New research asks why some stressors, like exercise, make us stronger

This mouse study looks at how some oxidative stress created by a particular enzyme is critical for the positive effects of exercise on metabolic health and performance.


The Study

Skeletal muscle NOX4 is required for adaptive responses that prevent insulin resistance
Published: Science Advances, December 2021
Where: Monash University, Australia

Read The Study

The Takeaway

Chronic levels of stress and inflammation are harmful to our health and contribute to insulin resistance. However, low-level acute metabolic and inflammatory stress (i.e., what happens during exercise) may elicit counterregulatory processes that actually lower our chronic levels of stress and inflammation. In short, a little bit of stress helps make our bodies stronger, an example of a process known as hormesis.

Specifically, exercise alters the homeostasis of the skeletal muscle by increasing the production of reactive oxygen species and thus changing the redox state of the cell. It is well recognized (reviewed here and here) that altering the redox state of skeletal muscle leads to many of the critical metabolic and performance adaptations in response to exercise training.

This study attempts to tease out the molecular mechanisms by which reactive oxygen species (ROS)—a cause of damaging oxidative stress at high enough levels—can act as molecular signals in skeletal muscle to improve metabolic health and exercise capacity in mice. Specifically, the authors found that particular ROS increase antioxidant defenses, increase mitochondrial biogenesis, improve endurance capacity, and prevent diet-induced insulin resistance.

Since this study is in mice, there is only limited circumstantial evidence to indicate that a similar mechanism with similar importance occurs in human skeletal muscle. However, the theory that our bodies adapt to specific stresses is well supported. Additional human studies may show that some amount of oxidative stress from exercise is necessary for the beneficial hormetic effect.

What It Looked At

Going into the study, the researchers drew on several essential observations about ROS, exercise, and their relationship to metabolic health. First, a single bout of exercise increases insulin sensitivity, enhances glucose uptake within skeletal muscle, and stimulates ROS production. These ROS have known signaling molecules that may be responsible for the benefits of exercise, but the full extent of their role in exercise adaptations is unknown. By characterizing where the ROS come from and clarifying their role in exercise adaptations, we will better understand which ROS are good and which may be bad.

In humans, one might imagine targeted therapeutics or lifestyle interventions that allow the “good” levels of ROS to be created, while “bad” or excessive ROS are targeted with antioxidant nutrients or lifestyle interventions. This kind of treatment would be nuanced and targeted beyond our current understanding of ROS and their role in hormesis.

The researchers in this study focused on NOX4, a specific enzyme known to produce ROS inside skeletal muscle cells during exercise. They used several genetic mouse models, cell culture approaches, and analytical techniques. Let’s start with some of the research questions and strategies they took to find answers.

  • Does exercise increase the NOX4 enzyme and acute production of ROS?
  • Is the NOX4 enzyme necessary for exercise training adaptations such as increases in capacity and mitochondria?
  • Can adding ROS restore the beneficial effects of exercise without NOX4?
  • What are the downstream signals that ROS activate in skeletal muscle during exercise?
  • Do conditions of poor metabolic health (from old age or diet-induced obesity) show abnormalities in the NOX4 enzyme?
  • Do the beneficial effects of a single bout of exercise in enhancing metabolic health require NOX4 enzyme in skeletal muscle?
  • Because our muscle tissues also contain other cell types (like those comprising the blood vessels and supporting tissue), researchers wanted to look at whether their results would replicate in muscle-only cell culture to rule out an effect in non-muscle cells of skeletal muscle tissue.

To summarize, a lot of work went into this paper, all of it in mouse and cell culture models. The researchers used several approaches to link NOX4 in skeletal muscle to many of the beneficial adaptations with exercise.

What It Found

Overall, researchers found the following linked events:

In skeletal muscle, the enzyme NOX4 generates ROS, which leads to common exercise adaptations, such as increased antioxidant defense, improved insulin signaling, and mitochondrial biogenesis. Furthermore, aging and diet-induced obesity lower NOX4 in skeletal muscle. Further tests deleting key enzymes proved that ROS, and not some other factor, was the active component in these benefits. Thus, the hormetic effect of exercise stress in mice is in part carried out by ROS from NOX4.

If you just want to know why all this acronym soup matters, skip to the next section. For those who love experimental details, here’s a bit more on the findings below.

  • After exercise, skeletal muscle NOX4 enzyme increases. Both a single bout of exercise and five weeks of exercise increases NOX4 (mRNA and protein).
  • The NOX4 enzyme in skeletal muscle produces ROS during exercise.
  • Mice that lack NOX4 in skeletal muscle have impaired exercise performance and less mitochondrial. Aerobic capacity (VO2max) was inherently lower without exercise training, while the time to fatigue at a constant speed was shorter both before and after exercise training. Researchers tried to reverse the adverse effects of NOX4 deletion by increasing ROS (H2O2) and were able to restore some of the increases in exercise performance—providing additional evidence that ROS are leading to improved exercise capacity and mitochondrial content.
  • ROS signal increases the activity of the transcription factor NFE2L2, which acts to increase antioxidant defense gene expression and mitochondrial biogenesis. Interestingly, researchers also used sulforaphane (found in cruciferous vegetables) to directly activate NFE2L2 and improve mitochondrial markers and time to fatigue in non-trained mice.
  • NOX4 levels decreased with age, and mice lacking NOX4 in muscle had worse metabolic health in old age, but not at younger ages. Older mice had impaired metabolic health measured by insulin tolerance tests and hyperinsulinemic-euglycemic clamps. Researchers could, however, partially restore insulin sensitivity with sulforaphane or by increasing ROS in these “NOX4 knockout” mice.
  • Diet-induced obesity reduces NOX4 in skeletal muscle, and lack of NOX4 in skeletal muscle made the effects of diet-induced obesity worse, including worse glucose disposal, insulin sensitivity, and lower levels of night energy expenditure. Furthermore, exercise training did not improve insulin sensitivity in mice lacking NOX4. However, exercise did enhance insulin sensitivity in mice with NOX4 in skeletal muscle—further evidence of the vital role of NOX4 in exercise adaptations.
  • Results from cell cultures were consistent with those from whole animals.

Why It Matters

Some stress is good; too much stress all the time is bad. That observation is what drives many of the complex relationships between stressors and health in which poor outcomes are at the extremes and optimal outcomes in the middle of the “U-shaped” curve (or bell-shaped depending on your labels). We know that chronic low-grade inflammation associated with oxidative stress is detrimental to overall health. Yet, transient increases in stress in the form of exercise, nutrient deprivation (fasting), cold showers, sauna, or consuming certain micronutrients lead to a sort of adaptation to the stress, or hormesis, which allow organisms to adapt to better deal with future stressors. This adaptation is a hallmark of many so-called “U-shaped” curves and may be an important concept for healthy aging.

Remaining Questions

We know that ROS can damage molecules in our cells, but also that ROS are necessary to activate the counterregulatory pathways to protect against future ROS or other oxidative damage. One initial question is, why are ROS generated from the NOX4 enzyme seemingly so important for signaling beneficial adaptations to this stress? Is it the levels of ROS produced? The location in the cell? The types of ROS? What is unique about NOX4 relative to other versions of NOX or other enzymes that also produce ROS?

What is the goldilocks (“just right”) amount of ROS for health adaptations, but no damage—and does this amount change as you adapt to higher and higher levels of ROS with more and more exercise? Besides exercise, what are some other similar inducers of hormesis: Phytochemicals in foods (such as broccoli, Brussel sprouts, and kale which have sulforaphane in them)? Ice baths? Heat? Small doses of radiation? And do other stressors result in similar metabolic adaptations?

A second central question is whether any work here will hold true in humans. NOX4 may be extremely important for mice living in their sterile cages, with their standard chow diets, but less so for humans living out in a real-world with a diverse diet. Many studies in mice are not replicated in humans. Perhaps in humans, NOX2 is the more critical enzyme, or there’s an entirely different mechanism.


The current paper shows that exercise-induced hormesis stems partly from the ROS generated by the enzyme NOX4 in mice. It provides additional support for the idea that exercise is stress that leads to adaptations that make us more resilient to future stress.