A surgery model that separates fear from anxiety
The study asks whether voluntary exercise before surgery can prevent the negative emotional states that surgery leaves behind, and if so, through which brain circuit. To model this, mice received auditory fear conditioning, in which a tone is paired with a footshock, and then, six hours later, exploratory laparotomy under sevoflurane anesthesia. The surgery manipulated the colon, liver, and spleen and exposed the abdominal contents to air, but preserved limb movement.
Two behavioral readouts capture different things. Conditioned defensive expression is measured as freezing time in the contextual and cued fear conditioning tests, where the context or the tone alone can trigger a learned fear response. Anxiety-like behavior is measured in the open field test as time spent in the center and in the elevated plus maze as time and entries into the open arms. These are separate measures: a change in freezing does not by itself establish a change in open-arm exploration.
The model produced both. Compared with naive controls, mice that received fear conditioning plus surgery froze more in both tests and spent less time in the open field center and in the elevated plus maze open arms. Surgery alone also reduced center time and open-arm time and entries. Notably, surgery did not increase freezing in a trimethylthiazoline-induced innate fear test, so the authors interpret the freezing increase as enhanced conditioned defensive expression rather than a general change in threat reactivity.
Three weeks of running, not one or two
Exercised mice had free access to running wheels in their home cages for three weeks before fear conditioning; the paper reports running distance as wheel revolutions multiplied by π and 21 cm. Daily distance varied moderately between mice, with a coefficient of variation of 19.55%, but was more stable within a mouse, averaging 10.49%.
Three weeks of voluntary exercise significantly reduced both contextual and cued freezing in mice that underwent fear conditioning plus surgery. The duration mattered: three weeks, but not one or two weeks, significantly reduced contextual freezing, and cumulative running distance was negatively correlated with contextual freezing across the one-, two-, and three-week cohorts. The authors note that this association partly reflects between-group differences in exercise duration, so it is not a clean dose-response estimate within animals.
The same three-week regimen, again but not one or two weeks, increased center time in the open field and open-arm time and entries in the elevated plus maze. Total distance traveled in the open field did not differ across groups, which the authors use to argue that locomotor function was preserved after laparotomy. Exercise did not affect fear acquisition, and mice given surgery alone or exercise alone showed no freezing, indicating that the behavioral relief required the combination of the fear-conditioning context and surgery.
M1 glutamatergic neurons gain activity after exercise
The primary motor cortex, M1, is the starting point of the proposed mechanism. The authors first used c-Fos, an immediate early gene marker of recent neuronal activation, and found more c-Fos expression in M1 of exercised mice that had undergone fear conditioning plus surgery than in non-exercised counterparts. About 90% of the c-Fos signal colocalized with glutamate antibody signal, and about 80% of the glutamatergic signal colocalized with c-Fos in the exercised group.
Slice recordings supported the same direction: M1 glutamatergic neurons from exercised mice fired more frequently and needed less current to reach threshold, with resting membrane potential unchanged. In freely moving mice, fiber photometry of calcium signals showed that exercised mice had higher M1 glutamatergic calcium event frequency in the home cage, and this frequency correlated positively with running distance after one, two, and three weeks of exercise. Mice without preoperative exercise showed no pre- to post-surgical change in that frequency, which the authors read as laparotomy not affecting M1 glutamatergic activity on its own.
During the contextual fear conditioning and elevated plus maze tests, exercised mice showed a rapid and sustained rise in M1 glutamatergic calcium signals. The paper's interpretation is that exercise raises the baseline activity of these neurons and that emotional stress pushes it further, providing the drive for the downstream circuit.
An inhibitory detour through the zona incerta
The circuit the paper proposes has three nodes. M1 glutamatergic neurons send excitatory input to GABAergic neurons in the zona incerta, an inhibitory population, which in turn project to glutamatergic neurons in the posterior thalamic complex. Because the zona incerta neurons are inhibitory, their activation suppresses posterior thalamic complex activity. The authors describe this as a dominant inhibitory M1→ZI→PO pathway that competes with a direct excitatory M1→PO projection.
The evidence combines activity measurements with loss- and gain-of-function manipulations. In exercised mice, posterior thalamic complex glutamatergic neurons were suppressed. Ablating zona incerta neurons that project to the posterior thalamic complex increased freezing in both contextual and cued tests and reduced open-arm exploration in the elevated plus maze, while total open field distance was unchanged. Ablating zona incerta GABAergic neurons within the activated M1 circuit produced similar behavioral effects. Conversely, activating those zona incerta GABAergic neurons reduced freezing and increased open-arm measures.
Electrophysiology in posterior thalamic complex glutamatergic neurons adds a cellular link: after ablating zona incerta neurons in the activated circuit, these neurons showed a higher firing rate and a higher rheobase. The authors conclude that the indirect inhibitory pathway is both necessary and sufficient for the exercise-induced relief of fear- and anxiety-like behavior. They also state a limit: the direct M1→PO projection was not functionally manipulated, so its contribution cannot be formally excluded.
Norepinephrine raises the circuit's sensitivity
The paper next asks what makes the circuit respond more strongly after exercise. Using a fluorescent norepinephrine sensor expressed in M1, the authors found that exercised mice had higher spontaneous norepinephrine signal during fear conditioning and during the elevated plus maze, measured as mean signal and area under the curve.
The signal tracked behavior within the tests. In exercised mice, norepinephrine fluorescence differed between mobility and freezing states during fear conditioning and between closed and open arms in the elevated plus maze; in non-exercised mice, these state differences were not significant. The authors interpret this as exercise enhancing the sensitivity of norepinephrine release rather than merely raising a constant level.
A pharmacological test supports a causal role. Injecting an α1 receptor blocker, prazosin, plus a β receptor blocker, propranolol, into M1 thirty minutes before testing changed behavior: it increased freezing in one fear conditioning measure and reduced open-arm time and entries in the elevated plus maze, while other measures, including cued freezing and open field behavior, were not significantly affected. The mixed pattern is important: the norepinephrine blockade did not reproduce the full exercise effect across every readout, and the authors present norepinephrine as a driver of the circuit rather than the only modulator.
What the circuit explains and what it does not
The paper's central claim is specific: in male mice, preoperative voluntary exercise activates an M1→ZI→PO inhibitory pathway whose activity is boosted by norepinephrine released in M1, and this pathway mediates relief of conditioned defensive expression and anxiety-like behavior after fear conditioning paired with laparotomy. The authors describe the pathway as converting motor-circuit activity into a top-down signal for emotional homeostasis.
Several boundaries sit close to that claim. All mechanistic experiments were conducted in male mice. A separate female cohort showed that fear conditioning plus surgery produced conditioned defensive expression and anxiety-like behavior and that three weeks of preoperative wheel running reduced them, with male and female data analyzed separately and not pooled; the authors therefore describe the behavioral benefit as not sexually dimorphic, but the circuit and norepinephrine manipulations were not repeated in females.
The manipulations also produced task-specific effects. The authors note that manipulating different brain regions gave different behavioral outcomes, which they attribute to several exercise-responsive regions acting through distinct downstream pathways, with the zona incerta integrating motor signals from M1 and other regions and engaging targets beyond the posterior thalamic complex. The paper's causal necessity and sufficiency claim is therefore anchored to the indirect M1→ZI→PO pathway and to the fear conditioning and elevated plus maze readouts, not to every exercise-related effect.
Why the indirect pathway is the priority
For readers weighing what this study changes, the most decision-relevant comparison is between the two routes from M1 to the posterior thalamic complex. The direct excitatory projection and the indirect inhibitory detour through the zona incerta push posterior thalamic complex output in opposite directions, and the exercised mice showed reduced posterior thalamic complex activity. That pattern, together with the ablation and activation results, is the paper's reason for prioritizing the inhibitory route.
The second comparison is timing and dose. The behavioral benefit appeared with three weeks of preoperative voluntary running, not with one or two weeks, and cumulative distance correlated with the outcome across cohorts. The paper does not establish a threshold distance or a minimum exercise intensity, and the correlation partly reflects group differences in duration, so the practical reading is that sufficient preoperative voluntary exercise, as operationalized here, was associated with relief rather than that any amount will do.
The third is the norepinephrine result. Exercise raised norepinephrine signal and made it track behavioral states, and blocking α1 and β receptors in M1 altered several, but not all, behavioral readouts. That partial pattern makes norepinephrine a plausible gain-control input to the circuit rather than a complete explanation. The paper's contribution is therefore a testable corticothalamic mechanism linking motor activity to postoperative emotional behavior in male mice, with the direct M1→PO projection and the female circuit left as open questions.
Sources used in this answer
Corticothalamic circuits drive voluntary physical exercise-mediated relief of surgery-induced negative emotions in mice
Three weeks of preoperative voluntary wheel running reduced conditioned freezing and anxiety-like behavior after fear conditioning paired with exploratory laparotomy in mice; in male mice, norepinephrine in M1 drove an inhibitory M1 glutamatergic→zona incerta GABAergic→posterior thalamic complex glutamatergic circuit that suppressed posterior thalamic complex activity.
