GM-CSF receptor/SYK/JNK/FOXO1/CD11c signaling promotes atherosclerosis

Summary Atherosclerosis complicates chronic inflammatory diseases, such as rheumatoid arthritis and systemic lupus erythematosus, suggesting that a shared physiological pathway regulates inflammatory responses in these diseases wherein spleen tyrosine kinase (SYK) is involved. We aimed to identify a shared therapeutic target for atherosclerosis and inflammatory diseases. We used Syk-knockout atherosclerosis-prone mice to determine whether SYK is involved in atherosclerosis via the inflammatory response and elucidate the mechanism of SYK signaling. The Syk-knockout mice showed reduced atherosclerosis in vivo, and macrophages derived from this strain showed ameliorated cell migration in vitro. CD11c expression decreased on Syk-knockout monocytes and macrophages; it was upregulated by forkhead box protein O1 (FOXO1) after stimulation with granulocyte-macrophage colony-stimulating factor (GM-CSF), and c-Jun amino-terminal kinase (JNK) mediated SYK signaling to FOXO1. Furthermore, FOXO1 inhibitor treatment mitigated atherosclerosis in mice. Thus, GM-CSF receptor/SYK/JNK/FOXO1/CD11c signaling in monocytes and macrophages and FOXO1 could be therapeutic targets for atherosclerosis and inflammatory diseases.


INTRODUCTION
Chronic inflammation promotes the development of atherosclerosis and cardiovascular disease. 1 Although hypercholesterolemia-in particular, high levels of low-density lipoprotein cholesterol (LDL-C)-is a significant risk factor for atherosclerosis, 2 lipid-lowering therapy, such as statin administration, does not completely avert the risks of cardiovascular disease. 3 There is a residual risk owing to factors, such as triglyceride, apolipoprotein B-100, and inflammation levels. 1,4,5 Previous studies have reported that inflammation is also involved in the development of atherosclerosis; for instance, the level of high-sensitivity C-reactive protein (CRP), a well-known serum inflammation marker in clinical use, can predict cardiovascular events, including myocardial infarction and stroke, and the need for a coronary revascularization procedure. 6,7 Inflammation plays a pivotal role in atherosclerotic development. LDL-C-induced hyperlipidemia results in the accumulation of LDL-C in vessel walls. Subsequently, LDL-C is oxidized and engulfed by macrophages, which secrete inflammatory cytokines, resulting in cellular infiltration from the bone marrow. 8 The process of cellular infiltration into a lesion is common in chronic systemic inflammatory diseases, such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), which are considered high-risk factors for cardiovascular events. 9,10 Furthermore, these disease activity scores correlate with an exacerbation of atherosclerotic development. 11,12 Herein, we focused on spleen tyrosine kinase (SYK), which is a non-receptor tyrosine kinase expressed in hematopoietic cells and regulates the downstream signaling of various cell surface receptors, 13 for the following reasons. A genome-wide association study revealed that patients with vascular dementia harbor a risk-related variant of the Syk gene. 14 In addition, SYK is highly expressed in human atherosclerotic plaques, 15 whereas an SYK inhibitor ameliorates RA and SLE in a murine model. 16,17 Although SYK carries out several functions, such as inflammatory cytokine production, phagocytosis of oxidized LDL, and cell differentiation, [18][19][20][21][22] its involvement in the shared pathophysiology between atherosclerosis and chronic systemic inflammatory diseases remains unclear.

SYK regulates cell migration via CD11c expression on macrophages
As we determined that Syk deletion reduced the macrophage area on the plaque, we hypothesized that Syk deletion attenuated cell migration. Therefore, we performed wound scratch and transwell migration assays. We used bone marrow-derived macrophages (BMDMs) and confirmed Syk deletion using PCR and western blotting ( Figures S2A and S2B). The wound scratch and transwell migration assays showed that Syk-deficient BMDMs exhibited decreased cell motility and migration capability owing to C-C motif chemokine ligand 2 (CCL2) (Figures 2A, 2B, and S2C).
To elucidate the cell migration-related genes downstream of Syk, we performed a comprehensive gene expression analysis of bone marrow monocytes using RNA sequencing (RNA-seq), which revealed the top 78 differentially expressed genes (DEGs), with a false discovery rate (FDR) < 0.05 (excluding immunoglobulin genes). Downregulated DEGs in Syk-deficient monocytes included the migration-related genes Itgam2 (CD11b isoform), Itgax (CD11c), and Itgad (CD11d)-also known as integrins (Figures 2C and 2D). Integrins are adhesion molecules that act against endothelial ligands, such as intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1, in cell migration. 23-25 However, Cd11b À/À bone marrow-transplanted mice were not protected from atherosclerosis, whereas Cd11cand Cd11d-knockout mice exhibited reduced atherosclerosis. 24,26,27 In addition, hypercholesterolemia induced CD11c surface expression on human peripheral monocytes. 28 Therefore, we focused on Cd11c in the context of the downstream signaling of SYK, and Cd11c expression was validated using qPCR ( Figure S2D). To determine the CD11c protein level, we measured CD11c expression in peripheral monocytes using flow cytometry. CD11c levels decreased significantly on Syk-deficient monocytes (p < 0.001), whereas those of CD11b did not decrease ( Figures 2E, S2E, and S2F). We then evaluated the adhesive capacity of CD11c against ICAM-1 using an adhesion assay. BMDMs were stimulated with granulocyte-macrophage colony-stimulating factor (GM-CSF) to increase CD11c expression (Figure S2G). The number of adhesive cells in the isotype-control Syk-deficient BMDMs decreased more than that in Syk-wild type BMDMs. Additionally, the number of adhesive cells in Syk-wild type BMDMs incubated with anti-CD11c antibody decreased more than that in Syk-wild type BMDMs incubated with the isotype control. CD11c and SYK increased the cell adhesive capacity of BMDMs ( Figure 2F).
Collectively, these results suggest that SYK promoted cell migration by regulating Cd11c gene and protein expression.

SYK translocates FOXO1 to the nucleus during CD11c expression
To determine the region of the Cd11c promoter associated with SYK, we performed a luciferase assay using various lengths of the promoter. SYK was associated with a region 917-1,164 bp upstream of the Cd11c transcription start site (TSS) ( Figure 3A). Next, we determined and sequenced the candidate transcription factors (TFs) that bound 917-1,164 bp upstream of the Cd11c TSS using in silico analysis (JASPAR, Table S1). 29 To identify candidate TFs, we analyzed CD11c expression on BMDMs by adding inhibitors against the TFs using flow cytometry. A forkhead box protein O1 (FOXO1) inhibitor (AS1842856) suppressed CD11c expression ( Figure 3B); the inhibitor hampers FOXO1 transcriptional activity. 30 iScience Article decreased CD11c expression on RAW264.7 cells, a murine macrophage cell line ( Figure S3A), and another FOXO1 inhibitor (AS1708727) suppressed BMDM CD11c expression ( Figure S3B).
FOXO1 shuttles between the cytosol and nucleus, and intranuclear FOXO1 activates gene expression. 32 Therefore, the subcellular localization of FOXO1 in GM-CSF-primed BMDMs was evaluated using western   Figure 3B), the intranuclear FOXO1 levels decreased in Syk-deficient BMDMs, whereas the FOXO1 level of total cell lysate remained the same in Syk-wild type and Syk-deficient BMDMs. These data indicate that SYK affects FOXO1 translocation to the nucleus ( Figure 3C). Western blotting revealed the nuclear translocation of FOXO1 in RAW264.7 cells without GM-CSF stimulation ( Figure S3C). Then, we evaluated whether FOXO1 binds to the predicted sequence in three ways. First, we mutated the predicted FOXO1 binding sequences present on the reporter plasmid to measure gene expression. The reporter assay performed with the mutated promoter did not show substantial differences between Syk-wild type and Syk-deficient cells ( Figure 3D), whereas a considerable difference was observed between them when performed with the original promoter ( Figure 3A). Second, chromatin immunoprecipitation (ChIP) PCR revealed that FOXO1 could bind to the predicted sequence of the Cd11c promoter ( Figure 3E). Third, FOXO1 overexpression increased Cd11c promoter activity, while mutation of the FOXO1 binding site in the Cd11c promoter reduced the activity ( Figure 3F). The FOXO1 overexpression in 293T cells was confirmed by western blotting ( Figure S3D).
These data suggest that SYK induced FOXO1 nuclear translocation, thereby upregulating Cd11c transcription.
Thus, with these data, it is reasonable to conclude that SYK and JNK mediated the GM-CSF receptor/ FOXO1/CD11c signaling pathway.

FOXO1 controls CD11c expression in human cell lines
The data presented above are based on animal experiments. However, the homology of the FOXO1 binding region in the Cd11c promoter between humans and mice is low. To evaluate whether FOXO1 mediated CD11c expression in THP-1 and HL-60 cells (human cell lines), we performed flow cytometry and western blotting. These cell lines, which were differentiated into macrophage/monocyte-like cells with phorbol myristate acetate (PMA), showed upregulated CD11c cell surface expression and PMA-mediated FOXO1 translocation into the nucleus ( Figure 5A). Moreover, AS1842856, a FOXO1 inhibitor, suppressed CD11c expression ( Figure 5B).
These data indicate that FOXO1 upregulated CD11c expression in both human cell lines and mice.   Figures 6C and 6D). No substantial difference was observed in the relative abundance of macrophages on atherosclerotic plaques. Nevertheless, the abundance of monocytes isolated from the dimethyl sulfoxide (DMSO) and FOXO1 inhibitor-administered groups of Syk +/+ mice decreased ( Figure 6E). Moreover, although the levels of LDL-C in Syk +/+ mice administered AS1842856 were lower than those in Syk del/ del mice administered AS1842856, no considerable differences were observed in the levels of the other serum cholesterols. However, a difference was observed between the DMSO and FOXO1 inhibitor groups of Syk +/+ mice in terms of body weight ( Figure S4A).
These findings indicate that the FOXO1 inhibitor suppressed mouse atherosclerosis and decreased CD11c expression on monocytes in vivo.

DISCUSSION
Herein, we focused on SYK, a key protein involved in chronic inflammation, and determined that it promoted atherosclerotic development via inflammatory cell migration (Figures 2A, 2B, 2F, and S2C). Additionally, we found that JNK, FOXO1, and CD11c mediated GM-CSF and SYK-related cell migration in monocytes and macrophages. Moreover, AS1842856, a FOXO1 inhibitor, reduced atherosclerosis by decreasing CD11c expression ( Figures 6A-6C), indicating its therapeutic potential.
Our finding that Syk deletion suppressed atherosclerosis is consistent with a previous study demonstrating that the SYK-specific inhibitor fostamatinib, approved by the FDA for chronic idiopathic thrombocytopenic purpura, 37 suppressed the development of atherosclerosis in mice. 38 These results support the application of fostamatinib for the treatment of atherosclerosis. However, further clinical trials in humans are needed. Other studies have also reported that SYK is required for integrin signaling during cell migration. 39,40 Here, we observed that SYK upregulated CD11c expression. To evaluate signal transduction between SYK and CD11c in vitro, we stimulated BMDMs with GM-CSF to upregulate CD11c expression. GM-CSF plays a role in proinflammatory cytokines in various diseases, and GM-CSF receptor signaling is SYK mediated. 41,42 In GM-CSF-knockout mice, atherosclerosis and collagen-induced arthritis are reportedly mitigated. 43,44 In humans, otilimab, an anti-GM-CSF monoclonal antibody, ameliorates RA. 45 The number of GM-CSF-secreting peripheral blood mononuclear cells in SLE increases more than that in healthy controls, positively correlating with anti-double stranded-DNA titers. 46 With reference to these findings, it is reasonable to suggest that the pathway identified herein contributes to both atherosclerosis and inflammation.
FOXO1 plays various roles, including insulin signaling, lipid metabolism, cellular differentiation, cell proliferation, apoptosis, and DNA repair 47 ; however, whether FOXO1 promotes atherosclerosis is unknown. In a previous study, endothelium-specific FOXO isoform (FOXO1, 3a, and 4) triple-knockout mice exhibited atherosclerosis alleviation via enhanced nitric oxide availability, inflammation, and superoxide iScience Article generation. 48 In contrast, although myeloid-specific FOXO isoform triple-knockout mice did not exhibit suppression of atherosclerosis, their inflammatory response was increased. 49 The differences between these reports and our findings suggest that the isoforms function differently because AS1842856 inhibits FOXO1 more specifically than FOXO3A and FOXO4, and analysis showed that the half maximal inhibitory concentration value of AS1842856 for FOXO1 (0.03 mM) is significantly lower than that for FOXO3A and FOXO4 (>1mM). 33 Although these isoforms have overlapping functions, single-knockout mice express distinct phenotypes; 50 for instance, atherosclerosis was not suppressed in FoxO4-systemic knockout mice. 51 In addition, FoxO3 knockdown increases reactive oxygen species production in vascular endothelial cells, contributing to atherosclerotic development. 52 Although FOXO3 mediates nucleotide-binding oligomerization domain, leucine rich repeat and pyrin domain containing 3 inflammasome activation, which accelerates atherogenesis, 53 the function of FOXO3 in atherosclerosis remains unclear. Furthermore, the function of each isoform differs in different organs; notably, skeletal muscle-specific FOXO1 overexpression in ApoE À/À mice suppresses atherosclerosis. 54 The signal transduction pathway from SYK to FOXO1 has been reported: FOXO1 initially binds to a 14-3-3 protein in the cytosol. Stress, including oxidative stress, can activate SYK-mediated JNK signaling, leading to the phosphorylation of 14-3-3 proteins and subsequent dissociation of the FOXO1-14-3-3 protein complex. This results in the translocation of free FOXO1 to the nucleus to upregulate transcription iScience Article activities. [55][56][57] Taken together, our study and previous reports suggested that the GM-CSF receptor/SYK/ JNK/FOXO1/CD11c signaling axis is involved in atherosclerotic development.
Our initial goal was to find a common pathway between atherosclerosis and chronic inflammatory diseases. CD11c, an integrin family member, mediates cell migration via cell-cell adhesion during inflammatory responses. 24 It is related to the pathogenesis of chronic inflammatory diseases; a previous study reported that Cd11c expression in peripheral blood mononuclear cells from patients with RA is higher than that in those from healthy donors and that Cd11c levels correlate with serum interleukin-2 (IL)-6 and IL-8 levels. 58 Furthermore, M1-like macrophages, which express CD11c, contribute to atherosclerosis in SLE. 59,60 Antiintegrin monoclonal antibody is approved for the treatment of ulcerative colitis and Crohn disease, 61 although it does not target CD11c. Additionally, we focused on the common pathway related to Syk and cell migration in atherosclerosis, RA, and SLE. However, there may be additional shared pathways. For instance, the anti-phosphorylcholine antibody, which targets a component of oxidized LDL, exerts an atheroprotective effect. 62 It has been reported that the lower titers of anti-phosphorylcholine antibodies in individuals with RA and SLE were probably associated with an increased risk of cardiovascular events. 63,64 Hence, the capacity to produce anti-phosphorylcholine antibodies might be related to the development of atherosclerosis.
To the best of our knowledge, our study is the first to identify that FOXO1 TF upregulates CD11c expression to induce cell migration in the downstream signaling of SYK. Our findings also suggest that the GM-CSF receptor/SYK/JNK/FOXO1/CD11c signaling axis is commonly involved in the development of atherosclerosis and chronic inflammation. The findings of the study will serve as a reference for future studies that focus on elucidating the role of FOXO1 inhibitors in ameliorating collagen-related diseases in model mice.

Limitations of the study
Our study has some limitations. First, we hypothesized that Syk deletion attenuated cell migration, which could explain decreased macrophage area on atherosclerotic plaques and reduced CD11c expression on monocyte-macrophage lineages in Syk À/À mice; however, the decreased macrophage area could also be due to impaired cell differentiation. 22 The evaluation of cell differentiation and the related genes using RNA-seq ( Figures 2C and 2D) could help identify other therapeutic targets. Second, although a downward trend was observed, the administration of FOXO1 inhibitor 5 days a week for 8 weeks did not considerably decrease the macrophage area in atherosclerotic mice that were fed a high-fat diet (Figure 6D). Therefore, a long-term experiment, e.g., 16 weeks, as shown in Figure 1C, could help improve

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
We would like to thank M. Muratani from the Department of Genome Biology, Faculty of Medicine, University of Tsukuba, and Tsukuba i-Laboratory for transcriptomics analysis. This work was supported by the Japan Society for the Promotion of Science KAKENHI (grant numbers 21K16294, 18K08418, 21K08480, and 24117719).