Lower Baseline Variability Gives Rise to Lower Detection Thresholds in Midbrain than Hindbrain Electrosensory Neurons

Electrosensory midbrain neurons optimally decode ascending input during object localization

Understanding how downstream brain areas decode sensory information represented by neural populations remains a central problem in neuroscience. While decoders that are optimized to extract the maximum amount of information have been extensively used in research, whether these are physiologically realistic remains at best unclear. Here we show that a physiologically realistic decoding scheme based on correlations between neural activities in the absence of stimulation can predict downstream neural responses as well as the optimal decoder. Simultaneous recordings were made from primary sensory neural populations and their downstream midbrain targets in the electrosensory system of Apteronotus leptorhynchus. We found that neural populations exhibited significant correlations in the absence of stimulation (i.e. 'baseline'), with downstream neural activity lagging primary sensory neural activity with a short latency. We then investigated how primary sensory neural activities were combined downstream. Overall, a decoder that assigned weights to each primary sensory neuron and was trained solely on baseline correlations performed as well as the optimal decoder trained on neural responses to stimulation. Interestingly, both decoders greatly outperformed schemes for which every neuron was assigned the same weight or when the weights were shuffled, indicating that neural identity is critical. Taken together, our results suggest that the brain uses decoding strategies that perform at optimal levels but are qualitatively different from those predicted from optimal solutions. KEY POINTS: How neural signals are decoded to give rise to perception remains poorly understood. We recorded from primary sensory neural populations and their downstream targets. A physiologically realistic decoder performed as well as the optimal solution to predict downstream responses. We found important qualitative differences between how information is decoded and the optimal solution. Our results demonstrate that the brain can do as well as an optimal decoder but uses a different strategy.

Descending pathways increase sensory neural response heterogeneity to facilitate decoding and behavior

The functional role of heterogeneous spiking responses of otherwise similarly tuned neurons to stimulation, which has been observed ubiquitously, remains unclear to date. Here, we demonstrate that such response heterogeneity serves a beneficial function that is used by downstream brain areas to generate behavioral responses that follows the detailed timecourse of the stimulus. Multi-unit recordings from sensory pyramidal cells within the electrosensory system of Apteronotus leptorhynchus were performed and revealed highly heterogeneous responses that were similar for all cell types. By comparing the coding properties of a given neural population before and after inactivation of descending pathways, we found that heterogeneities were beneficial as decoding was then more robust to the addition of noise. Taken together, our results not only reveal that descending pathways actively promote response heterogeneity within a given cell type, but also uncover a beneficial function for such heterogeneity that is used by the brain to generate behavior.

Serotonin increases population coding of behaviorally relevant stimuli by enhancing responses of ON but not OFF-type sensory neurons

How neural populations encode sensory input to generate behavioral responses remains a central problem in systems neuroscience. Here we investigated how neuromodulation influences population coding of behaviorally relevant stimuli to give rise to behavior in the electrosensory system of the weakly electric fish Apteronotus leptorhynchus. We performed multi-unit recordings from ON and OFF sensory pyramidal cells in response to stimuli whose amplitude (i.e., envelope) varied in time, before and after electrical stimulation of the raphe nuclei. Overall, raphe stimulation increased population coding by ON- but not by OFF-type cells, despite both cell types showing similar sensitivities to the stimulus at the single neuron level. Surprisingly, only changes in population coding by ON-type cells were correlated with changes in behavioral responses. Taken together, our results show that neuromodulation differentially affects ON vs. OFF-type cells in order to enhance perception of behaviorally relevant sensory input.