Dissecting ARL15 Function in Rheumatoid Arthritis: Insights from Ex Vivo and In Vitro Synovial Fibroblast Models

ARL15, coding for a small GTPase was identified as a non -HLA susceptibility gene in rheumatoid arthritis (RA) through a GWAS in a North Indian cohort. Serum adiponectin and ARL15 levels were higher in RA patients with the associated genotype. The present study aimed to delineate the functional role of ARL15 in RA pathobiology using gene knockdown (KD) combined with transcriptomic profiling in both ex-vivo RA synovial fibroblasts (RASF) and in vitro MH7A cell lines. In RASF, ARL15 KD led to the downregulation of COMP-an extracellular matrix stabilizer linked to severe RA-alongside upregulation of adiponectin and IFN response genes such as IFI6 and USP18. Furthermore, upregulation of NPTX1 and MX1, previously associated with disease modulation and treatment response was observed. Downregulation of CTGF, CD248, and PTX3 suggested involvement of ARL15 in inflammation and RA -associated cardiovascular risk. In contrast, ARL15 KD in MH7A cells displayed distinct gene signatures with upregulated cytokines ( IL1A, IL8, CXCL s) and downregulated inflammatory regulators (DOCK2, TLR4, TGFB2), reflecting an inflammatory bias distinct from the patient-derived RASF. This divergence highlights the limitations of immortalized cell models in capturing patient heterogeneity and disease complexity. However, the dual- system approach underscores the m ultifaceted role of ARL15 in regulating connective tissue architecture, inflammation, and immune response. These key findings position ARL15 as a promising therapeutic target, warranting furth er investigation in RA animal models and genomic medicine. Taken together, this work provides a compelling rationale to pursue ARL15 targeted interventions in RA management.

Rheumatoid arthritis (RA) is a common, chronic systemic autoimmune disease characterized by nonspecific inflammation of the peripheral joints, which leads to synovial damage, pain, stiffness, and reduced functional capacity (1, 2) . Synovial joints are .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint specialized organs designed to meet unique biomechanical and physiological demands at specific anatomical locations. These joints comprise structural elements such as articular cartilage, synovial membrane (a delicate structure consisting of one or two cel l layers of synovial fibroblasts and tissue -resident macrophages), ligaments and menisci, which collectively support joint function. The synovium contains a structurally distinct lining layer that interfaces with the synovial fluid(3). A hallmark feature of RA pathology is the marked hyperplasia of this synovial lining due to autoimmune activation. Synovial fibroblasts (SF), commonly referred to as rheumatoid arthritis synovial fibroblasts (RASF) in RA patients, are among the most dest ructive cell types in the synovium. These cells exhibit resistance to apoptosis, leading to an increase in their numbers and contributing to synovial hyperplasia, a critical factor in disease progression (4). RA manifests significant clinical heterogeneity, complicating diagnosis, prognosis and treatment. Progression of RA varies with patients as few achieve remissions or low disease activity whereas others develop severe joint deformity and functional disabili ty (5). The etiology of RA involves a complex interplay of genetic and non -genetic factors. Non-genetic factors implicated range from smoking, air pollution, obesity to alcohol consumption etc. (6, 7) but mechanism of their contribution or pathways adopted remain largely unexplained. Conversely, over 100 susceptibility genes have been identified based on meta -analyses of GWAS data in populations of European, Asian, and other ancestries (8-10). These findings clearly indicate genetic heterogeneity and likely population specific susceptibility in RA. In the only GWAS in RA from India to date, we identified ARL15, a non -HLA gene associated with the disease (11). Preliminary characterization of ARL15, a member of the ARF family of small GTPases using RASF and a knock down (KD) approach showed a significant decrease in the invasion and migration properties of RASF. It was also demonstrated that ARL15 action is independent of (12) (13). In animal models of RA, the inhibition of ARF proteins alleviated inflammation and disease severity (14). ARL15 has been associated with the age of onset for alcohol dependence (15), which, in turn, correlates with RA risk (16) and RA is closely linked to cardiovascular disease (CVD) risk factors. ARL15 is also associated with metabolic traits, including HDL cholesterol concentration, insulin resistance, coronary heart disease, fasting insulin, and triglyceride .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint levels (17). However, the precise mechanisms and pathways underlying the role of this novel susceptibility gene in RA and other complex traits remain unclear. In this study, we sought to gain deeper insights into the role of ARL15 in RA biology by employing ARL15 KD and hypothesis -free transcriptomics strategy using both patient derived RASF cells and MH7A, a RASF derived cell line. Analyses revealed i) differential expression of functionally relevant genes ii) differences in these between the two experimental RASF samples; iii) similarities and differences in findings between RASF samples and MH7A cells; and iv) most importantly, novel insights into the likely r ole of ARL15 in RA.

Recruitment of study subjects and collection of target tissues This study was carried with institutional ethical committee clearance from both participating centers. Patients with RA were diagnosed at Rheumatology unit All India Institute of Medical Sciences, New Delhi (by UK) in accordance with the American college of rheumatology/ EULAR criteria (18). Synovial fluid/tissue samples from study subjects undergoing total knee replacement surgery at the orthopedic surgery unit at the same hospital were collected with informed consent. Samples from osteoarthritis (OA) patients recruited the same way were used as controls. Generation of patient derived synovial fibroblasts (RASF) Tissue samples from five RA patients (synovial fluid from patient id # RA1, RA3 and RA6; synovial tissue from id # RA4 and RA6) and two OA patients (synovial fluid from id # OA1 and synovial tissue from id # OA2) were obtained during total knee replacement surgery. Synovial fibroblasts from these two tissues were cultured as described elsewhere (19) Briefly, synovial fluid was aspirated and transferred to a heparin vacutainer and immediately transported on ice to the laboratory. The sample was centrifuged at 1100 rpm for 10 min and the pellet was seeded in a T -25 flask and grown at 37 0C, in a 5% CO 2 incubator in complete medium (with 15% fetal bovine serum, 1% non -essential amino acids and 1% penicillin-streptomycin solution). Synovial tissue collected was transported on ice to the lab and first minced into small pieces and then incubated with collag enase (C2139, SIGMA) in DMEM (1199504, Thermofisher, USA) overnight at 37 0 C. The .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint suspension was then centrifuged at 1200 rpm for 15 and pellet was dissolved in complete medium (DMEM, 10% fbs, 1% penicillin -streptomycin solution) and seeded in a T -25 flask. Cells were then grown at 370 C and sub-passaging for both tissue and fluid derived synovial fibroblasts was done as required and aliquots were also frozen for future use. Cultures with ~80% confluence after third passage were used for all the experiments as previously described (13). Immortalized synovial fibroblast (MH7A cell line) culture MH7A synovial fibroblast cell line (RCB1512) was procured from RIKEN cell bank, Japan and used for ARL15 KD experiments. Briefly, cells were cultured in RPMI 1640 media (11875093, Thermo Fisher Scientific, USA) supplemented with 10% heat -inactivated Fetal Bovine Serum (10100147, Thermo Fisher Scientific, USA), 1% penicillin - streptomycin (10,000 U/mL) (151401 22, Thermo Fisher Scientific, USA), and were incubated at 37OC with 5% CO2 for maintenance and subsequent experiments. Knockdown of ARL15 using siRNA siRNA for ARL15 and scrambled siRNAs to be used as negative control (NC) were the same as described previously (Kashyap et al., 2018). RASF and OASF samples were seeded at a density of 1.5x106 in six well plates. 24 hours prior to transfection, cells were incubated in antibiotic and serum free medium. 10nM of ARL15 siRNA (A50284F6, Thermo Fisher Scientific, USA) and scrambled siRNA as NC (AM4611, Thermo Fisher Scientific, USA) were used for respective transfections using lipofectamine 2000 (11668019, Thermo Fisher Scientific, USA). Six hours later the medium containing siRNA was replaced by normal growth medium, and cells were grown for 24 hrs. at 37 OC and harvested for RNA isolation. RASF cultures with ~80% confluence treated with only lipofectamine, and Opti-MEM culture medium were considered as wild type controls. MH7A cells were grown in a T -25 flask to ~ 90 -95% confluency. Adherent cells were dislodged using 0.25 % trypsin-EDTA solution (25200072, Thermo Fisher Scientific, USA) and resuspended in 1000 μL of reduced serum medium (Opti -MEM media, 31985070, Thermo Fisher Scientific, USA). ~10 μL of resuspended cells stained with trypan blue (T10282, Thermo Fisher Scientific, USA) were used to count the total number of live/dead cells/mL using an automated cell counter (AMQAF1000, Thermo Fisher Scientific, USA). Around 80-100 μL (~0.2-0.3x105 cells/mL) of resuspended cells per well of a six-well plate .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint were used for reverse transfection. Like for RASF, 10nM of ARL15 siRNA and scrambled siRNA as NC were used. siRNA -Liposome mixture was prepared accordingly on a per well basis of a six -well plate using Lipofectamine 3000 and P3000 reagents (5 μL each) (L3000015, Thermo Fisher Scientific, USA) and used for KD experiments. For a good transfection efficiency, reverse transfection was performed as per the standard protocol by adding siRNA and Lipofectamine together followed by addition of resuspended cells to the mix. Briefly, Opti-MEM-Lipofectamine mix was added with the siRN A-P3000-Opti- MEM mix in equal (1:1) proportion and incubated for 5 minutes at room temperature. ~80- 100μL of resuspended cells was added to above mix, and the mixture was added to the respective wells having Opti-MEM media up to 2000 μL and incubated for 6 hours at 37OC with 5% CO 2 for surface attachment, followed by replacement of Opti -MEM media with complete media for further growth. Total RNA extraction Total RNA was extracted from wild -type (WT) and knockdown (KD) RASF/RAST cells using the standard Trizol method (TRI reagent, T9424, Sigma, USA). Briefly, cells were washed with PBS before treatment with Trizol for 5 minutes. RNA was isolated using chloroform extraction, followed by precipitation with isopropanol and a 75% ethanol wash. The final RNA pellet was dissolved in nuclease -free water, and its quantity and quality were assessed using a Nanodrop (Thermo, USA) and a Bioanalyzer (Agilent, USA). For MH7A cells, RNA isolation was performed using Trizol reagent as above but isolation was done with RNA isolation kit (R2070, Zymo Research, USA). Extracted RNA was quantified using a fluorometry -based method with an RNA quantification kit (Q32852, Thermo Fisher Scientific, USA). RNA samples thus isolated were used for transcriptome sequencing. Transcriptome sequencing RASF RNA isolated from RA3 (synovial fluid), RA6 (synovial tissue), and OASF from OA1 (synovial fluid) cells, assessed for quality and integrity (RIN ≥ 8), were used for transcriptome sequencing. Libraries were prepared using 400 ng of RNA per sample with the TruSeq RNA Library Prep Kit (Illumina, USA), following the manufacturer’s protocol and sequencing performed in paired -end mode (2×101 bp) on an Illumina HiSeq 2000, using a commercial facility (MedGenome, Bangalore, India). .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint For the MH7A cell line, RNA isolated from two independent samples from passages (P7 and P10) following confirmation of knockdown were used for transcriptome sequencing. All QC -passed samples with RIN ≥ 7.2 were used for library preparation. Library preparation was performed using the Ribo-Zero rRNA removal kit (Illumina, Inc., USA) as per the manufacturer’s protocol and sequencing was done as above. Transcriptome data analysis The raw sequencing data in FASTQ format were checked for quality using FastQC. Adapter removal and trimming of low-quality bases were performed using Cutadapt. For differential gene expression analysis in RASF samples, high -quality reads were aligned to the human reference genome (hg19, UCSC Genome Browser) using TopHat v2.0.8 (20). The aligned reads were assembled into transcripts without a reference genome using the Cufflinks software package (v2.0.2) (21). Differential expression between WT and KD samples was determined using the Cuffdiff module in Cufflinks and reported as Fragments Per Kilobase of exon per Million fragments mapped (FPKM). Gene ontology analysis was performed using DAVID Bioinformatics Re sources 6.8 (22). Additionally, transcriptome data from two synovial tissue-derived RASF samples—one from a patient and one from a control both of Caucasian origin (23) were included for comparative analyses. For MH7A cells, raw sequencing data underwent quality assessment using FastQC followed by data filtering, adapter trimming, and contamination removal using Trimmomatic (24). Read alignment, genome mapping, and gene expression quantification were performed using STAR (25). Gene expression levels were represented as FPKM, and differential expression was estimated using DESeq2 (26). Differentially expressed genes with an adjusted p-value (≤ 0.05) and a fold-change of ±1 were considered significant and taken forward for functional annotation. Functional annotation of significantly upregulated or downregulated genes was performed usin g reported literature in PubMed, as well as the UniProt and GeneCards databases (27). Pathway analysis was carried out as described above for RASF.

Confirmation of homogeneity of fluid and tissue derived RASF .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint RASF cultured from both synovial fluid and synovial tissue exhibited >99% specificity for CD90/Thy-1, with  0.5% contamination from other cell types, confirming test sample homogeneity (Supplementary Figure 1). RASF cells: ARL15 KD showed a significantly lower expression level of ARL15 in all three samples (two RA and one OA) as compared to the wild type. Differential expression analysis identified 217, 340, and 87 genes (P < 0.05) in the three samples, respectively. Only two genes, NPTX1 and MX1, were common between RA and OA samples, whereas 25 genes including ARL15 were shared between the two RA samples (Figure 1A, Supplementary tables 1-4). NPTX1 was found to be upregulated in all three samples, though its expression level was very low in OASF. In contrast, MX1 was upregulated in RASF but downregulated in OASF. Of the 16 genes common to both RA samples, COMP, CTGF, KCDN3, PTX3, PRELP, CD248, ACAN, CITED2, and NGF were downregulated, whereas IFI6, LRRC17, SLC45A2, USP18, COL14A1, CENPK, and GPR153 were upregulated. Of note, six other genes exhibited differential regulation between the two RA samples - ESM1, KRT19, SLN, and IL7R were downregulated, and MCF2L and GRIP1 were upregulated in fluid-derived RASF but showed the exact opposite trend in tissue-derived RASF. These are shown in (Supplementary table 1). The expression of the two most differentially expressed genes, NPTX1 and COMP was validated by qPCR. Significant (P < 0.0001) upregulation of NPTX1 and downregulation of COMP were confirmed in both RASF samples (Figure 1B) . Gene Ontology analysis revealed enrichment of genes associated with the extracellular matrix, and the PI3K -AKT pathway emerged as a common pathway in both samples (Figure 2 A, B). Increased apoptosis in RASF on ARL15 KD The Annexin V staining followed by flow-cytometry in RASF with ARL15 KD showed ~10% more cell death (P<0.01) as compared to RASF with WT ARL15 (Supplementary figure 2). MH7A cells A total of 56 differentially expressed genes (33 upregulated and 23 downregulated) including two long non-coding RNAs identified along with their functional annotation of all are presented (Supplementary table 5). Of these, expression of 21 genes of notable .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint functional relevance including DOCK2, TGFB2, TLR4, FPR1, TFPI, CD82, RARRES2, MSN, and TMED10 (downregulated, Figure 1C) ; and CCR1, CXCL2, CXCL3, ESM1, GRPR, HAS2, IL24, IL8, IL1A, MMP1, PAPPA, and S1PR1(upregulated, Figure 1D) was validated through qPCR using the same set of RNA that was used for transcriptome sequencing, with Actin and UBC as endogenous reference controls (See supplementary information). Network analysis of differentially expressed genes STRING based (Szklarczyk et al., 2019) network analysis was done with 54 differentially expressed genes eliminating PAX8 -AS1 and UCA1 (long non -coding RNAs) to identify the gene clusters and the pathways/processes . Two main cytokine-chemokine clusters and keratin protein clusters were observed and ARL15 is found to interact with DDIT4L (based on literature-based text mining). As STRING network analysis did not provide any conclusive idea about the mechanistic/downstream pathway through which ARL15 functions in the MH7A cell line, DAVID was used for GO and pathway analysis . Again, several inflammation related pathways were observed (Fig 2C). To get more in -depth insight into the role of ARL15, all the 56 differentially expressed genes were used for pathway analysis using the IPA tool (Ingenuity Systems, Redwood City, CA; www.qiagen.com/ingenuity). Results based on involvement of 56 genes showed disease and disorders, molecular and cellular functions, and physiological system development and function among the top canonical pathways.

ARL15 (rs255758 intronic, G>A) was identified as a non-HLA gene associated with RA in the first GWAS that was conducted on a north Indian cohort. Furthermore, in RA patients with the homozygous variant (AA), altered serum adiponectin level was observed (11). Subsequently lower levels of ARL15 in patients with the associated variant genotypes (GA+AA) were documented (28). With hypothesis testing efforts to characterize ARL15 in RASF using the KD approach, we demonstrated downregulation of IL6 and upregulation of adiponectin, adiponectin receptor I and TNF independent expression of ARL15 ; and a reduction in invasion and migration potential of RASF (13, 29). However, the mechanism of action of ARL15 in RA biology per se remains to be known. Therefore, .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint to identify the genes/pathways altered by this small GTPase with likely implications for disease biology, we employed a gene KD approach combined with the hypothesis free transcriptome sequencing in ex vivo (patient derived) RASF- both synovial tissue and fluid derived and in vitro (MH7A, a well-established cell line) samples. Patient derived RASF serves as a gold standard for studying disease pathology. RASF retains the genetic makeup and primary disease characteristics thus providing a disease comparable physiological milieu. Being derived from different patients, they have the advantage of representing the disease heterogeneity and variability. At the same time, such individual variability and limited life span pose limitations in the use of RASF. It also poses logistical challenges in obtaining and standardizing samples along with ethical and consent-related constraints. In contrast to RASF, immortalized synovial fibroblast (iSF) based cell lines such as MH7A have the advantage of unlimited proliferation, which helps in reproducibility and large-scale experiments; and are easier to manipulate genetically or experimentally as compared to RASF. However, iSF encounter a few limitations such as the potential loss of primary disease characteristics or differen tiation state, and importantly, it cannot capture patient -specific genetic or disease -specific clinical heterogeneity. With this rationale and to capture ARL15 biology to the extent possible, we characterized ARL15 through a knockdown approach followed by transcriptome analysis using both ex vivo (RASF) and in vitro (MH7A) systems, Insights from genes differentially expressed in these two relevant yet distinctly different experimental systems are discussed. RASF findings: Transcriptome data (Supplementary table 1) from the KD experiments in both the RASF samples (RASF3, RASF6) confirmed our previous findings of significant downregulation of IL6 in synovial fluid derived RASF and a similar trend in tissue derived fibroblast. Adiponectin was not detected in the WT controls but was identified in ARL15 KD RASF from synovial tissue as mentioned in our previous study (13) but not detected in synovial fluid derived RASF (Supplementary Table 2). Of the 24 genes which were differentially expressed in both the synovial fluid and tissue derived RASF KD samples ( Supplementary table 1) NPTX1 and MX1 were observed to be significantly differentially regulated. NPTX1 was upregulated in both RASF KD samples and negligibly in OASF KD sample. NPTX1 has been shown to be downregulated in RASF as compared to healthy SF in a microarray-based study (30) , .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint which supports our present findings. On the other hand, MX1 was upregulated in both RASF KD samples but down regulated in OASF. MX1 is an interferon (IFN) response gene, and it has been shown to be correlated with disease activity in RASF. It has also been demonstrated that MX1 may be an important gene for predicting the treatment response in RA patients (31). Of the remaining 22 genes which were differentially regulated between RASF ARL15 WT and KD, most significant was the downregulated COMP. This gene has been shown to play an important role for the stabilization of extracellular matrix. Higher amount of COMP has also shown to be correlated with severe disease in RA patients, implying its potential role as a biomarker for RA (32). Downregulation of COMP in both ARL15 KD RASF samples in our study reinforce the role of ARL15 in RA biology. In addition, CTGF, PTX3, IFI6, USP18 and CD248 are five other differentially regulated genes which seem to be functionally relevant for RA biology. CTGF has been shown to be involved in bone degradation of RA patients by activating osteoclastogenesis (33), and its inhibition in CIA mice model has been reported to ameliorate the disease (34) Our finding of CTGF being downregulated upon ARL15 KD may yet again support the likely disease causal role of ARL15. PTX3 has been reported to be overexpressed in RA cases (35). It has also been shown to be produced upon the activation of proinflammatory cytokines and is long thought to be involved in cardiovascular disease (CVD) prognosis (36). Of note, RA is associated with increased CVD risk (37) and PTX3 was downregulated upon ARL15 KD in both RA samples. These findings taken together amply support a likely role of ARL15 in metabolic syndromes as well. On the other hand, IFI6 and USP18 are major IFN signaling genes. Inhibition of USP18 has been correlated with inflammation (38) and upregulation of IFI6 is observed in good responders of TNF antagonists (39) . Both IFI6 and USP18 were upregulated in ARL15 KD RASF, yet again supporting a putative role of ARL15 in RA. A previous study on CD248 knockout mice showed less severe collagen induced arthritis compared to wild type with reduction in synovial hyperplasia and cartilage destruction (40) Our finding of reduced CD248 expression upon ARL15 KD further strengthens the druggability of ARL15 in decreasing severity of RA. MH7A cells : On analysis of transcriptome data in MH7A cells ARL15 was the most significantly downregulated (Log 2 FC = -3.9, P-adjusted value = 1.39E-49) validated by .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint qPCR (Figure 1C ). Most of the differentially expressed genes in MH7A KD were informative (Supplementary table 5) but were notably different compared to RASF findings. In MH7A samples major cytokine and chemokine genes such as IL1A, IL8, IL24, CXCL1, CXCL2, and CXCL3 were among the 12 significantly upregulated functionally relevant genes (Supplementary table 5) and validated through qPCR (Figure 1D) but are in stark contrast to expectation from a ARL15KD background. On the other hand, the nine downregulated genes included MSN, TMED10, RARRES2, CD82, DOCK2, TGFB2, TLR4, FPR1 and TFPI (Figure 1 C). Of these, MSN is part of the ERM (ezrin/radixin/moesin) proteins, where increased phosphorylation in response to cytokines promotes RASF proliferation (41). TMED10, acting as a receptor for ARF1 - GDP, participates in COPI -vesicle formation on the Golgi membrane, enhancing coatomer-dependent GTPase -activating activity of ARFGAP2. RARRES2 encodes chemerin, an adipokine involved in inflammation, adipogenesis, angiogenesis, and energy metabolism, also promoting adipocyte differentiation (42). CD82 functions as a tumor metastasis suppressor; its overexpression in SFs reduces RASF migration and adhesion under pro-inflammatory stimuli (43). DOCK2 acts as a guanine exchange factor (GEF), mediating Rac1/2 activation and regulating immune cell migration (44) . TGFB2, an anti-inflammatory cytokine, is implicated in RA pathogenesis (45). TLR4 upregulates inflammatory cytokines and chemokines ( IL-6, IL -8, and MMP3) in RASF (46). TFPI inhibits tissue factor (TF)-mediated inflammation in arthritis synovial joints (47). At this juncture it is important to discuss the notable differences between patient derived cells and the cell line. A large number of genes were differentially expressed between RASF and MH7A control samples (Figure 3A). Notable differences in the transcriptome profiles were also observed between these two sample types (Figure3B) which was effectively captured in the correlation plot (Figure 3C). Conversely, as expected, samples derived from synovial tissue in the study showed a higher correlation with the Caucasian RA synovial tissue sample which was used for comparison. These findings reiterate the basic differences between patient derived primary cells and an established cell line . It may be important to mention here that FPKM values observed in the two RASF WT samples in our study, were comparable with the values reported for Caucasian RASF but not for healthy SF samples (48) (Supplementary Table 6 ) further supporting our findings. .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint In addition, t his study using two different experimental samples mainly patient derived RASF and MH7A cell line uncovers a basic fact on the differences in their utility to characterize a susceptibility gene(s) of interest. RASF in early passages seem to represent/capture the key players in active disease while the cell line largely manifests inflammation related markers. Such differences underscore the effect of immortalization on gene expression and consequent under representation of disease related markers. Nevertheless, both these study designs together seem to provide useful insights into the role of ARL15 in RA. In summary, this study on ARL15 KD in RASF provides important insights into the role of this small GTPase in regulation of RA disease biology via maintenance of connective tissue architecture ( CTGF and COMP); reduction of inflammation ( PTX3 and USP18); and increase in apoptosis; together with some leads from MH7A KD cells through cytokine pathways. Despite significant therapeutic advances, RA treatment remains a major challenge encouraging alternate approaches such as genomic medicine. In this context, taken together the leads from this study using a combination of ex vivo and in vitro models are suggestive of a likely regulatory role of ARL15 in RA. They also highlight the need for further characterization of ARL15 to explore its druggability using animal models of RA.

1. Firestein GS. Evolving concepts of rheumatoid arthritis. Nature. 2003;423(6937):356- 61. 2. Sweeney SE, Firestein GS. Rheumatoid arthritis: regulation of synovial inflammation. Int J Biochem Cell Biol. 2004;36(3):372-8. 3. Noss EH, Brenner MB. The role and therapeutic implications of fibroblast ‐like synoviocytes in inflammation and cartilage erosion in rheumatoid arthritis. Immunological Reviews. 2008;223(1):252-70. 4. Bottini N, Firestein GS. Duality of fibroblast -like synoviocytes in RA: passive responders and imprinted aggressors. Nature Reviews Rheumatology. 2013;9(1):24-33. 5. Nell VPK. Benefit of very early referral and very early therapy with disease -modifying anti-rheumatic drugs in patients with early rheumatoid arthritis. Rheumatology. 2004;43(7):906-14. 6. Chang K, Yang SM, Kim SH, Han KH, Park SJ, Shin JI. Smoking and rheumatoid arthritis. Int J Mol Sci. 2014;15(12):22279-95. 7. Azizov V , Zaiss MM. Alcohol Consumption in Rheumatoid Arthritis: A Path through the Immune System. Nutrients. 2021;13(4):1324. .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint 8. Eyre S, Bowes J, Diogo D, Lee A, Barton A, Martin P , et al. High -density genetic mapping identifies new susceptibility loci for rheumatoid arthritis. Nature Genetics. 2012;44(12):1336-40. 9. Okada Y , Wu D, Trynka G, Raj T, Terao C, Ikari K, et al. Genetics of rheumatoid arthritis contributes to biology and drug discovery. Nature. 2014;506(7488):376-81. 10. Okada Y , Eyre S, Suzuki A, Kochi Y , Yamamoto K. Genetics of rheumatoid arthritis: 2018 status. Annals of the Rheumatic Diseases. 2019;78(4):446-53. 11. Negi S, Juyal G, Senapati S, Prasad P , Gupta A, Singh S, et al. A genome -wide association study reveals ARL15, a novel non -HLA susceptibility gene for rheumatoid arthritis in North Indians. Arthritis Rheum. 2013;65(12):3026-35. 12. Sahu D, Gupta C, Yennamalli RM, Sharma S, Roy S, Hasan S, et al. Novel peptide inhibitor of human tumor necrosis factor -α has antiarthritic activity. Scientific Reports. 2024;14(1). 13. Kashyap S, Kumar U, Pandey AK, Kanjilal M, Chattopadhyay P , Yadav C, et al. Functional characterisation of ADP ribosylation factor-like protein 15 in rheumatoid arthritis synovial fibroblasts. Clin Exp Rheumatol. 2018;36(4):581-8. 14. Zhu W, London NR, Gibson CC, Davis CT, Tong Z, Sorensen LK, et al. Interleukin receptor activates a MYD88 -ARNO-ARF6 cascade to disrupt vascular stability. Nature. 2012;492(7428):252-5. 15. Kapoor M, Wang J-C, Wetherill L, Le N, Bertelsen S, Hinrichs AL, et al. Genome -wide survival analysis of age at onset of alcohol dependence in extended high-risk COGA families. Drug and Alcohol Dependence. 2014;142:56-62. 16. Zegkos T, Kitas G, Dimitroulas T . Cardiovascular risk in rheumatoid arthritis: assessment, management and next steps. Therapeutic Advances in Musculoskeletal Disease. 2016;8(3):86-101. 17. Rocha N, Payne F , Huang-Doran I, Sleigh A, Fawcett K, Adams C, et al. The metabolic syndrome- associated small G protein ARL15 plays a role in adipocyte differentiation and adiponectin secretion. Scientific Reports. 2017;7(1). 18. Aletaha D, Neogi T, Silman AJ, Funovits J, Felson DT, Bingham CO, 3rd, et al. 2010 Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis and rheumatism. 2010;62(9):2569-81. 19. Stebulis JA, Rossetti RG, Atez FJ, Zurier RB. Fibroblast-like synovial cells derived from synovial fluid. J Rheumatol. 2005;32(2):301-6. 20. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differential gene and transcript expression analysis of RNA -seq experiments with TopHat and Cufflinks. Nature Protocols. 2012;7(3):562-78. 21. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, Van Baren MJ, et al. Transcript assembly and quantification by RNA -Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nature Biotechnology. 2010;28(5):511-5. 22. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature Protocols. 2009;4(1):44-57. 23. Wang X, Xia S, Fu B. RNA -seq analysis of synovial fibroblasts in human rheumatoid arthritis. Molecular Medicine Reports. 2014;10(1):241-7. .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint 24. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114-20. 25. Dobin A, Davis CA, Schlesinger F , Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15-21. 26. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. 27. Stelzer G, Rosen N, Plaschkes I, Zimmerman S, Twik M, Fishilevich S, et al. The GeneCards Suite: From Gene Data Mining to Disease Genome Sequence Analyses. Current protocols in bioinformatics. 2016;54(1). 28. Pandey AK, Saxena A, Dey SK, Kanjilal M, Kumar U, Thelma BK. Correlation between an intronic SNP genotype and ARL15 level in rheumatoid arthritis. J Genet. 2021;100. 29. Kashyap S, Chattopadhyaay P , Pandey AK, Kumar U, Yadav CS, Thelma B, editors. ARL15 expressed by rheumatoid arthritis synovial fibroblasts regulates IL6 and plays a role in apoptosis and hypoxia. ARTHRITIS & RHEUMATOLOGY; 2016: WILEY 111 RIVER ST, HOBOKEN 07030-5774, NJ USA. 30. Del Rey MJ, Izquierdo E, Usategui A, Gonzalo E, Blanco FJ, Acquadro F , et al. The transcriptional response of normal and rheumatoid arthritis synovial fibroblasts to hypoxia. Arthritis and rheumatism. 2010;62(12):3584-94. 31. Galligan CL, Baig E, Bykerk V , Keystone EC, Fish EN. Distinctive gene expression signatures in rheumatoid arthritis synovial tissue fibroblast cells: correlates with disease activity. Genes Immun. 2007;8(6):480-91. 32. Tseng S, Reddi AH, Di Cesare PE. Cartilage Oligomeric Matrix Protein (COMP): A Biomarker of Arthritis. Biomarker insights. 2009;4:33-44. 33. Nozawa K, Fujishiro M, Kawasaki M, Kaneko H, Iwabuchi K, Yanagida M, et al. Connective tissue growth factor promotes articular damage by increased osteoclastogenesis in patients with rheumatoid arthritis. Arthritis research & therapy. 2009;11(6):R174. 34. Nozawa K, Fujishiro M, Kawasaki M, Yamaguchi A, Ikeda K, Morimoto S, et al. Inhibition of connective tissue growth factor ameliorates disease in a murine model of rheumatoid arthritis. Arthritis Rheum. 2013;65(6):1477-86. 35. Luchetti MM, Piccinini G, Mantovani A, Peri G, Matteucci C, Pomponio G, et al. Expression and production of the long pentraxin PTX3 in rheumatoid arthritis (RA). Clin Exp Immunol. 2000;119(1):196-202. 36. Deyab G, Hokstad I, Aaseth J, Småstuen MC, Whist JE, Agewall S, et al. Effect of anti- rheumatic treatment on selenium levels in inflammatory arthritis. J Trace Elem Med Biol. 2018;49:91-7. 37. Dessein PH, Norton GR, Badenhorst M, Woodiwiss AJ, Solomon A. Rheumatoid arthritis impacts on the independent relationships between circulating adiponectin concentrations and cardiovascular metabolic risk. Mediators Inflamm. 2013;2013:461849. 38. Yim HY , Park C, Lee YD, Arimoto K, Jeon R, Baek SH, et al. Elevated Response to Type I IFN Enhances RANKL -Mediated Osteoclastogenesis in Usp18-Knockout Mice. J Immunol. 2016;196(9):3887-95. 39. Sanayama Y , Ikeda K, Saito Y , Kagami S, Yamagata M, Furuta S, et al. Prediction of therapeutic responses to tocilizumab in patients with rheumatoid arthritis: biomarkers .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint identified by analysis of gene expression in peripheral blood mononuclear cells using genome-wide DNA microarray. Arthritis Rheumatol. 2014;66(6):1421-31. 40. Maia M, de Vriese A, Janssens T, Moons M, van Landuyt K, Tavernier J, et al. CD248 and its cytoplasmic domain: a therapeutic target for arthritis. Arthritis Rheum. 2010;62(12):3595-606. 41. Huang H, Xiao Y , Lin H, Fu D, Zhan Z, Liang L, et al. Increased phosphorylation of ezrin/radixin/moesin proteins contributes to proliferation of rheumatoid fibroblast -like synoviocytes. Rheumatology. 2011;50(6):1045-53. 42. Helfer G, Wu Q -F . Chemerin: a multifaceted adipokine involved in metabolic disorders. Journal of Endocrinology. 2018;238(2):R79-R94. 43. Neumann E, Schwarz MC, Hasseli R, Hülser ML, Classen S, Sauerbier M, et al. Tetraspanin CD82 affects migration, attachment and invasion of rheumatoid arthritis synovial fibroblasts. Ann Rheum Dis. 2018;77(11):1619-26. 44. Chen Y , Meng F , Wang B, He L, Liu Y , Liu Z. Dock2 in the development of inflammation and cancer. European Journal of Immunology. 2018;48(6):915-22. 45. Marinova-Mutafchieva L, Gabay C, Funa K, Williams RO. Remission of collagen - induced arthritis is associated with high levels of transforming growth factor-β expression in the joint. Clinical and Experimental Immunology. 2006;146(2):287-93. 46. Liu J, Buisman -Pijlman F , Hutchinson MR. Toll -like receptor 4: innate immune regulator of neuroimmune and neuroendocrine interactions in stress and major depressive disorder. Frontiers in Neuroscience. 2014;8. 47. Busso N, Morard C, Salvi R, Péclat V , So A. Role of the tissue factor pathway in synovial inflammation. Arthritis Rheum. 2003;48(3):651-9. 48. Wang X, Xia S, Fu B. RNAseq analysis of synovial fibroblasts in human rheumatoid arthritis. Molecular medicine reports. 2014;10(1):241-7. .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint Unstained control Isotype control CD90Thy1 Supplementary figure1: Confirmation of synovial fibroblast homogeniety by flow staining. From left to right synovial fibroblast cells were run on FACS caliber (BD) with unstained control followed by Isotype and CD90Thy1 (FITC) antibody. .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint .CC-BY-ND 4.0 International licenseavailable under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (whichthis version posted July 4, 2025. ; https://doi.org/10.1101/2025.06.30.662131doi: bioRxiv preprint