YTHDF2 Regulates Macrophage Polarization through NF-κB and MAPK Signaling Pathway Inhibition or p53 Degradation

Macrophages are heterogeneous cells that can be polarized into M1 or M2 phenotype. m6A “reader” YTH domain family protein 2 (YTHDF2) has been the m6A binding protein with the highest activity, which can recognize and disturb m6A-containing mRNA in processing bodies to reduce mRNA stability. YTHDF2 is recently identified as an effective RNA binding protein that modulates inflammatory gene levels within inflammatory responses. However, the role of YTHDF2 in M1/M2 macrophage polarization has not been reported. We established a M1/M2 macrophage polarization model using bone-marrow-derived macrophages and found that the expression levels of YTHDF2 in M1/M2 macrophages were both elevated. YTHDF2-knockdown macrophage polarization model was then established, and through qPCR, ELISA, and FACS, we discovered that suppressing YTHDF2 encouraged M1 polarization but restrained M2 polarization. In M1 macrophages, YTHDF2 silencing had no significant effect on p53 expression; however, in YTHDF2 knockdown, M2 macrophage p53 expression was remarkably upregulated. p53 inhibitor PFT-α was then applied and revealed that suppressing p53 simultaneously promoted YTHDF2-silenced M1 polarization and facilitated M2 macrophage polarization. Actinomycin D assays were further utilized to examine the mRNA degradation level of different cytokines, and p53 mRNA degradation in YTHDF2-depleted M2 cells was discovered impeded. Western Blot analysis also implied that a deficit in YTHDF2 expression may activate MAPK and NF-κB pathways. In this study, YTHDF2 induces M2 macrophage polarization by promoting the degradation of p53 mRNA. YTHDF2 suppresses M1 macrophage polarization by inhibiting NF-κB, p38, and JNK signaling pathways, yet p53 remains unaffected in YTHDF2-silenced M1 macrophages.

N6-methyladenosine (m 6 A), methylated in adenosin's N6 position, shows the highest prevalence among internal RNA modifications in eukaryotes [8]. It has been reported and actively involved in many critical stages during the post-transcriptional course of RNA and regulates gene expression by modifying RNA processing, including localization, translation, and eventual decay [9,10]. m 6 A deposition functions co-transcriptionally through its methyltransferases ("m 6 A writers"), comprising the catalytic subunit methyltransferase-like 3 (METTL3) and METTL14, as well as its demethylases ("erasers") like ALKB homolog 5 (ALKBH5) and fat mass and obesity-associated protein (FTO) [11]. m 6 A modification can be "interpreted" through the binding of m 6 A "reader" proteins, like YTHdomain containing protein (YTHDC1, 2) and YTHdomain family proteins (YTHDF1-3) [12]. m 6 A is correlated with numerous biological activities, such as stem cell differentiation and pluripotency, embryogenesis, DNA damage response, and tumorigenesis [12][13][14][15]. Researches have been recently carried out to reveal the significant role of m 6 A in inflammatory responses, such as microglia inflammation, renal inflammation, and endothelial inflammation [16][17][18].
The m 6 A binding protein YTHDF2, belonging to the YTH domain family (YTHDF), can selectively bind m 6 Amethylated mRNA to destabilize or degrade mRNA [19]. YTHDF2 has been reported to have indispensable effect on various physiological courses, for example, neural development, cancer progression, and hematopoietic stem cell expansion [20][21][22]. Until recently, YTHDF2 has also been found to relate to inflammatory response progress [23]. YTHDF2 depletion within human hepatic cell carcinoma (HCC) cells or knockdown within mouse hepatocytes prompts metastasis, vascular reconstruction, and inflammation by mediating mRNA decay in cytokines that contain m 6 A [23]. Our previous study indicated that YTHDF2 regulated LPS-induced inflammatory response of macrophages through specifically mediating target mRNAs degradation [24]. As revealed by our RNA sequence results, differentially expressed genes within YTHDF2-silenced macrophages were predominantly enriched in the p53 signaling pathway (unpublished data). The transcription factor p53 is a well-known tumor suppressor and is recently reported to participate in the regulation of macrophage polarization [25][26][27]. In p53 deficient mice, LPS stimulation prompts proinflammatory cytokine production in macrophages by regulating NF-κB activity [25]. Another study on p53deficient mice revealed that in IL-4-stimulated peritoneal macrophages, M2 markers Arg1, interferon regulatory factor 4 (Irf4), Fizz1, and c-Myc were highly expressed [27]. In addition, through the mRNA analysis of p53 in the m 6 A modification database SRAMP (sequence-based RNA adenosine methylation site predictor) (http://www .cuilab.cn/sramp/), our team found that the mRNA of p53 has m 6 A modification sites (unpublished data), so it is speculated that p53 may be regulated by m 6 A and participate in the regulation of YTHDF2 on macrophage polarization. This present study focused on investigating YTHDF2 expression within M1/M2 polarization model as well as how YTHDF2 knockdown affects macrophage polarization. We further present evidence that YTHDF2 inhibited M1 polarization through the activation of MAPK and NF-κB pathways while promoted M2 polarization by destabilizing p53 mRNA.

Materials and Methods
2.1. Ethical Statement. This work gained approval from Ethical Review Board of Guanghua School of Stomatology of Sun Yat-sen University. Each animal experimental procedure was conducted in line with "Guide for the Care and Use of Laboratory Animals" formulated via the US National Institutes of Health. In addition, animal number adopted in this study was minimized.

Cell Culture and Macrophage Polarization.
In this work, C57BL/6 mouse (6-8-weeks-old) were sacrificed (Animal Center of Sun Yat-sen University) and immersed them in 75% ethanol. Dissect the tibias and femurs from the body, and wash them with 90% alpha-minimal complete medium (α-MEM; Gibco, New York, NY, USA) including the concentration of 10% fetal bovine serum (FBS; Gibco, Carlsbad, CA, USA) with supplementation of penicillin/streptomycin. Expel the bone marrow cells with a 0.5 ml syringe containing α-MEM complete medium. Thereafter, cells were isolated by 24-h cultivation within α-MEM and 10% FBS. Besides, this work collected the suspended cells for resuspension within the α-MEM that contained 10% FBS and 30 ng/ml M-CSF (Sino Biological, Beijing, China). At day 6 of postincubation under 37°C, 95% air, and 5% CO 2 conditions, bone-marrow-derived macrophages (BMDMs) were harvested. Logarithmic growth phase cells were utilized. After reaching 80% cell density, 0.25% trypsin/EDTA (Gibco; Thermo Fisher Scientific, Inc.) was utilized for detaching macrophages.

YTHDF2 Small
Interfering RNA (siRNA) Transfection. YTHDF2 was knocked down by siRNA transfection in macrophages. Cells (6 × 10 5 /well) were later inoculated before transfection into the 6-well plates that contained 2 ml α-MEM for a 24-h period. By adopting Lipofectamine ® 3000 transfection reagent (7 μL, Invitrogen; Thermo Fisher Scientific, Inc.), cells reaching 70% density were exposed to transfection with 50 nM siRNA against YTHDF2 or a nontargeting siRNA control (siYTHDF2 and the negative control NC; Invitrogen, Carlsbad, CA, USA). Macrophages were subject to 24-h incubation under 5% CO 2 and 37°C conditions following specific protocols. The present study . In addition, after being exposed to transcription suppression for 3 and 6 h, RNA samples were harvested for measuring the respective mRNA degradation. Finally, the half-life of corresponding mRNA was determined by degradation rate and level of obtained mRNA.
2.9. Statistical Analysis. Every assay was carried out in triplicate for 3 or more replicates. Results were shown by mean ± SD. In addition, this work utilized SPSS20.0 (SPSS Inc., Chicago, IL, USA) for One-way ANOVA and student's t -test in carrying out statistical analyses. Obviously, P < 0:05 stood for statistical significance.
For further verification, this work detected cell surface markers for macrophages. After M1 stimulation, the cells presented obvious M1 phenotype with a notable growth in CD86+ cells and CD16/32+ cells, while M2 stimulation increased M2 features with an increase in DECTIN-1+ and CD206+ cells (Figure 1(e)). We further used qRT-PCR and WB assays for assessing YTHDF2 level during polarization for confirmation. The level of YTHDF2 showed a notewor-thy increase after M1 (Figures 1(f) and 1(g)) and M2 stimulation (Figures 1(h) and 1(i)).

YTHDF2 Knockdown Promotes M1 but Inhibits M2
Polarization. To find out YTHDF2's effect on polarization in M1/M2 macrophages, siRNAs were designed for suppressing YTHDF2 expression in M0 cells. According to Figures 2(a) and 2(b), YTHDF2 mRNA and protein expression were notably decreased after siRNA treatment. The knockdown efficiency of siYTHDF2 #1 was the highest, so we selected it to continue the following experiments.

YTHDF2 Knockdown Increases p53 mRNA Stability but
Does Not Significantly Affect Pro-Inflammatory Cytokines mRNA Stability. For exploring the role of YTHDF2 in regulating macrophage polarization through modulating the mRNA degradation of p53 or pro-inflammatory factors, IL-6, IL-10, TNF-α, p53, and TGF-β mRNA stability were measured with actinomycin D. As indicated in Figure 4, YTHDF2 knockdown promoted the stability of p53 mRNA transcript in M2 macrophages, but no significant difference was detected among the mRNA stability of the cytokines. These results indicate that YTHDF2 has no notable influence on IL-6, p53, and TNF-α mRNA stability of M1 cells or TGF-β and IL-10 of M2 macrophages; YTHDF2 silencing upregulated the expression of p53 through stabilizing its mRNA, thereby inhibiting the polarization of M2 macrophages.
For further validating how both signaling pathways affected M1 macrophage polarization in YTHDF2-silenced cells, the inhibitors NF-κB (BAY 11-7082), p38 (SB203580) and JNK (SP600125) were applied sepaparately to impede the signaling. IL-6 and TNF-α mRNA expression were then evaluated. According to the results, the inhibitors of NF-

Discussion
N6 position of adenosine (m 6 A) shows the highest prevalence among internal epigenetic modifications in mRNA [28]. Methyltransferase serves as a "writer" of m 6 A modification, demethylases work as the "eraser," and m 6 A-selectivebinding proteins act as the "reader" by selective recognizing methylated RNA to perform regulations [13,14]. YTHDF2 is a well-known "reader" protein that targets and facilitates the degradation of m 6 A-containing RNAs [29]. Recent discoveries on YTHDF2 highlighted that YTHDF2 has a critical effect on regulating neural development, hematopoietic stem cell proliferation, cancer development, viral infection, and other physiological and pathological processes [23,[29][30][31].
Macrophages are heterogeneous cells endowed with great plasticity [32]. Upon exposure to different stimuli, recruited macrophages can be polarized into M1 or M2 phe-notypes [3,4,33]. While M1 macrophages mediate innate immune responses against pathogens and activate adaptive responses through antigen processing and presentation, M2 macrophages are important in eliminating inflammation, tissue repairing, and maintaining homeostasis [34,35]. Recently, studies in inflammatory and autoimmune diseases implicated that m 6 A modifications have regulatory roles in the activation of macrophages [36,37]. m 6 A methyltransferase promotes M1 polarization through methylating the mRNA of STAT1 [36]. Knockdown of demethylase FTO inhibits M1 polarization and restrains M2 activation at the same time [37]. However, the effect of m 6 A reader on macrophage activation still lingers to be elucidated.
To investigate the role of m 6 A reader YTHDF2 during macrophage polarization, BMDMs was used to establish the M1/M2 polarization system and investigated the expression of YTHDF2 after macrophage polarized. In our study, IL-4 increased the mRNA expressions of IL-10, TGF-β, Arg1, and Fizz in M2 cells, suggesting that M2 polarization model was established successfully. However, the changing expression levels of each inflammatory factor were not consistent at different time points. Secreted cytokines can bind to different receptors to induce activation of an intracellular   (e, f) M0 macrophages subject to transfection using YTHDF2 siRNA or NC-siRNA were further exposed to BAY 11-7082, SB203580, or SP600125 (inhibitors for NF-κB, p38, and JNK pathways, separately) for a 2-h period, while non-treated cells served as blank control. Thereafter, LPS/IFN-γ was added to stimulate cells for a 6-h period. IL-6 and TNF-α mRNA expression were determined through qRT-PCR, with GAPDH being an endogenous reference. Data are denoted by mean ± S:E:M: (n = 3). * P < 0:05 ; * * P < 0:01.

11
Disease Markers cascade of signal transduction, which leads to various cellular responses. However, not all cells within an organism are identical. They differ in the amount of proteins involved in signal transduction. These differences shape cellular communication and responses to intracellular signaling. So, there might be a negative regulation responsible for the expression level of TGF-β to begin to reduce after a sharp increase [38]. After confirming the activation of M1/M2 phenotypes, YTHDF2 mRNA and protein expression were determined. As presented in our result, YTHDF2 expression increased significantly within M1 and M2 polarized cells, indicating that YTHDF2 might be involved in regulating macrophage polarization.
YTHDF2 can selectively recognize m 6 A to regulate mRNA degradation [15,19]. Studies have shown that YTHDF2 enhances the capacity of self-renewal of the leukemia stem cells and neural stem/progenitor cells by suppressing the stability of multiple mRNAs critical for cell expansion [39]. YTHDF2 depletion in zebrafish embryos slows down the decline of maternal mRNAs that been m 6 A-modified and impedes the cell cycle, thereby restraining the growth development during vertebrate embryogenesis [29]. In the study of infectious diseases, YTHDF2 upregulation promotes HIV-1 and HBV levels as well as viral replication ability [40,41]. Most recently, a study by our team found that YTHDF2 negatively regulates the mRNA expression levels of MAP2K4 and MAP4K4 via destabilizing their mRNA transcripts, which inhibits the inflammatory response in LPS-stimulated inflammatory reactions [24]. For exploring YTHDF2's effect on macrophage polarization, YTHDF2 expression was silenced in BMDMs, and M1 and M2 markers levels were examined. IL-1β, IL-6, iNOS, TNF-α, CD86, as well as CD16/32 levels were upregulated within YTHDF2-silenced M1 cells. Meanwhile, in YTHDF2-silenced M2 cells, the secretion of M2 markers IL-10, TNF-α, ARG-1, FIZZ, CD206, and DECTIN-1 experienced a significant reduction. Therefore, these findings demonstrated that the expression of YTHDF2 is increased in both M1/M2 cells and that YTHDF2 might have different roles during the orientation of macrophages, with YTHDF2 inhibiting M1 but promoting the M2 phenotype.
Our preliminary RNA sequence results found that genes differentially expressed in YTHDF2-silenced macrophages were mainly enriched in the p53 signaling pathway. The gene p53 is the most common tumor suppressor gene in human cancer [42,43]. It functions as a crucial regulatory node through monitoring the expression of genes associated with metabolism, cell cycle arrest, and apoptosis [44][45][46]. Studies in cancer have found that inflammation is a vital aspect when it comes to determining its predisposition [47,48] and p53 has recently been discovered working as a regulator in various inflammatory diseases [25,26,27,49]. As a guardian of homeostasis, p53 plays a protective role by inhibiting the local inflammation of rheumatoid arthritis patients and collagen-induced osteoarthritis in mice [49].  Figure 6: Role of m6A "reader" YTHDF2 in polarization of M1/M2 macrophages. YTHDF2 suppression promotes polarization of M1 cells via MAPK and NF-κB pathway activation. Moreover, depleting YTHDF2 stabilizes p53 mRNA and upregulates its expression, thereby inhibiting the polarization of M2 macrophages. 12 Disease Markers p53 also controls immunity by directly impeding the activation of p65 promoter in the NF-κB signaling and negatively regulating its transcriptional expression of its downstream genes IL-6, Cox-2, and Nos2 [25]. Recent studies found that p53 can regulate macrophage polarization [26,27,[50][51][52].
Macrophages lacking p53 promoted the responses to LPS stimulation, producing more pro-inflammatory M1 marker genes, like IL-6, TNF-α, and MIP-2 [26]. When marrowderived macrophages activated towards to M2 phonotype, cells display endogenous p53 activity and the p53 activation can in turn inhibit expression of M2 genes [27]. The p53 activator Nutlin-3a in bone marrow-derived macrophages can reduce the expression of M2 subtype [51]. To investigate whether p53 involved in the activation of YTHDF2deficent macrophages, we examined the expression of p53 in macrophages after YTHDF2 knockdown. As shown in the result, YTHDF2 silencing did not significantly affect p53 level within M1 cells; however, p53 increased within YTHDF2-silenced M2 cells. To further determine the role of p53, the p53 inhibitor Pifithrin-α (PFT-α) was used to suppress the p53 expression in YTHDF2-knockdown cells. According to the above findings, PFT-α pretreatment further increased TNF-α and IL-6 levels within YTHDF2knockdown M1 cells, but the reduction of TGF-β and IL-10 within YTHDF2-knockdown M2 subtype was reversed. Collectively, YTHDF2 promoted M2 polarization by regulating p53, but did not repress the M1 polarization directly through p53. Accumulating evidences suggest that m 6 A may regulate the stability of the RNA through the effects of YTHDF2 [29]. The C-terminal YTH domain (YTHDF2-C) of YTHDF2 can selectively bind m 6 A, whereas P/Q/N enriched N-terminal region promote target mRNA migration into cytoplasmic foci (P bodies) while recruiting RNA import of CCR4-NOT deadenylase complex [53]. To investigate the role of YTHDF2 in destabilizing the gene transcripts of related cytokines in macrophage polarization, we measured the stability of IL-6, IL-10, TNF-α, p53, and TGF-β mRNAs in YTHDF2-depletion cells. According to our results, YTHDF2 silencing promoted p53 mRNA stability in M2 macrophages but had little effect on p53 stability in M1 macrophages. We further analyzed and predicted p53 mRNA on the m 6 A database SRAMP and found that p53 mRNA may have 11 m 6 A modification sites, while among other polarization-related factors, only IL-6 and ARG1 have a few m 6 A modification sites, which contains 3 and 1, respectively. This might explain why loss of YTHDF2 exhibited no detectable effect on the mRNA stability of TNF-α and IL-6 in M1 cells or TGF-β and IL-10 in M2 cells. Therefore, we concluded that YTHDF2 may encourage M2 polarization by promoting the degradation of p53 mRNA, but inhibit M1 polarization leaving the mRNA stability of inflammatory factors unaffected.
MAPK and NF-κB pathways have been identified as key pathways for inflammation and M1 macrophage polarization [54][55][56]. Recent studies discovered that YTHDF2 upregulation can inhibit ERK and MEK activation within liver cancer cells [23]. For verifying whether YTHDF2 regulates M1 polarization by deactivating MAPK and NF-κB pathways, we detected some critical molecules related to the above pathways for their phosphorylation levels after silencing YTHDF2. In our results, IKKα/β, p65, IκBα, p38, and JNK phosphorylation levels significantly elevated after YTHDF2 silencing, but ERK phosphorylation level was suppressed. Studies have found that ERK1/2 mainly function during cell growth, proliferation, and differentiation, while p38 and JNK play essential roles in inflammation, cytokine production, and apoptosis. Besides, the activation of ERK can be suppressed by JNK and p38 kinase. These might explain the inactivation of ERK in YTHDF2-knockdown M1 macrophages. Furthermore, we inhibited the above pathways to validate their effects on regulating M1 markers' expression. Our results stated that inhibitors of NF-κB, p38, and JNK pathways downregulated IL-6 and TNF-α levels in YTHDF2-depleted M1 cells, confirming that YTHDF2 impeded macrophage M1 polarization by inhibiting NF-κB, p38, and JNK signaling pathways.
To sum up, the m 6 A reader YTHDF2 increased its expression within M1 and M2 polarized cells. YTHDF2 silencing encouraged M1 but diminished M2 macrophage polarization. Mechanistically, silencing YTHDF2 promoted the stability of p53 mRNA and its expression level, thereby impeding M2 macrophage polarization; YTHDF2 depletion facilitated M1 polarization by triggering both MAPK and NF-κB pathways ( Figure 6). In this study, YTHDF2 was identified with its regulatory role during M1/M2 polarization. Our present research on m 6 A reader YTHDF2 may offer an alternative approach for the understanding of macrophage plasticity and probably a newfound target for treating inflammatory diseases.

Data Availability
All data utilized in the present work can be obtained from corresponding author upon reasonable request.

Conflicts of Interest
All authors declared no competing interest.