m6A mRNA Methylation Regulates Epithelial Innate Antimicrobial Defense Against Cryptosporidial Infection

Increasing evidence supports that N6-methyladenosine (m6A) mRNA modification may play an important role in regulating immune responses. Intestinal epithelial cells orchestrate gastrointestinal mucosal innate defense to microbial infection, but underlying mechanisms are still not fully understood. In this study, we present data demonstrating significant alterations in the topology of host m6A mRNA methylome in intestinal epithelial cells following infection by Cryptosporidium parvum, a coccidian parasite that infects the gastrointestinal epithelium and causes a self-limited disease in immunocompetent individuals but a life-threatening diarrheal disease in AIDS patients. Altered m6A methylation in mRNAs in intestinal epithelial cells following C. parvum infection is associated with downregulation of alpha-ketoglutarate-dependent dioxygenase alkB homolog 5 and the fat mass and obesity-associated protein with the involvement of NF-кB signaling. Functionally, m6A methylation statuses influence intestinal epithelial innate defense against C. parvum infection. Specifically, expression levels of immune-related genes, such as the immunity-related GTPase family M member 2 and interferon gamma induced GTPase, are increased in infected cells with a decreased m6A mRNA methylation. Our data support that intestinal epithelial cells display significant alterations in the topology of their m6A mRNA methylome in response to C. parvum infection with the involvement of activation of the NF-кB signaling pathway, a process that modulates expression of specific immune-related genes and contributes to fine regulation of epithelial antimicrobial defense.

Increasing evidence supports that N6-methyladenosine (m 6 A) mRNA modification may play an important role in regulating immune responses. Intestinal epithelial cells orchestrate gastrointestinal mucosal innate defense to microbial infection, but underlying mechanisms are still not fully understood. In this study, we present data demonstrating significant alterations in the topology of host m 6 A mRNA methylome in intestinal epithelial cells following infection by Cryptosporidium parvum, a coccidian parasite that infects the gastrointestinal epithelium and causes a self-limited disease in immunocompetent individuals but a life-threatening diarrheal disease in AIDS patients. Altered m 6 A methylation in mRNAs in intestinal epithelial cells following C. parvum infection is associated with downregulation of alpha-ketoglutarate-dependent dioxygenase alkB homolog 5 and the fat mass and obesity-associated protein with the involvement of NF-кB signaling. Functionally, m 6 A methylation statuses influence intestinal epithelial innate defense against C. parvum infection. Specifically, expression levels of immune-related genes, such as the immunity-related GTPase family M member 2 and interferon gamma induced GTPase, are increased in infected cells with a decreased m 6 A mRNA methylation. Our data support that intestinal epithelial cells display significant alterations in the topology of their m 6 A mRNA methylome in response to C. parvum infection with the involvement of activation of the NF-кB signaling pathway, a process that modulates expression of specific immune-related genes and contributes to fine regulation of epithelial antimicrobial defense.

INTRODUCTION
Increasing evidence supports that RNA methylation is a widespread phenomenon and a critical regulator of gene expression (1,2). The most prevalent RNA methylation, N6-methyladenosine (m 6 A), is a reversible RNA post-transcriptional modification and occurs in approximately 25% of transcripts at the genome-wide level (1). RNA m 6 A methylation regulates RNA splicing, translocation, stability, and translation into protein (3)(4)(5)(6). Dynamic regulation of the m 6 A epi-transcriptome is involved in diverse cellular functions, including heat shock, DNA damage, cancer, stem cell differentiation, circadian rhythm, spermatogenesis and oogenesis, response to interferon-g, and viral infections (2,3,7). m 6 A dynamics and functions are executed by three groups of proteins: methyltransferases or "writers", demethylases or "erasers", and m 6 A-binding proteins or "readers" (2,3,7). In most cell types, m 6 A methylation is catalyzed by the methyltransferase complex consisting of the methyltransferase-like 3 (METTL3) and METTL14, as well as their cofactors (3,7). The erasers include the fat mass and obesity-associated protein (FTO) and alphaketoglutarate-dependent dioxygenase alkB homolog 3 (ALKBH3) and ALKBH5 (2,3,7). Recent studies demonstrate that m 6 A methylation may play an important role in regulating immune responses (8,9). It has been associated with numerous physiological and pathological phenomena, including obesity, immunoregulation, yeast meiosis, plant development, and carcinogenesis (2,10). Specifically, m 6 A methylation has been recognized as crucial regulator in T cell homeostasis, inflammation, type I interferon production, and the immune response to bacterial or viral infection (3,(8)(9)(10)(11). Selectively altered m 6 A levels along with other types of immunotherapies may be efficient management strategies in a variety of immunological diseases.
Epithelial cells along the mucosal surface provide the front line of defense against luminal pathogen infection in the gastrointestinal tract and are an important component of gastrointestinal mucosal immunity (12). Intestinal epithelial cells generate various types of barriers to protect the intestinal mucosa from commensal microbes or invading pathogenic microorganisms. Upon microbial challenge, gastrointestinal epithelial cells quickly initiate a series of innate immune reactions, including production of antimicrobial molecules and release of inflammatory chemokines/cytokines. These chemokines/cytokines of epithelial cell origin may mobilize and activate immune effector cells to the infection sites (13). Therefore, intestinal epithelial cells not only create mucosal barriers to 'segregate' gut microbes and gut immune cells but also sense signals from both populations and secrete humoral factors to 'mediate' the balance between both populations (14). Failure to maintain the complex functional and anatomical features of the intestinal epithelium reduces the antimicrobial, immunoregulatory and regenerative ability of the epithelial barrier and might allow translocation of commensal bacteria from the intestinal lumen to the subepithelial tissue (13,14). Although much is known about the active role of epithelial innate immune stimulation in antimicrobial host defense and host-microbial homeostasis, how intestinal epithelial cells orchestrate gastrointestinal mucosal defense and homeostasis is still not fully understood.
Cryptosporidium spp, a coccidian parasite and an NIAID Category B priority pathogen, infects the gastrointestinal epithelium and causes a self-limited disease in immunocompetent individuals but a life-threatening diarrheal disease in AIDS patients (15)(16)(17). After rotavirus, Cryptosporidium is the most common pathogen responsible for moderate-to-severe diarrhea in children younger than 2 years (18). The majority of human cryptosporidial infections are caused by two species: C. parvum and C. hominis (15). C. parvum attaches to the apical membrane surface of intestinal epithelial cells (mainly enterocytes) and forms an intracellular but extra-cytoplasmic vacuole in which the organism remains (15). Thus, C. parvum is classified as a "minimally invasive" mucosal pathogen (15) and innate epithelial defense is critical to the host's defense against C. parvum infection (19). In this study, we present data demonstrating significant alterations in the topology of host m 6 A mRNA methylome in intestinal epithelial cells following C. parvum infection. C. parvum infection promotes m 6 A mRNA methylation in intestinal epithelial cells through downregulation of Alkbh5 with the involvement of NF-кB signaling. Functionally, m 6 A methylation statuses influence intestinal epithelial anti-C. parvum defense. Specifically, expression levels of immune-related genes, such as the immunity-related GTPase family M member 2 (Irgm2) and interferon gamma induced GTPase (Igtp, also called as Irgm3 in mice), are increased in infected cells with a decreased m 6 A mRNA methylation. Our data support that intestinal epithelial cells display significant alterations in the topology of their m 6 A mRNA methylome in response to C. parvum infection with the involvement of activation of the NF-кB signaling pathway, a process that modulates expression of specific immune-related genes and contributes to fine regulation of epithelial antimicrobial defense.

C. parvum and Cell Lines
C. parvum oocysts of the Iowa strain were purchased from a commercial source (Bunch Grass Farm, Deary, ID). The mouse intestinal epithelial cell line (IEC4.1) was received as a kind gift from Dr. Pingchang Yang (McMaster University, Hamilton, Canada). The HCT-8 cells were human intestinal epithelial cells from ATCC (Manassas, Virginia). The BV2 mouse microglia cells and RAW264.7 mouse macrophage cells were obtained from ATCC. Culture media were supplied with 10% FBS (Ambion, Austin, Texas) and antibiotics (100 IU/ml of penicillin and 100 µg/ml of streptomycin).

Infection Models and Infection Assays
Models of intestinal cryptosporidiosis using intestinal epithelial cell lines and enteroids were employed; infection was done with a 1:1 ratio between C. parvum oocysts and host cells as previously described (20)(21)(22). Intestinal epithelium and enteroids were isolated and cultured as previously described (22). Briefly, small intestines were opened longitudinally and washed with ice-cold Ca2 + and Mg2 + free PBS, then were cut into 1-2 mm fragments and washed with ice-cold Ca2 + and Mg2 + free PBS 3 times. The cut

rRNA Removal and Quality Analysis
Total RNA was isolated from IEC 4.1 cells with TRAZOL Reagent (Invitrogen). Contaminated rRNA was removed by using RiboMinus ™ Eukaryote Kit (Invitrogen, #A10837-08). The ribosomal RNA depleted RNA concentration was fragmented using RNA fragmentation reagents (Invitrogen, #AM8740). The RNA fragment was measured by NanoDrop and the quality of RNA was analyzed with Agilent 2100 bioanalyzer.

m 6 A Dot Blot
The ribosomal RNA depleted RNA concentration of each whole cell lysate was determined and subsequently analyzed by dotblot. Anti-m 6 A (Synaptic Systems, #202003) was used for blotting. Isolated RNA was first denatured by heating at 95°C for 3 min, followed by chilling on ice rapidly. Two-fold serial dilutions were spotted on an Amersham Hybond-N+ membrane optimized for nucleic acid transfer (GE Healthcare). After UV crosslinking in a Stratagene Stratalinker 2400 UV Crosslinker, the membrane was washed by 1×TBST buffer, blocked with 5% of non-fat milk in TBST, and incubated with anti-m 6 A antibody (1:1,000) overnight at 4°C. After incubating with HRPconjugated anti-rabbit IgG secondary antibody, the membrane was visualized by ECL Western Blotting Detection Kit (Thermo).

RNA Stability
RNA stability assay was performed by real-time PCR as previously reported (27); modifications are described in the Supplemental Experimental Procedures.

Luciferase Assay
The promoter region sequence of Alkbh5 or Fto (-2kb~0) was cloned into the pGL3 vector, and plasmids were transfected to IEC4.1 cells with Lipofectamine 2000 following the manufacturer's protocol (Santa Cruz Biotechnology). Transient transfected cells were harvested with Reporter lysis buffer (Progema). The activity of luciferase was then determined by Luciferase assay system (Progema) as previously reported (28). For specific details, see the Supplemental Information and Table S3.

ChIP Analysis
The formaldehyde crosslinking ChIP was performed as described (28)(29)(30). ChIP analysis was performed with a commercially available ChIP Assay Kit (Upstate Biotechnologies) in accordance with the manufacturer's instructions. For specific details, see the Supplemental Information and Table S3.

Bioinformatics and Statistical Analysis
For specific details about the bioinformatic analysis, see the Supplemental Experimental Procedures. All values are given as mean ± S.E. Means of groups were from at least three independent experiments and compared with Student's t test (unpaired) or the ANOVA test when appropriate. p values < 0.05 were considered statistically significant.

Elevated Level of Global mRNA m 6 A Methylation in Intestinal Epithelium Following Cryptosporidium Infection
We first characterized the global mRNA m 6 A status in intestinal epithelial cells following C. parvum infection. Using an in vitro infection model employing IEC4.1 cells, which are transformed but non-tumorigenic intestinal epithelial cells from neonatal mice (5-7 days old) (32) and received from Dr. Pingchang Yang (McMaster University, Hamilton, Canada), we measured the global m 6 A mRNA methylation levels using the m 6 A RNA methylation quantitation assay and dot blot as previously reported (33,34). We demonstrated a significant increase in the m 6 A mRNA methylation level in IEC 4.1 cells following C. parvum infection as revealed by m 6 A RNA methylation quantitation assay ( Figure 1A) and dot blot ( Figure 1B).
Using an ex vivo infection model employing 2D enteroid monolayers from neonatal mouse ileum (23,24), we detected an increase of global m 6 A mRNA methylation level in infected intestinal epithelial monolayers ( Figure 1C). Previous studies indicate that C. parvum infection activates NF-кB and IFN-a signaling in infected host cells (25,28,29,35). To define if elevated m 6 A mRNA methylation level in infected cells is due to activation of NF-кB and/or IFN-a signaling, we measured the m 6 A mRNA methylation levels in IEC4.1 cells following stimulation by TNF-a (to activate NF-кB signaling) and IFN-a. Indeed, induction of m 6 A mRNA methylation status was detected in IEC4.1 cells stimulated with TNF-a and IFN-a ( Figure S1).

C. parvum Infection Promotes m 6 A mRNA Methylation in Intestinal Epithelial Cells Through Downregulation of Alkbh5 and Fto With the Involvement of NF-кB Signaling
To explore the underlying mechanism of C. parvum-induced m 6 A mRNA methylation in infected intestinal epithelial cells, we analyzed the expression levels of key effectors regulating m 6 A methylation in general, including the major writers, erasers and readers (2,3,7). We previously performed a genome-wide transcriptome analysis of C. parvum-infected IEC4.1 cells (28). From this dataset, out of the genes coding these key effector molecules, we found out significant decreased RNA expression levels of Alkbh5 and Fto, while others showing no significant changes in their expression levels ( Figure 2A). We therefore focused on the two genes to test whether decrease of their expression levels contributes to C. parvum-associated m6A methylation in infected IEC4.1 cells. We confirmed the downregulation of Alkbh5 and Fto in infected IEC4.1 cells at the RNA level (using real-time PCR, Figure 2B) and at the protein level (using Western blot, Figure 2C), as well as at the RNA level in isolated intestinal epithelium from neonatal mice of intestinal cryptosporidiosis through oral administration of the parasite (23, 24) ( Figure 2D), and infected 2D enteroid monolayers from neonatal mouse ileum ( Figure 2E). Of note, the antibody detected multiple isoforms of Alkbh5 protein. Consistent with results from previous studies (23,24), upregulation of the inflammatory Cxcl2 gene was detected in infected IEC4.1 cells as a control ( Figure 2B). Expression levels of Mettl3 and Mettl14 were with a tendency of decrease in IEC4.1 cells following C. parvum infection at 24h, but without statistical significance and this tendency was not observed in other time points following infection ( Figure 2B). Downregulation of Alkbh5 or Fto was further detected in IEC4.1 cells treated with TNF-a ( Figure S2) or IFN-a ( Figure S3). Given the fact that activation of NF-кB signaling is a common cellular response in intestinal epithelial cells following C. parvum infection and upon TNF-a stimulation (25,36), we asked whether NF-кB signaling is involved in the suppression of Alkbh5 expression in cells following C. parvum infection. We exposed IEC4.1 cells deficient in MyD88 (MyD88-knockout, MyD88-KO), one of the key upstream adaptor for pathogeninduced NF-кB activation (25), to C. parvum infection and then measured the expression level of Alkbh5. No decrease of Alkbh5 expression level was observed in the MyD88-KO IEC4.1 cells following C. parvum infection ( Figure 3A) or TNF-a stimulation ( Figure S2). Based on TFSEARCH (http://www. cbrc.jp/research/db/TFSEARCH.html) and MOTIF (http:// motif.genome.jp/) database searches (31,37), putative NF-кB binding sites were identified within the potential promoter region of the Alkbh5 gene. We then cloned the potential promoter regions of the Alkbh5 and Fto genes and inserted the sequences into the pGL-luciferase reporter vector. C. parvum infection decreased the luciferase activity in cells transfected with the luciferase construct that encompassed the promoter regions of the Alkbh5 and Fto genes, but not in cells transfected with the empty vector control ( Figure 3B). Decreased luciferase activity associated with the promoter region of the Alkbh5 gene induced by C. parvum infection was not observed in the MyD88-KO IEC4.1 cells ( Figure 3C). Moreover, decreased luciferase activity associated with the promoter regions of the Alkbh5 and Fto genes was also detected in IEC4.1 cells following stimulation with TNF-a or IFN-a ( Figure S4). Since IFN-a stimulation also suppressed Fto expression in IEC4.1 cells, we measured luciferase activity associated with the promoter region of the Fto gene in IEC4.1 cells deficient in Ifnar1 (CRISPR/Cas9 stable knockout cell line, lack of Ifnar1 the receptor subunit for Type I IFN signaling) (38). No significant change in luciferase activity epithelium from neonates of 5 days old were isolated and cultured into 2D monolayers followed by exposure to C. parvum infection for 24 h, as shown by phase and immunofluorescence microscopy (C. parvum parasites were stained in green as indicated by arrows). Intestinal epithelial 2D monolayers were exposed to C. parvum for 24h and RNA was collected for m 6 Figure S4).
To define how NF-кB signaling suppresses Alkbh5 gene transcription, we performed chromatin immunoprecipitation (ChIP) analysis to measure the occupancy of NF-кB subunits, p65 and p50, to the Alkbh5 gene. An elevated occupancy of p65, but not p50, to the promoter region of Alkbh5 gene locus was detected in IEC4.1 cells following C. parvum infection (Figures 3D, E) or TNF-a stimulation ( Figure S5). Previous studies indicate that recruitment of NF-кB subunits to targeted gene promoters may promote occupancy of suppressive histone deacetylase 1 (HDAC1) to suppress gene transcription (39,40). However, no increase of HDAC1 occupancy was detected in the promoter region of Alkbh5 gene in infected IEC4.1 cells ( Figure 3E) or cells following TNF-a stimulation ( Figure S5). An enrichment of H3K9me3, but not H3K27me3, to the promoter region of Alkbh5 gene locus was observed in IEC4.1 cells following C. parvum infection ( Figure 3E) or TNF-a stimulation ( Figure S5). Consistent with results from previous studies (41,42), the enrichment of p65, H3K9me3, and H3K27me3 to the promoter region of ApoE gene locus, an NF-кBassociated downregulating gene, was detected in infected IEC4.1 cells as a control ( Figure 3E). These data suggest that NF-кB signaling may count for the suppression of Alkbh5 in cells following C. parvum infection or TNF-a stimulation. To address this possibility, we took the CRISPR/Cas9 knock-out approach to establish stable IEC4.1 cells deficient in Alkbh5 or Fto.
Knock-out of Alkbh5 and Fto in IEC4.1 cells were confirmed by real-time PCR and Western blot ( Figure 4A). Accordingly, knockout of Alkhb5 or Fto caused a significant increase of global m 6 A mRNA methylation in IEC4.1 cells ( Figure 4B). Cells were then exposed to C. parvum infection for measurement of attachment/invasion (after incubation for 4h) and host antiparasite defense (so called infection burden, after incubation for 24 or 48h), as previously reported (44). A decreased infection burden was detected in IEC 4.1 cells deficient in Alkbh5 or Fto ( Figure 4B). We then took the CRISPR/Cas9 knock-in approach to establish stable IEC4.1 cells to overexpress Alkhb5 or Fto ( Figure 4C). Cells expressing Alkbh5 showed an increase of infection burden ( Figure 4D). Intriguingly, an increase of infection burden was not detected in cells constitutively expressing Fto ( Figure 4D). No obvious difference in the attachment/invasion of C. parvum was observed in cells deficient in Alkbh5 or Fto and in cells overexpressing Alkbh5 or Fto, compared with that in the control IEC4.1 cells ( Figure S6). Moreover, siRNAs to knockdown Alkbh5 also decreased the burden of C. parvum infection in IEC4.1 cells ( Figure 4E) and in 2D intestinal monolayers derived from neonatal mice ( Figure 4F).

Alterations in the Topology of Host m 6 A mRNA Methylome in Intestinal Epithelial Cells Following C. parvum Infection
We next examined the topology of host m 6 A mRNA methylome in IEC4.1 cells following C. parvum infection by performing methylated RNA immunoprecipitation sequencing (MeRIP-seq) experiments. For this, IEC4.1 cells were exposed to C. parvum infection for 24h. Total mRNA transcripts were isolated and processed for m 6 A sequencing (m 6 A-seq) experiments, as previously reported (4,45). We first examined the abundance and distribution of m 6 A peaks on host mRNA transcripts from uninfected and C. parvum-infected cells. Metagene analysis Cells were transfected with the generated reporter constructs and then exposed to C. parvum infection for 12h, followed by measurement of luciferase activity. Cells transfected with the empty vector were used as control. (C) Luciferase activity associated with the promoter of Alkbh5 in MyD88-/-IEC4.1 cells following C. parvum infection or TNF-a stimulation. IEC4.1 cells deficient in Myd88 were exposed to C. parvum infection (for 12h) or TNF-a stimulation (for 6h). Luciferase activity was measured. (D) Recruitment of NF-кB p65 to the Alkbh5 promoter region in IEC4.1 cells following C. parvum infection or TNF-a stimulation. Cells were exposed to C. parvum infection (for 24 h) or TNF-a stimulation (for 4h), followed by ChIP analysis using anti-p65 and the PCR primer sets as designed. Increased recruitment of p65 was detected in the -1571~-1361 (Set-1) region of the Alkbh5 gene locus in cells following C. parvum infection or TNF-a stimulation. (E) Recruitment of NF-кB subunits and HDAC1, as well as enrichment of H3K9me3 and H3K27me3, at the Alkbh5 promoter region in intestinal epithelial cells following C. parvum infection. Cells were exposed to C. parvum infection for 24 h, followed by ChIP analysis using anti-p65, anti-p50, anti-HDAC1, anti-H3K9me3, or anti-H3K27me3 and the PCR primer Set-1. Recruitment of NF-кB subunits and HDAC1, as well as enrichment of H3K9me3 and H3K27me3, at the ApoE promoter region in cells following C. parvum infection were also measured as a positive control. Data represent three independent experiments. *p<.05 vs the non-infected control. showed that C. parvum infection caused significant alterations in m 6 A peaks in the 118 regions of 80 corresponding genes in the transcriptome ( Figure 5A and Table S1). The top ten genes with a significant alteration in their m 6 A peaks are shown in Figure 5A and the complete 118 regions and corresponding genes are listed in Table S1. Of these regions, the majority are at the promoter regions (<1kb 33.68% and 1-2kb 6.32%) and the coding sequence (CDS) regions (1st exon 21.05% and other exon 29.47%) of the target genes ( Figure 5A). The other peaks are at the 3'UTR (6.32%) and 5'UTR (3.16%) regions ( Figure 5A). Newly emerged m 6 A methylation sites were detected in 22 promoter regions, 10 UTR regions and 38 CDS regions ( Figure 5B). Loss of exiting m 6 A methylation sites was detected in 16 promoter regions, 9 UTR regions and 31 CDS regions ( Figure 5B and Table S2). It appears that no significant difference of alterations in m 6 A peaks was observed between the 5'UTR and 3'UTR regions ( Figure 5B). We also performed a motif analysis of the newly emerged and lost m 6 A peaks in cells following C. parvum infection. It revealed three top motifs ( Figure 5C). The distribution of motifs covered over the promoters, UTRs and CDS regions, with varies in different genes ( Figure S7). All sequence data were described in accordance with MIAME guidelines and deposited at NCBI database (with the NCBI accession numbers: SRR14029773 -SRR14029784).  Gene ontology analysis of the genes with changed m 6 A peaks identified a broad range of gene categories among the most enriched pathways in both the newly gained and lost m 6 A methylation sites, including immune-related genes, genes for RNA splicing and translation, mitochondrion functions, and cell proliferation ( Figure 5D and Table S1). These immune-related genes include Igtp, Irgm2, Cx3cl1, Crabp1, Iqgap1, and Jmid8. Genes involving with RNA translation and splicing include Rpl21, Rpl23q, Rpl17, Rpl12, Cep85, Hnrnpab, Rbm8a, and Sf3b1. Genes associated with mitochondrion functions include Ndufb10, Idh3b, Wdr90, Bcap31, and Uqcrb. Cell proliferationrelated genes include Sf3b1, Fosl2, Ccna2, Plcd3, Fanca, Flna, Btc, and Sipa1 ( Figure 5D and Table S1).

mRNA Expression Profile and Its Association With m 6 A Peaks in Intestinal Epithelial Cells Following C. parvum Infection
Of these mRNAs isolated from uninfected and infected IEC4.1 cells and processed for m 6 A-Seq analysis as described above, we also took a portion of the mRNA collections for whole genome transcriptome (RNA-Seq) analysis. Consistent with results from previous studies, we detected many genes that were upregulated or downregulated in cells following infection. The top 10 induced genes are listed in Figure 6A and a full list of upregulated and downregulated genes is provided in Table S2. These upregulated genes include immune-related genes (e.g., Mx2, Igtp, Iift1, Ddx58, and Cxcl1), stress-responsive genes (Usp18, Oas3, Ier3, Cox7a1, and Uba7), and metabolism-related genes (Dusp1, Fos, Beu1, Gbp2, Zfp36, Wnt4, and Dtx3l) ( Figure 6B and Table S2). All sequence data were described in accordance with MIAME guidelines and deposited at NCBI database (with the NCBI accession numbers: SRR14163429 -SRR14163434).
Interestingly, comparison of genes whose expression levels were altered and genes with altered m 6 A methylation in infected IEC4.1 cells revealed that only a small portion of the genes was overlaid ( Figure 6C). This represents 13.75% of the genes with altered m 6 A levels and 3.89% of the genes those expression levels are either upregulated or downregulated in cells following C. parvum infection for 24h ( Figure 6C). The majority of genes with increased or decreased m 6 A levels did not show a significant change in their expression levels in cells following C. parvum infection. Representative overlay genes include Irgm2, Igtp, Sf3b1, Rbm8a, and Idh3b ( Figure 6D). There was no obvious correlation between their expression levels and the m 6 A site locations, such as m 6 A in the promoters, UTRs or CDS regions (data not shown). Moreover, gene ontology analysis of these 11 genes with altered expression levels and altered m 6 A methylation revealed board biological processes, including cell adhesion, metabolic and immune processes ( Figure 6D).
Expression Levels of Two Immunity-Related GTPase Genes, Irgm2 and Igtp, Were Increased With a Decreased m 6 A mRNA Methylation in IEC4.1 Cells Following C. parvum Infection Interestingly, we found out that expression levels of Irgm2 and Igtp (also called as Irgm3 in mice) were increased in infected cells. Given the fact that suppression of m 6 A methylation in IEC4.1 cells results in an increase of C. parvum burden, coupled with the important role of Irgm2 and Igtp in innate epithelial immunity (46,47), we looked more details about m 6 A methylation for the genes of Irgm2 and Igtp in infected cells. We found out that both Irgm2 (NM_019440) and Igtp (NM_018738) mRNAs showed a decrease in their m 6 A methylation in cells following C. parvum infection, as revealed by m 6 A-seq ( Figure 7A). Relevant motif and distribution of decreased m 6 A peaks in Irgm2 and Igtp mRNAs are shown in Figure 7B. Moreover, increased stability of Irgm2 mRNA was observed in Alkbh5-/-IEC4.1 cells, compared with that in IEC4.1 cells ( Figure 7C).  Figure 8A). Decrease of ALKBH5 and FTO expression levels was detected in HCT-8 cells following C. parvum infection ( Figure 8B). The impact of knockdown ALKBH5 on C. parvum infection burden was further confirmed in HCT-8 cells. We took the siRNA approach to knockdown ALKBH5 in HCT-8 cells ( Figure 8C). When HCT-8 cells were treated with the siRNA-ALKBH5 and then exposed to C. parvum infection for 24h, a significant decrease of infection burden was observed ( Figure 8D). Our data support that intestinal epithelial cells display significant alterations in the topology of their m 6 A mRNA methylome in response to C. parvum infection with the involvement of activation of the NF-кB signaling pathway, which may contribute to fine regulation of epithelial anti-C. parvum defense. m 6 A dynamics are finely controlled by various methyl transferases (or writers) and demethylases (or erasers) (2,3,7). Key methyltransferases are METTL3 and METTL14 and important demethylases include FTO, ALKBH3, and ALKBH5 (2,3,7). Previous studies have demonstrated that mammalian cells have developed strategies to modulate cellular m 6 A RNA methylation statutes in response to heat shock (49) or viral infection (11,50). However, little is known about the molecular mechanisms of how extracellular stimuli may activate intracellular signals to modulate cellular m 6 A RNA methylation. Our data indicate that downregulation of Alkbh5 and Fto may account for the elevated m 6 A methylation in murine intestinal epithelial cells in response to C. parvum infection. Interestingly, downregulation of the Alkbh5 gene involves the activation of the NF-кB pathway in infected cells. Recruitment of NF-кB p65 subunit and enrichment of suppressive marker H3K9me3 to the promoter region of the Alkbh5 gene maybe associated with its downregulation. Activation of TLR/MYD88/NF-кB pathway has previously been demonstrated in epithelial cells following C. parvum infection (25). Indeed, knockdown of MyD88 blocked the downregulation of the Alkbh5 gene in infected cells. Moreover, since the TLR/MyD88/NF-кB pathway can be activated following infection by many pathogens, it is plausible that regulation of m 6 A methylation through activation of TLR/MyD88/NF-кB signaling may be a general cellular response to microbial infection. Similarly, downregulation of Alkbh5 was previously reported in epithelial cells in response to infection by Streptococcus suis (51), H1N1 influenza virus (52), and Chlamydia pneumoniae (53). An increase of global m 6 A methylation was also found in epithelial cells following infection by SARS-CoV-2 virus (54) or Kaposi's sarcoma-associated herpesvirus (55), and in immune cells by various pathogens (56). Moreover, we observed both elevated m 6 A methylation and lost m 6 A peaks at the UTR and CDS regions of target genes. The motif usage changes to these regions seem to occur on the overall level in cells following C. parvum infection. This suggests that not only the erasers but also the writers, including Mettl3 and Mettl14, may be involved in the regulation of m 6 A methylation in cells following C. parvum infection.
RNA m 6 A methylation regulates RNA splicing, translocation, stability, and translation into protein (3)(4)(5)(6). These genes with changed m 6 A peaks identified in C. parvum-infected cells cover a  (62,63). Therefore, C. parvum infection might affect gene translation, alternative splicing, and mRNA stability, as a consequence of differential deposition of m 6 A methylation. Particularly, the effects of m 6 A methylations on RNA stability would directly affect the expression levels of target RNAs (4). Interestingly, comparison of genes whose expression levels were altered and genes with an altered m 6 A levels in infected host cells revealed that only a small portion of the genes was overlaid. The majority genes with increased or decreased m 6 A levels did not show a significant change in their expression levels in cells following C. parvum infection. This clearly indicates that modulation of RNA stability may be one of the many mechanisms that m 6 A methylation can regulate cellular function.
Other mechanisms may involve RNA splicing and translation associated with m 6 A methylation of target mRNAs.
Another key finding of this study is the observation that elevated m 6 A methylation promotes intestinal epithelial innate defense against C. parvum infection both in mice and in humans. Manipulation of Alkbh5 expression levels through the CASPR/ Cas9 knock-out and knock-in approach caused reciprocal alterations in global m 6 A mRNA methylation in host cells, and consequently, infection dynamics of C. parvum in vitro. It is unclear why an increase of infection burden was not detected in cells constitutively expressing Fto, whereas a decreased infection burden was detected in cells deficient in Fto. Since the parasite attachment/invasion of host cells appears not affected by the genomic manipulation of Alkbh5 or Fto, it is plausible to speculate that m 6 A methylation may regulate innate intestinal epithelial anti-C. parvum defense. In this study, we did not include analysis of m6A methylation in the C. parvum RNA transcriptome, which may also undergo specific m 6 A methylations and thus, modulates host-parasite interactions. Multiple m 6 A methylation sites have been identified in the viral RNA genome and transcripts of DNA viruses in recent years (64). Several families in nonsegmented negative-sense RNA viruses acquire m 6 A in viral RNA as a common strategy to evade host innate immunity (65).
We identified Irgm2 and Igtp, two immunity-related GTPase genes whose expression levels were induced with a decreased RNA m 6 A methylation in C. parvum-infected murine intestinal epithelial cells. Both Irgm2 and Igtp proteins are members of the immunityrelated GTPases, a family of large, interferon-inducible GTPases implicated in resistance against a wide variety of intracellular pathogens, including Toxoplasma gondii, Leishmania major, Trypanosoma cruzi, Chlamydia trachomatis, C. psittaci, Mycobacterium tuberculosis, M. avium, Salmonella typhimurium, Listeria monocytogenes, and Legionella pneumophila (66)(67)(68)(69)(70)(71). Igtp/ Irgm3 knockout mice are significantly more susceptible to T. gondii Xia et al. m 6 A Modulates Anti-C. parvum Defense infection than their wild-type counterparts (67). Whereas little is known regarding potential functions of this family in the infection of extracellular pathogens, GTPase family members seem to have essential and pathogen-specific roles in resistance to infections (72). Irgm2 may play a role in the innate immune response by regulating autophagy formation in response to intracellular pathogens (70). In addition, increasing evidence supports that GTPase family represents a new IFN-g-dependent, nitric oxide synthase 2-independent pathway in the control of pathogen infection (68,73). Our data suggest that induction of Irgm2 and Igtp may also be associated with their m 6 A methylation in intestinal epithelial cells in response to C. parvum infection. Given the key role of the TLR/MyD88/NF-кB signal in epithelial innate antimicrobial defense (43), coupled with the modulation of m 6 A methylation through TLR/MyD88/NF-кB signaling in C. parvum-infected cells, our data implicate a new mechanism by which TLR/MyD88/NF-кB signaling coordinates intestinal epithelial antimicrobial defense. In addition, both Irgm2 and Igtp are critical modulators for IFN signaling (70). Their induction in intestinal epithelial cells following infection may provide a new cross-link for the network between m 6 A RNA methylation, TLR/MyD88/NF-кB and IFN signaling to modulate intestinal epithelial against C. parvum, relevant to fine regulation of epithelial antimicrobial defense in general.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

ETHICS STATEMENT
The animal study was reviewed and approved by Creighton University IACUC Committee.

AUTHOR CONTRIBUTIONS
ZX, GL, and X-MC designed experiments and wrote the manuscript. ZX, JX, WH, SD, A-YG performed experiments. ZX, JX, EL, JS-S, GM, GL, and X-MC performed data analysis. A-YG and X-MC directed and supervised the study. All authors contributed to the article and approved the submitted version.

FUNDING
This work was supported by funding from the National Institutes of Health (AI116323, AI136877, AI141325, and AI156370 to X-MC). The project described was also supported by Grant Number G20RR024001 from the National Center for Research Resources. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

ACKNOWLEDGMENTS
We thank Ms. Barbara L. Bittner (Creighton) for her assistance in writing the manuscript.