Skeletal muscle releases tiny bubbles into the circulation during exercise. These extracellular vesicles (EVs) preferentially travel to the liver, where they deliver protein and microRNA cargo. That finding came from a 2018 study of trained male cyclists, published in Cell Metabolism. The research team, led by Martin Whitham at the Garvan Institute of Medical Research, identified more than 300 proteins that rose in the EV fraction after a bout of cycling. Their work offered biological evidence for a long-suspected phenomenon: real-time signaling between muscle and metabolic organs during physical activity.
Why the muscle-to-liver signal has been hard to see
Skeletal muscle has been known for decades to release signaling molecules called myokines during exercise. These molecules mediate many benefits of physical activity, from improved insulin sensitivity to reduced inflammation and lower cardiovascular risk. But classical myokines were identified one at a time, and each required intact receptors on target cells. What has been harder to show is how the wider suite of exercise-induced signals is packaged and delivered to distant organs together. EVs offered a candidate mechanism. Before Whitham et al., however, no study had shown definitively that exercise increases circulating EVs in humans, or tracked where those vesicles end up.
What EVs are
EVs are small membrane-bound particles, typically 30 to 1,000 nanometers wide. Nearly all cell types release them. They fall into subclasses based on their origin. Exosomes form inside the cell, and are released when internal compartments fuse with the plasma membrane. Microvesicles bud directly off the cell surface. Both subclasses carry molecular cargo reflecting their parent cell. This cargo can include proteins, lipids, messenger RNA, and microRNAs. When an EV reaches a recipient cell, it can fuse with the cell membrane or be internalized. Either way, its contents are released into the cytoplasm. This makes them a candidate mechanism for coordinated communication between tissues. It is a different model from the classical one, in which a single molecule binds a single receptor.
How the study was designed
Whitham and colleagues studied a small group of healthy male cyclists in a laboratory setting. Participants completed a one-hour cycling test at increasing intensities. The protocol was 30 minutes at 55% of maximum oxygen consumption (VO2max), then 20 minutes at 70%, and finally about 10 minutes at 80%. Blood samples were drawn before and immediately after exercise. EVs were isolated from plasma using established laboratory techniques, and their protein content was characterized by high-resolution mass spectrometry. In separate experiments in mice, fluorescently labeled human plasma EVs collected after exercise were injected intravenously, and their tissue distribution was tracked using intravital imaging.
What the study found
Exercise produced a rapid, marked increase in circulating EVs. Proteomic analysis identified more than 300 proteins whose concentrations rose in the EV fraction after exercise. These spanned muscle-specific markers, metabolic enzymes, and immune signaling molecules. Bioinformatic analysis suggested that many of these proteins originated from skeletal muscle. In a separate experiment, labeled exercise-induced EVs from humans were injected into mice. The vesicles showed pronounced tropism for the liver, where they accumulated within minutes. Complementary experiments in cell culture showed that EV cargo could be taken up by hepatocytes, and that recipient cells changed their gene expression in response.
Subsequent studies from other groups have added to and refined this picture. A 2022 study, also of exercise-induced plasma EVs in humans, identified specific microRNAs whose EV-associated levels tracked with insulin sensitivity and body fat. One of these was miR-652-3p. Skeletal muscle-derived EVs, in animal models, have been shown to improve insulin sensitivity in recipient tissues. They also modulate lipid metabolism and glucose control.
Why this matters for diabetes and cardiometabolic disease
The link between exercise and improved blood sugar control has been clear for decades. Regular physical activity reduces the risk of developing type 2 diabetes by roughly a quarter, and it improves insulin sensitivity even before measurable weight loss. What has been unclear is how much of this benefit comes from the muscle itself. Some may come from fat lost through exercise. Some may come from circulating signals released during activity. The EV findings suggest that much of the exercise signal to the liver is packaged and delivered through vesicles. This signal is triggered acutely by physical activity, not just by long-term training adaptation.
How the effect size should be read
The 2018 study showed what exercise-induced EVs contain and where they go. It did not directly demonstrate that any specific EV cargo causes any specific metabolic benefit in humans. That causal chain is still being worked out. It runs from a muscle-released vesicle, to a liver-based response, to improved insulin sensitivity. Subsequent studies in mice have shown that engineered or exercise-derived EVs can improve glucose handling in recipient animals. Removing exercise-induced EVs also appears to blunt some benefits of training. Translation to humans is still preliminary. Individual variability in the EV response is substantial. The field is also still developing standard methods for isolating and characterizing plasma EVs.
What the evidence cannot yet answer
Several important questions remain open. Whether the acute EV response to a single exercise bout produces measurable long-term changes in liver or fat tissue is not established in humans. Whether EV-based tissue crosstalk contributes meaningfully to exercise-induced cardiovascular protection is also uncertain, independent of weight loss and blood pressure effects. And whether it will ever be possible to reproduce the metabolic benefits of exercise in a pill remains speculative. Such a therapy would be aimed at people who cannot exercise because of illness, injury, or age, delivering the right cargo of proteins and microRNAs without physical activity. A number of groups are trying to engineer synthetic vesicles for this purpose. No such therapy exists.
The practical takeaway
For anyone with normal access to physical activity, the practical implication is straightforward. Your muscles produce these signaling vesicles every time you exercise. Their effects appear to be part of why regular activity protects against type 2 diabetes and cardiovascular disease. Standard public health guidance still holds. About 150 minutes per week of moderate activity is enough to shift plasma biomarkers of metabolic health. So is 75 minutes of vigorous activity. Both can be spread across short bouts throughout the week. What the EV work adds is a mechanistic reason to believe that the benefit accumulates each time you exercise. It is not something that only builds up through slow adaptation.
The Whitham et al. authors concluded that exercise induces a substantial release of EVs into the circulation, and that these vesicles provide a mechanism for tissue crosstalk between skeletal muscle and metabolic organs during exercise. The observation has since seeded a growing area of research. The core question is whether the coordinated tissue signaling that exercise produces, together with the microRNA-carrying vesicles it releases, might one day be mimicked pharmacologically. If so, it could help people who cannot achieve those benefits through activity alone.
References
- Whitham M, Parker BL, Friedrichsen M, et al. Extracellular vesicles provide a means for tissue crosstalk during exercise. Cell Metabolism, 2018; 27(1): 237-251.e4. DOI: 10.1016/j.cmet.2017.12.001