Sensorimotor integration in the whisker somatosensory brain stem trigeminal loop
Abstract
The rodent’s vibrissal system is a useful model system for studying sensorimotor integration in perception. This integration determines the way in which sensory information is acquired by sensory organs and the motor commands that control them. The initial instance of sensorimotor integration in the whisker somatosensory system is implemented in the brain stem loop and may be essential to the way rodents explore and sense their environment. To examine the nature of these sensorimotor interactions, we recorded from lightly anesthetized rats in vivo and brain stem slices in vitro and isolated specific parts of this loop. We found that motor feedback to the vibrissal pad serves as a dynamic gain controller that controls the response of first-order sensory neurons by increasing and decreasing sensitivity to lower and higher tactile stimulus magnitudes, respectively. This delicate mechanism is mediated through tactile stimulus magnitude-dependent motor feedback. Conversely, tactile inputs affect the motor whisking output through influences on the rhythmic whisking circuitry, thus changing whisking kinetics. Similarly, tactile influences also modify the whisking amplitude through synaptic and intrinsic neuronal interaction in the facial nucleus, resulting in facilitation or suppression of whisking amplitude. These results point to the vast range of mechanisms underlying sensorimotor integration in the brain stem loop.
NEW & NOTEWORTHY Sensorimotor integration is a process in which sensory and motor information is combined to control the flow of sensory information, as well as to adjust the motor system output. We found in the rodent’s whisker somatosensory system mutual influences between tactile inputs and motor output, in which motor neurons control the flow of sensory information depending on their magnitude. Conversely, sensory information can control the magnitude and kinetics of whisker movement.
INTRODUCTION
Rodents actively explore their environment using their mystacial whiskers, which they move back and forth to scan the area surrounding their pads. These whisking movements serves a wide range of tasks (Diamond et al. 2008), including texture discrimination (Carvell and Simons 1990; von Heimendahl et al. 2007), distance and width judgment (Jenkinson and Glickstein 2000; Krupa et al. 2001), and object localization (Kleinfeld and Deschênes 2011; Knutsen et al. 2006; Mehta et al. 2007; O’Connor et al. 2010). To generate a coherent percept of the environment, sensory systems may require the involvement of a motor system that moves the receptor organs into the most effective position for probing the environment (Gibson 1962; Kleinfeld et al. 2006; Najemnik and Geisler 2005; Schroeder et al. 2010). In return, sensory signals affect the motor commands controlling whisker movement; e.g., rodents rapidly change whisking patterns upon active touch, presumably to enhance information flow and permit adaptive sensory sampling in both time and space (Crochet et al. 2011; Mitchinson et al. 2007). Understanding whisker tactile sensory perception requires the investigation of the interactions between the sensory and motor systems (Kleinfeld and Deschênes 2011), since it may rely on various sensorimotor loops. The sensorimotor vibrissal control system encompasses multiple closed loops that connect the sensory periphery to motor neurons through different levels in the nervous system (Ahissar and Kleinfeld 2003; Bosman et al. 2011; Deschênes et al. 2012; Diamond et al. 2008; Kleinfeld et al. 1999, 2006; Nguyen and Kleinfeld 2005).
In the rodent vibrissa somatosensory system, the vibrissa trigeminal loops in the brain stem are the first neuronal circuit in which sensorimotor integration occurs and may likely mediate reflexive or innate aspects of sensorimotor behaviors (Kis et al. 2004; Kleinfeld et al. 1999; Nguyen and Kleinfeld 2005). In these loops, multiple parallel sensory pathways provide whisker-related tactile sensory information to other brain stem circuits, cerebellar circuits, the superior colliculus, and the thalamocortical system (Bosman et al. 2011; Diamond et al. 2008; Petersen 2007). The input to the loop starts in the trigeminal ganglion (TG), which receives sensory signals from first-order neurons that innervate the vibrissae and the cutaneous skin on the face (Rice 1993, 1997), and terminate throughout the trigeminal nuclear complex (Arvidsson 1982; Aström 1953; Kerr 1963; Marfurt 1981; Olszewski 1950; Sakurai et al. 2013). The trigeminal complex contains both excitatory and inhibitory neurons (Avendaño et al. 2005; Furuta et al. 2008; Li et al. 1997) that project within and among the various brain stem nuclei (Bellavance et al. 2010; Jacquin et al. 1989a; Jacquin et al. 1989b). Several studies have examined the influence of anatomical and cellular properties on the functional significance of trigemino-facial sensorimotor loop; however, the results of these studies have not been consistent (Bellavance et al. 2017; Nguyen and Kleinfeld 2005).
In the current study, we used the well-studied brain stem loops (BL) to explore the mutual interactions between sensory inputs and motor output. We used anesthetized rats to examine the influences of mystacial pad motor output on sensory responses and the impact of tactile inputs on whisking behavior. Concomitantly, we recorded from the facial nucleus motor neurons to elucidate the neuronal mechanisms underlying these interactions.
MATERIALS AND METHODS
Recording and Stimulation
Surgical procedures.
Adult male Sprague-Dawley rats (250–350 g) were used. All experiments were conducted in accordance with international standards and were approved by the Institutional Animal Care and Use Committee. Surgical anesthesia was induced by urethane (1.5 g/kg ip) and maintained at a constant level by monitoring forepaw withdrawal and corneal reflex and administering extra doses (10% of original dose) as necessary. In some of the animals [motor cortex (M1) and superior colliculus (SC) stimulation protocols], after initial anesthesia with ketamine (100 mg/kg) and xylazine (10%), a tracheotomy was made following local subcutaneous injection of lidocaine. Rats were mounted in a stereotaxic device and respirated with a mixture of halothane (0.5–1.5%) and oxygen-enriched air.
Atropine methyl nitrate (AMN; 0.3 mg/kg im) was administered after general anesthesia to prevent respiratory complications (AMN may influence muscle tone). Therefore, in some of the experiments, we did not use AMN; no significant changes in the results were observed). Body temperature was maintained near 37°C using a servo-controlled heating blanket (Harvard, Holliston, MA). After placing the subjects in a stereotactic apparatus (TSE Systems, Bad Homburg, Germany), an opening was made in the skull overlying the TG, and tungsten microelectrodes (2 MΩ; NanoBio Sensors, Israel) were lowered, according to known stereotaxic coordinates of the TG [1.5–3 mm mediolateral, 0.5–2.5 mm anteroposterior to bregma (Shoykhet et al. 2000)]. Once a neuron was encountered, its receptive field was mapped. The recorded signals were amplified (×1000), band-pass filtered (1 Hz–10 kHz), digitized (25 kHz) and stored for off-line spike sorting and analysis. The data were then separated into local field potentials (LFP; 1–150 Hz) and isolated single-unit activity (0.5–10 kHz). All neurons could be driven by manual stimulation of one of the whiskers. Spike extraction and sorting implemented MClust (by A.D. Redish available from http://redishlab.neuroscience.umn.edu/MClust/MClust.html), which is MATLAB-based (MathWorks, Natick, MA) spike-sorting software. The extracted and sorted spikes were stored at a 0.1-ms resolution, and raster plots and peri-stimulus time histograms (PSTHs) were computed.
Vibrissa stimulation.
Receptive fields were initially determined by manually deflecting individual vibrissa. Vibrissa- evoking detectable responses were then individually attached to a computer-controlled galvanometer stimulator. In some of the experiments (EMG responses, SC and M1 stimulation), we connected the galvanometer to most of the vibrissae through a metal mesh (size of 1.5 × 1.5 cm; ~0.5-mm holes). The stimulus was frozen Gaussian filtered pseudo-random white noise (the size of the Gaussian window that was chosen resulted in a low-pass filter of ~250 Hz). Each stimulus was presented for 1 s and repeated 25–50 times in the preferred orientation. The stimuli were played at six different variances per session (between 20 and 400 µm, starting from the lowest variance to which the neuron responded). A period of 2 s separated each stimulus. We made sure that the small variances would produce a low but consistent firing rate, and the high variances would not cause saturation of the neuronal response. The common range was 50–250 μm and reached 100–350 μm and 1–10 μm in some cases. In the SC stimulation experiments, the sensory stimulus consisted of two variances, low (100 μm) and high (400 μm), with 30 repetitions each. In the M1 stimulation experiments, the sensory stimulus was similar to that of the SC experimental paradigm. Stimuli were delivered 3 mm from the mystacial pad. Accordingly, the angle of the whisker was given by θ = tan−1(x/3), where x is whisker displacement in millimeters.
Artificial motor output (Brown and Waite 1974; Szwed et al. 2003) was induced by stimulating the buccolabialis motor branch of the facial nerve (BL) (Semba and Egger 1986). The distal end of the nerve was cut and mounted on bipolar tungsten electrodes and was kept moist. Bipolar rectangle electrical pulses (duration of 100 µs at 143 Hz for the duration of the tactile stimuli) were applied through an isolated pulse stimulator (ISO-Flex; A.M.P.I., Israel). The stimulation magnitude was adjusted at the beginning of each recording session to the minimal value that reliably generated EMG signals without noticeable vibrissa movements (10–50 µA). Vibrissa movements were monitored using a noncontact optical displacement measuring system (resolution: 1 µm; LD1605-2; Micro-Epsilon, Ortenburg, Germany). A second type of artificial vibrissa movement was induced using SC and M1 stimulation. In these experiments, bipolar microelectrodes (0.1 MΩ, NanoBio Sensors, Israel) were inserted, according to the coordinates of the SC [0.5–2.5 mm lateral, −5.5 to −7.5 mm from bregma, depth 3–4 mm (Hemelt and Keller 2008)] and M1 [0.5–1.25 mm lateral, −0.5 to −2.0 mm from bregma, depth of 1.5 mm (Haiss and Schwarz 2005)]. Stimulation in the SC consisted of trains of stimuli, 200-μs pulse width, at 50–1,000 Hz delivered for 250–500 ms. The rhythmic subregion of M1 was identified by applying low-intensity current pulses (50 monophasic pulses at 50 Hz, 200-μs pulse width, 25 to 250 μA) to the area of M1 identified by Haiss and Schwarz (Haiss and Schwarz 2005).
To record the vibrissal electromyogram (EMGs), a pair of bipolar EMG electrodes (76 μm Teflon-coated stainless-steel wire) were tunneled subcutaneously into the deep intrinsic muscles through a small incision in the face, as previously described (Carvell et al. 1991; Fee et al. 1997). Microwires were also placed subcutaneously in the fibers of the extrinsic musculature (Dörfl 1982; Wineski 1985). EMG recordings were sampled at 25 kHz and filtered (0.1 Hz to 10 kHz).
Brain Stem Slice Preparation
The transversal slice containing the facial nucleus (FN) was obtained from 8- to 11-day old Sprague-Dawley rats, as previously described (Nguyen et al. 2004; Takahashi 1990). The facial motor neurons that innervate the mystacial pad were labeled by injection of 10–15 μl of 2% (weight/volume) Evans Blue dissolved in PBS in each mystacial pad. Two to three days after Evans Blue injection, the rats were anesthetized by 2% isoflurane inhalation and decapitated. The trigeminal nerves were dissected, and the brain was immersed in an ice-cold modified artificial cerebrospinal fluid (m-ACSF) solution (in mM: 252 sucrose, 5 KCl, 2 MgSO4, 26 NaHCO3, 1.25 NaH2PO4, 2 CaCl2+, and 10 mM d-glucose). The forebrain and cerebellum were removed, and the brain stem was glued dorsal side down on a cutting dish, which was placed on a Vibratome (VT1200; Leica). The first cut (~400 mm) exposed facial nucleus, and the blade was lowered by 500 μm. The separated slice was placed in a holding chamber containing oxygenated ACSF (in mM: 124 NaCl, 3 KCl, 2 CaCl2+, 2 MgSO4, 1.25 NaH2PO4, 26 NaHCO3, and 10 glucose; pH 7.4 when bubbled with 95% O2/CO2) at 32°C for 1 h.
The presence of vibrissa motoneurons in the slice was confirmed by retrograde transport of the fluorescent dye Evans Blue. The FN neurons were viewed with a ×40 water-immersion lens (Olympus) in a BX51WI microscope (Olympus). The labeled neurons were imaged by a charge-coupled device camera CoolSnap HQ2 (Photometrics) by using a 568-nm excitation wavelength and a 585-nm long-pass emission filter. Consistent with past work (Cajal 1955), no labeled trigeminal processes could be seen entering the facial nucleus.
Somatic whole cell recordings were made by using patch pipettes pulled from thick-walled borosilicate glass capillaries (1.5 mm outer diameter; Hilgenberg). The pipettes had resistances of 7 to 10 MΩ when filled with an intracellular solution (130 mM K-gluconate, 6 mM KCl, 2 mM MgCl2+, 4 mM NaCl, and 10 mM HEPES, adjusted to pH 7.25 with KOH) with a supplement of 3 mM QX314 (Alomone Laboratories, Israel). Current-clamp recordings were made by using an Axoclamp-2B amplifier (Axon Instruments). All recordings were conducted at physiological temperature (30 ± 1°C). Electrophysiological data analysis was performed using pCLAMP 10.4 (Axon Instruments), Origin 6.0 (Origin Laboratory). The junction potentials (−14 mV) between the pipette and the bath solution were corrected. Bipolar stimulating electrodes were placed on the IoN nerve to activate FN cells and were defined as the sensory activating input electrodes. Additional bipolar electrodes were placed near the FN to stimulate motor inputs. The intensity of the stimulus was adjusted to minimal responses. postsynaptic potentials (PSP) amplitudes were measured by averaging 7–15 trials. PSP latency was calculated by defining the time point when the amplitude of the PSP reached 5% of the peak (Markram et al. 1997). Data are given as means ± SE, unless otherwise stated. Pharmacological compounds included 1(S), 9(R)-(-)-bicucculine (Sigma, St. Louis, MO) and 3-[2′-phosphonomethyl[1,1′- biphenyl]-3-yl]alanine (PMBA) (Santa Cruz Biotechnology, Dallas, TX) and were perfused into the slice chamber.
Data Analysis
Electromyographic (EMG) analysis was performed using a custom-written program (Cramer et al. 2007); MATLAB). In brief, data were filtered (1 Hz to 1 kHz differential filter), and EMG envelopes were generated by smoothing the data, subsampled to 500 Hz, using a sliding box algorithm with a width of 6 points. For data sampled at 25 kHz, the smoothing algorithm had a corner frequency of 166 Hz. A width of 131 points was used for better visualization of Fig. 1B. We defined EMG onset as an increase in activity that significantly (99.5% confidence interval) exceeded the amplitude of baseline activity (Fig. 1B, arrow).
Fig. 1.Positive feedback in the brain stem loop. A: schematic diagram of the experimental paradigm. Modified from Nguyen and Kleinfeld (2005) B: recording EMG signals from the extrinsic (1–3) and intrinsic (4–6) muscles in response to whiskers stimuli (bottom). 1 and 4 denote the rectified and smoothed EMG signal, respectively. The threshold was set at 3 SDs (see materials and methods section); 2 and 5 denote raw EMG signal in response to whisker stimulus (bottom); 3 and 6 denote raw EMG signal in response to whisker stimulus, after nerve cut. The gray area indicates the 3 SDs of the control signal (without tactile stimulus). C: distribution of EMG response latency in all sessions (n = 5; 15.2 ± 2.6). D: simplified description of the experimental procedure in which a galvanometer stimulated most of the vibrissae through a mesh, whereas a single vibrissa (red) was left untouched, and its movement was monitored. E: influence of stimulus SD on EMG response latency. F: influence of stimulus SD on EMG response magnitude. G: tactile stimulus does not induce vibrissa motion. The gray area indicates the 3 SDs of the control signal. H: tactile stimulus does not induce significant vibrissa motion in five rats (control, no vibrissae stimuli). BL, buccolabialis branch of facial nerve; FN, facial nucleus; IoN, infraorbital branch of trigeminal nerve; PrV, principal trigeminal nucleus; RF, reticular formation; SpVc, spinal trigeminal nucleus pars caudalis; SpVi, spinal trigeminal nucleus interpolaris; TG, trigeminal ganglion; TN, trigeminal nuclear complex.
To calculate the influence of tactile stimuli on whisking, we calculated the area under the whisking signal (whisker movement). We initially applied a high-pass filter at 1 Hz and rectified the whisking signal. We then calculated the integral of the signals and compared the conditions.
To estimate the response variability of firing rates across trials, we calculated the response variance:
To examine the synaptic interactions between trigeminal and motor inputs, we calculated the ratio of the observed response to expected response. The expected response was calculated by obtaining the arithmetic sum of the two PSP amplitudes. To take into consideration the influence of membrane depolarization, we recorded PSP generated by motor inputs at different membrane voltages.
The significance of the differences between the parameters was evaluated using a one way ANOVA. When significant differences were indicated in the F ratio test (P < 0.05), Tukey’s method for multiple comparisons was used to determine those pairs of parameters that differed significantly within the pair (P < 0.05 or P < 0.01). Averaged data are expressed as means ± SE, unless stated otherwise. Error bars in all the figures indicate the standard error unless stated otherwise.
RESULTS
To ascertain the nature of sensorimotor interaction in the BL we used several approaches to stimulate and record from the vibrissae trigeminal loop (Fig. 1A). We initially repeated experiments done by Nguyen et. al. (Nguyen and Kleinfeld 2005) to examine the functional significance of sensorimotor feedback. We measured EMG responses during stimulation by passive deflection of the vibrissae. We induced vibrissae motion by stimulating with a galvanometer connected to a metal mesh to most of the vibrissae (Fig. 1D), which produces complex waveform stimuli; each stimulus was delivered for 500 ms at 5-s intervals and repeated 50 times (n = 5). EMG responses with variable amplitudes followed these mechanical stimuli in the extrinsic (Fig. 1B, 1–3), and intrinsic muscles (Fig. 1B, 4–6). The raw EMG signal, in response to whisker stimulation (Fig. 1B, bottom) is shown in panels 2 and 5. To quantify these responses, we rectified and smoothed the signal (panels 1 and 4; see materials and methods). To eliminate the possibility of contamination of EMG by tactile stimulus, we repeated the same procedure after severing the nerve (Fig. 1B, 3 and 6). To examine whether the significance of this contamination, we measured the spontaneous EMG activity for 30 s before stimulation and calculated the range of signal (means ± 3 SD, gray area). The vibrissae stimuli did not cause any significant EMG activity after nerve cut.
We initially quantified response latency by finding the first point that crossed the processed EMG response threshold (3 SDs). As has been shown, the EMG responses to passive vibrissa deflection were reduced almost immediately after the initial response (Fig. 1B, 1 and 4, initial surge and decline). The distribution of latencies for 50-µm stimulus variance after stimulus onset is shown in Fig. 1C. The average latency for this passive stimulation was 15.2 ± 2.6 ms. To examine the influence of tactile stimulus magnitude on muscular responses, we varied stimulus variance. Figure 1E shows that increasing the stimulus magnitudes reduced the response latency. More importantly, increases in tactile stimulus magnitude resulted in an elevation of EMG activity, as measured by the area under the curve of the EMG smoothed signal (Fig. 1F). Together, these results suggest that tactile inputs may cause further vibrissae movement. To examine this supposition, we applied mechanical stimuli through a metal mesh to most of the vibrissae, while monitoring the movement of a single vibrissa (Fig. 1D). An example of this type of experiment is depicted in Fig. 1G. We measured the spontaneous movements of the vibrissae for 2 min before stimulation (Control) and calculated the range of movements (means ± 3 SD, gray area). The vibrissae stimuli did not cause any significant movement of the nonattached vibrissa. Examining all rats (n = 5) revealed that tactile inputs did not induce significant further motion in the vibrissae. Thus, tactile stimuli-induced motor outputs to the pad do not cause further movement in the vibrissae (Fig. 1H).
Influence of Motor Output on Tactile Inputs
To examine the influence of motor output on sensory transformation, we took two approaches. In the first line of experiments, we disconnected the sensorimotor loop by severing the facial motor nerve (7th nerve; buccolabialis motor branch). We recorded from trigeminal ganglion neurons (n = 39; 29 rats) while stimulating the principal vibrissa with frozen Gaussian-filtered white noise stimuli at several variances using a galvanometer motor stimulator. We then cut the facial motor nerve to eliminate motor feedback and repeated the stimulus paradigm. To determine whether any changes in tactile response after nerve cutting was a result of pad movement afterward, we monitored adjacent vibrissa. We found only small and nonsignificant changes in vibrissa shifts in relation to baseline vibrissa position (Fig. 2C). Figure 2, B and D present an analysis of changes in TG neuronal firing rates before and after nerve cutting. This neuron exhibits a significant reduction of ~50% in firing rate after the nerve was cut, at different stimulus SDs. It should be noted that this neuron does not represent all of the neurons recorded. Figure 3A shows an example of three stimulus SDs. Each point represents the firing rates of each neuron before and after the nerve cut. The red points indicate the average firing rates. To quantify the changes in average firing rates after the nerve was severed in all neurons, we devised the amplification measure, which is the ratio between firing rates prior and after nerve cutting. Figure 3B reveals amplification values higher and lower than 1 for lower and higher stimulus SDs, respectively. Thus, motor nerve cut resulted in a marked reduction and increase in firing rates for low and high variances stimuli, respectively. To translate these results into actual population firing rates, we calculated the average firing rates for each stimulus intensity (Three values are shown in Fig. 3A) and created the graph in Fig. 3C (black dots and line). To determine the influence of nerve cutting on firing rates, we then multiplied each firing rate value by the amplification factor shown in Fig. 3B. We show a change in the slope (Control: slope = 0.0023, r2 = 0.98; muscular feedback: slope = 0.0012, r2 = 0.87). These results suggest that motor feedback exerts a powerful influence on tactile inputs by differentially controlling sensitivity to tactile inputs as a function of stimulus magnitude.
Fig. 2.Cutting the motor nerve influences the firing rates of trigeminal ganglion (TG) neurons. A: schematic diagram of the experimental paradigm. The seventh nerve was cut to eliminate motor feedback. B: raster plot and a peri-stimulus time histograms (PSTH) of the TG neuronal response before and after motor nerve cut. The scale bar near the PSTH indicate firing probability. C: changes in adjacent vibrissa movement during control and after nerve cut. D: reduction of firing rates of the neuron in B at all stimulus SDs range after nerve cutting. BL, buccolabialis branch of facial nerve; FN, facial nucleus; IoN, infraorbital branch of trigeminal nerve; PrV, principal trigeminal nucleus; RF, reticular formation; SpVc, spinal trigeminal nucleus pars caudalis; SpVi, spinal trigeminal nucleus interpolaris; TG, trigeminal ganglion; TN, trigeminal nuclear complex.
Fig. 3.Summary of the influence of motor nerve cutting on neuronal firing rates. A: distribution of neuronal firing rates before and after nerve cut in three separate stimulus SDs. The red dots indicate the mean firing rates before and after nerve cutting. B: amplification of neuronal firing rate by sensory feedback at all stimuli SD. Amplification was calculated by the ratio between firing rates prior and after nerve cutting. C: effects of muscular feedback on the input/output transformation of TG neurons. The black points represent the average firing rates for all neurons for all stimulus intensity after nerve cut. The black line is the linear regression to these points. The red dots and line are the hypothetical plot of the firing rate of a neuron as a function of tactile stimulus magnitude. Adding muscular feedback results in a gain modulation, which corresponds to a tipping of the firing-rate curve, resulting in a change of its slope. *P < 0.01.
In the second line of enquiry, we performed the opposite procedure by introducing motor output and examining its influence on tactile inputs. We induced subliminal facial muscle activation by BL stimulation (duration of 100 µs at 143 Hz for the duration of the tactile stimuli). We initially cut the BL nerve and stimulated it and then monitored the EMG and movement of adjacent vibrissa. We adjusted the stimulus strength to reach a point where we could see EMG activity but no vibrissa movement (Fig. 4B). An example of changes in the firing rate of one neuron is shown in Fig. 4C. The figure depicts the dramatic decrease in the firing rate during nerve stimulation (Fig. 4, C and D). In this experiment, we also saw a mixed trend with a clear tendency toward a decrease in the firing rate in 60–65% of the population of records (n = 22; Fig. 3E). Unlike the cutting experiment, we did not see a change in the ratio of low-to-high sensory stimulation intensity, suggesting that the influence of motor feedback on neuronal responses is intractably dependent on sensory stimulus magnitude.
Fig. 4.Stimulating the buccolabialis motor branch influences the firing rates of trigeminal ganglion (TG) neurons. A: schematic diagram of the experimental paradigm. The seventh nerve was cut to eliminate motor feedback. A hook stimulating bipolar electrode was placed on the nerve. EMG activity was recorded from the pad, and a single vibrissa was monitored for movement. B: raster plot and peri-stimulus time histograms (PSTH) of TG neuronal response before and during motor nerve stimulation. C: changes in adjacent vibrissa movement during control and buccolabialis motor branch stimulation. D: reduction of firing rates of the neuron in A at all stimulus variances during motor nerve stimulation. E: changes in neuronal firing rates, proportion of facilitated, suppressed, and no change at three stimulus variances, during motor nerve stimulation. BL, buccolabialis branch of facial nerve; FN, facial nucleus; IoN, infraorbital branch of trigeminal nerve; PrV, principal trigeminal nucleus; RF, reticular formation; SpVc, spinal trigeminal nucleus pars caudalis; SpVi, spinal trigeminal nucleus interpolaris; TG, trigeminal ganglion; TN, trigeminal nuclear complex.
Next, we examined the influence of motor feedback on response variability and reliability. First, we estimated the spike count variability across trials by calculating response variance. This measure was calculated in control conditions, during motor nerve stimulation, and after the nerve was severed. Typical examples of both of these paradigms are shown in Fig. 5, A and B. Cutting the facial motor nerve resulted in a marked decrease in response variability (Fig. 5A). Quantification of response variability of all the neurons in the current paradigm (Fig. 5C; n = 20) showed a significant decrease in response variability. By contrast, stimulation of the facial nerve following nerve cut resulted in an increase in response variance (Fig. 5D; n = 20).
Fig. 5.The influence of motor nerve cut and stimulation on response variability. A and B: raster plot and peri-stimulus time histograms (PSTH) of the TG neuronal response before and after nerve cut (A) and during buccolabialis motor branch stimulation (B). C: influence of nerve cut on response variance. D: influence of nerve stimulation on response variance.
Influence of Tactile Stimuli on Whisking
While these findings suggest that muscular feedback may influence sensory transformation during passive surface contact, they do not directly address the influence of tactile input on whisker movements. To examine these issues, we induced whisking in anesthetized rats by activation of the SC (Hemelt and Keller 2008; McHaffie and Stein 1982) and the M1 (Cramer and Keller 2006; Cramer et al. 2007; Haiss and Schwarz 2005). These two types of stimulations activate different anatomical pathways, which result in different whisking activation patterns (see discussion). To determine how tactile input may regulate vibrissae movements, we applied mechanical stimuli through a metal mesh to most of the vibrissae.
In the first set of experiments, we concomitantly induced vibrissae movement through microstimulation in the SC of anesthetized rats, while monitoring the movement of the vibrissa that was not connected to the mesh (Fig. 6A, red dot). Figure 6B depicts typical examples during SC stimulation-evoked vibrissa movement and the influence of vibrissae stimulation on this movement. Here, and in all cases (n = 12), stimulation in the SC produced a sustained protraction of the vibrissae, which frequently outlasted the stimulation period (Fig. 6, B and E). Application of filtered white noise stimuli to the vibrissae resulted in reduced SC stimulation-evoked vibrissa movement. We modified the magnitude of the tactile stimuli (Fig. 6B) and found that increasing tactile stimulus magnitude influenced the degree of whisking suppression, as indicated by the area under the curve (see materials and methods; Fig. 6C). When vibrissae stimulation occurs during whisking, a fast reflexive response termed the touch-induced pump (TIP) may be induced (Deutsch et al. 2012). To examine whether we could replicate TIP in the current experimental paradigm, we timed a short tactile stimulus during whisking and found the same effect of touch-induced suppression and facilitation, depending on the timing between the two stimuli (Fig. 6E). In earlier phases of the whisk, tactile stimulus caused slight facilitation followed by minor depression, whereas in later stages, tactile stimulus mainly caused depression. This was observed in more than 80% of the cases (Fig. 6, E and F).
Fig. 6.Attenuation of superior colliculus (SC) evoked whisking by vibrissa stimulus. A: simplified description of the experimental procedure in which a galvanometer stimulated most of the vibrissae through a mesh whereas a single vibrissa (red) was left untouched, and its movement monitored. B: tactile stimuli attenuate SC-evoked whisking. The degree of attenuation was dependent on tactile stimulus magnitude. C: mean changes in vibrissa tactile stimulus attenuated the area under the whisking curve. Dashed line indicates control conditions. D: stimulation of the vibrissae during whisking induces suppression (1) or facilitation (2) in whisk amplitude. Tactile stimulus magnitude influences the degree of suppression of SC-evoked whisking. E: proportion of affected sessions and the average suppression and facilitation in the area under the curve of SC-evoked whisking. BL, buccolabialis branch of facial nerve; FN, facial nucleus; IoN, infraorbital branch of trigeminal nerve; PrV, principal trigeminal nucleus; RF, reticular formation; SpVc, spinal trigeminal nucleus pars caudalis; SpVi, spinal trigeminal nucleus interpolaris; TG, trigeminal ganglion; TN, trigeminal nuclear complex.*Significantly different, P = 0.05.
Influence of Tactile Stimuli on Periodic Whisking
Compared with the protraction movements induced by SC stimulation, the characteristics of the M1-induced vibrissae movements differed fundamentally. In the second set of experiments, we introduced long-train stimulation in M1. These stimulation generated rhythmic movements at frequencies between 5 and 10 Hz (n = 6; Fig. 7), and a latency of around 80–100 ms in all cases. In particular, these movements were virtually indiscernible from natural whisking movements and from those shown in previous studies (Haiss and Schwarz 2005). We then repeated the previous experimental paradigm by pairing whisking with vibrissae stimuli. We employed two paradigms in which tactile stimulus followed M1-evoked whisking (Fig. 7A) and another example in which tactile stimulus preceded M1-evoked whisking (Fig. 7B). In both of these paradigms, tactile inputs substantially suppressed whisking (Fig. 7C). Thus, overall, these results indicate that tactile inputs may affect motor whisking output through influences to the rhythmic whisking circuitry that alters the whisking kinetics. Likewise, tactile influences may also modify the whisking amplitude through synaptic and intrinsic neuronal interaction in the facial nucleus, resulting in the facilitation or suppression of whisking amplitude.
Fig. 7.Suppression of motor cortex (M1)-evoked whisking by vibrissae stimulus. A: tactile stimuli suppressed M1-evoked whisking (protraction only). B: vibrissa stimulus preceding M1 stimulus suppressed M1-evoked whisking, but did not influence spontaneous whisking (protraction and retraction). C: degree of suppression of vibrissa stimuli on M1-evoked whisking in A and B.
Neuronal Mechanisms of Tactile Influences
To elucidate the neuronal mechanisms underlying these sensorimotor interactions, we used brain stem slices (Fig. 8B). We recorded from FN motor neurons, using the whole cell technique (soma size of 25–30 µm; Fig. 8, B and C), while stimulating the infraorbital branch (IoN) of the trigeminal nerve. The recorded neurons were loaded with QX-314 (3 mM) to block voltage-activated channels. Quantification of membrane properties revealed that the membrane resting potential was −67.2 ± 3 mV (n = 38), with an apparent input resistance of 70.6 ± 31 MΩ (n = 20).
Fig. 8.Synaptic inputs to facial nucleus neurons contain excitatory and inhibitory components. A: schematic of the experimental diagram. B: images of a transverse brain stem slice from the rat P10 and schematic localization of facial nucleus and electrode positioning for stimulation of the IoN and motor inputs. C: lower (left) and higher (right) image magnification view of the trigeminal nerve and facial nucleus neuron labeling. D: reversal potential of synaptic responses in FN neurons to trigeminal nerve stimulation reveals excitatory and inhibitory components (traces are the average of 10 trials). E: average V/I curve for the neuron in the upper panels at two time points marked by dashed lines. F: postsynaptic depression of the FN neuronal response at 5 and 20 Hz. Traces are the averages of 10 trials. Inset: synaptic summation at 20 Hz. G: supralinear and sublinear summation of synaptic inputs to FN neurons. Insets: comparisons of observed and predicted responses. EPSP, excitatory postsynaptic potentials; FN, facial nucleus; IoN, infraorbital branch of trigeminal nerve; IPSP, inhibitory postsynaptic potentials; PrV, principal trigeminal nucleus; RF, reticular formation; SpVc, spinal trigeminal nucleus pars caudalis; SpVi, spinal trigeminal nucleus interpolaris; TG, trigeminal ganglion; TN, trigeminal nuclear complex.
Ipsilateral IoN stimulation elicited delayed subthreshold PSPs (Fig. 8D). The stimulus evoked one or several PSPs (~10% of the neurons) when the stimulation current was increased (Fig. 8D, 2). We observed that the IoN-evoked PSPs had a voltage dependency. We found three synaptic response groups, which are summarized in Table 1: 1) a group with a shorter latency excitatory postsynaptic potentials (EPSPs) and a longer latency inhibitory postsynaptic potentials (IPSPs) (n = 8; Fig. 8D, 1), 2) a group in which the neurons only expressed EPSPs (n = 13; Fig. 8D, 3), and 3) neurons only expressing IPSPs (n = 19; Fig. 8D, 4). To confirm these observations, we calculated the reversal potential of each of these components and their intrinsic membrane properties. We compared both input impedance and membrane time constants. We found that these groups did not differ in their membrane properties (Table 1).
| EPSP | EPSP+IPSP | IPSP | |
|---|---|---|---|
| Time constant, ms | 19.3 ± 7.9 | 21.7 ± 11.4 | 23.5 ± 7.5 |
| n = 13 | n = 6 | n = 24 | |
| Input resistance, MΩ | 59.5 ± 34.5 | 65.2 ± 40.1 | 75.1 ± 31.2 |
| n = 13 | n = 6 | n = 24 | |
| Resting membrane potential, mV | −65 ± 8 | −69 ± 8 | −66 ± 10 |
| EPSP latency, ms | 4.5 ± 1.4* | 7.2 ± 4.2* | |
| n = 13 | n = 8 | ||
| EPSP reversal potential, mV | 20.1 ± 19.7 | −4.4 ± 29. | |
| n = 10 | n = 6 | ||
| IPSP latency, ms | 14.6 ± 8.0* | 6.3 ± 2.8* | |
| n = 8 | n = 19 | ||
| IPSP reversal potential, mV | −38.2 ± 11.9 | −47.6 ± 21 | |
| n = 6 | n = 20 |
The existence of excitatory and inhibitory components is further corroborated in Fig. 8E. Here, we show that the reversal potential of the excitatory component was around 0 mV, whereas the reversal potential of the inhibitory component was approximately −50 mV. These inhibitory currents could be mediated by GABAergic or glycinergic receptors. The contribution of GABA and Glycine channels was examined by pharmacological blockage. The addition of 10 μM of bicuculline methiodide, an antagonist of GABAA, through bath perfusion, slightly reduced the IPSP amplitude. Additional application of 8 μM PMBA, which in small concentrations blocks Glycine currents (Saitoh et al. 1994), blocked the IPSP completely (not shown). These results indicate that IPSPs are mediated by both GABA and glycine receptors. This finding is not surprising, since studies have reported the presence of GABAergic and glycinergic neural projections to the facial nucleus (Li et al. 1997; Zafra et al. 1995). Examination of IPSP properties in Table 1 reveals depolarized reversal potentials. This may result from the temporal overlap between IPSP/EPSP, leading to the contamination in our measurements. We wish to stress that while these calculations may not reflect the true reversal potential. Nevertheless, we find the comparison instructive, as it enables us to differentiate between the different pathways.
To examine the temporal evolution of the tactile inputs, we introduced IoN periodic stimuli at 5 and 20 Hz to mimic exploratory and foveal whisking, respectively (Berg and Kleinfeld 2003). At these frequencies, the responses were observed to depress after successive periodic stimuli (Fig. 8F). While each PSP showed reduced amplitude, temporal summation showed a buildup in membrane potential which resulted at times in spike discharge (Fig. 8F, inset). This result can be explained by the large membrane time constant (τ) 15.6 ± 5.6 (n = 25) of FN neurons, which is significantly longer than reported previously (Nguyen and Kleinfeld 2005). Finally, we paired IoN stimulation with electrical stimulation of short latency (1–2 ms) afferent pathways reaching the FN located rostral to it (Fig. 8B). The second electrode was placed in a location to stimulate separate fibers impinging on the FN. To exclude the role of driving force in these interactions, we recorded the PSP generated by motor inputs at different membrane voltages. As expected, the evoked motor PSP amplitude alone was not significantly affected by changes in the resting membrane potential (± 3 mV; not shown). We then examined the temporal interactions between the two responses (n = 14). We compared the magnitude of the observed responses and their arithmetic sum and calculated the ratio between the observed and expected responses (Fig. 8G). Traces of the observed and expected interactions are shown in the insets (Fig. 8G). When both PSPs arrived almost simultaneously (±5 ms), they summed supralinearly (Fig. 8G). By contrast, when motor PSPs followed IoN PSP (5–25 ms), the inputs summed sublinearly and show a depression of ~30%. Altogether, these results suggest that increases and decreases in whisking amplitude may be the result of intrinsic and synaptic interactions in facial nucleus neurons.
DISCUSSION
The present report provides a systematic analysis of sensorimotor integration occurring in the brain stem loop, which involves inputs from primary afferents through trigeminal brain stem nuclei to the facial nucleus to the muscles of the pad that control vibrissae movement. We found, as others have done, that passive tactile inputs result in depressing motor output to pad intrinsic and extrinsic muscles (Fig. 1). We show that these muscular activations stem from sensory signal-generated synaptic summation in FN neurons (Fig. 8F). Nevertheless, activation of these two antagonistic muscles did not lead to any vibrissae movement (Fig. 1, G and H). This motor feedback is tuned to the magnitude of sensory input (Fig. 1) and can serve as a dynamic range control system, which differentially reduces the firing rates of first-order sensory neurons to high inputs and increases low-magnitude inputs, thus reducing or compressing the output dynamic range (Figs. 2 and 3). The influence of motor feedback also emerged as an increase in response variability (Fig. 5). During active whisking, we found that tactile inputs influenced vibrissa motion depending on the activated whisking pathway. When whisking was induced by SC stimulation, tactile stimuli attenuated SC-induced whisking in a stimulus magnitude-dependent manner (Fig. 6). Precise temporal pairing of whisking with tactile stimulus revealed a phenomenon similar to TIP (Fig. 6). Sublinear summation can be attributed to longer latency inhibitory mechanisms, whereas the supralinear summation could result from voltage-dependent conductances or synaptic transmission (Fig. 7). Thus, some tactile influences could be the result of cellular synaptic interactions at the level of the FN. Stimulation of the motor cortex led to manifestations of periodic whisking similar to natural whisking. Tactile stimuli attenuated M1-induced whisking almost completely (Fig. 8), suggesting a direct influence of tactile stimuli on periodic whisking mechanisms. These results suggest that object touch may influence vibrissa motion, depending on the activated pathway and the behavioral paradigm that the animal engages in, and that the vibrissa movements evoked by M1 and SC are mediated through different pathways and probably serve distinct functional roles during behavior. Thus, overall, the whisker brain stem loop serves a dual role, which involves the motor control of sensory inputs, as well as the sensory control of motor output.
Circuitry Underlying Vibrissa Brain Stem Loop
Several seminal studies have revealed the circuitry underlying sensorimotor loops for reflex motion of the vibrissae upon contact with an object. There is ample evidence that the sensory branch of the vibrissa BL is excitatory (Guido et al. 2001; Lo et al. 1999; Minnery et al. 2003; Onodera et al. 2000; Sosnik et al. 2001; Zucker and Welker 1969). These neurons project to second-order sensory neurons in the trigeminal nuclear complex, in which both excitatory and inhibitory spinal trigeminal nucleus oralis (SpVo) neurons project to FN intrinsic vibrissa protractor muscle motoneurons (Pinganaud et al. 1999; Sreenivasan et al. 2015; Takatoh et al. 2013). In contrast, spinal trigeminal nucleus interpolaris (SpVi) neurons project to FN extrinsic vibrissa retractor muscle motoneurons (Bellavance et al. 2017; Sreenivasan et al. 2015; Takatoh et al. 2013). An additional pathway that may be implicated in this sensorimotor pathway is the trigeminal nucleus pars muralis (Matthews et al. 2015), which has been shown by the same group to exert a strong influence on motor output (Nguyen and Kleinfeld 2005). Thus, the known anatomy supports a mixed inhibitory and excitatory response (Deutsch et al. 2012; Sherman et al. 2013). Nevertheless, the specifics of this circuitry and the potential involvement of different muscle groups in this BL remain unclear.
In the current study, we corroborated most of these findings using brain stem slices (Fig. 8). By changing the resting membrane potential and calculating the synaptic reversal potentials of FN neurons, we found all of the combinations mentioned above. First, we found pure excitatory inputs, which may correspond to both SpVo and SpVi synaptic inputs. Second, in a subset of the neurons, we found short latency pure inhibitory synaptic inputs, which may reflect di-synaptic inhibitory inputs originating from the SpVo or from the reticular formation (RF). Finally, we found an additional synaptic input, which combines a short latency EPSP with a longer latency IPSP. This may result from an additional sensory pathway originating from the trigeminal brain stem nuclei through shorter latency direct excitation and longer latency disynaptic inhibition within the FN or through the RF (Li et al. 1997; Shammah-Lagnado et al. 1992; Ter Horst et al. 1991). These inhibitory inputs are mediated by GABAA and glycine receptors (Li et al. 1997; Zafra et al. 1995).
Further support for the influence of this well-established pathway comes from experiments in which we tested the influence of tactile stimulus on SC- and M1-induced whisking. In these experiments, we found the tactile stimulus only moderately reduced SC-induced whisking (Fig. 6) while suppressing M1-induced whisking completely (Fig. 7). It has been shown that the SC sends dense, direct projections to the FN, where the vibrissa motor neurons are located (Hattox et al. 2002; Miyashita et al. 1994; Miyashita and Mori 1995; Vidal et al. 1988), and stimulation of the superior colliculus has been reported to produce movements of the vibrissae (Hemelt and Keller 2008; McHaffie and Stein 1982). In contrast, anatomical and physiological data suggest that some premotor structures receive inputs from M1. The rhythmic whisker protraction evoked by M1 stimulation might be driven by excitation of excitatory and inhibitory premotor neurons in the brain stem reticular formation that innervate both intrinsic and extrinsic muscles (Alloway et al. 2010; Dauvergne et al. 2001; Matyas et al. 2010; Sreenivasan et al. 2016; Zerari-Mailly et al. 2001). Thus, a tactile stimulus should only reduce the magnitude of SC-induced whisking (Fig. 6), yet disrupt the M1-induced oscillatory whisking pattern generated in the RF (Fig. 7) (Deschênes et al. 2016; Moore et al. 2013). In contrast to our in vitro results, a previous study using horizontal brain stem slices (Nguyen and Kleinfeld 2005) did not find this plethora of trigemino-facial synaptic inputs and, presumably, the different pathways. This discrepancy may result from two main plausible explanations: First, the previous study used sharp intracellular recording techniques, in which the series resistance was high. Therefore, estimation of synaptic reversal potential and the magnitude of input impedance was misestimated. Additionally, the previous study used brain stem slices, which were 1 mm thick. We believe that this resulted in a lack of oxygenation and perfusion of the slices, thus rendering them less viable.
Influence of Tactile Stimuli on Motor Output
To examine the functional significance of the trigemino-facial sensorimotor loop, the effect of vibrissa contact on whisking kinematics needs to be observed in the course of different behavioral conditions in freely moving and head-restrained rodents during both active and passive touch (Deutsch et al. 2012; Grant et al. 2009; Kerekes et al. 2017; Mitchinson et al. 2007; Sachdev et al. 2003; Voigts et al. 2015). There are several simple explanations available for the control of the vibrissa pad that involve a combination of the major intrinsic and extrinsic muscle groups responsible for vibrissa protraction and retraction, respectively. Thus, touch-induced retraction may occur by inhibition of the intrinsic motoneurons and/or by excitation of extrinsic motoneurons. By contrast, touch-induced protraction may occur by excitation of the intrinsic motoneurons and/or by inhibition of extrinsic protraction motoneurons (Ashwell 1982; Courville 1966; Dörfl 1982; Herfst and Brecht 2008; Klein and Rhoades 1985; Komiyama et al. 1984).
Several studies have examined the functional significance of the trigemino-facial sensorimotor loop; however, the results are inconsistent. Upon contact with an object, vibrissae on the side of contact may rapidly and transiently protract (Sachdev et al. 2003), retract (Mitchinson et al. 2007), or exhibit a retraction followed by a protraction known as TIP (Deutsch et al. 2012), which was suggested as a feedback mechanism to facilitate object characterization. On a more general level, retraction and protractions may be related to the animal's current goals and focus (Deutsch et al. 2012; Gordon et al. 2014; Grant et al. 2009; Mitchinson et al. 2007; Sherman et al. 2017; Voigts et al. 2015). Thus, the nature of the vibrissa-object interaction is modulated by motor strategies and the context of the behavioral setting, all of which influence the spatiotemporal interactions between excitation and inhibition. These sensorimotor interactions could occur through parallel pathways in the brain stem (McElvain et al. 2018; Sherman et al. 2013) and the brain (Bosman et al. 2011; Kleinfeld et al. 1999) and be modulated and gated by numerous types of neuromodulators (Bosman et al. 2011; Timofeeva et al. 2005).
In the current study, we explored the putative principles governing vibrissa-object interactions. These interactions can be divided roughly into two different nonmutually exclusive entities. The first is the influence of touch on object-vibrissa interactions. The second is the influence of touch on periodic vibrissa movement. We, as well as others (Kleinfeld et al. 2002; Nguyen and Kleinfeld 2005), have found that passive vibrissa touch triggers a sensory signal-dependent increase in the drive to both the intrinsic and extrinsic vibrissa muscles (Fig. 1). Further support for the nature of this feedback comes from an additional study in which an increase in the amplitude of the mystacial EMG was detected when rats touched objects with their vibrissae (Sachdev et al. 2003). We found that by apparently activating these two antagonistic muscles, no significant vibrissa movement was detected (Fig. 1, G and H). These effects may, thus, stem from sensory signal-generated synaptic summation in FN neurons (Fig. 8F) such that these forces act on the vibrissa follicles to influence the transformation of tactile information in a stimulus-dependent manner (see below).
Studying the influence of touch on object-vibrissa interactions, we induced whisking by SC stimulation. We showed that tactile stimulation during SC-induced vibrissa movement that can cause a reduction or a facilitation of whisking amplitude depending on the temporal phase between them. That is, less or more pressure of the whiskers on contacted objects. Since SC-induced whisking is mediated through direct influences on the facial nucleus, any tactile influences on motor output may be the result of synaptic inputs impinging of FN neurons. To examine the cellular mechanisms underlying these differential influences, we paired IoN branch stimulation with other synaptic inputs. These afferents while of unknown origin sum supralinearly or sublinearly with sensory synaptic inputs (Fig. 8G; these are not IoN fibers, since we only get pair-pulse depression in that case). We found that sublinear summation could result from longer latency inhibitory mechanisms, whereas the supralinear summation can result from voltage-dependent conductances or synaptic transmission. These nonlinearities can arise from dendritic and somatic origin. Overall, our results tend to indicate that the effects of sensory inputs on whisker motion can also arise as a result of synaptic interaction. Note that all the interactions shown here, as well as observations in other studies, could be modulated by top-down pathways that are known to regulate numerous interactions, as well as, gate various loops. (Killackey et al. 1989; Lee et al. 2008; Pinganaud et al. 1989; Wise et al. 1979; Woolston et al. 1983). The brain stem loop itself can also be modulated by the cholinergic, serotonergic, and noradrenergic modulatory systems (Bosman et al. 2011).
Rhythmic motion of the vibrissae, denoted whisking, is driven by a neuronal oscillator located in the intermediate reticular formation (Deschênes et al. 2016; Moore et al. 2013) that can operate independently of direct sensory feedback (Berg and Kleinfeld 2003; Gao et al. 2003; Hattox et al. 2003; Moore et al. 2013; Welker 1964). Here, we showed a tactile stimulus disrupts M1-induced whisking mediated through the RF (Fig. 7). The rhythmic whisker protraction evoked by M1 stimulation might be driven by excitation of excitatory and inhibitory premotor neurons in the brain stem reticular formation that innervate both intrinsic and extrinsic muscles. This is reminiscent of the influence of the breathing pacemaker known as the pre-Bötzinger complex, which modulates whisking (Deschênes et al. 2016; Moore et al. 2013), and may also explain the changes that occur in whisking kinetics when a rodent touches an object (Knutsen et al. 2006; O’Connor et al. 2010). It is plausible that in awake behaving animals, object touch information through this pathway can modify and not disrupt whisking completely. These results suggest that vibrissa movements evoked by M1 and SC may be mediated through different pathways and probably serve distinct functional roles during behavior and that object touch may influence vibrissa motion as a function of the activated pathway and behavioral paradigm by which the animal engages.
Control of Tactile Responses by Motor Output
Multiple mechanisms are responsible for shaping the flow of tactile information through the trigeminal nuclei. Here, we reported additional mechanisms that precede the trigeminal nuclei and are mediated through muscular control of the follicle. Each vibrissa is held by a follicle embedded in the mystacial pad. The vibrissa follicle is populated by receptors with assorted morphologies and locations (Ebara et al. 2002; Rice et al. 1986). The relationship between the morphology and location of a receptor and its specific neuronal response properties remain unknown. However, the damping mechanism around the follicle known as the pressure plate that the mechanoreceptor depends on, is thought to play a key role in these properties (Dörfl 1982; Wineski 1985). More importantly, in the context of the current study four factors were shown to potentially contribute to the effects of this pressure plate: the differential blood flow and pressure regulation in the sinus surrounding each vibrissal follicle, contractions of the facial muscles, elastic rebound in the connective tissues, and finally the degree of vibrissa deflection (Whiteley et al. 2015). We also found muscular regulation of the flow of tactile information. This feedback mechanism is tuned to the magnitude of the sensory input and can serve as a dynamic gain control system, which is a change in the slope of the firing-rate curve, corresponding to a multiplicative or divisive scaling, which is distinct from these additive or subtractive shifts (Fig. 3C). These mechanisms can provide the whisker somatosensory system with a way to cope with a changing tactile environment whose intensities usually span several orders of magnitude, such as textures and objects. Furthermore, since muscular pad activity can be controlled by several mechanisms (see above) it is plausible that rodents may use this system to control the transformation of tactile vibrissa stimuli into neuronal discharge in the primary afferents. The cost of these mechanisms lies in the increase in response variability and a reduction in reliability, which is already high compared with other neurons higher up in the whisker somatosensory hierarchy. These mechanisms are similar to the signal compression problem in the peripheral auditory nervous system, as well as the muscles of the pupil in the eye.
GRANTS
This research was supported by Israel Science Foundation and the Leona M. and Harry B. Helmsley Charitable Trust (Helmsley Charitable Trust) to R. Azouz.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the authors.
AUTHOR CONTRIBUTIONS
R.A. conceived and designed research; O.T. and Y.K. performed experiments; O.T. and Y.K. analyzed data; O.T. and R.A. interpreted results of experiments; O.T. and R.A. prepared figures; R.A. drafted manuscript; R.A. edited and revised manuscript; R.A. approved final version of manuscript.
ACKNOWLEDGMENTS
We thank Milos Bogdanovic for help with the histological procedures.
REFERENCES
- . Closed-loop neuronal computations: focus on vibrissa somatosensation in rat. Cereb Cortex 13: 53–62, 2003. doi:10.1093/cercor/13.1.53.
Crossref | PubMed | ISI | Google Scholar - . Quantitative analysis of the bilateral brainstem projections from the whisker and forepaw regions in rat primary motor cortex. J Comp Neurol 518: 4546–4566, 2010. doi:10.1002/cne.22477.
Crossref | PubMed | ISI | Google Scholar - . Somatotopic organization of vibrissae afferents in the trigeminal sensory nuclei of the rat studied by transganglionic transport of HRP. J Comp Neurol 211: 84–92, 1982. doi:10.1002/cne.902110108.
Crossref | PubMed | ISI | Google Scholar - . The adult mouse facial nerve nucleus: morphology and musculotopic organization. J Anat 135: 531–538, 1982.
PubMed | ISI | Google Scholar - . On the central course of afferent fibers in the trigeminal, facial, glossopharyngeal, and vagal nerves and their nuclei in the mouse. Acta Physiol Scand 29, S106: 209–320, 1953. doi:10.1111/apha.1953.29.s106.209.
Crossref | Google Scholar - . Neuron numbers in the sensory trigeminal nuclei of the rat: a GABA- and glycine-immunocytochemical and stereological analysis. J Comp Neurol 493: 538–553, 2005. doi:10.1002/cne.20778.
Crossref | PubMed | ISI | Google Scholar - . Feedforward inhibition determines the angular tuning of vibrissal responses in the principal trigeminal nucleus. J Neurosci 30: 1057–1063, 2010. doi:10.1523/JNEUROSCI.4805-09.2010.
Crossref | PubMed | ISI | Google Scholar - . Parallel inhibitory and excitatory trigemino-facial feedback circuitry for reflexive vibrissa movement. Neuron 95: 673–682.e4, 2017. [Erratum in Neuron 95:722–723, 2017.] doi:10.1016/j.neuron.2017.06.045.
Crossref | PubMed | ISI | Google Scholar - . Rhythmic whisking by rat: retraction as well as protraction of the vibrissae is under active muscular control. J Neurophysiol 89: 104–117, 2003. doi:10.1152/jn.00600.2002.
Link | ISI | Google Scholar - . Anatomical pathways involved in generating and sensing rhythmic whisker movements. Front Integr Neurosci 5: 53, 2011. doi:10.3389/fnint.2011.00053.
Crossref | PubMed | Google Scholar - . Responses in the rat thalamus to whisker movements produced by motor nerve stimulation. J Physiol 238: 387–401, 1974. doi:10.1113/jphysiol.1974.sp010531.
Crossref | PubMed | ISI | Google Scholar - . Histologie du Systeme Nerveux de l’Homme et des Vertebres. Madrid, Spain: Instituto Rarnon y Cajal, 1955, vol 2.
Google Scholar - . Biometric analyses of vibrissal tactile discrimination in the rat. J Neurosci 10: 2638–2648, 1990. doi:10.1523/JNEUROSCI.10-08-02638.1990.
Crossref | PubMed | ISI | Google Scholar - . Electromyographic activity of mystacial pad musculature during whisking behavior in the rat. Somatosens Mot Res 8: 159–164, 1991. doi:10.3109/08990229109144740.
Crossref | PubMed | ISI | Google Scholar - . The nucleus of the facial nerve; the relation between cellular groups and peripheral branches of the nerve. Brain Res 1: 338–354, 1966. doi:10.1016/0006-8993(66)90126-0.
Crossref | PubMed | Google Scholar - . Cortical control of a whisking central pattern generator. J Neurophysiol 96: 209–217, 2006. doi:10.1152/jn.00071.2006.
Link | ISI | Google Scholar - . The whisking rhythm generator: a novel mammalian network for the generation of movement. J Neurophysiol 97: 2148–2158, 2007. doi:10.1152/jn.01187.2006.
Link | ISI | Google Scholar - . Synaptic mechanisms underlying sparse coding of active touch. Neuron 69: 1160–1175, 2011. doi:10.1016/j.neuron.2011.02.022.
Crossref | PubMed | ISI | Google Scholar - . Reticular premotor neurons projecting to both facial and hypoglossal nuclei receive trigeminal afferents in rats. Neurosci Lett 311: 109–112, 2001. doi:10.1016/S0304-3940(01)02150-4.
Crossref | PubMed | ISI | Google Scholar - . Sniffing and whisking in rodents. Curr Opin Neurobiol 22: 243–250, 2012. doi:10.1016/j.conb.2011.11.013.
Crossref | PubMed | ISI | Google Scholar - . Inhibition, not excitation, drives rhythmic whisking. Neuron 90: 374–387, 2016. doi:10.1016/j.neuron.2016.03.007.
Crossref | PubMed | ISI | Google Scholar - . Fast feedback in active sensing: touch-induced changes to whisker-object interaction. PLoS One 7:
e44272 , 2012. doi:10.1371/journal.pone.0044272.
Crossref | PubMed | ISI | Google Scholar - . ‘Where’ and ‘what’ in the whisker sensorimotor system. Nat Rev Neurosci 9: 601–612, 2008. [Erratum in Nat Rev Neurosci 9: 709, 2008.] doi:10.1038/nrn2411.
Crossref | PubMed | ISI | Google Scholar - . The musculature of the mystacial vibrissae of the white mouse. J Anat 135: 147–154, 1982.
PubMed | ISI | Google Scholar - . Similarities and differences in the innervation of mystacial vibrissal follicle-sinus complexes in the rat and cat: a confocal microscopic study. J Comp Neurol 449: 103–119, 2002. doi:10.1002/cne.10277.
Crossref | PubMed | ISI | Google Scholar - . Central versus peripheral determinants of patterned spike activity in rat vibrissa cortex during whisking. J Neurophysiol 78: 1144–1149, 1997. doi:10.1152/jn.1997.78.2.1144.
Link | ISI | Google Scholar - . Inhibitory gating of vibrissal inputs in the brainstem. J Neurosci 28: 1789–1797, 2008. doi:10.1523/JNEUROSCI.4627-07.2008.
Crossref | PubMed | ISI | Google Scholar - . Whisker motor cortex ablation and whisker movement patterns. Somatosens Mot Res 20: 191–198, 2003. doi:10.1080/08990220310001622924.
Crossref | PubMed | ISI | Google Scholar - . Observations on active touch. Psychol Rev 69: 477–491, 1962. doi:10.1037/h0046962.
Crossref | PubMed | ISI | Google Scholar - . Emergent exploration via novelty management. J Neurosci 34: 12646–12661, 2014. doi:10.1523/JNEUROSCI.1872-14.2014.
Crossref | PubMed | ISI | Google Scholar - . Active touch sensing in the rat: anticipatory and regulatory control of whisker movements during surface exploration. J Neurophysiol 101: 862–874, 2009. doi:10.1152/jn.90783.2008.
Link | ISI | Google Scholar - . Synaptic plasticity in the trigeminal principal nucleus during the period of barrelette formation and consolidation. Brain Res Dev Brain Res 132: 97–102, 2001. doi:10.1016/S0165-3806(01)00283-8.
Crossref | PubMed | Google Scholar - . Spatial segregation of different modes of movement control in the whisker representation of rat primary motor cortex. J Neurosci 25: 1579–1587, 2005. doi:10.1523/JNEUROSCI.3760-04.2005.
Crossref | PubMed | ISI | Google Scholar - . Serotonin regulates rhythmic whisking. Neuron 39: 343–352, 2003. doi:10.1016/S0896-6273(03)00391-X.
Crossref | PubMed | ISI | Google Scholar - . Functional circuitry involved in the regulation of whisker movements. J Comp Neurol 442: 266–276, 2002. doi:10.1002/cne.10089.
Crossref | PubMed | ISI | Google Scholar - . Superior colliculus control of vibrissa movements. J Neurophysiol 100: 1245–1254, 2008. doi:10.1152/jn.90478.2008.
Link | ISI | Google Scholar - . Whisker movements evoked by stimulation of single motor neurons in the facial nucleus of the rat. J Neurophysiol 99: 2821–2832, 2008. doi:10.1152/jn.01014.2007.
Link | ISI | Google Scholar - . Structure-function relationships in rat brain stem subnucleus interpolaris: IV. Projection neurons. J Comp Neurol 282: 45–62, 1989a. doi:10.1002/cne.902820105.
Crossref | PubMed | ISI | Google Scholar - . Structure-function relationships in rat brain stem subnucleus interpolaris: III. Local circuit neurons. J Comp Neurol 282: 24–44, 1989b. doi:10.1002/cne.902820104.
Crossref | PubMed | ISI | Google Scholar - . Whiskers, barrels, and cortical efferent pathways in gap crossing by rats. J Neurophysiol 84: 1781–1789, 2000. doi:10.1152/jn.2000.84.4.1781.
Link | ISI | Google Scholar - . Bilateral discrimination of tactile patterns without whisking in freely running rats. J Neurosci 37: 7567–7579, 2017. doi:10.1523/JNEUROSCI.0528-17.2017.
Crossref | PubMed | ISI | Google Scholar - . The divisional organization of afferent fibres of the trigeminal nerve. Brain 86: 721–732, 1963. doi:10.1093/brain/86.4.721.
Crossref | PubMed | ISI | Google Scholar - . Laminar and areal differences in the origin of the subcortical projection neurons of the rat somatosensory cortex. J Comp Neurol 282: 428–445, 1989. doi:10.1002/cne.902820309.
Crossref | PubMed | ISI | Google Scholar - . Facial nerve injury induces facilitation of responses in both trigeminal and facial nuclei of rat. Neurosci Lett 358: 223–225, 2004. doi:10.1016/j.neulet.2004.01.037.
Crossref | PubMed | ISI | Google Scholar - . Representation of whisker follicle intrinsic musculature in the facial motor nucleus of the rat. J Comp Neurol 232: 55–69, 1985. doi:10.1002/cne.902320106.
Crossref | PubMed | ISI | Google Scholar - . Active sensation: insights from the rodent vibrissa sensorimotor system. Curr Opin Neurobiol 16: 435–444, 2006. doi:10.1016/j.conb.2006.06.009.
Crossref | PubMed | ISI | Google Scholar - . Anatomical loops and their electrical dynamics in relation to whisking by rat. Somatosens Mot Res 16: 69–88, 1999. doi:10.1080/08990229970528.
Crossref | PubMed | ISI | Google Scholar - . Neuronal basis for object location in the vibrissa scanning sensorimotor system. Neuron 72: 455–468, 2011. doi:10.1016/j.neuron.2011.10.009.
Crossref | PubMed | ISI | Google Scholar - . Adaptive filtering of vibrissa input in motor cortex of rat. Neuron 34: 1021–1034, 2002. doi:10.1016/S0896-6273(02)00732-8.
Crossref | PubMed | ISI | Google Scholar - . Haptic object localization in the vibrissal system: behavior and performance. J Neurosci 26: 8451–8464, 2006. doi:10.1523/JNEUROSCI.1516-06.2006.
Crossref | PubMed | ISI | Google Scholar - . Somatotopic representation of facial muscles within the facial nucleus of the mouse. A study using the retrograde horseradish peroxidase and cell degeneration techniques. Brain Behav Evol 24: 144–151, 1984. doi:10.1159/000121312.
Crossref | PubMed | ISI | Google Scholar - . Behavioral properties of the trigeminal somatosensory system in rats performing whisker-dependent tactile discriminations. J Neurosci 21: 5752–5763, 2001. doi:10.1523/JNEUROSCI.21-15-05752.2001.
Crossref | PubMed | ISI | Google Scholar - . Motor modulation of afferent somatosensory circuits. Nat Neurosci 11: 1430–1438, 2008. doi:10.1038/nn.2227.
Crossref | PubMed | ISI | Google Scholar - . Distribution of GABAergic and glycinergic premotor neurons projecting to the facial and hypoglossal nuclei in the rat. J Comp Neurol 378: 283–294, 1997. doi:10.1002/(SICI)1096-9861(19970210)378:2<283:AID-CNE10>3.0.CO;2-T.
Crossref | PubMed | ISI | Google Scholar - . Electrophysiological properties and synaptic responses of cells in the trigeminal principal sensory nucleus of postnatal rats. J Neurophysiol 82: 2765–2775, 1999. doi:10.1152/jn.1999.82.5.2765.
Link | ISI | Google Scholar - . The central projections of trigeminal primary afferent neurons in the cat as determined by the tranganglionic transport of horseradish peroxidase. J Comp Neurol 203: 785–798, 1981. doi:10.1002/cne.902030414.
Crossref | PubMed | ISI | Google Scholar - . Physiology and anatomy of synaptic connections between thick tufted pyramidal neurones in the developing rat neocortex. J Physiol 500: 409–440, 1997. doi:10.1113/jphysiol.1997.sp022031.
Crossref | PubMed | ISI | Google Scholar - . Feedback in the brain stem: an excitatory disynaptic pathway for control of whisking. J Comp Neurol 523: 921–942, 2015. doi:10.1002/cne.23724.
Crossref | PubMed | ISI | Google Scholar - . Motor control by sensory cortex. Science 330: 1240–1243, 2010. doi:10.1126/science.1195797.
Crossref | PubMed | ISI | Google Scholar - . Circuits in the rodent brainstem that control whisking in concert with other orofacial motor actions. Neuroscience 368: 152–170, 2018. doi:10.1016/j.neuroscience.2017.08.034.
Crossref | PubMed | ISI | Google Scholar - . Eye movements evoked by electrical stimulation in the superior colliculus of rats and hamsters. Brain Res 247: 243–253, 1982. doi:10.1016/0006-8993(82)91249-5.
Crossref | PubMed | ISI | Google Scholar - . Active spatial perception in the vibrissa scanning sensorimotor system. PLoS Biol 5:
e15 , 2007. doi:10.1371/journal.pbio.0050015.
Crossref | PubMed | ISI | Google Scholar - . Response transformation and receptive-field synthesis in the lemniscal trigeminothalamic circuit. J Neurophysiol 90: 1556–1570, 2003. doi:10.1152/jn.00111.2003.
Link | ISI | Google Scholar - . Feedback control in active sensing: rat exploratory whisking is modulated by environmental contact. Proc Biol Sci 274: 1035–1041, 2007. doi:10.1098/rspb.2006.0347.
Crossref | PubMed | ISI | Google Scholar - . Input-output organization of the rat vibrissal motor cortex. Exp Brain Res 99: 223–232, 1994. doi:10.1007/BF00239589.
Crossref | PubMed | ISI | Google Scholar - . The superior colliculus relays signals descending from the vibrissal motor cortex to the facial nerve nucleus in the rat. Neurosci Lett 195: 69–71, 1995. doi:10.1016/0304-3940(95)11782-R.
Crossref | PubMed | ISI | Google Scholar - . Hierarchy of orofacial rhythms revealed through whisking and breathing. Nature 497: 205–210, 2013. doi:10.1038/nature12076.
Crossref | PubMed | ISI | Google Scholar - . Optimal eye movement strategies in visual search. Nature 434: 387–391, 2005. doi:10.1038/nature03390.
Crossref | PubMed | ISI | Google Scholar - . Positive feedback in a brainstem tactile sensorimotor loop. Neuron 45: 447–457, 2005. doi:10.1016/j.neuron.2004.12.042.
Crossref | PubMed | ISI | Google Scholar - . Developmental regulation of active and passive membrane properties in rat vibrissa motoneurones. J Physiol 556: 203–219, 2004. doi:10.1113/jphysiol.2003.060087.
Crossref | PubMed | ISI | Google Scholar - . Vibrissa-based object localization in head-fixed mice. J Neurosci 30: 1947–1967, 2010. doi:10.1523/JNEUROSCI.3762-09.2010.
Crossref | PubMed | ISI | Google Scholar - . On the anatomical and functional organization of the spinal trigeminal nucleus. J Comp Neurol 92: 401–413, 1950. doi:10.1002/cne.900920305.
Crossref | PubMed | ISI | Google Scholar - . Primary afferent synaptic responses recorded from trigeminal caudal neurons in a mandibular nerve-brainstem preparation of neonatal rats. J Physiol 524: 503–512, 2000. doi:10.1111/j.1469-7793.2000.00503.x.
Crossref | PubMed | ISI | Google Scholar - . The functional organization of the barrel cortex. Neuron 56: 339–355, 2007. doi:10.1016/j.neuron.2007.09.017.
Crossref | PubMed | ISI | Google Scholar - . Trigeminal projections to hypoglossal and facial motor nuclei in the rat. J Comp Neurol 415: 91–104, 1999. doi:10.1002/(SICI)1096-9861(19991206)415:1<91:AID-CNE7>3.0.CO;2-A.
Crossref | PubMed | ISI | Google Scholar - . Structure, vascularization, and innervation of the mystacial pad of the rat as revealed by the lectin Griffonia simplicifolia. J Comp Neurol 337: 386–399, 1993. doi:10.1002/cne.903370304.
Crossref | PubMed | ISI | Google Scholar - . Comprehensive immunofluorescence and lectin binding analysis of vibrissal follicle sinus complex innervation in the mystacial pad of the rat. J Comp Neurol 385: 149–184, 1997. doi:10.1002/(SICI)1096-9861(19970825)385:2<149:AID-CNE1>3.0.CO;2-1.
Crossref | PubMed | ISI | Google Scholar - . A comparative light microscopic analysis of the sensory innervation of the mystacial pad. I. Innervation of vibrissal follicle-sinus complexes. J Comp Neurol 252: 154–174, 1986. doi:10.1002/cne.902520203.
Crossref | PubMed | ISI | Google Scholar - . Unilateral vibrissa contact: changes in amplitude but not timing of rhythmic whisking. Somatosens Mot Res 20: 163–169, 2003. doi:10.1080/08990220311000405208.
Crossref | PubMed | ISI | Google Scholar - . A novel antagonist, phenylbenzene omega-phosphono-alpha-amino acid, for strychnine-sensitive glycine receptors in the rat spinal cord. Br J Pharmacol 113: 165–170, 1994. doi:10.1111/j.1476-5381.1994.tb16189.x.
Crossref | PubMed | ISI | Google Scholar - . The organization of submodality-specific touch afferent inputs in the vibrissa column. Cell Reports 5: 87–98, 2013. doi:10.1016/j.celrep.2013.08.051.
Crossref | PubMed | ISI | Google Scholar - . Dynamics of active sensing and perceptual selection. Curr Opin Neurobiol 20: 172–176, 2010. doi:10.1016/j.conb.2010.02.010.
Crossref | PubMed | ISI | Google Scholar - . The facial “motor” nerve of the rat: control of vibrissal movement and examination of motor and sensory components. J Comp Neurol 247: 144–158, 1986. doi:10.1002/cne.902470203.
Crossref | PubMed | ISI | Google Scholar - . Afferent connections of the parvocellular reticular formation: a horseradish peroxidase study in the rat. Neuroscience 50: 403–425, 1992. doi:10.1016/0306-4522(92)90433-3.
Crossref | PubMed | ISI | Google Scholar - . Tactile modulation of whisking via the brainstem loop: statechart modeling and experimental validation. PLoS One 8:
e79831 , 2013. doi:10.1371/journal.pone.0079831.
Crossref | PubMed | ISI | Google Scholar - . Attention robustly gates a closed-loop touch reflex. Curr Biol 27: 1836–1843.e7, 2017. doi:10.1016/j.cub.2017.05.058.
Crossref | PubMed | ISI | Google Scholar - . Coding of deflection velocity and amplitude by whisker primary afferent neurons: implications for higher level processing. Somatosens Mot Res 17: 171–180, 2000. doi:10.1080/08990220050020580.
Crossref | PubMed | ISI | Google Scholar - . Temporal frequency of whisker movement. I. Representations in brainstem and thalamus. J Neurophysiol 86: 339–353, 2001. doi:10.1152/jn.2001.86.1.339.
Link | ISI | Google Scholar - . Movement initiation signals in mouse whisker motor cortex. Neuron 92: 1368–1382, 2016. doi:10.1016/j.neuron.2016.12.001.
Crossref | PubMed | ISI | Google Scholar - . Parallel pathways from motor and somatosensory cortex for controlling whisker movements in mice. Eur J Neurosci 41: 354–367, 2015. doi:10.1111/ejn.12800.
Crossref | PubMed | ISI | Google Scholar - . Encoding of vibrissal active touch. Neuron 40: 621–630, 2003. doi:10.1016/S0896-6273(03)00671-8.
Crossref | PubMed | ISI | Google Scholar - . Membrane currents in visually identified motoneurones of neonatal rat spinal cord. J Physiol 423: 27–46, 1990. doi:10.1113/jphysiol.1990.sp018009.
Crossref | PubMed | ISI | Google Scholar - . New modules are added to vibrissal premotor circuitry with the emergence of exploratory whisking. Neuron 77: 346–360, 2013. doi:10.1016/j.neuron.2012.11.010.
Crossref | PubMed | ISI | Google Scholar - . Projections from the rostral parvocellular reticular formation to pontine and medullary nuclei in the rat: involvement in autonomic regulation and orofacial motor control. Neuroscience 40: 735–758, 1991. doi:10.1016/0306-4522(91)90009-D.
Crossref | PubMed | ISI | Google Scholar - . Cholinergic modulation of vibrissal receptive fields in trigeminal nuclei. J Neurosci 25: 9135–9143, 2005. doi:10.1523/JNEUROSCI.3073-05.2005.
Crossref | PubMed | ISI | Google Scholar - . Synaptic organization of the tectal-facial pathways in the cat. I. Synaptic potentials following collicular stimulation. J Neurophysiol 60: 769–797, 1988. doi:10.1152/jn.1988.60.2.769.
Link | ISI | Google Scholar - . Tactile object localization by anticipatory whisker motion. J Neurophysiol 113: 620–632, 2015. doi:10.1152/jn.00241.2014.
Link | ISI | Google Scholar - . Neuronal activity in rat barrel cortex underlying texture discrimination. PLoS Biol 5:
e305 , 2007. doi:10.1371/journal.pbio.0050305.
Crossref | PubMed | ISI | Google Scholar - . Analysis of sniffing of the albino rat. Behaviour 22: 223–244, 1964. doi:10.1163/156853964X00030.
Crossref | Google Scholar - . Deflection of a vibrissa leads to a gradient of strain across mechanoreceptors in a mystacial follicle. J Neurophysiol 114: 138–145, 2015. doi:10.1152/jn.00179.2015.
Link | ISI | Google Scholar - . Facial morphology and vibrissal movement in the golden hamster. J Morphol 183: 199–217, 1985. doi:10.1002/jmor.1051830208.
Crossref | PubMed | ISI | Google Scholar - . Somatotopic organization of corticospinal and corticotrigeminal neurons in the rat. Neuroscience 4: 65–78, 1979. doi:10.1016/0306-4522(79)90218-5.
Crossref | PubMed | ISI | Google Scholar - . Corticofugal influences in the rat on responses of neurons in the trigeminal nucleus interpolaris to mechanical stimulation. Neurosci Lett 36: 43–48, 1983. doi:10.1016/0304-3940(83)90483-4.
Crossref | PubMed | ISI | Google Scholar - . Glycine transporters are differentially expressed among CNS cells. J Neurosci 15: 3952–3969, 1995. doi:10.1523/JNEUROSCI.15-05-03952.1995.
Crossref | PubMed | ISI | Google Scholar - . Trigemino-reticulo-facial and trigemino-reticulo-hypoglossal pathways in the rat. J Comp Neurol 429: 80–93, 2001. doi:10.1002/1096-9861(20000101)429:1<80:AID-CNE7>3.0.CO;2-L.
Crossref | PubMed | ISI | Google Scholar - . Coding of somatic sensory input by vibrissae neurons in the rat’s trigeminal ganglion. Brain Res 12: 138–156, 1969. doi:10.1016/0006-8993(69)90061-4.
Crossref | PubMed | ISI | Google Scholar
