Heike Naumann

SUMMARY: The vertebrate hindbrain sustains functions essential for survival and behavioral control, yet how it operates as a whole remains poorly understood. Here we map its functional organization in larval zebrafish, whose transparency permits imaging of the entire hindbrain. We first curate established functional populations from the literature into an atlas annotated with putative mammalian counterparts. We then complement this library with ensembles of co-active neurons extracted without prior labels from whole-hindbrain light-field recordings in behaving larvae. We identify 70 ensembles that recur across individuals and recover most of them in independent light-sheet datasets. Ensembles also align with spatial gene-expression patterns and share their bilateral symmetry. Behavioral characterization identifies known eye- and tail-movement-related populations and reveals new associations with fin and mouth movements. Combining co-activity, molecular signatures, and neuronal projections suggests a candidate recurrent inter-regional circuit. Our atlas thus provides a common spatial reference for linking functional, molecular, and anatomical data to guide circuit discovery. HIGHLIGHTS: hindbrain ensembles recur across fish, and most are robustly detected across methods ensembles match gene expression and share its bilateral symmetry track movement of multiple body parts around swimming onset maps link activity, molecular and projection data to guide circuit discovery

Animals continuously extract and evaluate diverse sensory information from the environment to guide behavior. Yet, how neural circuits integrate multiple, potentially conflicting, inputs remains poorly understood. Here, we use larval zebrafish to address this question, leveraging their robust optomotor response to coherent random dot motion and phototaxis towards light. We demonstrate that animals employ an additive behavioral algorithm of three visual features: motion coherence, luminance level, and changes in luminance. Using brain-wide two-photon imaging, we identify the loci of these computations, with the anterior hindbrain emerging as a multifeature integration hub. Through single-cell neurotransmitter and morphological analyses of functionally identified neurons, we characterize potential connections within and across computational nodes. These experiments reveal three parallel and converging computational pathways, matching our behavioral results. Our study provides a mechanistic brain-wide account of how a vertebrate brain integrates multiple features to drive sensorimotor decisions, bridging behavioral algorithms with their neural implementation.

RNA ligases are essential for the repair, splicing, and editing of RNA across various biological systems. Recently, a new enzyme that catalyses 5’-3’ RNA ligation – RNA ligase 1 (Rlig1) – was identified in vitro . However, the in vivo biological functions of Rlig1 have remained elusive. Here, we reveal the role of Rlig1 during vertebrate development using embryonic and larval zebrafish as a model system. We found that rlig1 mRNA is maternally deposited and present ubiquitously during early embryogenesis, whereas at larval stages it localises to the brain and eyes. Interestingly, CRISPR/Cas9-generated rlig1 knockout zebrafish exhibited no overt morphological abnormalities, but showed reduced behavioural responsiveness to visual stimuli along with massively perturbed transcriptomes and widespread dysregulation of core metabolic and translational pathways. Brain-wide calcium imaging in rlig1 knockout larvae revealed decreased neuronal activity in key regions for visual processing, consistent with the observed behavioural defects. Together, our findings identify a role for Rlig1 in maintaining the integrity and function of the nervous system and uncover a new link between neuronal RNA processing, development, and sensory-motor computation.

SUMMARY: Animals continuously extract and evaluate diverse sensory information from the environment to guide behavior. Yet, how neural circuits integrate multiple, potentially conflicting, inputs during decision-making remains poorly understood. Here, we use larval zebrafish to address this question, leveraging their robust optomotor response to coherent random dot motion and phototaxis towards light. We demonstrate that animals employ an additive behavioral algorithm of three visual features: motion coherence, luminance level, and changes in luminance. Using brain-wide two-photon imaging, we identify the loci of these computations, with the anterior hindbrain emerging as a multifeature sensory integration hub. Through single-cell neurotransmitter and morphological analyses of functionally identified neurons, we characterize potential connections within and across computational nodes. These experiments reveal three parallel and converging pathways, matching our behavioral results. Our study provides a mechanistic brain-wide account of how a vertebrate brain integrates multiple features to drive sensorimotor decisions, bridging the algorithmic bases of behavior and its neural implementation.

To integrate noisy visual motion, vertebrate circuits must transform transient sensory signals into persistent, lateralized activity. The synaptic architecture supporting this computation remains unclear. We combined calcium imaging and correlative electron microscopy (EM) in larval zebrafish with an independent whole-brain EM dataset, reconstructing 1,246 neurons in and connected to the anterior hindbrain, a region previously implicated in the temporal integration of visual evidence. We identified four broad functional and morphological classes interconnected through recurrent excitation, interhemispheric and ipsilateral inhibitory motifs, as well as long-range inputs from rostral visual regions. We then trained bilateral neural networks constrained by the measured neuron-to-neuron connectivity as a fixed anatomical mask, while optimizing unknown physiological weights. These models reproduced population dynamics and predicted responses to previously untested motion stimuli. Shuffling connectivity or selectively removing recurrent connections impaired model performance, whereas removing an equivalent number of random connections across the network had much smaller effects. Our results identify an anatomical network compatible with visual-motion integration and implicate specific recurrent connections in its computational dynamics. Together, our results link functional dynamics with synaptic architecture in a vertebrate sensorimotor circuit and provide a framework for interpreting connectomic datasets through combined functional, anatomical, and computational analyses.

Cancer cells display distinct, recurrent phenotypic cell states. Metastatic spreading correlates with tumor cell state evolution. However, the molecular mechanisms underlying metastasis remain elusive. Here, we demonstrate that the quantitative dosage of oncogenic KRAS drives lung adenocarcinoma progression and metastasis via the integration of external signaling and pioneer transcription factor dynamics into qualitative cell states. Combining mouse models, in vivo CRISPR activation screens, and fate mapping, we show that even mild transcriptional amplification of KRAS significantly fuels tumor progression and metastasis. Chromatin profiling and transcriptomics reveal that high and low KRAS dosages supersede and integrate inflammatory and TGFβ signaling to dictate mouse cancer cell states. Patient data show that KRAS dosages correlate with distinct survival outcomes, transcription factor activity, and cell states. Direct KRAS inhibition in xenografts limits the KRAS-high “proliferative” cell state but spares a minimal residual state mimicking the KRAS-low “ciliated-like” state. Thus, oncogenic KRAS dosage fuels tumor heterogeneity at the cell state level and drives a bimodal tumor evolution during metastasis, with implications for prognosis and treatment.

Glioblastoma is a lethal brain tumor that exhibits heterogeneity and resistance to therapy. Our understanding of tumor homeostasis is limited by a lack of genetic tools to selectively identify tumor states and fate transitions. Here, we use glioblastoma subtype signatures to construct synthetic genetic tracing cassettes and investigate tumor heterogeneity at cellular and molecular levels, in vitro and in vivo. Through synthetic locus control regions, we demonstrate that proneural glioblastoma is a hardwired identity, whereas mesenchymal glioblastoma is an adaptive and metastable cell state driven by proinflammatory and differentiation cues and DNA damage, but not hypoxia. Importantly, we discovered that innate immune cells divert glioblastoma cells to a proneural-to-mesenchymal transition that confers therapeutic resistance. Our synthetic genetic tracing methodology is simple, scalable, and widely applicable to study homeostasis in development and diseases. In glioblastoma, the method causally links distinct (micro)environmental, genetic, and pharmacologic perturbations and mesenchymal commitment.Significance:Glioblastoma is heterogeneous and incurable. Here, we designed synthetic reporters to reflect the transcriptional output of tumor cell states and signaling pathways' activity. This method is generally applicable to study homeostasis in normal tissues and diseases. In glioblastoma, synthetic genetic tracing causally connects cellular and molecular heterogeneity to therapeutic responses.This article is highlighted in the In This Issue feature, p. 521

Supplementary Figures, legends and References

Supplementary Table S2. Phenotypic CRISPR-screen data analysis