Neuron最新文献

The profound autism community must support brain organoid research, and scientists must hurry up and build ethical guardrails so that this new technology can be put to work to improve lives.

Value-guided decisions are a cornerstone of cognition, yet the underlying circuit-level mechanisms remain elusive. We used reinforcement learning to train recurrent neural network models endowed with Dale's law on a battery of economic choice tasks, which revealed a two-stage computational framework. First, value estimation occurs at the input level, where learned weights store subjective preferences and approximate the non-linear multiplication of reward magnitude and probability to yield expected values. This feedforward mechanism enables generalization to novel choice options. Second, option values are compared within the recurrent network, where specific connectivity patterns mediate robust winner-take-all decisions, with both excitatory and inhibitory neurons exhibiting value and choice selectivity. By training a single network on multiple tasks, we show compositional representations combining a shared computational schema with specialized neural modules. Reproducing key neurophysiological findings from the primate orbitofrontal cortex, our model unifies value computation, comparison, and generalization into a coherent framework with testable predictions.

Synaptic responses adapt on millisecond-to-second timescales through short-term plasticity (STP), a key process that filters and transforms neuronal information. While STP is classically ascribed to presynaptic release mechanisms, postsynaptic receptor properties-particularly desensitization and surface diffusion-also shape synaptic responses. Here, we dissect pre- and postsynaptic contributions to synaptic adaptation using molecular tools to visualize glutamate release and manipulate AMPA receptor (AMPAR) diffusion in intact circuits. We find that synaptic gain during STP is tuned by synapse-specific regulation of AMPAR biophysics and diffusion-trapping. These features are determined constitutively by auxiliary subunit profiles and dynamically by activity-dependent signaling engaged during long-term plasticity. With modeling, we quantified how short-term synaptic dynamics are impacted by postsynaptic regulation of filtering properties, which broadened heterogeneity of filtering timescales to refine temporal selectivity in synaptic networks. By augmenting desensitization-mediated synaptic depression, AMPAR diffusion-trapping emerges as a fundamental regulatory mechanism of postsynaptic integration and circuit-level information processing.

A hexanucleotide repeat expansion in C9orf72 is the most common genetic cause of amyotrophic lateral sclerosis and frontotemporal dementia. While repeat RNAs are implicated in disease pathogenesis, their mechanisms of action remain incompletely understood. Here, we show that GGGGCC repeat RNA engages chromatin genome-wide preferentially at promoter regions in patient cells. This interaction obstructs RNA polymerase II and transcription factors with GC-rich motifs, leading to broad transcriptional repression. Biochemical assays, single-molecule imaging, and native bisulfite sequencing analyses demonstrate that GGGGCC repeat RNA intrinsically forms DNA:RNA hybrid G-quadruplexes (HQs) with cognate DNA, providing a structural basis for transcriptional interference. Stabilization of these G-quadruplex structures exacerbates neuronal vulnerability to metabolic stress in patient-derived motor neurons and cortical organoids, whereas restoring key gene dysregulation improves resistance. These findings uncover a previously unrecognized trans-acting mechanism whereby repetitive RNAs form hybrid structures with genomic DNA, disrupt gene regulation, and contribute to neurodegeneration.