Increased long noncoding RNA LINK-A contributes to rheumatoid synovial inflammation and aggression

Fibroblast-like synoviocytes (FLSs) play a key role in controlling synovial inflammation and joint destruction in rheumatoid arthritis (RA). The contribution of long noncoding RNAs (lncRNAs) to RA is largely unknown. Here, we show that the lncRNA LINK-A, located mainly in cytoplasm, has higher-than-normal expression in synovial tissues and FLSs from patients with RA. Synovial LINK-A expression was positively correlated with the severity of synovitis in patients with RA. LINK-A knockdown decreased migration, invasion, and expression and secretion of matrix metalloproteinases and proinflammatory cytokines in RA FLSs. Mechanistically, LINK-A controlled RA FLS inflammation and invasion through regulation of tyrosine protein kinase 6–mediated and leucine-rich repeat kinase 2–mediated HIF-1α. On the other hand, we also demonstrate that LINK-A could bind with microRNA 1262 as a sponge to control RA FLS aggression but not inflammation. Our findings suggest that increased level of LINK-A may contribute to FLS-mediated rheumatoid synovial inflammation and aggression. LINK-A might be a potential therapeutic target for RA.


Introduction
Rheumatoid arthritis (RA) is a common autoimmune disease characterized by chronic synovial inflammation and progressive articular damage (1). Fibroblast-like synoviocytes (FLSs), the active resident cell component in synovial intimal lining, play a critical role in aggressive action of synovial pannus tissue (2). Stable activated RA FLSs exhibit not only increased secretion of inflammatory mediators but also abnormal, aggressive behavior. It has been found that RA FLSs are epigenetically imprinted with an aggressive phenotype. Targeting activated FLS aggression is an attractive therapeutic approach for controlling joint damage of RA (3,4). Accumulating evidence suggests autonomous factors including epigenetic changes may have important roles in development of the aggressive phenotype of RA FLSs (5).
The majority of the mammalian genome is transcribed to noncoding RNAs (ncRNAs), and ncRNAs are important factors in modulating the physiological functions of mammalian cells (6,7). It is well-known that ncRNAs are classified into 2 groups based on their length. One group is long noncoding RNAs (lncRNAs), which are defined as ncRNAs longer than 200 nucleotides (nt) without protein coding ability, and the other group is microRNAs (miRNAs), whose length is between 20 and 24 nt. Both lncRNAs and miRNAs have been demonstrated to be crucial regulators of gene expression at the epigenetic, transcriptional, and posttranscriptional levels and can actively participate in various physiological and pathological processes (8,9). For instance, lncRNA is required for mouse acute leukemia maintenance by promoting leukemia stem cell signatures (10). LncRNA NEAT1 was proved to promote glioma cells' growth and invasiveness by promoting SOX2 expression through suppressing miRNA 132 (miR-132) (11). The lncRNA MCF2L-AS1 plays a critical role in the osteogenic differentiation of bone marrow mesenchymal stem cells, and targeting lncRNA MCF2L-AS1 is a promising strategy to promote osteogenic differentiation (12). LncRNA ZNNT1 acts as a potential tumor suppressor in uveal melanoma by inducing autophagy (13). LncRNA LINK-A modulates growth of breast cancer cells (14). Interestingly, recent studies show involvement of lncRNAs in autoimmune disorders, including RA (15)(16)(17)(18); for example, lncRNA LERFS negatively controls rheumatoid synovial Fibroblast-like synoviocytes (FLSs) play a key role in controlling synovial inflammation and joint destruction in rheumatoid arthritis (RA). The contribution of long noncoding RNAs (lncRNAs) to RA is largely unknown. Here, we show that the lncRNA LINK-A, located mainly in cytoplasm, has higher-than-normal expression in synovial tissues and FLSs from patients with RA. Synovial LINK-A expression was positively correlated with the severity of synovitis in patients with RA. LINK-A knockdown decreased migration, invasion, and expression and secretion of matrix metalloproteinases and proinflammatory cytokines in RA FLSs. Mechanistically, LINK-A controlled RA FLS inflammation and invasion through regulation of tyrosine protein kinase 6-mediated and leucine-rich repeat kinase 2-mediated HIF-1α. On the other hand, we also demonstrate that LINK-A could bind with microRNA 1262 as a sponge to control RA FLS aggression but not inflammation. Our findings suggest that increased level of LINK-A may contribute to FLS-mediated rheumatoid synovial inflammation and aggression. LINK-A might be a potential therapeutic target for RA.  (E) Expression of LINK-A in RA FLSs treated with IL-1β (10 ng/mL), TNF-α (10 ng/mL), IL-6 (10 ng/mL), IL-17 (10 ng/mL), LPS (10 ng/mL), methotrexate (MTX, 10 μg), and dexamethasone (DXM, 1 μg) for 24 hours. (F) Effect of MTX (10 μg) on TNF-α-induced LINK-A expression. (G and H) Cellular localization of LINK-A was measured by RNA FISH assay. Shown are representative images of LINK-A (red) and nuclei (blue) from 5 different RA patients and HCs. Graph (H) shows the quantification of staining intensity for 5 different RA patients and HCs. Original magnification, ×400. (I and J) LINK-A expression, evaluated by ISH staining, in synovial tissues from HCs and RA patients. Shown are representative images (I) and quantification of the percentage of LINK-A-positive cells (J) from 5 different RA patients and HCs. A scrambled probe was used as a negative control. White arrows indicate LINK-A-positive (red) cells. Original magnification, ×630. *P < 0.05, **P < 0.01, ***P < 0.001 versus HC FLSs or control (CON); ### P < 0.001 versus TNF-α, by Student's 2-tailed t test or 1-way ANOVA (for E and F).
JCI Insight 2021;6(23):e146757 https://doi.org/10.1172/jci.insight.146757 Figure 2D, RA FLSs transfected with control siRNA displayed flat or ruffling lamellipodia and filopodia at their leading edges, while cells transfected with si-LINK-A reduced lamellipodia and filopodia formations. Since Rho GTPases including CDC42, Rac1, and RhoA are critical proteins that control lamellipodia and filopodia, we determined the effect of LINK-A knockdown on the expression and activity of CDC42, Rac1, and RhoA. We found that LINK-A knockdown mitigated the protein expression and activity of RhoA but not Rac1 and CDC42 (Supplemental Figure 6). These data suggest that LINK-A regulates the formation of membrane protrusions in migrating cells by targeting RhoA.
Since MMPs have an important role in rheumatoid joint destruction, we investigated the role of LINK-A in modulating MMPs' expression. As shown in Figure 2E, transfection with si-LINK-A reduced TNF-α-induced gene expression of MMP-1, MMP-3, MMP-9, and MMP-13. Although MMP-14 plays an important role in regulating FLS invasion (22,23), we determined that LINK-A knockdown did not influence MMP-14 expression. We also demonstrated that LINK-A knockdown reduced secretion of MMP-1, MMP-3, MMP-9, and MMP-13 ( Figure 2F).
Finally, we determined the effect of LINK-A inhibition on in vivo invasion by RA FLSs. We coimplanted RA FLSs transfected with sh-LINK-A or control vector into the left or right flanks, respectively, of SCID mice. RA FLSs transfected with sh-LINK-A exhibited a significant reduction of invasion into cartilage as compared with control vector ( Figure 2G).
LINK-A regulates inflammation but not proliferation and apoptosis in RA FLSs. We first determined the role of LINK-A in regulating the expression of proinflammatory cytokines. We observed that si-LINK-A treatment reversed the increase of TNF-α-induced gene expression of IL-1β, IL-6, and IL-8 ( Figure 3A). Similarly, we also demonstrated the inhibitory effect of LINK-A knockdown on TNF-α-induced secretion of IL-1β, IL-6, and IL-8 ( Figure 3B).
Next, we observed the role of LINK-A in proliferation and apoptosis of RA FLSs. Cell Counting Kit-8 (CCK-8) and EdU assay were used to detect the growth and proliferation, respectively. We found that the cell viability and proliferation rate were not affected in si-LINK-A-treated RA FLSs compared with control siRNA-treated RA FLSs (Figure 3, C and D). Furthermore, we used annexin-FITC staining to detect the apoptosis of RA FLSs by flow cytometry. We observed that the total apoptotic cell number had no difference between RA FLSs treated with si-LINK-A and control siRNA ( Figure  3E). These data suggest LINK-A is involved in inflammation but not proliferation and apoptosis in RA FLSs. Collectively, our data suggest that increased LINK-A contributes to synovial inflammation and aggressiveness of RA.
LINK-A regulates the aggressiveness and inflammation of RA FLSs through PTK6-and LRRK2-mediated HIF-1α. To explore how LINK-A regulates aggressiveness and inflammation of RA FLSs, we sought to identify intracellular targeting proteins of LINK-A using RNA-Seq. Figure 4A shows the LINK-A silencing-induced top 10 mRNA KEGG pathway enrichment results, including HIF-1 signaling pathway. Consistent with our data, a recent study shows that lncRNA LINK-A activates normoxic HIF-1α signaling in breast cancer tumorigenesis (14); therefore, we speculated that LINK-A might regulate HIF-1α pathway in RA FLSs. As we expected, we demonstrated that LINK-A knockdown with siRNA significantly decreased the mRNA and protein levels of HIF-1α ( Figure 4, B and C), while LINK-A overexpression promoted HIF-1α expression (Figure 4, D and E). For exploring the role of LINK-A in regulating HIF-1α activation, we evaluated the effect of LINK-A knockdown or overexpression on nuclear HIF-1α expression. We demonstrated that LINK-A knockdown or overexpression reduced or increased nuclear HIF-1α expression in RA FLSs (Figure 4, C and E). On the other hand, we determined that HIF-1α knockdown did not influence LINK-A expression ( Figure 4F). These data suggest that HIF-1α is a downstream targeting protein of LINK-A.
Next, we explored how LINK-A modulates HIF-1α expression. Since it has been shown that LINK-A controls HIF-1α stabilization through interaction with PTK6 and LRRK2 in cancer cells (14), this prompted us to evaluate whether PTK6 and LRRK2 mediate increased LINK-A-induced HIF-1α expression in RA FLSs. We found that LINK-A knockdown decreased the mRNA and protein expression of PTK6 and LRRK2 ( Figure 5, A-D); however, knockdown with PTK6 or LRRK2 decreased total and nuclear HIF-1α expression ( Figure 5, E-H). Collectively, our data suggest that LINK-A regulates HIF-1α expression and activation via PKT6 and LRRK2 in RA FLSs.
We then investigated the role of HIF-1α in regulating migration, invasion, and inflammation of RA FLSs. We first observed that HIF-1α expression, measured by RT-qPCR, was increased in RA FLSs compared with HC FLSs (Supplemental Figure 7). We used specific siRNAs (si-HIF-1α-1 and si-HIF-1α-3) to inhibit HIF-1α

LINK-A interacts with miR-1262 to regulate the migration and invasion of RA FLSs.
Since it is mainly located in cytoplasm, in addition to interaction with cytoplasm proteins, LINK-A has a possibility of interaction with miRNAs and works as competing endogenous RNAs (ceRNAs) to promote miRNA exhaustion, thereby regulating gene expression at the posttranscriptional level (24,25). Therefore, to verify whether LINK-A acts as ceRNAs for a certain miRNA, we used miRBase and miRDB to predict the potential miRNA binding with LINK-A. The analysis showed that a total of 5 miRNAs, including miR-6736-5p, miR-4701-3p, miR-1262, miR-3915, and miR-5003-3p, had potential to bind LINK-A ( Figure 6A). We further demonstrated that LINK-A knockdown by siRNA increased expression of miR-6736-5p, miR-4701-3p, miR-1262, miR-3915, and miR-5003-3p; however, LINK-A overexpression decreased the expression of these miRNAs ( Figure 6, B and C). This confirmed the interaction of cytoplasmal LINK-A with miRNA in RA FLSs. Next, we determined the role of these miRNAs in regulating RA FLS biological functions. We observed that only transfection of miR-1262 mimics reduced the migration and invasion of RA FLS, and treatment with miR-1262 inhibitor increased migration and invasion of RA FLSs ( Figure 6, D and E). However, we found that treatment with mimics or inhibitors of miR-1262, miR-6736-5p, miR-4701-3p, miR-3915, and miR-5003-3p did not affect the expression of proinflammatory factors and MMPs in RA FLSs (Supplemental Figure 11). In addition, treatment with mimics or inhibitor of miR-1262 did not influence the secretion of proinflammatory factors and MMPs (Supplemental Figure 12). Moreover, we also found that treatment with mimics or inhibitor of miR-1262 did not influence the proliferation of RA FLSs (Supplemental Figure 13). Moreover, we observed increased levels of miR-1262 in RA FLSs compared with HC FLSs (Supplemental Figure 14). These data suggest that miR-1262 may mediate the role of LINK-A in regulating migration and invasion of RA FLSs.
Finally, to determine whether LINK-A binds directly with miR-1262, luciferase reporter assays were conducted. As shown in Figure 6F, LINK-A-WT and miR-1262 significantly reduced the luciferase activities, but the LINK-A-MUT and miR-1262 had little influence on the luciferase activity. Furthermore, we demonstrated that miR-1262 inhibition or mimics did not influence LINK-A expression ( Figure 6G). This suggests that miR-1262 was the directly targeting miRNA of LINK-A.
Since our above results show that HIF-1α mediates the role of LINK-A in regulating RA FLS functions, we observed the relationship between miR-1262 and HIF-1α. We found that HIF-1α knockdown did not affect miR-1262 expression in RA FLSs ( Figure 7A). Moreover, we showed that miR-1262 inhibitor or mimics did not influence the gene expression of HIF-1α (Figure 7, B and C). We also demonstrated that miR-1262 inhibitor or mimics did not affect the expression of total and nuclear protein of HIF-1α ( Figure  7, D and E), suggesting that LINK-A promotes RA FLS migration and invasion by sponging miR-1262 in an independent HIF-1α pathway.

Discussion
In the present study, we found increased expression of lncRNA LINK-A in FLSs and synovial tissues from patients with RA. LncRNA LINK-A functions as a positive regulator of migration, invasion, and inflammation in RA FLSs. Mechanistically, we identified that LINK-A regulates RA FLS functions through 2 independent pathways. One is through a PKT6-and LRRK2-mediated HIF-1α pathway to modulate aggressiveness and inflammation of RA FLSs; the other is a sponging interaction with miR-1262 to regulate RA FLS aggressiveness ( Figure 7F). Our data suggest that increased synovial LINK-A contributes to joint inflammation and destruction of RA. Recent studies show some lncRNAs are involved in regulation of RA FLS functions (17)(18)(19)26); however, the contribution of lncRNAs to the pathogenesis of RA is largely unknown. In this work, we observed increased expression levels of lncRNA LINK-A in FLSs and synovial tissues from patients with RA. LINK-A knockdown decreased migration, invasion, and proinflammatory cytokine secretion of RA FLSs. These data indicate that LINK-A positively regulates the FLS aggressiveness and inflammation and that the increase in LINK-A in RA FLSs might contribute to increased rheumatoid synovial inflammation and aggression, leading to joint destruction. In line with our findings, recent reports show that LINK-A modulates migration and invasion of osteosarcoma cell lines (27) and ovarian carcinoma cells (28). Although previous studies indicate that LINK-A is associated with proliferation and apoptosis in tumor cells (29) and podocytes (30), we found that LINK-A knockdown did not affect proliferation and apoptosis of RA FLSs, suggesting that LINK-A may possess a cell type-specific function in modulating cellular proliferation and apoptosis.
We further explored the underlying mechanism(s) by which LINK-A regulates the aggressiveness and inflammation of RA FLSs. LINK-A is located predominantly in the cytoplasm, which indicates that it might function by interacting with other cytoplasmic factors. A recent study shows that LINK-A controls HIF-1α stabilization and activation of HIF-1α transcriptional programs under normoxic conditions in triple-negative breast cancer (14). In our study, the KEGG pathway analysis of RNA-Seq data also revealed that LINK-A inhibition affects the HIF-1 signaling pathway.
We further demonstrated that HIF-1α is a downstream targeting protein of LINK-A in RA FLSs. On the other hand, it has been indicated that HIF-1α plays an important role in regulating RA FLS functions. For instance, HIF-1α is involved in regulating proinflammatory pathways and invasive behavior in RA FLSs (31-33). HIF-1α also modulates rheumatoid synovial bioenergetics, including glycolytic metabolism (34, 35), inflammatory cell recruitment, and angiogenesis (36). In addition, HIF-1α promotes the interactions between RA FLSs and synovial T cells and B cells (37). Here, we also observed the inhibitory effect of HIF-1α knockdown on the migration, invasion, and expression of proinflammatory cytokines and MMPs in RA FLSs. These findings suggest that HIF-1α mediates the role of LINK-A in regulating RA FLS invasion and inflammation.
Next, we explored interactions between LINK-A and HIF-1α in RA FLSs. It has been shown that LINK-A interacts with PKT6 and LRRK2 and thereby promotes HIF-1α phosphorylation and protein stabilization under normoxic conditions in breast cancer cells (14). In this work, we demonstrated that LINK-A knockdown decreased the expression of PTK6 and LRRK2, while PTK6 or LRRK2 inhibition by siRNA decreased HIF-1α protein expression in RA FLSs. Furthermore, PTK6 or LRRK2 knockdown decreased migration, invasion, and inflammation of RA FLSs (data not shown). In line with our findings, PTK6 promotes cancer cell aggressiveness (38,39). LRRK2 is also involved in inflammatory response and cell migration (40,41). Taken together, our data suggest that increased LINK-A promotes PTK6-and LRRK2-mediated HIF-1α protein expression and subsequently regulates aggressiveness and inflammation of RA FLSs. However, we do not eliminate the possibility that other signaling pathways mediate the role of LINK-A in controlling RA FLS functions.
Increasing evidence indicates that lncRNAs can be competitively binding to miRNAs and work as ceR-NAs (42). For instance, lncRNA PICSAR plays an important role in promoting rheumatoid synovial invasion and joint destruction by sponging miR-4701-5p (20). LncRNA GAPLINC controls tumor-like biological behaviors of RA FLSs by miRNA sponging (26). Thus, we determined whether LINK-A interacts with miR-NA using molecular biological analysis. In our study, miR-1262 was selected as a superior target of LINK-A according to its significantly increased expression after LINK-A inhibition and its role in regulating RA FLS function. Our subsequent analyses showed that miR-1262 inhibition or mimics did not affect LINK-A expression, and the targeting relationship between miR-1262 and LINK-A was confirmed using luciferase reporter assay. Furthermore, we observed that miR-1262 inhibition or mimics increased or reduced RA FLS migration and invasion but did not affect the expression of proinflammatory cytokines, suggesting miR-1262 is only involved in aggressive action of RA FLSs. Consistent with our findings, one research group has shown that miR-1262 regulates invasion of breast cancer cells (43). These data suggest that miR-1262 might be a new target for controlling synovial aggression and joint destruction of RA. However, it needs to be further explored how miR-1262 modulates RA FLS aggression in the future. Collectively, in this study, we propose a mechanism in which LINK-A acts as a miR-1262 sponge to regulate aggressive behavior of RA FLSs.
In summary, we have described the regulatory function of lncRNA LINK-A in the migration, invasion, and inflammation of FLSs through 2 independent pathways; one is dependent on HIF-1α, and the other is dependent on miR-1262. Our study suggests that increased expression of lncRNA LINK-A in FLSs may contribute to the synovial aggression and inflammation that characterize rheumatoid joint abnormalities.

Methods
Reagents and antibodies. Recombinant human IL-1β, IL-6, IL-17α, and TNF-α were obtained from R&D Systems (Bio-Techne), and LPS was purchased from MilliporeSigma. DMEM, FBS, antibiotics, trypsin EDTA, PBS, and other products for cell culture were obtained from Invitrogen, Thermo Fisher Scientific. The primary antibodies used were as follows: anti-HIF-1α (catalog number 36169), anti-LRRK2 (catalog number 5559), anti-CDC42 (catalog number 2466), and anti-RhoA (catalog number 2117) were purchased from Cell Signaling Technology. Anti-alpha Tubulin (catalog number ab7291) and anti-PTK6 (catalog number ab233392) were purchased from Abcam. Anti-Rac1 (catalog number 05-389) was purchased from MilliporeSigma. Anti-Lamin B (catalog number 66095-1-Ig) was obtained from Proteintech. or control siRNA (siC) or were infected with LINK-A-overexpressed lentiviruses (OE LINK-A) or control vector (Vector) or treated with miRNA inhibitors or mimics. (A) miRBase and miRDB were used to predict miRNAs that can bind to LINK-A. A total of 5 miRNAs, common to both databases, can bind to LINK-A. (B and C) Effect of LINK-A knockdown (B) or overexpression (C) on the expression of miRNAs. The miRNAs were detected using RT-qPCR. (D and E) Effects of miRNA mimics or inhibitors on migration (D) and invasion (E) of RA FLSs. Migration of RA FLSs was evaluated using a Transwell assay. Invasion was evaluated using inserts coated with Matrigel Basement Membrane Matrix. Representative images (original magnification, ×100) are shown. Graphs show the relative migration or invasion rates. (F) Luciferase reporter assay was conducted to verify the targeting effect of miR-1262 on LINK-A sequence. (G) Effect of miR-1262 inhibition or mimics on LINK-A expression. Data are presented as the mean ± SD from at least 3 independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001; ## P < 0.01 versus siC or vector or normal control (NC), by Student's 2-tailed t test (C and G) or 1- way ANOVA (B, D, E, and F)  Preparation of human synovial tissues and FLSs. Synovial tissues were obtained from active RA patients (25 women and 5 men, 35-68 years old) who were undergoing synovectomy of knee joint or total knee replacement. RA was diagnosed according to the 2010 American College of Rheumatology/European League Against Rheumatism classification criteria (44). The demographics of patients with RA are shown in Supplemental Table 1. We obtained HC synovial tissues from 22 people (18 women and 4 men, 31-57 years old) who underwent a traumatic single above-knee amputation and had no history of acute or chronic arthritis. RA synovitis was scored as follows (45): 0-1: no synovitis; 2-3: slight synovitis; 4-6: moderate synovitis; and 7-9: strong synovitis. The scores were calculated in a blinded fashion by 2 pathologists.
Synovial tissues were cut into small pieces and digested with collagenase (1 mg/mL) for 3 hours at 37°C to separate FLSs. FLSs were cultured in DMEM/F12 with 10% FBS at 5% CO 2 and 37°C. In the present study, we used cells from passages 4-6, during which time they were a homogeneous population of cells (1% CD11b positive, 1% phagocytic, and 1% FcgRII and FcgRIII receptor positive).
Microarray and data analysis. Total RNA was prepared using TRIzol (MilliporeSigma) from HC FLSs (n = 5) and RA FLSs (n = 5) and quantified by NanoDrop ND-1000 (Thermo Fisher Scientific). RNA integrity was evaluated by standard denaturing agarose gel electrophoresis. After elimination of ribosomal RNA (rRNA) using the mRNA-ONLY Eukaryotic mRNA Isolation Kit (Epicentre), the samples were amplified and transcribed into fluorescent complementary RNA (cRNA) along the entire length of the transcripts without 3′ bias, using a random priming method. The labeled cRNAs were hybridized onto the Human lncRNA Array v2.0 (8 × 60K, Arraystar). After hybridization and washing, the arrays were scanned using the Agilent Scanner G2505B (Agilent Technologies), and Agilent Feature Extraction software v10.7.3.1 (Agilent Technologies) was used to analyze the acquired array images. Quantile normalization and subsequent data processing were performed by GeneSpring GX v11.5.1 software (Agilent Technologies). We evaluated differentially expressed lncRNAs with statistical significance through volcano plot filtering and calculated a P value by Student's t test. The threshold set for up-and downregulated genes was a fold change more than 2.0 and P value less than 0.05. Heatmaps representing differentially regulated genes were produced using Cluster v3.0 software. We finally used hierarchical clustering to show the differential lncRNAs' expression pattern among the samples. The microarray data discussed in this article were deposited in the National Center for Biotechnology Information's (NCBI) Gene Expression Omnibus (GEO) database (GEO GSE181614; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE181614).
Rapid amplification of cloned complementary DNA. SMARTer RACE cDNA Amplification Kit (Clontech) was used for rapid amplification of 5′ and 3′ ends of LINK-A. The primer sequences for 5′ and 3′ RACE are listed in Supplemental Table 2. We performed RACE according to the manufacturer's instructions.
Fluorescence in situ hybridization. We performed FISH following the established protocol as described previously (46). FLSs were briefly rinsed in PBS and fixed in 4% formaldehyde in PBS (pH 7.4) for 15 minutes at room temperature. The cells were then permeabilized in freshly made 0.5 % v/v Triton X-100 in PBS containing 2 mM voriconazole on ice for 10 minutes. After being washed with 2 × saline-sodium citrate, the cells were hybridized with RNA FISH Probes (125 μM, Biosearch Technologies) at 37°C for 16 hours in a moist chamber. After incubation with RNA FISH wash buffer A for 30 minutes, the cells were counterstained with DAPI. Then, the cells were incubated with RNA FISH wash buffer B for 2 minutes and imaged by a confocal laser scanning microscope (LSM710; Carl Zeiss) equipped with LSM ZEN 2008 software.
In situ hybridization. To evaluate the expression pattern of LINK-A in synovial tissues, ISH was performed with RNA FISH Probe (Biosearch Technologies). Briefly, after deparaffinization and rehydration, the sections were digested with proteinase K (15 μg/mL) for 10 minutes at 37°C and subsequently dehydrated. Then the sections were hybridized with RNA FISH Probe (125 μM) at 37°C for 16 hours. After incubation with RNA FISH wash buffer A for 30 minutes, they were counterstained with DAPI for 30 minutes at 37°C. Finally, the sections were incubated with RNA FISH wash buffer B for 5 minutes, then imaged by a confocal laser scanning microscope (LSM710; Carl Zeiss) equipped with LSM ZEN 2008 software.
After being washed and blocked, the sections were incubated at room temperature for 1 hour with RNA FISH Probes. Finally, the sections were stained with nitroblue tetrazolium/5-Bromo-4-chloro-3-indolyl phosphate (Roche Life Science), counterstained with nuclear fast red, and mounted.
Quantitative real-time PCR. Total RNA was obtained using the Takara PrimeScript RT Reagent Kit (Takara Bio) according to the manufacturer's instructions, and RT-qPCR was performed using the Bio-Rad CFX96 system (Bio-Rad Laboratories). The primers employed for real-time PCR are listed in Supplemental Table 3. To quantify the relative expression of each gene, we normalized Ct values to the endogenous JCI Insight 2021;6(23):e146757 https://doi.org/10.1172/jci.insight.146757 reference (ΔCt = Ct target -Ct GAPDH) and compared them using a calibrator and the ΔΔCt method (ΔΔCt = ΔCt sample -ΔCt calibrator). All experiments were performed in triplicate.
Cell nuclear protein extraction. Nuclear protein was extracted using NE-PER Nuclear and Cytoplasmic Extraction Reagents kit (Thermo Fisher Scientific). RA FLSs were harvested to microcentrifuge tubes and pelleted by centrifugation at 500g for 2-3 minutes. The supernatant was carefully removed, and the cell pellet was dried and then suspended. Ice-cold CER II was added to the tube. The tube was vortexed and incubated, then centrifuged for 5 minutes at maximum speed in a microcentrifuge (~16,000g). The insoluble (pellet) fraction, which contains nuclei, was suspended in ice-cold Nuclear Extraction Reagent. We continued vortexing it for 15 seconds every 10 minutes, for a total of 40 minutes. The tube was centrifuged at maximum speed (~16,000g) in a microcentrifuge for 10 minutes. Finally, the supernatant (nuclear extract) fraction was harvested to a clean, prechilled tube.
Western blot. Protein concentrations were measured using the Bicinchoninic Acid Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of protein were solubilized in Laemmli buffer (62.5 mM Tris-HCl pH 6.8, 10% glycerol, 2% SDS, 5% β-mercaptoethanol, and 0.00625% bromophenol blue), boiled for 5 minutes, separated by SDS-PAGE, and transferred to nitrocellulose membranes. The membranes were probed with indicated primary antibodies in TBS/Tween 20 containing 5% nonfat milk at 4°C overnight. Then the membranes were incubated with the secondary antibodies (anti-rabbit IgG, 7074, Cell Signaling Technology) for 1 hour at room temperature. Enhanced chemiluminescence (ECL; GE Healthcare Life) was used to visualize immunoreactive bands. Each blot was representative of at least 3 similar independent experiments.
Infection of overexpression lentivirus. LINK-A overexpression lentivirus was purchased from Genechem; we used empty vector lentivirus expressing GFP as NCs only. Cells were cultured in 6-well plates until 60% confluent and infected with lentivirus particles in the presence of 10 g/mL polybrene at a MOI of 30. The cells were cultured for at least 3 days before further experiments.
Transfection of siRNA. The LINK-A, HIF-1α, PTK6, and LRRK2 siRNAs and nonsilence control siR-NAs were purchased from RiboBio. The target sequences of siRNAs are listed in Supplemental Table 4. The cells were seeded in 6-well plates at 60%-70% confluence and transiently transfected with the above siRNAs (100 nM) or the corresponding control siRNA using Lipofectamine 3000 (Thermo Fisher Scientific). Experiments were performed 48 hours after transfection.
Chemotaxis migration and invasion in vitro. Chemotaxis migration of FLSs was performed using the Boyden chamber method with a filter 6.5 mm in diameter and at a pore size of 8.0 μm (Transwell; Corning Labware Products). Briefly, DMEM containing 10% FBS was put in the lower wells, and suspended FLSs (6 × 10 4 cells/mL) in serum-free DMEM were put in the upper wells. The plate was incubated at 37°C in 5% CO 2 for 12 hours. Then the nonmigrating cells were removed from the filter's upper surface using a cotton swab, and the filters were fixed in methanol for 15 minutes and stained with 0.1% crystal violet for 15 minutes. The chemotaxis was quantified using an optical microscope to count the stained cells that had migrated to the lower side of the filter. The stained cells were counted as the mean number of cells per 10 random fields for each assay. For the measurement of invasion in vitro, similar experiments were conducted using inserts coated with BD Matrigel Basement Membrane Matrix (BD Biosciences).
Wounding migration. RA FLSs were covered to 70% confluence on 35 mm culture dishes, then were serum starved for 12 hours and subsequently wounded using 200 μL pipette tips. The culture dishes were washed 3 times with PBS to remove detached cells, and the remaining cells were grown in DMEM containing 10% FBS for 48 hours. The migration was quantified by counting the cells that had moved beyond a reference line.
Evaluation of in vivo invasion of RA FLSs into human cartilage implants. Cultured FLSs were suspended in sterile saline solution to a final volume of 60 mL for each sponge before implantation. The normal human cartilage, obtained from nonarthritic patients undergoing knee surgery for traumatic injuries, was cut into 5 to 8 mm 3 pieces. For implantation, 2 implants containing cartilage and the same population of RA FLSs were inserted under the skin at the left flank of 4-week-old female SCID mice (Guangdong Medical Laboratory Animal Center, Guangzhou, China).
The mice were raised and housed under standard conditions with air filtration (20 ± 2°C; 12 hours light/12 hours dark). Briefly, on the day of implantation, normal human cartilage was cut into pieces. A piece of cartilage (5-8 mm 3 ) was inserted with a cube of insert sponge (80 mm 3 ). The sponge was soaked with 4 × 10 5 FLSs suspended in sterile saline. Three pieces of sponge containing FLSs and cartilage were inserted into the skin of the anesthetized mouse under sterile conditions. After 50 days of implantation,