METTL16-m⁶A-PGC-1α axis contributes to exercise-induced mitochondrial adaptations in skeletal muscle of high-fat diet-fed insulin-resistant mice

Abstract

Background

The dynamic change of N6-methyladenosine (m⁶A) modification on substrate RNA molecules plays a critical role in different biological processes and disease pathogenesis. Although the beneficial effects of exercise training (ET) on skeletal muscle insulin resistance (IR) are well-established, the contribution of RNA m⁶A modification in ET-related adaptations in high-fat diet (HFD)-induced IR remains unclear.

Results

In this study, we show that exercise stimulation triggers a dynamic shift in skeletal muscle m⁶A modification levels during HFD consumption. As a key m⁶A methyltransferase, METTL16 was downregulated in HFD-fed mice and upregulated by ET at both the mRNA and protein levels. In vitro, METTL16 knockdown disrupted mitochondrial ultrastructure, reduced electron transport chain complex activities, and decreased the NAD⁺/NADH ratio, ATP content, and mitochondrial membrane potential, indicating impaired mitochondrial function. Concomitantly, METTL16 loss lowered m⁶A on PGC-1α mRNA, reducing its stability and protein abundance and blunting insulin signalling, whereas PGC-1α overexpression partially reversed these defects.

Conclusions

In conclusion, METTL16 functions as an exercise-responsive m⁶A methyltransferase that may modulate PGC-1α, mitochondrial function, and insulin-related signalling in HFD skeletal muscle, implicating the METTL16-m⁶A-PGC-1α axis in exercise-induced metabolic adaptations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12915-026-02519-5.

Background

As the body’s largest endocrine-related organ, skeletal muscle plays important roles in glucose clearance and utilization in the postprandial state in humans [1]. Under euglycemic-hyperinsulinemic conditions, nearly 80% of body glucose occurs in skeletal muscle via insulin-stimulated glucose disposal [2, 3]. Muscle insulin resistance (IR) is thought to be the initiating or primary defect in the pathogenesis of type 2 diabetes (T2D) [4]. Long-term diabetic conditions often lead to various diabetic complications including muscle myopathy, which is recognized as a primary risk factor for diabetic disability and is increasing in prevalence [5, 6]. Importantly, effective methods of preventing and reversing skeletal muscle defects in diabetic conditions still need to be further explored.

Exercise training (ET) protects skeletal muscle from pathological remodeling, thereby improving whole-body glucose disposal and insulin sensitivity, and ET is, therefore, an important cornerstone of treatment for the prevention of obesity and diabetes-related diseases [7, 8]. Several high-fat diet (HFD) studies in both human and animal models supported that ET led to significant weight loss and is protective against IR [9, 10]. ET can trigger complex molecular response across the continuum of IR by promoting glucose uptake and utilization in skeletal muscle [11]. This adaptive response usually is induced by complex cellular signalling events, including protein modifications and changes in gene activity.

RNA N6-methyladenosine (m⁶A) modification is a highly conserved pattern across mammalian cells in the transcriptional control of gene expression. The regulation of RNA m⁶A modification is highly dynamic and reversible. It is finely balanced through an interplay among m⁶A methyltransferases [12, 13], demethylases [14, 15], and binding proteins [16–18]. In skeletal muscle, the dysregulation of m⁶A balance is associated with the aberrant gene modification that contributes to a progressive loss of β-cell mass and function, as well as insulin sensitivity [19, 20]. In clinical studies, a significant upregulation of m⁶A demethylase FTO mRNA and protein levels has been detected in skeletal muscle from T2D patients as compared to the age-matched normal weight control group [21, 22]. FTO expression and FTO-mediated m⁶A demethylation modulated the activation of p70/86-S6 and PKB phosphorylation at both Thr308 and Ser473 sites, which maintains crucial role for skeletal muscle development [22]. Compared to the wild-type mice, an increased expression of tyrosine phosphorylation of IRS proteins was observed in the skeletal muscle of WTAP^± mice [23]. Despite emerging insights into how m⁶A imbalance contributes to skeletal muscle IR, the m⁶A-dependent molecular responses to exercise remain largely unexplored.

Here, we analyzed the expression patterns of major methyltransferases and demethyltransferases in response to ET in skeletal muscle of mice with T2D. Notably, we identified methyltransferase-like 16 (METTL16) as an m⁶A methyltransferase that is downregulated by HFD feeding and upregulated by ET at both the mRNA and protein levels. Moreover, we show that METTL16 deficiency in myoblasts reduced peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α) in an m⁶A-dependent way, which is associated with mitochondrial dysfunction and impaired insulin-related signalling. These observations suggest that METTL16-dependent regulation of PGC-1α may contribute to mitochondrial remodeling and altered muscle glucose metabolism under HFD conditions. By focusing on ET as an intervention in HFD-fed mice, this study provides a framework for investigating how m⁶A-dependent mechanisms may participate in exercise-related mitochondrial and metabolic adaptations in skeletal muscle.

Results

ET promotes m⁶A methyltransferase METTL16 expression in skeletal muscle of HFD-induced diabetic mice

To investigate the regulation of muscle m⁶A modification by ET under HFD conditions, we detected global mRNA m⁶A levels together with the expression of major m⁶A methyltransferases, demethylases, and reader proteins. We found that exercised HFD mice showed improved systemic insulin responsiveness in the ITT (Fig. 1A) and enhanced glucose tolerance in the OGTT compared with HFD mice (Fig. 1B). Histological analyses further revealed that ET presented obviously less obesity-induced loose arrangement of myofibers and muscle atrophy in the HFD + ET mice compared with those HFD mice (Fig. 1C). These results indicate that ET alleviates obesity-associated impairments in systemic glucose homeostasis and improves histopathology in HFD mice.

Fig. 1.

Exercise training (ET) is associated with increased methyltransferase METTL16 in skeletal muscle of HFD-induced diabetic mice. A Insulin tolerance test (ITT) in Chow, HFD, and HFD + ET mice (n = 5). B Oral glucose tolerance test (OGTT) in different groups (n = 5). C Representative microphotographs of hematoxylin and eosin (H&E) staining of gastrocnemius skeletal muscle tissue from mice in three groups. (n = 3) Scale bar: 100 µm. D The global mRNA m⁶A level in different groups. E qRT-PCR determination of mRNA levels of m⁶A methyltransferases and demethylases (n = 5). F Immunoblot detection and statistical data of METTL16 expression in Chow, HFD, and HFD + ET mice (n = 4). Data are expressed as the means ± SEM; ^*p < 0.05, ^**p < 0.01

Subsequently, we quantified global mRNA m⁶A level in skeletal muscle under HFD condition and observed that ET significantly reduced global m⁶A level in HFD-fed mice (Fig. 1D). The mRNA levels of several critical methyltransferases (METTL3, METTL14, METTL16, WTAP, and RBM15) and demethylases (FTO and ALKBH5) were also analyzed in the skeletal muscle. As shown in Fig. 1E, METTL3 was found to be increased and METTL16 was decreased in HFD mice as compared to the control group. Upon exercise stimulation, METTL14 and METTL16 indicated an obvious increase compared to HFD mice, suggesting the METTL16 mRNA level presented a change during ET intervention under HFD conditions. The expression of METTL16 protein in the muscle was further evaluated by Western blot assays (Fig. 1F). Similar to the qPCR assay, METTL16 protein expression presented a significant decrease in HFD mice as compared to the chow group, while its expression was promoted in HFD + ET mice as compared to HFD mice. These results suggest that METTL16 is an exercise-responsive m⁶A methyltransferase whose expression is suppressed by HFD and restored by ET in skeletal muscle.

METTL16 deficiency suppresses mitochondrial complex expression in myogenesis

Since mitochondria are essential for maintaining skeletal muscle energy homeostasis, we assessed the abundance of mitochondrial-related proteins in response to diet and exercise. In skeletal muscle from HFD-fed mice, PGC-1α protein expression was suppressed in HFD mice, which was accompanied by a lower expression of mitochondrial complex subunit (MCS) III and V. After ET stimulation, protein expression of PGC-1α as well as complexes III and V exhibited a markedly increased expression in HFD + ET mice in contrast to those in HFD mice (Fig. 2A). To further characterize mitochondrial adaptations during myogenesis, C2C12 myoblasts were stimulated to differentiate into myotubes for 6 days. As a marker of myotubes, MYH expression was dramatically promoted at 4 and 6 days during C2C12 myogenesis and was accompanied by increased MCSs (Fig. 2B). Our results are consistent with previous studies showing that myogenic differentiation is accompanied by mitochondrial remodeling and increased abundance of respiratory chain components [24, 25]. Furthermore, we found that METTL16 protein content sharply increased at 4 and 6 days during myogenesis (Fig. 2B).

Fig. 2.

C2C12 cell differentiation is accompanied by increased mitochondrial complex subunits and elevated METTL16 expression. A Immunoblot detection and statistical data of mitochondrial complex subunit (MCS) I, II, III, IV, and V after exercise and dietary intervention. B Protein expression of MYH, METTL16, and MCS I, II, III, IV, and V during myotube differentiation. C Protein levels of MYH, METTL16, and MCS I, II, III, IV, and V treated with or without siMETTL16 during myotube differentiation. Data are expressed as the means ± SEM from n = 4 independent experiments, ^*p < 0.05, ^**p < 0.01

Next, we evaluated the effects of METTL16 on the abundance of mitochondrial respiratory chain complex subunits. As shown in Fig. 2C, METTL16 deficiency led to a pronounced reduction in the protein levels of complex III and V, together with a more modest decrease in complexes I, II, and IV. These findings indicate that METTL16 appears to help maintain normal abundance of mitochondrial respiratory chain complex subunits, consistent with its role in supporting mitochondrial functional integrity.

METTL16 deficiency impairs mitochondrial function and insulin signalling

To investigate the mechanistic role of METTL16 in mitochondrial function, we conducted a functional assessment of electron transport chain (ETC) complexes in C2C12 cells. As shown in Fig. 3A and B, METTL16 knockdown significantly reduced the enzymatic activities of complexes I, III, and V, and was accompanied by a reduction in the NAD⁺/NADH redox ratio. Given the central role of mitochondria in adenosine triphosphate (ATP) production, we next assessed the effect of METTL16 on cellular ATP content. In contrast to control cells, METTL16-deficient C2C12 cells exhibited a marked decrease in ATP levels (Fig. 3C). Mitochondrial membrane potential (MMP), a key indicator of mitochondrial function [26], was also significantly reduced in METTL16 knockdown cells, as determined by a plate-based MMP assay and further confirmed by fluorescence imaging (Fig. 3D and E).

Fig. 3.

METTL16 deficiency disrupts mitochondrial homeostasis and impairs insulin-related signalling (IR). A Mitochondrial electron transport chain complex activities of myoblasts treated with or without METTL16 siRNA. B NAD⁺/NADH ratio of myoblasts treated with or without METTL16 siRNA. C Intracellular ATP level of myoblasts treated with or without METTL16 siRNA. D Relative MMP level of myoblasts treated with or without METTL16 siRNA. E Representative images of fluorescence staining quantification of the cellular MMP in myoblasts treated with or without METTL16 siRNA. F Glucose uptake and GLUT4 expression in myoblasts treated with or without METTL16 siRNA. G Immunoblot detection of p-AKT/AKT and p-GSK-3α/β/GSK-3α/β in myoblasts transfected with scrambled RNA and siMETTL16 siRNA for 48 h, followed by palmitic acid challenge for another 24 h with or without insulin for an additional 10 min. Data are expressed as the means ± SEM from n = 3 independent experiments, ^*p < 0.05, ^**p < 0.01

As mitochondrial dysfunction has been reported to impair insulin action [27, 28], we examined insulin-related signalling events in C2C12 cells. In the study, METTL16 knockdown resulted in a decreased GLUT4 expression and diminished glucose uptake (Fig. 3F). Moreover, insulin-stimulated phosphorylation of AKT at Thr308 and Ser473 as well as GSK-3α/β was attenuated in METTL16-deficient cells (Fig. 3G), indicating a broader impairment of insulin signalling. These observations collectively indicate that METTL16 knockdown is associated with reduced mitochondrial ETC complex activities, ATP production, and MMP, together with blunted GLUT4 expression, glucose uptake, and insulin-stimulated AKT/GSK-3α/β signalling.

METTL16 regulates PGC-1α mRNA stability in an m⁶A-dependent manner

METTL16-dependent m⁶A modification has been reported to influence the mRNA stability and translation efficiency of its target transcripts [29–31]. Putative METTL16-binding sites on the PGC-1α mRNA were detected via a predictor based on sequence-based m⁶A modification sites [32]. The potential predicted sites occur mostly enriched in regions of PGC-1α mRNA 3’UTRs (Fig. 4A). The four potential PGC-1α m⁶A modifications with very high confidence are shown in Fig. 4B. Given the identified PGC-1α mRNA methylation by the METTL16-based m⁶A RNA methylation complex, we then sought to examine the role of METTL16 gene knockdown on PGC-1α levels. To assess whether METTL16 affects m⁶A modification of PGC-1α mRNA, we performed m⁶A RNA immunoprecipitation followed by qPCR. In contrast to the control group, METTL16 knockdown reduced m⁶A enrichment on PGC-1α mRNA by approximately 33.5% (Fig. 4C).

Fig. 4.

METTL16 binds to the coding region of PGC-1α mRNA and decreases its stability. A Prediction sites of PGC-1α mRNA binding to METTL16 as produced from SRAMP server. B Schematic of the putative PGC-1α mRNA segments. C PGC-1α mRNA collected from a m⁶A RNA immunoprecipitation assay in myoblasts transfected with scrambled RNA and siMETTL16 siRNA. D The stability of PGC-1α mRNA in myoblasts transfected with scrambled RNA and siMETTL16 for 24 h with actinomycin D for another 24 h. E qRT-PCR determination of mRNA levels and F immunoblot detection of PGC-1α treated with or without METTL16 siRNA cells. Data are expressed as the means ± SEM from n = 3 independent experiments, ^**p < 0.01

Given that m⁶A methylation often stabilizes transcripts [16, 33], the reduced m⁶A enrichment on PGC-1α mRNA may contribute to its destabilization. Consistently, METTL16 knockdown shortened the half-life of PGC-1α mRNA (Fig. 4D). Furthermore, the downregulation of METTL16 significantly suppressed PGC-1α expression, at both mRNA level and protein level (Fig. 4E, F). Together, these above data suggest that METTL16-dependent m⁶A RNA methylation may contribute to the regulation of PGC-1α expression.

PGC-1α alleviates METTL16 deficiency-associated mitochondrial dysfunction in myoblasts

Given the established role of PGC-1α as a key regulator of mitochondrial gene expression and oxidative metabolism, we next investigated whether PGC-1α overexpression could mitigate the mitochondrial abnormalities associated with METTL16 deficiency. C2C12 cells were transfected with siMETTL16, pcDNA-PGC-1α, or the combination siMETTL16/pcDNA-PGC-1α and subjected to mitochondrial functional analyses. Overexpression of PGC-1α partially reversed the METTL16 knockdown-induced reduction in MCSs (Fig. 5A). Furthermore, METTL16 knockdown decreased cellular ATP levels to ~ 58.7% contrasted to the control group, whereas co-expression of PGC-1α significantly attenuated this reduction (Fig. 5B).

Fig. 5.

METTL16 regulates mitochondrial function via PGC-1α in myoblasts. A Immunoblot detection of METTL16, PGC-1α, and MCS I, II, III, IV, and V in myoblasts transfected with scrambled RNA, siMETTL16 siRNA, pcDNA-PGC-1α, and siMETTL16/pcDNA- PGC-1α for 48 h. B Intracellular ATP level and C relative MMP level in transfected myoblasts. D Representative images of fluorescence staining quantification of the cellular MMP. E Representative images of mitochondrial ultrastructure using TEM in myoblasts transfected with scrambled RNA, METTL16 siRNA, pcDNA-PGC-1α, and siMETTL16/pcDNA- PGC-1α for 48 h. Data are expressed as the means ± SEM from n = 3 independent experiments, ^*p < 0.05

Because of a decrease in mitochondrial membrane potential (MMP) being a hallmark of mitochondrial injury, we next assessed MMP using a microplate-based assay and fluorescence imaging. Interestingly, MMP level in METTL16-knockdown cells was a ~ 24.8% decrease compared with control cells, yet this decrease was attenuated in siMETTL16/pcDNA-PGC-1α cells when compared with siMETTL16-treated cells (Fig. 5C). Similar changes in MMP were observed by immunofluorescence analysis (Fig. 5D). In addition, transmission electron microscopy (TEM) revealed that METTL16 knockdown cells contained numerous mitochondria with swollen matrices and collapsed cristae, indicative of structural damage, whereas PGC-1α overexpression ameliorated these ultrastructural abnormalities (Fig. 5E). Collectively, the results suggest that METTL16 influences mitochondrial structure and function, at least in part, through a PGC-1α-dependent mechanism.

PGC-1α mitigates METTL16 deficiency-induced insulin signalling defects in myoblasts

In C2C12 myoblasts, METTL16 downregulation resulted in a decrease in expression of GLUT4 and a corresponding reduction in glucose uptake (Fig. 6A). Under insulin stimulation, METTL16 deficiency diminished phosphorylation of AKT at Thr308, Ser473, and GSK-3α/β at Ser21/9 (Fig. 6B). However, PGC-1α overexpression was able to attenuate the METTL16 knockdown-induced IR. Together, these results identify METTL16 as an exercise responsive regulator and suggest that it may participate in modulating insulin signalling in myotubes.

Fig. 6.

METTL16 promotes insulin sensitivity via PGC-1α in myoblasts. A Glucose uptake analysis and GLUT4 expression in transfected myoblasts treated with or without METTL16 siRNA. B Immunoblot analysis of p-AKT (Thr308/Ser473) and total AKT, and p-GSK-3α/β (Ser21/9) and total GSK-3α/β in myoblasts transfected for 48 h with scrambled RNA, siMETTL16 siRNA, pcDNA-PGC-1α, and siMETTL16 plus pcDNA- PGC-1α with or without insulin challenge for another 10 min. Data are expressed as the means ± SEM from n = 3 independent experiments,.^*p < 0.05

Discussion

The widespread benefits of exercise across multiple organ systems are well-recognized, and exercise intervention has shown excellent effects on many human common diseases, including T2D [34–36]. Several studies have revealed the inconsistent expression between mRNA level and its corresponding protein expression in mammals, indicating an important issue on post-transcriptional regulation in controlling gene expression [19, 37, 38]. This highlights that relying solely on mRNA changes is insufficient to fully capture post-transcriptional regulation underlying disease phenotypes. RNA m⁶A methylation is the most abundant and highly conserved epigenetic modification on RNA related to all mammals [39, 40]. A great deal of evidence indicates that there is a strong association between deregulation of m⁶A modification and progression of diabetes in recent years [39, 41]. However, the role of RNA m⁶A modification on skeletal muscle IR in response to exercise remains poorly understood.

Emerging evidence has demonstrated that ET can trigger m⁶A modification and m⁶A-associated regulatory factors to regulate different biological processes [42–45]. Endurance exercise led to reduced cardiac mRNA m⁶A modification, and a lowering of METTL16 levels upon exercise regulated PHLPP2 mRNA m⁶A to activate Akt-S473 in maintaining cardiac homeostasis [43]. Physical exercise supports the beneficial effects of m⁶A methylation status on medial prefrontal cortex neurons, thus conferring stress resilience [45]. It was also reported that HFD significantly enhanced the level of m⁶A and m⁶A-associated factor METTL3, YTHDF2, and FTO, while the relative level was reversed upon exercise stimulation [46]. Despite the incomplete understanding of the exact mechanism of m⁶A modulating the physiological functions upon exercise stimulation, however, some mechanistic cues are surfacing. In skeletal muscle IR, a remarkable elevation of m⁶A methylation RNA level has been found in diabetic mice [20, 47], which is consistent with our observations. How exercise intervention regulates skeletal muscle insulin resistance is an unresolved question. In the study, we evaluated the dynamic changes of major methyltransferases and demethyltransferases in response to ET in skeletal muscle of mice with T2D. Together, these data show that METTL16 exhibits diet- and exercise-sensitive regulation at both mRNA and protein levels, implicating dynamic control of this m⁶A writer during metabolic stress and ET. As an emerging m⁶A methyltransferase, METTL16 has been reported to target only a few transcripts [32, 48–50]. The effect of METTL16 on the cells is being investigated; however, growing evidence supports the important roles in the progression of most cells.

In this study, METTL16 levels increased with MCSs during myogenic differentiation, suggesting that METTL16 is linked to the mitochondrial remodeling that accompanies increased energetic demands. Loss of METTL16 blunted the myogenesis-associated rise in MCS expression. Beyond these changes in mitochondrial protein abundance, downregulation of METTL16 was associated with reduced ATP production, diminished MMP, and ultrastructural abnormalities, including cristae disorganization, indicating impaired mitochondrial function. Furthermore, METTL16 depletion was accompanied by a disrupted NAD⁺/NADH redox balance, consistent with the requirement of NAD⁺ as an essential cofactor for bioenergetics [51]. The concomitant reduction in ETC complex activities and NAD⁺/NADH ratio supports a role for METTL16 in sustaining mitochondrial bioenergetic competence, possibly through m⁶A-dependent regulation of PGC-1α and related transcripts. These findings point to METTL16 as an epitranscriptomic regulator of mitochondrial homeostasis with potential relevance to diabetic metabolic disturbances. However, the precise mechanisms linking METTL16-mediated RNA methylation to downstream metabolic reprogramming remain to be fully elucidated.

In addition, mitochondrial dysfunction has critical roles in muscle IR, which is related to many genes and molecular pathways [52, 53]. Thereby, we examined IR-related signalling and found that METTL16 downregulation attenuated insulin-stimulated phosphorylation of AKT at Thr308 and Ser473, as well as GSK-3α/β at Ser21/9 in C2C12 myoblasts. Mechanistically, METTL16 knockdown regulated m⁶A enrichment on PGC-1α mRNA and blunted insulin-signalling readouts, consistent with diminished glucose uptake. Previous studies showed that mitochondrial dysfunction may serve as a primary pathogenic driver in the development of insulin resistance, as impaired mitochondrial homeostasis can disrupt critical signalling pathways, including those involving AKT and GSK-3α/β phosphorylation, which are essential for glucose uptake and metabolic homeostasis [54–56]. In contrast, persistent glucose metabolic disorders, such as those characterized by reduced glucose uptake, may subsequently exacerbate mitochondrial damage. This intricate interplay between mitochondrial health and glucose metabolism underscores the progressive and multifaceted nature of insulin resistance. Collectively, the findings provide preliminary mechanistic insight into how METTL16-dependent m⁶A regulation of PGC-1α may link mitochondrial functional readouts to insulin-related signalling in skeletal muscle cells under HFD stress.

Despite the well-established importance of PGC-1α in the pathogenesis of T2D [57], the contribution of m⁶A modification to PGC-1α regulation remains poorly understood. Here, we showed that METTL16 downregulation significantly reduced PGC-1α mRNA stability and protein abundance in C2C12 cells, suggesting that METTL16 influences PGC-1α expression at the post-transcriptional level. Moreover, PGC-1α overexpression partially mitigated METTL16-deficiency-associated alterations in mitochondrial parameters and glucose uptake, which is consistent with previous reports that PGC-1α is a key regulator of mitochondrial oxidative metabolism and function [58, 59]. These findings support a model in which PGC-1α acts as one downstream effector through which METTL16-mediated m⁶A modification impacts mitochondrial homeostasis and insulin-related responses in skeletal muscle cells. Taken together, these data suggest that METTL16 contributes to the regulation of mitochondrial function and glucose metabolism via m⁶A-mediated mechanisms, with PGC-1α emerging as a plausible component of this regulatory axis. While PGC-1α m⁶A modification represents a significant mechanism in METTL16-mediated mitochondrial control, potential m⁶A contributions from other direct or indirect targets remain to be fully elucidated. This warrants comprehensive single-nucleotide resolution mapping of METTL16-dependent m⁶A epitranscriptome across mitochondria-associated genes to systematically delineate this intricate multilayered epigenetic network.

Conclusions

In summary, we identify METTL16 as an exercise-responsive m⁶A methyltransferase in skeletal muscle and suggest that it contributes to mitochondrial homeostasis and insulin-related signalling under HFD conditions. In vitro analyses revealed that METTL16-dependent m⁶A modification regulates PGC-1α mRNA stability and protein abundance, and METTL16 knockdown is associated with reduced ETC complex activities, decreased ATP production and mitochondrial membrane potential, and blunted insulin-stimulated AKT/GSK-3α/β signalling and glucose uptake; importantly, these alterations are partially ameliorated by PGC-1α overexpression. The findings offer preliminary mechanistic clues as to how METTL16-mediated m⁶A regulation of PGC-1α mRNA may modulate skeletal muscle mitochondrial function and insulin responses. At this stage, however, the METTL16-m⁶A-PGC-1α axis is regarded as a putative regulator of mitochondrial homeostasis and muscle glucose metabolism, and further tissue-specific validation and in vivo functional studies will be needed to clarify its pathophysiological significance.

Methods

Animal and treatment

Four-week-old male C57BL/6 J mice (SPF) were obtained from Shanghai SLAC Laboratory Animal Co. (Shanghai, China) and housed in a barrier facility at a temperature (22–26 ℃), humidity (40–50%), and light/dark (12 h/12 h) cycle-controlled animal room. All the methods were approved by the Animal Experiment Ethics Committee of Fujian University of Traditional Chinese Medicine (Permit #SYXK(Min)2020–0002).

The mouse model of HFD-induced diabetes was used in the study. All the mice were randomly divided into two groups: the mice fed on normal-diet (Chow, 10 kcal% (H10010, Research Diets)) and the mice fed on HFD (HFD, 45 kcal% (H10045, Research Diets)). For ET intervention, the HFD mice received a treadmill exercise (HFD + ET) or no training. Briefly, for the 1 st week, mice were acclimatized on a treadmill with 5 min at 5 m/min, 50 min at 8 m/min, and 5 min at 5 m/min. From 2 to 12 weeks, mice were placed on treadmills for 5 min at 5 m/min, 25 min at 12 m/min, 5 min for rest, 20 min at 12 m/min, and 5 min at 5 m/min. After intervention, mice were deeply anaesthetized by intraperitoneal injection of liquid isoflurane (0.2–0.8 ml per mouse), and depth was confirmed by loss of pedal withdrawal reflex and absence of response to a firm pinch. While still under deep anesthesia, mice were euthanized by cervical dislocation. All procedures followed the guidelines of the Animal Experiment Ethics Committee of Fujian University of Traditional Chinese Medicine. Gastrocnemius skeletal muscle tissues were rapidly excised, snap-frozen in liquid nitrogen, and stored at − 80 ℃ until analysis.

Insulin tolerance (ITT) and oral glucose tolerance test (OGTT) assays

The assays of ITT and OGTT were followed as previous studies. In brief, mice were fasted for 4 h (hrs) and 12vhrs prior to ITT and OGTT assays, respectively. For the ITT assay, 0.75 U/kg of insulin was injected, and blood was obtained at 0-, 15-, 30-, and 45-min later. For the OGTT assay, the mice were given glucose by oral gavage at 2.0 g/kg body weight. Blood was collected from the retrobulbar tail vein at indicated times 0-, 15-, 30-, 45-, and 60-min. The plasma glucose concentration was examined by the glucose oxidation method.

Cell culture, differentiation, and transfection

Mouse C2C12 myoblasts were cultured in the DMEM media with 10% FBS and 1% penicillin–streptomycin at 37 ℃ in 5% CO₂. In general, the medium was changed every 2–3 days. When cell density reached 90% confluence, we changed cell media from 10% FBS to 2% horse serum to induce differentiation. After 6 days, full cell differentiation was observed with a characteristic of myotube fusion and spontaneous twitching. The transfection of small interfering RNA (siRNA) and plasmids was performed using a Lipofectamine 3000 Transfection Reagent (Invitrogen) as per the manufacturer’s protocol. The siRNA targeting MTTL16 was synthesized from GenePharma (Shanghai, China), and the wild-type Pgc-1α gene (NM_008904.2) plasmid was cloned into a pcDNA3.1 vector.

Mitochondrial preparation and complex activity analysis

Mitochondrial isolation from freshly harvested C2C12 cells was performed using differential centrifugation. Cellular homogenization was carried out in ice-cold mitochondrial isolation buffer (250 mM sucrose, 10 mM Tris–HCl, 1 mM EDTA disodium salt, pH 7.5) followed by gentle mechanical disruption with 40 strokes in a Dounce homogenizer. Cell debris was removed by low centrifugation, and the supernatant was collected following a centrifuge (12 000 g × 15 min, 4 ℃). The precipitate was resuspended with extraction buffer. Purified mitochondrial protein was determined using a BCA protein assay kit (cat23225, Thermo Fisher Scientific). Enzymatic activities of the mitochondrial complex I, II, III, IV, and V were measured spectrophotometrically using conventional assays as described [60].

NAD⁺/NADH assay

The measurement of NAD⁺/NADH in myoblasts was performed using a chromogenic assay based on the WST-8 reaction (S0176S, Beyotime, Jiangsu, China). A 200 µl aliquot of NAD⁺/NADH extraction buffer was added to 200 µl of cell suspension, followed by centrifugation at 12,000 g at 4 ℃ for 10 min. The supernatant was then collected for analysis according to the manufacturer’s protocol.

ATP assay

Cellular ATP content was performed by ATP assay kit (FLAA, Sigma, MO, USA) following the supplier’s manual. Briefly, the treated-C2C12 myoblasts were maintained in a 6-well plate, lysed using RIPA buffer, and the supernatants were used. The cellular when firefly luciferase catalyzed the oxidation of D-luciferin, ATP was consumed, and the light was emitted.

Mitochondrial membrane potential (MMP) measurement

A cationic dye 5,5′,6,6-Tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) was used for MMP measurement as described in our previous study. The accumulation of JC-1 in mitochondria forms J-aggregates and emits red fluorescence at normal MMP levels, while the abnormal mitochondria exist in J-monomers and emit green fluorescence. The fluorescence intensities of the red/green ratio reflect cellular MMP levels. Five micrograms per milliliter of JC-1 solution was mixed into the cells for at least half an hour, rinsed with PBS, and then scanned on a microplate fluorometer.

Glucose uptake assay

Medium glucose was examined using a commercial kit (ab136955, Abcam, USA) following the manufacturer’s protocol. In brief, the intensity was examined by proportional to glucose concentration at 540 nm, and glucose uptake from cells was measured based on the calculation of fresh medium glucose minus diminished medium glucose.

RNA stability measurement

The actinomycin D (5 µg/ml; A9415, Sigma, MO, USA) was mixed into the cells for 24 h to suppress the mRNA generation. The associated mRNA level at three indicated time points (0, 2, and 4 h) was examined by a qPCR method.

Gene expression

RNA was obtained from gastrocnemius skeletal muscle tissues and myoblasts by a TRIzol reagent (Invitrogen, San Diego, CA) followed by the supplier’s protocol and treated for reverse transcription using a PrimeScript RT-PCR kit (TaKaRa, Shiga, Japan). The qPCR assay was performed, and RNA levels were examined using the 2^−ΔΔCT assay, and the data were exhibited as relative levels of the control.

Relative quantification of mRNA m⁶A

The relative mRNA m⁶A quantification was determined using a m⁶A RNA Methylation Quantification Kit (P-9005, Epigentek). Next, mRNAs were isolated from purified RNAs by oligo(dT) poly-styrene beads, and a total amount of 200 ng mRNA was mixed to a 96-well plate. After incubation, m⁶A RNA capture, signal detection, and the relative quantification were examined according to the supplier’s protocol.

Methylated-RNA-immunoprecipitation (MeRIP)‑qPCR analysis

Fragmented mRNA was purified as an input material using a GenElute mRNA Miniprep Kit (MRN70-1KT, Sigma), and then mixed with anti-m⁶A antibody bound protein G. After RNA binding, it was rinsed with washing buffer, blocked with BSA, and rotated overnight at 4 ℃. The m⁶A RNA was finally eluted, and mRNA m⁶A enrichment was examined by qPCR assay.

Immunoblotting

Samples of gastrocnemius skeletal muscle tissues and cells were rinsed and solubilized in RIPA buffer (Beyotime, China); the supernatant was examined by a BCA protein assay kit (cat23225, Thermo Fisher Scientific). Subsequently, samples with an equal aliquot were loaded onto SDS-PAGE, then electro-blotted onto nitrocellulose membranes. Protein was probed with the appropriate primary antibody, followed with the secondary antibody. Subsequently, the specific band was finally examined, and the intensity of the protein band was further quantified by ImageJ software. The primary antibodies applied in the study are the following: anti-METTL16 (87,538, Cell Signaling Technology (CST)), anti-NDUFB8 (PA5-52,011, Invitrogen), anti-SDHB (459,230, Invitrogen), anti-UQCRC1 (459,140, Invitrogen), anti-MTCO1 (459,600, Invitrogen), anti-ATP5A1 (459,240, Invitrogen), anti-PGC1α (2178, CST), anti-GSK (5676, CST), anti-phospho-GSK-3α/β-S21/9 (9331, CST), anti-AKT (9272, CST), anti-phospho-AKT-Ser473 (4060, CST), anti-phospho-AKT-Thr308 (13,038, CST), anti-MYH (22,281–1-AP, Proteintech), anti-GLUT4 (66,846–1-Ig, Proteintech), anti-β-actin (66,009–1-Ig, Proteintech), anti-α-Tubulin (66,031–1-Ig, Proteintech).

Statistical analysis

Data are presented as a mean ± standard error of the mean (SEM) using GraphPad Prism 8.0. The independent-sample t-test was applied for comparison between two groups, and two-way ANOVA with Tukey test or one-way ANOVA was used for comparison between multiple groups. p < 0.05 was regarded as statistically significant.

Abbreviations

m⁶A

N6-methyladenosine

ET

Exercise training

HFD

High-fat diet

T2D

Type 2 diabetes

IR

Insulin resistance

ATP

Adenosine triphosphate

ETC

Electron transport chain

OGTT

Oral glucose tolerance test

ITT

Insulin tolerance test

JC-1

5,5′,6,6-Tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide

MCS

Mitochondrial complex subunit

MMP

Mitochondrial membrane potential

H&E

Haematoxylin and eosin

qPCR

Quantitative polymerase chain reaction

NAD⁺/NADH

Nicotinamide adenine dinucleotide oxidized/reduced form

TEM

Transmission electron microscopy

C2C12

Mouse myoblast cell line

METTL3

Methyltransferase-like protein 3

METTL14

Methyltransferase-like protein 14

METTL16

Methyltransferase-like protein 16

RBM15

RNA-binding motif protein 15

FTO

Fat mass and obesity-associated protein

ALKBH5

AlkB homolog 5

WTAP

Wilms tumor 1-associating protein

PGC-1α

Peroxisome proliferator-activated receptor gamma coactivator 1-alpha

AKT

Protein kinase B

GSK-3

Glycogen synthase kinase-3

Authors’ contributions

C.C., C.J., Q.X. conceived the study and contributed equally to this work. C.J., Q.X., Y.H., H.J.W., C.X.H., H.H.Z., Y.X., M.J.H. performed the experiments and analyzed the data with the supervision of C.C., W.L.L., X.H. C.C., and W.L.L, X.H. instructed the experiments and drafted the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the grants from National Natural Science Foundation of China (32271175) and the Natural Science Foundation of Fujian Province (2025J01959).

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information, which is provided in Additional file 1 and 2.

Declarations

Ethics approval and consent to participate

All animal procedures were approved by the Animal Experiment Ethics Committee of Fujian University of Traditional Chinese Medicine (approval number FJTCM IACUC 2020092) and were performed in accordance with institutional and national guidelines for the care and use of laboratory animals.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.