Rheumatoid arthritis (RA) fibroblast-like synoviocytes (FLS) display unique aggressive behavior, invading the articular cartilage and promoting inflammation. Using an integrative analysis of RA risk alleles, the transcriptome and methylome in RA FLS, we recently identified the limb bud and heart development (LBH) gene as a key dysregulated gene in RA and other autoimmune diseases. Although some evidence suggests that LBH could modulate the cell cycle, the precise mechanism is unknown and its impact on inflammation in vivo has not been defined. Our cell cycle analysis studies show that LBH deficiency in FLS leads to S-phase arrest and failure to progress through the cell cycle. LBH-deficient FLS had increased DNA damage and reduced expression of the catalytic subunit of DNA polymerase α. Decreased DNA polymerase α was followed by checkpoint arrest due to phosphorylation of checkpoint kinase 1. Because DNA fragments can increase arthritis severity in preclinical models, we then explored the effect of LBH deficiency in the K/BxN serum transfer model. Lbh knockout exacerbated disease severity, which is associated with elevated levels of IL-1β and checkpoint kinase 1 phosphorylation. These studies indicate that LBH deficiency induces S-phase arrest that, in turn, exacerbates inflammation. Because LBH gene variants are associated with type I diabetes mellitus, systemic lupus erythematosus, RA, and celiac disease, these results suggest a general mechanism that could contribute to immune-mediated diseases.
Rheumatoid arthritis (RA) is a chronic inflammatory disease characterized by synovial infiltration by lymphocytes and macrophages, hyperplasia of the synovial lining, and destruction of cartilage and bone (1, 2). Current treatment strategies include the use of highly targeted biologics and, more recently, Janus kinase inhibitors. Despite improved outcomes, many patients do not respond to the available therapies and new approaches are needed (3, 4).
Fibroblast-like synoviocytes (FLS) are key players in rheumatoid synovial pathology and joint destruction through the production of cytokines and extracellular matrix–degrading enzymes (5). These synovial intimal lining cells contribute to cartilage damage and display a unique invasive phenotype. The aggressive behavior is maintained in tissue culture as well as in vivo transplantation into immunodeficient mice, indicating that RA FLS are imprinted (6–8). New therapies that target FLS could be beneficial, but the molecular mechanisms responsible for the abnormal phenotype are still poorly understood.
To address this, we recently performed an integrative analysis of methylome, transcriptome, and sequence variations in RA FLS (9). One of the genes identified through this analysis, limb bud and heart development (LBH) gene, encodes a transcription cofactor that regulates cell proliferation and differentiation during development and is linked to cardiac, skeletal, and mammary gland abnormalities in partial trisomy 2p syndrome (10–16). Several LBH polymorphisms have been associated with autoimmune diseases, including RA, celiac disease, and systemic lupus erythematosus, indicating a key role regulating immune function (17–20). We then showed that the RA-associated single-nucleotide polymorphism (SNP) decreases LBH transcription, and interacts with adjacent differentially methylated loci in regulatory regions, and that low levels of LBH alter the cell cycle (10, 21).
In the current study, we investigated the molecular mechanism of cell cycle regulation by LBH. We discovered that LBH deficiency causes S-phase arrest by decreasing the expression of the catalytic subunit of DNA polymerase α, increasing DNA damage, and activating the S-phase checkpoint. Because SNPs that lead to low levels of LBH are associated with RA, we evaluated Lbh-deficient mice and showed that a lack of LBH leads to S-phase arrest and increases inflammatory arthritis severity. These findings suggest that low LBH expression increases DNA damage, alters cell cycle kinetics, and enhances synovial inflammation.
Materials and Methods
FLS and culture conditions
This study was approved by the Institutional Review Board of the University of California, San Diego School of Medicine, and informed consent was obtained from all participants. Synovial tissue was obtained from patients with osteoarthritis and RA at the time of total joint replacement or synovectomy, as previously described. The diagnosis of RA conformed to American College of Rheumatology 1987 or 2010 criteria (22, 23). Enzymatic disaggregation of synovium was performed as previously described (21). Cells were allowed to adhere overnight and nonadherent cells were removed. Adherent FLS were split at 1:3 when 70–80% confluent and used from passages four through seven, when they are a homogeneous population of fibroblasts. Primary FLS were cultured (at 5% CO2, 37°C) in DMEM supplemented with l-glutamine, gentamicin, penicillin/streptomycin, and 10% heat-inactivated FBS (24). For synchronization experiments, cells were serum starved for 2 d and transfected with small interfering RNA (siRNA) in DMEM containing 1% FBS and supplements, followed by addition of 10% FBS.
Abs and reagents
The Abs and reagents included: anti–phosphorylated checkpoint 1 (CHK1) (S345), anti-CHK1, anti-phosphorylated CHK2 (T68), anti-CHK2, anti–phosphorylated cyclin-dependent kinase 2 (CDK2) (T160), anti-POLA1, anti-cyclin D1, anti-cyclin E1 (Cell Signaling Technology), anti-CDK2, anti-cyclin A, anti-GAPDH (Santa Cruz Biotechnology), anti-Rb (BD Pharmingen), anti-LBH, or anti–β-actin Abs (Sigma), and HRP-conjugated goat anti-rabbit or mice IgG (Cell Signaling Technology) as secondary Ab. Hydroxyurea (HU) was purchased from Sigma-Aldrich.
Cell cycle analysis
RA FLS were washed and resuspended in cold PBS, fixed in 70% ethanol, permeabilized, and stained with propidium iodide/0.1% Triton X-100 solution supplemented with RNase A. The cells were analyzed by flow cytometry using an LSR II Flow Cytometer (BD Biosciences). Data were analyzed using FlowJo software (Tree Star) and Watson Pragmatic algorithm, and the percentages of cells in the S phase were calculated. EdU assay was performed using Click-iT plus EdU Alexa Fluor 488 flow cytometry assay kit and FxCycle violet (Thermo Fisher Scientific) according to the manufacturer’s instructions.
Western blot analysis
Western blot analysis was performed as previously described (25). Protein was extracted from human FLS using a lysis buffer [150 mM NaCl, 50 mM Tris, 1% Triton X-100, 100 mM NaF, 100 mM β-glycerophosphate, protease inhibitor mixture (Roche), and phosphatase inhibitor mixture II and III (Sigma-Aldrich)]. Blots were developed using an Immun-Star WesternC Chemiluminescence kit, and analyzed using a VersaDoc imaging system and Quantity One software (Bio-Rad).
Gene knockdown was performed as previously described using the Amaxa system (10). Although siRNA could have off-target effects, we used scrambled siRNA as a control. In addition, our previous transcriptome studies showed that the only gene to be consistently and significantly suppressed across all cell lines by LBH siRNA was LBH (10). siRNA for LBH and control siRNA (ON-TARGET plus nontargeting pool) were purchased from Dharmacon. siRNAs (400 nM) were introduced into cultured cells using the Human Dermal Fibroblast Nucleofector Kit (Lonza).
Quantitative real-time PCR
Total RNA was isolated from FLS using RNASTAT-60 and reverse transcribed (Applied Biosystems, Foster City, CA). The cDNA served as template for amplification by quantitative real-time PCR (qPCR) using TaqMan Gene Expression assays (StepOnePlus Instruments; Applied Biosystems). The cycle threshold values were normalized to ACTB or 18S as previously described (26).
Alkaline comet assay
Cells were prepared by mixing 4 × 105 cells with 1 ml of LMAgarose. The cell suspension was layered on a glass microscope slide, and the slide was placed in lysis buffer at 4°C overnight. Electrophoresis was performed at 1 V per cm in an electrophoresis buffer for 40 min. Thereafter, the slide was rinsed with water and 70% ethanol, and stained with SYBR Gold for 30 min. The DNA damage levels were analyzed, and the slides were imaged using a fluorescence microscope. The comet assay is used to determine the tail moment measures of the smallest detectable size of migrating DNA (i.e., the comet’s tail length) and quantifies the number of fragments (i.e., the intensity of DNA in the tail). The Olive Tail Moment is calculated by measuring the [(Tail mean) − (Head mean) × percentage of DNA in the Tail]/100.
K/BxN serum transfer model of arthritis
Generation and phenotype of Rosa-26-Cre; Lbhloxp/loxp mice was previously described (27). Rosa-26-Cre; Lbh+/+ and matched Rosa-26-Cre; Lbhloxp/loxp littermate mice were injected i.p. with 80 μl of pooled adult K/BxN mouse serum on days 0 and 2. Arthritis was evaluated for each paw as described previously (28). For tissue protein assays, snap-frozen ankle joints were homogenized in a lysis buffer and assayed by Western blotting. The other hind paws were fixed in 10% formalin and decalcified in EDTA. For tissue qPCR assays, RNA was isolated from snap-frozen ankle joints by RNeasy Lipid Tissue Mini kit (Qiagen) and qPCR was performed.
The two-way ANOVA and two-tailed paired or unpaired t test were performed using GraphPad Prism software. A comparison was considered significant if the p value was < 0.05.
LBH deficiency blocks S-phase progression
To define the role of LBH in the cell cycle, we evaluated the effect of LBH deficiency on cell cycle progression by flow cytometry. siRNA-mediated LBH deficiency (LBHlow) was induced and cell cycle analysis was performed on synchronized FLS using additional methods and more detailed kinetics than previously reported. Based on flow cytometry studies to evaluate DNA content, control FLS progressed past the G1 checkpoint into the S phase, and then rapidly entered G2 when cultured in serum (Fig. 1A–C). In contrast, LBH-deficient cells remained in the S phase, with delayed progression through the cell cycle (Fig. 1A, 1B). The effect of LBH deficiency on the cell cycle persisted with S-phase arrest noted for at least 72 h, although cells slowly entered G2/M toward the end of the culture period (Fig. 1C, 1D). These data suggest that LBH is required for efficient progression through the S phase. To confirm the cell cycle data with a second technique, FLS were labeled with the thymidine analog EdU to quantify DNA synthesis in control and LBHlow cells. LBH deficiency increased the population of nonreplicating (EdU-negative) cells with DNA content corresponding to the S phase after 48 h (Fig. 1E, 1F).
LBH regulation of cell cycle proteins
Western blot analysis was then performed to determine the mechanism of LBH-mediated cell cycle regulation. The initial experiments evaluated the effect of LBH deficiency on CDK2 and demonstrated increased phosphorylation at the T160 site (Fig. 2A, 2B). Cyclin A, which activates CDK2 during the S phase, was increased in LBHlow cells compared with control (Fig. 2A, 2B). Finally, we evaluated the effect of LBH deficiency on retinoblastoma protein (Rb), which is a substrate for CDK2-cyclin A or cyclin E complex. As shown in Fig. 2, Rb in LBHlow cells was hyper-phosphorylated at 38 and 48 h compared with control cells that had progressed through S, G2, and M phases (Fig. 2C; p < 0.004 comparing LBHlow to control cells). We did not find a significant effect on cyclin D1 and E1 expression in LBHlow cells (Fig. 2D).
LBH deficiency increases accumulation of DNA damage
S-phase arrest is often due to DNA replication stress and damage. To address this possibility, we explored whether LBH deficiency allows DNA damage to accumulate. We used a comet assay that quantifies broken DNA fragments or damaged DNA migration away from the cell nucleus (see 2Materials and Methods). As shown in Fig. 3A, LBHlow FLS had comets with longer tails, whereas control cells had relatively shorter tails. There was a significant increase in the olive tail moment and tail length in LBHlow cells, indicating a significant increase in the amount of DNA damage in LBHlow cells compared with control (Fig. 3B).
LBH deficiency activates S-phase checkpoint
Evidence of S-phase arrest and DNA damage in LBHlow cells led us to assess the activation status of the intra–S-phase checkpoint that allows DNA repair and prevents propagation of mutated or fragmented DNA. The checkpoint kinases, CHK1 and CHK2, are activated in response to DNA damage or stalled replication forks during the S phase. Fig. 3C shows that CHK1 phosphorylation increased 4-fold more in LBH-deficient cells than control cells within 24 h. Phospho-CHK1 accumulation occurred before the increase in the percentage of cells in the S phase was detected in LBHlow cells (Fig. 3C–E). However, CHK2 was not activated (Fig. 3C, 3D).
LBH deficiency delays replication by decreasing DNA polymerase α
Three main polymerases are responsible for repair and propagation in the S phase and were then assayed in control and LBHlow cells. LBH deficiency reduced POLA1 mRNA by nearly 40% (Fig. 4A) and POLA1 protein by 86% (p < 0.01) compared with control (Fig. 4B). However, POLD1 and POLE were not affected by LBH silencing (Fig. 4A). We then determined whether POLA1 regulation is specific for DNA damage associated with LBH deficiency or whether the general S-phase checkpoint activation can participate. FLS were exposed to HU, which is a DNA replication inhibitor that decreases the elongation and initiation phases of replication and triggers the intra–S-phase checkpoint. The percentage of cells in the S phase decreased from 16 to 4% in control cells, whereas the percentage of HU-treated FLS in S phase increased from 2 to 22% from 24 to 38 h (Supplemental Fig. 1A). HU treatment significantly increased CHK1 phosphorylation (Supplemental Fig. 1B). However, HU treatment did not reduce POLA1 (Supplemental Fig. 1C).
LBH deficiency increases arthritis severity
Low LBH expression and functional SNPs that decrease LBH expression are associated with RA (17–21). Based on those studies and our data indicating that LBH regulates the cell cycle, we hypothesized that LBH deficiency would exacerbate inflammatory arthritis. To examine the contribution of LBH in the pathogenesis of RA, passive K/BxN arthritis was studied in wild type (WT) and LBH-deficient mice (27). Arthritis severity was significantly increased in the Lbh−/− mice, primarily in the later stages of the model (Fig. 5A). In addition, articular IL-1β gene expression was significantly higher in the joints of the Lbh−/− mice with arthritis compared with WT (Lbh+/+) (Fig. 5B). There was a trend toward increased expression of IL-6 and MMP-3 in Lbh−/− mice compared with WT, which did not reach statistical significance. To determine whether S-phase arrest occurred in the arthritic mice, Western blot analysis was performed to quantify phospho-CHK1 levels. There was a significant increase in ankle CHK1 phosphorylation in Lbh−/− mice compared with WT (Fig. 5C). These results confirm that LBH deficiency led to exacerbation of joint inflammation and that DNA damage and S-phase checkpoint activation occur in vivo.
RA FLS display an aggressive phenotype and invade into the articular extracellular matrix. However, no current therapies target this cell lineage in inflammatory arthritis. To address this unmet need, we previously integrated RA risk alleles, differentially methylated genes, and differentially expressed genes to identify key pathogenic pathways in RA (10). LBH, a gene encoding a highly conserved transcription coregulator, emerged from this analysis and is particularly interesting because it is associated with several autoimmune diseases, including systemic lupus erythematosus, type I diabetes, RA, and celiac disease. Subsequent studies showed that a causative mutation in RA is located in an upstream enhancer and decreases LBH gene transcription. Furthermore, this SNP occurs adjacent to a nearby differentially methylated locus in RA (21). We hypothesized that the disease-associated polymorphism in regulatory regions decreases LBH expression and contributes to RA pathogenesis.
The precise function of LBH, especially in immune-mediated diseases, has been elusive. Forced expression of LBH in embryos increases cardiomyocyte growth but blocks cell differentiation and alters bone formation (11, 14). The phenotype of Lbh knockout mice is relatively mild, with reduced breast epithelial progenitor cell proliferation (16). LBH has been implicated in Wnt signaling (29) and possibly MAP kinase signaling (13, 30). However, we did not observe these effects in RA FLS (10). Our recent functional studies showed that LBH expression in FLS is induced by platelet-derived growth factor and that low levels of LBH expression alter cell cycle dynamics. However, the precise impact on cell proliferation was not defined. In the current study, we explored the mechanisms of cell cycle regulation by LBH and its impact on genomic integrity.
Our initial interpretation of cell cycle data suggested that LBH can potentially increase progression through the G1 checkpoint. More detailed kinetic analysis confirmed normal progression through G1, but also showed that LBH-deficient cells accumulate in S phase. Delay in the cell cycle was associated with Rb hyper-phosphorylation, CDK2 activation and elevated cyclin A expression. Recent evidence suggests that cyclin A-CDK2 complex activation is important for drug-induced S-phase arrest and that hyper-phosphorylated Rb is associated with an accelerated S-phase entry (31, 32). However, we observed that the cyclin A-CDK2 complex is activated in LBHlow cells after S-phase arrest occurs, which implicates alternative mechanisms.
The lack of an effect on cyclins differs from the previous reports indicating that low LBH promotes cyclin D1 and E1 expression in nasopharyngeal cancer cells, leading to G1-S transition (33). In contrast to transformed cells, LBH silencing did not affect these cyclins in FLS. The differences could be due to culture conditions, variations resulting from the biology of cell lineages, or, most likely, the role of LBH in primary nontransformed cells compared with prior studies using malignant cells.
Subsequent studies were designed to understand how LBH regulates S-phase arrest in FLS. Previous studies in human cancer cell lines showed that DNA damage activates the S-phase checkpoint, which enables DNA repair (34–37). To test this hypothesis, we showed that LBH deficiency in FLS leads to accumulation of DNA damage followed by CHK1 phosphorylation and S-phase checkpoint activation. These observations suggest that LBH could play a role in genomic surveillance and that low levels prevent cells with damaged DNA from progressing through the cell cycle.
Accumulation of DNA damage in LBH-deficient cells led us to evaluate whether DNA synthesis is impaired. LBH deficiency selectively reduced POLA1 expression, but not other DNA polymerases. The effect was selective for endogenous DNA damage because it was not observed when S-phase checkpoint activation was initiated by HU. POLA1 initiates de novo DNA synthesis at replication forks and discontinuously replicates DNA during lagging strand DNA synthesis (38, 39). The other DNA polymerases appear to replicate lagging and leading DNA synthesis. Limiting lagging strand DNA synthesis due to LBH deficiency might cause the replication fork to stall, leading to fork collapse, which is consistent with activation of CHK1 but not CHK2 (34). This result is also consistent with reports that POLA1 is a component of the signaling that activates the S-phase checkpoint through CHK1 activation (40–42). Several studies have demonstrated that LBH may be a transcriptional cofactor, although the transcription factor partners are not known (11–13).
Because low LBH is associated with RA and altered cell cycle dynamics, we explored whether Lbh deficiency promotes inflammation in a serum transfer arthritis model. This hypothesis was supported by previous studies showing that DNA fragment accumulation increases arthritis severity (43). Lbh deficiency in this model increased arthritis severity, especially in the later stages, and was associated with higher IL-1β expression. The effect is relatively modest, but is reproducible and statistically significant. The degree of synovitis is similar to other target genes that increase arthritis severity, including p53, p38, and Bax/Bim (44–46). Given the relatively modest effect of LBH SNPs on the odds ratio for autoimmune disease susceptibility, it is not surprising that the effect is subtle. Increased CHK1 phosphorylation was observed and suggests that S-phase arrest occurs in vivo and exacerbates inflammation. It is also interesting to speculate that persistent DNA damage caused by low LBH might contribute to accumulation of somatic mutations in RA as described by several groups (47–50).
The K/BxN serum transfer arthritis model is dependent on FLS and innate immunity, especially IL-1β (51, 52). The model also involves other cells that contribute to arthritis progression, including macrophages, neutrophils, and osteoclasts. Understanding the relative role of FLS will involve future studies on conditional knockout mice when they are available.
In conclusion, we identified LBH as a critical gene that regulates immune-mediated diseases. The mechanism of action is related to the altered cell cycle through accumulation of DNA damage and S-phase arrest in the presence of continued cyclin A–CDK2 activity. The data suggest that the LBH pathway participates in RA and could be mined for potential new therapeutic targets.
This work was supported by the Rheumatology Research Foundation (Disease Targeted Research), National Institutes of Health (National Institute of Arthritis and Musculoskeletal and Skin Diseases) Grant R01AR065466 (to G.S.F.), and National Institutes of Health Grant R01 GM113256 (to K.J.B.).
The online version of this article contains supplemental material.
Abbreviations used in this article:
The authors have no financial conflicts of interest.