Exerkines: How Exercise Makes Our Organs Communicate

Key points

• A myokine is produced and released by muscle,
  whereas an exerkine is any signalling factor altered by
  exercise, regardless of its tissue of origin.
• Contraction-induced interleukin-6 remains the
  best-demonstrated human example of a muscle-derived systemic
  signal.
• Exerkines are not limited to proteins. They include lactate, BAIBA,
  bioactive lipids, microRNAs and extracellular vesicles.
• Exercise creates multidirectional communication among skeletal
  muscle, liver, adipose tissue, bone, heart, brain and immune cells.
• Many candidate exerkines have compelling mechanisms in animals but
  lack causal evidence of clinical benefit in humans.
• No isolated molecule reproduces the combined mechanical, metabolic,
  vascular and neural demands of exercise.

Muscle does more than
contract

Skeletal muscle was once viewed mainly as a mechanical organ that
consumed energy to generate movement. That description is no longer
sufficient. During contraction, muscle changes its metabolism, gene
expression and secretome. It releases signals that can act locally and,
in some cases, enter the circulation and influence distant organs.

Irisin brought this idea to public attention. As discussed in our
previous analysis, the existence of irisin in humans is now supported by
mass spectrometry, yet the magnitude of its exercise response, its
circulating concentrations and its clinical effects remain difficult to
establish. Irisin is only one part of a much larger signalling
system.

Multi-omic studies show that a single bout of exercise produces
coordinated changes in hundreds of proteins, metabolites, lipids,
hormones and immune mediators within minutes. In a densely sampled human
treadmill study, early responses involved energy metabolism, oxidative
stress and immunity, followed by signals related to repair and
remodelling. Exercise is therefore better understood as a
molecular choreography than as the action of one
hormone [1].

Myokines and
exerkines are not synonymous

A myokine is a protein or peptide synthesised and
released by muscle cells. It may act on the producing cell itself, on
neighbouring cells or, if it reaches the circulation in biologically
relevant amounts, on distant organs.

An exerkine is a broader concept: any signalling
molecule whose production or concentration changes in response to
exercise and that may contribute to exercise adaptation. Exerkines may
originate from:

• skeletal muscle;
• liver;
• white or brown adipose tissue;
• bone;
• heart and vascular endothelium;
• immune cells;
• or several tissues at the same time.

They may be cytokines, hormones, metabolites, bioactive lipids or
cargo carried by extracellular vesicles. Myokines are therefore one
subset of exerkines.

This distinction matters. Finding that a molecule rises in blood
after exercise proves neither that it came from muscle nor that it
mediates a health benefit.

Contraction-induced
IL-6: the strongest human model

Interleukin-6 is often described as a pro-inflammatory cytokine. That
description is appropriate in some chronic settings, but it is
incomplete during exercise.

A landmark study using arterial and venous sampling across an
exercising leg showed that contracting skeletal muscle could account for
the exercise-induced rise in plasma IL-6 [2]. The increase can be rapid
and is particularly marked during prolonged exercise and when muscle
glycogen availability is low. Concentrations then fall after exercise
stops.

In this acute context, IL-6 helps coordinate fuel availability. In a
human infusion experiment, IL-6 increased insulin-stimulated glucose
disposal, while cell experiments linked its metabolic effects to AMPK
activation [3]. Exercise-associated IL-6 also occurs alongside an
immunoregulatory response that includes IL-1 receptor antagonist and
IL-10.

Two situations must therefore be separated:

• a short-lived IL-6 pulse generated during muscle
  contraction;
• a chronic low-grade elevation associated with
  adiposity, inflammatory disease or production by other tissues.

The molecule is the same, but its source, duration and signalling
environment differ. It is too simplistic to label every rise in IL-6 as
harmful—or every exposure as beneficial.

Confidence level: established for acute muscle release during
human exercise; probable for some systemic metabolic effects;
insufficient to attribute the overall health benefits of exercise to
IL-6 alone.

Myostatin:
muscle can also send an inhibitory signal

Not every myokine promotes a desired adaptation. Myostatin, also
known as GDF-8, is produced mainly by skeletal muscle and restrains
muscle growth. A loss-of-function mutation in a child was associated
with striking muscular hypertrophy, providing unusually strong human
evidence for its biological role [4].

Exercise, inactivity, ageing and disease can alter myostatin–activin
signalling. Pharmacological inhibition often increases lean mass, but
more muscle volume does not automatically produce a proportional
improvement in strength, mobility or clinical outcomes.

Myostatin illustrates a broader principle: muscle-derived signals can
restrain or fine-tune a system rather than simply amplify an exercise
benefit.

Confidence level: the role in muscle-mass regulation is
firmly established; its contribution to the systemic benefits of
exercise is much less certain.

BAIBA: when a
metabolite becomes a messenger

Beta-aminoisobutyric acid, or BAIBA, is not a
cytokine. It is a small metabolite derived from pathways including
valine and thymine catabolism. Experimental work connected BAIBA with
muscle PGC-1α signalling, browning-related changes in white adipose
tissue and increased hepatic fatty-acid oxidation [5].

An experiment using human skeletal muscle subsequently showed that
BAIBA can be released during contraction [6]. This supports its status
as a candidate muscle-derived systemic metabolite. However, circulating
BAIBA is also shaped by hepatic and renal metabolism, and its
associations with cardiometabolic risk do not demonstrate that BAIBA
itself protects humans from disease.

Confidence level: release from contracting human muscle has
been shown experimentally; distant metabolic effects remain
predominantly preclinical; clinical efficacy is unproven.

Apelin,
myonectin and musclin: credible but unevenly validated

Apelin is produced by several tissues, including
skeletal muscle. In obese men completing endurance training, muscle
apelin expression increased and the change was associated with improved
insulin sensitivity. Resting plasma levels did not increase, however,
leaving open the possibility that much of its action was local [7].
Restoring apelin signalling improves several muscle outcomes in aged
mice, but this does not yet establish a therapy for human sarcopenia
[8].

Myonectin, or CTRP15, was identified in mice as a
muscle signal linking nutritional state with fatty-acid uptake in liver
and adipose tissue [9]. Its human physiology is substantially less
certain.

Musclin, also called osteocrin, rises with activity
in some animal models and may support endurance and muscle–bone–heart
communication. Translation to humans remains preliminary.

IL-15, Metrnl, SPARC, decorin and hundreds of other proteins have
also been proposed as myokines. For many, the evidence consists only of
increased muscle RNA, secretion from cultured myotubes or an effect
produced by a high recombinant dose. These findings generate hypotheses;
they do not establish physiological endocrine function.

Lactate and
succinate are not merely waste products

Lactate is a circulating fuel and a signalling molecule. It is
generated in large amounts during some forms of exercise, shuttled among
tissues and recognised by receptors including HCAR1. In mice, brain
HCAR1 signalling contributed to VEGF expression and exercise-induced
cerebral angiogenesis [10]. This supports metabolic muscle-to-brain
communication, but it does not show that lactate alone accounts for the
cognitive benefits of exercise in humans.

Succinate, an intermediate in the citric acid cycle,
can also leave contracting muscle. Experimental work suggests that
exercise-associated acidification facilitates its release and that
signalling through SUCNR1 on stromal cells contributes to muscle
remodelling [11]. The action described is mainly local or paracrine,
illustrating that an exerkine need not behave as a long-distance
circulating hormone.

Lac-Phe:
an exercise signal does not have to come from muscle

N-lactoyl-phenylalanine, or Lac-Phe, is formed from
lactate and phenylalanine. Its blood concentration rises markedly after
some forms of exercise, particularly when lactate production is high. In
obese mice, pharmacological Lac-Phe reduced food intake for a limited
period and attenuated weight gain [12].

An exercise-associated rise has also been observed in humans, but
important limitations remain:

• causal experiments on appetite and body weight are predominantly
  animal studies;
• Lac-Phe can be produced by cells expressing CNDP2 and is not
  established as a muscle-specific secretion;
• pharmacological dosing may not reproduce physiological
  concentrations;
• a transient change in appetite cannot summarise the metabolic
  effects of training.

Lac-Phe is therefore a promising exerkine, but neither a demonstrated
myokine nor a validated obesity treatment.

12,13-diHOME:
brown adipose tissue responds to exercise

12,13-diHOME is a bioactive lipid whose circulating
concentration rises rapidly with exercise. Experimental studies suggest
that it is released in part by brown adipose tissue and promotes
fatty-acid uptake by working muscle [13].

Here, the direction of communication is reversed: muscle is not
simply instructing other tissues. Adipose tissue helps meet the
metabolic demands of muscle. Exercise creates bidirectional coordination
among organs.

Confidence level: a circulating response has been observed in
humans and the mechanism is plausible; its quantitative importance and
long-term clinical relevance remain uncertain.

Liver and blood
signals may reach the brain

Two sets of experiments strengthened the idea that plasma from an
exercised organism carries biologically active information.

In aged mice, plasma from animals that had exercised improved
neurogenesis and selected cognitive outcomes in sedentary recipients.
The investigators identified GPLD1, produced mainly by
the liver, as a candidate mediator [14]. In another model, plasma from
exercised mice reduced neuroinflammation and improved memory, with
clusterin contributing to the effect [15].

These are important mechanistic studies, but they do not show that
exercise plasma, GPLD1 or clusterin improves cognition in humans. They
demonstrate that the brain can receive systemic consequences of exercise
without every mediator necessarily crossing the blood–brain barrier
itself.

Muscle
may protect the brain by transforming a metabolite

Inter-organ communication does not always require secretion of a new
hormone. Muscle can instead change the composition of circulating
metabolites.

In mice, muscle PGC-1α1 increased kynurenine aminotransferase
expression. These enzymes converted circulating kynurenine, which can
enter the brain, into kynurenic acid, which crosses the blood–brain
barrier much less readily. This muscle-based metabolic barrier protected
the animals from selected behavioural effects of chronic stress
[16].

The mechanism is elegant and biologically plausible. It does not
establish that exercise treats human depression through this pathway
alone. Depressive disorders are heterogeneous, and causal human evidence
remains limited.

Extracellular
vesicles: biological parcels

Cells release membrane-bound vesicles capable of carrying proteins,
lipids and RNA. During acute human exercise, the abundance and cargo of
some circulating extracellular vesicles change rapidly. Proteomic work
identified hundreds of associated proteins and subsequently showed
uptake of exercise-related vesicles by the mouse liver [17].

This provides a credible mechanism for long-distance communication,
but major challenges remain:

• several tissues release vesicles simultaneously;
• markers assigning a vesicle specifically to skeletal muscle lack
  sensitivity and specificity;
• isolation and quantification methods remain heterogeneous;
• altered circulating cargo does not by itself prove a functional
  effect in humans.

MicroRNAs carried by these vesicles might alter gene expression in
distant tissues. This is experimentally plausible, but its physiological
importance remains under investigation.

An expanding whole-body map

The Molecular Transducers of Physical Activity Consortium, or
MoTrPAC, examined the time course of endurance-training responses across
18 solid tissues and blood in male and female rats. After eight weeks,
adaptations were detected in almost every tissue studied, with distinct
time courses and, in some cases, sex-specific responses [18].

This map confirms that exercise is an organised systemic stress. It
also warns against simple conclusions: a pathway may rise in muscle but
fall in liver, emerge only after several weeks, or differ between males
and females. These animal data generate valuable hypotheses, but
confirmation in the consortium’s human cohorts is essential.

What
would prove that an exerkine is functionally important?

A convincing demonstration should ideally proceed through several
stages:

1. show a reproducible change during or after exercise;
2. identify the producing tissue rather than measuring plasma
   alone;
3. demonstrate actual secretion, not merely increased tissue RNA;
4. establish concentrations compatible with biological activity;
5. identify a receptor or target pathway in another organ;
6. block the signal and show loss of an exercise adaptation;
7. restore that adaptation by reintroducing the signal;
8. confirm the mechanism and clinical relevance in humans.

Very few candidates currently satisfy all these criteria.

Why are findings
difficult to reproduce?

Exerkine concentrations depend on many variables:

• endurance exercise, resistance training, intervals or eccentric
  work;
• intensity, duration and the muscle mass recruited;
• training status and muscle glycogen availability;
• age, sex, adiposity and metabolic health;
• sampling time, fasting, hydration and haemoconcentration;
• the delay between exercise and blood collection;
• analytical method and antibody specificity.

A molecule may peak during exercise, return to baseline within 30
minutes and trigger a tissue response hours later. Conversely, long-term
training may improve sensitivity to a signal without increasing its
resting concentration.

Physiological exposure must also be distinguished from
pharmacological administration. Many animal studies report effects after
a recombinant protein or metabolite dose far above the concentration
measured during exercise.

Can one molecule reproduce
exercise?

The idea of an “exercise pill” is appealing, particularly for people
who are immobilised, frail or affected by neuromuscular disease.
Exerkine analogues may eventually reproduce a targeted adaptation:
preserving a tissue, stimulating one metabolic pathway or reducing a
complication of inactivity.

Exercise, however, simultaneously combines:

• mechanical tension;
• energy expenditure;
• altered blood flow and vascular shear stress;
• autonomic nervous-system activation;
• mitochondrial remodelling;
• immune signals;
• interactions with sleep, food intake and circadian rhythms.

No isolated molecule is likely to reproduce this architecture.
Amplifying a supposedly beneficial exerkine or blocking an apparently
negative one could also disrupt a useful feedback system.

There is currently no exerkine measurement recommended for routine
clinical care, no validated dietary supplement that reproduces these
effects and no single biomarker capable of quantifying the “molecular
benefit” of an exercise session.

Evidence summary

Confidence level	Examples	Reasonable conclusion
Established	Contraction-induced IL-6; myostatin for muscle regulation	Skeletal muscle produces biologically active signals, and IL-6 enters the circulation during human exercise.
Probable	Lactate, BAIBA, apelin, selected extracellular vesicles	Systemic communication is credible and partly demonstrated, but the independent clinical importance of each signal remains uncertain.
Emerging	Lac-Phe, 12,13-diHOME, GPLD1, clusterin, myonectin, musclin	Mechanisms are promising and sometimes causal in animals, but human validation remains limited.
Hypothetical in clinical practice	A supplement or drug that globally reproduces exercise	No candidate currently reproduces the multisystem effects of physical activity.

Conclusion

Exercise does more than make muscle work. It changes the circulating
signals that temporarily reorganise whole-body physiology. Myokines
introduced this concept; exerkines expand it to metabolites, lipids,
vesicles and mediators produced by many organs.

Muscle-derived IL-6 remains the strongest human example. Lactate and
BAIBA show that metabolites can also carry information. Lac-Phe,
12,13-diHOME, GPLD1 and clusterin reveal communication among muscle,
adipose tissue, liver and brain, although their clinical roles remain to
be demonstrated.

The key message is not that one hidden molecule explains exercise. It
is almost the opposite: the benefits of physical activity
probably emerge from the timed coordination of many signals, produced by
several tissues and integrated across the whole organism.

Key references

1. Contrepois K, et al. Molecular Choreography of Acute Exercise.
   Cell. 2020;181:1112-1130.e16. PubMed
2. Steensberg A, et al. Production of interleukin-6 in contracting
   human skeletal muscles can account for the exercise-induced increase in
   plasma interleukin-6. J Physiol. 2000;529:237-242. Article
3. Carey AL, et al. Interleukin-6 increases insulin-stimulated glucose
   disposal in humans and glucose uptake and fatty acid oxidation in vitro
   via AMPK. Diabetes. 2006. PubMed
4. Schuelke M, et al. Myostatin mutation associated with gross muscle
   hypertrophy in a child. N Engl J Med. 2004;350:2682-2688. PubMed
5. Roberts LD, et al. β-Aminoisobutyric acid induces browning of white
   fat and hepatic β-oxidation and is inversely correlated with
   cardiometabolic risk factors. Cell Metab. 2014;19:96-108. PubMed
6. Barlow JP, et al. Beta-aminoisobutyric acid is released by
   contracting human skeletal muscle. Metabol Open. 2020;7:100053.
   PubMed
7. Besse-Patin A, et al. Effect of endurance training on skeletal
   muscle myokine expression in obese men: identification of apelin as a
   novel myokine. Int J Obes (Lond). 2014;38:707-713. PubMed
8. Vinel C, et al. The exerkine apelin reverses age-associated
   sarcopenia. Nat Med. 2018. PubMed
9. Seldin MM, et al. Myonectin (CTRP15), a novel myokine that links
   skeletal muscle to systemic lipid homeostasis. J Biol Chem.
   2012;287:11968-11980. PubMed
10. Morland C, et al. Exercise induces cerebral VEGF and angiogenesis
    via the lactate receptor HCAR1. Nat Commun. 2017;8:15557. PubMed
11. Reddy A, et al. pH-Gated Succinate Secretion Regulates Muscle
    Remodeling in Response to Exercise. Cell. 2020;183:62-75.e17.
    PubMed
12. Li VL, et al. An exercise-inducible metabolite that suppresses
    feeding and obesity. Nature. 2022;606:785-790. PubMed
13. Stanford KI, et al. 12,13-diHOME: An Exercise-Induced Lipokine that
    Increases Skeletal Muscle Fatty Acid Uptake. Cell Metab.
    2018;27:1111-1120.e3. PubMed
14. Horowitz AM, et al. Blood factors transfer beneficial effects of
    exercise on neurogenesis and cognition to the aged brain.
    Science. 2020;369:167-173. PubMed
15. De Miguel Z, et al. Exercise plasma boosts memory and dampens brain
    inflammation via clusterin. Nature. 2021;600:494-499. PubMed
16. Agudelo LZ, et al. Skeletal muscle PGC-1α1 modulates kynurenine
    metabolism and mediates resilience to stress-induced depression.
    Cell. 2014;159:33-45. PubMed
17. Whitham M, et al. Extracellular Vesicles Provide a Means for Tissue
    Crosstalk during Exercise. Cell Metab. 2018;27:237-251.e4. PubMed
18. MoTrPAC Study Group. Temporal dynamics of the multi-omic response to
    endurance exercise training. Nature. 2024;629:174-183. PubMed

---

This article summarises current scientific knowledge and does not
provide individual diagnostic or treatment advice.