H3K4me3‐Mediated FOXJ2/SLAMF8 Axis Aggravates Thrombosis and Inflammation in β2GPI/Anti‐β2GPI‐Treated Monocytes

Abstract Antiphospholipid syndrome (APS) is characterized by thrombus formation, poor pregnancy outcomes, and a proinflammatory response. H3K4me3‐related monocytes activation are key regulators of APS pathogenesis. Therefore, H3K4me3 CUT&Tag and ATAC‐seq are performed to examine the epigenetic profiles. The results indicate that the H3K4me3 signal and chromatin accessibility at the FOXJ2 promoter are enhanced in an in vitro monocyte model by stimulation with β2GPI/anti‐β2GPI, which mimics APS, and decreases after OICR‐9429 administration. Furthermore, FOXJ2 is highly expressed in patients with primary APS (PAPS) and is the highest in patients with triple‐positive antiphospholipid antibodies (aPLs). Mechanistically, FOXJ2 directly binds to the SLAMF8 promoter and activates SLAMF8 transcription. SLAMF8 further interacts with TREM1 to stimulate TLR4/NF‐κB signaling and prohibit autophagy. Knockdown of FOXJ2, SLAMF8, or TREM1 blocks TLR4/NF‐κB and provokes autophagy, subsequently inhibiting the release of inflammatory and thrombotic indicators. A mouse model of vascular APS is established via β2GPI intraperitoneal injection, and the results suggest that OICR‐9429 administration attenuates the inflammatory response and thrombus formation by inactivating FOXJ2/SLAMF8/TREM1 signaling. These findings highlight the overexpression of H3K4me3‐mediated FOXJ2 in APS, which consequently accelerates APS pathogenesis by triggering inflammation and thrombosis via boosting the SLAMF8/TREM1 axis. Therefore, OICR‐9429 is a promising candidate drug for APS therapy.


Introduction
Antiphospholipid syndrome (APS) is a heterogeneous disease characterized by autoimmune disorders, chronic inflammation, venous or arterial thrombosis, adverse pregnancy outcomes (APOs), and the presence of antiphospholipid antibodies (aPLs). [1]aPLs are autoantibodies with a heterogeneous profile and include anti-beta2-glycoprotein I (anti-2GPI), anticardiolipin antibody (aCL), and lupus anticoagulant (LAC). [2]These autoantibodies activate endothelial cells (ECs), monocytes, and platelets, contributing to the activation of hypercoagulable and hyperinflammatory states in APS. [3]Current therapeutic options for APS are limited, and long-term anticoagulation therapy is commonly used to prevent recurrent thrombosis.However, this is unfavorable for patients who have not yet suffered from thrombosis but have aPLs. [4]Thus, the molecular mechanisms of APS development are worth exploring and would provide great opportunities for advanced treatment of APS.
Several studies have implicated epigenetic mechanisms in autoimmune disease pathogenesis.Epigenetic changes mainly involve histone modifications, non-coding RNAs (ncRNAs), and DNA methylation, which regulate chromatin remodeling and downstream genes expression. [5]Trimethylation of histone H3K4 (H3K4me3), a major histone modification mark that is strongly associated with active transcription, is mostly found around transcription start sites (TSSs). [6,7]In monocytes from patients with systemic sclerosis (SS), H3K4me3-mediated gene alterations are enriched in immune, interferon (IFN), and antiviral pathways by providing binding sites for interferon regulatory factors (IRF) and signal transducer and activator of transcription (STAT) transcription factors in their promoters. [8]In addition, H3K4me3 has been linked to the activation of the specificity protein 1 (SP1), activating transcription factor 3 (ATF3), and high mobility group A1 (HMGA1) promoters in T cells, B cells, and monocytes of patients with systemic lupus erythematosus (SLE) and is involved in SLE pathogenesis and progression. [9]A previous study found elevated H3K4me3 signals at the matrix metalloproteinase (MMP)−1, −3, −9, and −13 promoters in synovial fibroblasts of patients with rheumatoid arthritis (RA), and upregulated MMPs are potential biomarkers of disease activity. [10]OICR-9429, a high-affinity small-molecule compound, potently suppresses the trimethylation of histone H3K4. [11,12]Nevertheless, the role of OICR-9429 in inhibiting H3K4me3 and whether it can be used for the treatment of APS remain to be further investigated.
The 2GPI/anti-2GPI complex binds and activates monocytes. [13]Activated monocytes cause APS development by inducing proinflammatory and prothrombotic responses, [14] which might be attributed to epigenetic abnormalities of monocytes. [13,15]For example, DNA methylation at the interleukin (IL)−8 promoter and first intron of tissue factor 3 (TF3) increases at 4 h in a 2GPI/anti-2GPI-stimulated monocyte model that mimics APS, followed by a decrease at 6 h and a return to basal levels at 24 h post-stimulus.Changes in the dynamic methylation of IL-8 and TF3 cause transcriptional activation. [13]Significantly, these epigenetically altered genes are responsible for aggravating coagulation formation, proinflammatory reactions, and poor pregnancy outcomes in patients with APS.
Considering the vital role that the epigenetic distortion of monocytes plays in APS, H3K4me3 CUT&Tag and ATAC-seq were designed to examine H3K4me3 signaling and chromatin accessibility at the whole-genome level in an in vitro monocyte model mimicking APS.Integrative analysis identified augmented H3K4me3 signaling and chromatin accessibility at the forkhead box J2 (FOXJ2) promoter.Moreover, FOXJ2 was highly expressed in patients with primary APS (PAPS) and in an in vivo mouse model mimicking APS.The level of FOXJ2 was highest in patients with PAPS with triple-positive aPLs, indicating the pivotal function of FOXJ2 in APS pathogenesis.Therefore, a series of experiments was conducted to investigate the corresponding molecular mechanisms and clarify whether OICR-9429 can alleviate APS pathogenesis, which could provide a novel therapeutic target for patients with APS.

Epigenetic Profiles of the In Vitro Monocyte Model Mimic APS
H3K4me3 serves as a canonical epigenetic mark and represents transcription initiation and elongation. [16,17]OICR-9429, an inhibitor targeting the MLL1-WDR5 interaction, can repress the H3K4me3 mark and related gene transcription. [11,12]In order to explore H3K4me3-mediated APS pathogenesis, we first built an in vitro APS model via stimulating monocytes or THP-1 cells with the 2GPI/anti-2GPI IC.Samples were divided into NC, IC, and OICR-9429+IC groups for both monocytes and THP-1 cells (Figure 1A).H3K4me3 CUT&Tag-seq was used to detect H3K4me3 signatures at the whole-genome level in an in vitro monocyte model that mimicked APS.In line with the described canonical H3K4me3 pattern, the H3K4me3 signal was mainly enriched in the TSS region (Figure 1B).Moreover, the specific H3K4me3 signal at the FOXJ2 promoter was increased in IC-stimulated monocytes mimicking APS but was remarkably reduced upon treatment with OICR-9429 (Figure 1C).CUT&Tag-qPCR further indicated increased enrichment of H3K4me3 at the FOXJ2 promoter in the in vitro APS model (Figure 1D).
Chromatin accessibility plays a crucial role in determining gene expression profiles by opening or closing accessibility.Open accessibility predisposes to autoimmune diseases by upregulating the expression of inflammatory genes. [18]Therefore, this study utilized ATAC-seq to test chromatin accessibility at the whole-genome level and found that highly open accessibility was mainly located at the TSS site (Figure 1E).Moreover, chromatin accessibility at the FOXJ2 promoter was specifically augmented in the in vitro APS model but was attenuated by OICR-9429 treatment (Figure 1F).
FOXJ2, as a member of forkhead/HNF3 family, has the capability to maintain LPS-induced production of inflammatory cytokines, such as IL-6 and tumor necrosis factor (TNF)-. [19]herefore, FOXJ2 was the focus of subsequent experiments.FOXJ2 was overexpressed in the IC-stimulated APS model, whereas its expression was dramatically reduced upon exposure to OICR-9429 (Figure 1G,H).Therefore, IC administration in the in vitro APS model specifically strengthened the H3K4me3 signal and chromatin accessibility at the FOXJ2 promoter, which facilitated FOXJ2 expression in APS.

Elevated FOXJ2 Expression in Patients with PAPS
Since FOXJ2 was overexpressed in the in vitro APS model, we explored the clinical relationship between FOXJ2 and patients with PAPS.Primary monocytes from 64 patients with PAPS and CUT&Tag-Seq at the whole genome level in an in vitro APS model using THP-1 cells and monocytes, respectively.C) The relative enrichment levels of H3K4me3 at the promoter of FOXJ2 are visualized using IGV.D) H3K4me3 CUT&Tag-qPCR tested the relative enrichment levels of H3K4me3 at the FOXJ2 promoter in an in vitro APS model of THP-1 cells and monocytes, respectively.E) Heatmaps of chromatin accessibility at the whole-genome level from ATAC-seq analysis of an in vitro APS model of THP-1 cells and monocytes, respectively.F) The chromatin accessibility at the FOXJ2 promoter is displayed using IGV.G) The level of FOXJ2 in vitro APS model is detected using western blotting.H) The expression of FOXJ2 in the in vitro APS model is determined using RT-qPCR.Data are expressed as the mean ± SEM of three independent experiments.IGV, Integrative Genomics Viewer; *P < 0.05; **P < 0.01; ***P < 0.001.32 HDs were collected to examine the expression of FOXJ2, and their clinical baseline data were obtained (Data S1, Supporting Information).FOXJ2 expression was higher in patients with PAPS than in HDs (Figure 2A).The mRNA level of FOXJ2 was higher in aPL-positive patients with PAPS than in aPL-negative patients (Figure 2B-D) and was the highest among patients with PAPS with triple-positive aPLs (Figure 2E).Notably, PLT and complement C3 and C4 levels were associated with APS progression.Among the patients with PAPS who had laboratory data for C3, C4, and PLT, the levels of C3, C4, and PLT showed a decreasing tendency in patients with PAPS with triple-positive aPLs compared with those of patients with single-or double-positive aPLs (Figure 2F-H).However, the difference was only significant between aCL-positive and aCL-negative patients with PAPS (Table 1).The positivity rate of other aPLs did not correlate with a history of thrombosis or adverse pregnancy delivery (Table 1).The above findings indicated that the expression of FOXJ2 was increased in patients with PAPS and the level of FOXJ2 positively correlated with the positive rate of aPLs (Figure 2I).Moreover, aPLs may have affected the levels of C3, C4, and PLT.

FOXJ2 Accelerates Inflammation and Thrombosis by Transcriptionally Activating SLAMF8 In Vitro
After confirming the elevated FOXJ2 expression in patients with PAPS, we next examined the effect of FOXJ2 on the inflammatory response and thrombotic formation in APS.ELISA quantification was used to determine the levels of related inflammatory and thrombotic cytokines.In particular, the levels of IL-6, IL-8, TNF-, and tissue factor (TF) were increased in the IC group, while the use of OICR-9429 limited their production (Figure 3A).Further, knockdown of FOXJ2 in THP-1 cells dramatically restrained IC-induced IL-6, IL-8, TNF-, and TF secretion, whereas FOXJ2 overexpression promoted their secretion (Figure 3B).Therefore, overexpressed FOXJ2 facilitates APS progression by stimulating IL-6, IL-8, TNF-, and TF production, and OICR-9429 may mitigate APS pathogenesis by downregulating FOXJ2.
To further explore the underlying mechanisms of FOXJ2 in APS, RNA-seq was used to screen for FOXJ2-modulated genes.Significantly, the results showed 549 downregulated and 123 upregulated genes in FOXJ2-knockdown THP-1 cells (Figure 3C,D; Data S2, Supporting Information).Meanwhile, anti-FOXJ2 ChIP-Seq was conducted, and the results showed that 9.4% of the FOXJ2 peaks were located on the promoters (Figure 3E).Integrative analysis confirmed that 109 promoter peak-annotated genes were downregulated after FOXJ2 downregulation, including guanosine monophosphate reductase (GMPR), dynein axonemal heavy chain 8 (DNAH8), mast cell-expressed membrane protein1 (MCEMP1), and signaling lymphocyte activating molecule family member 8 (SLAMF8) (Figure 3F).The relationship between GMPR, DNAH8, MCEMP1, and autoimmune diseases has not yet been reported.SLAMF8 is an essential surface receptor that is highly expressed on monocytes and has been implicated in the progression of autoimmune diseases such as RA. [20]Moreover, FOXJ2 regulated SLAMF8 by directly binding to the SLAMF8 promoter (Figure 3G; Data S3, Supporting Information), with a 21.6-fold increase in the binding of FOXJ2 to the SLAMF8 promoter compared with the nonspecific binding of IgG to the SLAMF8 promoter (Figure 3H).RT-qPCR and western blotting revealed the downregulated SLAMF8 after FOXJ2 knockdown and the upregulated SLAMF8 after FOXJ2 overexpression (Figure 3I,J).In addition, the luciferase activity reporter assay further determined that the mutated promoter reporter resulted in decreased luciferase reporter activity for the SLAMF8 promoter (Figure 3K).Taken together, these results indicate that FOXJ2 directly induces SLAMF8 promoter transcription.
To explore whether the protective effect of SLAMF8 inhibition on APS pathogenesis is associated with reduced inflammation and thrombosis, we compared the levels of inflammatory and thrombotic indicators in SLAMF8-OE and SLAMF8knockdown THP-1 cells.Qualification results showed increased production of IL-6, IL-8, TNF-, and TF in SLAMF8-OE THP-1 cells and reduced production in SLAMF8-knockdown THP-1 cells (Figure 3L).We then cotransfected siSLAMF8 and FOXJ2-OE into THP-1 cells to conduct rescue experiments.The data showed that SLAMF8 knockdown attenuated the production of IL-6, IL-8, TNF-, and TF in FOXJ2-OE THP-1 cells after IC stimulation (Figure 3M).Thus, FOXJ2 plays an important role in ICinduced inflammation and thrombosis via SLAMF8 overexpression.

SLAMF8 Directly Interacts with TREM1 to Modulate Thrombosis and Inflammatory Responses Related to TLR4/NF-𝜿B
SLAMF8 acts as a novel regulator of macrophage-mediated inflammation by activating TLR4/NF-B signaling but does not depend on the direct regulation of TLR4. [20,21]Triggering receptor expressed on myeloid cells 1 (TREM1) is an active receptor expressed on human monocytes and macrophages that facilitates the LPS-induced inflammatory response by directly binding to TLR2 and TLR4. [22]Considering that TERM1 was significantly downregulated in FOXJ2-knockdown THP-1 cells (Figure 3C), we hypothesized that SLAMF8 participates in the TLR4/NF-B axis by stabilizing TREM1 expression through molecular interaction.To validate this hypothesis, we performed co-IP and double-labeled immunofluorescence assays.Double-labeled immunofluorescence showed that SLAMF8, TREM1, and TLR4 were mainly expressed on the cell membrane in THP-1 cells (Figure 4A).Nevertheless, co-IP data indicated that SLAMF8 interacted with TREM1 but not with TLR4 and that TREM1 coimmunoprecipitated with TLR4 (Figure 4B).

SLAMF8/TREM1 Complex Blocks Autophagy in the In Vitro APS Model
The 2GPI/anti-2GPI complex impairs autophagy, and deficient autophagy plays proinflammatory and prothrombotic roles in APS development. [23,24]Moreover, SLAMF8/TREM1 exerts crucial functions in restraining the autophagic response. [25,26]o investigate whether SLAMF8/TREM1 participates in APS pathogenesis by autophagy, cells were subjected to TEM to visualize the formation of autophagic vacuoles.As shown in Figure 5A, the number of autophagic vacuoles significantly decreased upon IC stimulation but increased after SLAMF8 knockdown, indicating the protective effect of SLAMF8 deletion on autophagy.We also examined the expression of related autophagic molecules.In the IC group, the expression levels of LC3II/I, ATG5, and Beclin-1 were reduced, whereas those of p62/SQSTM1, TREM1, and SLAMF8 were increased, indicating impaired autophagy (Figure 5B,C).Significantly, SLAMF8 deletion or OICR-9429 administration rescued their expression, indicating that autophagy was provoked (Figure 5B,C).Additionally, TREM1 knockdown rescued SLAMF8-injured autophagy after IC stimulation (Figure 5D).Taken together, SLAMF8 impairs autophagy by positively regulating TREM1 expression, further exacerbating the inflammatory response and APS pathogenesis.

Intervention of TREM1 Rescues FOXJ2/SLAMF8-Induced Inflammation and Thrombosis In Vitro
Based on the verification of the FOXJ2/SLAMF8/TREM1 pathway, we explored whether TREM1 deletion ameliorated SLAMF8or FOXJ2-mediated APS pathogenesis in vitro.Cells cotransfected with siTREM1 and SLAMF8-OE or FOXJ2-OE were established to test the concentrations of IL-6, IL-8, TNF-, and TF.Quantification ELISA results showed that TREM1 deletion reduced the secretion of IL-6, IL-8, TNF-, and TF in SLAMF8-OE and FOXJ2-OE THP-1 cells after IC administration, suggesting that TREM1 knockdown alleviated SLAMF8-or FOXJ2-induced proinflammatory and thrombotic cytokine secretion (Figure 6A,B).Taken together, these results suggest that FOXJ2/SLAMF8 facilitates APS deterioration by increasing the protein level of TREM1.

FOXJ2 Promotes APS Pathogenesis by Stimulating the SLAMF8/TREM1 Pathway In Vivo
In order to further confirm the in vivo function of FOXJ2, two groups of mice were used to establish a model of APS using intraperitoneal injections of 100 ug 2GPI thrice; one of 2GPI groups received OICR-9429 for 7 consecutive days (Figure 7A).On day 21, the 2GPI group had the increased anti-2GPI levels, indicating that the mouse model with APS was successfully established (Figure 7B).The 2GPI group had the longest APTT (Figure 7C), while PLT was reduced in the 2GPI group but increased under OICR-9429 treatment (Figure 7D).On day 28, the 2GPI group had decreased ascending aortic blood velocity, which increased after OICR-9429 treatment (Figure 7E).The thrombus size was greatly increased in the 2GPI group, and the use of OICR-9429 remarkably decreased the thrombus formation (Figure 7F).Furthermore, the APS model had increased serum concentrations of IL-6, IL-8, TNF-, and TF that significantly decreased with OICR-9429 administration (Figure 7G).In addition, western blot analysis validated the enhanced FOXJ2/SLAMF8/TREM1/TLR4 signaling and impaired autophagy in bone marrow-derived monocytes from mice with APS, whereas the use of OICR-9429 produced the opposite effect (Figure 7H).Therefore, FOXJ2-mediated inflammation and thrombus formation are activated but autophagy is suppressed in a mouse model of APS, and OICR-9429 may relieve Figure 3. FOXJ2 accelerated inflammation and thrombosis by transcriptionally activating SLAMF8 in vitro.A) ELISA is utilized to determine the concentrations of IL-6, IL-8, TNF-, and TF in an IC-induced APS model and their levels with OICR-9429 administration.B) The effect of FOXJ2 knockdown or overexpression on the secretion of IL-6, IL-8, TNF-, and TF is examined using ELISA.C) A heatmap shows the normalized expression of several deregulated genes between shNC and shFOXJ2-THP-1 cells.D) Volcano plot of differentially expressed genes between shNC and shFOXJ2-THP-1 cells.E) The distribution of FOXJ2 ChIP-Seq peaks.F) The integrative analysis between FOXJ2 ChIP-Seq and RNA-Seq.G) IGV representation of anti-FOXJ2 ChIP-Seq at the SLAMF8 promoter in THP-1 cells.H) ChIP-qPCR indicates the relative enrichment level of FOXJ2 within the SLAMF8 promoter in THP-1 cells, with control IgG as a negative control.I) mRNA levels of FOXJ2 and SLAMF8 in FOXJ2-knockdown and FOXJ2-OE THP-1 cells are detected with RT-qPCR.J) The protein levels of FOXJ2 and SLAMF8 in FOXJ2-knockdown and FOXJ2-OE THP-1 cells, with -actin as a loading control.K) Schematic representation of the FOXJ2 target sequence within the SLAMF8 promoter; the luciferase activities of WT-SLAMF8 and Mut-SLAMF8 are measured after cotransfection with FOXJ2-OE or FOXJ2-NC.L) IL-6, IL-8, TNF-, and TF levels in siSLAMF8 and SLAMF8-OE THP-1 cells are detected using an ELISA.M) IL-6, IL-8, TNF-, and TF concentrations in siSLAMF8 and FOXJ2-OE cotransfected THP-1 cells are detected using an ELISA.Data are expressed as the mean ± SEM of three independent experiments.TSS, transcription start site.*P < 0.05; **P < 0.01; ***P < 0.001.APS pathogenesis by inactivating FOXJ2/SLAMF8/TREM1 signaling (Figure 7I).

Discussion
APS pathogenesis is partly orchestrated by epigenetic aberrance. [6,27]Monocytes play a key role in aPLs-mediated APS pathogenesis by upregulating TF, vascular cell adhesion molecule-1 (VCAM-1), IL-6, IL-8, and TNF- expression. [13,28]ccumulating evidence suggests the therapeutic potential of epigenetics targeting monocytes.Therefore, this study sought evidence of epigenetic alterations in monocytes in APS. [29]Considering the association of extended and enhanced H3K4me3 peaks with genes modulating immune responses, [30] H3K4me3 CUT&Tag and ATAC-seq were used to assess the profiles of H3K4me3 peaks and chromatin accessibility at the genome-wide level, respectively.Specifically, the enhanced H3K4me3 peak and chromatin accessibility at the FOXJ2 promoter were closely aligned with increased FOXJ2 transcription in patients with PAPS, as well as in an in vivo mouse model and an in vitro monocyte model mimicking APS.However, the mechanisms of FOXJ2 involvement in APS remain unclear.
FOXJ2 plays a crucial role in regulating the expression of immune response-related genes through recognizing type A and B DNA sequences. [31,32]Type A sequences possess a core element that also exists in other forkhead factors, whereas type B sequences do not hold this core element.The frequency of type B sequences was lower than that of type A sequences in the site selection experiments; however, type A and B sequences have comparable affinities for fork head homologous X (FHX) sites. [33][36] In this study, we identified the upregulation of FOXJ2 in APS for the first time, demonstrating the important role of FOXJ2 in monocytes for regulating APS pathogenesis.Considering the essential roles of activated ECs and platelets in APS, FOXJ2 might also epigenetically regulate the function of ECs and platelets during the progression of APS, which is worth investigating.
Furthermore, we identified the binding sites of FOXJ2 in the promoter region of SLAMF8 and verified that SLAMF8 is a transcriptional target of FOXJ2 using ChIP-seq and luciferase activity reporter assays.As a cell surface receptor, SLAMF8 is highly expressed upon exposure to proinflammatory stimuli. [21]SLAMF8 upregulation has also been observed in autoimmune inflammation, such as RA. [20,37]Previous reports have demonstrated the important role of SLAMF8 in altering LPS-mediated inflammation and autophagy by regulating mitogen-activated protein kinase (MAPK) and TLR4 signaling. [20,21]On the other side, some studies have indicated that SLAMF8 attenuates the inflammation process in bacterial infected-macrophages by inhibiting Nox2 activity through the PKC and PI3K pathways, these results might be involved in macrophage's microbicidal mechanisms based on bacterial infection-induced models. [38,39]Thereby, different stimuli for constructing different disease models will trigger distinct pathological responses, which may lead to different or even conflicting results.Of importance, our studies confirmed that FOXJ2 participates in IC-induced APS pathogenesis by positively modulating the expression of SLAMF8.Furthermore, the expression of SLAMF8 was increased in APS and SLAMF8 knockdown was demonstrated to ameliorate FOXJ2-mediated inflammation and thrombosis in APS.
TREM and TREM-like receptors are associated with inflammatory responses in both innate and adaptive immunity. [40]Among these, TREM1 partakes in TLR4-mediated inflammation and autophagy. [26,41]A study reported that TREM1 directly binds with TLR4 to boost TLR4/NF-B signaling, [22] which plays an essential role in the pathogenesis of various autoimmune diseases, including RA, SLE, IBD, and type 1 diabetes (T1D). [42]Most importantly, elevated TREM1 expression has been verified in patients with thrombotic PAPS and shows great potential for predicting thrombotic events and inflammatory activity. [43]In addition, TLR4 exacerbates the pathogenesis of APS, and TLR4 deficiency prevents inflammation, thrombus formation, and TF activity in a murine model of APS. [44]Taken together, both SLAMF8 and TREM1 are correlated with the amplification of TLR4 signaling.Therefore, we assessed the interaction of SLAMF8 and TREM1 with TLR4 and showed that SLAMF8 binds to TREM1 but not to TLR4, and TREM1 binds to TLR4.Moreover, SLAMF8 deletion inactivates TREM1/TLR4/NF-B signaling, and TREM1 knockdown abates the production of FOXJ2/SLAMF8-induced inflammatory and thrombotic indicators.Of note, knockdown of SLAMF8 did not influence the TREM1 mRNA level (Figure S2A, Supporting Information), but repressed the protein expression of TREM1 (Figure S2B, Supporting Information) and attenuated the effective interacting of SLAMF8 with TREM1 (Figure S2C, Supporting Information).Thereby, we proposed that SLAMF8 might maintain the high protein level of TREM1 via interacting with TREM1, the relative molecular mechanisms are worthy of exploring in the next study.
Monocyte-macrophage autophagy is linked to inflammation and thrombosis, and its dysfunction is a feature of autoimmune activation. [45]In a previous study, impaired autophagy was detected in in a murine APS model. [23]In the in vivo and in vitro APS models of our study, we tested downregulated LC3II/I, Beclin-1, and ATG5 and upregulated p62/SQTM1, indicating a deficiency of autophagy.SLAMF8 or TREM1 knockdown, or OICR-9429 administration induces autophagy.TREM1 accelerates thrombin generation by inducing TF secretion and is an interesting target for investigating novel inhibitors of thrombotic activity. [46]In addition, TREM1 intervention partially compensates for FOXJ2-and SLAMF8-induced TF secretion.These results provide novel evidence that FOXJ2 activation contributes to the pathogenesis of APS by repressing SLAMF8/TREM1-Figure 5. SLAMF8/TREM1 complex blocks autophagy in an in vitro APS model.A) Representative TEM images of autophagic vacuoles (red arrows) in THP-1 cells.B) The expression of SLAMF8, TREM1, LC3II/I, ATG5, Beclin-1, and p62/SQSTM1 is detected in siSLAMF8-transfected THP-1 cells after IC stimuli.C) Quantification of SLAMF8, TREM1, LC3II/I, ATG5, Beclin-1, and p62/SQSTM1 in IC-treated and OICR-9429 plus IC-treated cells.D) The expression of TREM1, LC3II/I, ATG5, Beclin-1, and p62/SQSTM1 in THP-1 cells cotransfected with siTREM1 and SLAMF8-OE, with -actin as a loading control.Data are expressed as the mean±SEM of three independent experiments.TEM, transmission electron microscopy; *P < 0.05; **P < 0.01; ***P < 0.001.Recurrent arterial or venous thrombosis is the main symptom of APS.Treatment of APS mainly depends on long-term anticoagulation therapy to prevent thrombosis.We found that the thrombi of the carotid artery are larger and the blood velocity of the ascending aorta is slower in the in vivo APS model.Significantly, OICR-9429 administration shrinks the thrombus plaque area, increases the blood velocity, and reduces anti-2GPI levels.Meanwhile, FOXJ2-mediated inflammation is inhibited and autophagy is markedly induced.These results suggested that OICR-9429 has the potential to prevent inflammation and thrombosis in APS by inactivating FOXJ2/SLAMF8/TREM1 signaling.In order to more accurately explore the role of FOXJ2 in APS, FOXJ2 need to be specifically deleted or overexpressed in monocytes to construct FOXJ2 genetic mouse model, which might be more specific for investigating the function of FOXJ2 in monocytes for regulating APS pathogenesis.
Clinically, aPLs contribute to pregnancy loss, thrombosis, and thrombocytopenia in APS. [6]Therefore, we analyzed the correlation between aPLs and FOXJ2 expression and confirmed that FOXJ2 is highly expressed in patients with PAPS and is positively correlated with the positive rate of aPLs.Complement activation is associated with autoimmune disease activity [47,48] and emerging evidence has shown low levels of C3 and C4 in patients with PAPS with thrombosis and poor pregnancy outcomes. [49,50]onsistent with these findings, we found that aPL-positive patients with PAPS have lower C3, C4, and PLT levels.Therefore, aPLs may influence APS pathogenesis by affecting the amounts of PLT, complement C3 and C4.Nevertheless, aPL positivity is not significantly related to a history of thrombosis or adverse pregnancy deliveries, and more samples should be collected to explore the relationship between aPL positivity, a history of thrombosis, and APOs.
Taken together, this study is the first to demonstrate that FOXJ2 is highly expressed in the pathogenesis of APS and is related to the presence of aPLs.FOXJ2 directly augments SLAMF8/TREM1 signaling, consequently accelerating APS pathogenesis by repressing autophagy and promoting the release of proinflammatory and prothrombotic cytokines.Furthermore, OICR-9429 boosts autophagy and blocks inflammation and thrombus formation by downregulating FOXJ2/SLAMF8/TREM1.These findings provide a novel rationale for targeting FOXJ2 using OICR-9429 and indicate the potential of epigenetic medicine to improve the treatment of APS.

Experimental Section
Animal Models of Vascular APS: Female BALB/c mice (8-10 weeks old) were purchased from Beijing Vital River Laboratory Animal Technol-ogy.The experimental mice were bred in a specific pathogen-free (SPF) environment.Mice were randomly divided into three groups (four per group): a) the negative control (NC) group received intraperitoneal injections of bovine serum albumin (BSA) in Freund's adjuvant (F5881, Sigma-Aldrich); b) the 2GPI group received intraperitoneal injections of 100 ug 2GPI protein (11221-H08H, Sino Biological Inc.) in Freund's adjuvant on days 1, 8, and 14; [51] and c) the OICR-9429-treated APS (OICR-9429+2GPI) group received 5 mg kg −1 of OICR-9429 daily, starting on day 15 and lasting for 7 consecutive days.On day 22, blood samples were collected from the inner canthus to determine the anti-2GPI levels, platelet counts (PLT), and activated partial thromboplastin time (APTT).On day 28, Doppler imaging was used to detect the blood velocity in the ascending aorta.Whatman filter paper (3 mm × 1 mm) was soaked with 10% FeCl 3 and placed under the carotid artery for 5 min.Then, the arteries were removed, and the thrombosis size was measured following hematoxylin-eosin (HE) staining.In addition, monocytes from thigh bone marrow of mice were extracted using the EasySep Mouse Monocyte Isolation Kit (#19 861, STEMCELL Technologies), and the levels of functional molecules in bone marrow-derived monocytes were directly examined.
Participants: Sixty-four adult patients diagnosed with PAPS according to the Sydney classification criteria between November 2022 and July 2023 were enrolled in this retrospective study. [52]Thirty-two healthy donors (HDs) were selected as the control group and were age-and sex-matched with patients with PAPS.Peripheral blood monocytes were collected from blood samples of patients with PAPS and HDs.Demographic, clinical, and laboratory features were recorded, including age, sex, history of venous or arterial thrombosis and APOs, complement C3 and C4 levels, PLT, normalized silica clotting time (SCT), dilute Russell viper venom time (dRVVT), and titers of aCL and anti-2GPI.Antiphospholipid antibodies titers were defined as positive for aCL and anti-2GPI titers ≥20 units and for normalized SCT >1.16 and dRVVT >1.11.
Monocytes Isolation: Peripheral blood mononuclear cells were isolated from three HDs using 1.077 g mL −1 Lymphoprep density gradient medium (StemCell Technologies, Canada).Monocytes were isolated from peripheral blood mononuclear cells using APC anti-human CD14 antibody (301 807, Biolegend, USA) and the EasySep Human Monocyte Isolation Kit (#19 359, Stemcell Technologies).The monocyte purity was routinely assessed by flow cytometry.The cell suspension was seeded onto 48-well plates and allowed to adhere overnight.On the second day, non-adherent cells were removed and starved for 16 h.Monocytes were pretreated with OICR-9429 for 24 h and exposed to the IC for 4 h for RNA and DNA analyses and for 6 h for protein analysis.
Statistical Analysis: Data are presented as mean ± standard error of the mean (SEM) of three independent experiments and were statistically analyzed by a two-tailed unpaired t-test or Mann-Whitney U test.The 2test was utilized to compare categorical variables.Differences between groups were considered significant at *P < 0.05, **P < 0.01, and ***P < 0.001 and were nonsignificant (ns) at P > 0.05.All statistical data were analyzed using GraphPad Prism 7.0.
Ethics Approval and Consent to Participate: All experiments involving animals were approved by the Ethics Committee of Peking University Third Hospital (Approval Form: 060-02).
The studies involving human participants were approved by Ethics Committee of Peking University Third Hospital (Approval Form: 053-01).Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

Figure 1 .
Figure 1.Epigenetic profiles in an in vitro monocyte model mimicking APS.A) Workflow of IC treatment and OICR-9424 exposure in ex vivo monocytes or THP-1 cells model partially mimic APS.B) Heatmaps of H3K4me3CUT&Tag-Seq at the whole genome level in an in vitro APS model using THP-1 cells and monocytes, respectively.C) The relative enrichment levels of H3K4me3 at the promoter of FOXJ2 are visualized using IGV.D) H3K4me3 CUT&Tag-qPCR tested the relative enrichment levels of H3K4me3 at the FOXJ2 promoter in an in vitro APS model of THP-1 cells and monocytes, respectively.E) Heatmaps of chromatin accessibility at the whole-genome level from ATAC-seq analysis of an in vitro APS model of THP-1 cells and monocytes, respectively.F) The chromatin accessibility at the FOXJ2 promoter is displayed using IGV.G) The level of FOXJ2 in vitro APS model is detected using western blotting.H) The expression of FOXJ2 in the in vitro APS model is determined using RT-qPCR.Data are expressed as the mean ± SEM of three independent experiments.IGV, Integrative Genomics Viewer; *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 2 .
Figure 2. Elevated FOXJ2 expression in patients with PAPS.A) Relative FOXJ2 mRNA level in patients with PAPS and HDs.B) Relative FOXJ2 mRNA level in LAC-positive and -negative patients with PAPS.C) Relative FOXJ2 mRNA level in aCL-positive and -negative patients with PAPS.D) Relative FOXJ2 mRNA level in anti-2GPI-positive and -negative patients with PAPS.E) The expression level of FOXJ2 in patients with PAPS and triple-, double-, and single-positive aPLs.F) C3 levels in single-, double-, and triple-positive patients with PAPS.G) C4 levels in single-, double-, and triple-positive patients with PAPS.H) PLT levels in single-, double-, and triple-positive patients with PAPS.I) The correlation of FOXJ2 with the aPL positive rate and the levels of C3, C4, and PLT.PAPS, primary APS; HDs, healthy donors; *P < 0.05; **P < 0.01; ***P < 0.001.
mediated autophagy and inducing related inflammation and thrombosis.

Table 1 .
The association of clinical findings with aPLs in patients with PAPS.