Single Unit publications

With practice, rapid early gains in performance are followed by a slower phase marked by kinematic refinement, automaticity and enhanced cortical and striatal interactions. While sleep is known to support early learning, its causal role in the slow phase of learning is not known. Here we recorded neuronal activity in primary motor cortex (M1) and the dorsolateral striatum (DLS) during long-term skill acquisition and interleaved sleep. Surprisingly, the slow phase of learning was marked by the emergence of a previously unrecognized 5–10 Hz oscillatory activity during NREM sleep that was coherent across M1 and DLS. This oscillation resulted in repeated joint reactivation of task-specific information in cortex and striatum. Strikingly, during later stages of training, such joint reactivation of task activity increased over the course of NREM sleep, suggesting that sleep-dependent processing strengthens cortico-striatal interactions. The strength of M1–DLS coherence was predictive of next-day performance gains and increased cortico-striatal coupling during task performance. Targeted closed-loop disruption of this oscillation during NREM sleep abolished performance gains, whereas switching to a dose-matched random stimulation paradigm enabled performance improvements in the same animals. Importantly, the same cortico-striatal 5–10 Hz rhythm was also found in sleeping non-human primates, where it was selectively enhanced following learning. Together, we identify, across species, a novel NREM sleep oscillation that is important for sleep-dependent performance gains which depend on cortico-striatal processing.

Neuronal variability is a fundamental feature of neuronal coding. In spiking activity, across-trial variance (ATV) normalized by mean spike count (i.e., the Fano factor) shows stimulus-induced quenching. ATV has also been studied in electro/magneto-encephalography (E/MEG) without normalization, revealing effects of stimulation and cognition. Here we show that outside of event-related potentials (ERPs), ATV for both EEG and local field potential (LFP) is nearly identical to intra-trial variance (ITV), which equals signal power. ATV–ITV correlation decreases during ERPs, proportional to how much the ERP explains total power. While EEG ATV shows post-stimulus quenching, LFP ATV does not, particularly in the gamma band, where power increases after stimulus onset. To provide an independent variability measure, we introduce CV(power), the coefficient of variation of signal power, as a mean-normalized metric. CV(power) is the inverse of the signal-to-noise ratio of power and thus related to decodability, offering a variability measure applicable to E/MEG and LFP.

The lateral geniculate nucleus (LGN) relays color information from the retina to the cerebral cortex via anatomically segregated pathways. Although accumulating evidence suggests that neurons conveying signals from short-wavelength-sensitive (S) cones reside in the koniocellular (K) layers, it has been technically challenging to definitively correlate physiological recordings with their histological locations in macaques, which are routine trichromatic primate. In this study, we investigated the distribution of blue-responsive (S+) cells in the macaque LGN using extracellular single-unit recordings combined with a novel, high-resolution, and minimally invasive technique for marking recording sites. Recorded units were classified using cone-isolating stimuli to identify their cone inputs, and the recording site of S+ units were marked for histological verification. We found that the majority of cells responsive to S+ stimulation were located either within the K-layers or at the border between the K- and parvocellular (P) layers. We also found one instance located in the pregeniculate nucleus. Our results provide strong evidence that both New World and Old World monkeys share this color-segregating pathway in the LGN.

The inferior colliculus (IC) is a critical auditory processing center. IC neurons exhibit diverse responses to different sound features, but how IC neurons encode co-varying sound features—such as both the direction and intensity of a frequency-modulated (FM) sweep—is poorly understood. Here, we show that individual IC neurons and neuronal populations multiplex information about FM sweep direction, frequency range, and speed through distinct coding strategies, including both spike timing patterns and overall firing rate.

Objectives. Spatial mapping of sound frequencies along the auditory pathway is refined through sound-evoked activity during early development. However, how neonatal auditory deprivation alters this organization remains unclear. This study examined frequency-specific spatial representations in the inferior colliculus (IC) using targeted low-frequency (apical) and high-frequency (basal) cochlear stimulation, enabling direct evaluation of low-frequency pathway integrity after neonatal deafness.Methods. Sixteen Sprague-Dawley rats (8 normal-hearing [NH] and 8 neonatally deafened [ND]) were studied. Neonatal deafness was induced by daily kanamycin injections from postnatal days 4 to 17. At 8–10 weeks of age, all animals underwent unilateral cochlear implantation with bipolar electrode pairs positioned in apical and basal cochlear regions to activate low- and high-frequency pathways. Multiunit responses were recorded from the contralateral IC using a multichannel linear electrode array. Spatial tuning curves (STCs) based on cumulative discriminability were used to assess thresholds, spread of excitation, and tonotopic precision. Spiral ganglion neuron (SGN) density was quantified histologically.Results. Despite 48%–67% SGN loss across cochlear turns, electrically evoked auditory brainstem response thresholds and IC thresholds were comparable between groups, indicating preserved peripheral excitability. However, ND rats exhibited significantly broader STCs (P=0.002) and reduced dorsoventral separation of apical and basal best sites (1.18± 0.67 mm vs. 1.80±0.40 mm, P=0.039), reflecting degraded central tonotopic precision. The effects were most pronounced for apical (low-frequency) stimulation, consistent with the selective vulnerability of low-frequency pathways to early auditory deprivation.Conclusion. These findings provide direct physiological evidence that early auditory experience is critical for refining frequency-organized connectivity in the auditory midbrain. The preferential degradation of apical (low-frequency) spatial tuning offers translational insights for optimizing stimulation strategies in early-onset deafness.

The detection of unexpected sounds, or auditory deviance detection, is thought to be achieved by comparing prediction signals and sensory inputs in the auditory cortex. This process is markedly impaired in patients with schizophrenia. However, the local circuit properties underlying deviance detection, together with their dysfunction in schizophrenia-related conditions, remain poorly understood. We used two-photon calcium imaging to monitor the population activity of cortical neurons during auditory deviance detection. The deviance-detecting component was predominant in layer 2/3 neurons of the higher-order auditory cortex. The deviance-detecting cells were spatially clustered, exhibited robust synchronization, and constitutively expressed immediate-early genes. In the Del(1.5 megabase)/+ mouse model of 22q11.2 deletion syndrome, a model of schizophrenia, deviance-detecting activity in the upper-layer network was disrupted, accompanied by dysregulated gene expression and selective loss of dendritic spines in immediate-early gene–positive neurons. Our findings support the hypothesis that ensemble activation of this neuronal subpopulation amplifies unexpected inputs, driving error signals that propagate to higher-order cortical areas.

Introduction Sensory processing depends on interactions between neural circuits that convey and regulate information across cortical and subcortical networks. Classical frameworks distinguish driving inputs, which transmit sensory content via suprathreshold activation, from modulatory inputs, which alter neuronal excitability without directly eliciting spiking. However, physiological signatures of these circuit types that generalize widely across distributed brain regions remain unclear. Methods We functionally differentiated driving and modulatory circuits in the awake macaque brain by jointly quantifying suprathreshold multiunit activity (MUA) and oscillatory phase coherence (inter-trial coherence, ITC) across eight cortical and thalamic structures during auditory, visual, and eye-movement conditions. Results Preferred sensory stimuli elicited broadband ITC increases accompanied by robust MUA, yielding relatively uniform spectral distributions across adjacent frequency bands, consistent with driving inputs. In contrast, non-preferred sensory and eye-movement related events produced narrowband, frequency-specific ITC modulation without concurrent firing, characterized by dominant peaks at stimulation or event rates, consistent with modulatory inputs. These two response types were observed within individual regions, revealing two separable modes of neural activity. Discussion This narrowband ITC modulation is indicative of coordinated phase alignment, capable of dynamically regulating information transfer mediated by driving inputs across thalamocortical circuits. These findings identify distinct spectrotemporal signatures of driving and modulatory circuit properties across cortical and thalamic levels and demonstrate that subthreshold oscillatory modulation is a widespread phenomenon reflecting cross-modal and saccadic-related influences on neural excitability.

In nature, animals learn to replace predisposed behaviors with new strategies, yet the neural constraints on these transitions are unclear. Using an ethological search task in mice, we reveal medial prefrontal cortical (mPFC) neural correlates of a predisposed win-stay strategy that decays as animals learn to follow a more reliable auditory cue. Auditory cortex (ACx) activity predicts correct trial-by-trial sound-guided search, even on day one of training. This prognostic coding strengthens with learning and emerges from suppressed spiking, most pronounced in neurons tuned laterally to the cue’s spectrum. Chemogenetic disruption reveals ACx contributions to improving performance. Unexpectedly, the global silencing of mPFC accelerates successful usage of sound-tracking, contrary to its canonical role in flexible or stimulus-dependent behavior. Instead, a decentralized multiexpert competition model best predicts behavior and causal perturbations. These findings suggest that mPFC implements a default strategy based on prior knowledge, which actively hinders the expression of more efficient strategies.

Brain activity consists of neural signals dynamically coordinated across spatial and temporal scales. To sample distributed brain-wide activity, we combine chronically implanted ultra-flexible electrodes for subcortical recordings with simultaneous two-photon calcium imaging in mouse neocortex. Flexible electrodes preserve optical access even at steep insertion angles and enable weeks-long single-neuron tracking. With these combined recordings, we demonstrate how subcortical-cortical coupling is modulated across slow brain states and around fast ripple events.

Sleep is critical for consolidating recent experiences into memories. A key mechanism underlying this process is hippocampal sharp-wave ripples (SWRs). Although SWR-dependent memory consolidation is often viewed as an "offline" process, sounds heard during sleep are processed by a highly active auditory system that projects to medial temporal lobe structures, providing a pathway by which sounds may modulate hippocampal activity. While processing salient sounds during sleep can be adaptive, whether ongoing processing of environmental sounds interferes with SWR-dependent memory consolidation remains unknown. To address this question, we recorded neural activity in the CA1 region of the hippocampus in sleeping rats and used a closed-loop system to deliver non-waking broadband noise (BBN) stimuli during or outside of SWRs. We found that BBN suppressed ripple power and reduced SWR rates. BBN delivered during SWRs (On-SWR) produced a greater reduction in ripple power than BBN delivered outside SWRs (Off-SWR). This suppression was accompanied by substantial changes in SWR-associated spiking. We next tested the influence of BBN on memory consolidation using a conditioned place preference paradigm. On-SWR BBN presentation during post-learning sleep abolished memory retention 24 h after learning while sparing memory immediately after sleep. By contrast, Off-SWR BBN weakened memory at both time points. Notably, On-SWR BBN induced a significantly larger impairment in memory 24 h after learning as compared with Off-SWR BBN. These findings suggest that exposure to noise during sleep disrupts hippocampal activity and impairs memory consolidation in a manner that depends on the timing of sounds relative to SWRs.

Brain function relies on coordinated activity across spatial and temporal scales. Single neurons integrate local and long-range connectivity and reflect activity across brain-wide networks. Understanding integrated brain function requires tools capable of recording from anatomically connected populations in distributed brain areas to bridge local and global dynamics. Here, we present high-density, micro-electrocorticography arrays that facilitate multi-scale studies of brain activity. The hybrid arrays integrate the desirable features of silicone elastomers and polyimide films: silicone provides optical transparency and permits repeated penetration with intracortical arrays, while polyimide enables fine photolithographic feature definition. This combination facilitates high-throughput functional mapping to identify targets and insertion of intracortical arrays for dense local sampling. We demonstrate functional mapping in rats, cats, and marmosets, showing how functional maps guide multi-area laminar recordings. Finally, we demonstrate local and feedforward optogenetic stimulation to investigate cortico-cortical interactions. These capabilities establish the hybrid μECoG as a compelling tool for systems neuroscience.

How does the human brain detect and respond to disruptions in breathing? While animal studies have advanced our understanding of respiratory control, breathing distress in humans remains difficult to treat. It often arises not only from pulmonary lesions or brainstem dysfunction but also from how higher brain regions interpret breathing signals shaped by emotion and experience. We recorded intracranial cortical activity in neurosurgical patients during an interoceptive task involving transient breathing challenges. Conscious detection of these disruptions was predicted by early responses in the anterior insula, which routed signals to orbitofrontal and premotor cortices for appraisal and compensation. These cortical regions preferentially encoded inspiratory effort or airflow, revealing signal-specific processing that echoes functional segregation in brainstem centers. The present findings identify a dynamic insular-frontal circuit for sensing and adapting to respiratory challenges, offering insight into the neural basis of breathing awareness and its disruption in disease.

Neuromodulation is widely employed to enhance neural activity and restore impaired circuitry, yet current methods suffer from invasiveness, limited depth, or poor spatial resolution. Temporal interference stimulation (TIS) has emerged as a non-invasive approach to modulate deep brain regions via interference of high-frequency electric fields, achieving spatial selectivity. Despite its potential, the physiological effects remain unexplored. Here, we present a customized liquid metal-based neural interface to record electrophysiological responses to TIS across diverse neural tissues, from brain organoids to mouse brains. The softness of the interface minimizes invasiveness while ensuring stable access to deep neural regions. Electrophysiological analyses reveal TIS-induced dynamics in neurodevelopment and functional connectivity. Furthermore, we investigated latent neural dynamics, presenting population-level neural activity, to compare responses across models of differing complexity and maturity. By combining soft neural interfaces with multi-scale analyses, this study uncovers how neuromodulation shapes neural dynamics and supports the development of future therapeutic strategies.

The frontopolar cortex (FPC) is thought to coordinate the more posterior lateral prefrontal cortex (LPFC) during complex, nonroutine behaviors through high-level functions such as management of multiple goals, exploration, and self-generated decision-making. However, direct neurophysiological comparisons with other prefrontal regions are lacking, leaving the FPC’s putative dominance untested. Contrary to this view, our comparison of neuronal activity across the full anteroposterior LPFC in macaques during six distinct tasks probing these functions revealed a posterior-to-mid LPFC dominance, with resource-allocation signals, novelty- and reward-related learning signals, and modality-invariant decision-monitoring signals all showing a common posterior bias. In contrast, regardless of the task demands, FPC exhibited inflexible encoding of just-executed actions and consistent task-related deactivation resembling default-mode network activity. We identified a turning point in this graded posterior-to-anterior transition from task-positive to task-negative regions around the border between the anterior and middle thirds of the LPFC. These findings challenge the prevailing notion that the LPFC is anterior dominant across primate species and provide evolutionary constraints on theories of human prefrontal organization.

Auditory decision-making requires transforming sensory inputs into meaningful choices by integrating external stimuli with internal cognitive processes. Classical hierarchical models propose sequential processing, where primary sensory areas relay sensory information to higher-order regions that compute decision-related processes. However, whether this sequential hierarchy fully captures cortical dynamics remains unclear. Here, we analyzed simultaneous neuronal recordings from primary auditory cortex (A1) and dorsolateral prefrontal cortex (PFC) in two male macaque monkeys performing a spatial auditory decision-making task. Our analyses revealed a clear temporal hierarchy: sensory information was unidirectionally transferred from A1 to PFC, while decision-related information exhibited a bidirectional transfer, first from PFC to A1, and crucially back from A1 to PFC. These findings extend classical hierarchical assumptions, demonstrating a recurrent loop between A1 and PFC during decision-making, and offer a dynamic perspective on the integration of sensory and cognitive processes in the cerebral cortex.

Purpose: In anesthesia research and clinical practice, the level of consciousness is often assessed by stimulating subjects and observing their responses. It is possible that such stimulation alters the very state of consciousness being assessed. Whether this phenomenon is accompanied by specific cortical activities remains unknown. Here, we tested the effects of behavioral assessments on neuronal activity in the rat primary auditory cortex (A1) and perirhinal cortex (PrC). Methods: Neuronal spiking activity was recorded from microwire arrays targeting A1 and/or PrC in rats (n = 9), whilst presenting 40 Hz click-trains and gradually increasing infusion rates of propofol, remifentanil, ketamine or dexmedetomidine. We assessed righting reflex, pain withdrawal, whisking reflex, and corneal reflex every 12 minutes (every 360 auditory trials). We searched for significant changes in baseline firing rate and stimulus-evoked activities (where present) due to the behavioral assessments, by comparing 4 min intervals before and after the behavioral tests. Results: Out of 1004 units in 20 sessions, we found significantly elevated activity (p < 0.01) following behavioral tests in 32% and 20% of A1 auditory-responsive units and 8.6% and 5.3% of auditory-unresponsive units under propofol and dexmedetomidine, respectively, but not under remifentanil or ketamine. A higher proportion of units modulated their firing following behavioral tests in A1 compared to PrC, with median gains ranging from +0.23 to +1.06, depending on the specific response feature. Conclusion: Behavioral testing increases ongoing firing rate and responses to sounds in some cortical neurons. This increase is more prevalent in A1 than in PrC, mirroring clinical observations that patients exhibit basic responses under sedation, while complex processing, such as memory encoding, is impaired. The degree of neuronal modulation due to behavioral testing varies between anesthetic agents, mirroring their clinical phenotypes.

Platinum electrodes are used in many bionic devices; however, they have limited charge injection capacity (CIC) and signal‐to‐noise ratio (SNR), restricting performance. Iridium oxide has substantially larger CIC and reduced impedance compared to platinum. While these types of electrochemical measurements are typically used to infer improved electrophysiological performance, there are no studies demonstrating this correlation. This article compares the electrochemical and acute electrophysiological recording performance of iridium and electrochemically activated IrOx. Increased iridium activation reduced total impedance at low‐intermediate frequencies and steady‐state electroactive area but had no impact on linear diffusion electroactive area. The total impedance had a poor correlation with background noise, SNR and spike count. The results contribute to the previous literature indicating linear diffusion electroactive area determines the number of neurons within recording distance of the electrode, impacting SNR and spike count. Measurements of impedance are mainly of benefit as an indirect measure of linear diffusion electroactive area, in which case the measurement and analysis must be performed appropriately. A wide range of IrOx activation states can be produced with increased CIC, without concern of neural recording performance impacts. However, higher oxidation states resulting in electrode dissolution or increase in electrode resistance may limit IrOx chronic performance.

Introduction: Transcranial focused ultrasound (tFUS) can noninvasively modulate sensory pathways, but the cell-type-specific mechanisms underlying excitatory or inhibitory effects remain unclear. Here, we investigate how tFUS applied to the somatosensory cortex (S1) influences S1 and posterior medial thalamic nucleus (POm) responses to hind paw vibration-tactile stimulation and which neuronal populations mediate these effects. Methods: Vibration-tactile stimulation evoked potentials (TEPs) and multi-unit activity (MUA) in S1 and POm were recorded from male rats. Optogenetic tagging was used to identify S1 CaMKII-positive, PV-positive, and SST-positive neurons, while waveform features were used to classify putative excitatory (i.e., regular-spiking units - RSUs) and inhibitory neurons (i.e., fast-spiking units - FSUs) in POm. Results: We found that only S1 CaMKII-positive neurons and POm RSUs responded robustly to tactile stimulation. When tFUS was applied to S1, high pulse repetition frequency (PRF), high duty cycle, and high-pressure stimulation (etFUS) produced excitatory modulation of the sensory pathway, whereas low PRF, low duty cycle, and low-pressure stimulation (itFUS) induced inhibitory effects. Further analyses suggested that excitatory modulation may involve activation of S1 CaMKII-positive neurons, while the inhibitory effect may be associated with their deactivation. Conclusions: These findings demonstrate that tFUS exerts bidirectional, parameter-dependent modulation of a sensory pathway and highlight the critical role of CaMKII-positive neurons in mediating these effects. This study provides mechanistic insight into cell-type-specific neuromodulation by tFUS, particularly in bidirectional modulation of a sensory pathway, and informs the optimization of stimulation parameters for targeted therapeutic interventions.

Current disease-sensing devices primarily focus on distinguishing between healthy and diseased states, effective for diagnosis but limited in guiding optimal intervention timing for prevention. We developed a tripartite framework identifying pre-disease state in depression, a reversible phase preceding irreversible onset. Using complex systems theory, we analyzed early-warning signals emerging as biological systems approach critical transitions. Continuous monitoring of nine multimodal biomarkers—spanning electrophysiological, behavioral, and biological—enabled classification into normal, pre-disease, and disease states by quantitatively defining critical points. An artificial intelligence agent classified disease states with 95.2% accuracy using multimodal data, enabled by ultrasoft neural probes for stable, low-damage recordings. Therapeutic validation with a skin-attachable wireless vagus nerve stimulator integrating soft three-dimensional electrodes demonstrated superior efficacy during pre-disease states. Subjects treated during pre-disease showed faster recovery and greater therapeutic responses, while those treated after disease onset failed to achieve full recovery. This framework provides evidence-based rationale for early intervention.

In mammals, slow-wave sleep is characterized by synchronized neuronal activity that alternates between on and off periods. Slow-wave activity (SWA) and synchrony reflect sleep need, are correlated with synaptic strength in cortical circuits and promote synaptic downselection and memory consolidation. Here we assessed whether these core benefits of sleep can be obtained during waking. We locally induced alternating on/off periods during wakefulness using optogenetics in mice. This led to a local ipsilateral reduction in SWA and synchrony during subsequent sleep, and to reduced markers of synaptic strength. Moreover, bilateral induction of off periods over sensorimotor cortex during sleep deprivation restored memory consolidation. Thus, inducing on/off activity during wakefulness is sufficient to reduce local sleep need and fulfill core functions of sleep.