AAV-only targeting of ventral tegmental area dopamine neurons for optical self-stimulation studies in mice Authors: Ezequiel Marron Fernandez de Velasco, John C Brent IV, Alex L Welter, Praseuth Yang, Annelise Wickman, Eric H Mitten, Kevin Wickman Published: 2026 ABSTRACT AbstractOptogenetic studies in rodents have highlighted the importance of ventral tegmental area (VTA) dopamine (DA) neurons to reward, learning, and motivation. Selective manipulation of VTA DA neurons is generally achieved with transgenic rodent lines that use a DA neuron-specific promoter to express Cre recombinase, combined with intra-VTA infusion of adeno-associated virus (AAV) vectors harboring Cre-dependent expression cassettes. These studies incur expenses related to line acquisition and husbandry and are typically rate-limited by in-house breeding capacity. Here, we describe a dual AAV approach that permits genetic access to VTA DA neurons and supports robust optical self-stimulation in mice. One vector employs a mouse tyrosine hydroxylase (TH) promoter fragment to drive Cre expression (AAV8-mTH-Cre). Intra-VTA co-infusion of AAV8-mTH-Cre with another vector harboring a Cre-dependent reporter labeled TH-positive VTA neurons in C57BL/6J mice with an efficiency (82%) and selectivity/fidelity (73%) comparable to studies involving TH-Cre knock-in mice. Co-infusion of AAV8-mTH-Cre with a Cre-dependent channelrhodopsin (ChR2) expression vector permitted optical regulation of VTA neurons exhibiting electrophysiological features consistent with conventional DA neurons and supported robust optical self-stimulation. Thus, this AAV-only approach allows investigation of reward-related behavior when transgenic Cre lines are not available, and in studies involving mutant mouse lines or inbred mouse strains.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-55353-8.Keywords: adeno-associated virus, ventral tegmental area, dopamine neuron, optogenetic, self-stimulation, miceSubject terms: Biological techniques, Neuroscience Introduction Drugs of abuse share an ability to enhance dopamine (DA) transmission in the mesocorticolimbic system1, an interconnected network of brain regions including the nucleus accumbens (NAc) and medial prefrontal cortex (mPFC) that mediates the processing and learning related to rewarding stimuli2. DA neurons of the ventral tegmental area (VTA) are an integral component of the mesocorticolimbic system3–5. Optogenetic manipulation of VTA DA neurons is a powerful means to assess the contribution of these neurons to reward-related behavior5,6. For example, optogenetic stimulation of VTA DA neurons enhances locomotor activity in mice7–9, mimicking the unconditioned DA-dependent response evoked by drugs of abuse. Optogenetic stimulation of VTA DA neurons can also induce a conditioned place preference7. Studies employing optogenetic tools have shown that rodents will engage in instrumental responding to self-stimulate VTA DA neurons8,10. Optical self-stimulation (OSS) of VTA DA neurons allows for response-contingent activation of VTA DA neurons, mimicking the ability of addictive drugs to enhance DA transmission in the mesocorticolimbic system. OSS of VTA DA neurons is sufficient to induce key cellular and behavioral hallmarks of repeated drug exposure in mice, including potentiation of glutamatergic neurotransmission in the NAc and resistance of reward-seeking behavior to punishment11. OSS has also highlighted critical roles of discrete VTA DA neuron projections in reinforcement and cue-induced reinstatement of reward-seeking behavior12,13. Genetic manipulation of VTA DA neurons in support of OSS and other addiction-relevant paradigms generally involves a transgenic rodent line expressing Cre recombinase under the control of the promoter for tyrosine hydroxylase (TH)14,15 or the DA transporter (DAT)16,17, together with intra-VTA infusion of a Cre-dependent adeno-associated virus (AAV) vector18,19. Reliance on transgenic Cre driver lines for studies of VTA DA neurons limits investigations to genetically tractable species and comes with the associated costs of acquiring, maintaining, and genotyping mutant strains. In-house breeding capacity is a rate-limiting factor in these studies, and study costs and delays are compounded if breeding the Cre driver line with a different inbred strain or mutant strain (e.g., a knockout line) is desired. Given the practical limitations associated with the use of transgenic Cre driver lines, we sought to develop an “AAV-only” approach to VTA DA neuron manipulation in support of OSS studies in mice. Our approach involves two AAV vectors, one of which harbors an expression cassette containing a fragment (2571 bp) of the mouse TH (mTH) promoter20 and Cre recombinase. This vector, when co-infused into the VTA with another AAV harboring a Cre-dependent transgene expression cassette, permits genetic manipulation of VTA DA neurons in C57BL/6J mice. We further show that this dual-virus approach can be used for optogenetic regulation of VTA DA neurons and OSS studies in C57BL/6J mice. Results AAV-only targeting of VTA dopamine neurons in C57BL/6J mice To gain selective optogenetic access to VTA DA neurons in C57BL/6J mice, we used a relatively short (2571 bp) fragment of the mouse TH promoter (mTH). An AAV harboring this mTH fragment was shown to target TH-positive neurons in the mouse VTA with moderate efficiency (86% of TH-positive VTA neurons expressed the reporter) and selectivity/fidelity (81% of VTA neurons expressing the reporter were TH-positive)20. Efficient packaging of AAV vectors generally requires that the genetic payload does not exceed 4.7-5.0 kb; titers decrease sharply with larger payloads21,22. As the combined size of the mTH promoter fragment (~ 2.6 kb), coding sequence for the channelrhodopsin/EYFP fusion protein (~ 1.7 kb), regulatory elements for enhanced expression (Woodchuck Hepatitis Virus Posttranscriptional Regulatory Element and polyadenylation signal; ~0.8 kb), and Inverted Terminal Repeats (~ 0.3 kB) is ~ 5.4 kb, we used a dual-virus approach to promote transgene expression in VTA DA neurons. One vector contained the coding sequence for Cre recombinase positioned downstream of the mTH promoter fragment (AAV8-mTH-Cre). The other vector harbored a Cre-dependent cassette for channelrhodopsin (AAV8-EF1α-DIO-hChR2(H134R)-EYFP) or EYFP (AAV8-EF1α-DIO-EYFP). A similar dual-virus strategy involving a short fragment of the rat TH promoter has been used to achieve Cre-dependent manipulation of VTA DA neurons in rats and mice23,24, as well as rhesus macaques25. To evaluate the targeting of VTA DA neurons using the AAV-only approach, C57BL/6J mice were given an intra-VTA infusion of pre-mixed (1:1 ratio by volume; 0.4 µL total) cocktail of high-titer AAV8-mTH-Cre and AAV8-EF1α-DIO-EYFP (Fig. 1A). We targeted the parabrachial pigmented nucleus of the VTA (PBP) to minimize off-target labeling of DA neurons of the substantia nigra. After a 2-3-wk incubation period, we detected virus-driven EYFP expression throughout the VTA of C57BL/6J mice that received the virus mixture (Fig. S1; Fig. 1B), but not in mice that received either vector alone (not shown). We used immunohistochemistry and fluorescence microscopy to quantify the percentage of TH-positive neurons in the PBP that express EYFP (targeting efficiency) and the percentage of EYFP-expressing neurons in the PBP that express TH (targeting fidelity/specificity); 82.0 ± 0.4% of TH-positive neurons expressed EYFP and 73.1 ± 3.1% of EYFP-positive VTA neurons expressed TH in this VTA sub-region (Fig. 1B, C). AAV-only genetic manipulation of VTA DA neurons in C57BL/6J mice. (A) Male and female C57BL/6J mice were given intra-VTA infusions of AAV8-mTH-Cre and either AAV8-EF1α-DIO-hChR2(H134R)-EYFP or AAV8-EF1α-DIO-EYFP. Following a 2–4 wk recovery and incubation period, mice were used in immunohistochemistry or slice electrophysiology experiments. (B) Fluorescent image showing a coronal section containing the VTA (-3.16 mm from bregma) of a mouse treated with intra-VTA AAV8-mTH-Cre and AAV8-EF1α-DIO-EYFP. EYFP fluorescence is denoted in yellow, TH immunolabeling is shown in fuchsia, and nuclear (DAPI) labeling is shown in blue; scale: 100 μm. The 3 images below show separate EYFP and TH labeling, and the image overlay (merge), within the region defined by the blue rectangle; scale: 50 μm. White circles highlight cell bodies with EYFP and TH labeling. White arrows highlight cell bodies with TH but not EYFP labeling, and white arrowheads show cell bodies with EYFP but not TH labeling. Abbreviations: fr – fasciculus retroflexus; IF – interfascicular nucleus; IPF – interpenducular fossa; IPR – interpeduncular nucleus; ml – medial lemniscus; mp – mammilary peduncle; MT – medial terminal nucleus; mtg – mammilotegmental tract; PBP – parabrachial pigmented nucleus, VTA; PN – paranigral nucleus, VTA; Rli – rostral linear nucleus; SNc – substantia nigra pars compacta; SNr – substantia nigra pars reticulata. (C) Summary of targeting efficiency (left: percentage of TH-positive neurons exhibiting viral-driven EYFP fluorescence) and targeting fidelity (right: percentage of EYFP-positive neurons with TH labeling) in the PBP. Data were obtained from 2 male and 2 female mice; 6 sections containing the VTA were evaluated per mouse. In total, more than 10,000 TH-positive cells were evaluated for EYFP fluorescence and more than 10,000 EYFP-positive cells were evaluated for TH labeling. (D) Representative traces showing action potentials induced by optical stimulation (1, 5, 10, 25 Hz) in VTA neurons expressing ChR2, but not VTA neurons expressing EYFP; scale: 20 mV/1 s. (E) Summary of optically induced spike fidelity, or the percentage of optical stimuli that evoked a single action potential, as a function of stimulation frequency in ChR2-expressing VTA neurons (n = 8 recordings from 2 male and 3 female mice). To test whether the dual-virus approach permits optogenetic modulation of VTA DA neurons, C57BL/6J mice received an intra-VTA infusion of pre-mixed (1:1 ratio by volume; 0.4 µL total) cocktail of high-titer AAV8-mTH-Cre and AAV8-EF1α-DIO-hChR2(H134R)-EYFP. After a 2-4-wk incubation period, we employed slice electrophysiological approaches in horizontal slices to characterize ChR2-expressing (EYFP-positive) VTA neurons. VTA DA neurons are heterogeneous in their molecular and physiological properties3,26,27, which can differ in a VTA sub-region and projection-specific manner28,29. We targeted EYFP-positive neurons in the lateral PBP, in the region medial to the medial terminal nucleus of the accessory optical tract. In our prior studies involving DAT-Cre mice30–32, we found that most VTA DA neurons (~ 75%) in this VTA sub-region exhibit spontaneous activity (~ 2 Hz), along with other properties often ascribed to “conventional” DA neurons including prominent Ih conductance (> 100 pF) and broad action potentials (2–4 ms)3. Consistent with these published reports, ChR2-expressing VTA neurons evaluated in C57BL/6J mice from the AAV-only treatment group exhibited large Ih currents (139±37 pA) and long action potential durations (3.0±0.3 ms), and most (6/8; 75%) exhibited spontaneous activity (2.1±0.7 Hz). These VTA neurons also responded with high fidelity to brief pulses (5 ms) of blue light delivered at frequencies up to 25 Hz (Fig. 1D, E). In contrast, control VTA neurons evaluated in mice treated with AAV8-mTH-Cre and AAV8-EF1α-DIO-EYFP did not respond to optical stimulation. Thus, the AAV-only approach permits optogenetic regulation of VTA DA neurons in C57BL/6J mice. Optical self-stimulation in C57BL/6J mice We next tested whether the AAV-only approach to ChR2 expression in VTA DA neurons would support optical self-stimulation (OSS) in C57BL/6J mice (Fig. 2A). ChR2-expressing C57BL/6J mice, EYFP-expressing control C57BL/6J mice, and DAT-Cre mice treated with AAV8-EF1α-DIO-hChR2(H134R)-EYFP, were evaluated in parallel. Mice began acquisition sessions 16–19 d after viral infusion and fiber optic light guide implantation in the VTA (Fig. S2); conditions were modeled after those in a study involving DAT-Cre mice33(Fig. 2B). Acquisition sessions (1 h/d) were conducted over 11 d using a fixed-ratio 1 (FR1) schedule of reinforcement. During these sessions, a nosepoke in the active port resulted in a 3-s optical stimulus delivered to the VTA at 50 Hz, with a 5-ms pulse duration. Optical self-stimulation in C57BL/6J mice. (A) C57BL/6J mice were given intra-VTA infusions of AAV8-mTH-Cre and AAV8-EF1α-DIO-hChR2(H134R)-EYFP or AAV8-EF1α-DIO-EYFP. DAT-Cre mice were treated with AAV8-EF1α-DIO-hChR2(H134R)-EYFP. (B) Following viral infusion and fiber implantation, subjects recovered for 16–19 d before beginning 11 daily optical self-stimulation sessions (1 h) using an FR1 schedule of reinforcement. On day 12, subjects were evaluated in an optical frequency preference test. Following a 2-d rest period and 1 additional FR1 session, subjects began a 4-d extinction test. The accuracy of viral targeting and fiber placement was assessed in all mice (Fig. S2). (C) Active (filled symbols) and inactive (open symbols) port nosepokes during acquisition, for C57BL/6J (n = 19) and DAT-Cre (n = 15) mice expressing ChR2 in VTA DA neurons. Control (EYFP-expressing) C57BL/6J mice (yellow circles; n = 6) exhibited low-level responding on the active port over 5 sessions. Two-way repeated measures ANOVA of active-port nosepokes for ChR2-expressing C57BL/6J and DAT-Cre groups revealed a main effect of session (F2.63,84.25=108.5, p < 0.0001) and genotype (F1,32=8.54, **p = 0.0063), but no interaction (F2.63,84.25=1.47, p = 0.2327). Symbol: #p < 0.05 (within session). (D) Discrimination index (active-port nosepokes divided by total nosepokes) during acquisition sessions for C57BL/6J and DAT-Cre mice expressing ChR2 in VTA DA neurons. Two-way repeated measures ANOVA revealed a main effect of session (F1.59,50.95=46.15, p < 0.0001) but not genotype (F1,32=3.35, p = 0.0766), and a session x genotype interaction (F1.59,50.95=6.63, p = 0.0051). Pairwise comparison did not detect significant differences between groups for any session. (E) Acquisition day for C57BL/6J and DAT-Cre mice (t1.25=27.51, p = 0.22; Student’s t-test with Welch’s correction). Male and female data are denoted by open and filled symbols, respectively.(F) Optical frequency test. For each animal, active-port nosepokes measured during each 15-min frequency session was normalized to active-port nosepokes measured at the frequency that evoked maximal responding. Two-way repeated measures ANOVA revealed a main effect of frequency (F3.04,97.17=249.0, p < 0.0001) but not genotype (F1,32=1.45, p = 0.24) or frequency x genotype interaction (F5,160=1.39, p = 0.23). (G) Extinction test. The number of active-port nosepokes recorded during the preceding acquisition session is shown for comparison only. Two-way repeated measures ANOVA revealed a main effect of session (F2.25,71.92=36.4, p < 0.0001) but not genotype (F1,32=1.00, p = 0.32). Moreover, no session x genotype interaction was detected (F2.25,71.92=1.59; p = 0.21). While EYFP-expressing C57BL/6J control mice failed to develop significant active port responding over 5 sessions, all ChR2-expressing mice in the C57BL/6J and DAT-Cre groups developed robust and stable active-port responding (Fig. 2C), with a high degree of active: inactive port discrimination (Fig. 2D), over this same period. Interestingly, ChR2-expressing C57BL/6J mice exhibited higher levels of active-port responding than DAT-Cre mice. While ChR2-expressing C57BL/6J mice reached stable and high levels of active-port discrimination faster than DAT-Cre mice (Fig. 2D), the rate of acquisition – defined as the first of 3 consecutive sessions during which active port responding did not vary by more than 10% and discrimination index was > 0.75 – did not differ significantly between genotypes (Fig. 2E). Following the acquisition sessions, active-port responding as a function of optical stimulation frequency was evaluated in all subjects. Active-port responding was frequency-dependent for ChR2-expressing C57BL/6J and DAT-Cre mice. Maximal responding in both groups was seen between 25 and 100 Hz, and no genotype-dependent differences in sensitivity to optical stimulation frequency were detected (Fig. 2F). Finally, we evaluated the rate of extinction of active-port responding over 4 sessions wherein a nosepoke in either port had no programmed response. ChR2-expressing C57BL/6J and DAT-Cre mice rapidly reduced active-port responding during the extinction sessions, with no evident genotype-dependent differences in rate or extent of extinction (Fig. 2G). Thus, the AAV-only approach is suitable for studies investigating factors that impact acquisition of self-administration behavior, as well as extinction. Discussion Optogenetic manipulation of VTA DA neurons is a powerful strategy to investigate the relevance of these neurons to motivation, reinforcement learning, and addiction-like behavior. Most studies involving optogenetic manipulation of VTA DA neurons have relied on transgenic DAT-Cre and TH-Cre rodent strains to confer neuronal specificity. These studies are typically rate-limited by in-house breeding capacity and they incur costs associated with the acquisition, husbandry, and genotyping of mutant strains. These practical constraints are exacerbated if breeding of the Cre driver line with another inbred or mutant strain is desired. Our goal was to develop a flexible approach to VTA DA neuron manipulation in mice that does not rely on transgenic Cre driver lines and can be used to support OSS studies. To circumvent the genetic payload limitations of AAV, we used a dual-vector strategy that has been used successfully to drive expression of ChR2 and other transgenes in VTA DA neurons from mice, rats, and rhesus macaques23–25. In our approach, access to DA neurons is achieved with a 2571 bp fragment of the mTH promoter20. This promoter was derived based on alignment with the rat TH upstream promoter that was shown to drive transgene expression in catecholaminergic neurons, including DA neurons of the substantia nigra and VTA34,35. A prior study that involved systemic administration of an AAVPHB.eB vector containing this mTH promoter fragment showed that it could drive reporter expression in TH-positive neurons in the mouse VTA with moderate efficiency (86%) and selectivity/fidelity (81%)20. Our assessment of targeting efficiency (82%) and selectivity/fidelity (73%) for TH-positive neurons in the PBP using the dual-virus approach in C57BL/6J mice yielded comparable values; these metrics are similar to those reported in rodent studies employing rat TH promoter fragments23,35. In TH-Cre knock-in mice, Cre-dependent transgene expression has been reported in neurons within and surrounding the VTA that exhibit little-or-no TH protein expression18. Significant variability (50–98+%) in the targeting fidelity metric has been noted across studies involving TH-Cre knock-in mice18,19,36,37. This variability likely reflects in part study-specific differences in the VTA sub-region(s) selected for quantitative analysis, as TH expression varies by region in the rodent VTA38,39. Particularly high targeting fidelities (> 98%) have been reported in the lateral VTA in studies involving TH-Cre mice that used TH immunolabeling as a marker for DA neurons36,37; lower fidelity is seen in midline VTA sub-regions18. Consistent with this latter report, we observed virus-driven fluorescence in midline VTA sub-regions where TH immunolabeling was less prominent in the AAV-only model (Fig. S1). This may be an issue of detection threshold, as some VTA neurons express TH mRNA but produce little-or-no evident TH protein40; this would tend to lower the assessment of targeting fidelity as defined by the fraction of neurons expressing viral transgene that also exhibit TH immunolabeling. As the AAV-only model relies on a TH promoter fragment to gain access to VTA DA neurons, off-target manipulation of neurons within and surrounding the VTA at levels comparable to or higher than those reported in studies involving TH-Cre knock-in mice is not surprising. A notable limitation of this study is that we did not systematically evaluate the impact of several AAV-related parameters on VTA DA neuron targeting metrics or behavioral outcomes. Viral load (titer and volume) can impact targeting fidelity when using AAV vectors with cell-specific promoters41–43. The ratio of AAV-Cre and Cre-dependent AAV vector employed in the dual-virus approach may also be impactful. Enhanced targeting efficiency and/or fidelity for VTA DA neurons may be achieved with a different AAV serotype(s). While all vectors in this study were packaged using the AAV8 serotype, a similar study employing a comparable dual virus approach, a rat TH promoter fragment, and vectors packaged in AAV9 and AAV5 serotypes yielded 95% targeting fidelity for VTA DA neurons in the rhesus macaque, albeit with lower targeting efficiency (~ 50%)25. Thus, systematic optimization of these AAV-related parameters could lead to improved targeting efficiency and fidelity, or behavioral outcomes. Using the AAV-only strategy to manipulate VTA DA neurons, we show that ChR2 expression in VTA DA neurons supports robust OSS behavior in C57BL/6J mice. Interestingly, parallel comparison of ChR2-expressing C57BL/6J and DAT-Cre mice revealed greater active-port responding among subjects in the AAV-only cohort. While DAT-Cre mice are thought to facilitate highly selective targeting of VTA DA neurons18, some TH-positive VTA neurons exhibit little-or-no DAT expression29,44–46. Indeed, DAT expression is low or undetectable in TH-positive neurons in the medial VTA that project to the mPFC, NAc core and medial shell, and basolateral amygdala29. Furthermore, several discrete VTA DA projections, including those targeting the NAc core or shell and the infralimbic region of the mPFC, can independently support optical self-stimulation in mice12,13. Thus, the enhanced active-port responding associated with the AAV-only model may reflect the fact that TH-driven transgene expression provides access to more VTA DA neuron sub-populations than transgene expression driven by the DAT promoter. Most neurons in the rodent VTA are dopaminergic (~ 60%), but there are significant sub-populations of GABAergic (~ 30%) and glutamatergic (~ 10%) neurons that exert influence on reward-related behaviors47–54. Given that a significant fraction (~ 30%) of VTA neurons from mice in the AAV-only treatment group exhibit viral-driven fluorescence but are TH-negative, the enhanced active-port responding seen with the AAV-only model may reflect the influence of ChR2 expression in non-DA neurons of the VTA. Consistent with this premise, stimulation of VTA glutamate neurons is rewarding in mice51. In addition, TH-Cre knock-in mice permit genetic access to a unique population of VTA neurons that project to the lateral habenula (LHb) and release GABA but not DA; optogenetic stimulation of VTA terminals in the LHb induced a GABAA receptor-dependent real-time place preference37. As ChR2 expression and activation in VTA GABA and/or glutamate neurons is likely to impact OSS, identifying the EYFP-positive/TH-negative VTA neuron sub-population(s) accessed with the AAV-only approach will be important and may shed light on the contribution of non-DA neurons in the VTA to reward-related behaviors. Future imaging studies involving expression of genetically encoded biosensors for DA, GABA, and glutamate in downstream targets of the VTA, or studies exploring the impact of receptor-selective antagonists on active-port responding, may help tease apart the relative contributions of DA and non-DA neurotransmission to this reward-related behavior. Lower responding levels in the hemizygous DAT-Cre mice used in this study may also reflect the reduced DAT expression and function in the mesocorticolimbic system17,55. Indeed, these hemizygous DAT-Cre mice were reported previously to exhibit 17% less DAT protein in the striatum than wild-type counterparts17. Reduced DA reuptake capacity following evoked release has also been detected in this and other DAT-Cre lines55–57. Thus, VTA DA neuron stimulation should yield higher peak and more prolonged elevation of extracellular DA levels in the mesocorticolimbic system of hemizygous DAT-Cre mice. Consistent with this premise, hemizygous DAT-Cre mice on a C57BL/6J background exhibited increased basal locomotor activity and decreased sensitivity to the locomotor-stimulatory effects of amphetamine, whereas these behaviors were normal in hemizygous TH-Cre mice on the same genetic background56. The AAV-only approach described herein allows for manipulation of VTA DA neurons in support of OSS studies without reliance on existing transgenic Cre driver mouse lines. While this approach may not be advisable for studies requiring strict cell-type specificity in the manipulation of VTA DA neurons, it can be beneficial in cases where transgenic Cre driver lines are not available or when proof-of-concept data are desired prior to acquiring a transgenic Cre driver line. The AAV-only approach can also be used to facilitate comparative assessments of reward circuit function across inbred mouse strains and to explore genetic influences on reward circuit function using knockout and transgenic mouse lines. Methods Animals C57BL/6J mice (stock 000664) were purchased from The Jackson Laboratories (Bar Harbor, ME) and allowed to acclimate for at least 1 wk prior to experimental handling. DAT-Cre (B6.SJL-Slc6a3tm1.1(cre)Bkmn/J; stock 006660; RRID: IMSR_JAX:006660) mice were purchased from The Jackson Laboratories and were crossed with C57BL/6J mice to generate the DAT-Cre hemizygous offspring used in this study. Mice were group-housed and maintained on a 12 h light/dark cycle, and they had ad libitum access to food and water. All animal use was reviewed and approved by the Institutional Animal Care and Use Committee of the University of Minnesota. All experiments involving animals were designed, performed, and reported in accordance with institutional regulations and ARRIVE Guidelines (2.0)58. Reagents pAAV-mTH-Cre plasmid was generated by the VIC using standard cloning techniques and pAAV-mTH-GFP (RRID: Addgene_99128; a gift from Viviana Gradinaru) as the backbone20. The pAAV-EF1α-DIO-hChR2(H134R)-EYFP plasmid was a gift from Karl Deisseroth (RRID: Addgene_20298) and was used as the backbone to generate pAAV-EF1α-DIO-EYFP. Multiple batches of high-titer AAV8-mTH-Cre (3.1–9.6 × 1012 genocopies/mL), AAV8-EF1α-DIO-EYFP (1.2–1.3 × 1013 genocopies/mL), AAV8-EF1α-DIO-hChR2(H134R)-EYFP (0.9–1.5 × 1013 genocopies/mL) vectors were packaged and purified by the University of Minnesota Viral Innovation Core (VIC; Minneapolis, MN). Intracranial manipulations Procedures for intra-VTA infusion of virus in mice have been described previously30–32,59. Briefly, mice (6–10 wk) were anesthetized with isoflurane and placed in a stereotaxic frame (David Kopf Instruments; Tujunga, CA). A microinjector was lowered through a burr hole in the skull and used to deliver a unilateral infusion (0.4 µL at 0.1 µL/min) of undiluted high-titer virus or pre-mixed high-titer virus cocktail (1:1 ratio by volume), targeting the center of the parabrachial nucleus of the VTA (PBP; from bregma, AP: -3.3 mm, ML: ±0.55 mm, DV: -4.60 mm). For mice used in optical self-stimulation studies, a small screw was placed in the skull (from bregma: AP: -1.25 mm, ML: ±2.00 mm) prior to viral infusion to aid in stabilization of a dental cement cap used to close the surgical site and hold the fiber optic light guide in place. After viral infusion, a fiber optic light guide implant (400 μm – NA0.66, 5 mm; Doric Lenses Inc; Quebec, QC) was positioned 250 μm above the viral infusion site (from bregma, AP: -3.3 mm, ML: ±0.55 mm, DV: -4.35 mm). Unilateral viral infusion and fiber implantation were counterbalanced by hemisphere within each group. Mice recovered for 2–4 wk after surgery before processing for immunohistochemical or electrophysiological analysis, and 16–19 d before behavioral testing. Immunohistochemistry Brains were harvested and drop-fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) at 4 °C overnight, rinsed and equilibrated in a 30% sucrose solution at 4 °C overnight, and then flash-frozen in an isopentane/dry ice bath. Coronal Sect. (50 μm) containing the VTA were obtained by cryostat and collected in PBS in a 24-well plate. Free-floating sections were permeabilized in PBS containing 0.5% Triton X-100 and 2% donkey serum for 10 min and then transferred to blocking solution (PBS containing 0.1% Triton X-100 and 2% donkey serum) for 30 min at room temperature. Sections were incubated at 4 °C overnight in blocking solution with 1:500 dilutions of primary polyclonal antibodies from Sigma-Aldrich (Burlington, MA) targeting TH (AB152, rabbit anti-TH; RRID: AB_390204) and from Abcam (Waltham, MA) targeting GFP (AB6673, goat anti-GFP; RRID: AB_305643). Sections were then incubated in 1:500 dilutions of CyTM3 donkey anti-rabbit (711-165-152; RRID: AB_2307443) and Alexa Fluor® 488 donkey anti-goat (705-405-003; RRID: AB_2340428) secondary antibodies from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA) for 2 h at room temperature. Sections were then mounted using ProLong™ Gold Antifade Mountant with DNA stain DAPI (P36941, Invitrogen; Chicago, IL) and imaged using a BZ-X810 fluorescence microscope (Keyence Corporation of America; Itasca, IL). TH and EYFP expression (and their overlap) in the PBP were quantified in 6 coronal sections (from bregma, AP: -3.00 to -3.60) per mouse, from 2 male and 2 female mice. The boundaries of the PBP in these sections were defined using the Allen Brain Atlas60. We used CellProfiler™ (RRID: SCR_007358)61 and a modified version of a published analysis pipeline for quantification62. In brief, CellProfiler™ was used to subtract background fluorescence and segment cells based on nuclear (DAPI) staining in each section. Initial cell segmentation was expanded to include somatic fluorescent labeling to identify expression-positive cells. The probability that a cell was positive for TH or EYFP was calculated using a binomial distribution test comparing the area of expression within a cell against the area of background fluorescence. Data from the 6 sections were then aggregated and organized using a Python script (GitHub 10.5281/zenodo.18511410) to calculate targeting efficiency and fidelity for each mouse, based on the number of cells positive for 1 or both targets. Final targeting efficiency and fidelity metrics are derived from analysis of more than 10,000 TH-positive and 10,000 EYFP-positive cells. Slice electrophysiology Horizontal slices (225 μm) containing the VTA were prepared as described63. Whole-cell recordings from EYFP-positive cells in the VTA sub-region medial to the medial terminal nucleus of the accessory optical tract were acquired in an oxygenated aCSF bath at 33–35 °C, using a Multiclamp 700B amplifier and pCLAMPv.10.6 software (Molecular Devices, LLC; San Diego, CA), as described31. After whole-cell access was achieved, Ih current was measured in voltage-clamp mode (Vhold = -45 mV) using a 200-ms voltage step to − 120 mV, and spontaneous activity was monitored for 1-min in current-clamp mode. Optically evoked action potentials were measured in current-clamp mode (I = 0) using 5 ms pulses of 470 nm wavelength light (2–5 mW), delivered for 3 s at 1, 5, 10, 25 Hz through a 40x objective. All measured and command potentials factored in a liquid junction potential of − 15 mV. Optogenetic self-stimulation (OSS) Operant conditioning boxes constructed with ARDUINO boards (Arduino Uno or Elegoo Uno R3) were built and controlled as described33. Each box has 2 nosepoke ports and green LED cue lights mounted above the ports. Prior to each OSS session, mice were allowed at least 30 min to acclimate to the testing room. Mice were then hand-restrained and a fiber optic patchcord (MFP 400/430/1100-NA0.57–1 m) from Doric Lenses Inc. (Quebec, Canada) was connected to the fiber optic light guide implant. The fiber optic patchcord was connected to a fiber-coupled LED (M470F4) from Thorlabs, Inc. (Newton, NJ) through a fiber optic rotary joint (FRJ 1 × 1 400/430/LWMJ-0.57; Doric Lenses Inc.). During each 60-min acquisition session, a nosepoke in the active port resulted in illumination of the green cue light above the port and concurrent 3-s optical stimulus (LED 470 nm/20 mW). The optical stimulus for acquisition sessions was delivered at 50 Hz, with a 5-ms pulse duration. A nosepoke in the inactive port resulted only in illumination of the green cue light (3 s) above the port. Discrimination index was defined as the number of active-port nosepokes divided by the total number of nosepokes during a session. Acquisition of OSS was defined as the first of 3 consecutive days when the number of active-port nosepokes varied by less than 10% and discrimination index exceeded 0.75. The stimulation frequency preference test was conducted 1 d after mice finished the 11 acquisition sessions. During this 90-min test, the frequency of optical stimulus elicited by an active-port nosepoke was changed every 15 min (900 s) in the following order: 100 Hz, 50 Hz, 25 Hz, 10 Hz, 5 Hz, 1 Hz. The number of active-port nosepokes was measured at each frequency for each subject. These values were then normalized to the number of active-port nosepokes measured at the stimulation frequency that generated the highest level of responding by that mouse. Extinction of OSS behavior was assessed over 4 consecutive days in 60-min sessions during which responding in active and inactive ports had no programmed consequences. Following extinction, brains were harvested and 225 μm horizontal sections containing the VTA were obtained by vibratome. Brightfield and fluorescent images were acquired using a BZ-X810 fluorescence microscope (Keyence). Light guide placement and the scope of AAV-driven EYFP fluorescence were evaluated using the Mouse Brain Atlas as a reference60(Fig. S2). Data from mice with mistargeted light guide implants, or from mice that lacked fluorescence or exhibited significant fluorescence outside of the VTA (e.g., in the adjacent substantia nigra pars compacta), were excluded from analysis. Data analysis Data were analyzed using Prism v.10.0.2 software (GraphPad Software; Boston, MA). All studies included balanced numbers of male and female mice, and data were evaluated first for sex differences. We did not observe any effect of sex or interaction involving sex for any measured endpoint in this study. As such, data from male and female mice were pooled and are presented throughout as mean SEM. Pooled behavioral data were evaluated for session and/or genotype effects using Student’s t-test or 2-way repeated measures ANOVA. Sidák’s multiple comparisons test was used for pairwise comparisons if warranted. Differences were considered significant if p < 0.05. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements The authors thank members of the Wickman lab (McKinzie Frederick, Anna Souders) for care of the mouse colony and technical support, and Dr. Sarah Mulloy for providing guidance related to the analysis pipeline for the immunohistochemistry approach. Author contributions Ezequiel Marron Fernandez de Velasco: conceptualization, data curation, formal analysis, investigation, methodology, resources, supervision, validation, visualization, writing – review & editingJohn C. Brent IV: data curation, formal analysis, investigation, methodology, visualization, writing – review & editingAlex L. Welter: formal analysis, investigation, methodology, validation, visualization, writing – review & editingPraseuth Yang: formal analysis, investigation, methodology, validation, visualization, writing – review & editingAnnelise Wickman: methodology, resources, writing – review & editingEric H. Mitten: methodology, writing – review & editingKevin Wickman: conceptualization, funding acquisition, project administration, supervision, validation, visualization, writing-original draft, writing – review & editing. Funding This study was supported by the following NIH awards: R01 AA027544 and R01 DA034696 (KW), R25 DA057802 (JCB), F31 DA062412 (ALW), T32 DA007234 (ALW), and T32 NS105604 (EHM). Financial support for viral vector production came from the Viral Innovation Core, part of the University of Minnesota Center for Neural Circuits in Addiction (P30 DA048742). Data availability The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Ezequiel Marron Fernandez de Velasco, Email: marro014@umn.edu. Kevin Wickman, Email: wickm002@umn.edu. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.