Emerging role of RNA m6A modification in aging regulation

To date, more than 170 types of chemical RNA modifications have been identified, among which N⁶-methyladenosine (m⁶A) is the most abundant posttranscriptional modification in messenger RNA (mRNA) (Huang et al. 2020a). m⁶A is deposited co-transcriptionally by a methyltransferase complex (known as “writers”) mainly composed of METTL3, METTL14 and WTAP, and removed by “erasers” such as FTO and ALKBH5. Recognition of m⁶A is executed by m⁶A-binding proteins called “readers”, such as members of the YTH family and IGF2 binding proteins (Huang et al. 2020a; Zhao et al. 2017; Deng et al. 2018). m⁶A has been implicated in many aspects of mRNA metabolism, including alternative splicing, nuclear export, stability, and translation, thus playing dynamic roles in reversible regulation of a variety of biological processes ranging from embryonic development to tumorigenesis (Zhao et al. 2017; Deng et al. 2018; Roignant and Soller 2017; Niu et al. 2013; Fu et al. 2014; Zaccara et al. 2019). More recently, emerging evidence has identified the m⁶A epitranscriptome in the process of aging or cellular senescence.

To our best knowledge, Li et al. were the first to report m⁶A regulation in cellular senescence in 2017, with data demonstrating that METTL3/METTL14-mediated m⁶A methylation promoted translation efficiency of cyclin-dependent kinase inhibitor (CKI) P21 (Fig. 1) during H₂O₂-induced cellular senescence of TP53-deficient human colon carcinoma cells (HCT116) (Li et al. 2017). In the same year, Lewinska et al. found that low-dose treatment with sulforaphane (SFN) led to upregulation of P53 and P21, concomitant with increased cellular senescence in three types of human breast cancer cells (MCF-7, MDA-MB-231, SK-BR-3). In this study, global reduction in RNA m⁶A abundance was detected, suggesting a potential involvement of m⁶A in regulation of SFN-induced cancer cell senescence (Fig. 1), while the specific mechanism remains unclear (Lewinska et al. 2017). These studies identified m⁶A-mediated senescence regulation of cancer cells as a mechanism with potential for developing therapeutic strategies.

Apart from the regulatory role of m⁶A in cancer cell senescence, several recent studies implicate m⁶A and its enzymes in the aging process of human blood cells, fibroblasts, and stem cells. In 2018, Min et al. observed that, relative to a young population, m⁶A sites in human peripheral blood mononuclear cells (PBMCs) were overall declined in the aged cohorts. Using a replicatively senescent human fibroblast cell model (IMR-90), the authors detected a decrease in both AGO2 mRNA methylation and expression levels, recapitulating the observations in aged PBMCs. Similarly, knockdown of METTL3 or METTL14 accelerated the senescence of human fibroblasts, whereas overexpression of METTL3 or METTL14 promoted the stability of AGO2 mRNA (Fig. 1), thus implicating that m⁶A-dependent AGO2 mRNA decay contributes to cellular senescence by affecting the abundance of target miRNAs (Min et al. 2018). In 2020, Wu et al. described a geroprotective role for METTL3 and m⁶A in counteracting cellular senescence through stabilizing the mRNA of the cell cycle regulator MIS12 (Fig. 1) using prematurely senescent human mesenchymal stem cells (MSCs) for modeling two progeroid syndromes, Hutchinson-Gilford progeria syndrome (HGPS) and Werner syndrome (WS). They first found diminished levels of both global m⁶A modification and METTL3 expression in prematurely senescent human MSCs. Subsequently, they proved that METTL3 depletion led to declined m⁶A modification and MIS12 expression as well as accelerated senescence in human MSCs. Conversely, overexpression of METTL3 rescued m⁶A loss and prevented premature senescence in progeroid human MSCs. Finally, the authors demonstrated that the m⁶A reader IGF2BP2 bound and stabilized m⁶A-modified MIS12 mRNA to prevent human MSCs from premature senescence (Fig. 1), thus providing mechanistic insights into the epitranscriptomic regulation of premature senescence in human stem cells (Wu et al. 2020). Most recently, Zhang et al. revealed an m⁶A-independent role of FTO in stabilizing MIS12 protein and counteracting human MSC senescence, further supporting the importance of MIS12 in antagonizing stem cell aging (Zhang et al. 2022). In related work in human dental pulp stem cells (DPSCs), Luo et al. reported that knockdown of METTL3 resulted in disrupted cell cycle and accelerated senescence via upregulating the expression of PLK1, another critical cell cycle modulator, in m⁶A-dependent manners (Fig. 1) (Luo et al. 2021). This study together with the work by Wu et al. suggests an indispensable role of METTL3 and m⁶A in counteracting human stem cell senescence via regulation of cell cycle progression. Supporting the geroprotective function of m⁶A and its core enzymes, Zhang et al. found that nuclear lamina protein Lamin A interacted with and safeguarded the proper localization of METTL14 in nuclear speckles to prevent cellular senescence in human and murine fibroblasts (Fig. 1). The authors demonstrated that Lamin A interacted with and stabilized both METTL3 and METTL14, and that knockdown of METTL3 or METTL14 accelerated senescence while overexpression of METTL14 alleviated replicative senescence in normal human fibroblasts and rescued premature senescence in progeroid murine fibroblasts. To be noted, METTL14 overexpression in senescent human fibroblasts also attenuated nuclear envelope abnormalities and restored the expression of H3K9me3 (Fig. 1), a well-known heterochromatin marker (Zhang et al. 2020). In addition, Mapperley et al. identified the role of m⁶A reader YTHDF2 in repression of proinflammatory pathways and long-term maintenance of hematopoietic stem cell (HSC) upon aging (Mapperley et al. 2021).

In other types of cells or tissues, namely in aged human follicular fluid and granulosa cells or during ovarian aging of mouse, m⁶A abundance was instead found to be increased due to declined expression of the m⁶A eraser FTO (Sun et al. 2021). Here, downregulation of FTO increased overall m⁶A levels and stability of FOS mRNA (Fig. 1), thus contributing to the senescence of ovarian granulosa cells (Jiang et al. 2021). In human nucleus pulposus cells (NPCs), Zhu et al. reported that knockdown of METTL14 alleviated TNF-α-induced cellular senescence. They revealed that METTL14 interacted with DGCR8, and promoted the production of miR-34a-5p in an m⁶A-dependent manner (Fig. 1), which subsequently accelerated the senescence of human NPCs by targeting SIRT1 (Zhu et al. 2021). In oncogenic RAS-induced senescence of human fibroblasts, Liu et al. revealed that genomic redistribution of METTL3 and METTL14 mediated the promoter-enhancer (P-E) loop formation of senescence-associated secretory phenotype (SASP)-associated genes (Fig. 1), thus promoting their transcriptional activation via an m⁶A-independent way. The authors also proved that knockdown of METTL3 or METTL14 reduced the expression of SASP genes in RAS- or etoposide-induced senescent fibroblasts (Liu et al. 2021).

The m⁶A modification has also been implicated in age-associated degenerative diseases. For instance, in rat models of Parkinson’s disease, Chen et al. observed globally reduced m⁶A modification and increased dopaminergic neuronal death, which was potentially mediated by the upregulation of N-methyl-d-aspartate receptor 1 (NMDAR1), oxidative stress, and Ca²⁺ influx (Chen et al. 2019). In mouse model of Alzheimer’s disease (AD), the findings have been varied; Han et al. reported an increase in overall m⁶A abundance, accompanied by increased METTL3 expression and decreased FTO expression (Han et al. 2020), whereas Shafik et al. detected a significant decrease in m⁶A levels in AD mice compared to wildtype controls, concomitant with FTO upregulation and METTL3 downregulation (Shafik et al. 2021). What causes the differences between these two studies remains to be further investigated. Notably, dysregulated expression of METTL3 and IGF2BP2 in AD patients was also reported to be potentially associated with the occurrence of AD (Huang et al. 2020b; Deng et al. 2021). In the visual system, Li et al. dissected the m⁶A modification in age-related cataracts. In patients with age-related cataracts, circRNA m⁶A abundance was globally reduced and ALKBH5 upregulated in lens epithelium cells relative to control subjects, thus providing potential molecular targets to develop intervention strategies for the treatment of ophthalmic degeneration diseases (Li et al. 2020).

In summary, accumulating studies point to a role for m⁶A regulation in cellular senescence or aging-related processes, but the use of specific cell lines and rodent models and focus on phenotypic analysis prevents making strong conclusions about roles and mechanisms of the m⁶A regulatory network and potential clinical applications. Furthermore, considerable inconsistency exists in the conclusions from different studies on m⁶A-mediated senescence or aging regulation, likely due to distinct research contexts, implying species-, tissue-, cell type-, or substrate-specific manners. Therefore, development of suitable animal models (e.g., non-human primates) for studying context-specific aging-related m⁶A regulations and associated in-depth mechanisms for these physiological and pathological process is needed. Such models would allow investigations of cellular senescence in vivo at the tissue/organ level, and if supplemented by human stem cell-based platforms, could significantly advance our understanding at the cellular level, and ultimately contribute to m⁶A-based therapeutic advances for age-related diseases.