According to MODOMICS, an RNA modification database constructed by Boccaletto et al. few years ago, documented RNA modifications have now raised to 144 (Dunin-Horkawicz et al., 2006), and the upsurge continues due to improved sequencing techniques and other technological advancements. To date, discussions on RNA modifications mainly revolve around the well-characterized ones, including N⁶-methyladenosine (m⁶A), 5-methylcytosine in RNA (m⁵C), N¹-methyladenosine (m¹A), and pseudouridine (Ψ). Others like 5-hydroxymethylcytosine (5-hmC), N⁴-acetylcytidine (ac⁴C), and adenosine-to-inosine editing (A-to-I) are only registered with unknown or unspecified functions. Moreover, MODOMICS covers the related diseases and pathways (Dunin-Horkawicz et al., 2006), with sequential updates at regular intervals (Machnicka et al., 2013) (Boccaletto et al., 2018), leading to more attention directed to the rising role of RNA modifications contributing to the nuanced transcriptomic homeostasis from clinicians and scientists (Song et al., 2020).

The fate of an mRNA transcript is determined by a series of events posttranscriptionally, and one of such crucial processes is epitranscriptomic modifications. In general, the process of mRNA epitranscriptomic editing relies on three major types of RNA-modifying proteins (RMPs):

1) writers that deposit RNA modifications, for e.g., methyltransferase-like (METTL) enzyme family members, zinc finger CCCH-type containing 13 (ZC3H13), and VIRMA/KIAA1429 for m⁶A, TRMT family members for m¹A, ADARs for A-to-I editing, and NSUNs for m⁵C;

RNAWRE, which was constructed in 2020 by Nie et al. (2020) and apropos to mention, comprises more than 2000 manually curated writers, erasers, and readers. RMP regulation determines whether the previously mentioned epitranscriptomic signatures are installed, removed, or recognized. By dint of Table 1 summary and Figure 1 illustration, types of epitranscriptomic marks and their respective RMPs will not be outlined thoroughly in paragraphs. The concept of how these epitranscriptomic marks and RMP expression affect the existence and severity of ferroptosis will be discussed in later parts and illustrated in the compiled figures.

Even though the adjustments on nucleotides seem slight and minuscule, finding a way to elucidate the epitranscriptomic marks is never simple and uncomplicated. Consecutive efforts are required owing to these nanoscopic modifications down to nucleotides. This review will not focus on the in-depth discussion of epitranscriptomic mark detection, given that such an issue has already been brilliantly reviewed elsewhere (Helm & Motorin, 2017) (Sarkar et al., 2021). Nevertheless, we shall highlight the important ones, including NGS-based techniques or mass spectrometry-based techniques.

Ferroptosis starts with the production of lipid peroxides as a general cellular suicidal program with an iron-mediated oxidative mechanism. Cellular reactions exhibit redox equilibrium, and disruption of redox equilibrium is attributed to the synthesis and accumulation of reactive oxygen species. Definitive ROS, including superoxide anion (O₂ ⁻•), hydrogen peroxide (H₂O₂), hydroxyl radicals (HO•), hydroperoxides (ROOH), and hydroxyl radicals (ROO•), are formed by partial reduction of oxygen. ROS are generated inevitably from oxidative phosphorylation in mitochondria to cellular respiration, and the endogenous antioxidant system is instrumental to remove the oxidative stress. It has been held as an axiom that ROS accumulation also lays the groundwork of multiple pathologies, given its roles in cellular damage in diabetic cardiomyopathy (Kaludercic & Di Lisa, 2020), atherosclerosis (Yang et al., 2017), neurological complications (Manoharan et al., 2016), and in cell growth, especially in cancers (Aggarwal et al., 2019) (Dias Amoedo et al., 2020) (Tien Kuo & Savaraj, 2006) (Zeng et al., 2021). A detailed mechanistic overview of ferroptosis is illustrated in Figure 2.

What lies at the cardiac part of this cellular iron-mediated killing is lipid ROS. The most abundant ROS, superoxide, is generated by cytochrome P450 and NADPH oxidases (NOXs) partial reduction, forming H₂O₂ by superoxide dismutase (SOD), and the anions proceed to the production of hydroxyl radicals with the catalytic role of iron. In fact, published articles only documented the iron participation as the Fenton reaction, and the final Haber–Weiss reaction, obtained after balancing chemical equations from Fenton and the others, was less recognized than the Fenton reaction. Ferrous ions (Iron (II) or Fe²⁺) are mainly produced from the labile iron pool (LIP) and upon radical attack to heme groups with iron–sulfur (Fe–S) clusters (Gomez et al., 2014). Oxidation of ferrous to ferric ion (iron (III) or Fe³⁺) facilitates free radical formation from H₂O₂, whilst the O₂ ⁻• radicals are also oxidized to harmless O₂ as a net Haber–Weiss reaction. Taken together, the iron-mediated production of hydroxyl radicals is a “superoxide-driven Fenton-catalyzing Haber–Weiss reaction,” or Fenton/Haber–Weiss reaction, as illustrated in Figure 3.

After the Paleoproterozoic Great Oxygenation Event (GOE), lives on the earth were subjected to oxidation readily, especially for polyunsaturated lipids with bis-allylic carbons (Wagner et al., 1994). The victim of such ROS attack in ferroptosis after all the aforementioned series of events is, therefore, polyunsaturated fatty acids (PUFAs). Under normal physiology, PUFAs, including arachidonic acid, eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA), are situated in the cell membrane. The attack from accumulating free radicals to PUFAs, otherwise named peroxidation reaction, generates phospholipid free radicals (PL•) and, therefore, PUFA-containing-phospholipid hydroperoxides (PL-PUFA (PE)-OOH, PLOOH in short) (Forcina & Dixon, 2019), facilitated by different lipoxygenases (LOXs). It was also demonstrated that depletion of an acyl-CoA synthetase ACSL4 and LPCAT3 esterification enzyme inhibited ferroptosis (Doll et al., 2017) (Yuan et al., 2016). PLOOHs execute the unelucidated last hit to the cell membrane and initiate disruption to cellular integrity, leading to ferroptosis.

The transmembrane cystine/glutamate exchanger commences the work to initiate a ferroptosis-specific antioxidant system. System x_c ⁻, which was found to be inhibited by erastin, serves as an amino acid homeostatic control with the exchange of extracellular L-cystine and intracellular L-glutamate. Dissecting the antiporter, it consists of two subunits, a light chain solute carrier family 7 member 11 (SLC7A11) and a heavy chain subunit SLC family 3 member 2 (SLC3A2), which are targeted by respective inhibitors. Intracellular cysteine from cystine reduction facilitates the production of glutathione (GSH) that is catalyzed by glutamate–cysteine ligase catalytic subunit (GCLC) (which is inhibited by buthionine sulfoximine (BSO)) and then by glutathione synthetase (GSS). The classical redox-associated glutathione system (GSH and oxidized GSH disulfide (GSSG)) comes in to play a role in antioxidant defense, proven back in the 90s (Ceballos-Picot et al., 1996). Glutathione peroxidase 4 (GPX4) protects the cells from ferroptotic death by reducing toxic PLOOHs to PUFA-containing-phospholipid hydroxides (PL-PUFA (PE)-OH, PLOH in short), with the presence of selenium (Liu et al., 2021) and GSH (Ursini & Maiorino, 2020). While PLOHs appear to be non-ferroptogenic (not ferroptosis-inducing), this marks the end of the brief ferroptosis mechanisms as the homeostasis is achieved.

Ferroptosis has been observed in different pathologies. For example, in Alzheimer’s disease that is characterized by prominent brain cell death, β-amyloid plaques and neurofibrillary tangles were investigated, and excess iron accumulation and downregulation of iron exporter, ferroportin1, were observed, thereby explaining the oxidative stress exerted and promoting the AD cognitive impairment (Bao et al., 2021). In renal ischemia/reperfusion injury (IRI), ferroptosis is proven in the mediation of renal tubule-synchronized necrosis, and a novel third-generation ferrostatin 16–86 could rescue or protect the tubular damage that contributes to IRI (Linkermann et al., 2014). In cancer, particularly in colorectal cancer, it was evident that ferroptosis promotes metabolic rewiring, or the Warburg effect, which favors cancer cell growth, as well as suppresses ferroptosis sensitivity by inducing ROS production and activating nuclear factor erythroid 2-related factor 2 (NRF2) (Yuan et al., 2021). These are just few examples that ferroptosis correlates with disease progression, and more details about various pathologies can be found in other good articles such as Jiang et al. (2021b) and Yan et al. (2021) for readers’ reference.

Cellular signaling in biological systems evolved with harmonized crosstalk and attempting to inspect the entirety via a single chronological representation remains laborious to reach the finality. It becomes interesting whether the accumulating lipid ROS being non-eliminated construct a feedback influence on the epitranscriptomic marks. Oxygen atoms in –OH groups and phosphodiester backbone are the most vulnerable to be subjected to chemical damage or oxidation (Liu et al., 2022), and ROS onslaught has demonstrated evidently in mutations (Niedernhofer et al., 2003), cell arrest (Dixon & Stockwell, 2014), and epitranscriptomic induction (Kumar & Mohapatra, 2021). Particularly in cancer, m⁶A induction has been studied and reviewed in response to the production of ROS, and a biphasic and conflicting effect on tumor growth, intriguingly, has been noticed (Chio & Tuveson, 2017) (Yang & Chen, 2021). The potential ROS effect in ferroptosis via epitranscriptomic mediation is hence plausible.

Since the concept of “global m⁶A level can be ROS-induced” was revealed, one ROS-induced post-translational regulation on m⁶A demethylase was discovered recently (Yu et al., 2021). In this study by Yu et al., human cell lines with high m⁶A induced by ROS and determined by m⁶A-seq, were employed to survey the intrinsic mechanism that contributed to the elevation, where SUMOylation in m⁶A demethylase ALKBH5 was found to be associated using comet analysis, a single-cell gel electrophoresis assay that helps determine DNA damage and repair equilibrium at a single cell level. Particularly, SUMOylation-deficiency in ALKBH5 led to weakened DNA repair in H₂O₂-induced DNA damage, in other words, SUMOylation in ALKBH5 is essential in the increase of global m⁶A level by limiting the activity of m⁶A erasers. As ROS also joins the ferroptotic signaling and can lead to ferroptotic cell death, how ROS can potentially construct a stressful environment and add on epitranscriptomic modifications of ferroptosis proteins remains to be extrapolated. Having a feedback loop discovered that thrusts in the cell death process offer a great potential to manipulate the pathways, and the prospects of targeting ferroptosis in therapeutic settings await.

Dixon et al. (2012) extensively acknowledged the importance of iron in its mediation to the PCD event by coining the “ferro-” in the nomenclature of the iron-driven cell death, ferroptosis, assisted by the Nomenclature Committee of Cell Death (NCCD). Before then, prominent iron overload was observed among pathologies, such as hereditary hemochromatosis, along with the complications manifested, including organ damage, hypothyroidism, and hypogonadism. Managing iron homeostasis, thence, is necessitated from a medical standpoint, combined with the fact that ferroptosis is also dependent on intracellular iron status. In addition, en route to the research on how important iron to ferroptosis is, we also discovered some connections between iron and epitranscriptomic marks upon rummaging articles. We aimed to address the potential association of iron status and epitranscriptomics in ferroptosis and provided upcoming possible research directions to facilitate the elucidation of this mystery.

The fact that heme groups and Fe–S clusters are frequently under the attack of various kinds of ROS is well known (Imlay, 2006). This increases the intracellular level of Fe²⁺ apart from the LIP, though the LIP serves as the predominant source of Fe²⁺. In fact, perturbations of epitranscriptomics that affect the iron level or iron metabolism have been shown via some direct studies. In a hypopharyngeal squamous cell carcinoma (HPSCC) study by Ye et al. (2020), transcriptomic analyses including m⁶A-seq, RNA-seq, and RIP-seq identified m⁶A reader YTHDF1’s downstream target transferrin receptor (TFRC), simultaneously linking to poor prognosis in postoperative platinum-based chemoradiotherapy (CCT) or radiation patients in an m⁶A-dependent manner. HPSCC patients with intratumorally elevated Fe²⁺ were also shown upregulated YTHDF1 expression, and knockdown YTHDF1 in HPSCC cells proved the suppression of cell proliferation and migration ability. Taken together, as YTHDF1 modifies TFRC mRNA in cytosol and modulates transcriptomic stability and fate, relationships between an RMP and an iron metabolism participant were speculated by this pioneering work. Additionally, a pancreatic ductal adenocarcinoma (PDAC) study by Huang et al. (2021) aimed at elucidating the connection of ALKBH5 and iron metabolism, concretely on mRNAs encoding ubiquitin ligase FBXL5 and iron importers SLC25A28 and SLC25A37. ALKBH5 was identified to be mechanistically associated to the RNA decay event for FBXL5, and the team has divulged its unique prognostic ability among multiple m⁶A regulators analyzed in the study. Considering FBXL5-IRP2 serves as the cardinal part to iron metabolism (Wang et al., 2020), this study adds on the evidence of connecting epitranscriptomic-mediated iron metabolism since the bridge between FBXL5 and ALKBH5 can now be surmised through transcriptomic analyses, and further validation work awaits to confirm.