A machine learning approach for real-time cortical state estimation.

Objective.Cortical function is under constant modulation by internally-driven, latent variables that regulate excitability, collectively known as 'cortical state'. Despite a vast literature in this area, the estimation of cortical state remains relatively ad hoc, and not amenable to real-time implementation. Here, we implement robust, data-driven, and fast algorithms that address several technical challenges for online cortical state estimation.Approach. We use unsupervised Gaussian mixture models to identify discrete, emergent clusters in spontaneous local field potential signals in cortex. We then extend our approach to a temporally-informed hidden semi-Markov model (HSMM) with Gaussian observations to better model and infer cortical state transitions. Finally, we implement our HSMM cortical state inference algorithms in a real-time system, evaluating their performance in emulation experiments.Main results. Unsupervised clustering approaches reveal emergent state-like structure in spontaneous electrophysiological data that recapitulate arousal-related cortical states as indexed by behavioral indicators. HSMMs enable cortical state inferences in a real-time context by modeling the temporal dynamics of cortical state switching. Using HSMMs provides robustness to state estimates arising from noisy, sequential electrophysiological data.Significance. To our knowledge, this work represents the first implementation of a real-time software tool for continuously decoding cortical states with high temporal resolution (40 ms). The software tools that we provide can facilitate our understanding of how cortical states dynamically modulate cortical function on a moment-by-moment basis and provide a basis for state-aware brain machine interfaces across health and disease.

Dynamic corticothalamic modulation of the somatosensory thalamocortical circuit during wakefulness.

The feedback projections from cortical layer 6 (L6CT) to sensory thalamus have long been implicated in playing a primary role in gating sensory signaling but remain poorly understood. To causally elucidate the full range of effects of these projections, we targeted silicon probe recordings to the whisker thalamocortical circuit of awake mice selectively expressing Channelrhodopsin-2 in L6CT neurons. Through optogenetic manipulation of L6CT neurons, multi-site electrophysiological recordings, and modeling of L6CT circuitry, we establish L6CT neurons as dynamic modulators of ongoing spiking in the ventro-posterior-medial nucleus of thalamus (VPm), either suppressing or enhancing VPm spiking depending on L6CT neurons' firing rate and synchrony. Differential effects across the cortical excitatory and inhibitory sub-populations point to an overall influence of L6CT feedback on cortical excitability that could have profound implications for regulating sensory signaling across a range of ethologically relevant conditions.

Membrane potential dynamics of excitatory and inhibitory neurons in mouse barrel cortex during active whisker sensing.

Neocortical neurons can increasingly be divided into well-defined classes, but their activity patterns during quantified behavior remain to be fully determined. Here, we obtained membrane potential recordings from various classes of excitatory and inhibitory neurons located across different cortical depths in the primary whisker somatosensory barrel cortex of awake head-restrained mice during quiet wakefulness, free whisking and active touch. Excitatory neurons, especially those located superficially, were hyperpolarized with low action potential firing rates relative to inhibitory neurons. Parvalbumin-expressing inhibitory neurons on average fired at the highest rates, responding strongly and rapidly to whisker touch. Vasoactive intestinal peptide-expressing inhibitory neurons were excited during whisking, but responded to active touch only after a delay. Somatostatin-expressing inhibitory neurons had the smallest membrane potential fluctuations and exhibited hyperpolarising responses at whisking onset for superficial, but not deep, neurons. Interestingly, rapid repetitive whisker touch evoked excitatory responses in somatostatin-expressing inhibitory neurons, but not when the intercontact interval was long. Our analyses suggest that distinct genetically-defined classes of neurons at different subpial depths have differential activity patterns depending upon behavioral state providing a basis for constraining future computational models of neocortical function.

Ipsilateral Stimulus Encoding in Primary and Secondary Somatosensory Cortex of Awake Mice.

Lateralization is a hallmark of somatosensory processing in the mammalian brain. However, in addition to their contralateral representation, unilateral tactile stimuli also modulate neuronal activity in somatosensory cortices of the ipsilateral hemisphere. The cellular organization and functional role of these ipsilateral stimulus responses in awake somatosensory cortices, especially regarding stimulus coding, are unknown. Here, we targeted silicon probe recordings to the vibrissa region of primary (S1) and secondary (S2) somatosensory cortex of awake head-fixed mice of either sex while delivering ipsilateral and contralateral whisker stimuli. Ipsilateral stimuli drove larger and more reliable responses in S2 than in S1, and activated a larger fraction of stimulus-responsive neurons. Ipsilateral stimulus-responsive neurons were rare in layer 4 of S1, but were located in equal proportion across all layers in S2. Linear classifier analyses further revealed that decoding of the ipsilateral stimulus was more accurate in S2 than S1, whereas S1 decoded contralateral stimuli most accurately. These results reveal substantial encoding of ipsilateral stimuli in S1 and especially S2, consistent with the hypothesis that higher cortical areas may integrate tactile inputs across larger portions of space, spanning both sides of the body.SIGNIFICANCE STATEMENT Tactile information obtained by one side of the body is represented in the activity of neurons of the opposite brain hemisphere. However, unilateral tactile stimulation also modulates neuronal activity in the other, or ipsilateral, brain hemisphere. This ipsilateral activity may play an important role in the representation and processing of tactile information, in particular when the sense of touch involves both sides of the body. Our work in the whisker system of awake mice reveals that neocortical ipsilateral activity, in particular that of deep layer excitatory neurons of secondary somatosensory cortex (S2), contains information about the presence and the velocity of unilateral tactile stimuli, which supports a key role for S2 in integrating tactile information across both body sides.

Inferring thalamocortical monosynaptic connectivity in vivo.

As the tools to simultaneously record electrophysiological signals from large numbers of neurons within and across brain regions become increasingly available, this opens up for the first time the possibility of establishing the details of causal relationships between monosynaptically connected neurons and the patterns of neural activation that underlie perception and behavior. Although recorded activity across synaptically connected neurons has served as the cornerstone for much of what we know about synaptic transmission and plasticity, this has largely been relegated to ex vivo preparations that enable precise targeting under relatively well-controlled conditions. Analogous studies in vivo, where image-guided targeting is often not yet possible, rely on indirect, data-driven measures, and as a result such studies have been sparse and the dependence upon important experimental parameters has not been well studied. Here, using in vivo extracellular single-unit recordings in the topographically aligned rodent thalamocortical pathway, we sought to establish a general experimental and computational framework for inferring synaptic connectivity. Specifically, attacking this problem within a statistical signal detection framework utilizing experimentally recorded data in the ventral-posterior medial (VPm) region of the thalamus and the homologous region in layer 4 of primary somatosensory cortex (S1) revealed a trade-off between network activity levels needed for the data-driven inference and synchronization of nearby neurons within the population that results in masking of synaptic relationships. Here, we provide a framework for establishing connectivity in multisite, multielectrode recordings based on statistical inference, setting the stage for large-scale assessment of synaptic connectivity within and across brain structures.NEW & NOTEWORTHY Despite the fact that all brain function relies on the long-range transfer of information across different regions, the tools enabling us to measure connectivity across brain structures are lacking. Here, we provide a statistical framework for identifying and assessing potential monosynaptic connectivity across neuronal circuits from population spiking activity that generalizes to large-scale recording technologies that will help us to better understand the signaling within networks that underlies perception and behavior.

State-aware detection of sensory stimuli in the cortex of the awake mouse.

Cortical responses to sensory inputs vary across repeated presentations of identical stimuli, but how this trial-to-trial variability impacts detection of sensory inputs is not fully understood. Using multi-channel local field potential (LFP) recordings in primary somatosensory cortex (S1) of the awake mouse, we optimized a data-driven cortical state classifier to predict single-trial sensory-evoked responses, based on features of the spontaneous, ongoing LFP recorded across cortical layers. Our findings show that, by utilizing an ongoing prediction of the sensory response generated by this state classifier, an ideal observer improves overall detection accuracy and generates robust detection of sensory inputs across various states of ongoing cortical activity in the awake brain, which could have implications for variability in the performance of detection tasks across brain states.

State-dependent cell-type-specific membrane potential dynamics and unitary synaptic inputs in awake mice.

The cellular and synaptic mechanisms driving cell-type-specific function during various cortical network activities and behaviors are poorly understood. Here, we targeted whole-cell recordings to two classes of inhibitory GABAergic neurons in layer 2/3 of the barrel cortex of awake head-restrained mice and correlated spontaneous membrane potential dynamics with cortical state and whisking behavior. Using optogenetic stimulation of single layer 2/3 excitatory neurons we measured unitary excitatory postsynaptic potentials (uEPSPs) across states. During active states, characterized by whisking and reduced low-frequency activity in the local field potential, parvalbumin-expressing neurons depolarized and, albeit in a small number of recordings, received uEPSPs with increased amplitude. In contrast, somatostatin-expressing neurons hyperpolarized and reduced firing rates during active states without consistent change in uEPSP amplitude. These results further our understanding of neocortical inhibitory neuron function in awake mice and are consistent with the hypothesis that distinct genetically-defined cell classes have different state-dependent patterns of activity.

Diverse Long-Range Axonal Projections of Excitatory Layer 2/3 Neurons in Mouse Barrel Cortex.

Excitatory projection neurons of the neocortex are thought to play important roles in perceptual and cognitive functions of the brain by directly connecting diverse cortical and subcortical areas. However, many aspects of the anatomical organization of these inter-areal connections are unknown. Here, we studied long-range axonal projections of excitatory layer 2/3 neurons with cell bodies located in mouse primary somatosensory barrel cortex (wS1). As a population, these neurons densely projected to secondary whisker somatosensory cortex (wS2) and primary/secondary whisker motor cortex (wM1/2), with additional axon in the dysgranular zone surrounding the barrel field, perirhinal temporal association cortex and striatum. In three-dimensional reconstructions of 6 individual wS2-projecting neurons and 9 individual wM1/2-projecting neurons, we found that both classes of neurons had extensive local axon in layers 2/3 and 5 of wS1. Neurons projecting to wS2 did not send axon to wM1/2, whereas a small subset of wM1/2-projecting neurons had relatively weak projections to wS2. A small fraction of projection neurons solely targeted wS2 or wM1/2. However, axon collaterals from wS2-projecting and wM1/2-projecting neurons were typically also found in subsets of various additional areas, including the dysgranular zone, perirhinal temporal association cortex and striatum. Our data suggest extensive diversity in the axonal targets selected by individual nearby cortical long-range projection neurons with somata located in layer 2/3 of wS1.

In vivo measurement of cell-type-specific synaptic connectivity and synaptic transmission in layer 2/3 mouse barrel cortex.

Intracellular recordings of membrane potential in vitro have defined fundamental properties of synaptic communication. Much less is known about the properties of synaptic connectivity and synaptic transmission in vivo. Here, we combined single-cell optogenetics with whole-cell recordings to investigate glutamatergic synaptic transmission in vivo from single identified excitatory neurons onto two genetically defined subtypes of inhibitory GABAergic neurons in layer 2/3 mouse barrel cortex. We found that parvalbumin-expressing (PV) GABAergic neurons received unitary glutamatergic synaptic input with higher probability than somatostatin-expressing (Sst) GABAergic neurons. Unitary excitatory postsynaptic potentials onto PV neurons were also faster and more reliable than inputs onto Sst neurons. Excitatory synapses targeting Sst neurons displayed strong short-term facilitation, while those targeting PV neurons showed little short-term dynamics. Our results largely agree with in vitro measurements. We therefore demonstrate the technical feasibility of assessing functional cell-type-specific synaptic connectivity in vivo, allowing future investigations into context-dependent modulation of synaptic transmission.

Membrane potential dynamics of neocortical projection neurons driving target-specific signals.

Primary sensory cortex discriminates incoming sensory information and generates multiple processing streams toward other cortical areas. However, the underlying cellular mechanisms remain unknown. Here, by making whole-cell recordings in primary somatosensory barrel cortex (S1) of behaving mice, we show that S1 neurons projecting to primary motor cortex (M1) and those projecting to secondary somatosensory cortex (S2) have distinct intrinsic membrane properties and exhibit markedly different membrane potential dynamics during behavior. Passive tactile stimulation evoked faster and larger postsynaptic potentials (PSPs) in M1-projecting neurons, rapidly driving phasic action potential firing, well-suited for stimulus detection. Repetitive active touch evoked strongly depressing PSPs and only transient firing in M1-projecting neurons. In contrast, PSP summation allowed S2-projecting neurons to robustly signal sensory information accumulated during repetitive touch, useful for encoding object features. Thus, target-specific transformation of sensory-evoked synaptic potentials by S1 projection neurons generates functionally distinct output signals for sensorimotor coordination and sensory perception.