The 8-week HFHF feeding substantially increased body weight of mice by 62% than those fed with NC (Fig. 1a, b). HFHF also increased fasting blood glucose, detrimentally affected glucose tolerance and insulin resistance, as evidenced by the significant increases in AUCs of OGTT and IST (Fig. 1c-g). HFHF negatively impaired the anti-fatigue capacity, as evidenced by the reduction in the running distances and time stayed spent on the instrument (Fig. 1h, i), alongside lower skeletal muscle mass (Fig. 1m) compared with the NC group. HFHF also increased the muscle glycogen content (Fig. 1l). Histopathological results confirmed that HFHF reduced the fiber cross-section area of quadriceps muscle while increased the inflammatory infiltrate, compared with NC group (Fig. 1j, k).
The 3-week exercise reduced body weight, improved glucose tolerance, insulin sensitivity and anti-fatigue activity, successfully alleviated inflammatory infiltration in skeletal muscle while improved muscle fiber cross-sectional area of HFHF-induced obese mice (Fig. 1j, k). Notably, consuming sucrose, TGS or IMO-contained water during exercise had no impact on the exercise induced weight loss efficacy, but differently affected physiological indicators, independent of food and water intake. Specifically, IMO augmented beneficial effects of exercise on improving glucose metabolism and skeletal muscle morphology, which outperformed other treatments in restoring HFHF-induced muscle loss (Fig. 1h-m). Drinking TGS-sweetened water failed to restore HFHF-caused abnormal losses in muscle mass and fiber cross-section area, while increased the inflammatory infiltrate, compared with mice drinking tap water. Drinking sucrose-sweetened water eliminated exercise induced improvements on anti-fatigue capacity and glucose tolerance.
In addition, we observed that the 3-week Diet-replacement (i.e., changing daily diet from HFHF to normal chow) without any exercise could also reduce body weight of HFHF-induced obese mice (Fig. 1a). However, no benefit was seen for glucose regulation and skeletal muscle functions HFHF-induced obese mice.
3.2 Skeletal muscle transcriptome reveals key biological pathways involved in exercise- induced weight loss
RNA sequencing on skeletal muscle was performed in order to identify the biological pathways and hub genes involved in weight loss. We detected 5713 genes with a fragments per kilobase of exon per million fragments mapped (FPKM) value and qualified annotations in skeletal muscles. Among them, 335 genes were upregulated by HFHF (e.g., Atf5, Elk4, Vegfa) compared with NC, mostly involving in PI3K-AKT pathways, while 305 were downregulated (e.g., Mapk, Ppar, p53, Myh4, Jun, Cyc1 and Yap), which were involved in PI3K-AKT, NF-κB, JAK-STAT and MAPK pathways.
The 3-week exercise intervention remarkably influenced the skeletal muscle transcriptome, and enabled to restore gene profiles that were altered by HFHF, leading to a similar skeletal muscle transcriptome compared with the NC group (Fig. 2a, d, e). In compared with the HFHF group, EX upregulated 143 genes while down-regulated 788 genes (Fig. 2b). DEGs were predominately related to fatty acid metabolism, amino acid metabolism, insulin resistance, and Hippo, AMPK, and mTOR signaling pathways (Fig. 2f, g).
Additionally, DR induced 261 DEGs however, such regulatory effects on a large number of genes in relation with obesity were distinct from exercise (Fig. 2c). For instance, DR uniquely upregulated genes involved in amino acid metabolism, Diet-replacement was not effective to restore gene profiles that were altered by HFHF, and to some extent unfortunately worsen these abovementioned genes disrupted by HFHF-induced disorder, thereby resulting in the skeletal muscle transcriptome that was substantially distinct from NC. Results from GESA analysis revealed hub genes that may play critical roles in the obesity (i.e., AMPK, MAPK, JAK-STAT) and exercise -induced weight loss, including MAPK, PPAR, p53 and Myh4.
3.3 Sucrose, TGS and IMO unequally affect muscular function and transcriptomics-identified pathways related with exercise-induced weight loss
We hypothesize that sucrose, TGS and IMO may cause varied impacts on exercise induced regulation on skeletal muscle transcriptome, which could explain the observed differential influences on glucose metabolism, anti-fatigue capacity and skeletal muscle morphology during weight loss. We then assessed the expressions of transcript genes and proteins in skeletal muscle that were related with HFHF-induced obesity and were regulated by exercise, as revealed by comparative transcriptomics analysis.
Specifically, as shown in Fig. 3a, compared with NC, HFHF increased mRNA expressions of Yap (Fold change HFHF/NC =3.07), IL-6 (Fold change HFHF/NC =4.63), TNF-α (Fold change HFHF/NC =2.20), Fis1 (Fold change HFHF/NC =3.41), Cytc (Fold change HFHF/NC =2.51), Myh2 (Fold change HFHF/NC =3.05) and Myh4 (Fold change HFHF/NC =3.35). By contrast, HFHF lowered AMPK, PI3K and Akt mRNA expressions relative to NC group (Fig. S2). These findings confirmed the severe inflammation and deteriorated muscle fiber composition induced by HFHF.
Exercise effectively downregulated gene expressions of Yap, IL-6, TNF-α, PI3K, Fis1, Cytc, Myh2 and Myh4 alongside with weight loss (Fig. 3a). Notably, drinking TGS- and sucrose- sweetened water exacerbated inflammation, as evidenced by higher expressions of Yap, IL-6 and TNF-α compared with HFHF (p < 0.05). Besides, in comparison with mice drinking tap water, mice drinking TGS- and sucrose- sweetened water showed higher expressions of Fis1, Cyct, Myh2 and Myh4, as well as lower PI3K. Consistently, TGS and sucrose intake eliminated exercise induced reduction in protein expressions of IL-6 (Fig. 3b, c) and FABP4 (Fig. 3b, e).
Moreover, drinking IMO-sweetened water had no additional impacts on mRNA expressions beyond the exercise, except for the synergistic effect with exercise on inhibiting HFHF induced abnormal changes in protein expressions of IL-6 (Fig. 3b, c), mTOR and HIF1α (Fig. 3d, f). Diet-replacement restored HFHF-induced inflammation and muscular dysfunction, but the effect size on regulating the genes and proteins was much weak than EX. In addition, strong correlations between biochemical indicators, proteins and genes were demonstrated.
We further validated distinct effects of sucrose, TGS and IMO on the lipid metabolism in C2C12 skeletal muscle cells. We found that cells treated with palmitic acid had higher TG and TC compared with NC. The addition of ALCAR with and without adding TGS did not influence TG and TC contents, while the addition of IMO reduced TG and TC, compared with HFHF (Fig. 3g, h). The Oil Red O staining results revealed that AICAR with and without IMO could alleviated the lipid accumulation in palmitic acid treated C2C12 myoblasts (Fig. 3i).
3.4 Sucrose and TGS weaken protection of exercise against weight regain-exacerbated inflammation and muscular dysfunctions
We investigated whether consuming sweeteners or sucrose- sweetened water during exercise may affect metabolic status and skeletal muscle of mice at the following weight regain. As expected, compared with NC, a 4-week follow-up HFHF feeding significantly increased body weight while worsen glucose homeostasis in mice underwent weight loss interventions (Fig. 4a-d).
We observed exercise-treated mice had better anti-fatigue capacity and larger cross-sectional area of skeletal muscle compared with mice continuously fed with HFHF, although no difference in body weight, fasting glucose and glucose tolerance was observed (Fig. 4e-i). The exercise-treated mice showed lower AMPK, IL-6, TNF-α, PI3K, Fis1, Cytc, Yap and Myh4 upon the follow-up HFHF compared with mice continuously fed with HFHF (Fig. 4o, and Fig. S3). Such regulatory effects were eliminated or weaken by TGS- and sucrose-sweetened water. Besides, drinking TGS, sucrose or IMO sweetened water did not influence exercise induced reduction in protein expression of FABP4 during weight regain (Fig. 4j, m). Consuming IMO was found to inhibit the protein expression of IL-6 during the weight regain period (Fig. 4k).