Regulation of type I interferon signature by VGLL3 in the fibroblast-like synoviocytes of rheumatoid arthritis patients via targeting the Hippo pathway

Synovial tissues were collected from ten RA patients (all females, mean age 60.2 ± 11.8 years) and five osteoarthritis (OA) patients (all females, mean age 64.2 ± 7.1 years) diagnosed according to the1987 ACR classification criteria for RA [29] or 1986 revised ACR classification criteria for knee OA [30] during total knee arthroplasty in the department of orthopedic surgery of Shanghai Jiao Tong University Affiliated Sixth People’s Hospital. All the patients were not receiving any antirheumatic drug therapy at the time of surgery. Informed consent was obtained from all patients before sample collection. This study complied with the Declaration of Helsinki (1964), and the research was approved by the Ethic Committee of Shanghai Jiao Tong University Affiliated Sixth People’s Hospital. FLS were isolated from synovial tissues by digesting with 1 mg/mL type II collagenase (Merck Millipore, MA, USA) as previously described [31]. Cells were collected and cultured in Dulbecco’s Modified Eagle Medium (DMEM, Hyclone, UT, USA) containing 10% fetal bovine serum (FBS, Gibco, NY, USA). FLS at passages 3–8 were used in the following experiments.

Synovial tissues were fixed in paraformaldehyde (PFA) and embedded in paraffin. For immunohistochemistry, paraffin-embedded sections were blocked with 10% donkey serum (Jackson ImmunoResearch Labs, PA, USA) and incubated with rabbit anti-VGLL3 antibody (1:500, HPA054983, Merck Millipore, MA, USA), rabbit anti-MX1 antibody (1:100, 13750-1-AP, Proteintech, Wuhan, China), rabbit anti-STAT1 antibody (1:250, ab109320, Abcam, Cambridge, UK), and rabbit IgG (isotype control, 12-370, Merck Millipore, MA, USA) at 4 °C overnight. After washing with phosphate-buffered saline (PBS) three times, sections were incubated with horseradish peroxidase (HRP)-conjugated donkey anti-rabbit antibody (1:1000, Jackson ImmunoResearch Labs, PA, USA) and further stained with diaminobenzidine (DAB, Beyotime Biotechnology, Shanghai, China). For immunofluorescence staining, cultured FLS were fixed in PFA and permeabilized with 0.5% Triton X-100. After blocking with 10% donkey serum, cells were incubated with rabbit anti-VGLL3 antibody (1:500, ab83555, Abcam, Cambridge, UK), rabbit anti-STAT1 antibody ((1:250, ab109320, Abcam, Cambridge, UK), rabbit anti-phospho-STAT1 antibody (1:1000, 7649, Cell Signaling Technology, MA, USA), and rabbit anti-IRF3 antibody (1:100, ab76493, Abcam, Cambridge, UK) at 4°C overnight. The next day, cells were washed with PBS and incubated with Alexa Fluor 555-conjugated donkey anti-rabbit antibody (1:500, ab150130, Abcam, Cambridge, UK) and mounted with Antifade Mounting Medium with DAPI (4’,6-diamidino-2-phenylindole) (Beyotime Biotechnology, Shanghai, China). Images were captured with a microscope (Olympus, Tokyo, Japan) (magnification ×200) with an equal time of exposure.

VGLL3 siRNA or WWTR1 siRNA were transfected into RA-FLS to knock down the expression of these genes. Scramble siRNA was used as the negative control. All the siRNA were synthesized by Ribobio (Guangzhou, China). The target sequences are as follows: scramble siRNA, TTCTCCGAACGTGTCACGT; VGLL3 siRNA, GGTCAGTAGTGGATGAACA; WWTR1 siRNA, CGATGAATCAGCCTCTGAA. siRNA was transfected into FLS using Lipofectamine 2000 (Thermo Fisher Scientific, MA, USA) transfection reagent following the manufacturer’s instructions. For VGLL3 overexpression, full-length VGLL3 was cloned into lentiviral vectors (pHBLV-CMV-MCS-3FLAG-EF1-ZsGreen-T2A-PURO) via standard molecular biology techniques. Lentivirus was produced by co-transfecting 293T cells with VGLL3 lentiviral vectors or control vectors together with packaging plasmids. The lentivirus solution was concentrated by ultracentrifugation. The titers of control and overexpression lentivirus solutions were determined by infecting 293T cells. Briefly, 293T cells were seeded in a 96-well plate at a density of 10⁴ cells/well and added with gradient diluted virus solutions in different wells. The titers of viruses were calculated after 72 h with 10%-50% GFP-positive wells using the following formula:

FLS were infected with control or VGLL3-overexpressing lentiviruses at an MOI of 300 with 5μg/ml polybrene.

Total RNA was extracted by TRIzol reagent (Thermo Fisher Scientific, MA, USA) according to the manufacturer’s instructions. cDNA was generated using the ReverTra Ace qPCR RT Master Mix with gDNA Remover kit (Toyobo, Osaka, Japan). RT-PCR was conducted by 2×Hieff^TM PCR Master Mix (Yeasen, Shanghai, China) followed by agarose gel electrophoresis. qPCR was performed using SYBR Green Realtime PCR Master Mix kit (Toyobo, Osaka, Japan). The primer sequences are listed in Table 1.

The purity of extracted RNA was examined by the kaiaoK5500®Spectrophotometer (Kaiao, Beijing, China). RNA integrity and concentration were evaluated using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, CA, USA). Sequencing libraries were generated using NEBNext® Ultra™ RNA Library Prep Kit for Illumina® (#E7530L, NEB, USA) with 2 μg of total RNA per sample according to the manufacturer’s instructions. The clustering of the index-coded samples was performed on a cBot cluster generation system using HiSeq PE Cluster Kit v4-cBot-HS (Illumina, CA, USA) following the manufacturer’s recommendations. After cluster generation, the libraries were sequenced on an Illumina platform and 150 bp paired-end reads were generated. Reads Count for each gene in each sample was counted by HISAT v2.0.5, and FPKM (fragments per kilobase per million mapped reads) was then calculated to estimate the expression level of genes in each sample. DEGseq v1.18.0 was used for differential gene expression analysis between two samples with nonbiological replicates. Genes with log2 fold change (log₂(FC)) ≥ 1 and q value (false discovery rate or FDR) ≤ 0.05 were identified as differentially expressed genes (DEGs). Heatmaps for the selected DEGs were mapped by R package “pheatmap” in R studio. The expression levels of the DEGs were shown as log₂(FC). Gene set enrichment analysis (GSEA) for DEGs was performed using the RNA-seq data through WebGestalt as indicated [32], and the top 10 enriched gene sets were demonstrated.

Cells were lysed by RIPA buffer (Beyotime Biotechnology, Shanghai, China) on ice. Protein concentration was determined using a BCA protein assay kit (Beyotime Biotechnology, Shanghai, China). Denatured proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–PAGE) and electrotransferred onto polyvinylidene fluoride (PVDF) membranes (Merck Millipore, MA, USA). Membranes were blocked by 10% bovine serum albumin (BSA) and incubated at 4°C overnight with rabbit anti-VGLL3 antibody (1:500, ab83555, Abcam, Cambridge, UK), rabbit anti-STAT1 antibody (1:10000, ab109320, Abcam, Cambridge, UK), rabbit anti-MX1 antibody (1:1000, 13750-1-AP, Proteintech, Wuhan, China), rabbit anti-IRF3 (phosphor S386) antibody (1:1000, ab76493, Abcam, Cambridge, UK), rabbit anti-IRF3 antibody (1:1000, ab68481, Abcam, Cambridge, UK), rabbit anti-YAP/WWTR1 antibody (1:1000, 8418, Cell Signaling Technology, MA, USA), rabbit anti-AMOTL2 antibody (1:500, 23351-1-AP, Proteintech, Wuhan, China), mouse anti-GAPDH antibody (1:5000, T0004, Affinity, OH, USA), and mouse anti-β-Tubulin (1:5000, T0023, Affinity, OH, USA). Next, membranes were washed with Tris-buffered saline (TBS) containing 0.1% Tween 20 and incubated with HRP-conjugated goat anti-rabbit antibody (1:5000, S0001, Affinity, OH, USA) or HRP-conjugated goat anti-mouse antibody (1:5000, S0002, Affinity, OH, USA). Protein bands were detected with ECL substrate (New Cell & molecular Biotech, Suzhou, China) and analyzed using ImageJ software.

All the experiments were conducted at least three times independently. Data were analyzed by IBM SPSS Statistics 22.0 and plotted using GraphPad Prism 8 software. Data were presented as mean ± SD. Student’s t-test was used for statistical comparison between two groups. One-way analysis of variance (ANOVA) followed by Dunnett’s T3 post hoc test was adopted for multiple comparisons. p-values < 0.05 were considered statistically significant.

Immunohistochemistry staining was performed to examine the expression of VGLL3 in RA and OA synovium using anti-VGLL3 antibodies and rabbit IgG (isotype). Hematoxylin and eosin (HE) staining was conducted to evaluate the pathological changes. There were significant hyperplasia and leukocyte infiltration in the RA intimal lining layer comparing to OA, and VGLL3 expression was substantially upregulated in RA intimal lining layer (Fig. 1A). However, there was no significant difference in VGLL3 expression in the sublining layer between RA and OA (Fig 1A). Then, the FLS were isolated from RA synovium and subjected to immunofluorescence staining. The results showed that VGLL3 was expressed in both the cytoplasm and nucleus of RA-FLS (Fig. 1B). RT-PCR and qPCR analysis of RA-FLS and OA-FLS at passage 3 also indicated that the mRNA expression level of VGLL3 in RA-FLS was increased approximately by 100% compared with OA-FLS (Fig. 1C, D).

To address the contribution of VGLL3 to ISGs expression in RA-FLS, VGLL3 was overexpressed in RA-FLS via a lentiviral vector. After 4 days of transfection, Flag-VGLL3-overexpressed RA-FLS and vector-transfected RA-FLS were subjected to RNA sequencing (n = 1). The expression levels of all the DEGs were shown as log2(FC) in the heatmap. The RNA expression levels of 1047 genes in total were altered by VGLL3 overexpression, including 357 genes upregulated and 690 genes downregulated (Fig. 2A). Subsequent GSEA revealed that response to type I IFN was the most altered biological process in RA-FLS overexpressing VGLL3 (normalized enrichment score, 2.57, Fig. 2B). Differential gene expression analysis showed that many ISGs were upregulated by VGLL3 overexpression, especially the type I ISGs (Fig. 2C). Real-time PCR verified the upregulation of ISGs (e.g., MX1, OAS1, STAT1, TNFSF13B, IRF7, TLR3, and CCL5) by VGLL3 (Fig. 2D). The protein expression levels of STAT1 and MX1 were also increased by VGLL3 overexpression (Fig. 2F). To exclude the influence on ISGs expression by virus infection, VGLL3 was further silenced by transfection of synthetic siRNA in RA-FLS. The mRNA expression of multiple ISGs and protein expression of STAT1 and MX1 were significantly inhibited after VGLL3 knockdown (Fig. 2E, G).

Next, the expression of STAT1 and MX1 in the OA and RA synovium was examined via IHC staining using anti-STAT1 antibody and anti-MX1 antibody. Isotype IgG was used as the negative control. Compared to OA, STAT1 and MX1 were substantially upregulated in the RA synovium (Fig. 3A, C). In both OA and RA patients, the percentage of VGLL3-positive cells had moderate correlations with that of STAT1-positive cells (r = 0.5212, p < 0.05, Fig. 3B) and was weakly correlated with the proportion of MX1-positive cells (r = 0.2786, p < 0.05, Fig. 3D). These data indicated that VGLL3 could determine the expression of STAT1 and MX1 in RA synovium to a certain extent.

The mechanism that mediates the VGLL3-enhanced ISGs expression in RA-FLS remains unelucidated. Given that response to type I IFN was the predominantly modified biological process by VGLL3 overexpression, it was then explored whether VGLL3 changed the expression level of type I IFN. VGLL3 was overexpressed via a lentiviral vector or silenced by transfected with VGLL3 siRNA in RA-FLS. The mRNA expression of IFN-α1 (gene name, IFNA1) and IFN-β1 (gene name, IFNB1) in RA-FLS after overexpression or knockdown of VGLL3 was examined via qPCR. VGLL3 overexpression significantly increased the expression of IFN-β1 in RA-FLS, but the expression of IFN-α1 was unaffected. Correspondingly, VGLL3 knockdown considerably decreased the IFN-β1 expression level, and the expression of IFN-α1 was only moderately inhibited (Fig. 4A). These results indicated that VGLL3 might enhance the expression of ISGs in RA-FLS by facilitating IFN-β1 production. IRF3 is a master transcription factor for IFN-β expression. The nuclear translocation and phosphorylation of IRF3 were further examined by immunofluorescence staining after overexpression of VGLL3. Nuclei were stained with DAPI. The results showed that overexpression of VGLL3 markedly increased the protein level of IRF3 in the nuclei of RA-FLS (Fig. 4B), and the phosphorylation of IRF3 in RA-FLS was also enhanced by overexpression of VGLL3 as shown by Western blotting (Fig. 4C). STAT1 could be phosphorylated and activated upon IFN-β stimulation and translocated into the nuclear to activate the expression of ISGs. It was found that overexpression of VGLL3 promoted the phosphorylation of STAT1 (Fig. 4D), indicating the activation of STAT1. To verify that the elevated level of IFN-β1 after overexpression of VGLL3 was accountable for the increased expression of ISGs, a Janus kinase (JAK) inhibitor tofacitinib was used to block the intracellular signaling transduction of the IFN-β1 receptor IFNAR. RA-FLS were transfected with Flag-VGLL3-expressing lentivirus with/without the treatment of JAK inhibitor tofacitinib (250 nM). The results demonstrated that VGLL3 overexpression-augmented mRNA expression levels of ISGs (STAT1, BAFF, MX1, IRF7, OAS1, TLR3, and CCL5) were suppressed by tofacitinib (250 nM) (Fig. 4E). And immunofluorescence staining showed that tofacitinib (250 nM) significantly inhibited the expression of STAT1 induced by VGLL3 overexpression (Fig. 4F). The above findings indicated that VGLL3 facilitated the IRF3-mediated IFN-β1 production in RA-FLS, and IFN-β1 might stimulate the expression of ISGs in an autocrine manner.

To explore the mechanism by which VGLL3 regulates the IFN-β1 production in RA-FLS, proteins involved in the Hippo pathway were examined. Western blotting showed that the protein expression level of WWTR1 was enhanced by VGLL3 knockdown, and the expression of AMOTL2 was inhibited (Fig. 5A). In contrast, overexpression of VGLL3 markedly suppressed the expression of WWTR1 and increased the expression of AMOTL2 (Fig. 5B). Transfection of WWTR1 siRNA into RA-FLS silenced the expression of WWTR1, as verified by Western blotting (Fig. 5C). The reduction of IFN-β1 mRNA expression level induced by VGLL3 silencing was partially rescued by WWTR1 siRNA transfection (Fig. 5D). Furthermore, WWTR1 siRNA transfection could also restore the nuclear translocation of IRF3 and the expression of STAT1 which were inhibited by VGLL3 silencing in RA-FLS (Fig. 5E, F), suggesting that VGLL3 regulates the IRF3 nuclear translocation and IFN-β1 expression through the Hippo pathway.