The present study provides a potential mechanism by which exercise may attenuate the influence of the FTO rs9939609 polymorphism on obesity risk. To the best of our knowledge, this is the first preliminary report on the effects of two isocaloric bouts of high and low intensity exercise on skeletal muscle FTO gene expression. The current findings demonstrate that an acute bout of high intensity exercise significantly downregulates skeletal muscle FTO mRNA during the early stages of recovery. This was not observed for lower intensity exercise. Downregulation of FTO mRNA expression was associated with elevated muscle glucose levels, but only in those individuals carrying the risk AA genotype. Despite higher intensity exercise inducing greater metabolic perturbations compared to the low intensity trial, metabolomics analysis was unable to identify any unique metabolic differences between the FTO genotypes. This study suggests that in addition to nutritional regulation, FTO is also regulated by exercise and may be involved in exercise’s role in reducing obesity risk.
The acute and significant downregulation of FTO mRNA following high intensity exercise is a major novel finding of this investigation. A weak trend for genotype by time interaction was also identified suggesting greater downregulation of FTO mRNA in the AA genotype compared to the other genotypes (AT and TT). Indeed, AA genotypes demonstrated a 0.32-fold decrease in FTO mRNA expression compared to a 0.21-fold and 0.11-fold decrease for AT and TT genotypes respectively, at 10 minutes post-exercise. Previous studies have shown that FTO gene activity is nutritionally regulated with high fat diet, fasting and glucose ingestion all having effects on FTO mRNA levels [24–26]. Only one other study has investigated the effect of high intensity training on FTO mRNA expression and showed lifestyle changes (diet and exercise) did not impact FTO gene expression in peripheral blood mononuclear cells . However, when FTO genotype was considered, FTO expression was up-regulated in AA genotype carriers and down-regulated in AG/GG genotype carriers in the intervention group . Though the direction of FTO mRNA change was opposite to the present study observations, different cell type, age, sex and intervention period, may explain such differences.
The downregulation of FTO mRNA following exercise in the present study could be due to its role as an ‘energy sensor’. FTO gene activity is sensitive to the energy status of the cell , and it is possible that FTO is responding to the change in energy status and increased energy demand of the muscle as a result of exercise. While a change energy status is quite complex, the current study used an untargeted metabolomics approach to see if the high and/or low intensity bout of exercise could unmask any differences in metabolic responses between FTO genotypes that would otherwise not be seen at rest. High intensity exercise induced a significant increase in muscle alanine, erythronate, fumarate, GHB, glucose, glycolate, lactate, malate, maltose, mannose, monopalmitoglycerol and tyrosine during the first 10 minutes of recovery. LO exercise caused similar but not as many metabolite changes. Despite greater metabolic perturbations following the high intensity compared to the low intensity exercise, O2PLS-DA multivariate regression analysis was unable to distinguish between FTO allelic variants based on metabolic profiles following the exercise bouts. A limitation of the O2PLS-DA model is that baseline (resting) and post-exercise measurements are grouped together. Incorporating a time point in which energy demand is at a minimum and under tight homeostatic regulation, the ability to identify differences may be confounded. Indeed, previous investigations have found similar metabolic profiles between allelic variants of FTO in at rest [29, 30]. While the current study was unable to identify any specific metabolite(s) and/or metabolic by-product(s), previous research has suggested kreb cycle intermediate fumarate as a potential modifier of FTO. Gerken and colleagues  demonstrated inhibition of Fto-catalyzed 1-meA demethylation by fumarate within 2OG decarboxylation assays. While the Gerken study  examined FTO function and not gene regulation, it is possible that elevated levels of fumarate (as seen during exercise) are also inhibiting its expression. It is clear that further functional studies are needed to explore other metabolites, especially those significantly impacted by exercise, as possible modifier candidates.
Recent studies have suggested that AMPK may also regulate FTO expression and function in skeletal muscle and could explain another mechanism by which exercise downregulates FTO mRNA. Using C2C12 cells, Wu and colleagues  showed that inhibition of AMPK upregulates FTO expression and activity and lipid accumulation, while activation of AMPK downregulates FTO expression and activity and reduces lipid accumulation. The current study showed that phosphorylated AMPKα was significantly increased during the early stages of recovery following high intensity exercise, however, no genotype by time interaction was identified. Further, no relationship was found between elevated AMPKα levels and the downregulation of FTO. While phosphorylated AMPKα may not be driving the downregulation of FTO mRNA in AA genotypes, it could still be impacting the FTO function. Observations from Wu et al.  and others [4, 5] suggest that inhibition of FTO function drives higher fat oxidation and lower fat accumulation possibly via FTO-dependent demethylation of mRNA m6A. In the current study, fat oxidation and/or markers of lipid accumulation were not measured post-exercise, however, individuals homozygous for the risk A-allele demonstrated greater muscle glucose levels compared to non-risk T-allele at 10 minutes recovery following both high and low intensity exercise. Higher intramuscular glucose levels post-exercise could reflect a metabolic shift towards greater lipid oxidation and away from glucose oxidation potentially via the AMPK activation and FTO-dependent demethylation of N6-methyladenosine mechanism as previously mentioned above. However, acute higher post-exercise intramuscular glucose levels observed in AA genotypes could be due to a number of processes involved in glucose metabolism including glucose delivery and uptake into the muscle, and the resynthesis of glycogen levels post-exercise. Plasma glucose concentrations were measured in the present study and were found to be similar between FTO genotypes in response to high intensity exercise and thus it is also unlikely that differences in plasma glucose levels could be responsible for the greater muscle glucose uptake. Glucose uptake into the sarcoplasm depends on the skeletal muscle expression of GLUT4 (an insulin and contraction regulated glucose transport isoform) , which is normally increased following exercise and can facilitate post-exercise glucose uptake . Although the current study did not measure GLUT4 translocation directly, we did measure phosphorylation of AS160, an insulin dependent and independent regulator of GLUT4 vesicle movement to, and/or fusion with, the plasma membrane . Despite exercise increasing AS160/Total AS160 in the early stages of recovery, there were no significant differences between genotypes. Though not statistically significant, the AA genotype group did complete their trial on average by about 3–7 minutes faster and produce higher total workloads in both exercise trials compared to AT and TT genotypes. Further research is needed to determine whether higher intramuscular glucose levels are due to genetic factors inherent in AA genotypes or influences from the aforementioned factors.
Several limitations do exist in the current study. Firstly, we acknowledge that our final sample size (n = 28) is relatively low. The average partial eta-squared for observed skeletal muscle variables (VIP metabolites, and protein and mRNA expression levels) was found to be of a medium effect size when performing high and low intensity exercise data analysis (η2 = 0.068 and η2 = 0.051 respectively). Secondly, we studied both males and females that were young and of “healthy” weight range (BMI range 24–26 kg/m2). It is apparent that substrate metabolism is subject to gender-specific regulation. Gender differences in muscle fibre type distribution and substrate availability to, and in, skeletal muscle , which also includes molecular differences in glucose and lipid metabolism of skeletal muscle. We used gender as well as age (another known factor) as covariates within our ANCOVA analysis. When gender or age were considered, the reported significant findings were still present. The influence of age was not anticipated given our relatively similar age distribution across alleles. Gender has shown to influence the effect of the FTO polymorphism on obesity related traits. However, it was clear from our study that despite gender differences within our cohort (nearly a 50/50 gender split), downregulation of FTO mRNA still occurred in the at-risk AA allele and within both genders. Thirdly, while the vastus lateralis muscle is the most common muscle of choice for biopsies because of its accessibility, it is of mixed fibre type composition and thus we cannot comment on fibre type specific differences. Furthermore, it is possible that the sampling window of up to 90 minutes was insufficient to detect any significant changes in FTO protein content resulting from exercise, with previous research demonstrating changes in the expression levels of other muscle proteins occurring greater than 3 hours .