Abstract

   Protein citrullination modification plays a pivotal role in the
   pathogenesis of rheumatoid arthritis (RA), and anti-citrullinated
   protein antibodies (ACPAs) are extensively employed for clinical
   diagnosis of RA. However, there remains limited understanding regarding
   specific citrullinated proteins and their implications in the
   progression of RA. In this study, we screen and verify insulin-like
   growth factor-2 mRNA binding protein 1 (IGF2BP1) as a novel
   citrullinated protein with significantly elevated citrullinated level
   in RA. Autoantibodies against citrullinated IGF2BP1 are further
   detected in serum and synovial fluid samples from RA patients, which
   are positively correlated with erythrocyte sedimentation rate (ESR) and
   disease activity score 28 (DAS28). Transcriptomic sequencing and
   functional verification show that citrullination at the R167 site of
   IGF2BP1 promotes the proliferation, migration, and invasion of RA
   fibroblast-like synoviocytes (RA-FLSs) by improving the mRNA stability
   of Semaphorin 3D (SEMA3D). Experiments in collagen-induced arthritis
   (CIA) mice, the classical animal model of RA, show that IGF2BP1 R176K
   point mutation (Igf2bp1^R167K/R167K) mice exert reduced inflammatory
   response, clinical scores, and joint destruction. At a molecular level,
   citrullination of IGF2BP1 promotes the stability of SEMA3D mRNA by
   promoting the interaction between IGF2BP1 and its cofactor ELAV-like
   protein 1 (ELAVL1), thereby promoting the invasiveness of RA-FLSs. In
   this study, a new citrullinated protein of IGF2BP1 is discovered, and
   the molecular mechanism of its citrullinated modification promoting the
   progression of RA disease is elucidated, which provides theoretical
   basis for the diagnosis and treatment of RA.

   graphic file with name 42003_2025_7492_Figa_HTML.jpg

   Subject terms: Post-translational modifications, Rheumatoid arthritis,
   Post-translational modifications
     __________________________________________________________________

   Citrullinated IGF2BP1 enhances the aggressiveness of fibroblast-like
   synoviocytes by increasing the mRNA stability of SEMA3D, autoantibodies
   against citrullinated IGF2BP1 are detected in patients with rheumatoid
   arthritis and positively correlate with disease severity.

Introduction

   Rheumatoid arthritis (RA) is the most common chronic autoimmune
   disease, affecting approximately 0.5% to 1% of the global
   population^[38]1. Its primary clinical feature is erosive joint
   inflammation, which eventually leads to joint deformity and functional
   impairment, significantly impacting the living quality of RA
   patients^[39]1,[40]2. Fibroblast-like synoviocytes (FLSs) are the
   predominant cell type in the RA inflamed synovium and act as key
   players in RA pathogenesis^[41]3,[42]4. Recent studies have shown these
   cells are heterogeneous groups performing distinct and specialized
   functions^[43]5,[44]6. According to their anatomical location, FLSs in
   the lining layer (THY1^- subset) act as bone effector cells,
   contributing to joint damage by secreting matrix metalloproteinases
   (MMPs) and RANKL, while FLSs in the sublining layer (THY1^+ subset),
   which are the predominant subset expanded in RA, play immune effector
   roles by producing a distinct repertoire of chemokines and
   cytokines^[45]7,[46]8. Throughout RA pathogenesis, the morphology and
   biological activity of FLSs shows a unique transformation and
   continuous activation^[47]4. However, the underlying mechanisms by
   which FLSs become imprinted aggressors in RA remain fully elucidated.

   As an irreversible post-translational modification of proteins,
   citrullination plays an important role in the occurrence and
   development of RA^[48]9. Under the catalysis of peptidyl arginine
   deiminase (PAD) enzymes, arginine residues of proteins are converted to
   citrulline residues^[49]10. On the one hand, it increases the
   immunogenicity of the proteins and induces the corresponding
   autoantibodies^[50]11. As a clinical diagnostic indicator of RA,
   anti-citrullinated protein antibodies (ACPAs) are autoantibodies
   against citrullinated proteins, and the diagnostic sensitivity and
   specificity are 67% and 95%, respectively^[51]12. On the other hand,
   protein citrullination, like other protein modifications such as
   protein phosphorylation and ubiquitination, is involved in a variety of
   biological functions through cell signaling transduction^[52]10,[53]13.
   For instance, citrullination of NF-κB p65 facilitates its interaction
   with importin α3, thus promoting the p65 nuclear transfer^[54]14.
   Similarly, citrullination of MMP9 enhances its affinity and enzymatic
   activity for gelatin^[55]15. Furthermore, citrullination of C-X-C motif
   chemokine 5 (CXCL5), a neutrophil recruitment chemokine, enables its
   new capacity for recruiting monocytes^[56]16. Therefore, in this study,
   we screened and verified the citrullinated proteins in RA-FLSs,
   identified insulin-like growth factor-2 mRNA binding protein 1
   (IGF2BP1) as a new citrullinated protein, and further explored its role
   in the development and progression of RA.

   IGF2BP1, a member of the highly conserved IGF2BP RNA-binding protein
   family alongside IGF2BP2 and IGF2BP3, has been implicated in promoting
   multiple oncogenic processes like cell proliferation and invasion
   through stabilizing several mRNA targets such as KI67 and
   MYC^[57]17,[58]18. However, current studies predominantly focus on the
   role of IGF2BPs in malignant tumors. Recent studies have revealed that
   IGF2BP3 promotes the invasiveness of RA-FLSs^[59]19,[60]20.
   Nevertheless, the specific role of IGF2BP1 in RA-FLSs remains unclear.
   In addition, some studies have found that post-translational
   modification of IGF2BP1 plays an important role in tumorigenesis and
   development^[61]21. Phosphorylation of IGF2BP1 promotes proper splicing
   of mRNA, methylation promotes its recognition and binding to mRNA, and
   ubiquitination promotes its carcinogenic activity^[62]22–[63]24.
   However, the biological function of IGF2BP1 citrullination modification
   and its role in the occurrence and development of RA need to be further
   explored.

   In this study, we found that IGF2BP1 was a novel citrullinated protein,
   and its citrullinated level was significantly increased in RA-FLSs and
   promoted the proliferation and invasion function of RA-FLSs. Meanwhile,
   we also found that autoantibodies against citrullinated IGF2BP1
   (cit-IGF2BP1) were elevated in RA patients, which was a potential
   clinical diagnostic indicator of RA.

Result

The citrullinated modification of IGF2BP1 was identified and its
citrullinated level was significantly increased in RA patients

   In this study, we assessed the global citrullination level of proteins
   in RA-FLSs and HC-FLSs, finding a significant increase in RA-FLSs,
   particularly around 35–40 kD and 70 kD (Fig. [64]1A). Mass spectrometry
   was then used to identify specific citrullinated proteins
   immunoprecipitated by the pan-anti-citrulline antibody F95 in RA-FLSs
   (Fig. [65]1B). A total of 51 proteins were screened (Supplementary
   Table [66]1), in which four were identified as citrullinated proteins
   by a mass increase of 0.98 Da and neutral loss (isocyanic acid of 43 Da
   and ammonia of 17 Da in peptide collision-induced dissociation) on the
   peptide. In addition to three confirmed citrullinated proteins
   (Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), annexin A2 (ANXA2)
   and Alpha-enolase (ENO1)), a novel citrullinated protein of IGF2BP1 was
   identified at Arg 167 (R167) (Fig. [67]1C). Notably, the molecular
   weight of IGF2BP1 (72 kD) corresponded to the region of significantly
   elevated citrullination observed in RA-FLSs (Fig. [68]1A). Further
   analysis demonstrated that citrullination of IGF2BP1 was elevated in
   RA-FLSs (Fig. [69]1D), while its overall protein levels remained
   unchanged (Supplementary Fig. [70]1A). Additionally, when RA patients
   were stratified based on ACPA positivity, we found that the
   citrullination of IGF2BP1 was higher in ACPA^+ RA patients compared to
   ACPA^− RA patients (Fig. [71]1E). Furthermore, cell sorting of FLSs
   using THY1 magnetic beads showed that citrullination of IGF2BP1 was
   elevated in both THY1^+ and THY1^− RA-FLSs compared to HC
   (Supplementary Fig. [72]1B, C). To identify additional citrullination
   sites of IGF2BP1, in vitro citrullination was carried out using PAD2
   enzyme (Fig. [73]1F), followed by mass spectrometry analysis. Three
   additional citrullination sites of IGF2BP1 were identified at R168,
   R174, and R575 (Fig. [74]1G, Supplementary Fig. [75]1D, E). Taken
   together, the novel citrullinated protein of IGF2BP1 was found that
   increased in RA-FLSs.

Fig. 1. IGF2BP1 was identified and verified as a new citrullinated protein in
RA-FLSs.

   [76]Fig. 1
   [77]Open in a new tab

   A The global citrullination level of proteins in FLSs from HC (n = 4)
   and RA (n = 4) were analyzed by western blot. B MS1 chromatograms of
   citrullinated proteins from RA-FLSs. C The MS2 fragmentation spectra of
   IGF2BP1 indicating the citrullination site of R167. A mass increase of
   0.98 Da, together with neutral loss of NHCO (43 Da) and NH3 (17 Da)
   were marked in red, which are diagnostic criterions for protein
   citrullination. D Citrullination levels of IGF2BP1 in FLSs from HC
   (n = 9) and RA (n = 12) were detected by western blot using F95
   antibody. E Citrullination levels of IGF2BP1 in RA-FLSs from ACPA^+
   (n = 8) and ACPA^− (n = 4) patients were detected by western blot using
   F95 antibody. F The citrullination level of recombinant human IGF2BP1
   was detected by western blot. G Schematic illustration of the
   citrullination sites of IGF2BP1 identified by Mass spectrometry. All
   the reactions were performed in triplicate. The data were presented as
   mean ± SD. Unpaired t-test (A, D, E) was used as parametric test.
   *P < 0.05, **P < 0.01.

Citrullination of IGF2BP1 at R167 promotes the pathogenicity of RA-FLSs

   To investigate the specific citrullination sites of IGF2BP1, IGF2BP1
   was firstly knocked down in RA-FLSs and rescued with various IGF2BP1
   mutants, including IGF2BP1^R167K, IGF2BP1^R168K, IGF2BP1^R174K,
   IGF2BP1^R575K, or a quadruple mutant IGF2BP1^4R-K (Fig. [78]2A). Three
   short hairpin RNA (shRNA) were designed to achieve IGF2BP1 knockdown,
   among which shRNA-1 was selected for its best suppressing effect
   (Fig. [79]2B). The IGF2BP1-KD FLSs were then rescued with the mutant
   plasmids above. Results showed that the IGF2BP1^R167K effectively
   attenuated the citrullination of IGF2BP1 (Fig. [80]2C), indicating that
   the R167 residue is the primary citrullination site of IGF2BP1 in
   RA-FLSs. Considering that FLSs exhibited the characteristics of
   excessive proliferation, migration, and invasion in RA progression, we
   further explored the effects of IGF2BP1 citrullination on these
   pathogenic phenotypes in RA-FLSs. Results showed that the IGF2BP1^R167K
   decreased the proliferation of RA-FLSs than the other IGF2BP1 mutants
   in the Edu assay (Fig. [81]2D). Additionally, the IGF2BP1^R167K
   suppressed the migration and invasion abilities of RA-FLSs in the wound
   healing assay (Fig. [82]2E) and Transwell invasion assay (Fig. [83]2F).
   Moreover, IGF2BP1^R167K significantly reduced the formation of invasive
   pseudopodia (Fig. [84]2G). We next examined the effects of IGF2BP1
   citrullination on the secretion of cytokines and MMPs by RA-FLSs.
   Results showed that IGF2BP1^R167K led to a decrease in IL-6, TNF-α,
   MMP3, and MMP9 secretion of RA-FLSs (Supplementary Fig. [85]2A, B, D,
   E), and it showed no effects on IL-1β and MMP13 (Supplementary
   Fig. [86]2C, F). To further clarify this, we constructed an
   IGF2BP1^3R-K mutant (R168K, R174K and R575K), leaving R167 as the only
   potential citrullination site, and performed corresponding experiments.
   Results showed that the IGF2BP1^3R-K group had similar proliferation,
   migration, invasion, and pro-inflammatory abilities as the IGF2BP1^WT
   group (Supplementary Fig. [87]2G−N), demonstrating that the
   IGF2BP1^3^R-K mutant, with R167 as the only citrullination site, is
   sufficient to sustain the pathogenic function of citrullinated IGF2BP1.
   Collectively, these results suggested that the citrullination of
   IGF2BP1 at the R167 residue promotes the pathogenicity of RA-FLSs.

Fig. 2. Citrullination of IGF2BP1 at R167 promotes the aggressive behaviors
of RA-FLSs.

   [88]Fig. 2
   [89]Open in a new tab

   A Schematic illustration of IGF2BP1^R167K, IGF2BP1^R168K,
   IGF2BP1^R174K, IGF2BP1^R575K, and IGF2BP1^4R-K mutants. B The effect of
   IGF2BP1 knockdown in RA-FLSs was detected by western blot. ShIGF2BP1-1
   showed best effect and was used for following experiments. C The
   citrullination level of IGF2BP1 in RA-FLSs was detected by western
   blot. The IGF2BP1-KD RA-FLSs were rescued with IGF2BP1 mutants, then
   the citrullination level of IGF2BP1 was detected using F95 antibody. D
   Cell proliferation of RA-FLSs was detected using EdU experiment. Total
   cells were detected by Hoechst staining (blue), and cells with active
   DNA replication were detected by EdU staining (red). E Cell migration
   of RA-FLSs detected by scratch assay. Statistical significance was
   determined by ANOVA of repeated measurements. The wound closure ratio
   of each group was calculated using the following equation: migrated
   cellular area/scratched area × 100%. F Cell invasion of RA-FLSs
   detected by Transwell assay. Images of RA-FLSs on the undersurface of a
   filter were captured under the microscope (original magnification ×
   100) and the number of FLSs per field in each group was quantified by
   ImageJ. G Formation of invasive pseudopodia were detected by F-actin
   staining with TRITC-Phalloidin. All the experiments were performed in
   triplicate. The data were presented as mean ± SD. One-way ANOVA (C, D,
   F, G) followed by Dunnett test was used as parametric test. *P < 0.05,
   **P < 0.01, ***P < 0.001, ****P < 0.0001.

Cit-IGF2BP1 promoted SEMA3D expression by maintaining its mRNA stability

   To screen the downstream target genes of cit-IGF2BP1, the RNA
   sequencing (RNA-seq) was conducted. A total of 538 differentially
   expressed genes (DEGs) were identified ( | log2FC | > = 1 & P < 0.05),
   including 280 upregulated genes and 258 downregulated genes in the
   IGF2BP1-KD RA-FLSs rescued with IGF2BP1^WT than that rescued with
   IGF2BP1^R167K (Fig. [90]3A). Gene Ontology (GO) enrichment analysis
   revealed that DEGs were significantly enriched in cell-cell adhesion,
   extracellular matrix organization, and cell-cell signaling
   (Fig. [91]3B). Additionally, KEGG pathway enrichment analysis indicated
   the DEGs were enriched in the ECM-receptor interaction, Chemokine
   signaling and Rheumatoid arthritis signaling pathway (Fig. [92]3C).
   Based on the fact that IGF2BP1 is a classical RNA binding protein, we
   screened the mRNAs that could interact with IGF2BP1 by the ENCORI,
   RM2Target, and RNA INTER database combined with RNA-seq data in this
   experiment. In summary, 12 candidate genes were screened (Fig. [93]3D,
   E), with Semaphorin 3D (SEMA3D) selected for further validation due to
   its similar function with IGF2BP1, such as enhancing cell
   proliferation, migration, and invasion^[94]25–[95]27. In the subsequent
   validation experiments, we found that the citrullination of IGF2BP1
   promoted the expression of SEMA3D (Fig. [96]3F, G). Furthermore, the
   interaction between IGF2BP1 and SEMA3D mRNA was also verified by RNA
   immunoprecipitation (RIP)-qPCR, and this interaction was impaired by
   IGF2BP1 R167K mutation (Fig. [97]3H). Moreover, citrullination of
   IGF2BP1 improved the mRNA stability of SEMA3D (Fig. [98]3I). Taken
   together, cit-IGF2BP1 promoted SEMA3D expression by maintaining its
   mRNA stability.

Fig. 3. Cit-IGF2BP1 promoted SEMA3D expression by maintaining its mRNA
stability.

   [99]Fig. 3
   [100]Open in a new tab

   A Volcano plots of differently expressed genes (DEGs) between
   IGF2BP1-KD RA-FLSs rescued with IGF2BP1^WT and IGF2BP1^R167K (n = 3 per
   group). B The enriched Gene Ontology (GO) analysis of the DEGs. C The
   enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway
   analysis of the DEGs. D The screening of mRNAs interacting with IGF2BP1
   by the ENCORI, RM2Target, and RNA INTER database, three databases that
   supported by large-scale CLIP-Seq for exploring protein-RNA
   interactions. E The overlapping of the mRNAs that interacted with
   IGF2BP1 and the DEGs of the RNA-seq data from this experiment. F, G The
   verification of SEMA3D as a target gene of cit-IGF2BP1 by qPCR and
   western blot. H The interaction of IGF2BP1 and SEMA3D mRNA was detected
   by RIP-qPCR. I The stability of SEMA3D mRNA was detected by q-PCR.
   Statistical significance was determined by ANOVA of repeated
   measurements. RA-FLSs were treated with actinomycin D and harvested at
   0, 1, 2, 4, and 8 h. All the experiments were performed in triplicate.
   The data were presented as mean ± SD. Unpaired t-test (F, G, H) was
   used as parametric test. *P < 0.05, **P < 0.01.

Cit-IGF2BP1 enhances the pathogenicity of RA-FLSs via upregulating the
expression of SEMA3D

   Since there has no report on the function of SEMA3D in RA disease
   progression, we found that the expression of SEMA3D was elevated in
   RA-FLSs compared to HC-FLSs (Fig. [101]4A, B), which positively
   correlated with IGF2BP1 citrullination (Fig. [102]4C). SEMA3D was then
   knocked down by shRNAs (Supplementary Fig. [103]3A). Results showed
   SEMA3D knockdown attenuated the proliferation, migration, and invasion
   of RA-FLSs (Fig. [104]4D, F, H, J). Additionally, SEMA3D overexpression
   enhanced the proliferation, migration, and invasion of RA-FLSs
   (Supplementary Fig. [105]3B–E). Moreover, the reduced function of FLSs
   induced by IGF2BP1 R167K mutation was reversed by SEMA3D overexpression
   (Fig. [106]4E, G, I, K). Collectively, our data suggested that
   cit-IGF2BP1 enhances the pathogenicity of RA-FLSs via upregulating the
   expression of SEMA3D.

Fig. 4. Cit-IGF2BP1 promotes aggressive behaviors of RA-FLSs via upregulation
of SEMA3D.

   [107]Fig. 4
   [108]Open in a new tab

   A, B The expression of SEMA3D in FLSs from HC (n = 9) and RA (n = 12)
   was detected by qPCR and western blot. C The correlation between
   IGF2BP1 citrullination and SEMA3D mRNA in FLSs (n = 21) was evaluated
   by Pearson correlation coefficient. D, E Cell proliferation of RA-FLSs
   detected by EdU experiment. Total cells were detected by Hoechst
   staining (blue), and cells with active DNA replication were detected by
   EdU staining (red). F, G Cell migration of RA-FLSs detected by scratch
   assay. Statistical significance was determined by ANOVA of repeated
   measurements. The wound closure ratio of each group was calculated
   using the following equation: migrated cellular area/scratched area ×
   100%. H, I Cell invasion of RA-FLSs detected by Transwell assay. Images
   of RA-FLSs on the undersurface of a filter were captured under the
   microscope (original magnification × 100) and the number of FLSs per
   field in each group was quantified by ImageJ. J, K Formation of
   invasive pseudopodia were detected by F-actin staining with TRITC -
   Phalloidin. All the experiments were performed in triplicate. The data
   were presented as mean ± SD. Unpaired t-test (A, B) and one-way ANOVA
   (D, E, H–K) followed by Dunnett test were used as parametric test.
   *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Citrullination modification of IGF2BP1 promotes SEMA3D mRNA stability by
facilitating the IGF2BP1–ELAVL1 interaction

   In order to further clarify the molecular mechanism of IGF2BP1
   citrullination promoting SEMA3D expression, given that the binding of
   IGF2BP1 to its cofactors is crucial for maintaining the stability of
   its target mRNA, we pulled down the proteins interacting with
   IGF2BP1^WT and IGF2BP1^R167K. A differential protein band at nearly
   35–40 kD in the IGF2BP1^R167K group was identified by silver-staining
   and subsequently detected by mass spectrometry (Fig. [109]5A). Among
   the 208 potential binding proteins identified, we selected the top
   three RNA-binding proteins for validation: ELAV-like protein 1
   (ELAVL1), Heterogeneous nuclear ribonucleoproteins C1/C2 (HNRNPC), and
   Heterogeneous nuclear ribonucleoprotein A0 (HNRNPA0). Co-IP assay
   demonstrated that ELAVL1 exhibited specific interaction with IGF2BP1 as
   opposed to the other two proteins (Fig. [110]5B). Furthermore, both the
   forward and reverse Co-IP assays verified that the point mutation at
   R167, rather than other citrullination sites, reduced the interaction
   between IGF2BP1 and ELAVL1 (Fig. [111]5C, Supplementary Fig. [112]4A).
   We further investigate the impact of IGF2BP1–ELAVL1 interaction on
   SEMA3D expression. Structural docking revealed the conformation of
   ELAVL1 (Arg37, Phe279, and Lys323) direct contacted with IGF2BP1
   (Glu157, Glu324, and Glu461) (Fig. [113]5D). Co-IP assays confirmed
   that mutation of ELAVL1 (R37A/F279E/K323A) reduced its binding capacity
   to IGF2BP1 (Fig. [114]5E). Subsequently, ELAVL1 was knocked down in
   RA-FLSs (Supplementary Fig. [115]4B) and rescued with ELAVL1^WT or
   ELAVL1^MUT, respectively. Results demonstrated that ELAVL1^MUT reduced
   the expression of SEMA3D (Fig. [116]5F, G). RIP-qPCR assay showed
   ELAVL1^MUT impaired the interaction between IGF2BP1 and SEMA3D mRNA
   (Fig. [117]5H), thus decreased the mRNA stability of SEMA3D
   (Fig. [118]5I). These results indicated that IGF2BP1 citrullination
   enhances its binding capacity to the co-factor of ELAVL1, thus promotes
   the expression of SEMA3D.

Fig. 5. Citrullination modification of IGF2BP1 promotes SEMA3D mRNA stability
by facilitating the IGF2BP1–ELAVL1 interaction.

   [119]Fig. 5
   [120]Open in a new tab

   A Proteins that interacted with the HA-IGF2BP1^WT and HA-IGF2BP1^R167K
   were determined by silver-staining and identified by Mass Spectrometry.
   B The interaction between IGF2BP1 and HNRNPC, HNRNPA0, or ELAVL1 in
   RA-FLSs was analyzed by Co-IP and western blot. C The interaction
   between HA-IGF2BP1^WT or HA-IGF2BP1^R167K with FLAG-ELAVL1 in RA-FLSs
   was analyzed by forward and reverse Co-IP assays and western blot. D
   Structural docking between IGF2BP1 and ELAVL1. IGF2BP1 and ELAVL1 were
   colored using blue and light pink, respectively. E The interaction
   between FLAG-ELAVL1^WT or FLAG-ELAVL1^MUT with HA-IGF2BP1 in RA-FLSs
   was analyzed by forward and reverse Co-IP assays and western blot. F, G
   The expression of SEMA3D was detected by qPCR and western blot. H The
   interaction between IGF2BP1 and SEMA3D mRNA was detected by RIP-qPCR. I
   The stability of SEMA3D mRNA was detected by q-PCR. Statistical
   significance was determined by ANOVA of repeated measurements. RA-FLSs
   were treated with actinomycin D and harvested at 0, 1, 2, 4, and 8 h.
   All the experiments were performed in triplicate. The data were
   presented as mean ± SD. Unpaired t-test (F, G, H) was used as
   parametric test. *P < 0.05.

IGF2BP1 R167K point mutation alleviated the joint destruction in
collagen-induced arthritis (CIA) mice

   We further generated IGF2BP1 R167K point mutation mice, and constructed
   the CIA model (Fig. [121]6A). Results showed that the citrullination
   level of IGF2BP1 and the expression of SEMA3D were decreased in the
   Igf2bp1^R167K/R167K mice than that in Igf2bp1^WT/WT mice (Fig. [122]6E,
   G). Morphological observation of CIA mice showed that
   Igf2bp1^R167K/R167K mice exhibited less paw swelling and had lower
   clinical scores (Fig. [123]6B, C). Consistently, H&E and safranin
   O-fast green staining showed that Igf2bp1^R167K/R167K mice exhibited
   less synovial hyperplasia, joint damage and bone erosion (Fig. [124]6C,
   D). In addition, Igf2bp1^R167K/R167K mice displayed decreased
   concentration of pro-inflammatory factors, including TNF-α, IL-6,
   IL-1β, and IL-17A, and increased concentration of anti-inflammatory
   factor of IL-10 (Fig. [125]6H–L). These findings suggested that
   cit-IGF2BP1 promotes the progression of RA.

Fig. 6. IGF2BP1 R167K point mutation alleviated the joint destruction in
collagen-induced arthritis (CIA) mice.

   [126]Fig. 6
   [127]Open in a new tab

   A Timeline for the construction of CIA models in the Igf2bp1^WT/WT mice
   and Igf2bp1^R167K/R167K mice. n = 6 per group. B Arthritis clinical
   scores of the CIA mice. Statistical significance was tested using the
   analysis of variance of repeated measurement. C H&E and safranin O-fast
   green staining of the knee joint sections in CIA mice. D Histological
   scores based on analysis of synovial hyperplasia, inflammatory cell
   infiltration, bone destruction, and cartilage destruction. E
   Citrullination level of IGF2BP1 in the knee joints of CIA mice was
   detected by western blot using F95 antibody. F, G The expression of
   SEMA3D in the knee joints of CIA mice was detected by qPCR and western
   blot. H–L The level of cytokines in serum of CIA mice were measured
   using ELISA, including TNF-α, IL-6, IL-1β, IL-17A, and IL-10. All the
   reactions were performed in triplicate, n = 6 per group. The data were
   presented as mean ± SD. One-way ANOVA (D–L) followed by Dunnett test
   was used as parametric test. *P < 0.05, **P < 0.01, ***P < 0.001,
   ****P < 0.0001.

The autoantibodies against cit-IGF2BP1 were detected in RA patients

   Due to the production of autoantibodies, induced by post-translational
   modification of proteins, is implicated in the pathogenesis of RA, we
   initially detected autoantibodies against cit-IGF2BP1 in a small cohort
   of RA patients using immunodot blot. The results demonstrated that
   autoantibodies against cit-IGF2BP1 were present in 26.7% of RA serum
   samples (n = 15, 4 of 15), and 33.3% of RA synovial fluid (SF) samples
   (n = 12, 4 of 12), but were absent in the HC serum (n = 15) and SF
   (n = 6) (Fig. [128]7A, B, Supplementary Fig. [129]5A, B). We then
   performed a custom ELISA to analyze anti-cit-IGF2BP1 antibodies in more
   RA patients (n = 92) and HCs (n = 82). This analysis revealed that
   39.1% of RA serum (n = 92, 36 of 92) were positive for anti-cit-IGF2BP1
   antibodies (Fig. [130]7C). Furthermore, we evaluated the clinical
   relevance of anti-cit-IGF2BP1 antibodies with RA clinical parameters.
   The results indicated that serum anti-cit-IGF2BP1 antibodies were
   weakly positively correlated with erythrocyte sedimentation rate (ESR)
   and Disease Activity Score 28 (DAS28) in RA patients (Fig. [131]7D, E).

Fig. 7. The autoantibodies against cit-IGF2BP1 were detected in RA Serum and
SF.

   [132]Fig. 7
   [133]Open in a new tab

   A, B Autoantibodies against cit-IGF2BP1 were detected by immunodot blot
   in RA serum (n = 15), HC serum (n = 15), RA SF (n = 12), and HC SF
   (n = 6) (1:5000 dilution). Representative images were shown.
   Noncit-IGF2BP1, BSA, and cit-BSA were used as control antigens. C
   Autoantibodies against cit-IGF2BP1 were detected by ELISA in RA serum
   (n = 92), HC serum (n = 82). The red line indicates the mean value and
   the blue line indicates SD. Black dashed line represents the cutoff
   value used to determine positive reactivity (white dots) in
   experimental samples. The 95% percentile of HC samples was used to
   define the cutoff for positive antibody levels. Unpaired t-test was
   used as parametric test. D, E Correlation between the serum
   anti-cit-IGF2BP1 antibodies with the erythrocyte sedimentation rate
   (ESR) and Disease Activity Score 28 (DAS28) in RA patients (n = 92)
   were evaluated by Pearson correlation coefficient. ****P < 0.0001. ns,
   no significance.

Discussion

   RA is an autoimmune disease characterized by synovitis^[134]1,[135]28.
   The disease course is accompanied by the production of a large number
   of autoantibodies induced by protein post-translational
   modification^[136]29,[137]30. Protein citrullination, as a protein
   post-translational modification, induces the production of ACPAs, which
   is currently used in the clinical diagnosis of RA^[138]31. Furthermore,
   recent studies have revealed that changes in protein function caused by
   protein citrullination play a crucial role in the progression of
   RA^[139]32,[140]33. Aripova et al. reported that citrullinated vimentin
   and fibrin inhibit apoptosis in RA-FLSs cells and stimulate
   pro-inflammatory cytokine secretion^[141]34–[142]36. Additionally,
   citrullinated autoantigens form immune complexes with ACPAs, leading to
   their deposition in joints and attraction of immune cells, resulting in
   chronic inflammation and joint destruction^[143]37,[144]38. In this
   study, we identified a novel citrullinated protein, IGF2BP1, and found
   that its excessive citrullination promoted disease progression in RA.

   The IGF2BP family, which were discovered by Nielsen et al. in
   1999^[145]39, is a class of RNA-binding proteins that can bind and
   stabilize IGF2 mRNA. This family includes IGF2BP1, 2, and 3, which have
   been found to promote tumor development by binding and stabilizing mRNA
   targets such as PTEN, KI67, c-MYC, and CD44^[146]18,[147]40,[148]41.
   Until now the correlation between IGF2BPs and RA progression remains
   controversial. Previous studies have reported that IGF2BP3 promotes
   proliferation, migration, and invasion of FLSs through Akt/MMP9 pathway
   while IGF2BP2 inhibits the invasive phenotypes of RA-FLSs and its
   secretion of pro-inflammatory cytokines such as IL-1β and
   IL-6^[149]19,[150]20,[151]42. Additionally, post-translational
   modifications like phosphorylation, ubiquitination, and methylation of
   IGF2BP1 play a carcinogenic role by regulating its ability to recognize
   target mRNA for binding^[152]22–[153]24. In this study, we observed
   elevated levels of citrullinated IGF2BP1 in both RA patients and CIA
   mice, confirming that inhibition of IGF2BP1 citrullination
   significantly reduced the proliferation, migration, and invasion of
   FLSs cells, the key pathogenic cell type of RA, thereby reducing its
   pathogenicity. Furthermore, R167K point mutation in IGF2BP1 reduced
   joint swelling, joint destruction, and bone erosion, as well as the
   production of pro-inflammatory cytokines in CIA mice. Overall, our
   findings suggested that cit-IGF2BP1 was involved in the development of
   RA, indicating its potential as a therapeutic target.

   Then, we performed RNA-seq and subsequently screened and validated
   SEMA3D as a downstream target gene of cit-IGF2BP1. SEMA3D belongs to
   the semaphorin family, which has been implicated in angiogenesis,
   cancer cell growth, invasion, and survival^[154]25,[155]26,[156]43. On
   the role of semaphorin family in RA disease progression, previous
   studies have demonstrated that SEMA3A promotes MMPs secretion and
   cartilage invasion capacity in RA-FLSs^[157]44,[158]45, while SEMA3G
   enhances macrophage migration and proliferation, thereby promoting
   synovial infiltration^[159]46. Additionally, our findings indicate that
   SEMA5A stimulates RA-FLSs proliferation and angiogenesis through its
   unique TSP-1 domain, consequently exacerbating pannus formation and
   worsening the course of RA^[160]47,[161]48. However, no previous report
   has investigated the role of SEMA3D in RA pathogenesis. In this study,
   we observed overexpression of SEAM3D in RA-FLSs along with its
   promotion of cell proliferation, migration, and invasion in these
   cells; thus suggesting its potential involvement in the pathogenesis of
   RA.

   We further elucidated the molecular mechanism underlying IGF2BP1
   citrullination-mediated regulation of SEMA3D expression. We first found
   that cit-IGF2BP1 prolonged the mRNA stability of SEMA3D. Subsequently,
   we observed that it regulated the binding affinity between IGF2BP1 and
   the cofactor ELAVL1. ELAVL1 is an RNA-binding protein involved in
   nuclear-cytoplasmic mRNA transport, as well as mRNA stability and
   translation^[162]49,[163]50. In cellular contexts, ELAVL1 colocalizes
   with IGF2BP1 within P-bodies, which are crucial sites determining mRNA
   fate^[164]51. Notably, IGF2BP1 recruits ELAVL1 to safeguard mRNA
   against degradation and facilitate translation^[165]51,[166]52.
   Silencing of ELAVL1 largely abrogates the effect of IGF2BP1 in
   maintaining mRNA stability^[167]53,[168]54. Here, our findings
   demonstrated the significance of the interaction between IGF2BP1–ELAVL1
   for SEMA3D mRNA stability and highlight how citrullination enhances
   this interaction. These results propose a novel mechanism by which
   citrullination impacts the function of RNA-binding protein IGF2BP1.

   In this study, we identified cit-IGF2BP1 as a novel autoantigen in RA.
   Using custom ELISA, we analyzed the antibodies against cit-IGF2BP1 in
   serum samples from 92 RA patients (72 ACPA^+ and 20 ACPA^−). Among the
   ACPA^+ RA patients, 48.6% (35 of 72) exhibited anti-cit-IGF2BP1
   antibodies, while only 5% (1 of 20) of ACPA^− RA patients exhibited
   these antibodies, suggesting that in these ACPA^+ RA patients exhibited
   anti-cit-IGF2BP1 antibodies, targeted therapies against cit-IGF2BP1 may
   enhance precision medicine approaches.

   In conclusion, our findings demonstrate that IGF2BP1 is a novel
   citrullinated protein, thereby confirming the underlying molecular
   mechanism through which cit-IGF2BP1 promotes the invasiveness of
   RA-FLSs by maintaining the stability of SEMA3D mRNA, ultimately
   promoting the progression of rheumatoid arthritis. This study unveils a
   promising target for both diagnosis and therapeutic intervention in RA.

Methods

   Details of the cell sorting of THY1^- and THY1^+ FLSs, western blot,
   identification of additional citrullination sites of IGF2BP1, point
   mutation of IGF2BP1 and ELAVL1, lentiviral infection and cell
   transfection, EdU assay, wound healing assay, Transwell assay, RNA
   immunoprecipitation (RIP)-qPCR, CIA model establishment, histological
   analysis, cytokines ELISA, structural docking, dot blot assay, custom
   ELISA are provided in the Supplementary Materials.

Patients and samples

   The synovium samples were collected from normal subjects undergoing
   high amputation or undergoing joint arthroplasty with femoral-neck
   fracture, as well as from patients with RA who underwent synovectomy,
   or joint replacement surgery, at the First Affiliated Hospital of
   Wenzhou Medical University. Peripheral blood was obtained from healthy
   participants undergoing routine check-ups and from RA patients. All RA
   patients met the 2010 American College of Rheumatology (ACR) criteria.
   Serum and synovial fluid were obtained from patients with RA who were
   newly diagnosed in the outpatient department before drug treatment. RA
   synovium samples were obtained from RA patients who had discontinued
   drug treatment for at least 15 days. The study was approved by the
   research ethics committee and fulfilled the ethical guidelines of the
   First Affiliated Hospital of Wenzhou Medical University and all
   participants submitted consent forms. All ethical regulations relevant
   to human research participants were followed. Clinical data for all
   participants listed in the Supplementary Table [169]2. FLSs were
   isolated from the synovium as previously described^[170]55, and the
   third to fifth passage of FLSs were used for subsequent experiments.

Ethics approval

   This study involves human participants. The human study was adhered to
   the Declaration of Helsinki and approved by the Clinical Research
   Ethics Committee of the First Affiliated Hospital of Wenzhou Medical
   University (No.2016157). All animal experiments were approved by the
   Institutional Animal Care and Use Committee of Wenzhou Medical
   University (ID Number: wydw2024-0215).

The screening of citrullinated proteins by mass spectrometry

   For the screening of citrullinated proteins, immunoprecipitated samples
   by pan-anti-citrulline antibody F95 from RA-FLSs were reduced by 10 mM
   DTT for 50 min at 57 °C, digested by trypsin at 37 °C overnight and
   then identified by LC-MS/MS. The criteria used to distinguish
   citrullinated proteins were as follows: the fragmentation of
   citrullinated peptides were accompanied by a mass increase of
   0.9840 Da, meanwhile, the neutral losses of isocyanic acid (HNCO,
   43 Da) and ammonia (NH3, 17 Da) in peptide collision-induced
   dissociation.

Immunoprecipitation (IP)

   To extract citrullinated proteins or other target proteins in RA-FLSs,
   300 µg cell lysates were firstly immunoprecipitated using
   anti-pan-citrulline antibody (F95), anti-IGF2BP1, anti-FLAG, anti-HA or
   isotype antibody overnight, respectively. On the second day, the
   proteins were incubated with protein L magnetic beads (HY-K0205, MCE)
   or protein A/G magnetic beads (HY-K0202, MCE) for 1 h. After that, the
   beads were separated by a magnetic field, washed with PBS, boiled in
   SDS loading buffer for 5 min. The proteins in the supernatant were
   further analyzed by Mass Spectrometry and western blot. All details of
   antibodies used for IP were listed in the Supplementary Table [171]3.

RNA isolation, quantitative PCR (qPCR), and RNA stability assay

   Total RNA from FLSs were extracted using Trizol reagent (T9108, Takara)
   and reversed transcribed using Primescript^TM RT reagent (RR037,
   Takara). Gene expression of SEMA3D was detected by qPCR using SYBR qPCR
   Master Mix (Q511, Vazyme). The relative mRNA expression was normalized
   to GAPDH, and the relative fold change was calculated using the 2^−ΔΔCT
   method as previously described^[172]56. The primer sequences used were
   listed in Supplementary Table [173]4.

   For RNA stability assay, RA-FLSs were treated with 5 μg/ml actinomycin
   D and harvested for RNA extraction at 0, 1, 2, 4, and 8 h. The
   expression of SEMA3D mRNA was detected by qPCR and half-life of SEMA3D
   mRNA was estimated as previously decried^[174]51.

Transcriptome sequencing and bioinformatics analysis

   To investigate the transcriptome changes induced by cit-IGF2BP1 in
   RA-FLSs, we extracted total RNA from IGF2BP1-KD RA-FLSs that rescued
   with either IGF2BP1^WT plasmid (high citrullinated IGF2BP1) or
   IGF2BP1^R167K plasmid (low citrullinated IGF2BP1). The extracted RNA
   were then subjected to Illumina HiSeq deep sequencing (Illumina HiSeq
   6000) for analysis of the transcriptome changes. The DEGs were
   identified as screened (fold change > 2 and p < 0.05) (Supplementary
   Data [175]1). Then, bioinformatics analyses of the DEGs were applied
   including Gene Onotology (GO) ([176]http://david.abcc.ncifcrf.gov/) and
   Kyoto Encyclopedia of Genes and Genomes (KEGG)
   ([177]http://www.kegg.jp). The mRNAs interacting with IGF2BP1 were
   screened by [178]ENCORI, RM2Target
   ([179]http://rm2target.canceromics.org/), and RNA [180]INTER database,
   three databases that supported by large-scale CLIP-Seq for exploring
   protein-RNA interactions^[181]57–[182]59.

Statistical analysis

   Analyses of all data were carried out using Graphpad Prism 9.0. First,
   the Shapiro-Wilk test was used to analyze the normality of data and the
   Levene test was used to analyze the homogeneity of variance of data.
   Then, Student’s t-test was used to analyze two sets of data which met
   the normal distribution and homogeneity of variance. One-way analysis
   of variance (ANOVA) test was used to analyze multigroup comparisons of
   the means. ANOVA of repeated measurements were used to analyze
   statistical significance of clinical scores. Kruskal-Wallis and
   Mann-Whitney non-parametric tests were used to analyze data that did
   not meet the normal distribution or the homogeneity of variance.

Reporting summary

   Further information on research design is available in the [183]Nature
   Portfolio Reporting Summary linked to this article.

Supplementary information

   [184]Transparent Peer Review file^ (540.3KB, pdf)
   [185]Supplementary Information^ (6.4MB, pdf)
   [186]42003_2025_7492_MOESM3_ESM.pdf^ (59.8KB, pdf)

   Description of Additional Supplementary File
   [187]Supplementary Data 1^ (33.4KB, xlsx)
   [188]Supplementary Data 2^ (34.7KB, xlsx)
   [189]Reporting Summary^ (5.1MB, pdf)

Acknowledgements