Introduction

The brain-derived neurotrophic factor (BDNF) belongs to the neurotrophins family [[1](https://www.nature.com/articles/s41386-025-02274-1#ref-CR1 “Miranda M, Morici JF, Zanoni MB, Bekinschtein P. Brain-Derived Neurotrophic Factor: A Key Molecule for Memory in the Healthy and the Pathological Brain. Front Cell Neurosci. 2019;13:363 https://doi.org/10.3389/fncel.2019.00363

.“)]. BDNF plays a vital role in neuronal differentiation and maturation [[2](https://www.nature.com/articles/s41386-025-02274-1#ref-CR2 “Waterhouse EG, An JJ, Orefice LL, Baydyuk M, Liao GY, Zheng K, et al. BDNF promotes differentiation and maturation of adult-born neurons through GABAergic transmission. J Neurosci. 2012;32:14318–30. https://doi.org/10.1523/JNEUROSCI.0709-12.2012

.“)], synaptic plasticity, learning, and memory [[3](https://www.nature.com/articles/s41386-025-02274-1#ref-CR3 “Colucci-D’Amato L, Speranza L, Volpicelli F. Neurotrophic Factor BDNF, Physiological Functions and Therapeutic Potential in Depression, Neurodegeneration and Brain Cancer. Int J Mol Sci. 2020;21, https://doi.org/10.3390/ijms21207777

.“)]. BDNF is highly expressed in the adult rodent brain [[4](https://www.nature.com/articles/s41386-025-02274-1#ref-CR4 “Hofer M, Pagliusi SR, Hohn A, Leibrock J, Barde YA. Regional distribution of brain-derived neurotrophic factor mRNA in the adult mouse brain. EMBO J. 1990;9:2459–64. https://doi.org/10.1002/j.1460-2075.1990.tb07423.x

.“), [5](https://www.nature.com/articles/s41386-025-02274-1#ref-CR5 “Timmusk T, Palm K, Metsis M, Reintam T, Paalme V, Saarma M, et al. Multiple promoters direct tissue-specific expression of the rat BDNF gene. Neuron. 1993;10:475–89. https://doi.org/10.1016/0896-6273(93)90335-o

.“)]. The majority of BDNF is stored in presynaptic dense core vesicles and is released in an activity-dependent manner upon neuronal depolarization [[6](https://www.nature.com/articles/s41386-025-02274-1#ref-CR6 “Huang EJ, Reichardt LF. Trk receptors: roles in neuronal signal transduction. Annu Rev Biochem. 2003;72:609–42. https://doi.org/10.1146/annurev.biochem.72.121801.161629

.“)]. BDNF binds to tropomyosin receptor kinase B (TrkB) receptor, which activates PI3K/AKT, and/or PLC/PKC, and/or ERK1/2 signaling, leading to the activation of transcription [[7](https://www.nature.com/articles/s41386-025-02274-1#ref-CR7 “Reichardt LF. Neurotrophin-regulated signalling pathways. Philos Trans R Soc Lond B Biol Sci. 2006;361:1545–64. https://doi.org/10.1098/rstb.2006.1894

.“)] or translation [[8](https://www.nature.com/articles/s41386-025-02274-1#ref-CR8 “You H, Lu B. Diverse Functions of Multiple Bdnf Transcripts Driven by Distinct Bdnf Promoters. Biomolecules. 2023;13, https://doi.org/10.3390/biom13040655

.“)]. Dysregulation of BDNF signaling has been implicated in psychiatric disorders, such as depression [[9](https://www.nature.com/articles/s41386-025-02274-1#ref-CR9 “Autry AE, Monteggia LM. Brain-derived neurotrophic factor and neuropsychiatric disorders. Pharmacol Rev. 2012;64:238–58. https://doi.org/10.1124/pr.111.005108

.“)], schizophrenia [[9](https://www.nature.com/articles/s41386-025-02274-1#ref-CR9 “Autry AE, Monteggia LM. Brain-derived neurotrophic factor and neuropsychiatric disorders. Pharmacol Rev. 2012;64:238–58. https://doi.org/10.1124/pr.111.005108

.“)], and addiction [[10](https://www.nature.com/articles/s41386-025-02274-1#ref-CR10 “Barker JM, Taylor JR, De Vries TJ, Peters J. Brain-derived neurotrophic factor and addiction: Pathological versus therapeutic effects on drug seeking. Brain Res. 2015;1628:68–81. https://doi.org/10.1016/j.brainres.2014.10.058

.“)].

Only 10–15% of alcohol users develop AUD [[11](https://www.nature.com/articles/s41386-025-02274-1#ref-CR11 “National Survey on Drug Use and Health. https://www.samhsa.gov/data/sites/default/files/reports/rpt42728/NSDUHDetailedTabs2022/NSDUHDetailedTabs2022/NSDUHDetTabsSect5pe2022.html

.“)], implying that there are protective mechanisms that prevent the development of the disorder in the majority of the population. Using rodents as a model system, we and others, presented data to suggest that BDNF is part of a protective mechanism that gates the development of heavy alcohol intake and abuse (reviews [[12](#ref-CR12 “Ron D, Barak S. Molecular mechanisms underlying alcohol-drinking behaviours. Nat Rev Neurosci. 2016;17:576–91. https://doi.org/10.1038/nrn.2016.85

.“),[13](#ref-CR13 “Egervari G, Siciliano CA, Whiteley EL, Ron D. Alcohol and the brain: from genes to circuits. Trends Neurosci. 2021;44:1004–15. https://doi.org/10.1016/j.tins.2021.09.006

.“),[14](https://www.nature.com/articles/s41386-025-02274-1#ref-CR14 “Ron D, Berger A. Targeting the intracellular signaling “STOP” and “GO” pathways for the treatment of alcohol use disorders. Psychopharmacology. 2018;235:1727–43. https://doi.org/10.1007/s00213-018-4882-z

.“)]). For example, we found that the activation of BDNF/TrkB/ERK1/2 but not PI3K/AKT, or PLC/PKC signaling in the DLS, a region involved in habitual behavior [[15](https://www.nature.com/articles/s41386-025-02274-1#ref-CR15 “Lipton DM, Gonzales BJ, Citri A. Dorsal Striatal Circuits for Habits, Compulsions and Addictions. Front Syst Neurosci. 2019;13:28. https://doi.org/10.3389/fnsys.2019.00028

.“)], keeps alcohol intake in moderation by activating the transcription machinery [[16](#ref-CR16 “Jeanblanc J, He DY, McGough NN, Logrip ML, Phamluong K, Janak PH, et al. The dopamine D3 receptor is part of a homeostatic pathway regulating ethanol consumption. J Neurosci. 2006;26:1457–64. https://doi.org/10.1523/JNEUROSCI.3786-05.2006

.“),17,[18](#ref-CR18 “Jeanblanc J, He DY, Carnicella S, Kharazia V, Janak PH, Ron D. Endogenous BDNF in the dorsolateral striatum gates alcohol drinking. J Neurosci. 2009;29:13494–502. https://doi.org/10.1523/JNEUROSCI.2243-09.2009

.“),[19](https://www.nature.com/articles/s41386-025-02274-1#ref-CR19 “Jeanblanc J, Logrip ML, Janak PH, Ron D. BDNF-mediated regulation of ethanol consumption requires the activation of the MAP kinase pathway and protein synthesis. Eur J Neurosci. 2013;37:607–12. https://doi.org/10.1111/ejn.12067

.“)].

The prefrontal cortex (PFC) is the major source of BDNF in the striatum [[20](#ref-CR20 “Altar CA, Cai N, Bliven T, Juhasz M, Conner JM, Acheson AL, et al. Anterograde transport of brain-derived neurotrophic factor and its role in the brain. Nature. 1997;389:856–60. https://doi.org/10.1038/39885

.“),[21](#ref-CR21 “Baquet ZC, Gorski JA, Jones KR. Early striatal dendrite deficits followed by neuron loss with advanced age in the absence of anterograde cortical brain-derived neurotrophic factor. J Neurosci. 2004;24:4250–8. https://doi.org/10.1523/JNEUROSCI.3920-03.2004

.“),[22](#ref-CR22 “Baydyuk M, Xu B. BDNF signaling and survival of striatal neurons. Front Cell Neurosci. 2014;8:254. https://doi.org/10.3389/fncel.2014.00254

.“),[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)], and we and others, reported that escalation of alcohol intake results from a breakdown in BDNF signaling. Specifically, chronic high alcohol consumption attenuates BDNF expression in corticostriatal regions of rodents [[24](#ref-CR24 “Logrip ML, Janak PH, Ron D. Escalating ethanol intake is associated with altered corticostriatal BDNF expression. J Neurochem. 2009;109:1459–68. https://doi.org/10.1111/j.1471-4159.2009.06073.x

.“),[25](#ref-CR25 “Darcq E, Warnault V, Phamluong K, Besserer GM, Liu F, Ron D. MicroRNA-30a-5p in the prefrontal cortex controls the transition from moderate to excessive alcohol consumption. Mol Psychiatry. 2015;20:1219–31. https://doi.org/10.1038/mp.2014.120

.“),[26](https://www.nature.com/articles/s41386-025-02274-1#ref-CR26 “Tapocik JD, Barbier E, Flanigan M, Solomon M, Pincus A, Pilling A, et al. microRNA-206 in rat medial prefrontal cortex regulates BDNF expression and alcohol drinking. J Neurosci. 2014;34:4581–8. https://doi.org/10.1523/JNEUROSCI.0445-14.2014

.“)], which is mediated by the microRNA machinery in the medial prefrontal cortex (mPFC) [[25](https://www.nature.com/articles/s41386-025-02274-1#ref-CR25 “Darcq E, Warnault V, Phamluong K, Besserer GM, Liu F, Ron D. MicroRNA-30a-5p in the prefrontal cortex controls the transition from moderate to excessive alcohol consumption. Mol Psychiatry. 2015;20:1219–31. https://doi.org/10.1038/mp.2014.120

.“), [26](https://www.nature.com/articles/s41386-025-02274-1#ref-CR26 “Tapocik JD, Barbier E, Flanigan M, Solomon M, Pincus A, Pilling A, et al. microRNA-206 in rat medial prefrontal cortex regulates BDNF expression and alcohol drinking. J Neurosci. 2014;34:4581–8. https://doi.org/10.1523/JNEUROSCI.0445-14.2014

.“)], resulting in escalation of alcohol consumption [[25](https://www.nature.com/articles/s41386-025-02274-1#ref-CR25 “Darcq E, Warnault V, Phamluong K, Besserer GM, Liu F, Ron D. MicroRNA-30a-5p in the prefrontal cortex controls the transition from moderate to excessive alcohol consumption. Mol Psychiatry. 2015;20:1219–31. https://doi.org/10.1038/mp.2014.120

.“), [26](https://www.nature.com/articles/s41386-025-02274-1#ref-CR26 “Tapocik JD, Barbier E, Flanigan M, Solomon M, Pincus A, Pilling A, et al. microRNA-206 in rat medial prefrontal cortex regulates BDNF expression and alcohol drinking. J Neurosci. 2014;34:4581–8. https://doi.org/10.1523/JNEUROSCI.0445-14.2014

.“)]. We further found that the transition from moderate to heavy alcohol intake is also mediated by the alterations in the membranal localization of the BDNF receptors, TrkB and p75NTR in the DLS of rats [[27](https://www.nature.com/articles/s41386-025-02274-1#ref-CR27 “Darcq E, Morisot N, Phamluong K, Warnault V, Jeanblanc J, Longo FM, et al. The Neurotrophic Factor Receptor p75 in the Rat Dorsolateral Striatum Drives Excessive Alcohol Drinking. J Neurosci. 2016;36:10116–27. https://doi.org/10.1523/JNEUROSCI.4597-14.2016

.“)]. Finally, we reported that transgenic mice carrying a polymorphism within the BDNF gene that disrupts BDNF release [[28](https://www.nature.com/articles/s41386-025-02274-1#ref-CR28 “Egan MF, Kojima M, Callicott JH, Goldberg TE, Kolachana BS, Bertolino A, et al. The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell. 2003;112:257–69. https://doi.org/10.1016/s0092-8674(03)00035-7

.“)], compulsively drink alcohol, exhibit a reduction in the anxiolytic actions of alcohol and an increase in alcohol preference over social interaction in mice [[29](https://www.nature.com/articles/s41386-025-02274-1#ref-CR29 “Warnault V, Darcq E, Morisot N, Phamluong K, Wilbrecht L, Massa SM, et al. The BDNF Valine 68 to Methionine Polymorphism Increases Compulsive Alcohol Drinking in Mice That Is Reversed by Tropomyosin Receptor Kinase B Activation. Biol Psychiatry. 2016;79:463–73. https://doi.org/10.1016/j.biopsych.2015.06.007

.“), [30](https://www.nature.com/articles/s41386-025-02274-1#ref-CR30 “Moffat JJ, Sakhai SA, Hoisington ZW, Ehinger Y, Ron D. The BDNF Val68Met polymorphism causes a sex specific alcohol preference over social interaction and also acute tolerance to the anxiolytic effects of alcohol, a phenotype driven by malfunction of BDNF in the ventral hippocampus of male mice. Psychopharmacology. 2023;240:303–17. https://doi.org/10.1007/s00213-022-06305-3

.“)]. Together, these data suggest that BDNF in corticostriatal regions gates alcohol drinking behaviors, and that AUD-like phenotypes develop in part when BDNF signaling ceases to function.

The cortical regions that release BDNF into the DLS have not been carefully mapped out. Using a combination of transgenic mouse lines together with a viral-mediated retrograde tracing strategy, we characterized BDNF-expressing cortical neurons that project to the DLS [[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)]. We found that a small ensemble of BDNF-positive neurons in the vlOFC extend axonal projections to the DLS [[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)]. The OFC plays a critical role in decision-making, reward-prediction error [[31](#ref-CR31 “Bromberg-Martin ES, Matsumoto M, Hikosaka O. Dopamine in motivational control: rewarding, aversive, and alerting. Neuron. 2010;68:815–34. https://doi.org/10.1016/j.neuron.2010.11.022

.“),[32](#ref-CR32 “Schultz W. Dopamine reward prediction-error signalling: a two-component response. Nat Rev Neurosci. 2016;17:183–95. https://doi.org/10.1038/nrn.2015.26

.“),[33](https://www.nature.com/articles/s41386-025-02274-1#ref-CR33 “Schultz W. Reward prediction error. Curr Biol. 2017;27:R369–R371. https://doi.org/10.1016/j.cub.2017.02.064

.“)], reward information [[34](https://www.nature.com/articles/s41386-025-02274-1#ref-CR34 “Wallis JD. Orbitofrontal cortex and its contribution to decision-making. Annu Rev Neurosci. 2007;30:31–56. https://doi.org/10.1146/annurev.neuro.30.051606.094334

.“)], and stimulus-outcome behaviors [[35](#ref-CR35 “Ostlund SB, Balleine BW. Orbitofrontal cortex mediates outcome encoding in Pavlovian but not instrumental conditioning. J Neurosci. 2007;27:4819–25. https://doi.org/10.1523/JNEUROSCI.5443-06.2007

.“),[36](#ref-CR36 “Rolls ET, Critchley HD, Mason R, Wakeman EA. Orbitofrontal cortex neurons: role in olfactory and visual association learning. J Neurophysiol. 1996;75:1970–81. https://doi.org/10.1152/jn.1996.75.5.1970

.“),[37](https://www.nature.com/articles/s41386-025-02274-1#ref-CR37 “Stalnaker TA, Cooch NK, Schoenbaum G. What the orbitofrontal cortex does not do. Nat Neurosci. 2015;18:620–7. https://doi.org/10.1038/nn.3982

.“)]. The OFC has also been identified as a critical region in AUD [[38](https://www.nature.com/articles/s41386-025-02274-1#ref-CR38 “Moorman DE. The role of the orbitofrontal cortex in alcohol use, abuse, and dependence. Prog Neuropsychopharmacol Biol Psychiatry. 2018;87:85–107. https://doi.org/10.1016/j.pnpbp.2018.01.010

.“), [39](https://www.nature.com/articles/s41386-025-02274-1#ref-CR39 “Bracht T, Soravia L, Moggi F, Stein M, Grieder M, Federspiel A, et al. The role of the orbitofrontal cortex and the nucleus accumbens for craving in alcohol use disorder. Transl Psychiatry. 2021;11:267. https://doi.org/10.1038/s41398-021-01384-w

.“)]. Specifically, in humans, alcohol dependence reduces white matter and neuronal density in the OFC [[40](#ref-CR40 “Pfefferbaum A, Sullivan EV. Disruption of brain white matter microstructure by excessive intracellular and extracellular fluid in alcoholism: evidence from diffusion tensor imaging. Neuropsychopharmacology. 2005;30:423–32. https://doi.org/10.1038/sj.npp.1300623

.“),[41](#ref-CR41 “Atmaca M, Tabara MF, Koc M, Gurok MG, Baykara S, Korkmaz S, et al. Cortical Thickness of the Orbitofrontal Cortex in Patients with Alcohol Use Disorder. Brain Sci. 2023:13. https://doi.org/10.3390/brainsci13040552

.“),[42](https://www.nature.com/articles/s41386-025-02274-1#ref-CR42 “Miguel-Hidalgo JJ, Overholser JC, Meltzer HY, Stockmeier CA, Rajkowska G. Reduced glial and neuronal packing density in the orbitofrontal cortex in alcohol dependence and its relationship with suicide and duration of alcohol dependence. Alcohol Clin Exp Res. 2006;30:1845–55. https://doi.org/10.1111/j.1530-0277.2006.00221.x

.“)], and the connectivity between OFC and striatum is altered in abstinent subjects [[43](https://www.nature.com/articles/s41386-025-02274-1#ref-CR43 “Volkow ND, Wang GJ, Telang F, Fowler JS, Logan J, Jayne M, et al. Profound decreases in dopamine release in striatum in detoxified alcoholics: possible orbitofrontal involvement. J Neurosci. 2007;27:12700–6. https://doi.org/10.1523/JNEUROSCI.3371-07.2007

.“)]. In rodents, alcohol affects the activity of OFC neurons [[44](https://www.nature.com/articles/s41386-025-02274-1#ref-CR44 “Nimitvilai S, Lopez MF, Mulholland PJ, Woodward JJ. Chronic Intermittent Ethanol Exposure Enhances the Excitability and Synaptic Plasticity of Lateral Orbitofrontal Cortex Neurons and Induces a Tolerance to the Acute Inhibitory Actions of Ethanol. Neuropsychopharmacology. 2016;41:1112–27. https://doi.org/10.1038/npp.2015.250

.“), [45](https://www.nature.com/articles/s41386-025-02274-1#ref-CR45 “Gioia DA, Woodward JJ. Altered Activity of Lateral Orbitofrontal Cortex Neurons in Mice following Chronic Intermittent Ethanol Exposure. eNeuro. 2021;8:ENEURO.0503–20.2021. https://doi.org/10.1523/ENEURO.0503-20.2021

.“)]. Furthermore, OFC lesions or chemogenetic inhibition increase alcohol drinking [[46](#ref-CR46 “den Hartog C, Zamudio-Bulcock P, Nimitvilai S, Gilstrap M, Eaton B, Fedarovich H, et al. Inactivation of the lateral orbitofrontal cortex increases drinking in ethanol-dependent but not non-dependent mice. Neuropharmacology. 2016;107:451–9. https://doi.org/10.1016/j.neuropharm.2016.03.031

.“),[47](#ref-CR47 “Schuh KM, Sneddon EA, Nader AM, Muench MA, Radke AK. Orbitofrontal cortex subregion inhibition during binge-like and aversion-resistant alcohol drinking. Alcohol. 2022;99:1–8. https://doi.org/10.1016/j.alcohol.2021.11.004

.“),[48](https://www.nature.com/articles/s41386-025-02274-1#ref-CR48 “Ray MH, Hanlon E, McDannald MA. Lateral orbitofrontal cortex partitions mechanisms for fear regulation and alcohol consumption. PLoS One. 2018;13:e0198043. https://doi.org/10.1371/journal.pone.0198043

.“)] and decrease context and cue-induced reinstatement [[49](https://www.nature.com/articles/s41386-025-02274-1#ref-CR49 “Bianchi PC, Carneiro de Oliveira PE, Palombo P, Leão RM, Cogo-Moreira H, Planeta C, et al. Functional inactivation of the orbitofrontal cortex disrupts context-induced reinstatement of alcohol seeking in rats. Drug Alcohol Depend. 2018;186:102–12. https://doi.org/10.1016/j.drugalcdep.2017.12.045

.“), [50](https://www.nature.com/articles/s41386-025-02274-1#ref-CR50 “Hernandez JS, Binette AN, Rahman T, Tarantino JD, Moorman DE. Chemogenetic Inactivation of Orbitofrontal Cortex Decreases Cue-induced Reinstatement of Ethanol and Sucrose Seeking in Male and Female Wistar Rats. Alcohol Clin Exp Res. 2020;44:1769–82. https://doi.org/10.1111/acer.14407

.“)]. Together, these data suggest that the OFC is an important target of alcohol. However, whether BDNF in vlOFC-to-DLS projecting neurons affects alcohol drinking behaviors is unknown. We report that BDNF expression is reduced in the vlOFC of mice following 7 weeks of intermittent access to 20% alcohol in a 2-bottle choice procedure (IA20%2BC). We further show that overexpression of BDNF in vlOFC-to-DLS circuit limits alcohol intake, seeking, and relapse. Finally, we show that the systemic administration of a TrkB agonist reverts habitual to goal-directed alcohol seeking.

Materials and methods

Reagents, preparation of solutions, collection of brain samples, real-time PCR, purchasing of viruses, stereotaxic viral infection, confirmation of viral expression, and behavioral procedures can be found in the supplementary material.

Animals

Male (152) and female (38) C57BL/6J mice (6–8 weeks) were purchased from Jackson Laboratory and were allowed one week of habituation before experiments began. Mice were individually housed on paper-chip bedding, under a reverse 12-h light-dark cycle. Temperature and humidity were kept constant at 22 ± 2°C, and relative humidity was maintained at 50 ± 5%. Mice were allowed access to food and tap water ad libitum. All animal procedures were approved by the University’s Institutional Animal Care and Use Committee (IACUC) and were conducted in agreement with the Association for Assessment and Accreditation of Laboratory Animal Care.

Behavioral procedures

Intermittent access to 20% alcohol two-bottle choice (IA20%2BC)

IA20%2BC was conducted as previously described [[51](https://www.nature.com/articles/s41386-025-02274-1#ref-CR51 “Ehinger Y, Zhang Z, Phamluong K, Soneja D, Shokat KM, Ron D. Brain-specific inhibition of mTORC1 eliminates side effects resulting from mTORC1 blockade in the periphery and reduces alcohol intake in mice. Nat Commun. 2021;12:4407. https://doi.org/10.1038/s41467-021-24567-x

.“)]. Briefly, mice were given one bottle of 20% alcohol (v/v) in tap water and one bottle of water for 24 h on Monday, Wednesday, and Friday, with 24 or 48-h (weekend) of alcohol withdrawal periods during which mice consumed only water. The placement of water or alcohol bottles was alternated between each session to avoid side preference. Alcohol and water bottles were weighed at the beginning and end of each alcohol drinking session, and alcohol intake (g/kg of body weight), water intake (ml/kg) and total fluid intake (ml/kg) were calculated. Two bottles containing water and alcohol in an empty cage were used to evaluate the spillage. Alcohol preference ratio was calculated by dividing the volume of alcohol consumed to the total volume of fluid intake.

Operant alcohol self-administration

Alcohol operant self-administration training and habitual alcohol seeking training was performed as described previously [[52](https://www.nature.com/articles/s41386-025-02274-1#ref-CR52 “Hoisington ZW, Salvi A, Laguesse S, Ehinger Y, Shukla C, Phamluong K, et al. The Small G-Protein Rac1 in the Dorsomedial Striatum Promotes Alcohol-Dependent Structural Plasticity and Goal-Directed Learning in Mice. J Neurosci. 2024;44:e1644232024. https://doi.org/10.1523/JNEUROSCI.1644-23.2024

.“)]. First, mice underwent 7 weeks IA20%2BC. Mice drinking more than 12.5 g/kg were selected for the experiment. Alcohol operant self-administration training was initiated under a fixed-ratio (FR) 1 schedule, i.e., one lever press resulted in the delivery of one reward, for four 6-h FR1 sessions, followed by four 4-h FR1 sessions, and finally four 2-h FR1 sessions. Reward deliveries were paired with a 3-s tone and illumination of a cue light. All remaining sessions lasted 2 h. Mice were then trained on a random-interval (RI) schedule, during which rewards were delivered with random delays following active lever presses according to previous studies [[53](#ref-CR53 “Gremel CM, Costa RM. Orbitofrontal and striatal circuits dynamically encode the shift between goal-directed and habitual actions. Nat Commun. 2013;4:2264. https://doi.org/10.1038/ncomms3264

.“),[54](#ref-CR54 “Gremel CM, Chancey JH, Atwood BK, Luo G, Neve R, Ramakrishnan C, et al. Endocannabinoid Modulation of Orbitostriatal Circuits Gates Habit Formation. Neuron. 2016;90:1312–24. https://doi.org/10.1016/j.neuron.2016.04.043

.“),[55](https://www.nature.com/articles/s41386-025-02274-1#ref-CR55 “Corbit LH, Nie H, Janak PH. Habitual alcohol seeking: time course and the contribution of subregions of the dorsal striatum. Biol Psychiatry. 2012;72:389–95. https://doi.org/10.1016/j.biopsych.2012.02.024

.“)]. Timepoints were pseudo-randomly assigned by the computer program. Mice first underwent 5 sessions on an RI30 schedule (delays averaging 30 s after lever press, with intervals ranging from 0 to 60 s). Mice were then subjected to 5 sessions of RI60 training (intervals ranging from 30 to 90 s). Mice were divided into 2 groups with similar numbers of active lever presses (106.41 ± 40.78 and 116.35 ± 58.35), port entries (96.35 ± 45.73 and 72.61 ± 14.36), and amount of self-administered alcohol (2.55 ± 0.70 and 2.46 ± 0.79 g/kg/2 h). Three mice were excluded from the operant self-administration procedure due to low pressing numbers. BDNF or mCherry control was then overexpressed in the vlOFC-to-DLS circuit. Three weeks following surgery, RI30 was resumed for 5 sessions, followed by RI60 until the end of the experiment. The number of active lever presses and reward port entries, as well as the number of reward deliveries, were recorded during each session.

LM22A-4 administration: Thirty minutes prior to a degradation session, mice received intraperitoneal (i.p.) administration of saline or LM22A-4 (100 mg/kg) [[29](https://www.nature.com/articles/s41386-025-02274-1#ref-CR29 “Warnault V, Darcq E, Morisot N, Phamluong K, Wilbrecht L, Massa SM, et al. The BDNF Valine 68 to Methionine Polymorphism Increases Compulsive Alcohol Drinking in Mice That Is Reversed by Tropomyosin Receptor Kinase B Activation. Biol Psychiatry. 2016;79:463–73. https://doi.org/10.1016/j.biopsych.2015.06.007

.“)].

Contingency degradation

Contingency degradation was used as previously described [[56](https://www.nature.com/articles/s41386-025-02274-1#ref-CR56 “Gourley SL, Olevska A, Zimmermann KS, Ressler KJ, Dileone RJ, Taylor JR. The orbitofrontal cortex regulates outcome-based decision-making via the lateral striatum. Eur J Neurosci. 2013;38:2382–8. https://doi.org/10.1111/ejn.12239

.“), [57](https://www.nature.com/articles/s41386-025-02274-1#ref-CR57 “Morisot N, Phamluong K, Ehinger Y, Berger AL, Moffat JJ, Ron D. mTORC1 in the orbitofrontal cortex promotes habitual alcohol seeking. Elife. 2019;8. https://doi.org/10.7554/eLife.51333

.“)] to test the sensitivity of mice to changes in the response-outcome association. The procedure was conducted across nondegraded (ND) and degraded (D) sessions. During the 2-h degraded session, a lever was extended, but lever presses produced no consequences. In total, 2-3 degradation sessions were performed. During the nondegraded sessions, alcohol deliveries occurred at a rate that was determined based on each animal’s average reward rate during the four RI60 schedules of nondegraded sessions, prior to degradation.

Extinction and reacquisition

Extinction and reacquisition procedures were conducted following completion of the RI training phase and contingency degradation test. Mice underwent 13 daily extinction sessions (2-h each), during which both levers were available, but lever presses were not paired with alcohol deliveries and no cues associated with reward delivery were presented.

The first extinction session was used to assess alcohol-seeking behavior in the absence of reinforcement, as measured by the number of active lever presses and port entries. Subsequent extinction sessions continued until responding stabilized across sessions and decreased to at least half of lever presses compared to previous RI sessions.

Following the last extinction session, reacquisition training (2-h) was conducted during which lever presses were paired with cues and alcohol deliveries to evaluate the reinstatement of operant responding. The session parameters were identical to FR1 schedule.

Statistical analysis

D’Agostino–Pearson normality test, Shapiro–Wilk normality test and F-test/Levene tests were used to verify the normal distribution of variables and the homogeneity of variance, respectively. Data were analyzed using the appropriate statistical test, including two-tailed unpaired t-test, one-way ANOVA, two-way ANOVA with and without repeated measures followed by post-hoc test. GraphPad Prism 9 was used for statistical analyses. All data are expressed as mean +/− SEM. Significance was set at p < 0.05.

Results

High alcohol drinking reduces BDNF expression in the vlOFC

We identified a small ensemble of vlOFC neurons expressing BDNF that project to the DLS of mice [[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)]. Since BDNF signaling in the DLS is a locus for keeping alcohol drinking in moderation [[18](https://www.nature.com/articles/s41386-025-02274-1#ref-CR18 “Jeanblanc J, He DY, Carnicella S, Kharazia V, Janak PH, Ron D. Endogenous BDNF in the dorsolateral striatum gates alcohol drinking. J Neurosci. 2009;29:13494–502. https://doi.org/10.1523/JNEUROSCI.2243-09.2009

.“), [19](https://www.nature.com/articles/s41386-025-02274-1#ref-CR19 “Jeanblanc J, Logrip ML, Janak PH, Ron D. BDNF-mediated regulation of ethanol consumption requires the activation of the MAP kinase pathway and protein synthesis. Eur J Neurosci. 2013;37:607–12. https://doi.org/10.1111/ejn.12067

.“)], we hypothesized that normal levels of BDNF in the OFC are required to control the development of heavy alcohol intake. We further hypothesized that breakdown in BDNF signaling in OFC-to-DLS circuit promotes the escalation of alcohol intake. To address these hypotheses, we first tested whether high alcohol drinking alters BDNF levels in the vlOFC. Mice underwent 7 weeks of IA20%-2BC procedure [[52](https://www.nature.com/articles/s41386-025-02274-1#ref-CR52 “Hoisington ZW, Salvi A, Laguesse S, Ehinger Y, Shukla C, Phamluong K, et al. The Small G-Protein Rac1 in the Dorsomedial Striatum Promotes Alcohol-Dependent Structural Plasticity and Goal-Directed Learning in Mice. J Neurosci. 2024;44:e1644232024. https://doi.org/10.1523/JNEUROSCI.1644-23.2024

.“)] (Fig. 1a). Male mice consumed an average of 14.47 ± 1 g/kg/24 h (Supplementary Fig. 1a–c; Supplementary Table 1), and female mice drank an average of 19.27 ± 1.1 g/kg/24 h (Supplementary Fig. 1d–f; Supplementary Table 1). Mice were sacrificed 4 h after the beginning of the last drinking session (“binge”) and 24 h after the last drinking session (“withdrawal”), and BDNF expression was measured. We found that BDNF mRNA in the vlOFC was significantly decreased in male mice that were subjected to 7 weeks of IA20%2BC during both binge and withdrawal as compared to mice consuming water only (Fig. 1b) (One-way ANOVA: effect of treatment, F(2,15) = 5.705, p = 0.014; Post hoc Dunnett’s multiple comparison test indicates significant differences between water and binge drinking and between water and withdrawal). We further discovered that alcohol-mediated reduction in BDNF mRNA levels is localized to the vlOFC since no changes in BDNF expression were detected in the medial OFC (mOFC) (Fig. 1c) (One-way ANOVA: effect of treatment, F(2,15) = 0.846, p = 0.448). As BDNF neurons in M2 motor cortex send dense projections to the DLS [[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)], we examined the level of BDNF mRNA in this region in response to high alcohol consumption. As shown in Fig. 1d, BDNF mRNA levels in M2 motor cortex of male mice were unchanged by alcohol (One-way ANOVA: effect of treatment, F(2,15) = 0.354, p = 0.707). Interestingly, the alterations of BDNF levels were sex specific as BDNF mRNA levels in the vlOFC, mOFC or M2 of female mice were unaltered after binge drinking and withdrawal from 7 weeks of IA20%2BC (Fig. 1e–g) (one-way ANOVA: vlOFC, effect of treatment, F(2,16) = 1.13, p = 0.347; mOFC, effect of treatment, F(2,15) = 2.88, p = 0.087; M2 motor cortex, effect of treatment, F(2,17) = 0.484, p = 0.624). Together, these data suggest that chronic alcohol consumption reduces BDNF expression specifically in the vlOFC of male mice. As a result, all subsequent experiments were performed in male mice.

Fig. 1: BDNF expression is attenuated in the vlOFC but not in the mOFC and M2 of male but not female mice in response to chronic high alcohol intake and withdrawal.

a Timeline of experiments: Female and male mice underwent 7 weeks of IA20%2BC or water only (Supplementary Table 1). Four hours after the beginning of the last drinking session (“binge”) and 24 h after the last drinking session (“withdrawal”), the vlOFC, mOFC and motor cortex M2 were dissected and harvested. The expression of BDNF was measured by RT-qPCR using GAPDH an internal control. Data are presented as the average ratio of BDNF to GAPDH ± SEM and are expressed as percentage of water control. BDNF mRNA in the vlOFC (b), mOFC (c) and M2 motor cortex (d) of male mice. BDNF mRNA in the vlOFC (e), mOFC (f) and M2 motor cortex (g) of female mice. *p < 0.05; ns non-significant. n = 6–8 per group.

DIO-Cre-dependent overexpression of BDNF in vlOFC neurons projecting to the DLS

As described above, high alcohol drinking downregulates BDNF expression in the vlOFC. If BDNF in vlOFC-to-DLS circuit is gating alcohol intake, then replenishing its levels in this circuitry will revert high alcohol intake to moderate levels. To test this possibility, we utilized a circuit-specific strategy to overexpress BDNF in vlOFC neurons that project to the DLS by using the DIO/Cre system, enabling BDNF expression only in the presence of Cre recombinase (Fig. 2a) [[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)]. AAV2-DIO-BDNF-mCherry virus (1 ×1012 gc/ml) was bilaterally infused into the vlOFC and AAVretro-Cre-GFP virus (3 ×1012 vg/ml) was bilaterally infused into the DLS (Fig. 2b, c). This allowed Cre expression in vlOFC neurons projecting to the DLS visualized by GFP (Fig. 2c), thereby activating Cre-mediated expression of BDNF visualized by mCherry in the vlOFC (Fig. 2c). We further analyzed BDNF signal in vlOFC neurons that project to the DLS in the presence of Cre recombinase (Supplementary Fig. 2a–d). In addition, we assessed the spread of both viruses in the DLS and OFC. As shown in Supplementary Fig. 3, the AAVretro-Cre-GFP virus is localized 1 mm from bregma on the AP axis in the DLS. The AAV2-DIO-BDNF-mCherry virus infection site is localized at 2.1 mm from bregma on the AP axis in the vlOFC, confirming that BDNF is overexpressed in the vlOFC to DLS circuit. Control Mice were bilaterally infected with AAV2-DIO-mCherry in the vlOFC and AAVretro-Cre-GFP in the DLS. Finally, we tested whether BDNF mRNA levels are elevated over baseline in vlOFC neurons that project to the DLS. As shown in Fig. 2d, there was a significant increase of BDNF mRNA levels in the vlOFC of mice infected with AAV2-DIO-BDNF-mCherry as compared to mice infected with AAV2-DIO-mCherry (Unpaired t-test: t (13) = 2.990, p = 0.010).

Fig. 2: Overexpression of BDNF in vlOFC-to-DLS circuit moderates alcohol but not sucrose intake.

a Viral strategy of Cre-dependent overexpression of BDNF: BDNF and mCherry coding sequences are floxed by a pair of loxP (blue triangles) and lox2272 (red triangles) sites. In the absence of Cre recombinase, the BDNF and mCherry coding sequences are inverted relative to the EF1a promoter. When expressed, Cre recombinase inverts the BDNF and mCherry sequences into a correct orientation, allowing their expression. b Schematic representation of BDNF overexpression in vlOFC-to-DLS circuit. Mice received bilateral injections of AAV2-DIO-BDNF-mCherry or AAV2-DIO-mCherry in the vlOFC and AAVretro-Cre-GFP in the DLS. c Representative images depicting targeting of AAV2-DIO-BDNF-mCherry (red) in the vlOFC and AAVretro-Cre-GFP (green) in the DLS. Red signal indicates expression of Cre and infection of AAV2-DIO-BDNF-mCherry, confirming the overexpression of BDNF in vlOFC neurons projecting to the DLS. Top left panel depicts mCherry (red) and DAPI (cyan). Bottom left panel depicts GFP (green) and DAPI (cyan). Top and bottom right panels depict mCherry (red) and GFP (green). d The vlOFC was dissected 10 weeks after the infusion of AAV2-DIO-BDNF-mCherry or AAV2-DIO-mCherry in the vlOFC and retroAAV-Cre-GFP in DLS and BDNF expression was measured by RT-qPCR. GAPDH was used as an internal control. Data are presented as the average ratio of BDNF to GAPDH ± SEM and expressed as the percentage of water control. e Timeline of experiments: Three weeks after the surgery, mice underwent IA20%2BC for 7 weeks in the home cage (Supplementary Table 1). f Alcohol intake was recorded. g Alcohol preference was calculated as the ratio of alcohol intake relative to total fluid intake. h Timeline of experiments: Mice received a bilateral injection of AAV2-DIO-BDNF-mCherry or AAV2-DIO-mCherry in the vlOFC and AAVretro-Cre-GFP in the DLS and three weeks after the viral injection, mice were subjected to 2-bottle choice with 0.3% sucrose drinking for 7 sessions in the home cage (Supplementary Table 1). i Sucrose intake was recorded. j Sucrose preference was calculated as the ratio of sucrose intake relative to total fluid intake. Data are represented as mean ± SEM. *p < 0.05, ns: non-significant. n = 6–8 per group.

BDNF in vlOFC-to-DLS projecting neurons gates alcohol but not sucrose intake

Three weeks following viral infection enabling maximal BDNF expression, mice were subjected to 7 weeks of IA20%2BC or water only (Fig. 2e). We found that overexpression of BDNF in vlOFC to DLS projecting neurons significantly reduces alcohol drinking and preference as compared to control mice (Fig. 2f, g, Supplementary Table 1) (Alcohol intake: Two-Way mixed-effect ANOVA, effect of BDNF overexpression, F(1, 13) = 5.89, p = 0.030, effect of session, F(5.873, 73.71) = 3.016, p = 0.011, main effect of interaction, F(20,251) = 0.925, p = 0.554. Alcohol preference: effect of BDNF overexpression, F(1, 13) = 5.91, p = 0.020, effect of session, F(20,240) = 4.35, ****p < 0.0001, effect of interaction, F(20, 240) = 0.96638, p = 0.507). Furthermore, alcohol intake of control mice escalated over time, whereas progressive increase of intake was not detected in mice infected with BDNF in vlOFC neurons projecting to the DLS (Fig. 2f, g). Overexpression of BDNF in vlOFC to DLS projecting neurons did not affect water and total fluid intake (Supplementary Fig. 4a, b) (Water consumption: Two-Way mixed-effect ANOVA, effect of BDNF overexpression, F(1, 13) = 2.906, p = 0.11, effect of session, F(20, 245) = 4.437, ****p < 0.0001, effect of interaction F(20,245) = 0.92, p = 0.55; Total intake consumption: effect of BDNF overexpression, F(1, 13) = 1.168, p = 0.2994, effect of session, F(20, 245) = 1.875, *p = 0.0148, effect of interaction F(20,245) = 1.536, p = 0.0701). Together, these data indicate that escalation of alcohol intake is due in part to the attenuation of BDNF levels in the vlOFC, which is rescued by replenishing BDNF in vlOFC-to-DLS projecting neurons. Our data further suggests that BDNF in OFC to DLS circuitry gates escalation of high alcohol intake.

Next, we set out to determine whether BDNF in vlOFC-to-DLS circuit gates the consumption of sucrose, a natural rewarding substance. To do so, a new cohort of mice was subjected to bilateral injections of AAV2-DIO-BDNF-mCherry in the vlOFC and AAVretro-Cre-GFP in the DLS. Control mice were infected with AAV2-DIO-mCherry in the vlOFC and AAVretro-Cre-GFP in the DLS. Three weeks post-viral infusion, mice underwent intermittent access to 0.3% sucrose 2BC procedure for 2 weeks (Fig. 2h). We found that BDNF overexpression in vlOFC-to-DLS circuit does not alter sucrose intake and preference compared to control mice (Fig. 2i, j, Supplementary Table 1) (Sucrose intake: Two-Way ANOVA, effect of BDNF overexpression, F(1, 11) = 0.233, p = 0.638, effect of session, F(6, 63) = 1.426, p = 0.218, effect of interaction, F(6,63) = 1.605, p = 0.160; Sucrose preference: effect of BDNF overexpression, F(1, 11) = 0.011, p = 0.915, effect of session, F(6, 58) = 1.333, p = 0.257, effect of interaction, F(6, 58) = 3.069, p = 0.012). BDNF overexpression in vlOFC-to-DLS circuit did not alter water consumption and total fluid consumption (Supplementary Fig. 4c, d) (Water intake: Two-Way mixed-effect ANOVA, effect of BDNF overexpression, F(1, 11) = 0.2527, p = 0.62, effect of session, F(6, 58) = 0.8752, p = 0.51, effect of interaction, F(6, 58) = 2.479, *p = 0.03; Total fluid consumption: effect of BDNF overexpression, F(1, 11) = 2.491, p = 0.1428, effect of session, F(6, 58) = 1.541, p = 0.1811, effect of interaction, F(6, 58) = 0.4101, p = 0.8694). Thus, the attenuation of alcohol drinking by BDNF in the vlOFC-to-DLS circuit is not due to changes in palatability and is specific to alcohol. As the striatum plays a role in motor skills [[58](https://www.nature.com/articles/s41386-025-02274-1#ref-CR58 “Calabresi P, Picconi B, Tozzi A, Ghiglieri V, Di Filippo M. Direct and indirect pathways of basal ganglia: a critical reappraisal. Nat Neurosci. 2014;17:1022–30. https://doi.org/10.1038/nn.3743

.“)], we determined whether overexpression of BDNF in vlOFC-to-DLS projecting neurons is due to attenuation of locomotion. As shown in Supplementary Fig. 5, total distance traveled (Supplementary Fig. 5b, c) and velocity **(**Supplementary Fig. 5d) were similar in the two groups (Total distance traveled: Mann–Whitney Test: U = 15, p = 0.445; Average velocity: Mann–Whitney Test: U = 15, p = 0.445). These results indicate that the behavioral difference in alcohol consumption is not due to changes in motor behavior.

BDNF in vlOFC-to-DMS or M2-to-DLS projecting neurons does not moderate alcohol intake

A small population of BDNF projecting neurons from the vlOFC also projects to the DMS [[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)]. Therefore, we investigated whether BDNF in vlOFC-to-DMS projecting neurons modulates alcohol consumption. We used the same circuit-specific viral approach, consisting of a bilateral infusion of AAV2-DIO-BDNF-mCherry in the vlOFC and AAVretro-Cre-GFP in the DMS (Fig. 3a), enabling overexpression of BDNF specifically in vlOFC-DMS projecting neurons (Fig. 3b). Control mice were infected with AAV2-DIO-mCherry in the vlOFC and AAVretro-Cre-GFP in the DMS. Three weeks post-viral administration, mice underwent IA20%2BC for 7 weeks, and alcohol intake was assessed. We found no significant difference in alcohol (Fig. 3c, Supplementary Table 1), water, total fluid intake (Supplementary Fig. 6a, b), and alcohol preference (Fig. 3d) between BDNF-overexpressing mice and control groups (Alcohol intake: Two-Way mixed-effect ANOVA, effect of BDNF overexpression, F(1, 17) = 1.033, p = 0.323, effect of session, F(20, 330) = 3.750, ****p < 0.0001, effect of interaction, F(20,330) = 1.592, p = 0.052; Alcohol preference: effect of BDNF overexpression, F(1, 17) = 2.577, p = 0.1268, effect of session, F(20, 327) = 6.605, ****p < 0.0001, effect of interaction, F(20,327) = 0.5599, p = 0.9377; Water consumption: effect of BDNF overexpression, F(1, 17) = 2.14, p = 0.16, effect of session, F(20, 329) = 7.296, ****p < 0.0001, effect of interaction, F(20, 329) = 0.38, p = 0.99; Total fluid consumption: effect of BDNF overexpression, F(1, 17) = 0.2219, p = 0.6436, effect of session, F(20, 324) = 4.744, ****p < 0.0001, effect of interaction, F(20, 324) = 1.523, p = 0.0713). Together, our results suggest that BDNF in vlOFC-to-DLS but not vlOFC-to-DMS circuit moderates alcohol drinking.

Fig. 3: Overexpression of BDNF in vlOFC-to-DMS or M2-to-DLS neurons does not alter alcohol intake.

a–d Overexpression of BDNF in vlOFC-to-DMS projecting neurons. a Schematic representation of BDNF overexpression in vlOFC-to-DMS circuit. Mice received bilateral injections of AAV2-DIO-BDNF-mCherry or AAV2-DIO-mCherry in the vlOFC and AAVretro-Cre-GFP in the DMS. b Representative image depicting targeting of AAV2-DIO-BDNF-mCherry in vlOFC and AAVretro-Cre-GFP in the DMS. Red cells indicate the expression of Cre and infection of AAV2-DIO-BDNF-mCherry, confirming the overexpression of BDNF in vlOFC neurons projecting to the DMS. Top left panel depicts mCherry (red) and DAPI (cyan). Bottom left panel depicts GFP (green) and DAPI (cyan). Top and bottom right panels depict mCherry (red) and GFP (green). c Three weeks after the surgery, mice underwent IA20%2BC for 7 weeks in the home cage (Supplementary Table 1), and alcohol intake was recorded. d Alcohol preference was calculated as the ratio of alcohol intake relative to total fluid intake. e–h Overexpression of BDNF in M2-to-DLS projecting neurons. e Schematic representation of BDNF overexpression in M2-to-DLS circuit. Mice received bilateral injections of AAV2-DIO-BDNF-mCherry or AAV2-DIO-mCherry in M2 and AAVretro-Cre-GFP in the DLS. f Representative image depicting targeting of AAV2-DIO-BDNF-mCherry in M2 and AAVretro-Cre-GFP in the DLS. Red cells indicate expression of the Cre and infection with AAV2-DIO-BDNF-mCherry confirming the overexpression of BDNF in M2 neurons projecting to the DLS. Top left panel depicts mCherry (red) and DAPI (cyan). Bottom left panel depicts GFP (green) and DAPI (cyan). Top and bottom right panels depict mCherry (red) and GFP (green). g) Three weeks after the surgery, mice underwent IA20%2BC for 7 weeks in the home cage (Supplementary Table 1), and alcohol intake was recorded. h Alcohol preference was calculated as the ratio of alcohol intake relative to total fluid. Data are represented as mean ± SEM. n = 8–10 per group.

BDNF-expressing neurons in M2, a region essential for motor learning and behaviors [[59](https://www.nature.com/articles/s41386-025-02274-1#ref-CR59 “Barthas F, Kwan AC. Secondary Motor Cortex: Where ‘Sensory’ Meets ‘Motor’ in the Rodent Frontal Cortex. Trends Neurosci. 2017;40:181–93. https://doi.org/10.1016/j.tins.2016.11.006

.“)], extend dense projections to the DLS [[23](https://www.nature.com/articles/s41386-025-02274-1#ref-CR23 “Ehinger Y, Soneja D, Phamluong K, Salvi A, Ron D. Identification of Novel BDNF-Specific Corticostriatal Circuitries. eNeuro 2023;10. https://doi.org/10.1523/ENEURO.0238-21.2023

.“)]. We, therefore, assessed whether BDNF in M2-to-DLS projecting neurons alters alcohol consumption. To do so, AAV2-DIO-BDNF-mCherry was infused into the M2 and AAVretro-Cre-GFP int