Publications

Adaptation is a common neural phenomenon wherein sustained stimulation evokes fewer action potentials (spikes) over time. Rather than simply reduce firing rate, adaptation may help neurons form better (i.e. more discriminable) representations of sensory input. To study effects of adaptation in low-threshold mechanoreceptors (LTMRs), we recorded single unit LTMR responses to 30-second-long vibrotactile stimuli with different intensities and frequencies applied to the hind paw of rats. To assess the impact of adaptation on somatosensory encoding, decoders were applied to the initial and late (adapted) phases of population-level responses to assess the decodability (discrimination) of stimulus intensity and frequency. Adaptation-mediated changes in the rate and timing (phase-locking) of spikes were quantified. Rate coding of stimulus intensity was improved by the nonuniform reduction in firing rate across responses to different stimuli, and across neurons. This improvement was absent in simulations with uniform reductions in firing rate, thus revealing the necessity of stimulus-dependent variability in adaptation effects. Spike timing (quantified as interspike intervals) remained highly informative about stimulus frequency throughout stimulation despite the progressive reduction in spike count over time. When the drop in spike count was accounted for, adaptation was found to improve temporal coding of stimulus frequency by increasing the precision of phase-locking. In other words, adaptation improved the precision of spike timing, and this increased the information about stimulus frequency conveyed by each spike. These results show that adaptation, by modulating spiking in different ways, can improve encoding of different stimulus features using different coding schemes.

Touch is mistakenly perceived as painful when inhibition in the spinal dorsal horn (SDH) is weakened. Disinhibition un-gates a polysynaptic spinal circuit but why mechanical allodynia is predominantly evoked by certain stimuli, like dynamic brushing, remains unclear. To answer this, we incorporated receptive fields (RFs) into a computational model of the SDH to study the processing of stimuli with different spatiotemporal features. Broad stimuli normally suppress output spiking by engaging inhibition from the RF surround, but the efficacy of inhibition depends on the input’s temporal pattern. This is critical since excitatory and inhibitory spinal neurons are preferentially sensitive to synchronous and asynchronous input, respectively. Furthermore, spikes driven by synchronous input are resistant to feedforward inhibition. The combined results explain why broad dynamic touch (e.g. brush or vibration) evokes more allodynia than punctate static touch. Our results also show that asynchronous and spatially disordered input evoked by kilohertz-frequency spinal cord stimulation preferentially activates inhibitory neurons, thus explaining its anti-allodynic effects.

Our sense of touch starts with activation of nerve fibres in the skin. Although response properties of various fibre types are well-established in other species (e.g. primates), quantitative characterization in rats and mice is limited. To fill this gap, we performed a comprehensive electrophysiological investigation into the coding properties of tactile fibres in rodent non-hairy skin and then simulated these fibres to explain differences in their responses. We show that rodent tactile fibres resemble those from other species, but that their heterogeneity at the population level may differ, with potentially important implications for encoding of touch. Simulations reveal intrinsic mechanisms that support this heterogeneity and provide a useful tool to explore somatosensation in rodents.

To achieve smooth motor performance in a changing sensory environment, motor outputs must be constantly updated in response to sensory feedback. Inhibitory interneurons in the spinal cord play an essential role in shaping motor activity by gating the transmission of sensory information and setting the pattern and rhythm of motor neurons. Here, in mice, we identify the medial deep dorsal horn of the spinal cord as a “hot zone” of convergent proprioceptive and cutaneous information from the hindlimb, where inhibitory neurons show increased responsiveness to sensory input and are more prominently recruited during locomotion in comparison to excitatory neurons. We identify a novel population of glycinergic inhibitory neurons within the deep dorsal horn that express parvalbumin (dPV) and receive convergent proprioceptive and cutaneous input from the paw. We show that dPVs possess intrinsic properties that support spontaneous discharge, even in the absence of synaptic input. However, a drug cocktail mimicking descending input (5-HT, dopamine, NMDA) amplifies dPV output, while cutaneous and proprioceptive inputs shape the temporal dynamics of dPV activity. These findings suggest dPV-mediated inhibition is modulated by behavioral state and can be fine-tuned by sensory input. Using intersectional genetic strategies, we selectively target spinal cord dPVs and demonstrate their capacity to provide widespread ipsilateral inhibition to both pre-motor and motor networks of the ventral horn, thereby gating sensory-evoked muscle activity. Manipulating the activity of dPVs during treadmill locomotion results in altered limb kinematics at the transition of stance to swing and altered step cycle timing at increased speeds. To investigate the effects of manipulating dPV activity on broader sets of motor behaviors, we used depth vision and machine learning to quantify and scale spontaneous behavior. We find that although sub-movements remain stable, the transitions between sub-movements are reduced, suggesting a role in movement switching. In sum, our study reveals a new model by which sensory convergence and inhibitory divergence produce a surprisingly flexible influence on motor networks to increase the diversity of mechanisms by which sensory input facilitates smooth movement and context-appropriate transitions.

Electrically activating mechanoreceptive afferents inhibits pain. However, paresthesia evoked by spinal cord stimulation (SCS) at 40–60 Hz becomes uncomfortable at high pulse amplitudes, limiting SCS “dosage.” Kilohertz-frequency SCS produces analgesia without paresthesia and is thought, therefore, not to activate afferent axons. We show that paresthesia is absent not because axons do not spike but because they spike asynchronously. In a pain patient, selectively increasing SCS frequency abolished paresthesia and epidurally recorded evoked compound action potentials (ECAPs). Dependence of ECAP amplitude on SCS frequency was reproduced in pigs, rats, and computer simulations and is explained by overdrive desynchronization, spikes desychronize when axons are stimulated faster than their refractory period. Unlike synchronous spikes, asynchronous spikes fail to produce paresthesia because their transmission to somatosensory cortex is blocked by feedforward inhibition. Our results demonstrate how stimulation frequency impacts synchrony based on axon properties and how synchrony impacts sensation based on circuit properties.

After nerve injury, light touch can become painful. To help understand how this occurs, we built a computer model of the spinal cord circuitry responsible for integrating touch and pain signals.

We built a computer model of subfornical organ (SFO) neurons to better understand their unique firing patterns and role in energy homeostasis.

In my MSc thesis I built a computer model of subfornical organ (SFO) neurons to better understand their unique firing patterns and role in energy homeostasis.