Studies investigating the neural mechanisms by which associations between cues and predicted outcomes control behavior often use associative learning frameworks to understand the neural control of behavior. These frameworks do not always account for the full range of effects that novelty can have on behavior and future associative learning. Here, in mice, we show that dopamine in the nucleus accumbens core is evoked by novel, neutral stimuli, and the trajectory of this response over time tracked habituation to these stimuli. Habituation to novel cues before associative learning reduced future associative learning, a phenomenon known as latent inhibition. Crucially, trial-by-trial dopamine response patterns tracked this phenomenon. Optogenetic manipulation of dopamine responses to the cue during the habituation period bidirectionally influenced future associative learning. Thus, dopamine signaling in the nucleus accumbens core has a causal role in novelty-based learning in a way that cannot be predicted based on purely associative factors.
This is a preview of subscription content, access via your institution
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$189.00 per year
only $15.75 per issue
Rent or buy this article
Get just this article for as long as you need it
Prices may be subject to local taxes which are calculated during checkout
All data in the manuscript or the supplementary material are available from the corresponding author upon request. Correspondence and requests for materials should be addressed to Erin S. Calipari.
Codes used for the analysis of the fiber photometry data are available at github.com/kutlugunes.
Harris, J. D. Studies on Nonassociative Factors Inherent in Conditioning (Williams & Wilkins, 1943).
Lubow, R. E. & Moore, A. U. Latent inhibition: the effect of nonreinforced pre-exposure to the conditional stimulus. J. Comp. Physiol. Psychol. 52, 415 (1959).
Kamprath, K. & Wotjak, C. T. Nonassociative learning processes determine expression and extinction of conditioned fear in mice. Learn. Mem. 11, 770–786 (2004).
Harris, J. D. Habituatory response decrement in the intact organism. Psychol. Bull. 40, 385–422 (1943).
Rescorla, R. A. Effects of US habituation following conditioning. J. Comp. Physiol. Psychol. 82, 137 (1973).
Pavlov, I. P. Conditioned Reflexes: An Investigation of the Physiological Activity of the Cerebral Cortex (Oxford Univ. Press, 1927).
Kutlu, M. G. et al. Dopamine release in the nucleus accumbens core signals perceived saliency. Curr. Biol. 31, 4748–4761.e8 (2021).
Lubow, R. E. Latent inhibition. Psychol. Bull. 79, 398 (1973).
Rescorla, R. & Wagner, A. in Classical Conditioning II: Current Research and Theory Vol. 2 (Appleton-Century-Crofts, 1972).
Mackintosh, N. J. A theory of attention: variations in the associability of stimuli with reinforcement. Psychol. Rev. 82, 276 (1975).
Pearce, J. M. & Hall, G. A model for Pavlovian learning: variations in the effectiveness of conditioned but not of unconditioned stimuli. Psychol. Rev. 87, 532 (1980).
Schmajuk, N. A., Lam, Y.-W. & Gray, J. A. Latent inhibition: a neural network approach. J. Exp. Psychol. Anim. Behav. Process. 22, 321 (1996).
Saunders, B. T., Richard, J. M., Margolis, E. B. & Janak, P. H. Dopamine neurons create Pavlovian conditioned stimuli with circuit-defined motivational properties. Nat. Neurosci. 21, 1072–1083 (2018).
Oleson, E. B., Gentry, R. N., Chioma, V. C. & Cheer, J. F. Subsecond dopamine release in the nucleus accumbens predicts conditioned punishment and its successful avoidance. J. Neurosci. https://doi.org/10.1523/JNEUROSCI.3087-12.2012 (2012).
Young, A. M. J., Joseph, M. H. & Gray, J. A. Latent inhibition of conditioned dopamine release in rat nucleus accumbens. Neuroscience 54, 5–9 (1993).
Joseph, M. H., Peters, S. L. & Gray, J. A. Nicotine blocks latent inhibition in rats: evidence for a critical role of increased functional activity of dopamine in the mesolimbic system at conditioning rather than pre-exposure. Psychopharmacol. (Berl.) 110, 187–192 (1993).
Morrens, J., Aydin, Ç., Janse van Rensburg, A., Esquivelzeta Rabell, J. & Haesler, S. Cue-evoked dopamine promotes conditioned responding during learning. Neuron 106, 142–153.e7 (2020).
Hall, G. & Channell, S. Context specificity of latent inhibition in taste aversion learning. Q. J. Exp. Psychol. 38, 121–139 (1986).
Killcross, S. & Balleine, B. Role of primary motivation in stimulus preexposure effects. J. Exp. Psychol. Anim. Behav. Process. 22, 32 (1996).
Westbrook, R. F., Jones, M. L., Bailey, G. K. & Harris, J. A. Contextual control over conditioned responding in a latent inhibition paradigm. J. Exp. Psychol. Anim. Behav. Process. 26, 157 (2000).
Hart, A. S., Rutledge, R. B., Glimcher, P. W. & Phillips, P. E. M. Phasic dopamine release in the rat nucleus accumbens symmetrically encodes a reward prediction error term. J. Neurosci. https://doi.org/10.1523/JNEUROSCI.2489-13.2014 (2014).
de Jong, J. W. et al. A neural circuit mechanism for encoding aversive stimuli in the mesolimbic dopamine system. Neuron https://doi.org/10.1016/j.neuron.2018.11.005 (2019).
Schultz, W., Dayan, P. & Montague, P. R. A neural substrate of prediction and reward. Science https://doi.org/10.1126/science.275.5306.1593 (1997).
Menegas, W., Babayan, B. M., Uchida, N. & Watabe-Uchida, M. Opposite initialization to novel cues in dopamine signaling in ventral and posterior striatum in mice. eLife https://doi.org/10.7554/eLife.21886 (2017).
Nolan, S. O. et al. Direct dopamine terminal regulation by local striatal microcircuitry. J. Neurochem. https://doi.org/10.1111/jnc.15034 (2020).
Cragg, S. J., Rice, M. E. & Greenfield, S. A. Heterogeneity of electrically evoked dopamine release and reuptake in substantia nigra, ventral tegmental area, and striatum. J. Neurophysiol. https://doi.org/10.1152/jn.1922.214.171.1243 (1997).
Mohebi, A. et al. Dissociable dopamine dynamics for learning and motivation. Nature https://doi.org/10.1038/s41586-019-1235-y (2019).
Patriarchi, T. et al. Ultrafast neuronal imaging of dopamine dynamics with designed genetically encoded sensors. Science 360, eaat4422 (2018).
Pearce, J. M. A model for stimulus generalization in Pavlovian conditioning. Psychol. Rev. 94, 61–73 (1987).
Budygin, E. A. et al. Aversive stimulus differentially triggers subsecond dopamine release in reward regions. Neuroscience https://doi.org/10.1016/j.neuroscience.2011.10.056 (2012).
Young, A. M. J. Increased extracellular dopamine in nucleus accumbens in response to unconditioned and conditioned aversive stimuli: studies using 1 min microdialysis in rats. J. Neurosci. Methods https://doi.org/10.1016/j.jneumeth.2004.03.003 (2004).
Lubow, R. E., Rifkin, B. & Alek, M. The context effect: the relationship between stimulus preexposure and environmental preexposure determines subsequent learning. J. Exp. Psychol. Anim. Behav. Process. 2, 38–47 (1976).
Day, J. J., Roitman, M. F., Wightman, R. M. & Carelli, R. M. Associative learning mediates dynamic shifts in dopamine signaling in the nucleus accumbens. Nat. Neurosci. 10, 1020–1028 (2007).
Bouton, M. E. Context, time, and memory retrieval in the interference paradigms of Pavlovian learning. Psychol. Bull. 114, 80–99 (1993).
Hall, G. & Honey, R. C. Contextual effects in conditioning, latent inhibition, and habituation: associative and retrieval functions of contextual cues. J. Exp. Psychol. Anim. Behav. Process. 15, 232 (1989).
Sokolov, E. N. Neuronal models and the orienting reflex. In The Central Nervous System and Behavior, 3rd Conference (ed. M. A. B. Brazier) 187–276 (1960).
Jacob, P. F. et al. Prior experience conditionally inhibits the expression of new learning in Drosophila. Curr. Biol. 31, 3490–3503.e3 (2021).
Corbit, L. H., Muir, J. L. & Balleine, B. W. The role of the nucleus accumbens in instrumental conditioning: evidence of a functional dissociation between accumbens core and shell. J. Neurosci. 21, 3251–3260 (2001).
Westbrook, R. F., Bond, N. W. & Feyer, A.-M. Short- and long-term decrements in toxicosis-induced odor-aversion learning: the role of duration of exposure to an odor. J. Exp. Psychol. Anim. Behav. Process. 7, 362–381 (1981).
Ayres, J. J. B., Philbin, D., Cassidy, S., Bellino, L. & Redlinger, E. Some parameters of latent inhibition. Learn. Motiv. 23, 269–287 (1992).
Schnur, P. & Lubow, R. E. Latent inhibition: the effects of ITI and CS intensity during preexposure. Learn. Motiv. 7, 540–550 (1976).
Rankin, C. H. et al. Habituation revisited: an updated and revised description of the behavioral characteristics of habituation. Neurobiol. Learn. Mem. 92, 135–138 (2009).
Delacasa, G. & Lubow, R. E. Latent inhibition in conditioned taste aversion: the roles of stimulus frequency and duration and the amount of fluid ingested during preexposure. Neurobiol. Learn. Mem. 64, 125–132 (1995).
Eshel, N., Tian, J., Bukwich, M. & Uchida, N. Dopamine neurons share common response function for reward prediction error. Nat. Neurosci. 19, 479–486 (2016).
Killcross, A. S., Dickinson, A. & Robbins, T. W. The on-baseline latent inhibition effect is not counterconditioning. Psychopharmacol. (Berl.) 118, 104–106 (1995).
Redgrave, P., Prescott, T. J. & Gurney, K. Is the short-latency dopamine response too short to signal reward error? Trends Neurosci. 22, 146–151 (1999).
Kakade, S. & Dayan, P. Dopamine: generalization and bonuses. Neural Netw. 15, 549–559 (2002).
Horvitz, J. C. Mesolimbocortical and nigrostriatal dopamine responses to salient non-reward events. Neuroscience 96, 651–656 (2000).
McDiarmid, T. A., Bernardos, A. C. & Rankin, C. H. Habituation is altered in neuropsychiatric disorders—a comprehensive review with recommendations for experimental design and analysis. Neurosci. Biobehav. Rev. 80, 286–305 (2017).
Teo, C. et al. Decreased habituation of midlatency auditory evoked responses in Parkinson’s disease. Mov. Disord. 12, 655–664 (1997).
Rey, R. D. et al. The effect of levodopa on the habituation of the acoustic-palpebral reflex in Parkinson’s disease. Electromyogr. Clin. Neurophysiol. 36, 357–360 (1996).
Kutlu, M. G. & Schmajuk, N. A. Solving Pavlov’s puzzle: attentional, associative, and flexible configural mechanisms in classical conditioning. Learn. Behav. https://doi.org/10.3758/s13420-012-0083-5 (2012).
Stalnaker, T. A. et al. Dopamine neuron ensembles signal the content of sensory prediction errors. eLife 8, e49315 (2019).
Sharpe, M. J. et al. Dopamine transients are sufficient and necessary for acquisition of model-based associations. Nat. Neurosci. https://doi.org/10.1038/nn.4538 (2017).
Parker, K. E. et al. A paranigral VTA nociceptin circuit that constrains motivation for reward. Cell 178, 653–671.e19 (2019).
Paxinos, G. & Franklin, K. B. J. The Mouse Brain in Stereotaxic Coordinates (Academic Press, 2008).
Sharpe, M. J. et al. Dopamine transients do not act as model-free prediction errors during associative learning. Nat. Commun. 11, 1–10 (2020).
Sharpe, M. J., Batchelor, H. M., Mueller, L. E., Gardner, M. P. H. & Schoenbaum, G. Past experience shapes the neural circuits recruited for future learning. Nat. Neurosci. 24, 391–400 (2021).
Yorgason, J. T., España, R. A. & Jones, S. R. Demon voltammetry and analysis software: analysis of cocaine-induced alterations in dopamine signaling using multiple kinetic measures. J. Neurosci. Methods 202, 158–164 (2011).
This work was supported by NIH grants no. KL2TR002245 to M.G.K.; nos. DA055380 and DA048931 to E.S.C.; no. GM07628 to J.E.Z.; and no. DA045103 to C.A.S.; as well as by funds from the VUMC Faculty Research Scholar Award to M.G.K.; the Pfeil Foundation to M.G.K.; the Brain and Behavior Research Foundation to M.G.K., E.S.C. and C.A.S.; the Whitehall Foundation to E.S.C.; and the Edward Mallinckrodt, Jr. Foundation to E.S.C. We thank J. Dunning for his technical support. The opinions expressed in this article are the authors’ own and do not reflect the views of the NIH/DHHS.
The authors declare no competing interests.
Peer review information
Nature Neuroscience thanks the anonymous reviewers for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended Data Fig. 1 Analysis of dopamine dynamics using fiber photometry.
a, Diagram showing the methods used for calculating area under the curve, peak height, time to baseline and tau. These analyses have been used extensively for defining the kinetics and dynamics of dopamine signals previously59. Area under the curve (AUC) is the total area from stimulus onset to the return to baseline. Peak height is the maximal amount of dopamine that is evoked by the stimulus over the entire trace. Time to baseline is the time in seconds that it takes for the signal to return to baseline following the peak. Tau is the time it takes to return to 2/3 of peak height. b, Representative traces for 470-nm excitation (dLight) and 405-nm excitation (isosbestic control) channels in an individual animal at baseline. c, Representative ΔF/F trace showing dopamine transients in the nucleus accumbens core.
Extended Data Fig. 2 Dopamine response to neutral cue during the second day of exposure.
Session 2 dopamine signal to repeated white noise presentations (6–7 presentations per animal; n = 5 mice). The first presentation of the neutral stimulus in session 2 evoked a smaller dopamine response compared to the first presentation of the neural cue in the first session (peak height for the first presentation of session 1 versus session 2; two-sided paired t-test, t4 = 2.429, P = 0.07, n = 5 mice). # P = 0.07. Data represented as mean ± s.e.m.
Extended Data Fig. 3 Pre-exposure to stimuli decreases positive dopamine responses during subsequent fear conditioning without affecting shock responses.
a, Averaged dopamine response (z-scores) during the CS+ and pre-exposed CS+ cues and footshocks in the first fear conditioning session. The music note represents the cue onset and the lightning symbol denotes the footshock onset. b, Fold change (in AUC) from average CS− values across 6 trials (two-sided nested ANOVA, F(2, 83) = 2.10, P = 0.1287). c, Percent change (in peak dopamine response) from CS− values across 6 trials (two-sided nested ANOVA, F(2, 83)= 3.91, P = 0.0239). Pre-exposure to the predictive cue does not affect dopamine response to the subsequent footshock. d, Averaged dopamine signal to footshocks following the CS+ and pre-exposed CS+ on fear conditioning session 1. e, Peak dopamine response to the footshock following a pre-exposed or non-pre-exposed cue during session 1 (two-sided nested ANOVA F(1,54) = 0.13, P = 0.3738), f, time for the signal to return to baseline following peak evoked by the footshock across trial types did not differ (two-sided nested ANOVA F(1,54) = 0.10, P = 0.7475) and g, tau also did not differ between groups (two-sided nested ANOVA F(1,54) = 0.71, P = 0.4040). Data represented as mean ± s.e.m. * P < 0.05. ns = not significant.
Extended Data Fig. 4 Latent inhibition: Dopamine responses to non-pre-exposed and pre-exposed stimuli do not differ in the absence of latent inhibition and converge following extensive experience.
a, Dopamine responses did not differ between the CS+ and pre-exposed CS+ for the animals that did not show latent inhibition. b, The peak heights (two-sided nested ANOVA, F(1, 21) = 0.61, P = 0.4449, n = 30 presentations; n = 5 mice), c, the time to return to baseline (two-sided nested ANOVA, F(1, 21) = 0.30, P = 0.5888, n = 30 presentations; n = 5 mice) and d, tau were not different between the CS+ and pre-exposed CS+ (two-sided nested ANOVA, F(1, 21) = 0.60, P = 0.4467, n = 30 presentations; n = 5 mice). e, In the mice that showed latent inhibition, the behavioral and dopamine differences disappeared. f, Freezing responses to the pre-exposed CS+, non-pre-exposed CS+ (CS+) and non-pre-exposed CS− (CS−) were measured on session 2 of a two session fear conditioning paradigm (RM ANOVA pre-exposure main effect, F(1.466,5.863) = 19.99, P = 0.0032), the difference between the CS+ and pre-exposed CS+ disappeared on the second conditioning session (Tukey post-hoc, P = 0.9979). Both the CS+ (Tukey post-hoc, P = 0.0034) and the pre-exposed CS+ (Tukey post-hoc, P = 0.0037) yielded a stronger freezing response compared to the CS−. g, Averaged dopamine responses to the CS+ and pre-exposed CS+ during session 2 over all trials. h, Dopamine responses did not differ between the CS+ and pre-exposed CS+ (nested ANOVA, F(1, 54) = 0.42, P = 0.8901, n = 30 presentations; n = 5 mice). i, The time to return to baseline was not different (nested ANOVA, F(1, 54) = 0.07, P = 0.7864, n = 30 presentations; n = 5 mice). j, Tau is another measure of dopamine clearance and is defined by the time in seconds for the signal to return to 2/3 of peak height. Tau was not different between the CS+ and pre-exposed CS+ (unpaired t-test, t58 = 0.27, P = 0.78, n = 30 presentations; n = 5 mice). In the absence of the latent inhibition effect, dopamine response to the pre-exposed and novel CS+ do not differ. Data represented as mean ± s.e.m. ** P < 0.01, ns = not significant.
Extended Data Fig. 5 Fear conditioning with additional trials yielded a negative dopamine response to the fear cues.
a, Averaged dopamine signal to fear cues during the first two versus last two CS+ trials in a separate group of C57BL6/J mice (n = 4). b, Dopamine response to the CS+ (area under the curve, AUC) following 6 trials of the latent inhibition experiment compared to the dopamine response to the CS+ in an additional group with extensive fear conditioning trials did not differ for the first 6 trials (RM ANOVA group × trial interaction F(2, 14)= 0.52, P = 0.60; main effect of group F(1, 7) = 0.12, P = 0.20) before becoming a negative response after the 9th trial. Data represented as mean ± s.e.m. ns = not significant.
Extended Data Fig. 6 Fewer pre-exposure presentations result in latent inhibition.
a, Mice (n = 8; 4 males, 4 females) received two sessions of pre-exposure rather than four. b, Fewer pre-exposure sessions still produced a latent inhibition effect (two-sided paired t-test t7 = 3.314, P = 0.0129). Data represented as mean ± s.e.m., * P < 0.05.
Extended Data Fig. 7 Validation of TH + cell-specific opsin expression.
Optogenetics studies were designed to test whether the latent inhibition effect is controlled by the NAc core dopamine response to the pre-exposed fear cue. a, Representative images showing the expression of ChR2 and TH in the VTA dopamine cell bodies. AAV9.rTH.PI.Cre.SV40 and AAV5.Ef1a.DIO.hchR2.eYFP or AAV5.Ef1a.DIO.eYFP was injected into the VTA to achieve specific expression of Chr2 in dopamine neurons. Specifically, AAV9.rTH.PI.Cre.SV40 injections resulted in Cre expression in all Tyrosine Hydroxylase (TH) positive cells within the VTA. By placing a fiberoptic above the NAc core, we were able to stimulate dopamine release from VTA projecting dopamine terminals in the NAc core. b, Representative images showing the expression of NpHR and TH in the VTA dopamine cell bodies using the same approach as described. AAV9.rTH.PI.Cre.SV40 and AAV5.hSyn.eNpHR.3.0.eYFP or AAV5.Ef1a.DIO.eYFP were injected into the VTA and a fiberoptic was placed in the NAc core. c, Schematic showing histologically verified fiber optic placements for all mice (n = 21 mice, 9 males, 12 females). d, Cell counts were completed within the VTA from the experiments using the TH-specific excitatory/inhibitory opsin strategy. About 75% of the Cre+ cells in the VTA were also TH + suggesting a significant portion of the ChR2 and NpHR cells were dopaminergic (two-sided paired t-test t22 = 8.96, P = 0.00000001). Data represented as mean ± s.e.m., **** P < 0.0001.
Extended Data Fig. 8 Optogenetic stimulation, but not inhibition, of dopaminergic terminals during inter-trial interval abolishes latent inhibition.
a, Mice (n = 12; 5 males, 7 females) underwent four sessions of pre-exposure where they received unpaired stimulations (ChR2) or inhibitions (NpHR) during inter-trial interval windows. b, Unpaired stimulation of the NAc core dopamine response abolished latent inhibition (two-way ANOVA cue × group interaction F(1,10) = 4.078, P = 0.071; Bonferroni multiple comparisons: ChR2 pre-exposed versus non-pre-exposed P = 0.973; NpHR pre-exposed versus non-pre-exposed P = 0.023) while inhibition of the terminals resulted in a latent inhibition effect. Data represented as mean ± s.e.m., * P < 0.05, ns = not significant.
Extended Data Fig. 9 The effect of the optogenetic inhibition and excitation of dopaminergic terminals disappears with additional fear conditioning training.
Freezing response to the CS+ and pre-exposed CS+ did not differ in the eYFP or ChR2 groups on the second session of fear conditioning (multiple comparison Ps > 0.05). Freezing to the CS+ was still greater than the freezing response to the pre-exposed CS+ at the end of the session 2 (two-way ANOVA cue × group interaction F(2,36) = 4.31, P = 0.02; multiple comparisons: NpHR pre-exposed CS+ versus CS+ P = 0.04). This suggest that the freezing response to all cues (pre-exposed and non-pre-exposed) reached the asymptotic level with additional training but the enhancing effect of dopamine inhibition during pre-exposure on latent inhibition persisted beyond the initial fear conditioning session. Data represented as mean ± s.e.m., * P < 0.05, ns = not significant.
Extended Data Fig. 10 Optogenetically stimulating VTA dopamine cell bodies during cue pre-exposure enhances subsequent associative learning for that stimulus.
a, Representative histology showing ChR2 expression in the VTA dopamine cells in the TH-Cre rats. Histology maps showing ChR2 and eYFP expression and fiber placements in the VTA. b, These experiments were designed to look at the effects of dopamine stimulations during the pre-exposure period when the cues are novel and have not yet acquired value. Ventral tegmental area (VTA) dopamine neurons were stimulated using a blue laser at the time of the cue presentation during pre-exposure sessions. c, Rats received 2 sessions of stimulus pre-exposure followed by a single session of appetitive conditioning without any stimulation. In the pre-exposure session, the auditory cue was presented in the absence of an outcome whereas in the conditioning sessions, both the pre-exposed and non-pre-exposed cues were followed by the delivery of a food pellet. d, Averaged responses (appetitive response = CS response − preCS response) for the eYFP group throughout the 6 conditioning trials (repeated measures session × group interaction ANOVA F(5,80) = 0.78, P = 0.56). e, The difference between the first trial responses to the pre-exposed and non-pre-exposed cues trended towards significance in the eYFP group (paired t-test, t8 = 2.13, P = 0.06, n = 9 rats). f, There was no difference between pre-exposed versus non-pre-exposed cue responses during the last 3 trials of the conditioning session in the eYFP group (paired t-test, t8 = 0.25, P = 0.80, n = 9 rats). g, Averaged responses for the ChR2 group throughout the 6 conditioning trials (repeated measures session × group interaction ANOVA F(5,70) = 2.42, P = 0.04). h, The difference between the first trial responses to the pre-exposed and non-pre-exposed cues did not differ in the ChR2 group (paired t-test, t7 = 1.11, P = 0.30, n = 8 rats). i, The pre-exposed cue responses were significantly higher compared to the non-pre-exposed cue responses during the last 3 trials of the conditioning session in the ChR2 group (paired t-test, t7 = 0.008, P = 0.02, n = 8 rats). This demonstrates that stimulation of the VTA dopamine cell body response to stimuli during pre-exposure enhances the learning of cue-reward associations in the subsequent appetitive conditioning training. Data represented as mean ± s.e.m., # P = 0.056, ** P < 0.01, ns = not significant.
Rights and permissions
About this article
Cite this article
Kutlu, M.G., Zachry, J.E., Melugin, P.R. et al. Dopamine signaling in the nucleus accumbens core mediates latent inhibition. Nat Neurosci 25, 1071–1081 (2022). https://doi.org/10.1038/s41593-022-01126-1