Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 77 of the 77 references it cites were found in a registry or an index; 44 of the 54 sources its claims cite were checked for support.

Dissociable roles of prefrontal plasticity in decision-making strategy and execution of habitual behavior

Open access
Asaoka, Nozomi, Pagano, Diane, Hayashi, Yasunori · 2026 · Nature Communications

Abstract

Habits are essential for sustaining adaptive behaviors but can also lead to maladaptive behaviors such as compulsive and addictive disorders. They emerge through a shift in decision-making from motivation-driven strategy to an automatic one. During this, the level of habit execution, such as frequency and duration, is superficially maintained despite a decline in motivational drive, raising the question of how the amount of execution is maintained even when decision-making strategies undergo substantial changes. By developing a unique paradigm capable of inducing habit formation within a defined time window in male mice, we found that shift in decision-making and amount of habit execution are controlled by plasticity in distinct cortical pathways. Erasure of each plasticity selectively altered the decision-making strategy or the execution level without affecting the other. These findings reveal a dual regulatory model for habit, providing insights into the neurocircuit mechanisms underlying both adaptive and maladaptive habits. Neural mechanisms underlying habits are not fully understood. By identifying specific synaptic adaptations, the authors demonstrate that distinct prefrontal cortical neurons differentially govern decision-making strategies and habit execution.

Overview

CoChat Summary

TL;DR

This mouse study argues that habitual behavior has two dissociable components: the decision-making strategy that becomes habitual, and the execution level of the behavior. Using a sequential VR→VI operant paradigm, the authors link goal-directed strategy to plasticity in RSC-projecting ACC neurons and execution level to plasticity in CS-projecting LOFC neurons.

Key findings

  • In the sequential VR→VI task, mice shifted from goal-directed to habitual behavior: after VR training, devaluation reduced lever pressing, but after 4 days of VI training, devaluation indices were close to zero; in contrast, 13 days of VR training (VR+VR) still showed goal-directed responding.
  • The execution index varied across mice in the VR+VI group, with a median of −0.16 (max 0.44, min −0.58, SD 0.21), whereas the VR+VR group had a median of 0.11 (max 0.71, min −0.09, SD 0.18), indicating that execution level changed independently of strategy.
  • Synaptic potentiation shifted across prefrontal regions: after CRF, AMPA/NMDA increased in MOFC L2/3; after VR training, increases were seen in ACC L5 and LOFC L5; after VR→VI switching, ACC L5 potentiation returned to control, while LOFC L5 potentiation remained elevated in the high-execution group and correlated with execution index.
  • Manipulating these circuits changed different aspects of habit: erasing LTP in RSC-projecting ACC neurons made behavior habitual (smaller devaluation index), while erasing LTP in CS-projecting LOFC neurons lowered execution index and reduced the proportion of VR+VIHigh mice; activating ACC could revert habitual behavior back to goal-directed behavior.

Methods

  • Male C57BL/6J mice were trained in a two-step operant paradigm: 3 CRF sessions, then 6 VR sessions (VR10 then VR20), followed by 4 VI60 sessions; the key comparison was the last VR vs the last VI session, and execution index was defined from the change in lever-press rate.
  • Behavioral readouts included devaluation tests, omission tests, plus additional T-maze spatial discrimination/reversal and marble burying assays; neural approaches included ex vivo electrophysiology, DREADD manipulations, optogenetic CALI/LTP erasure, and in vivo Ca2+ imaging.
  • Main analyses used AMPA/NMDA ratios, Ca2+ event classification, and linear mixed effects models with mouse ID as a random intercept; p-values were adjusted by Benjamini–Hochberg FDR.

Limitations

  • The paper states that in vivo Ca2+ imaging could not longitudinally track the same neurons across training days because of image-alignment and mechanical displacement issues.
  • The authors note potential off-target effects of CNO via conversion to clozapine, though they argue this is mitigated by consistency with CALI results.
  • It is evident from the text that the study used only male mice, which limits generalization to females; this is stated as a design choice to reduce estrous-cycle variability.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41467-026-75706-1

Cites

77 of 77 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 Individual differences in the execution of lever pressing during the two-step training. a Schematic diagram of the two-step operant training protocol. On day 10, the training was seamlessly switched from a variable ratio (VR) to a variable interval (VI) task (VR + VI). Mice without the VR → VI switching (VR + VR) are considered as the control group. b Transitions of lever pressing rate in the two-step training. VR + VR, N = 19; VR + VI, N = 79. c Mice underwent the first devaluation test after the 6th VR session. Then, after the 4th day of VI or VR training, the second devaluation test was performed. On each test day, mice were fed with regular chow, or sucrose solution. With free access to sucrose solution, its relative value declined (devalued condition). After 30-min free-access, mice were allowed to press the lever for 5 min, during which no sucrose was delivered. Numbers of lever presses on each test day in the VR + VR (d) and VR + VI (e) groups. f Within-subject differences in lever pressing between valued and devalued conditions, are shown as the devaluation index. d–f VR + VR, N = 13; VR + VI, N = 17. g Lever pressing activities before and after habit formation were compared in a within-subject manner. h Within-subject differences in the lever press rate between days 9 and 13 are plotted as the execution index. The VR + VI group was subdivided into the VR+VIHigh and VR+VILow groups by the median value of the execution index. i Lever press rate in the VR+VIHigh and VR+VILow groups. j Single session lever press rate on days 9 and 13 in the VR+VIHigh and VR+VILow groups. k Comparison of the lever press rate on days 9 and 13, reveal a significant difference in the distribution only on day 13. l The number of lever press bouts on days 9 and 13. VR → VI switching reduced lever pressing bout in the VR+VILow group. m, n Reward presentation and nose poke rates in the VR+VIHigh and VR+VILow groups. o Relations between the execution index and within-subject differences in the nose poke rate between days 9 and 13 are plotted in the XY-plane. h–o VR+VIHigh N = 40; VR+VILow N = 39. p, q Number of lever presses in each devaluation test in the VR + VI group (e) are separated into the VR+VIHigh and VR+VILow groups. p N = 9, q N = 8. r Within-subject differences in lever pressing between the valued and devalued conditions are shown as the devaluation index. Relations between the execution index and the devaluation index in the first and second devaluation tests are plotted in XY-planes. N = 17. **P < 0.01 and ***P < 0.001; N.S., not significant. For detailed statistics, see Supplementary Data 1.
Fig. 1 Individual differences in the execution of lever pressing during the two-step training. a Schematic diagram of the two-step operant training protocol. On day 10, the training was seamlessly switched from a variable ratio (VR) to a variable interval (VI) task (VR + VI). Mice without the VR → VI switching (VR + VR) are considered as the control group. b Transitions of lever pressing rate in the two-step training. VR + VR, N = 19; VR + VI, N = 79. c Mice underwent the first devaluation test after the 6th VR session. Then, after the 4th day of VI or VR training, the second devaluation test was performed. On each test day, mice were fed with regular chow, or sucrose solution. With free access to sucrose solution, its relative value declined (devalued condition). After 30-min free-access, mice were allowed to press the lever for 5 min, during which no sucrose was delivered. Numbers of lever presses on each test day in the VR + VR (d) and VR + VI (e) groups. f Within-subject differences in lever pressing between valued and devalued conditions, are shown as the devaluation index. d–f VR + VR, N = 13; VR + VI, N = 17. g Lever pressing activities before and after habit formation were compared in a within-subject manner. h Within-subject differences in the lever press rate between days 9 and 13 are plotted as the execution index. The VR + VI group was subdivided into the VR+VIHigh and VR+VILow groups by the median value of the execution index. i Lever press rate in the VR+VIHigh and VR+VILow groups. j Single session lever press rate on days 9 and 13 in the VR+VIHigh and VR+VILow groups. k Comparison of the lever press rate on days 9 and 13, reveal a significant difference in the distribution only on day 13. l The number of lever press bouts on days 9 and 13. VR → VI switching reduced lever pressing bout in the VR+VILow group. m, n Reward presentation and nose poke rates in the VR+VIHigh and VR+VILow groups. o Relations between the execution index and within-subject differences in the nose poke rate between days 9 and 13 are plotted in the XY-plane. h–o VR+VIHigh N = 40; VR+VILow N = 39. p, q Number of lever presses in each devaluation test in the VR + VI group (e) are separated into the VR+VIHigh and VR+VILow groups. p N = 9, q N = 8. r Within-subject differences in lever pressing between the valued and devalued conditions are shown as the devaluation index. Relations between the execution index and the devaluation index in the first and second devaluation tests are plotted in XY-planes. N = 17. **P < 0.01 and ***P < 0.001; N.S., not significant. For detailed statistics, see Supplementary Data 1.CC BY-NC-ND · PubMed Central
Fig. 2 Cortical region-specific plasticity is induced at each training stage. Acute brain slices of the anterior cingulate cortex (ACC), lateral orbitofrontal cortex (LOFC) and medial OFC (MOFC) were prepared within 30 min after the 3rd CRF and 6th VR training. Control (Ctrl) mice received food restriction but were not trained. AMPA/NMDA ratios were recorded from pyramidal neurons in the ACC layer 2/3 (a), ACC layer 5 (b), LOFC layer 2/3 (c), LOFC layer 5 (d), MOFC layer 2/3 (e) and MOFC layer 5 (f). a N = 10, 16 & 12; b N = 10, 18 & 9; c N = 9, 10 & 9; d N = 10, 9 & 9; e N = 9, 11 & 9; f N = 10, 9 & 10 for the Ctrl, CRF & VR groups, respectively. Acute brain slices of the ACC, LOFC and medial OFC were prepared within 30 min after the 4th VI training session. The VR + VR group was introduced as a control matched for total training days. AMPA/NMDA ratios were recorded from pyramidal neurons in the ACC layer 2/3 (g), ACC layer 5 (h), LOFC layer 2/3 (i), LOFC layer 5 (j), MOFC layer 2/3 (k) and MOFC layer 5 (l). The black and blue dashed lines show the mean value of the Ctrl and VR groups in (a–f), respectively. g N = 12, 9 & 9; h N = 9, 9 & 9; i N = 9, 9 & 10; j N = 12, 9 & 10; k N = 9, 9 & 9; l N = 8, 9 & 9 for theVR+VR, VR+VIHigh & VR+VILow groups, respectively. m–r Correlation between the execution index and AMPA/NMDA ratios in the VR+VIHigh and VR+VILow groups. N = 18–19. Individual data points are shown in (m–r). *P < 0.05; **P < 0.01; N.S., not significant. For detailed statistics, see Supplementary Data 1.
Fig. 2 Cortical region-specific plasticity is induced at each training stage. Acute brain slices of the anterior cingulate cortex (ACC), lateral orbitofrontal cortex (LOFC) and medial OFC (MOFC) were prepared within 30 min after the 3rd CRF and 6th VR training. Control (Ctrl) mice received food restriction but were not trained. AMPA/NMDA ratios were recorded from pyramidal neurons in the ACC layer 2/3 (a), ACC layer 5 (b), LOFC layer 2/3 (c), LOFC layer 5 (d), MOFC layer 2/3 (e) and MOFC layer 5 (f). a N = 10, 16 & 12; b N = 10, 18 & 9; c N = 9, 10 & 9; d N = 10, 9 & 9; e N = 9, 11 & 9; f N = 10, 9 & 10 for the Ctrl, CRF & VR groups, respectively. Acute brain slices of the ACC, LOFC and medial OFC were prepared within 30 min after the 4th VI training session. The VR + VR group was introduced as a control matched for total training days. AMPA/NMDA ratios were recorded from pyramidal neurons in the ACC layer 2/3 (g), ACC layer 5 (h), LOFC layer 2/3 (i), LOFC layer 5 (j), MOFC layer 2/3 (k) and MOFC layer 5 (l). The black and blue dashed lines show the mean value of the Ctrl and VR groups in (a–f), respectively. g N = 12, 9 & 9; h N = 9, 9 & 9; i N = 9, 9 & 10; j N = 12, 9 & 10; k N = 9, 9 & 9; l N = 8, 9 & 9 for theVR+VR, VR+VIHigh & VR+VILow groups, respectively. m–r Correlation between the execution index and AMPA/NMDA ratios in the VR+VIHigh and VR+VILow groups. N = 18–19. Individual data points are shown in (m–r). *P < 0.05; **P < 0.01; N.S., not significant. For detailed statistics, see Supplementary Data 1.CC BY-NC-ND · PubMed Central
Fig. 3 Chemogenetic manipulation of the ACC but not LOFC neurons affects habitual decision-making. a Mice received injection with AAV-CaMKII-hM4Di-mCherry in the ACC. On the next two days after the 6th VR training, the decision-making strategy was assessed by the devaluation test (first test). Before the free-access sessions, clozapine-N-oxide (CNO; 0.5 mg/kg) or its vehicle (VEH) was injected. After the first devaluation test, mice were re-trained with the VR schedule for 1 day. Then the second devaluation test was conducted with VEH or CNO injection in a cross-over design. b A representative image of hM4Di-mCherry expression in the ACC. c Effects of ACC inhibition on lever pressing in the devaluation test. N = 11. d Within-subject changes in the devaluation index of (c). e Mice received an injection with AAV-CaMKII-hM3Dq-mCherry in the ACC. The first devaluation test was performed following the 4th VI training session. After the first test, mice were re-trained with the VI schedule for one additional day, after which the second devaluation test was performed. Same as (a), drug injections were conducted in a cross-over design. f Representative image of hM3Dq-mCherry expression in the ACC. g Effects of ACC activation on lever pressing in the devaluation tests. N = 10. h Within-subject changes in the devaluation index of (g). i Mice received injection with AAV-CaMKII-hM4Di-mCherry in the LOFC. The experimental protocol was same as (a), except for the AAV injection sites. j Representative image of hM4Di-mCherry expression in the LOFC. N = 13. k Effects of the LOFC inhibition on lever pressing in the devaluation tests. l Within-subject changes in the devaluation index of (k). Individual data points are shown in (d, h, l). *P < 0.05; **P < 0.01. For detailed statistics, see Supplementary Data 1.
Fig. 3 Chemogenetic manipulation of the ACC but not LOFC neurons affects habitual decision-making. a Mice received injection with AAV-CaMKII-hM4Di-mCherry in the ACC. On the next two days after the 6th VR training, the decision-making strategy was assessed by the devaluation test (first test). Before the free-access sessions, clozapine-N-oxide (CNO; 0.5 mg/kg) or its vehicle (VEH) was injected. After the first devaluation test, mice were re-trained with the VR schedule for 1 day. Then the second devaluation test was conducted with VEH or CNO injection in a cross-over design. b A representative image of hM4Di-mCherry expression in the ACC. c Effects of ACC inhibition on lever pressing in the devaluation test. N = 11. d Within-subject changes in the devaluation index of (c). e Mice received an injection with AAV-CaMKII-hM3Dq-mCherry in the ACC. The first devaluation test was performed following the 4th VI training session. After the first test, mice were re-trained with the VI schedule for one additional day, after which the second devaluation test was performed. Same as (a), drug injections were conducted in a cross-over design. f Representative image of hM3Dq-mCherry expression in the ACC. g Effects of ACC activation on lever pressing in the devaluation tests. N = 10. h Within-subject changes in the devaluation index of (g). i Mice received injection with AAV-CaMKII-hM4Di-mCherry in the LOFC. The experimental protocol was same as (a), except for the AAV injection sites. j Representative image of hM4Di-mCherry expression in the LOFC. N = 13. k Effects of the LOFC inhibition on lever pressing in the devaluation tests. l Within-subject changes in the devaluation index of (k). Individual data points are shown in (d, h, l). *P < 0.05; **P < 0.01. For detailed statistics, see Supplementary Data 1.CC BY-NC-ND · PubMed Central
Fig. 4 LTP in the RSC-projecting ACC neurons during VR training is necessary for the maintenance of goal-directed decision-making. a Representative images of retrograde labeling. The retro AAV was injected in the central striatum (CS), retrosplenial cortex (RSC) and basolateral amygdala (BLA). For all three groups, tdTomato-positive neurons were detected in the ACC. Acute brain slices of ACC were prepared within 30 min after the 3rd CRF and 6th VR training. Control (Ctrl) mice received food restriction but were not trained. AMPA/NMDA ratios were recorded from CS-projecting ACC layer 5 neurons (b), RSC-projecting ACC layer 5 neurons (c) and BLA-projecting ACC layer 5 neurons (d). b N = 10, 9 &10; c N = 9, 9 & 9; d N = 9, 9 & 9 for the Ctrl, CRF & VR groups, respectively. Acute brain slices of the ACC were prepared within 30 min after the 4th VI training. The VR + VR group was introduced as a control matched for total training days. AMPA/NMDA ratios were recorded from CS-projecting ACC layer 5 neurons (e), RSC-projecting ACC layer 5 neurons (f) and BLA-projecting ACC layer 5 neurons (g). Black and blue dashed lines show the mean value of Ctrl and VR groups in (b–d), respectively. e N = 9, 10 & 9; f N = 10, 9 & 10; g N = 9, 9 & 9 for the VR + VR, VR+VIHigh & VR+VILow groups, respectively. h–j Correlation between execution index and AMPA/NMDA ratios in the VR+VIHigh and VR+VILow groups. N = 18–19. k Schematic diagram of VR training with CALI. Mice received CALI (illumination with 1–2 mW amber light for 1 min) within 5-min after each VR training session ended (days 4–9). l Representative image of CFL-SN expression and optic fiber implantation in the ACC. retroAAV-CaMKII-CFL-SN or retroAAV-CaMKII-SN were injected into the RSC and optic fibers implanted into the ACC. m Lever press rate during training with CALI. n Effects of CALI in RSC-projecting ACC neurons on lever pressing in the devaluation tests. o Within-subject changes in the devaluation index of (n). SN, N = 7; CFL-SN, N = 9. Individual data points are shown in (h–j). *P < 0.05; **P < 0.01; ***P < 0.001; N.S. not significant. For detailed statistics, see Supplementary Data 1.
Fig. 4 LTP in the RSC-projecting ACC neurons during VR training is necessary for the maintenance of goal-directed decision-making. a Representative images of retrograde labeling. The retro AAV was injected in the central striatum (CS), retrosplenial cortex (RSC) and basolateral amygdala (BLA). For all three groups, tdTomato-positive neurons were detected in the ACC. Acute brain slices of ACC were prepared within 30 min after the 3rd CRF and 6th VR training. Control (Ctrl) mice received food restriction but were not trained. AMPA/NMDA ratios were recorded from CS-projecting ACC layer 5 neurons (b), RSC-projecting ACC layer 5 neurons (c) and BLA-projecting ACC layer 5 neurons (d). b N = 10, 9 &10; c N = 9, 9 & 9; d N = 9, 9 & 9 for the Ctrl, CRF & VR groups, respectively. Acute brain slices of the ACC were prepared within 30 min after the 4th VI training. The VR + VR group was introduced as a control matched for total training days. AMPA/NMDA ratios were recorded from CS-projecting ACC layer 5 neurons (e), RSC-projecting ACC layer 5 neurons (f) and BLA-projecting ACC layer 5 neurons (g). Black and blue dashed lines show the mean value of Ctrl and VR groups in (b–d), respectively. e N = 9, 10 & 9; f N = 10, 9 & 10; g N = 9, 9 & 9 for the VR + VR, VR+VIHigh & VR+VILow groups, respectively. h–j Correlation between execution index and AMPA/NMDA ratios in the VR+VIHigh and VR+VILow groups. N = 18–19. k Schematic diagram of VR training with CALI. Mice received CALI (illumination with 1–2 mW amber light for 1 min) within 5-min after each VR training session ended (days 4–9). l Representative image of CFL-SN expression and optic fiber implantation in the ACC. retroAAV-CaMKII-CFL-SN or retroAAV-CaMKII-SN were injected into the RSC and optic fibers implanted into the ACC. m Lever press rate during training with CALI. n Effects of CALI in RSC-projecting ACC neurons on lever pressing in the devaluation tests. o Within-subject changes in the devaluation index of (n). SN, N = 7; CFL-SN, N = 9. Individual data points are shown in (h–j). *P < 0.05; **P < 0.01; ***P < 0.001; N.S. not significant. For detailed statistics, see Supplementary Data 1.CC BY-NC-ND · PubMed Central
Fig. 5 Reward-related activation of RSC-projecting ACC neurons is potentiated during goal-directed action. a Schematic diagram of in vivo Ca2+ recording and a representative image of jGCaMP7s expression and GRIN lens implantation. retroAAV-syn-jGCaMP7s was injected into the RSC and GRIN lens was implanted into the ACC. Recordings were performed in the last CRF (day 3), VR (day 9) and VI (day 13) training sessions. b Representative heat map of Ca2+ transients around the first, second (left) and tenth (right) rewarded nose pokes during VR training (day 9). A subset of RSC-projecting ACC neurons was activated during reward consumption. c The red and blue highlighted zones show the time windows for calculation of reward and lever press-related neural response, respectively (see Methods). d The average activity of all recorded cells during lever pressing, rewarded nose pokes and non-rewarded nose pokes during VR training. e Population-level distribution of reward-related responses. The AUC of reward-related responses normalized by the duration of events (from the start to the end of rewarded nose pokes). Each dot represents the mean normalized AUC of an individual cell recorded within each training stage. CRF, N = 130; VR, N = 195; VR + VI, N = 197. The average temporal dynamics of reward-excited (f) and -inhibited (g) cells at the start and end of rewarded nose pokes and lever pressing during each training stage. h Fractions of reward-excited and reward-inhibited cells recorded at each training stage. i, j For reward-excited and reward-inhibited cells, the population-weighted normalized AUC were calculated by multiplying normalized AUCs of positive (i) or negative (j) reward-related responses by the fraction of the reward-excited and reward-inhibited cells. i CRF, N = 47; VR, N = 99; VR + VI, N = 69; j CRF, N = 71; VR, N = 59; VR + VI, N = 69. *P < 0.05; **P < 0.01; ***P < 0.001. For detailed statistics, see Supplementary Data 1.
Fig. 5 Reward-related activation of RSC-projecting ACC neurons is potentiated during goal-directed action. a Schematic diagram of in vivo Ca2+ recording and a representative image of jGCaMP7s expression and GRIN lens implantation. retroAAV-syn-jGCaMP7s was injected into the RSC and GRIN lens was implanted into the ACC. Recordings were performed in the last CRF (day 3), VR (day 9) and VI (day 13) training sessions. b Representative heat map of Ca2+ transients around the first, second (left) and tenth (right) rewarded nose pokes during VR training (day 9). A subset of RSC-projecting ACC neurons was activated during reward consumption. c The red and blue highlighted zones show the time windows for calculation of reward and lever press-related neural response, respectively (see Methods). d The average activity of all recorded cells during lever pressing, rewarded nose pokes and non-rewarded nose pokes during VR training. e Population-level distribution of reward-related responses. The AUC of reward-related responses normalized by the duration of events (from the start to the end of rewarded nose pokes). Each dot represents the mean normalized AUC of an individual cell recorded within each training stage. CRF, N = 130; VR, N = 195; VR + VI, N = 197. The average temporal dynamics of reward-excited (f) and -inhibited (g) cells at the start and end of rewarded nose pokes and lever pressing during each training stage. h Fractions of reward-excited and reward-inhibited cells recorded at each training stage. i, j For reward-excited and reward-inhibited cells, the population-weighted normalized AUC were calculated by multiplying normalized AUCs of positive (i) or negative (j) reward-related responses by the fraction of the reward-excited and reward-inhibited cells. i CRF, N = 47; VR, N = 99; VR + VI, N = 69; j CRF, N = 71; VR, N = 59; VR + VI, N = 69. *P < 0.05; **P < 0.01; ***P < 0.001. For detailed statistics, see Supplementary Data 1.CC BY-NC-ND · PubMed Central
Fig. 6 LTP in the CS-projecting LOFC neurons during VI training determined the execution of habitual responses. a Representative images for retrograde labeling of LOFC neurons. The retro AAV was injected in the central striatum (CS), retrosplenial cortex (RSC) and basolateral amygdala (BLA). Each image was taken from the same mouse shown in Fig. 4a. tdTomato-positive neurons were not detected in the LOFC, when the retroAAV vector was injected in the RSC. Acute brain slices of LOFC were prepared within 30 min after the 3rd CRF and 6th VR training. Control (Ctrl) mice underwent food restriction but were not trained. AMPA/NMDA ratios were recorded from CS-projecting LOFC layer 5 neurons (b) and BLA-projecting LOFC layer 5 neurons (c). b N = 10, 11 & 12; c N = 9, 9 & 9 for the Ctrl, CRF & VR groups, respectively. Acute brain slices of the LOFC were prepared within 30 min after the 4th VI training. The VR + VR group was introduced as a control matched for total training days. AMPA/NMDA ratios were recorded from the CS-projecting LOFC layer 5 neurons (d) and BLA-projecting LOFC layer 5 neurons (e). Black and blue dashed lines show the mean value of the Ctrl and VR groups in (b, c), respectively. d N = 10, 10 & 9; e N = 9, 9 & 9 for the VR + VR, VR+VIHigh & VR+VILow groups, respectively. f, g Correlation between the execution index and AMPA/NMDA ratios in the VR+VIHigh and VR+VILow groups. N = 18–19. h Schematic diagram of the two-step training protocol with CALI. Mice received CALI within 5-min after each VI training (day 10–13). i Representative image of CFL-SN expression and optic fiber implantation in the LOFC. The retroAAV-CaMKII-CFL-SN or retroAAV-CaMKII-SN was injected into the CS and optic fibers were implanted into the LOFC. j Execution index during VI training with CALI. k Execution index on day 13 and the fraction of VR+VIHigh and VR+VILow mice in each group. SN, N = 26; CFL-SN, N = 29. l Effects of CALI in the CS-projecting LOFC neurons on lever pressing in the devaluation test. SN, N = 10; CFL-SN, N = 10. m Within-subject changes in the devaluation index of (l). *P < 0.05; **P < 0.01; N.S., not significant. For detailed statistics, see Supplementary Data 1.
Fig. 6 LTP in the CS-projecting LOFC neurons during VI training determined the execution of habitual responses. a Representative images for retrograde labeling of LOFC neurons. The retro AAV was injected in the central striatum (CS), retrosplenial cortex (RSC) and basolateral amygdala (BLA). Each image was taken from the same mouse shown in Fig. 4a. tdTomato-positive neurons were not detected in the LOFC, when the retroAAV vector was injected in the RSC. Acute brain slices of LOFC were prepared within 30 min after the 3rd CRF and 6th VR training. Control (Ctrl) mice underwent food restriction but were not trained. AMPA/NMDA ratios were recorded from CS-projecting LOFC layer 5 neurons (b) and BLA-projecting LOFC layer 5 neurons (c). b N = 10, 11 & 12; c N = 9, 9 & 9 for the Ctrl, CRF & VR groups, respectively. Acute brain slices of the LOFC were prepared within 30 min after the 4th VI training. The VR + VR group was introduced as a control matched for total training days. AMPA/NMDA ratios were recorded from the CS-projecting LOFC layer 5 neurons (d) and BLA-projecting LOFC layer 5 neurons (e). Black and blue dashed lines show the mean value of the Ctrl and VR groups in (b, c), respectively. d N = 10, 10 & 9; e N = 9, 9 & 9 for the VR + VR, VR+VIHigh & VR+VILow groups, respectively. f, g Correlation between the execution index and AMPA/NMDA ratios in the VR+VIHigh and VR+VILow groups. N = 18–19. h Schematic diagram of the two-step training protocol with CALI. Mice received CALI within 5-min after each VI training (day 10–13). i Representative image of CFL-SN expression and optic fiber implantation in the LOFC. The retroAAV-CaMKII-CFL-SN or retroAAV-CaMKII-SN was injected into the CS and optic fibers were implanted into the LOFC. j Execution index during VI training with CALI. k Execution index on day 13 and the fraction of VR+VIHigh and VR+VILow mice in each group. SN, N = 26; CFL-SN, N = 29. l Effects of CALI in the CS-projecting LOFC neurons on lever pressing in the devaluation test. SN, N = 10; CFL-SN, N = 10. m Within-subject changes in the devaluation index of (l). *P < 0.05; **P < 0.01; N.S., not significant. For detailed statistics, see Supplementary Data 1.CC BY-NC-ND · PubMed Central
Fig. 7 Reward-related activity of CS-projecting LOFC neurons is positively correlated with performance changes during habit formation. a The retroAAV-syn-jGCaMP7s was injected into the CS and GRIN lens was implanted into the LOFC. Recordings were performed in the last CRF (day 3), VR (day 9) and VI (day 13) training sessions. b Representative heat map of Ca2+ transients during VR training (day 9). The right panel shows the enlarged version of the dotted area on the left. (c) Representative trace of Ca2+ transients around lever pressing and reward events. Cells A and B were activated during lever presses, but not at reward consumption. In contrast, cells C and D showed higher activity during reward consumption than at lever presses. The red and blue highlighted zones show the time windows for calculation of reward and lever press-related neural responses, respectively (see Methods). d The AUC of event-related responses was normalized by the duration of lever press bouts, or reward consumption. Each dot represents the mean normalized AUC of an individual cell recorded within each training stage. CRF, N = 306; VR, N = 387; VR + VI, N = 460. e The XY-planes represent lever press and reward-related normalized AUCs for each cell. Cells were classified into three subgroups according to the magnitude and preference of their responses toward lever press and reward events (see Methods). The pie charts represent fractions of each functional subgroup within the population at each training stage. f Average activity of lever press cells, reward cells and non-responsive cells during each type of event during VR training. g Average temporal dynamics and normalized AUCs of lever press cells to lever press events during each training stage. CRF, N = 121; VR, N = 132; VR + VI, N = 116. h Correlation between the execution index and population-weighted AUCs of the lever press cells to lever press events during each training stage. CRF, N = 121 cells/8 mice; VR, N = 132 cells/8 mice; VR + VI, N = 116 cells/9 mice. i Average temporal dynamics and normalized AUCs of the reward cells to reward events during each training stage. CRF, N = 80; VR, N = 138; VR + VI, N = 144. j Correlation between the execution index and population-weighted AUCs of the reward cells to reward events during each training stage. CRF, N = 80 cells/9 mice; VR, N = 138 cells/9 mice; VR + VI, N = 144 cells/9 mice. *P < 0.05; **P < 0.01; ***P < 0.001; N.S., not significant. For detailed statistics, see Supplementary Data 1.
Fig. 7 Reward-related activity of CS-projecting LOFC neurons is positively correlated with performance changes during habit formation. a The retroAAV-syn-jGCaMP7s was injected into the CS and GRIN lens was implanted into the LOFC. Recordings were performed in the last CRF (day 3), VR (day 9) and VI (day 13) training sessions. b Representative heat map of Ca2+ transients during VR training (day 9). The right panel shows the enlarged version of the dotted area on the left. (c) Representative trace of Ca2+ transients around lever pressing and reward events. Cells A and B were activated during lever presses, but not at reward consumption. In contrast, cells C and D showed higher activity during reward consumption than at lever presses. The red and blue highlighted zones show the time windows for calculation of reward and lever press-related neural responses, respectively (see Methods). d The AUC of event-related responses was normalized by the duration of lever press bouts, or reward consumption. Each dot represents the mean normalized AUC of an individual cell recorded within each training stage. CRF, N = 306; VR, N = 387; VR + VI, N = 460. e The XY-planes represent lever press and reward-related normalized AUCs for each cell. Cells were classified into three subgroups according to the magnitude and preference of their responses toward lever press and reward events (see Methods). The pie charts represent fractions of each functional subgroup within the population at each training stage. f Average activity of lever press cells, reward cells and non-responsive cells during each type of event during VR training. g Average temporal dynamics and normalized AUCs of lever press cells to lever press events during each training stage. CRF, N = 121; VR, N = 132; VR + VI, N = 116. h Correlation between the execution index and population-weighted AUCs of the lever press cells to lever press events during each training stage. CRF, N = 121 cells/8 mice; VR, N = 132 cells/8 mice; VR + VI, N = 116 cells/9 mice. i Average temporal dynamics and normalized AUCs of the reward cells to reward events during each training stage. CRF, N = 80; VR, N = 138; VR + VI, N = 144. j Correlation between the execution index and population-weighted AUCs of the reward cells to reward events during each training stage. CRF, N = 80 cells/9 mice; VR, N = 138 cells/9 mice; VR + VI, N = 144 cells/9 mice. *P < 0.05; **P < 0.01; ***P < 0.001; N.S., not significant. For detailed statistics, see Supplementary Data 1.CC BY-NC-ND · PubMed Central
Fig. 8 Dynamic regulation of cortical region-specific synaptic potentiation orthogonally controls habitual behavior. (1) LTP in RSC-projecting ACC neurons is necessary for the predominance of goal-directed system for action selection, and as it weakens, behavior becomes habitual. (2) By contrast, LTP in CS-projecting LOFC neurons serves to maintain the execution of the habitual behavior at a comparable level to its previous, goal-directed execution.
Fig. 8 Dynamic regulation of cortical region-specific synaptic potentiation orthogonally controls habitual behavior. (1) LTP in RSC-projecting ACC neurons is necessary for the predominance of goal-directed system for action selection, and as it weakens, behavior becomes habitual. (2) By contrast, LTP in CS-projecting LOFC neurons serves to maintain the execution of the habitual behavior at a comparable level to its previous, goal-directed execution.CC BY-NC-ND · PubMed Central

Verification

Claims the paper makes

21 Valid · 0 Not valid · 1 Not checked · 91 Own results and interpretations

52 more claims

Create a free account to see all 113 claims and every source verdict.

See all claims — free account
  • It may state two separate claims as one. 40
  • Probably not a substantial claim. 22
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Citations

Cites

Show all 76 references

Metadata

Authors
Asaoka, N., Pagano, D., & Hayashi, Y.
Affiliations
Kyoto University (JP)
Corresponding author
  • Asaoka, Nozomi
  • Hayashi, Yasunori
Venue
Nature Communications
Published
2026-07-27
Volume
17
Issue
1
ISSN
2041-1723 (ISSN-L)
OpenAlex
W7171429318
Cited by
0
References
76
FWCI
0.00
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • Institutional Repositories DataBase (IRDB) (submittedVersion, cc-by-nc-nd)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • PubMed Central (submittedVersion, cc-by-nc-nd)
  • Kyoto University Research Information Repository (Kyoto University) (submittedVersion, cc-by-nc-nd)
Funders
  • MEXT | Japan Society for the Promotion of Science (JP22K15291, JP25K02556, JP18H05434, JP22H04981)
  • Japan Agency for Medical Research and Development (JP24wm0625507)
  • Uehara Memorial Foundation
  • Takeda Science Foundation
  • Astellas Foundation for Research on Metabolic Disorders
  • Naito Foundation
  • NOVARTIS Foundation (Japan) for the Promotion of Science (NOVARTIS Foundation
  • Human Frontier Science Program (RGP0022/2013)
  • MEXT | JST | Core Research for Evolutional Science and Technology (JPMJCR20E4)
Topics
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neurotransmitter Receptor Influence on Behavior
  • Life Sciences › Neuroscience › Endocrine and Autonomic Systems › Neuroendocrine regulation and behavior
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory and Neural Mechanisms
Keywords
habit formation, prefrontal cortex, synaptic plasticity, decision strategies, automatism, compulsive behavior, addiction, cortical circuits
MeSH terms
Animals, Behavior, Animal, Decision Making, Habits, Male, Mice, Inbred C57BL, Motivation, Neuronal Plasticity, Prefrontal Cortex

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Dissociable roles of prefrontal plasticity in decision-making strategy and execution of habitual behavior. https://app.cochat.ai/papers/10.1038/s41467-026-75706-1
The paper
Asaoka, N., Pagano, D., & Hayashi, Y. (2026). Dissociable roles of prefrontal plasticity in decision-making strategy and execution of habitual behavior. Nature Communications. https://doi.org/10.1038/s41467-026-75706-1

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.