Juha J. Hulmi

Glucose is traditionally viewed as a substrate for ATP production and glycogen storage in skeletal muscle. Here, we review evidence that glucose also serves as a building block for biomass synthesis in proliferating muscle satellite (stem) cells and hypertrophying skeletal muscle fibers, drawing parallels to anabolic metabolic reprogramming in cancer cells. In cancer and other growing cells, increased glucose uptake, aerobic glycolysis (Warburg effect), and the TCA cycle provide substrates for anabolic pathways that generate macromolecules required for growth and proliferation. Radiotracer studies in mammalian cancer and muscle cells demonstrate that approximately 8%–15% of the cell dry mass originates from glucose. Mechanistic insights, obtained predominantly from cell culture models, indicate that glucose-derived glycolytic and TCA cycle intermediates provide substrates for serine synthesis and the pentose phosphate pathway, glycine and one-carbon metabolism, nonessential amino acid synthesis, nucleotide and lipid synthesis, as well as for epigenetic methylation and acetylation. We further review evidence that human resistance training, hypertrophy through muscle-specific expression of Akt1 in mice, loss or inhibition of myostatin/activin signaling, and other hypertrophy-inducing interventions improve glucose homeostasis under conditions of obesity, insulin resistance, and type 2 diabetes. We discuss the possibility that glucose incorporation into biomass contributes to this effect.

Background: Aberrant NAD ⁺ metabolism has been implicated in the pathogenesis of cancer cachexia, highlighting this pathway as a potential therapeutic target to mitigate skeletal muscle wasting. However, it remains unclear whether chemotherapeutic agents contribute to the onset of cachexia by disrupting NAD ⁺ metabolism. Here, we investigated the effects of commonly used chemotherapy regimens on NAD ⁺ metabolism in skeletal muscle and liver of healthy mice. Methods: Healthy mice were treated with either 2-week regimens of folfiri or cisplatin, or 5-week regimens of folfiri or folfox, with vehicle-treated mice serving as controls. Cachexia-related outcomes were assessed, while skeletal muscle and liver tissues were analyzed for NAD metabolites and markers of NAD ⁺ metabolism. Given the consistent downregulation of the NAD ⁺ biosynthetic enzyme Nrk2 in cachectic chemotherapy-treated mice, we examined skeletal muscle Nrk2 / NRK2 expression across published murine and human cachexia datasets, and in additional models of muscle wasting and hypertrophy. Results: NAD ⁺ loss was observed in atrophic muscle following administration of cisplatin (2-week treatment; −14% vs controls, p=0.047) and folfiri (5-week treatment; −18%, p=0.069). In contrast, muscle NAD ⁺ levels were preserved in non-atrophic groups (2-week folfiri and 5-week folfox). Muscle Nrk2 was the most responsive NAD ⁺ biosynthetic enzyme, showing consistent downregulation across chemotherapy models with ongoing or developing muscle loss: cisplatin (−93%, p+ levels declined after prolonged 5-week folfiri (−20%, p=0.013) and folfox (−15%, p=0.043) treatments. These changes were accompanied by distinct alterations in NAD ⁺ biosynthesis pathways, indicating treatment-specific reorganization of hepatic NAD ⁺ metabolism. Cross-study analyses revealed early and consistent skeletal muscle Nrk2 downregulation across multiple murine cachexia models and human inactivity studies, whereas cachexia-targeted interventions in rodents and resistance training in humans increased its expression. Conclusions: These findings demonstrate that chemotherapy distrupts tissue NAD ⁺ metabolism, with skeletal muscle NAD ⁺ loss accompanying muscle atrophy and hepatic NAD ⁺ levels declining after prolonged treatment. The early and robust responsiveness of muscle Nrk2 expression to changes in muscle mass underscores its potential as a dynamic indicator for predicting treatment-induced changes in muscle mass. Together, these results provide new molecular insight into the metabolic basis of chemotherapy-induced muscle wasting and support further investigation of NAD ⁺ -targeted strategies in this context.

Skeletal muscles and blood vessels are continuously exposed to mechanical forces, particularly during exercise. We subjected human endothelial and skeletal muscle cells to cyclic mechanical stretch to mimic exercise and investigated acute molecular responses. Mechanical loading elicited both shared and cell type‐specific alterations in transcriptomic and metabolomic profiles, several of which mirrored changes observed in vivo following exercise. Both cell types released acetate in response to mechanical loading, at least partly via reactive oxygen species–dependent mechanism. Interestingly, transcriptomic changes occurred in opposite directions in endothelial and muscle cells. For example, genes associated with the electron transport chain were repressed in endothelial cells but upregulated in skeletal muscle cells. In endothelial cells, mechanical loading promoted a transcriptomic shift indicative of increased barrier integrity and attenuated proliferation. Metabolic changes were more pronounced in endothelial cells, which exhibited increased serine biosynthesis from glucose, as demonstrated by ¹³ C‐(U)‐glucose tracing. Targeting phosphoglycerate dehydrogenase (PHGDH), a key enzyme in the serine synthesis pathway, underscored the role of serine biosynthesis in endothelial cell anabolism. These findings suggest that mechanical loading recapitulates several exercise‐induced effects in endothelial and muscle cells, and highlights a potential link between mechanical stimuli, serine synthesis, and endothelial cell quiescence.

The Inter-Individual Variation in Resistance Training Response Conference was hosted at the University of Jyväskylä, Finland November 19–21, 2025. This paper summarizes key themes that emerged across lectures and discussions. First, resistance training induces multidimensional adaptations at the tissue, muscle fiber, and ultrastructural levels, including radial muscle fiber hypertrophy through increased myofibril number, longitudinal growth through sarcomere addition throughout the length (not ends) of muscle fibers, and metabolic adaptations that emulate other models of rapid cell growth. Second, training program variables including weekly sets, volume-load, rest interval duration, and training proximity to failure meaningfully influence hypertrophic outcomes in the general population, whereas exercise selection can be flexible. Third, age as well as molecular signatures before and in response to training influence interindividual response heterogeneity. Finally, while interindividual variability in observed hypertrophic responses is considerable, delineating true inter-individual variability from random variation remains challenging. Hence, study design considerations that can be thoughtfully applied to enhance rigor include repeat validation trials, unilateral within-subject designs, minimum clinically important difference thresholds, and multivariate composite responder classifications. This paper aims to summarize conference highlights while also providing meaningful implications for both researchers and practitioners and advancing current thinking on heterogeneity in the resistance training response.

Human skeletal muscle cells increase glucose utilization for protein, lipid, and especially RNA synthesis during their growth. Resistance training-induced muscle growth increases the abundance of enzymes and metabolites needed for nucleotide, serine, and glycine synthesis. The results suggest an underappreciated role of glucose utilization for biomass production during human muscle cell growth.

Skeletal muscle atrophy in response to pathophysiological stimuli or disuse includes loss of muscle mass and strength. Targeting signaling pathways regulating muscle growth can counteract muscle loss, but also unwanted side-effects on muscle vascularization, oxidative metabolism and exercise tolerance have been reported. Here, we investigated if combined induction of angiogenesis and muscle hypertrophy can promote physiological muscle growth and improve muscle function to overcome the limitations of current hypertophic treatments. We used myostatin propeptide (Pro-MSTN) and vascular endothelial growth factor B (VEGF-B) gene therapies to increase muscle size and angiogenesis, respectively. Intramuscular and systemic adeno-associated viral vector (AAV) delivery was used to study their effects alone and in combination in healthy and diabetic mice. Single-cell RNA sequencing was used to investigate the effects on different cell types and on intercellular communication in the healthy mice. We demonstrate that in the healthy mice the i.m. delivery of VEGF-B rescued Pro-MSTN -induced capillary rarefaction and enhanced muscle growth in the combination group (VEGF-B+Pro-MSTN). The systemic combination treatment also improved body composition in the healthy mice and increased muscle mass and grip strength in the diabetic mouse model. The single-cell RNA sequencing data showed that among the non-myocytes, endothelial cells and pericytes responded the most to both treatments resulting in enhanced intercellular communication. Our findings demonstrate beneficial effects of the combined gene delivery of Pro-MSTN and VEGF-B on muscle growth and body composition. The results also decipher the contribution of various cell types and their crosstalk in skeletal muscle growth.

Skeletal muscles and blood vessels are continuously exposed to mechanical forces, particularly during exercise. We subjected human endothelial and skeletal muscle cells to cyclic mechanical stretch to mimic exercise and investigated acute molecular responses. Mechanical loading elicited both shared and cell type-specific alterations in transcriptomic and metabolomic profiles, several of which mirrored changes observed in vivo following exercise. Both cell types released acetate in response to mechanical loading, at least partly via reactive oxygen species -dependent mechanism. Interestingly, transcriptomic changes occurred in opposite directions in endothelial and muscle cells. For example, genes associated with the electron transport chain were repressed in endothelial cells but upregulated in skeletal muscle cells. In endothelial cells mechanical loading remodelled intercellular junctions, promoted a transcriptomic shift indicative of increased barrier integrity and attenuated proliferation. Metabolic changes were more pronounced in endothelial cells, which exhibited increased serine biosynthesis from glucose, as demonstrated by ¹³ C-(U)-glucose tracing. Targeting phosphoglycerate dehydrogenase (PHGDH), a key enzyme in the serine synthesis pathway, underscored the role of serine biosynthesis in endothelial cell anabolism. These findings suggest that mechanical loading recapitulates several exercise-induced effects in endothelial and muscle cells, and highlights a potential link between mechanical stimuli, serine synthesis, and endothelial cell quiescence.

We aimed to characterize and compare elite strength athletes and physique athletes and to investigate potential sex differences and the contribution of muscle size to strength in this unique population. We examined male and female strength athletes (n=12), physique athletes (n=13) and non-trained controls (n=14). Anthropometry, maximal strength and strength endurance, and muscle cross-sectional area (ultrasound) were measured. Statistical significance was set at p<0.05. Although strength athletes excelled in their own sports, generic dynamic (leg press and arm curl one repetition maximum) or isometric knee extension torque did not differ compared to physique athletes. However, strength athletes had better muscle power in vertical jumps, while the cross-sectional area of biceps brachii was higher and body fat was lower in physique athletes. Males were more muscular and stronger, whereas females performed more repetitions in a multiple-set leg press protocol. Athletes had better strength/cross-sectional area ratios than controls only in complex tasks, but not in simple strength tasks. Partial correlation analysis with sex and training background as covariates showed that the cross-sectional area moderately explained the variance in maximal strength. In conclusion, strength athletes and physique athletes differ in sport-specific muscle size and power. In these athletes, the muscle size contributes to muscle strength. Finally, females have better strength endurance than males, independent of the training background.

Investigating repeated resistance training (RT) separated by a training break enables exploration of the potential for a proteomic memory of RT‐induced skeletal muscle growth, i.e. retained protein adaptations from the previous RT. Our aim was to examine skeletal muscle proteome response to 10‐week RT (RT1) followed by 10‐week training cessation (i.e. detraining, DT), and finally, 10‐week retraining (RT2). Thirty healthy, untrained participants conducted either periodic RT (RT1‐DT‐RT2, n = 17) or a 10‐week no‐training control period (n = 13) followed by 20 weeks of RT (n = 11). RT included twice‐weekly supervised whole‐body RT sessions, and resting vastus lateralis biopsies were obtained every 10 weeks for proteomics analysis using high‐end dia‐PASEF's mass spectrometry. The first RT period altered 150 proteins (93% increased) involved in, for example, energy metabolism and protein processing compared to minor changes during the control period. The proteome adaptations were similar after the second RT compared to baseline demonstrating reproducibility in proteome adaptations to RT. Many of the proteins induced by RT1 were reversed towards baseline after detraining and increased again after retraining. These reversible proteins were especially involved in aerobic energy metabolism. Interestingly, several proteins which increased after RT1 remain elevated (i.e. retained) after detraining, including carbonyl reductase 1 (CBR1) and proteins involved in muscle contraction, cytoskeleton and calcium binding. Among the latter, calcium‐activated protease calpain‐2 (CAPN2) has been recently identified as an epigenetic muscle memory gene. We show that resistance training evokes retained protein levels even after 2.5 months of no training, which demonstrates a potential proteomic memory of resistance training‐induced muscle growth in human skeletal muscle. image Key points: Repeated resistance training in humans separated by a training break (i.e. detraining) enables the identification of temporal protein signatures over the training, detraining and retraining periods, as well as studying reproducibility of protein changes to resistance training. Muscle proteome adaptations were similar after a second period of resistance training, demonstrating reproducibility in proteome adaptations to earlier resistance training. Many of the proteins induced by resistance training were reversed towards baseline after detraining and increased again after retraining. These reversible proteins were especially involved in aerobic energy metabolism. Several proteins increased after resistance training remain elevated (i.e. retained) after detraining, including carbonyl reductase 1 (CBR1) and calcium‐binding proteins such as calpain‐2 (CAPN2), a recently identified epigenetic muscle memory gene. Human skeletal muscle experiences retained protein changes following resistance training persisting over 2 months, demonstrating a potential proteomic memory of resistance training‐induced muscle growth.

Competitive sports and sports nutrition, popular among amateur athletes aiming for a lean physique, have limited research on gut microbiota. We conducted a 46-week study to analyze the consequences of fat loss and diet restrictions in 23 fitness athletes who prepared for a physique competition. Body composition, dietary intakes, serum cytokines and chemokines, and fecal samples were analyzed. Fat loss through caloric restriction and aerobic exercise led to an increased phylogenetic diversity of gut microbiota and changes in the composition of gut microbiota, with Faecalibacterium, Lachnospiraceae, Bacteroides, and Intestinimonas showing altered abundances. Fat loss also changed the predicted microbial functions responsible for the metabolism of carbohydrates and amino acids. Consumption of energy, carbohydrates, fiber, vitamins and minerals, and various fatty acids decreased during the preparation for the competition, which was partly associated with changes in gut microbiota. Several cytokine levels decreased (IL1a, IL1b, IL10, and TFNα), and certain chemokine levels increased (GROa and RANTES). During the 23-week regain period after the competition, gut microbiota showed signs of recovery, with increased diversity compared to pre- and post-competition measurements. Most taxonomic changes returned to their baseline levels after the regain period. The study highlights the dynamic nature of gut microbiota and its response to fat loss and regain in non-obese fitness/physique competitors and provides novel insights into how competitive sports and sports nutrition can influence the gut ecosystem.