Analysis of responses to L-3,4-dihydroxyphenylalanine (L-DOPA) in hemiparkinsonian rodent models can enhance understanding of Parkinson's disease (PD). Traditionally, these responses have been classified into medication OFF and ON states, with ON defined as the period after sufficient time has elapsed following drug administration. However, this binary classification may lead to neurophysiological analyses that overlook interindividual variability in the onset and duration of therapeutic effects. Although measuring plasma L-DOPA concentrations has been proposed to quantify this variability, technical difficulties in direct measurement limit the application of this approach. Here, we investigated whether acute L-DOPA responses can be quantitatively distinguished into behavioral states without direct plasma measurements, and whether these states correspond to neurophysiological dynamics. Using quantitative analysis of rotation direction in hemiparkinsonian rat model, we classified L-DOPA responses into four temporally distinct states: OFF, early contralateral rotation (ECR), rotational pause (RP), and late contralateral rotation (LCR). These behaviorally defined states exhibited statistically distinct neurophysiological profiles, characterized by simultaneously measured subthalamic firing rates and striatal dopamine levels. Although single-unit activity was insufficient to decode state transitions, neural population activity enabled robust decoding via a support vector machine (SVM). Furthermore, linear dynamical system modeling revealed transition-specific trajectories in a low-dimensional state space, suggesting that state transitions are represented by coordinated population dynamics. These findings demonstrate that temporal variability in L-DOPA responses can be identified behaviorally, providing a neurophysiologically validated framework that extends the conventional binary classification into a multi-state transition for PD research.

Adaptive sequential decision-making requires animals to learn how actions shape future states and rewards while adjusting behavioral strategies as experience accumulates. Although the hippocampus is implicated in state-structure representation and the striatum in value learning, how learning strategies evolve from exploratory sampling to efficient exploitation and how hippocampal-striatal interactions reorganize alongside this transition remain unclear. We trained six pigeons of undetermined sex to perform a probabilistic two-step decision task while simultaneously recording behavioral choices and hippocampal-striatal local field potentials. Strategy evolution was tracked using a dynamic reinforcement-learning framework in which model-based, model-free, and weighted model-based/model-free models were fitted across sliding windows. Hippocampal-striatal functional coordination was assessed using phase-locking analysis, and directed predictive influence was quantified using kernel Granger causality. The weighted model best explained choice dynamics, indicating a shift from model-based-like control during early acquisition toward model-free-like control during later stable performance. This behavioral shift was accompanied by two complementary neural changes: hippocampal-striatal phase coordination increased across learning, whereas hippocampus-to-striatum directed predictive influence decreased and covaried positively with model-based-like strategy weight. Direction-ratio analysis showed that hippocampus-to-striatum remained the predominant direction throughout learning, indicating reduced influence rather than directional reversal. After acquisition, hippocampus-to-striatum directed influence was not uniformly reduced but remained relatively preserved during rare, low-value, unrewarded, or less expected events. Together, these findings reveal a learning-dependent reorganization of hippocampal-striatal dynamics, in which increasing functional coordination is accompanied by reduced but condition-dependent hippocampal-to-striatal influence as behavior becomes more stable and efficient. Adaptive learning requires coordinated changes in behavioral strategy and neural interactions as animals progress from exploration to stable performance. Here, we show that pigeons gradually shifted from flexible, model-based-like control toward more stable, model-free-like behavior during sequential learning. In parallel, hippocampal-striatal interactions underwent a complementary reorganization: phase coordination progressively increased, whereas hippocampus-to-striatum directed predictive influence decreased with declining model-based-like control. Importantly, this directed influence remained relatively preserved during rare, low-value, or unrewarded events. These findings link behavioral strategy transition to the dynamic reorganization of hippocampal-striatal interactions, revealing how neural systems may simultaneously support efficient learned behavior while retaining sensitivity to behaviorally informative events.

We present an enhancer-AAV toolbox for accessing and perturbing striatal cell types and circuits. Best-in-class vectors were curated for accessing major striatal neuron populations including medium spiny neurons (MSNs), direct- and indirect-pathway MSNs, Sst-Chodl, Pvalb-Pthlh, and cholinergic interneurons. Specificity was evaluated by multiple modes of molecular validation, by three different routes of virus delivery, and with diverse transgene cargos. Importantly, we provide detailed information necessary to achieve reliable cell-type-specific labeling under different experimental contexts. We demonstrate direct pathway circuit-selective optogenetic perturbation of behavior and multiplex labeling of striatal interneuron types for targeted analysis of cellular features. Lastly, we show conserved in vivo activity for exemplary MSN enhancers in rats and macaques. This collection of striatal enhancer AAVs offers greater versatility compared to available transgenic lines and can readily be applied for cell type and circuit studies in diverse mammalian species beyond the mouse model.

We previously reported, using fiber photometry, that operant responding reinforced by access to a peer (social self-administration) is associated with phasic dopamine increases in nucleus accumbens (NAc) core following lever insertion (reward-availability cue) and gradual increases preceding lever-pressing. Here, we sought to replicate these findings and determine whether dopamine signals (1) generalize to responding for high-carbohydrate palatable food, (2) show opposite patterns during negative reinforcement (shock avoidance/escape), and (3) depend on whether reinforcers are experienced alone or together. We trained male and female rats to lever-press for access to a same-sex peer (15 s/trial) and palatable food (45-mg pellet/trial), followed by shock avoidance/escape (0.18-0.28 mA). After training, we expressed the dopamine sensor GRAB-DA2m and implanted optic fibers into NAc core. We measured dopamine activity during sessions with either one- or three-reinforcers. During social self-administration, dopamine activity showed phasic increases following lever insertion and gradual increases preceding lever-pressing; responses were moderately greater during sessions with all three reinforcers. Palatable food self-administration showed a similar pattern, but responses were approximately twofold greater during single-reinforcer sessions. During shock avoidance/escape, dopamine activity showed phasic decreases at warning onset, lever insertion, and shock onset; responses were also greater during single-reinforcer sessions. Results suggest that NAc core dopamine signaling distinguishes positive from negative reinforcement and is modulated by reinforcer availability. Compared with single-reinforcer sessions, dopamine responses during food self-administration and shock avoidance/escape were reduced during sessions with all three reinforcers, whereas responses during social self-administration modestly increased. Nucleus accumbens (NAc) dopamine is critical for processing the valence of positive reinforcers, whereas its role in negative reinforcement is less well understood. Here, we extended our prior work on operant social self-administration to determine whether these findings generalize to food reinforcement, show the opposite pattern during operant negative reinforcement (shock avoidance/escape), and depend on whether reinforcers are experienced alone or together. Results suggest that NAc core dopamine activity differentiates positive and negative reinforcement and is modulated by the presence of other available reinforcers. Specifically, dopamine responses during food self-administration and shock avoidance/escape were reduced when all three reinforcers were available together compared with when each reinforcer was experienced alone, whereas responses during social self-administration showed the opposite pattern.

A central framework in basal ganglia neuroscience is that the cerebral cortex sends excitatory information to the striatum, where it is integrated with thalamic and neuromodulatory signals to shape action selection, learning, and behavioral flexibility [1-10]. In this classical view, the striatum acts as the major input structure of the basal ganglia, whereas processed output returns to the cortex indirectly through pallidal, nigral, and thalamocortical pathways [1-4, 10, 11]. However, recent evidence indicates that striatal and pallidal circuits can influence cortical processing through more direct, specialized, and neurochemically diverse routes than previously appreciated [12-16], revealing an unexpected diversity of basal ganglia return pathways. In particular, the identification of a previously uncharacterized cholinergic projection from the dorsal tail of the striatum to the auditory cortex suggests that specialized striatal circuits can directly modulate cortical sensory processing [15]. These findings provide a new entry point for reconsidering basal ganglia-cortical communication beyond the classical indirect loop model. In this review, we use this striato-cortical cholinergic pathway as a recent example, considered alongside the more established pallidocortical literature, to discuss the emerging concept of striato-pallidal-cortical return signaling. We first revisit the historical organization of corticostriatal and basal ganglia-thalamocortical loops, discuss the dorsal tail of the striatum as a sensory-associated basal ganglia node, and finally integrate recent evidence for direct and reciprocal pallidocortical pathways. Together, we propose an expanded model in which basal ganglia circuits may influence cortical gain, timing, sensory salience, and behavioral relevance through multiple return channels. Rather than replacing the classical model, this emerging view expands it: the striatum is not only a recipient of cortical input, but also a participant in return pathways that can contribute to shaping cortical computation.

Information flow in neurons proceeds by integrating inputs in dendrites, generating action potentials near the soma, and releasing neurotransmitters from nerve terminals in the axon. We found that in the striatum, acetylcholine-releasing neurons induce action potential firing in distal dopamine axons. Spontaneous activity of cholinergic neurons produced dopamine release that extended beyond acetylcholine-signaling domains, and traveling action potentials were readily recorded from dopamine axons in response to cholinergic activation. In freely moving mice, dopamine and acetylcholine covaried with movement direction. Local inhibition of nicotinic acetylcholine receptors impaired dopamine dynamics and affected movement. Our findings uncover an endogenous mechanism for action potential initiation independent of somatodendritic integration and establish that this mechanism segregates the control of dopamine signaling between axons and somata.

The brain's response to sensory input is strikingly modulated by behavioral state. Notably, the visual response of mouse primary visual cortex (V1) is enhanced by locomotion, a tractable and accessible example of a time-locked change in cortical state. The neural circuits that transmit behavioral state to sensory cortex to produce this modulation are unknown. In vivo calcium imaging of behaving animals revealed that locomotion activates vasoactive intestinal peptide (VIP)-positive neurons in mouse V1 independent of visual stimulation and largely through nicotinic inputs from basal forebrain. Optogenetic activation of VIP neurons increased V1 visual responses in stationary awake mice, artificially mimicking the effect of locomotion, and photolytic damage of VIP neurons abolished the enhancement of V1 responses by locomotion. These findings establish a cortical circuit for the enhancement of visual response by locomotion and provide a potential common circuit for the modulation of sensory processing by behavioral state.

Transmission from striatal cholinergic interneurons (CINs) controls dopamine release through nicotinic acetylcholine receptors (nAChRs) on dopaminergic axons. Anatomical studies suggest that cholinergic terminals signal predominantly through non-synaptic volume transmission. However, the influence of cholinergic transmission on electrical signaling in axons remains unclear. We examined axo-axonal transmission from CINs onto dopaminergic axons using perforated-patch recordings, which revealed rapid spontaneous EPSPs with properties characteristic of fast synapses. Pharmacology showed that axonal EPSPs (axEPSPs) were mediated primarily by high-affinity α6-containing receptors. Remarkably, axEPSPs triggered spontaneous action potentials, suggesting that these axons perform integration to convert synaptic input into spiking, a function associated with somatodendritic compartments. We investigated the cross-species validity of cholinergic axo-axonal transmission by recording dopaminergic axons in macaque putamen and found similar axEPSPs. Thus, we reveal that synaptic-like neurotransmission underlies cholinergic signaling onto dopaminergic axons, supporting the idea that striatal dopamine release can occur independently of somatic firing to provide distinct signaling.

Goal-directed behaviors involve distributed brain networks. The small size of the mouse brain makes it amenable to manipulations of neural activity dispersed across brain areas, but existing optogenetic methods serially test a few brain regions at a time, which slows comprehensive mapping of distributed networks. Laborious operant conditioning training required for most experimental paradigms exacerbates this bottleneck. We present an autonomous workflow to survey the involvement of brain regions at scale during operant behaviors in mice. Naive mice living in a home-cage system learned voluntary head-fixation (>1 hr/day) and performed difficult decision-making tasks, including contingency reversals, for 2 months without human supervision. We incorporated an optogenetic approach to manipulate activity in deep brain regions through intact skull during home-cage behavior. To demonstrate the utility of this approach, we tested dozens of mice in parallel unsupervised optogenetic experiments, revealing multiple regions in cortex, striatum, and superior colliculus involved in tactile decision-making.