Progress in Neurobiology

Recent studies report rotational population dynamics in spinal cord neuronal activity during rhythmic movements, suggesting computational principles shared with motor cortex. Here we show that primate cervical spinal cord activity does not exhibit rotational dynamics during an alternating single-joint isometric wrist task, instead it displays low-dimensional alternating population patterns. Positive controls confirm presence of rotational structure in motor cortex activity during the same task, indicating distinct computational strategies across the motor axis. Cortical neurons with post-spike effects on motoneurons had activity with dynamics resembling cortical rather than spinal populations.

Adaptive decision-making under ambiguity requires constant integration of reward- and loss-related information to guide behaviour. In humans and rodents, not all individuals maximise gains in decision-making tasks, such as the Iowa Gambling task or its rodent version, the Rat Gambling task (rGT). While the prefrontal and insular cortices have each been shown independently to support optimal probabilistic decision-making, how they interact functionally to shape individual differences in performance remains unclear. Here, we investigated the consequences of bilateral baclofen/muscimol-mediated inactivation of the prelimbic cortex (PLC) or the anterior insular cortex (AIC) vs. their functional disconnection on the performance of Sprague Dawley rats identified as safe (SDMs) or risky decision makers (RDMs) in the RGT. AIC inhibition decreased advantageous choice in SDMs, whereas it increased win-stay responding in RDMs. In contrast, PLC inhibition primarily affected lose-shift behaviour, reducing sensitivity to losses in SDMs while enhancing adaptive switching in RDMs. Functionally disconnecting the PLC from the AIC, which had no effect on the performance of SDMs, improved decision-making in RDMs by increasing loss-guided behavioural adaptation. Together, these findings identify parallel versus serial AIC-PLC processing as a potential neural mechanism underlying the tendency some individuals have to make suboptimal decisions.

Throughout the preclinical phase of Alzheimers disease (AD) {beta}-amyloid (A{beta}) accumulates preferentially within the default mode network (DMN), yet the functional and behavioural consequences of this pathological burden remain poorly understood. Using task-based fMRI combined with A{beta}, tau, and dopamine PET in cognitively normal older adults, we show that A{beta} burden impairs learning independent of tau, but this learning performance is recovered with higher dorsolateral striatal dopamine synthesis capacity. Investigating the neural mechanisms that support this learning, we observe that A{beta} positive individuals show attenuated DMN activity to error related feedback, a metric that relates to poorer learning. When estimating the effective connectivity during feedback, computational modelling reveals that A{beta} induces dis-inhibition of the DMN during error processing. Critically, dopamine synthesis capacity in the dorsolateral striatum rebalances effective connectivity between the DMN and frontostriatal network, thereby opposing A{beta} related disruption. These findings establish a systems-level framework in which A{beta} impairs learning by disrupting dynamic DMN modulation during feedback, a disruption for which dopaminergic function can partially compensate. This suggests that learning in the presence of A{beta} may be subserved by dopamine-dependent network rebalancing, a candidate mechanism of cognitive resilience to support learning in preclinical AD.

Attentional functions enable the nervous system to regulate incoming information, selecting what is behaviorally relevant while filtering out distractions. To investigate attentional control in rats, we developed a paradigm in which animals are required to ignore an irrelevant tactile stimulus (a vibration) and categorize the relevant one as weak or strong. Stimuli were separated in time but delivered to the same set of vibrissae, making this a temporal rather than spatial attention task. In the first task version, the irrelevant stimulus was presented first and the relevant second. Across animals, we observed substantial variability in how this task was learned and performed, consistent with emerging work showing that learning unfolds along idiosyncratic trajectories rather than converging on a single behavioral solution. While the irrelevant stimulus typically exerted an attractive bias on judgments, some rats progressively reduced this influence and achieved near-complete suppression, transitioning from a proficient to an expert stage of performance. Other animals, however, adopted alternative stable strategies and did not exhibit comparable levels of attentional filtering. To probe behavioral flexibility, we designed a version of the task in which the relevant stimulus could appear in either temporal position. Under these conditions, rats generally struggled to flexibly allocate attention across time, again with substantial variability across individuals. Importantly, behavioral analyses indicate that this variability is not random but reflects a small number of reproducible strategy classes, characterized by distinct patterns of sensitivity to stimulus order. These findings suggest that attentional control in this task does not rely on a single canonical algorithm, but instead emerges from a constrained set of alternative solutions. Electrophysiological recordings from a limited number of animals revealed neural correlates consistent with these behavioral differences. In expert animals performing the original task, neuronal populations in motor cortex showed differential encoding of the irrelevant and relevant stimuli. In a proficient animal implanted in both vS1 and M1/M2, outcome-dependent modulation was prominent in vS1, whereas M1/M2 transformed sensory inputs into categorical representations with partial suppression of the irrelevant stimulus. Local field potential analyses further indicated that effective task performance was associated with increased high-gamma synchronization between vS1 and M1/M2, alongside low-beta modulation consistent with top-down interactions. Taken together, these results suggest that learning to ignore irrelevant information reshapes interactions between sensory and motor cortices, but that this process unfolds heterogeneously across individuals. Rather than reflecting a single mechanism of attentional control, the data support a framework in which multiple, identifiable strategies coexist within a shared task structure, underscoring the importance of individual differences in understanding the neural basis of attention.

Intelligent behavior depends on the brain's ability to represent multiple features of the environment simultaneously while keeping those representations independent [1,2]. Patients with Alzheimer's disease often mix up objects, people, and events [3-7], raising the possibility that disease mixes up the way that information is represented in the brain. Here we show that the independence of visual representations progressively breaks down during early stages of disease progression in a rhesus macaque model of Alzheimer's disease and related dementias [8-10]. In visual area V4, representations of different visual features become progressively less independent, such that the representation of one feature is increasingly influenced by the value of another. We term this loss of independence neuronal feature confusion. This neuronal change predicts a specific behavioral consequence: because feature representations become less independent, preferences associated with one visual feature increasingly influence visually guided choices associated with other, independent features. Using an analogous image-selection task, we found the same behavioral signature in people with mild cognitive impairment, distinguishing them from age-matched controls. These results identify a specific and measurable alteration in neuronal population representations that predicts a behavioral change observed across species. More broadly, these findings demonstrate that neuronal population representations can guide the development of sensitive, non-invasive behavioral methods for early detection of functional changes associated with Alzheimer's disease.

Visuospatial attention enables the selective allocation of cognitive resources to relevant stimuli. A well-established neural signature of attentional shifts is the lateralised modulation of occipito-parietal alpha power, with decreases over the hemisphere contralateral to the attended location and increases over the ipsilateral hemisphere. However, growing evidence suggests that multiple oscillatory mechanisms contribute to attentional deployment, including beta-band activity. A key unresolved question that remains is whether the same neural rhythms support the deployment of attention and the perceptual decisions that follow. Here, we recorded EEG in 26 participants (22 females) during covert visuospatial orienting and investigated how alpha- and beta-band dynamics relate to behavioural measures, namely perceptual sensitivity (d') and decision criterion (c), and whether attended location could be preferentially decoded from alpha- or beta-band activity. We found that pre-target beta phase significantly predicted decision criterion at earlier pre-target intervals, whereas perceptual sensitivity was predicted closer to target onset, suggesting that beta is related to both sensory gain and the perceptual decision. In contrast, decoding analyses revealed that attended location was most strongly discriminable from alpha-band activity, as confirmed by time-frequency analysis of decoding accuracy. Together, these findings suggest a functional dissociation between oscillatory mechanisms supporting attentional orienting and perceptual decision-making. Whereas alpha-band activity primarily reflects the allocation of attention, beta-band dynamics predict trial-by-trial variability in perceptual decisions.

Derived from previous observations on equal and cross-frequency coupling, we evaluated the proposal that equal and cross-frequency phase synchronization may characterize the integration-segregation perspective of cerebral sensory-motor processing. Using brain recordings obtained in normal conditions and in conditions of diminished sensory input (eyes closed wakefulness, sleep and coma, when there is presumably less functional segregation of sensory-motor processing in neural networks), we assessed potential differences in partitioning of the synchrony state space linked to cross-frequency synchronization. More partitions were found in conditions of decreased sensory input. In addition, there was a less complex synchrony state space in cross-frequency as compared with equal-frequency coupling, in terms of fewer connectivity configurations. These results support the idea that equal-frequency coupling favours integration from multiple brain regions occurring in a complex synchrony state space rich in possible connectivity configurations, whereas cross-frequency coupling contributes to segregation, or localized sensory-motor transformations taking place in specific brain areas. This evidence may contribute to new considerations about the much-discussed role of multi-frequency relations in neuronal activity, and how the structural and functional modular organization of the nervous system is able to generate the coordinated activity needed for conscious and appropriate cognitive behaviors in complex environments.

Voluntary movement requires the central nervous system to transform and integrate visual and somatosensory information into coordinated motor outputs. Although mirrorlike neuronal activity during both action execution and observation has been extensively described in premotor, motor, and parietal cortices, it remains unknown whether the primary somatosensory cortex (S1) also participates in the action observation network. Here, we recorded single-unit activity from cytoarchitectonically defined areas 3a, 3b, 1, and 2 in macaque S1 while monkeys either executed or observed grasping movements. Approximately one-third of neurons across S1 modulated their firing during action observation, with the proportion of responsive neurons increasing from area 3 to areas 1 and 2, consistent with the hierarchical organization of somatosensory processing. Most action observation neurons showed congruent activity during action execution and observation, suggesting that these responses may reflect top-down motor-related or integrated visuomotor signals and are unlikely to be explained by visual input alone. The higher prevalence of action observation neurons in areas 1 and 2 suggests that action observation-related signals preferentially influence later stages of somatosensory processing, potentially via cortico-cortical interactions with motor and parietal regions.

Complex reasoning depends on flexible coordination among frontal, parietal, and thalamic systems, but the circuit mechanisms that support increasing relational demands remain unclear. We combined EEG with biologically grounded corticothalamic neural field modelling while participants solved relational problems of graded complexity. Successful reasoning was associated with dissociable frontoparietal dynamics. Frontal regions showed increased theta-band power, whereas parietal regions showed reduced alpha- and beta-band power. Theta-band phase synchronisation across frontoparietal-network nodes increased with problem complexity but was not associated with performance. By contrast, stronger beta-band synchronisation across the same network was associated with slower and less accurate responses as demands approached the highest complexity, suggesting that stronger coordination is not uniformly beneficial. Neural field modelling indicated that these regional spectral dynamics reflected specific complexity-dependent circuit adaptations. Parietal regions showed modulation of intracortical and corticothalamic gains, intrathalamic inhibition, prolonged loop delays, and faster synaptic filtering, whereas frontal regions primarily adjusted intracortical gains to maintain local excitatory-inhibitory balance and supported longer temporal integration windows. Together, these empirical and model-derived findings reveal complementary frontoparietal and corticothalamic mechanisms for relational reasoning.

Motor planning and control is executed by different motor areas within the neocortex. Despite their distinct functions these areas are built by the same archetypes of neurons as the rest of the cortex, with the pyramidal neurons (PNs) as their principal building blocks. Recent results suggest that the synapses of the PNs are modeled and adapted to their required functions in an area specific manner. PN synapses in a cortical area engaged in higher order functions, the prefrontal cortex (PFC), were found to operate with loose microdomain calcium-influx-to-release coupling and showed short-term facilitation, whereas synapses processing sensory information in a lower order cortical area, the primary somatosensory cortex (S1), featured tight nanodomain coupling and showed short-term depression. In the present study, we asked for the functional coupling configuration of an intermediate processing area. We focused on PN synapses in the premotor cortex M2 and compared their properties to those of PN synapses in the primary motor cortex M1. In both areas we found tight nanodomain coupling and high release probability, but a significant difference in short-term plasticity. Synapses in M1 showed paired-pulse depression similar to S1. In contrast, synapses in M2 exhibited paired-pulse facilitation. Our data suggest that this facilitation results from an accelerated recruitment of synaptic vesicles to the readily releasable pool from an enlarged replenishment pool. Thus, PN synapses in M2 appear to have properties intermediate between those in PFC and M1.

Parkinsons disease (PD) is known to alter the intrinsic properties of striatal cholinergic interneurons (ChIs). However, how PD shapes ChI control of intrastriatal GABAergic circuits regulating principal spiny projection neurons (SPNs) is unknown. To fill this gap, striatal circuits in healthy and parkinsonian mice were interrogated. In ex vivo brain slices from healthy mice, optogenetic stimulation of ChIs evoked GABAA receptor currents in both indirect and direct pathway SPNs that were attributable to nicotinic acetylcholine receptor (nAChR)-mediated activation of GABAergic interneurons (GIs). Simulations suggested that this circuit exerts a state-dependent control of SPN dendritic integration that was modulated by concomitant muscarinic receptor signaling. Surprisingly, in mouse models of prodromal and parkinsonian states, the ability of ChIs to engage this intrastriatal circuitry was disrupted because interneurons down-regulated nicotinic AChRs. Taken together, these studies suggest that impaired ChI control of GABAergic interneurons contributes to behavioral deficits in both prodromal and clinical PD states.

The hippocampus supports spatial memory by dynamically integrating external sensory inputs with intrinsic neural circuit dynamics during novel experience. In Alzheimers disease (AD), despite impaired hippocampal spatial remapping, spatial learning and memory abilities can remain partially preserved, a phenomenon consistent with cognitive resilience (Gomez-Isla and Frosch 2022, Jia, Xu et al. 2025). However, the hippocampal ensemble coding patterns associated with these preserved learning and memory abilities remain remains unclear. We hypothesize that intrinsic temporal structures of neuronal firing continue to facilitate the encoding of new spatial information. Using the AppNL-G-F rat model, we longitudinally tracked hippocampal CA1 activity during a familiar-novel context alternating task. We found a dissociation between impaired explicit spatial coding and preserved implicit temporal coding in the AD hippocampal network. Explicit spatial coding was impaired, as place cells showed weak discrimination between distinct contexts and failed to improve with learning. In contrast, implicit temporal coding exhibited learning-dependent refinement, with cofiring dynamic becoming increasingly context-specific across long-term experience. Further analysis suggested that the enhancement of implicit cofiring may be associated with the increased consistency of neural ensemble reactivation during sharp wave ripples in awake rest. Taken together, these findings reveal an explicit-implicit dissociation in the AD hippocampal network, suggesting that the learning-dependent refinement of implicit temporal coding may support preserved learning capacity despite impaired spatial remapping.

Attention is a fundamental mechanism enabling the brain to overcome its limited capacity for parallel processing. In non-human primates, invasive electrophysiology has shown that attentional selection operates rhythmically, primarily within the alpha ([~]8-12 Hz) and theta ([~]4-5 Hz) bands. Whether such finely resolved control signals can be captured non-invasively in humans, and how they adapt to changing task demands, remains unclear. Using high-precision magnetoencephalography (MEG) combined with machine learning, we decoded the spatial locus of covert attention in humans performing three variants of a spatial cueing task that manipulated cue validity as well invalid trial switching rules. Spatial attention could be decoded from whole-brain MEG activity at both static and time-resolved scales, with accuracies significantly above chance (N = 30). Decoding performance decreased as cue validity was reduced, indicating that task structure shapes attentional engagement. Analysis of decoding trajectories revealed rhythmic fluctuations at [~]8-12 Hz across all tasks, demonstrating alpha-band sampling of attention. Pre-target attention became increasingly focused on the cued side, especially in the 100% Valid condition, consistent with proactive orienting. Furthermore, individual and task-specific differences in decoding strength correlated with task-variations in behavioral performance, linking the accuracy of neural attention codes to both discrimination accuracy and reaction time. These findings demonstrate that MEG can non-invasively capture dynamic, task-dependent fluctuations in spatial attention that parallel those observed in non-human primates. They reveal that attentional demands reshape the neural code for attention, modulate rhythmic sampling, and influence behavioral efficiency. This work bridges invasive primate and non-invasive human research and establishes MEG-based decoding of attention as a promising tool for mechanistic and clinical applications, including neurofeedback and attention-related interventions.

Humans and non-human animals adaptively boost their encoding speed when they expect limited sensory exposure time, so that they can capture essential information before it is gone. However, it is unclear how the brain implements this crucial adaptation. Using multivariate pattern analysis of human EEG data, we found that an accelerated neural code underlies adaptations in visual working memory encoding speed.

Whether proprioception is necessary for upper-limb motor control has been debated for decades. Classic studies in deafferented animals and humans suggested that proprioception may be dispensable for rapid, goal-directed movements. However, chronic sensory loss conflates the absence of proprioceptive input with years of compensatory adaptation. As a result, the field has lacked a strong causal test of proprioceptions contribution to motor control. Here, we leveraged a clinical trial of cervical spinal cord stimulation (SCS) in individuals with chronic post-stroke hemiparesis to study if electrical stimulation of sensory afferents causally perturbs proprioception and affects arm reaching. We found that turning SCS ON causally impaired proprioceptive perception and postural stabilization in response to force perturbations, and enhanced adaptation to visual errors during implicit learning. Yet visually and non-visually rapid, goal-directed reaching improved in smoothness, straightness and spatial accuracy. These findings provide strong causal evidence that proprioception is not required for rapid, goal-directed action, helping resolve a decades-long debate regarding its necessity for effective movement.

Theta oscillations are thought to provide a temporal scaffold for short-term memory (STM), organizing item encoding and maintenance into successive phases to reduce representational conflict. Whether this rhythm also determines encoding fidelity, that is, how precisely items are encoded in human cortical activity, remains unclear. Here, we show that STM encoding fidelity fluctuates rhythmically at theta frequency. EEG was recorded while participants encoded arrays of colored, oriented objects under two memory loads and, after a delay, reported the features of a retrospectively cued item on a continuous scale. Using time-resolved multivariate pattern analysis, we predicted subsequent recall error from encoding- and maintenance-period activity. Fronto-parietal theta- and beta-band activity predicted subsequent memory error. This prediction was not sustained but fluctuated at theta frequency and was not modulated by memory load. Cross-temporal generalization indicated that the same neural pattern recurred across encoding and maintenance, underlying the rhythmic fluctuations in memory-error prediction. Prediction fluctuations were temporally offset across spatial positions and object features. These findings characterize STM encoding as a rhythmic, recurrent process and link behavioral theta fluctuations to a distributed neural mechanism.

The neural correlates of conscious perception remain debated, particularly regarding the role of extra-sensory regions such as the prefrontal cortex. One promising approach is to study the dynamical properties of neural processing in task-related and task-free contexts. A previous EEG study showed that conscious perception is associated with all-or-none late activations, giving rise to bifurcation dynamics even without a task. Here we used fMRI and near-threshold auditory stimulation to ask which brain networks give rise to these bifurcations, and how they differ depending on task. In both contexts, stimulus intensity modulated activity within broad networks spanning sensory and extra-sensory regions, including the prefrontal cortex. These networks showed both shared and distinct components. Single-trial modelling further revealed bifurcation dynamics beyond primary sensory cortices and enabled single-trial prediction of conscious perception in both contexts. These findings resolve previous conflicting results and reveal common networks and dynamics underlying conscious perception irrespective of task.

The medial entorhinal cortex (MEC) has classically been viewed as a spatial coding region, but growing evidence indicates that it also contains temporal representations. However, how spatial and temporal codes are organized within the MEC remains unclear. Here we recorded MEC neurons with high-density silicon probes while mice performed a MEC-dependent timing task and a virtual reality spatial navigation task in a similar head-fixed setup. We found that spatial and temporal representations were partially overlapping but systematically biased across the MEC population. Grid cells and non-grid cells with strong spatial tuning were less likely to show reliable time-locked activity during the timing task. In contrast, neurons with weaker spatial tuning more flexibly shifted their coding scheme to temporal coding during timing task. Moreover, spatial tuning strength and its negative relationship with temporal tuning were preserved in a distinct open field environment, indicating that the coding preferences of individual neurons are constrained by a stable network-level organization. Together, these findings suggest MEC is organized along a coding gradient, ranging from dedicated stable spatial coding neurons to more flexible spatial or temporal coding neurons which represent information according to cognitive demands.

The neuronal mechanisms driving progression in neuroinflammatory disorders from early relapse-remitting phases to later neurodegenerative phases remain largely elusive. Functional brain state shifts towards hyperactivity, persisting beyond relapses, represent an early maladaptive response. Here, in remission stage of an experimental autoimmune encephalitis (EAE) mouse model of RRMS, we identified a reduced excitability upon optogenetic stimulation in the brain stem, the area of active disease, while in the cortex a persistent cortical neuronal hyperactivity and synaptic remodeling emerged, accompanied with an increase of markers of early apoptosis. In contrast, hippocampal circuits, which undergo a functional state shift without hyperactivity, do not show increased apoptosis. Visual cortical networks showed a deterioration of the accuracy of encoding visual information and a decrease in the behavioural visual discrimination ability in mice. In RRMS patients in remission, we identified a reduced visual colour discrimination, indicating both the presence and the clinical relevance of early brain state maladaptation that may contribute to progression independent from relapse activity (PIRA). SummaryIn a RRMS model and in patients, impaired visual processing was reported, indicating brain state maladaptations, associated with persistent cortical hyperactivity, brain stem hypoactivity, synaptic remodeling, and apoptosis. These maladaptations might contribute to relapse-independent disease progression through sustained network dysfunction.

The brains ability to consolidate a wide range of memories while maintaining their distinctiveness across experiences remains poorly understood. Sharp-wave ripples, neural oscillations that occur predominantly within CA1 of the hippocampus during immobility and sleep, have been shown to play a critical role in the consolidation process. More recently, evidence has uncovered functional heterogeneity of pyramidal neurons within distinct sublayers of CA1 that display unique properties during ripples, potentially contributing to memory specificity. Despite this, it remains unclear exactly how ripples shift the activity of CA1 neuronal populations to accommodate the consolidation of specific memories and how sublayer differences manifest. Here, we studied interactions between the anterior cingulate cortex (ACC) and CA1 neurons during ripples and discovered a reorganization of their communication following learning. Specifically, using a generalized linear model decoder, we demonstrated the pre-existence of ACC-to-CA1 communication, which is weakened during post-training sleep following learning, suggesting that ACC activity reallocates the contribution of CA1 neurons during memory formation. Interestingly, the reorganization appeared unique for a subset of CA1 superficial (CA1sup) neurons that were task inactive, whereas communication between the ACC and CA1 deep neurons remained largely stable across pre- and post-training sleep. Consistent with this sublayer-selective reorganization, we found that optogenetic stimulations of the ACC preferentially suppressed CA1sup neurons while activating a unique subset of CA1 interneurons. Overall, these findings highlight an important role of the ACC in rebalancing CA1 neuronal populations contribution in learning and memory consolidation.