ± SD low-responders 0.29 ± 0.11 vs. highresponders 1.01 ± 0.28, p < 0.01) (Fig. 2A). Actually, the significant elevation of NF-B activity in TNF␣ high-responder FLS lines than that in low-responders was confirmed by EMSA (Fig. 2B and Fig. S3). These data indicated that high-responder group has characteristics of the enhanced activation of NF-B in the early signaling events induced by TNF␣. 3.4. Positive effect of A20/ABINs on pro-inflammatory cytokine induction in spite of suppression of NF-B However, it seemed difficult to explain why induced A20 and ABIN family could not affect to NF-B activity even at 24 h after stimulation of TNF␣. In order to know the direct effect of A20/ABINs for the activation of NF-B and the induction of pro-inflammatory cytokines, GFP-tagged protein of each was over-expressed in RAFLS whether they can suppress the intrinsic NF-B activity and
basal expression level of pro-inflammatory cytokines. As shown in Fig. 3A, these transfection experiments revealed distinct intracellular distribution of A20/ABINs. In contrast to the diffuse distribution in both nucleus and cytoplasm by the introduction of vector alone, A20 resides not only in the nucleus but also in cytoplasm along with plasma membrane as dotted pattern. ABIN-1 was preferentially localized at peri-nuclear region. ABIN-3 accumulated in the nucleus. Similar distribution patterns were also observed in two other lines of RA-FLS (Fig. S4). These unique distributions suggest that A20 and ABIN family proteins play their functions at distinct portion in the cell. Then, GFP+ cells were enriched by cell sorting and compared to GFP− cells with respect to the transcription level of TNF␣, IL1 and IL-6, and the phosphorylation status of IB␣. The extent of gene induction by each transfection is shown in Supplementary Fig. 5A and B. Surprisingly, the single transfection of each A20/ABINs rather enhanced the transcription of pro-inflammatory cytokines that are positively regulated usually by NF-B, clearly indicating that these “inhibitors” exert positive effects on transcription of pro-inflammatory genes in RA-FLS (Fig. 3B). This unexpected effect of A20/ABINs for pro-inflammatory cytokines was finally confirmed by the enhanced production of IL-6 in the supernatant from the culture of transfectants (Fig. 3C). On the other hand, in the same cells, the phosphorylation of IB␣ reflecting the NF-B activation was significantly reduced (Fig. 3D). Furthermore, EMSA assay revealed that forced expression of A20/ABINs significantly suppressed the canonical NF-B activation pathway even after the stimulation with TNF␣, although some comparison did not reach the statistic significance, indicating that the known A20/ABINmediated inhibitory pathway, at least in the earliest step, is intact (Fig. 3E). To ensure the positive function of A20/ABINs for proinflammatory cytokine production, we tried to suppress the expression of A20/ABINs by siRNA technology in primary RA-FLS lines. However, it was not successful due to the difficulties for the efficient delivery of siRNA to primary FLSs. Thus, we used a transformed RA-FLS line, MH7A, which is more competent for this purpose. Importantly, MH7A showed the very similar phenotype with high responder RA-FLS, indicating the compatibility of this cell line as a representative of high responder RA-FLSs (Fig. S6). As expected,
250
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
Fig. 3. GFP-tagged A20 and ABIN family proteins were expressed in three different lines of RA-FLS. (A) The localization of each GFP-tagged A20 and ABIN family proteins in H4 RA-FLS observed under fluorescence microscope at 24 h post transfection. Scale bar indicates 100 m. (B) Increased expression levels of TNF˛, IL-1ˇ and IL-6 in the test fraction (GFP+ ) over-expressing A20 and ABIN family proteins compared with control fraction (GFP− ). Data of three high-responder RA-FLS lines are shown; H2 (left), H3 (middle) and H4 (right). Dotted line indicates the standard expression level “1”. (C) The amount of IL-6 in the culture supernatant at the time of 24 h post-sorting between control fraction (GFP− ) and test fraction (GFP+ ) over-expressing A20 and ABIN family proteins in H2, H3 and H4 RA-FLS was assessed by measuring the production of IL-6 on ELISA in triplicate, and shown as the mean ± SEM with P values calculated by Student’s t-test. *p < 0.05, **p < 0.01. Two independent experiments (Exp. 1 and Exp. 2) are exhibited. (D) Endogenous IB␣ phosphorylation was examined by western blot analysis with phospho-IB␣ specific mAb either in control fraction (GFP− ) or test fraction (GFP+ ) over-expressing A20, ABIN-1 and ABIN-3 in H2, H3 and H4 RA-FLS. (E) Each transfectant was unstimulated or stimulated with 10 ng/ml of TNF␣ for 3 h prior to cell harvest, then sorted into control fraction (GFP− ) and test fraction (GFP+ ). After sorting, cells were immediately lysed and nuclear protein was extracted for EMSA. The intensity of the bands corresponding to NF-B were measured on the densitometer and normalized to that of Oct-1. The NF-B/Oct-1 ratio among three high-responder (H2, H3 and H4) are shown as the mean ± SEM with P values calculated by Student’s t-test at the right.
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
251
Fig. 3. (Continued )
siRNA significantly reduced the expression level of each target genes due to the more efficient delivery of reagent in comparison with the case of primary FLS (Fig. S7). The suppressed expression of each A20, ABIN-1 and ABIN-3 resulted in the insufficient
induction of pro-inflammatory cytokines by the stimulation with TNF␣ (Fig. 4), clearly demonstrating that A20, ABIN-1, and ABIN3 have positive roles in TNF␣-dependent cytokine induction in RA-FLS.
252
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
Fig. 4. Knock-down of A20/ABINs reduced expression levels of pro-inflammatory cytokine genes in FLS stimulated with TNF␣. MH7A cells were transfected with siRNA for each A20/ABINs, then 24 h later, cells were stimulated with 10 ng/ml of TNF␣ for further 24 h. The mRNA of inflammatory cytokines were quantified (relative quantification) in triplicate by real-time PCR and normalized to housekeeping gene, ˇ-actin. The fold induction over the expression level of unmanipulated and unstimulated sample was calculated by CT method. The results are shown as the mean ± SEM with P values calculated by Student’s t-test. The statistic significance of the changes in the expression levels of cytokine genes were evaluated between MH7A cells untreated or treated with siRNA. *p < 0.01. Dotted line indicates the standard expression level “1”. The results from two independent experiments are exhibited.
4. Discussion In this study, we demonstrated that the TNF␣ mediated induction of inflammatory cytokines that are regulated by NF-B activity in RA-FLS showed the same kinetics with NF-B inhibitory molecules, A20 and ABINs not only at early phase (3H) but also at late phase (24H). It is no wonder that A20 is induced by TNF␣ stimulation, because it was originally identified as TNF␣ inducible gene in endothelial cell [8]. Likewise, there is a NF-B binding site at the promoter region of both ABIN-1 and ABIN-3 gene, and TNF␣ signals induce the expression of these genes [16,17]. Thus, at early time point after stimulation with TNF␣, it is reasonable that the expression of A20/ABINs were correlated with the induction of pro-inflammatory cytokines by NF-B. However, at later time point such as 24 h post stimulation, it does not make sense that both NF-B and its inhibitors were up-regulated simultaneously. This unexpected result drove us to think that A20 and ABINs do not work as inhibitors in RA-FLS rather may augment the production of pro-inflammatory cytokines via enhancement of NF-B activity. Importantly, single transfection of each A20, ABIN-1 and ABIN-3 was sufficient to induce the transcription of pro-inflammatory cytokine genes in RA-FLS, whereas knock-down of each molecule conversely reduced it (Figs. 3B and 4), definitively demonstrating the novel, positive regulatory roles for A20 and ABINs in terms of cytokine gene induction in the TNF␣ signaling. In spite of the positive end effect to induce cytokine gene expression, the biochemical data prevented us from clear understanding of the mechanism, because TNF␣ stimulation under the over-expression of A20/ABINs induced the simultaneous reduction of phospho-IB␣ and increase of IB␣ protein, which are usually thought to lead inhibition of NF-B activity. One possible reason for this discrepancy is the alteration of signaling pathway to induce pro-inflammatory cytokines other than the conventional
NF-B activation. For example, the enhanced recruitment of other transcription factors, such as c-Fos, c-Jun/AP-1, p53 or NF-B2, which is related to non-canonical NF-B pathway, to the promoter region of pro-inflammatory cytokines might occur by the function of A20/ABINs as an adapter protein. Indeed, these transcription factors are known to be ubiquitinated [18–21], and A20/ABINs may bind with them through its ubiquitin-binding domain, leading to successfull recruitment of those transcription factors to the promoter region of pro-inflammatory cytokines. Elsby et al. reported that the expression level of A20 in RA-FLS is relatively weaker than that in OA-FLS and the relationship to the polymorphisms at 6q23, although its difference did not reach to statistical significance [22]. Based on the known function of A20, this finding seems reasonable to account for the enhanced NF-B activation leading to production of pro-inflammatory cytokines in RA-FLS, but down-regulation of A20 does not necessarily mean NF-B activation in the synovium if the heterogeneity of FLS is considered. As we are showing here, some subtypes of RA-FLS have potential to express A20 as well as ABINs by the secondary or continuous stimulation with TNF␣ and they alter the intracellular signaling cascade, which leads to pro-inflammatory cytokine production without affecting the conventional inhibitory pathway for NF-B. One may think that the clinical backgrounds of individual patients, especially the kind of medicines used for the treatment, might affect to the response of FLS to TNF␣. However, the patients examined in this study were medicated with similar drugs, such as NSAIDS, DMARDS and corticosteroids, and no bias of the usage for these drugs was found in patients with either low-responder or high-responder FLSs. It will be informative if we could obtain and analyze the FLS from the naïve patients. Although precise mechanism is currently unknown and deserves further investigation, the results demonstrating proinflammatory roles for A20 and ABINs have brought us one of
H. Igarashi et al. / Immunology Letters 141 (2011) 246–253
possible mechanisms for chronic inflammation in RA. Furthermore, our approach to categorize the RA-FLS subset is beneficial to reveal the unexpected abnormality of RA-FLS that could cause the pathophysiology in the synovium of RA. Acknowledgments
[9]
[10]
We thank Ms. Reina Tanaka and Ms. Yuka Kenmotsu for their technical assistance. This work was supported by Research Project Grants (20-418I to H.I. and 19-406M to K.I.) from Kawasaki Medical School and Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, Culture and Technology of Japan (19590390 to. K.I.).
[11]
Appendix A. Supplementary data
[14]
[12]
[13]
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.imlet.2011.10.011.
[15]
References
[16]
[1] Taylor PC, Feldmann M. Anti-TNF biologic agents: still the therapy of choice for rheumatoid arthritis. Nat Rev Rheumatol United States 2009:578–82. [2] van Vollenhoven RF. Treatment of rheumatoid arthritis: state of the art 2009. Nat Rev Rheumatol United States 2009:531–41. [3] Igarashi H, Hashimoto J, Tomita T, Yoshikawa H, Ishihara K. TP53 mutations coincide with the ectopic expression of activation-induced cytidine deaminase in the fibroblast-like synoviocytes derived from a fraction of patients with rheumatoid arthritis. Clin Exp Immunol England 2010:71–80. [4] Handel ML, McMorrow LB, Gravallese EM. Nuclear factor-kappa B in rheumatoid synovium. Localization of p50 and p65. Arthritis Rheum United States 1995:1762–70. [5] Marok R, Winyard PG, Coumbe A, Kus ML, Gaffney K, Blades S, et al. Activation of the transcription factor nuclear factor-kappaB in human inflamed synovial tissue. Arthritis Rheum United States 1996:583–91. [6] Brown KD, Claudio E, Siebenlist U. The roles of the classical and alternative nuclear factor-kappaB pathways: potential implications for autoimmunity and rheumatoid arthritis. Arthritis Res Ther England 2008:212. [7] Okamoto T. NF-kappaB rheumatic F diseases. Endocr Metab Immune Disord Drug Targets United Arab Emirates 2006:359–72. [8] Dixit VM, Green S, Sarma V, Holzman LB, Wolf FW, O’Rourke K, et al. Tumor necrosis factor-alpha induction of novel gene products in human endothelial
[17]
[18] [19]
[20]
[21] [22]
253
cells including a macrophage-specific chemotaxin. J Biol Chem United States 1990:2973–8. Huang L, Verstrepen L, Heyninck K, Wullaert A, Revets H, De Baetselier P, et al. ABINs inhibit EGF receptor-mediated NF-kappaB activation and growth of EGF receptor-overexpressing tumour cells. Oncogene England 2008:6131– 40. Vereecke L, Beyaert R, van Loo G. The ubiquitin-editing enzyme A20 (TNFAIP3) is a central regulator of immunopathology. Trends Immunol England 2009:383–91. Wagner S, Carpentier I, Rogov V, Kreike M, Ikeda F, Lohr F, et al. Ubiquitin binding mediates the NF-kappaB inhibitory potential of ABIN proteins. Oncogene England 2008:3739–45. Oshima S, Turer EE, Callahan JA, Chai S, Advincula R, Barrera J, et al. ABIN-1 is a ubiquitin sensor that restricts cell death and sustains embryonic development. Nature England 2009:906–9. Weaver BK, Bohn E, Judd BA, Gil MP, Schreiber RD. ABIN-3, a molecular basis for species divergence in interleukin-10-induced anti-inflammatory actions. Mol Cell Biol United States 2007:4603–16. Dieguez-Gonzalez R, Calaza M, Perez-Pampin E, Balsa A, Blanco FJ, Canete JD, et al. Analysis of TNFAIP3, a feedback inhibitor of nuclear factor-kappaB and the neighbor intergenic 6q23 region in rheumatoid arthritis susceptibility. Arthritis Res Ther England 2009:R42. Orozco G, Hinks A, Eyre S, Ke X, Gibbons LJ, Bowes J, et al. Combined effects of three independent SNPs greatly increase the risk estimate for RA at 6q23. Hum Mol Genet England 2009:2693–9. Gallagher J, Howlin J, McCarthy C, Murphy EP, Bresnihan B, FitzGerald O, et al. Identification of Naf1/ABIN-1 among TNF-alpha-induced expressed genes in human synoviocytes using oligonucleotide microarrays. FEBS Lett Netherlands 2003:8–12. Verstrepen L, Adib-Conquy M, Kreike M, Carpentier I, Adrie C, Cavaillon JM, et al. Expression of the NF-kappaB inhibitor ABIN-3 in response to TNF and toll-like receptor 4 stimulation is itself regulated by NF-kappaB. J Cell Mol Med Romania 2008:316–29. Petroski MD, Deshaies RJ. Function and regulation of cullin-RING ubiquitin ligases. Nat Rev Mol Cell Biol England 2005:9–20. Vasilescu J, Zweitzig DR, Denis NJ, Smith JC, Ethier M, Haines DS, et al. The proteomic reactor facilitates the analysis of affinity-purified proteins by mass spectrometry, application for identifying ubiquitinated proteins in human cells. J Proteome Res United States 2007:298–305. Gu Q, Tan M, Sun Y. SAG/ROC2/Rbx2 is a novel activator protein-1 target that promotes c-Jun degradation and inhibits 12-O-tetradecanoylphorbol13-acetate-induced neoplastic transformation. Cancer Res United States 2007:3616–25. Lee JT, Gu W. The multiple levels of regulation by p53 ubiquitination. Cell Death Differ England 2010:86–92. Elsby LM, Orozco G, Denton J, Worthington J, Ray DW, Donn RP. Functional evaluation of TNFAIP3 (A20) in rheumatoid arthritis. Clin Exp Rheumatol Italy 2010:708–14.