GALNT1 drives aggressive phenotypes of rheumatoid synoviocytes via NEK9 O‑glycosylation

Figures

• 

  Figure 1. Enhanced O-glycosylation and GALNT1 expression in RA FLSs and synovial tissues. (A) An overview of the numbers of DEGs with a fold-change of 2.0 or greater and P less than 0.05 identified by microarray analysis of RA FLSs (n = 5) and HC FLSs (n = 5) shown in heatmap. (B) KEGG enrichment analysis of the upregulated DEGs. (C) Expression levels of indicated genes in RA FLSs (n = 5) and HC FLSs (n = 5) groups. Differences between RA and HC groups were analyzed by 2-tailed unpaired t test with Benjamini-Hochberg FDR correction for multiple testing. (D) Differential expression of the indicated genes was validated by qRT-PCR. Differences between RA (n = 12) and HC (n = 14) groups were analyzed using a 2-tailed unpaired t test. (E) The protein expression level of GALNT1 in RA FLSs (n = 4) and HC FLSs (n = 4) was measured by Western blot. Differences were analyzed using 2-tailed Welch’s t test. (F) Expression and subcellular localization of GALNT1 in RA FLSs and HC FLSs were visualized by IF. (G) GALNT1 expression in RA synovial tissues (n = 66) and HC synovial tissues (n = 10) measured by IHC. Original magnification, ×1,000. (H) Semiquantitative analysis was performed for the evaluation…

• 

  Figure 2. GALNT1 promotes aggressiveness of RA FLSs and severity of CIA in rats. (A) The cell scratch assay was performed for the measurement of in vitro migration of RA FLSs. Original magnification, ×50. n = 5 independent experiments. Cell migration assays (B) and invasion assays (C) by Transwell chambers were carried out in RA FLSs. Scale bars: 100 μm. Original magnification, ×200. n = 5 independent experiments. Data were normalized to the siNC control and analyzed by a 2-tailed 1-sample t test (theoretical value = 1). (D) The actin cytoskeleton in RA FLSs was stained with phalloidin and visualized under a fluorescence microscope. The white arrow indicates the formation of lamellipodia; yellow arrow indicates the formation of filopodia. Original magnification, ×1,000. Secretion of IL-6 (E) and CCL-2 (F) by RA FLSs was measured by assays. Data were normalized to the siNC control and analyzed by a 2-tailed 1-sample t test (theoretical value = 1). n = 4 independent experiments. For in vivo CIA experiments, rats were divided into normal control group (n = 3), CIA+vector group (n = 10), and CIA+shGalnt1 group (n = 10). (G) Effect of adenovirus-mediated knockdown of Galnt1 on joint swelling…

• 

  Figure 3. GALNT1 functions by interacting with NEK9. (A). Identification of co-IP products by GALNT1 antibody by mass spectrometry and GO analysis of the immunoprecipitated products. (B) GALNT1 antibody co-IP followed by Western blot was conducted to validate the binding of the indicated proteins to GALNT1. (C) The binding of GALNT1 to NEK9 was confirmed by NEK9 antibody co-IP. Visualization of the colocalizations between GALNT1 and NEK9 (D) and NEK9 and actin (E) are shown with representative images. Original magnification, ×1,000. RA FLSs were transfected with NEK9 siRNA and the rates of migration (F) and invasion (G) were evaluated by Transwell assays. n = 5 independent experiments. Secretion of IL-6 (H) and CCL-2 (I) by RA FLSs was measured by ELISA assays. n = 5 (IL-6) and n = 6 (CCL-2) independent experiments. Data were normalized to the siNC control and analyzed by a 2-tailed 1-sample t test (theoretical value = 1). GALNT1 overexpression RA FLSs were transfected with NEK9 siRNA and subjected to migration assays (J) and invasion assays (K). n = 5 independent experiments. Data were normalized to the siNC/vector control. One-way repeated-measures ANOVA with Greenhouse-Geisser…

• 

  Figure 4. GALNT1-mediated O-glycosylation of NEK9 in RA FLSs promotes migration and invasion. (A) Expression of NEK9 in RA synovial tissues (n = 8) and HC synovial tissues (n = 8) by IHC. Original magnification, ×400. (B) Semiquantitative NEK9 analysis; Mann-Whitney U test. (C) Expression of NEK9 in RA FLSs (n = 6) and HC FLSs (n = 3) by Western blot; 2-tailed 1-sample t test. (D) GALNT1 siRNA-transfected RA FLSs analyzed by Western blot; 2-tailed 1-sample t test (n = 5 independent experiments, normalized to siNC). (E) Lysates of RA FLSs were treated with neuraminidase or PNGase F, pulled down by VVL agarose beads, and then analyzed by Western blot with an anti-NEK9 antibody. Total lysate was used as a loading control. (F) Lysates of HC FLSs and RA FLSs were pulled down by VVL agarose beads and then analyzed by Western blot. (G) Lysates of GALNT1 knockdown or GALNT1 overexpression RA FLSs were pulled down by VVL agarose beads and then analyzed by Western blot. (H) Lysates of GALNT1 knockdown or GALNT1 overexpression RA FLSs were immunoprecipitated with an anti-NEK9 antibody, and the products were analyzed with VVL lectin blot. NEK9 overexpression RA FLSs were treated with O-glycosylation…

• 

  Figure 5. NEK9 regulates cytoskeleton organization in RA FLS. (A) RNA-Seq was performed on the NEK9-silenced group and the control group, followed by GO clustering analysis of the differentially downregulated genes. (B) Differentially expressed genes enriched in the GO term of actin cytoskeleton organization. (C) Expression levels of indicated genes in RA (n = 5) and HC (n = 5) groups (microarray). Data are log2-transformed normalized intensity. Differences were analyzed by 2-tailed t test with Benjamini-Hochberg FDR correction. (D) After silencing of NEK9 expression, qRT-PCR was used to detect the expression levels of mRNAs. Data were normalized to siNC and analyzed by 2-tailed 1-sample t test against 1. n = 3 independent experiments. (E) Staining of actin cytoskeleton by phalloidin. White arrows indicate lamellipodia; yellow arrows indicate filopodia. Scale bars: 20 μm (left). Original magnification, ×1000 (left). (F) RA FLSs were infected with vector or NEK9-OE or NEK9-mut and then analyzed by qRT-PCR. Data were normalized to vector control. NEK9-OE and NEK9-mut were compared with vector by 2-tailed paired t tests. n = 3 independent experiments. Data are mean ± SD. *P < 0.05, **P <…

• 

  Figure 6. NEK9 regulates the migration and invasion of RA FLSs through the regulation of the ER stress pathway. (A) GO clustering analysis of the differentially upregulated genes in the siNEK9 group. (B) Upregulated genes enriched in the GO term of response to ER stress after NEK9 knockdown. (C) HC FLSs and RA FLSs were transfected with siNC, siGALNT1, and siNEK9 and were observed under a transmission electron microscope. (D) RA FLSs were infected with vector or NEK9-OE or NEK9-mut and were observed with a transmission electron microscope. (E) Relative expression of HSPA5 mRNA in RA FLSs transfected with siNEK9 or siGALNT1. n = 4 independent experiments. (F) Relative expression of HSPA5 mRNA in RA FLSs transfected with NEK9-OE or NEK9-mut. n = 3 independent experiments. Data were normalized to the respective control (siNC for E, vector for F), and comparisons between each treatment and control were performed using paired t tests. (G) GRP78 expression in CIA rat synovium detected by IHC. (H) Semiquantitative analysis was performed for the evaluation of GRP78 expression in vector control group synovial tissues (n = 8) and shGalnt1 group synovial tissues (n = 10). Differences were analyzed…