Phillip Bokiniec

Thermal sensation is a fundamental sense initiated by the activity of primary afferent thermoreceptors. While considerable attention has been paid to the encoding of noxious temperatures by thermoreceptors, it is far less clear how they encode innocuous cool and warm which are more commonly encountered in the environment. To address this, we sampled the entire thermoreceptor population using in vivo two-photon calcium imaging in the lumbar dorsal root ganglia of awake and anesthetized mice. We found that the vast majority of thermoreceptors respond bidirectionally, with an enhanced response to cool and a suppressed response to warm. Using in vivo pharmacology and computational modelling, we demonstrate that conductance changes in the cool-sensitive TRPM8 channel are sufficient to explain this bidirectional response type. Our comprehensive dataset reveals the fundamental principles of the peripheral encoding of innocuous temperatures and suggests that the same population of thermoreceptors underlie the distinct sensations of cool and warm.

The first phase of feeding consists in acquiring solid foods from the environment by biting, and their preparation for swallowing by chewing. These actions require the precise coordination of tens of orofacial muscles for the jaw and tongue. The siege for this motor patterning is known to be in the reticular formation, a complex and poorly mapped region of the hindbrain, but the neuron groups involved are still elusive. Here, we characterize a group of reticular interneurons located in the supratrigeminal area that express the homeodomain transcription factor Phox2b . This nucleus — Sup5 ^Phox2b — is premotor to both jaw-closing and jaw-opener motoneurons and receives direct input from cranial sensory afferents, motor cortex and satiation related nuclei. Its activity differentially tracks lapping, biting and chewing movements, suggesting its involvement in the elaboration of distinct orofacial motor patterns in vivo. Acute global activation or inhibition of Sup5 ^Phox2b by optogenetics both interrupt volitional feeding sequences. Thus, Sup5 ^Phox2b is an obligatory subcortical node, topologically and genetically defined, in the neural circuits that control the oral phase of feeding. Teaser: A genetically defined cluster of neurons in the hindbrain is an essential relay for biting and chewing food.

Although distinct thalamic nuclei encode sensory information for almost all sensory modalities, the existence of a thalamic representation of temperature with a role in thermal perception remains unclear. To address this, we performed high-density electrophysiological recordings across the entire forelimb somatosensory thalamus in awake mice, and identified an anterior and a posterior representation of temperature that spans three thalamic nuclei. We found that these parallel representations show fundamental differences in the cellular encoding of temperature which reflects their cortical output targets. While the anterior representation encodes cool only and the posterior both cool and warm; in both representations cool was more densely represented and showed shorter latency, more transient responses as compared to warm. Moreover, thalamic inactivation showed a major role in thermal perception. Our comprehensive dataset identifies the thalamus as a key structure in thermal processing and highlights a novel posterior pathway in the thalamic representation of warm and cool.

In the thermal system, skin cooling is represented in the primary somatosensory cortex (S1) and the posterior insular cortex (pIC). Whether S1 and pIC are nodes in anatomically separate or overlapping thermal sensorimotor pathways is unclear, as the brain-wide connectivity of the thermal system has not been mapped. We address this using functionally targeted, dual injections of anterograde viruses or retrograde tracers into the forelimb representation of S1 (fS1) and pIC (fpIC). Our data show that inputs to fS1 and fpIC originate from separate neuronal populations, supporting the existence of parallel input pathways. Outputs from fS1 and fpIC are more widespread than their inputs, sharing a number of cortical and subcortical targets. While, axonal projections were separable, they were more overlapping than the clusters of input cells. In both fS1 and fpIC circuits, there was a high degree of reciprocal connectivity with thalamic and cortical regions, but unidirectional output to the midbrain and hindbrain. Notably, fpIC showed connectivity with regions associated with thermal processing. Together, these data indicate that cutaneous thermal information is routed to the cortex via parallel circuits and is forwarded to overlapping downstream regions for the binding of somatosensory percepts and integration with ongoing behavior.

In the thermal system, skin cooling is represented in the primary somatosensory cortex (S1) and the posterior insular cortex (pIC). Whether S1 and pIC are nodes in anatomically separated or overlapping thermal sensorimotor pathways is unclear, as the brain-wide connectivity of the thermal system has not been mapped. We address this using functionally targeted, dual injections of anterograde viruses or retrograde tracers into S1 and pIC. Our data show that inputs to S1 and pIC originate from two non-overlapping populations, supporting the existence of parallel input pathways. While outputs from S1 and pIC were more widespread and share a number of cortical and subcortical regions, within target structures axonal projections were often separable. We observed a high degree of reciprocal connectivity with thalamic and cortical regions from both S1 and pIC, but output to the mid- and hind-brain was unidirectional. Notably, pIC showed exclusive connectivity with regions associated with thermal processing. Together, these data indicate that cutaneous thermal information is routed to the cortex via multiple, parallel streams of information which are forwarded to overlapping downstream regions for the binding of complex somatosensory percepts and integration with ongoing behavior.

It has long been known that orofacial movements for feeding can be triggered, coordinated, and often rhythmically organized at the level of the brainstem, without input from higher centers. We uncover two nuclei that can organize the movements for ingesting fluids in mice. These neuronal groups, IRt^Phox2b and Peri5^Atoh1, are marked by expression of the pan-autonomic homeobox gene Phox2b and are located, respectively, in the intermediate reticular formation of the medulla and around the motor nucleus of the trigeminal nerve. They are premotor to all jaw-opening and tongue muscles. Stimulation of either, in awake animals, opens the jaw, while IRt^Phox2b alone also protracts the tongue. Moreover, stationary stimulation of IRt^Phox2b entrains a rhythmic alternation of tongue protraction and retraction, synchronized with jaw opening and closing, that mimics lapping. Finally, fiber photometric recordings show that IRt^Phox2b is active during volitional lapping. Our study identifies one of the subcortical nuclei underpinning a stereotyped feeding behavior.

Key points: Spinally‐projecting neurons of the rostral ventrolateral medulla (RVLM) determine sympathetic outflow to different territories of the body. Previous studies suggest the existence of RVLM neurons with distinct functional classes, such as neurons that target sympathetic nerves bound for functionally‐similar tissue types (e.g. muscle vasculature). The existence of RVLM neurons with more general actions had not been critically tested. Using viral tracing, we show that a significant minority of RVLM neurons send axon collaterals to disparate spinal segments (T₂ and T₁₀). Furthermore, optogenetic activation of sympathetic premotor neurons projecting to lumbar spinal segments also produced activation of sympathetic nerves from rostral spinal segments that innervate functionally diverse tissues (heart and forelimb muscle). These findings suggest the existence of individual RVLM neurons for which the axons branch to drive sympathetic preganglionic neurons of more than one functional class and may be able to produce global changes in sympathetic activity. We investigate the extent of spinal axon collateralization of rat rostral ventrolateral medulla (RVLM) sympathetic premotor neurons and its functional consequences. In anatomical tracing experiments, two recombinant herpes viral vectors with retrograde tropism and expressing different fluorophores were injected into the intermediolateral column at upper thoracic and lower thoracic levels. Histological analysis revealed that ∼21% of RVLM bulbospinal neurons were retrogradely labelled by both vectors, indicating substantial axonal collateralization to disparate spinal segments. In functional experiments, another virus with retrograde tropism, a canine adenovirus expressing Cre recombinase, was injected into the left intermediolateral horn around the thoracolumbar junction, whereas a Cre‐dependent viral vector encoding Channelrhodopsin2 under LoxP control was injected into the ipsilateral RVLM. In subsequent terminal experiments, blue laser light (473 nm × 20 ms pulses at 10 mW) was used to activate RVLM neurons that had been transduced by both vectors. Stimulus‐locked activation, at appropriate latencies, was recorded in the following pairs of sympathetic nerves: forelimb and hindlimb muscle sympathetic fibres, as well as cardiac and either hindlimb muscle or lumbar sympathetic nerves. The latter result demonstrates that axon collaterals of lumbar‐projecting RVLM neurons project to, and excite, both functionally similar (forelimb and hindlimb muscle) and functionally dissimilar (lumbar and cardiac) preganglionic neurons. Taken together, these findings show that the axons of a significant proportion of RVLM neurons collateralise widely within the spinal cord, and that they may excite preganglionic neurons of more than one functional class.

The anatomical and functional characterization of somatostatin (SST) and somatostatin receptors (SSTRs) within the spinal cord have been focused in the dorsal horn, specifically in relation to sensory afferent processing. However, SST is also present within the intermediolateral cell column (IML), which contains sympathetic preganglionic neurons (SPN). We investigated the distribution of SSTR2 within the thoracic spinal cord and show that SSTR2A and SSTR2B are expressed in the dorsal horn and on SPN and non-SPN in or near the IML. The effects of activating spinal SSTR and SSTR2 were sympathoinhibition, hypotension, bradycardia, as well as decreases in interscapular brown adipose tissue temperature and expired CO₂, in keeping with the well-described inhibitory effects of activating SSTR receptors. These data indicate that spinal SST can decrease sympathetic, cardiovascular and thermogenic activities. Unexpectedly blockade of SSTR2 revealed that SST tonically mantains sympathetic, cardiovascular and thermogenic functions, as activity in all measured parameters increased. In addition, high doses of two antagonists evoked biphasic responses in sympathetic and cardiovascular outflows where the initial excitatory effects were followed by profound but transient falls in sympathetic nerve activity, heart rate and blood pressure. These latter effects, together with our findings that SSTR2A are expressed on GABAergic, presumed interneurons, are consistent with the idea that SST2R tonically influence a diffuse spinal GABAergic network that regulates the sympathetic cardiovascular outflow. As described here and elsewhere the source of tonically released spinal SST may be of intra- and/or supra-spinal origin.