Low- and high-level coordination of orofacial motor actions

Abstract

Orofacial motor actions are movements that, in rodents, involve whisking of the vibrissa, deflection of the nose, licking and lapping with the tongue, and consumption through chewing. These actions, along with bobbing and turning of the head, coordinate to subserve exploration while not conflicting with life-supporting actions such as breathing and swallowing. Orofacial and head movements are comprised of two additive components: a rhythm that can be entrained by the breathing oscillator and a broadband component that directs the actuator to the region of interest. We focus on coordinating the rhythmic component of actions into a behavior. We hypothesize that the precise timing of each constituent action is continually adjusted through the merging of low-level oscillator input with sensory-derived, high-level rhythmic feedback. Supporting evidence is discussed.

Figure 1.. Relation of orofacial motor actions to breathing.

A. Vibrissa movement (blue) recorded with one-dimensional videography during whisking and non-whisking intervals. Lesioning the vibrissa oscillator (vIRt) in one hemisphere quenches rhythmic whisking yet does not interfere with changes in baseline (set-point) protraction (red). The baseline protraction is seen to track between the two sides of the face before the lesions and after, even though rhythmic whisking is lost on the lesioned side. Adapted from [13]. B. Whisking under two states in a head-fixed rat recorded with one-dimensional videography. The upper trace shows rapid, exploratory whisking (blue) that is phase-locked to sniffing (red), as recorded with a thermocouple implanted in the nasal cavity. The lower trace shows whisking during slow, basal breathing; note the 3 to 5 whisks per breath that are referred to as intervening whisks and the reset of whisking by the onset of inhalation. Adopted from [11]. C. Compilation of the timing of all rhythmic orofacial motor actions in rat with the behavior and underlying muscle activity plotted relative to phase in the breathing cycle. A phase of zero corresponds to the onset of inhalation. The data are segregated by behavioral state. Rearing corresponds to free-ranging animals that are elevated on their hind paws and breath with a mean rate of 8 Hz; the timing of motor actions appears to be the same for head-fixed animals as for rearing animals. Foraging refers to animals with their nose close to the ground that locomote on all four limbs and breath with a mean rate of 11 Hz. Abbreviation: NA, data not available. Adopted from [4] and [17].

Figure 2.. The circuit for the whisking oscillator and proposed extension for other orofacial motor actions.

A. Cartoon of the basic circuit derived from the experimental evidence [11, 12, 23]. The preBötzinger complex ($\text{pB}\ddot{\mathrm{o}}\text{tC}$) provides a rhythmic inhibitory input to entrain or reset the whisking oscillator (Figure 1B), formed by inhibitory neurons in the vIRt^pro and vIRt^ret clusters. The neurons in each cluster interact through synaptic connections with conductance g_intra, while those in different clusters interact with conductance ginter. All neurons receive an external excitatory input I_ext. Neurons in the vIRt^ret cluster project to motoneurons in the vibrissa region of the facial motor nucleus (vFMN). These cells receive an excitatory external input I^F_ext that protracts the vibrissae. Adopted from [24]. B. Schematic of the reduced vIRt oscillator for mean-field analysis. The individual synapses have been replaced by net currents, with J_intra as the current within each of the vIRt^pro and vIRt^ret clusters and J_inter as the current between the vIRt^pro and vIRt^ret clusters. Adopted from [24]. C. Results from the mean-field analysis, which yields equations for the average rate of spiking, whisking amplitude, and whisking frequency (green). We also show the results from numerical simulation of the conductance-based equations using detailed parameters for the network (light and dark gray). The left column highlights the three regions of operation as a function of the differential synaptic current J_inter - J_intra. The right column highlights the change in performance as a function of the input current. The simulations yield the same three regions of operation, but with a broader range for the oscillatory state that now permeates the bistable region of operation. Adopted from [24]. D. Compilation of the known and hypothesized circuitry for entrainment of all rhythmic orofacial motor actions and head movement to breathing during epochs of exploration and, for the case of the tongue and jaw, to a putative feeding oscillator during chewing. Whisking involves oscillatory drive from the vibrissa intermediate reticular zone (vIRt) to the vibrissa facial motoneurons (vFMN) that project to the intrinsic (Int) muscles in the mystacial pad that protract the follicles. During sniffing, facial muscles also drive the nasalis (NA), the extrinsic protractor muscle, and maxillolabialis (ML) and nasolabiallis (NL), the extrinsic retractor muscles, that move the pad in coordination with whisking. Nose twitching involves a yet undescribed oscillator, likely in the gigantocellular (Gi) region of the reticular formation, that projects to motoneurons in the facial motor nucleus that innervate the deflector nasi (DN). Head turning involves a yet undescribed oscillator, also likely in the Gi region of the reticular formation, that projects to motoneurons that form part of the spinal accessory nucleus (SAN) and drive the sternomastoid (SM) and clediomastoid (CM) muscles in the neck and projects to motoneurons in spinal level C1 to C5 ganglia that drive the clavotrapezius (CT), the splenius (SP), and the biverter cervicis (BC) muscles in the neck. Licking and chewing involve an oscillator under study in the tongue-jaw region of the parvocellular reticular formation (tjPCRt), with a potential accessory nucleus in the intermediate reticular zone (IRt), that presumably drives the hypoglossal motonucleus (hMN) and the trigeminal motonucleus (MotV). Trigeminal MotV drives the masseter (Mas), temporalis, and medial pterygoid jaw closing muscles and the lateral pterygoid LtPG) jaw opening muscle. The hypoglossal nucleus drives the genioglossus (GG), the protractor muscle, the hyoglossus and styloglossus (SG), the retractor muscles, and the palatoglossus, the elevation muscle. Extension of the summary in [17]. E. Anatomy of neural circuits involved in generating orofacial actions. The left panel is the three-dimensional reconstruction of the pons and medulla showing the pools of cranial motoneurons that control the jaws (orange), face (red), airway (yellow), and tongue (green). The middle and right panels are brainstem oscillators (marked as “~”) and their connections. Breathing-related regions are shown in black. Additional abbreviations: caudal/rostral ventral respiratory groups (cVRG and rVRG, respectively), parafacial respiratory group (pFRG), and lateral paragigantocellular reticular formation (LPGi).

Figure 3.. Hypothetical scheme for concurrent low-level and high-level control of the timing of orofacial motor actions.

A. Schematic. "Low-level" neuronal computation corresponds to orofacial motor oscillators, shown here as premotor nuclei, that are entrained by output by the $\text{pB}\ddot{\mathrm{o}}\text{tC}$ and drive a particular group of muscles. Peripheral reafference and/or proprioception lead/s to inputs to midbrain and forebrain and "high-level" computations that feedback to the premotor oscillators and potentially the motor nuclei. The high-level computations adjust the detailed timing, or equivalently the phase, of the individual motor actions that comprise a behavior. Head turning involves a yet undescribed oscillator, likely in the gigantocellular (Gi) region of the reticular formation, that projects to motoneurons that form part of the spinal accessory nucleus (SAN) and drive the sternomastoid (SM) and clediomastoid muscles in the neck and projects to motoneurons in spinal level C1 to C5 ganglia that drive the clavotrapezius, the splenius, and the biverter cervicis muscles in the neck. Whisking involves oscillatory drive from the vibrissa intermediate reticular zone (vIRt) to the facial motoneurons that project to the intrinsic (Int) muscles in the mystacial pad that protract the follicles. During sniffing, facial muscles also drive the nasalis, the extrinsic protractor muscle, and maxillolabialis and nasolabiallis, the extrinsic retractor muscles, that move the pad in coordination with whisking. Licking and chewing involve an oscillator under study in the parvocellular reticular formation (tjPCRt) that presumably drives the hypoglossal motonucleus and the trigeminal motonucleus (MotV). The hypoglossal nucleus drives the genioglossus (GG), the protractor muscle, the hyoglossus and styloglossus, the retractor muscles, and the palatoglossus, the elevation muscle. Trigeminal MotV drives the masseter (Mas), temporalis, and medial pterygoid jaw closing muscles and the lateral pterygoid jaw opening muscle. B. Schematic of a minimal phase-coupled oscillator model to demonstrate how combined low- and high-level inputs can tune the phase of a motor action. ${\Psi }_{\text{pB}\ddot{\mathrm{o}}\text{tC}}(\mathrm{t})$ and ${\Psi }_{\text{OMO}}(\mathrm{t})$ are the phases of the $\text{pB}\ddot{\mathrm{o}}\text{tC}$ output and the orofacial motor oscillator, respectively, $\Gamma$ is a coupling constant and the high-level feedback is presumed to alter a delay-time, $\tau$, that is separately optimized for each action. The experimental observation of both inhibitory as well as excitatory projections from the $\text{pB}\ddot{\mathrm{o}}\text{tC}$ [22] accounts for the choice of sign of the $\text{pB}\ddot{\mathrm{o}}\text{tC}$ to orofacial oscillator connection. C. Plot of the calculated phase difference, $\Delta \varphi$, between the output of the $\text{pB}\ddot{\mathrm{o}}\text{tC}$ and that of the orofacial oscillator as a function of delay-time, coupling constant, and sign of the $\text{pB}\ddot{\mathrm{o}}\text{tC}$ to orofacial oscillator connection. We used ${\mathrm{f}}_{\text{pB}\ddot{\mathrm{o}}\text{tC}}=11\text{Hz}$, ${\mathrm{f}}_{\text{OMO}}=8\text{Hz}$, and note that ${\Gamma }_0\equiv 2\pi ({\mathrm{f}}_{\text{pB}\ddot{\mathrm{o}}\text{tC}}-{\mathrm{f}}_{\text{OMO}})$.

Figure 4.. Oscillations in neuronal activity in forebrain and midbrains structures that are linked to breathing.

A. The top two traces show an example of a concurrent measurement of breathing, with a thermistor embedded in the nasal cavity, and the local field potential in vibrissa primary sensory (vS1) cortex. The bottom trace shows the cross-correlation, which has essentially no time-lag. Adopted from [43]. B. Recording of vibrissa position using one-dimensional videography concurrent with single unit recording in vibrissa primary motor (vM1) cortex. The blue lines indicate an epoch of spikes that are tightly timed to whisking. Note that breathing was not recorded in these animals, but the animal whisked in a manner that was subsequently associated with breathing [11, 18]. Adopted from [46]. C. Schematic that illustrates the transformation of the measured vibrissa motion, in terms of angle relative to the midline, into phase coordinates. A Hilbert transform converts the angle, denoted $\theta (\mathrm{t})$, into set-point $({\theta }_{\text{set}})$, amplitude $({\theta }_{\text{amp}})$, and phase $(\varphi )$ components defined by $\theta (\mathrm{t})={\theta }_{\text{set}}(\mathrm{t})\text{-}{\theta }_{\text{amp}}(\mathrm{t})[1+\cos \varphi (\mathrm{t})]$. Adopted and modified from [46]. D. The result of a Hilbert decomposition for the coding of vibrissa position by three different units in vM1 cortex. Units 1 and 3 show a dependence on phase in the whisk cycle in addition to the more slowly varying set-point and amplitude. Adopted from [46]. E. Sketch of a rodent licking. Original art by Julia Kuhl. F. Recordings from three different cortical areas related to movement of the tongue, including primary sensory (tjS1), primary motor (tjM1), and anterior lateral motor (ALM) cortices. Animal performed a sequential licking task across an array of spouts. Note that breathing was not recorded in these animals, but the animal licked in a manner that is associated with breathing [11, 18]. Shown are raster plots for three units per area with time-zero aligned to a lick in the middle of a sequence. The top three traces show responses for right-to-left licking across the array of spouts and the bottom three traces show responses for left-to-right licking. Adopted from [36]. G. Activity in the major output nucleus of the basal ganglia, the substantia nigra pars reticularis (SNr), is time-locked to the lick cycle in mice. The left trace is the rate histogram of spikes that occurred within a bout of licking. The oscillations in the neuronal activity are tightly coupled and positively correlated with the oscillations of the lick cycle; while breathing was not recorded in these animals, the animal licked in a manner that is associated with breathing [11, 18]. Dashed line indicates the time of spout contact. The right trace is the spike density for a population of neurons. Each row corresponds to the activity of one cell. Adopted from [50]. H. Activity in the superior colliculus (SC) is time-locked to the lick cycle in mice. All conditions the same as in panel G, except that here the oscillations in the neuronal activity are negatively correlated with the oscillations of the lick cycle. Adopted from [50]. I. The spike triggered average, i.e., the correlation and standard error of the mean, of neural activity in the SNr and SC at the start of each lick show that neurons in the SNr and SC exhibit antiphase oscillations during licking. Adopted from [50].