Karl Deisseroth

Microbial genomes encode non-coding RNAs (ncRNAs) and nucleic-acid-interacting proteins essential to biology and biotechnology. However, discovery of these systems remains protein-centric and based on similarity to known sequences. Here, we introduce Minerva, a framework for coevolutionary mining that uses genome language models to predict pairwise interactions directly from sequence. Minerva enables fast, alignment-free prediction of base pairing, monomeric protein contacts, and repetitive sequence motifs. We recover known systems such as CRISPR, detect open-reading-frame signatures at the DNA level without supervision, and predict hundreds of unannotated regions per bacterial genome containing multiple hairpins or repetitive motifs. In Pseudomonadota, we find that the widespread TwoAYGGAY structured RNA family carries large secondary-structure extensions and is often flanked by repetitive motifs. In prophages, we discover that Unknown Group 27 reverse transcriptase systems encode variable arrays of structurally conserved yet sequence-diverse ncRNAs that template complementary DNA hairpin products. Coevolutionary mining with Minerva enables efficient biological discovery across microbial genomes and metagenomes.

White matter fiber bundles are the structural conduits of information flow in the human brain and thereby mediate the spatial trajectories and timings of the electrophysiological signaling. Here, we combined diffusion magnetic resonance imaging (dMRI) and stereoelectroencephalography (SEEG) recordings in neurosurgical patients to develop an integrated framework for predicting and interpreting causal electrophysiological connectivity patterns between pairs of brain regions. We used repeated single-pulse electrical stimulation in 40 participants implanted with a total of 5794 intracranial electrodes throughout the human brain, encompassing both cortical and multiple thalamic nuclei. A nonlinear, time-frequency manifold learning approach was used to define electrophysiological connectivity, which was then compared with subject-specific and atlas-based structural connectivity. Across 150,000 electrode pairs, we found that the presence of a structural connection predicted causal electrophysiological connectivity with a probability of ≈ 0.95; and its absence predicted the lack of the direct electrophysiological connectivity with a probability of ≈ 0.8. We also show evidence of indirect/polysynaptic pathways supported by both modalities and reported neural features from time-frequency decomposition that distinguished between direct and indirect signaling. We demonstrated that an early phase-locked broadband component (10–70 ms) marked direct structural pathways, whereas delayed and slower components reflected indirect propagation. Notably, we reported that thalamic involvement within an indirect pathway results in increased latency ( > 200 ms) and enhanced late oscillatory behavior, despite increased conduction velocity measures along thalamo-cortical pathways. Therefore, our multimodal framework maps human brain connectivity, bridging structural architecture, causal electrophysiological dynamics, and network-level communication.

Sleep and wakefulness are associated with distinct cortical patterns of rhythmic activity ¹ . During low-arousal states such as slow-wave sleep, synchronous low-frequency rhythms dominate activity across widespread cortical regions. Although inhibitory neurons are increasingly recognized as key regulators of cortical state ^2–4 , the circuit mechanisms that coordinate synchronized activity across local and distant neocortical networks in vivo remain poorly understood. Here we show in mice that cells co-expressing somatostatin (Sst) and chondrolectin (Chodl)—which constitute a sparse and genetically distinct class of neocortical GABAergic inhibitory neurons—are selectively active during low-arousal states and mostly silent during periods of high arousal. In contrast to most neocortical inhibitory neurons, Sst-Chodl cells, despite being extremely sparse, exert widespread influence across the neocortex, through long-range axons that target multiple regions simultaneously. Selective activation of Sst-Chodl cells is sufficient to promote the multi-region cortical synchronization that is characteristic of low-arousal states and to induce sleep. Together, these findings show that long-range Sst-Chodl inhibitory neurons not only track behavioural state, but can also actively promote synchronized cortical activity and sleep behaviour, highlighting that cortical circuits have a key role in sleep regulation, alongside established subcortical mechanisms.

Encoding novel experiences and retrieving stored information rely on the hippocampal theta rhythm in several mammalian species. Coordination of putative pacemakers in the medial septum (MS) by the GABAergic hippocampo-septal (HS) feedback is hypothesized to be essential for theta genesis. In order to reveal how the HS-feedback controls the MS circuit, and through it hippocampal oscillations in general, we set out to uncover the behavior-related activity of HS neurons as well as how different components of the MS network are regulated by the HS-feedback. Utilizing computational modelling and in vivo optogenetic manipulation of the HS pathway, we show that the HS feedback is not required for theta generation, which is primarily governed by intra-MS connections. HS blockade did not alter the phase coupling of MS neurons to hippocampal theta rhythm nor the synchrony between rhythmic MS neurons but augmented ongoing theta and coupled mid-gamma oscillations in the hippocampus. By contrast, correlated activity between non-rhythmic MS neurons was sensitive to HS blockade. Moreover, the HS feedback scaled with hippocampal activity and running speed. In turn, modulation of MS neurons by running and ripples was diminished by the transient suppression of HS terminals. These findings suggest that the HS feedback exerts inhibitory control of the MS depending upon hippocampal output and movement speed.

Dopamine neurons in the ventral tegmental area (VTA) have roles in motivation, learning, and psychiatric disorders. We found that genetically defined VTA dopamine neuron subtypes had distinct electrophysiological properties, neuronal signaling dynamics in response to reward- and aversion-related stimuli, and roles in somatic optogenetically induced reward dependent on the neurotransmitters they released. Nonglutamate-dopamine neurons increased activity after reward-related stimuli and decreased activity after the omission of an expected reward (negative reward prediction error) and aversion-related stimuli. Glutamate-dopamine and glutamate-only (nonGABAergic and nondopaminergic) neurons were activated by both rewarding and aversive events, but only glutamate-dopamine neurons had sustained cue-induced reward signaling. Recordings of all dopamine neurons without considering glutamate cotransmission showed mixed population responses during prediction error and aversive stimuli that obscured the distinct signaling patterns of its constituent subpopulations. Although the examined cell types largely differ in mediolateral location, cell-type identity better accounted for functional differences than mediolateral location. Glutamate-dopamine neurons were more excitable than nonglutamate-dopamine neurons. Glutamate-dopamine and nonglutamate-dopamine axons had similar but not identical dopamine release dynamics in the nucleus accumbens. Only nonglutamate-dopamine neurons supported somatic optogenetically induced reward and reinforcement. In glutamate-dopamine neurons, the dopamine synthetic enzyme TH contributed to associative learning of aversive or less-beneficial outcomes, whereas the vesicular glutamate transporter VGLUT2 contributed to reward- and exploration-related vigor. Our results suggest that glutamate cotransmission is a distinguishing feature of VTA dopamine neuron signaling patterns and roles in natural reward- or aversion-motivated behavior.

The catalytic synthesis of many abiotic organic compounds is beyond the native capabilities of mammalian cells. Broadening the range of physiologically compatible, bond-forming catalytic reactions could enable the in situ synthesis of diverse molecules within living systems. By controlling these reactions in both specific cells and in physical space, this approach could offer transformative ways to modulate cellular behaviour and treat cells as distinct reaction vessels within organisms. Here we use genetic targeting to incorporate organic photocatalytic dyes onto and within cells to enable cell-type-specific and spatially defined catalysis. Photocatalysis can be restricted to individual cells using mild, green-to-red visible stimulation. The dyes catalyse cross-dehydrogenative C–H functionalization for the light- and cell-specific construction of C–C bonds in situ. Nuclear-localized assembly directs protein translocation with reversed transport directionality, illustrating the utility of this approach to both generate and enhance the function of chemical biology tools. We anticipate that the integration of synthetic catalysts with cell-specific targeting may lay an important foundation for future precision-directed abiotic chemical synthesis in living organisms.

The neocortex covers a vast expanse of the mammalian brain and represents the principal target of clinical neuromodulation; however, the global principles of neocortical operation have been challenging to identify. In this regard, a limitation has been tracking activity with cortex-wide spatial coverage while maintaining access to the millisecond temporal resolution of neuronal firing— a crucial combination not achievable with existing recording technologies. Here we introduce and apply conformal immersion microscopy, enabling activity tracking across the entire dorsal cortex with millisecond temporal resolution and 100 μm spatial resolution (thus spanning five orders of magnitude in time and four in space), at sufficient sensitivity to resolve single-trial activity beyond 100 Hz. Drawing on physics-based frameworks, we apply multiscale analysis to identify a fundamental frequency-dependent coherence length that partitions the neocortex into discrete dynamical elements with well-defined propagation speeds, boundaries, and scale-invariant dynamics. These dynamical elements were found to be conserved from sub-threshold to suprathreshold (neuronal firing) regimes of neural activity, and were robust to diverse pharmacological, optogenetic, and genetic interventions. However, it was possible to identify and establish conditions allowing elemental boundaries to be selectively overridden, and to allow perturbation of specific elements even while conserving global dynamical architecture. Together, these findings enable measurement of intrinsic spatiotemporal parameters governing the dynamical organization of neocortex, which may provide a foundation for mechanistically-informed basic and translational understanding.

Mammalian cognition appears to involve coordinated neural activity within and across brain-spanning networks. However, the governing principles of large-scale integration remain largely unknown. Here we developed an approach to record spontaneous cortex-wide neural activity (including with novel genetically-activity sensors), over very long timescales, and applied unbiased computation to capture conserved spatiotemporal regularities in these data. Initial screening extracted a set of directional elements: consistent spatiotemporal patterns of cortical activity that generalized across excitatory and inhibitory neuronal cell types as revealed by genetic targeting of activity sensors. In particular, novel genetically-encoded voltage-sensing strategies revealed that the directional elements were not only highly conserved, but also were represented across the full neural activity frequency spectrum including the fastest timescales (gamma rhythms) enabled by voltage-sensing. Exploiting the spatiotemporal structure revealed by these directional elements, we tested for causal rules governing cortical network activation, using patterned optogenetic stimulation combined with activity imaging. We found that the directional propagation structure of these elements encoded a causal control hierarchy, as source regions (but not sink regions) sufficed to drive full element recruitment---a principle that held across all tested directional elements. Employing a panel of psychotropic drugs, we showed that directional element structure and excitability were robust to manipulations of neural and behavioral state, even as spontaneous network dynamics were reshaped in compound-specific ways. Finally, we developed and applied an all-optical sensing/control approach targeting the directional elements in a behavioral visual detection paradigm, revealing contributions of these conserved large-scale dynamics to elevated sensorimotor behavioral performance.