Among the diverse layers of epigenetic regulation, modifications of DNA and proteins have been explored in depth; however, RNA modification is a relatively new field (He, 2010). More than 100 posttranscriptional covalent modifications have been identified on RNA transcripts from various organisms, including viruses, yeast, and mammals (Meyer et al., 2012; Mamadou et al., 2013; Tan and Gao, 2018). The most prevalent modification of internal messenger RNA in eukaryotes and in nuclear-replicating viruses is the addition of a methyl group to the N⁶ position of adenosine, known as m⁶A (Dominissini et al., 2012; Meyer and Jaffrey, 2014). In 1975, m⁶A was first reported to exist in cellular mRNAs, and it was reported that there are three internal m⁶A residues on the average ∼2.2 kb cellular transcript (Desrosiers et al., 1975; Kennedy et al., 2017). Furthermore, highly decorated and regulated transcripts might include ten or more m⁶A groups (Linder et al., 2015). Similar to DNA methylation, the m⁶A modification is reversibly catalyzed by corresponding enzymes (Li et al., 2018). Despite the discovery of m⁶A decades ago (Desrosiers et al., 1974; Adams and Cory, 1975; Wei et al., 1975), the related signaling pathways it is involved in and the biological roles it plays were not fully described until recently (Nilsen, 2014; Cao et al., 2016). Meanwhile, with the development of efficient methods of m⁶A detection and subsequent analysis, increasing biological functions are being elucidated (Meyer et al., 2012; Li et al., 2015). m⁶A has been reported to control the fate of modified RNAs at multiple steps, including RNA splicing (Dominissini et al., 2012), mRNA stability (Li and Mason, 2014), cap-independent translation (Meyer et al., 2015), and miRNA biogenesis (Alarcon et al., 2015b). m⁶A-decorated RNAs participate in many biological processes, such as stress responses (Zhou et al., 2015), cellular reprogramming (Chen T. et al., 2015), circadian cycle (Fustin et al., 2013), stem cell differentiation (Yue et al., 2015), fertility (Zhao and He, 2017), and cancer (Pan et al., 2018).

Viruses, as a type of organism, were found to be modified by m⁶A in their genomic RNAs (Zhou et al., 2010; Courtney et al., 2017). Earlier studies have shown that some viruses, such as simian virus 40 (Lavi and Shatkin, 1975), influenza A virus (Krug et al., 1976), adenovirus (Sommer et al., 1976), avian sarcoma virus (Dimock and Stoltzfus, 1977), and RSV (Kane and Beemon, 1985) have m⁶A residues in their mRNAs. Narayan and colleagues performed the biochemical analysis of different influenza virus mRNAs, to detect the distribution of m⁶A and showed that the amount of m⁶A in different hemagglutinin (HA) mRNAs varied (Narayan et al., 1987). Those authors attempted to explore the functions of m⁶A in RNA splicing and translation, but the experimental conditions and relevant knowledge of m⁶A at that time were not enough. The specific sites of m⁶A were not mapped, and the functions of m⁶A in IAV remained vague for decades since the restriction with techniques and knowledge of the methyltransferases, demethylases, and m⁶A reader proteins then. Transcriptome-wide mapping of m⁶A was available after N⁶-methyladenosine-sequencing (m⁶A-seq) was developed by two independent research teams in Dominissini et al. (2012) and Meyer et al. (2012). In the following years, m⁶A has attracted the attention of scientists to elucidate the role it plays in viral epitranscriptomics (Gokhale and Horner, 2017; Kennedy et al., 2017). Immediately following m⁶A-seq, new technologies to unravel the m⁶A epitransctiptome were developed, including PA-m⁶A-seq, miCLIP, m⁶A-LAIC-seq, microarray, and SELECT (Chen K. et al., 2015; Li et al., 2015; Linder et al., 2015; Molinie et al., 2016; Xiao et al., 2018). The advantages and disadvantages of these techniques were well summarized in a review by Tan and Gao (2018). Based on these sequencing data, some bioinformatics tools were developed to help researchers investigate the potential functions and mechanisms of m⁶A modification. At present, most of the m⁶A decoration profiles have been available in some databases. A brief description of these databases related to m⁶A research are summarized in Table 1. Moreover, some computational m⁶A site predictors have been constructed, such as SRAMP (Zhou et al., 2016), m6Apred (Chen W. et al., 2015) and RNA-MethylPred (Jia et al., 2016). Although few studies have been done using these bioinformatics tools on the role of m⁶A in viral infection currently, they would be helpful for m⁶A prediction prior to experimental study. With the help of all these advanced technologies, a number virus types were subjected to transcriptome-wide mapping of m⁶A. Gokhale et al. (2016) mapped the m⁶A sites within the viral RNA genomes of the flaviviridae family of viruses, including HCV, ZIKV, DENV, YFV, and WNV. These researchers verified that m⁶A negatively regulated HCV infection by using m⁶A-abrogating mutations in HCV E1 (Gokhale et al., 2016). In addition, other viruses, such as influenza A virus (IAV) (Courtney et al., 2017), Kaposi’s Sarcoma-Associated Herpesvirus (KSHV) (Ye et al., 2017), Human Immunodeficiency Virus-1 (HIV-1) (Tirumuru et al., 2016), Simian Virus 40 (SV40) (Tsai et al., 2018), Hepatitis B Virus (HBV) (Imam et al., 2018), and Enterovirus 71 (EV71) (Hao et al., 2019) have been subjected to transcriptome-wide mapping of m⁶A and reported for their different roles in the life cycles of the viruses.

Our review summarizes the recent advances in related literature in this relatively new emerging field and discusses the different functions and relevant mechanisms of m⁶A in the biological processes of different types of viral infections.

The associated mechanisms that regulate m⁶A involve proteins working as writers, erasers, readers, and anti-readers (Edupuganti et al., 2017; Li et al., 2018). M⁶A decoration is not randomly distributed in RNA transcripts (Batista, 2017), which was found to occur on the consensus RNA motif of RRACH (R = A or G;H = A, U, or C), and preferentially center on specific transcript landmarks such as near 3′ untranslated regions (3′ UTRs) and stop codons or in long exons (Yang et al., 2018). m⁶A is added cotranscriptionally to nuclear pre-mRNAs by a multicomponent protein complex consisting of catalytic subunit Methyltransferase Like 3 (METTL3) (Yao et al., 2018), RNA-binding platform METTL14 (Kobayashi et al., 2018), cofactors Wilms tumor 1-associated protein (WTAP) and KIAA1429 (Schwartz et al., 2014; Scholler et al., 2018), two novel subunits of the methyltransferase complex RBM15 (Patil et al., 2016), and zinc finger CCCH domain-containing protein 13 (ZC3H13) (Guo et al., 2018; Knuckles et al., 2018; Wen et al., 2018). As the core methyltransferase subunit, METTL3 is a strongly conserved protein (Bujnicki et al., 2002; Yang et al., 2018) and has been demonstrated to selectively methylate the GAC or AAC motifs in synthetic single-stranded RNA in vitro (Rottman et al., 1994; Kan et al., 2017). METTL14 is also highly conserved in mammals and can form a stable protein heterodimer with METTL3 (Liu et al., 2014; Ping et al., 2014; Wang Y. et al., 2014). Wang et al. studied the crystal structure of Mettl3-Mettl14 complex and formulated that Mettl14 has a degenerate active site and is unavailable for catalysis (Wang et al., 2016). Mettl14 has a structural role that could offer an RNA-binding scaffold, allosterically activating and enhancing Mettl3’s catalytic function (Sledz and Jinek, 2016). As a regulatory subunit of the m⁶A methyltransferase complex, WTAP allows METTL3/METTL14 to interact with messenger RNAs in the nucleus to improve m⁶A modification efficiency (Liu et al., 2014; Ping et al., 2014). WTAP is required for the localization of the METTL3-METTL14 complex into nuclear speckles that are enriched with various precursor messenger RNA (pre-mRNA) processing factors (Ping et al., 2014). The depletion of KIAA1429 in human A549 cells results in a fourfold reduction of m⁶A abundance (Schwartz et al., 2014), suggesting a significant regulatory effect in the writer complex. Recently, RBM15 and its paralog RBM15B were shown to be members of the m⁶A methyltransferase complex that recruit the METTL3/14 protein complex to specific sites in RNA for the selective methylation (Patil et al., 2016). More recently, three research teams demonstrated that ZC3H13 was another member of the m⁶A writer complex and modulated m⁶A methylation (Guo et al., 2018; Knuckles et al., 2018; Wen et al., 2018). METTL16 is cognate with METTL3, was reported to control the cellular SAM level and catalyze the m⁶A group onto the U6 small nuclear RNA (Pendleton et al., 2017). More subunits of the m⁶A methyltransferase complex might be explored to achieve accurate posttranscriptional RNA regulation through selectively recognizing candidate m⁶A sites.

The removal of m⁶A from the decorated mRNA is catalyzed by the demethylase FTO or a-ketoglutarate-dependent dioxygenase AlkB homolog 5 (ALKBH5) (Jia et al., 2011; Zheng et al., 2013). Jia et al. (2011) found that FTO could efficiently demethylate m⁶A in RNA in vitro. They used siRNA mediated knockdown of FTO and then detected an increased level of m⁶A in mRNA, whereas the overexpression of FTO resulted in a decreased level of m⁶A in human HeLa cells. In 2013, another m⁶A demethylase, ALKBH5 was identified and it exhibited m⁶A demethylation efficiency comparable to that of FTO (Zheng et al., 2013). A recent report showed that ALKBH5-mediated m⁶A elimination in the nucleus of spermatocytes and round spermatids was elementary for correct splicing and the production of longer 3′-UTR mRNAs (Tang et al., 2018).

The function of m⁶A modification on target mRNAs is thought to be mediated by “reader” proteins (Riquelme-Barrios et al., 2018). To date, several m⁶A reader proteins have been identified in mammalian cellular extracts using affinity chromatography combined with mass spectrometry (Dominissini et al., 2012). Among the “reader” proteins, the YT521-B homology (YTH) family of proteins have been very well studied (Meyer and Jaffrey, 2017; Patil et al., 2018). There are three YTHDF (YTH domain family) members localized in the cytoplasm, including YTHDF1, YTHDF2, and YTHDF3 (Dominissini et al., 2012; Wang X. et al., 2014; Wang X. et al., 2015; Shi et al., 2017), and two YTHDC (YTH domain containing) proteins, YTHDC1 located in the nucleus (Xu et al., 2014) and YTHDC2 located in the cytoplasm (Morohashi et al., 2011). YTHDF proteins contain a conserved YTH RNA-binding domain that is inclined to bind the m⁶A-targeted RNAs and a N/P/Q-rich region that is correlated with different RNA-protein complexes (Fu et al., 2014). YTHDF1 has been demonstrated to interact with the translation initiation machinery and enhance the translational efficiency of its mRNA targets (Wang X. et al., 2015). The function of YTHDF2 was the induction of mRNA degradation (Wang X. et al., 2014). It has been shown that YTHDF2 accelerates the degradation of m⁶A-containing RNAs by directly recruiting the CCR4-NOT deadenylase complex (Du et al., 2016). YTHDF3 has been shown to promote the function of both YTHDF1 and YTHDF2. When it cooperates with YTHDF1, YTHDF3 can favor mRNA translation by the interactions with some ribosomal proteins (Li et al., 2017; Shi et al., 2017). Moreover, when it is associated with YTHDF2, YTHDF3 could participate in mRNA decay through direct relations with YTHDF2 (Shi et al., 2017). The nuclear m⁶A reader YTHDC1 was demonstrated to promote exon inclusion by assisting the splicing factor SRSF3 during its recruitment, while blocking the binding of SRSF10 (Kasowitz et al., 2018). Another study in HeLa cells showed that the association of YTHDC1 with SRSF3 and NXF1 could promote the nuclear export of m⁶A-target mRNAs (Roundtree et al., 2017b). YTHDC2, was identified to favor translational efficiency while decreasing the abundance of m⁶A-containing mRNAs (Hsu et al., 2017). YTHDC2 is a relatively large protein molecule (∼160 kDa) and contains many helicase domains and two ankyrin repeats. These special structural features might allow YTHDC2 to possess multiple functions, including regulating effects on RNA binding and RNA structure, and binding with or the recruitment of other interacting proteins (Tanabe et al., 2016; Hsu et al., 2017). A study reported that YTHDC2 lacked the m⁶A binding activity in HEK cells (Patil et al., 2016, 2018). This finding suggests that YTHDC2 might have indirect effects in regulating m⁶A-modified RNAs through interaction with other factors. In addition to the YTH family members, there are other proteins that have been identified to recognize and bind to m⁶A. The eukaryotic initiation factor 3 (eIF3) complex, interacts with m⁶A-containing 5′UTR through a multisubunit interface to directly recruit the 40S preinitiation complex to the 5′UTR of target-mRNAs to stimulate translation initiation (Meyer et al., 2015). hnRNPA2/B1 was shown to bind m⁶A-modified RNAs to regulate splicing and microRNA maturation (Alarcon et al., 2015a). Fragile X Mental Retardation Protein (FMRP) was also demonstrated to bind m⁶A-containing transcripts in a sequence context-dependent manner (Edupuganti et al., 2017). Recently, insulin growth factor 2 binding proteins, IGF2BP 1, 2, and 3, were shown to promote stability and storage of mRNAs through binding target m⁶A mRNAs under normal and stress conditions (Huang et al., 2018).

Furthermore, researchers have found a new m⁶A-regulated protein, referred to as m⁶A-repelled proteins or anti-readers, during the study of proteomics that interact with m⁶A. These researchers found that m⁶A disrupts RNA binding by the stress granule proteins G3BP1/2, CAPRIN1, USP10 and RBM42 (Arguello et al., 2017). A study has confirmed that the presence of m⁶A in mRNA could decrease mRNA stability by interfering with the binding of G3BP1 and G3BP2 (Edupuganti et al., 2017). This phenomenon reveals an additional function of m⁶A in RNA metabolism. We provide a picture (Figure 1) as an overview of the associated machinery and molecular functions of m⁶A.

The specific functions and mechanisms of m⁶A in RNA viruses are not well-known, although m⁶A has long been known to be present in some RNA viruses, such as IAV (Krug et al., 1976), RSV (Kane and Beemon, 1985), B77 avian sarcoma virus (Stoltzfus and Dimock, 1976; Dimock and Stoltzfus, 1977) and feline leukemia virus (Thomason et al., 1976). In recent years, technological advances have made m⁶A a focus of research in elucidating the role of this RNA modification in viral epitranscriptomics. Herein we summarize and describe the role of m⁶A in RNA viruses including IAV, HIV-1, HCV, ZIKV, VSV, and EV71.

As the first virus found to express mRNAs bearing internal m⁶A groups, Influenza A virus (IAV) was mapped for the sites of m⁶A by Courtney et al. in both the IAV mRNA and vRNA strands, and it was demonstrated that m⁶A modification increases viral RNA expression in cis (Courtney et al., 2017). Those authors used two methods to inhibit m⁶A modification in A549 cells infected by IAV. One method used a non-toxic dose of DAA treatment (an inhibitor of m⁶A addition), and the other used knockout of METTL3 through gene editing with CRISPR/Cas. The results showed that both methods could reduce the expression of the IAV proteins NS1 and M2. However, when the m⁶A reader YTHDF2 was ectopically overexpressed, increased IAV replication and infectious particle production were found. Moreover, they used synonymous mutations to remove m⁶A on both strands of the hemagglutinin (HA) segment, and found that IAV HA m⁶A mutants revealed reduced pathogenicity in mice. The findings confirmed that the addition of m⁶A residues in IAV transcripts could enhance viral gene expression.

Prior to this review, four reports have studied the involvement of m⁶A in HIV-1 infection. Three excellent reviews comprehensively compared and summarized the functions and mechanisms of m⁶A modification in the HIV-1 life cycle (Gonzales-van Horn and Sarnow, 2017; Riquelme-Barrios et al., 2018; Tan and Gao, 2018). Despite some agreed-upon conclusions, there are unconformities in the locations, effects, and mechanisms of m⁶A in HIV-1 RNA in these studies. Herein, a table was created to allow readers to understand functions of m⁶A more clearly and intuitively in the four different articles (Table 2).

RNA-based regulation of HCV plays an essential role in its infection (Wang X. et al., 2012), for example the liver-specific microRNA 122 has been proved to facilitate replication of the viral RNA (Jopling et al., 2005; Cuypers et al., 2016). Recently, Gokhale et al. carried out m⁶A analyses in cells infected with HCV and demonstrated that the HCV RNA genome was modified by m⁶A (Gokhale et al., 2016). The abundance of HCV NS5A protein was significantly increased through siRNA mediated simultaneous depletion of m⁶A methyltransferases METTL3 and METTL14 in Huh7 cells infected with HCV. In contrast, HCV NS5A levels were decreased when the m⁶A demethylase FTO was depleted. Considering that CD81 is an essential entry factor but has no effects on other steps of HCV’S life cycle (Zhang et al., 2004), Huh7.5 CD81 knockout (KO) cells were used to test whether the m⁶A machinery affects viral RNA replication or viral particle production. The results indicated that the m⁶A machinery modulated HCV particle production but did not regulate HCV translation or RNA replication. Furthermore, the authors found that YTHDF proteins relocalized at viral assembly sites to retain HCV RNA with m⁶A, leading to reduced HCV particle production. To further confirm this conclusion, mutations in the E1 gene of HCV to inactivate m⁶A modification were constructed marked as HCV-E1^mut. Like the above findings, these mutations increased viral particle assembly. In brief, this article revealed that m⁶A, as a conserved regulatory mark across the HCV genome, negatively regulates HCV infection.

Like HCV, ZIKV is also a member of the Flaviviridae family. To date, there has been only one research article that reports the relevance of ZIKV to m⁶A modification. Twelve discrete m⁶A peaks were identified spanning the full length of ZIKV RNA, among which half were present in the region encoding the NS5 protein and the 3′ UTR region (Lichinchi et al., 2016b). Like the role of m⁶A in regulating the HCV life cycle, the knockdown of host methyltransferases METTL3 and METTL14 increased ZIKV production while the silencing of demethylases ALKBH5 and FTO suppressed ZIKV production. The difference was that the m⁶A machinery decreased HCV production by affecting the assembly of virus, yet m⁶A modification inhibited ZIKV production by impacting replication of ZIKV. Furthermore, YTHDF family proteins bound to ZIKV RNA and decreased the viral titer. Remarkably, YTHDF2 showed the strongest effect compared with YTHDF1 and YTHDF3. In total, these results identified that the m⁶A methylome was a new mechanism by which ZIKV interacted with the host cells and inhibited viral infection.

Because of its relatively simple structure and high replication capacity, VSV has been widely used in the study of innate immunity. After VSV infection, pattern-recognition receptors trigger activation of the type I interferon signaling pathway (Schlee and Hartmann, 2016). Since m⁶A is involved in regulating the life cycles of many viruses, whether m⁶A modification participates in the regulation of VSV infection is a question. The Cao group demonstrated in 2017 that knockdown of the m⁶A eraser ALKBH5 noticeably increased the production of type I interferons triggered by VSV infection (Zheng et al., 2017). This finding suggested that m⁶A modifications might play a negative role in the life cycle of VSV. To investigate the detailed mechanisms related to m⁶A in regulating the antiviral effects, the authors performed numerous biological experiments. They found that nuclear DDX member DDX46 recruited ALKBH5 via DDX46’s DEAD helicase domain to erase the m⁶A modification of three important antiviral transcripts MAVS, TRAF3, and TRAF6 and therefore caused their retention in the nucleus and reduced their translation into protein; this result further caused decreased production of type I interferons. In effect, elimination of the m⁶A modification suppressed the exportation of mRNA from the nucleus and induced retention of the mRNA encoding proteins in the nucleus such as those involved in circadian rhythm (Fustin et al., 2013; Zheng et al., 2013); this phenomenon was well verified in this article. This work might promote research progress in m⁶A-modification-based regulation of gene expression in innate immunity.

As a single-stranded RNA virus which belongs to the genus Enterovirus within the family Picornaviridae, Enterovirus 71 (EV71) is one of the main pathogens that causes hand-foot-and-mouth disease (HFMD) (Wang and Li, 2018). Lately, a study reported that EV71 RNA contained m⁶A residues and the expression and localization of the m⁶A methyltransferase and demethylase were altered during viral infection (Hao et al., 2019). The genomic copy numbers of EV71 RNA were significantly decreased by silencing METTL3 gene and increased by FTO gene depletion. When two m⁶A sites were mutated in the EV71 RNA, the virus titer was significantly decreased. Besides, the authors found that METTL3 interacted with viral RdRp 3D protein and overexpression of METTL3 induced enhanced sumoylation and ubiquitination of 3D. As it has been reported that sumoylation and ubiquitination levels can enhance self-stability of 3D and facilitate EV71 replication (Liu et al., 2016), this suggested us that the replication of EV71 were also influenced through METTL3-mediated sumoylation and ubiquitination of viral 3D protein. Collectively, these results suggest that m⁶A modifications in EV71 RNA played a positive role in viral replication.

DNA viruses use the host machinery to replicate in the nucleus, are likely to usurp m⁶A machinery to regulate their lifecycle, and have been reported to have m⁶A modifications in viral mRNAs such as Kaposi sarcoma–associated herpesvirus (KSHV), Simian virus 40 (SV40) and Hepatitis B virus (HBV) (Ye et al., 2017; Hesser et al., 2018; Imam et al., 2018; Tsai et al., 2018).

Kaposi’s sarcoma-associated herpesvirus (KSHV) is a carcinogenic virus associated with a variety of malignant tumors including Kaposi’s sarcoma (KS), primary effusion lymphoma (PEL), and multicentric Castleman’s disease (MCD) (Chang et al., 1994; Cesarman et al., 1995; Soulier et al., 1995). Like all herpesviruses, KSHV has two phases of the lifecycle, latent infection and lytic replication (Zhao et al., 2015; Liu et al., 2017). During KSHV latent infection, most of the viral genome is suppressed through DNA methylation, repressive histone modifications and other regulatory mechanisms which negatively regulate gene expression (Pantry and Medveczky, 2009; Gunther and Grundhoff, 2010; Lu et al., 2010; Toth et al., 2010; Purushothaman et al., 2015). Changes in the host cell microenvironment could reactivate the virus from latency to the lytic cycle, wherein the inhibitory epigenetic marks are replaced by active ones to allow for transcription of viral lytic genes (Ye, 2017; Uppal et al., 2018). Three recent studies revealed that the KSHV life cycle was affected by RNA N⁶-adenosine methylation epigenetic modification (Ye et al., 2017; Hesser et al., 2018; Tan et al., 2018). In the first report, the authors found that most KSHV transcripts undergo m⁶A modifications by experiments of MeRIP combined with qRT-PCR (Ye et al., 2017). Knockdown (KD) or functional inhibition of FTO using a reagent, named meclofenamic acid (MA), could enhance the expression of lytic gene ORF50 and ORF57, while KD of METTL3 or the blocking of m⁶A could abolish the expression of the two lytic genes and virion production. The splicing of pre-mRNA of ORF50, a key KSHV lytic switch gene, was inhibited when m⁶A deposition was blocked. Further research suggested that m⁶A positively regulates ORF50 (RTA) RNA splicing through binding of the YTH domain containing 1 (YTHDC1) to identified m⁶A sites in RTA pre-mRNA and the cooperation with serine/arginine-rich splicing factor 3 (SRSF3) and SRSF10. RTA itself could induce m⁶A decoration and enhance its own pre-mRNA splicing. In short, m⁶A marks in the KSHV genome promoted lytic replication. In the second report, the KSHV-positive renal carcinoma cell line iSLK.219 during lytic reactivation went through transcriptome-wide m⁶A-sequencing, and the results revealed that the m⁶A modification was present across most viral transcripts. Depletion of the m⁶A writer METTL3 and the reader YTHDF2 significantly impaired virion production in iSLK.219 and iSLK.BAC16 cells, suggesting that the m⁶A pathway functioned in a pro-viral manner. In contrast, ORF50 protein expression was increased upon depletion of METTL3 in KSHV infected B cells, reflecting the anti-viral impacts of m⁶A in the KSHV life cycle (Hesser et al., 2018). The second report emphasized the result that the m⁶A pathway might play different roles in promoting or inhibiting viral gene expression depending on the cell-type analyzed. The last report found that KSHV transcripts contain abundant and highly conserved m⁶A modifications among different cell types and infection systems. The m⁶A reader protein YTHDF2 inhibited KSHV lytic replication by facilitating the degradation of viral lytic transcripts which is consistent with the second report. The lytic replication period of KSHV induced dynamic reprogramming of the epitranscriptome, regulating relevant pathways that control lytic replication (Tan et al., 2018). This report provided insights into the mechanism of KSHV-induced disease by helping us understand the changes of viral and cellular m⁶A modifications during KSHV latent and lytic infection. The three articles together suggest that m⁶A modifications in KSHV might play a positive role through different mechanisms or play different roles owing to different cell types during lytic replication.

As a member of the polyomavirus family, the gene expression of simian virus 40 (SV40) is regulated in an early phase, encoding the viral regulatory proteins, and a late phase, encoding the viral structural proteins. In 1979, SV40 was reported to contain some m⁶A groups in transcripts in the “late” region of the virus (Canaani et al., 1979). However, the specific location of these m⁶A groups was not identified and their functional roles have remained unclear. Recently, a report demonstrated that the overexpression of YTHDF2 induced faster viral replication, and larger viral plaques in BSC40 cells infected with SV40, whereas mutational inactivation of the endogenous YTHDF2 gene, or the m⁶A writer METTL3, had the contrary effect, suggesting a positive influence for m⁶A modification in the regulation of the SV40 life cycle (Tsai et al., 2018). The authors also mapped the sites of m⁶A residues on SV40 transcripts and identified two m⁶A sites on the viral early transcripts and eleven m⁶A sites within the SV40 late region. The authors observed that the mutant virus replicated more slowly than wild type SV40 when they inactivated most of the m⁶A addition sites on the SV40 late mRNAs using synonymous mutations. Together, these results suggest that the addition of m⁶A residues to the late SV40 transcripts played a positive role in viral gene expression and replication.

Hepatitis B virus infection is the leading cause of chronic hepatitis (Fu et al., 2016) and plays a role in the development of cirrhosis (Zhao et al., 2011) and hepatocellular carcinoma (Chen J. et al., 2009; Chen L. et al., 2009; Levrero et al., 2009). Imam et al. (2018) reported that HBV transcripts detected from both liver tissues of chronic HBV patients and HBV-expressing cells contain m⁶A. They observed that m⁶A located in HBV 3′UTRs reduced the stability of these RNAs, ultimately affecting the expression of their corresponding proteins. YTHDF proteins were found to bind HBV transcripts, and the depletion of them increased HBV protein expression, likely by changing the stability of the HBV transcript. Similar effects were observed when the m⁶A site within the 3′ epsilon loop of all HBV transcripts was inactivated. However, the authors found that m⁶A at the 5′epsilon loop, which is present only in pregenomic RNA (pgRNA), positively regulates pgRNA reverse transcription. Therefore, m⁶A regulates HBV RNAs in more than one way, hinging on its position in the RNA. The roles of m⁶A in different viral infections are summarized in Table 3.

The addition of m⁶A in viral mRNAs has both pro-viral and antiviral functions in distinct viral life cycles (He et al., 2011). The m⁶A modification enhanced the translation of viral late transcripts in SV40 while decreasing infectious HCV particle production (Gokhale et al., 2016; Tsai et al., 2018). Different reports studying the m⁶A effects in the same virus indicated that there exist discrepancies in the distribution of m⁶A sites and roles of m⁶A in the viral life cycle such as the abovementioned studies focusing on HIV-1, KSHV and HBV. The reasons for the contradictory results remain unclear in this field. One reason is the suggestion that m⁶A machinery functions in a cell type specific manner to either promote or inhibit KSHV gene expression (Hesser et al., 2018). The use of different cell types to study the m⁶A modification is one of the reasons for the different results. Some other reasons might contribute to different results obtained. First, the m⁶A decoration itself is dynamic in controlling mRNA fate and its ability to impact virtually every stage of host gene expression (Gonzales-van Horn and Sarnow, 2017; Roundtree et al., 2017a). Therefore, different experimental results are likely to result from different time points of sampling or different viral infection times. Second, the intracellular localizations of m⁶A modified by viral RNAs might influence the availability of host proteins, which might further result in different functions of m⁶A in different viruses. Third, the distinct m⁶A distribution sites of HIV-1 might be attributed to the sequencing-based methodologies used by each report which had different resolutions. For example, neither the m⁶A-seq nor the PA-m⁶A-seq can discriminate between m⁶A and m⁶Am (Tan and Gao, 2018), therefore leading to comprehensive distribution of the two types of modification when using the abovementioned m⁶A transcriptome-wide profiling techniques. Finally, the binding sites of the three m⁶A reader proteins YTHDF1, YTHDF2, and YTHDF3 identified by CLIP-seq analysis were different in the reports about HIV-1 (Ao et al., 2016; Riquelme-Barrios et al., 2018). Certain binding sites for YTHDF1, 2, and 3 do not line up with predicted m⁶A sites (Riquelme-Barrios et al., 2018), suggesting that cytoplasmic readers would also bind the HIV-1 gRNAs in an m⁶A-independent manner or that the viral transcripts possess additional m⁶A sites that have not yet been mapped. In general, the different approaches, detection strategies, cell lines and other reagents used in m⁶A-related studies might contribute to the distinct results obtained, which deserve to be clarified in the future.

Given that the presence of m⁶A participates in the life cycles of multiple viruses, drugs that target this pathway could have the potential to be used to fight a range of viral illnesses. A drug, 3-deazaadenosine (DAA), which inhibits m⁶A addition through depleting intracellular levels of the methyl donor SAM (Bader et al., 1978; Fustin et al., 2013), is a potent inhibitor of IAV replication at doses that do not exert any apparent cytotoxic effects (Fischer et al., 1990). DAA has been reported to suppress a wide variety of viruses, not only in vitro culture but also in vivo in mice and in rats (Bader et al., 1978; Wyde et al., 1990; Bray et al., 2000; Kennedy et al., 2016; Courtney et al., 2017). In addition, a reagent, known as meclofenamic acid (MA), which can inhibit FTO demethylation by competition on m⁶A-containing substrate binding, has been demonstrated to enhance the expression of KSHV lytic genes (Ye et al., 2017). It is not difficult to conclude that developing new small molecule inhibitors or drugs targeting m⁶A “writers” or “erasers” might prove to be of great significance for antiviral therapies.

All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. WD analyzed the literatures and studies and wrote the manuscript. JL designed this review and revised the manuscript. The rest of the authors assisted with the process of writing the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.