The rodent whisker-to-barrel pathway constitutes a major model system for studying experience-dependent brain development. Yet little is known about responses of neurons to whisker stimulation in young animals. Response properties of trigeminal ganglion (NV) neurons in 2-, 3-, and 4-week-old and adult rats were examined using extracellular single-unit recordings and controlled whisker stimuli. We found that the receptive field size of NV neurons is mature in 2-week-old animals while response latencies, magnitudes, and angular tuning continue to develop between 2 weeks of age and adulthood. At the earliest time recorded, NV neurons respond to stimulation of only one whisker and can be characterized as slowly or rapidly adapting (SA, RA). The proportion of SA and RA neurons remains constant during development. Consistent with known on-going myelination of NV axons, response latencies decrease with age, becoming adult-like during the third and fourth postnatal weeks for RA and SA neurons, respectively. Unexpectedly, we found that evoked response magnitudes increase several-fold during development becoming adult-like only during the fourth postnatal week. In addition, RA neurons become less selective for whisker deflection angle with age. Maturation of response magnitude and angular tuning is consistent with developmental changes in the mechanical properties of the whisker, the whisker follicle, and the surrounding tissues. The findings indicate that whisker-derived tactile inputs mature during the first postnatal month when whisker-related cortical circuits are susceptible to long-term modification by sensory experience. Thus normal developmental changes in sensory input may influence functional development of cortical circuits.
The whisker-to-barrel pathway in the rat somatosensory system undergoes substantial postnatal development, and sensory experience during this time impacts its structural and functional organization. The pathway receives sensory input from the array of whiskers on the animal's mystacial pad (Woolsey and Van der Loos 1970). Whiskers are specialized sinus hairs and are embedded in richly vascularized follicles in the skin (Ebara et al. 2002; Rice et al. 1986). Each follicle is innervated by 100–150 large myelinated axons (Lee and Woolsey 1975), which originate in the trigeminal ganglion (NV). Each axon innervates only one follicle, resulting in single-whisker receptive fields of NV neurons (Dorfl 1985; Gibson and Welker 1983a). Responses of NV neurons to whisker movements have been well characterized in adult animals (Dykes 1975; Gibson and Welker 1983b; Gottschaldt et al. 1973; Lichtenstein et al. 1990; Shoykhet et al. 2000). In contrast, limited knowledge exists regarding the development of response properties in NV neurons during the first postnatal month when whisker-derived sensory experience has a long-term effect on the function of central neuronal circuits (Simons and Land 1987).
An adult-like pattern of segregated whisker follicle innervation is established by embryonic day 16 in the rat (Rhoades et al. 1990). NV primary afferent neurons at this age respond to mechanical stimulation of a single follicle (Chiaia et al. 1993). In rodent skin, several subtypes of cutaneous nerve endings are physiologically distinct at birth, including those that respond in a tonic [slowly adapting (SA)] or phasic [rapidly adapting (RA)] manner to sustained stimulation (Fitzgerald 1987; Woodbury et al. 2001; Woodbury and Koerber 2003). Myelination of NV primary afferent axons, on the other hand, is known to continue for several weeks postnatally (Cabanes et al. 2002).
Detailed response properties of NV neurons in developing animals remain largely unknown as does the time course of their maturation. We therefore recorded responses of NV neurons to controlled whisker deflections in 2-, 3-, and 4-wk-old and adult rats. Consistent with previous morphological findings, the proportions of SA and RA neurons were similar across age groups, and response latencies decreased, indicative of ongoing myelination. We observed unexpectedly a substantial increase in evoked response magnitudes in both cell types. We also found that RA neurons became slightly less selective for the angular direction of whisker movement with age. Maturation of response properties among NV neurons likely reflects age-dependent changes in the mechanical properties of the whisker/follicle system. Current concepts of experience-dependent maturation in the whisker/barrel system may need to take into account the changing nature of peripheral input during development.
Sprague-Dawley rats aged P12–14 (n = 4), P19–21 (n = 3), P26–28 (n = 4), and > P65 (n = 4) were used in the experiments (P0 = day of birth). These groups will be referred to as P14, P21, P28, and P65, respectively. Male and female pups were from litters born on a specified date (Harlan Sprague-Dawley, Indianapolis, IN) and were housed with their mother. Data from adult female rats (HillTop) were obtained during a separate study (Minnery and Simons 2003). All procedures were approved by the Institutional Animal Care and Use Committee.
Surgery and physiologic maintenance
Details of the surgical procedure are available as supplementary information at http://www.neurobio.pitt.edu/barrels/. These procedures are identical to those used for studying responses of thalamic and cortical neurons in young and mature rats, thus facilitating direct comparisons among the data sets (Shoykhet and Simons 2002). Briefly, under halothane anesthesia, the external jugular vein and the femoral artery were catheterized for drug delivery and blood pressure monitoring, respectively. A tracheotomy and a craniotomy were performed. The animal was transferred to a recording table, halothane anesthesia was discontinued, and artificial ventilation was initiated. Neuromuscular blockade was induced and maintained with pancuronium bromide to prevent spontaneous movement of the whiskers. Thereafter, the animal was maintained in a lightly narcotized state via continuous infusion of a synthetic opiate, fentanyl. The animal's physiologic state was monitored using mean arterial pressure, peak end inspiratory pressure, pulse rate, and glabrous skin perfusion.
Recordings, stimulus control, and data acquisition
Extracellular recordings from NV neurons were obtained using stainless steel microelectrodes (4–6 MΩ at 1 kHz). Single units were identified on the basis of spike amplitude and waveform criteria and digitized using a time/amplitude discriminator. Spike times were collected with 100-μs resolution during a 500-ms time period symmetrically bracketing a 200-ms ramp-and-hold stimulus. The whisker activating the unit, the principal whisker (PW), was identified with a hand-held probe. The PW was then deflected with a piezoelectric stimulator (Simons 1983) at 3-s intervals (2 s in P65 rats) in eight angular directions in 45° increments (0° = caudal, 90° = dorsal). Each deflection was repeated 10 times, and all 80 deflections were presented in random sequence. In P65 rats, the stimulator was attached 10 mm away from the face, delivering a 1-mm deflection at the tip. Because whiskers in young animals are less stiff, the stimulator was attached 5 mm away from the face in P14, P21, and P28 animals, and whisker was deflected 500 μm. The stimuli thus produced equivalent angular deflections of the whisker hair in young and adult animals. In addition, we calibrated stimulator movement using a photodiode circuit to ensure that stimuli in young and adult animals had comparable velocities.
Spike time stamps were converted to peristimulus time histograms (PSTHs) with 1-ms resolution. Individual neuron PSTHs were combined into population PSTHs, and population PSTHs in each age group were used to determine 25-ms-long time windows appropriate for measuring responses to deflection onset (on) and offset (off). on and off response magnitudes are defined as the number of spikes discharged by a neuron during these time windows. Spontaneous firing rates were measured over a 100-ms epoch during the period of data collection prior to whisker deflection. Responses to sustained whisker deflection (plateau) were evaluated during the latter 100 ms of the plateau at the deflection angle evoking the greatest plateau response. A neuron was classified as SA if the plateau response significantly exceeded spontaneous firing (t-test, 1-tail P < 0.025) (Simons and Carvell 1989). All other neurons were classified as RA.
Neurons were included in the analyses if their on responses exceeded spontaneous activity using either one of the following two measures. 1) Average on responses at each deflection angle were examined to identify the deflection angle evoking the largest number of spikes (onmax), and onmax was compared with spontaneous activity (t-test, 1-tailed P < 0.025). 2) on response PSTHs were used to identify a deflection angle evoking the largest number of spikes in a 1-ms bin, and the spike count in that bin was compared with the average of spike counts in 100 bins during spontaneous activity assuming a Poisson distribution of firing rates (1-tailed P < 0.025). The first bin at this deflection angle having a spike count that exceeded spontaneous activity was defined as response latency. In all age groups, 100% of the recorded whisker-responsive NV neurons satisfied at least one of the preceding criteria.
Two other temporal properties of the on response were measured in addition to latency. For each unit, the time to the 50th percentile spike (T50) in on response was computed from a PSTH compiled for PW deflections in all eight angles. T50 is defined as the time required to discharge 50% of spikes in the on response averaged over all deflection angles (Brumberg et al. 1999). It is measured from the 25-ms-long time window beginning with the first PSTH bin in the on response to exceed spontaneous activity (P < 0.025). Initial firing rates were computed as the number of spikes discharged during the first 5 ms of the on response averaged over all deflection angles.
Angular tuning was assessed using two related but somewhat different measures. A neuron was classified into one of seven categories using a statistically based tuning index (TI). TI is calculated as the number of angles (0–7) that evoke an on response significantly smaller than the onmax (t-test, 1-tailed P < 0.05) (see Simons and Carvell 1989). Neurons with a TI of 5–7 will be referred to as well-tuned. As a second measure, we computed the ratio of onmax to the on response magnitude averaged over all eight deflection angles (onmean). Larger ratios denote a greater degree of angular tuning.
One-way ANOVAs were employed for parametric analyses; unless otherwise noted, P values in the text refer to these analyses. Sample means were compared using Student's t-test, and sample variances were compared using Fisher's F test. The χ2 test was employed for nonparametric analysis of discrete variables, e.g., tuning index. The Bonferroni correction was applied to P values when more than three multiple comparisons were performed. Two-tail probability values <0.05 were considered significant unless differently noted. All averages are presented as means ± SE.
Responses of NV neurons to whisker deflections were examined in 2-, 3-, and 4-week-old and adult rats (n =33, 35, 46, and 64, respectively). NV neurons respond to deflection of one, and only one, vibrissa at all ages examined. Population PSTHs show that SA and RA neurons, distinguished on the basis of plateau activity, are present in all age groups (Fig. 1). The relative proportion of the two populations (≈75% SA, ≈25% RA) remains unchanged with age (χ2, P = 0.99). Response properties of both SA and RA neurons, however, continue to develop into the fourth postnatal week.
Development of response magnitudes
Spontaneous firing rates and evoked response magnitudes increase in SA neurons between 2 wk and adulthood. Spontaneous activity (see discussion) increases from 7.3 ± 1.9 to 16 ± 2.1 Hz (Fig. 2C, P = 0.001). Plateau responses of SA neurons to whisker deflection angles evoking the highest plateau firing rates triple in magnitude from 41 ± 4.6 to 121 ± 1.1 Hz (Fig. 2C, P < 10–6). on responses averaged over all deflection angles approximately double in magnitude from 1.5 ± 0.2 to 3.4 ± 0.2 spikes/25 ms becoming adult-like only after P28 (Fig. 2A, P < 10–7; t-test, P28 vs. P65, P = 0.03). Similarly, average off response magnitudes increase twofold from 1.0 ± 0.1 to 2.2 ± 0.2 spikes/25 ms between P14 and P65 (Fig. 2A, P < 10–4). The developmental increases in on and off response magnitudes of SA neurons are monotonic and occur in parallel. Hence, the average off-on ratio in SA neurons remains constant with age (Fig. 2D, P = 0.6).
Evoked responses of RA neurons also increase in size during development (Fig. 2B). Mean on response magnitudes triple from P14 to P65 (0.7 ± 0.1 to 2.4 ± 0.2 spikes/25 ms, P < 0.001) as do average off response magnitudes (0.8 ± 0.2 spikes/25 ms to 2.2 ± 0.3 spikes/25 ms, P = 0.002). These increases are somewhat greater than those observed in SA neurons. Also in contrast to SA neurons, RA neurons display a rather abrupt increase in response magnitude between P21 and P28. Further, on and off response magnitudes of RA neurons are equivalent at all ages examined. off-on ratios of RA neurons are ≈1.0, compared with a value of ≈0.7 for SA neurons. Spontaneous firing rates are lower in RA than SA neurons.
Development of temporal response properties
NV neurons encode salient stimulus features early in their response (Shoykhet et al. 2000). We therefore examined how response latency, time to 50th percentile spike and initial firing rates of NV neurons change between 2 wk and adulthood. Average response latency of SA neurons decreases from 5.5 ± 0.3 to 3.7 ± 0.2 ms (Fig. 3A, P < 10–4), becoming adult-like only by P28 (t-test, P21 vs. P65 P < 0.001; P28 vs. P65 P = 0.052). Mean response latency of RA neurons also decreases (P = 0.047), reaching an adult value by P21 (t-test, P14 vs. P65 P = 0.011; P21 vs. P65 P = 0.23). The variance of latency values decreases slightly with age for SA neurons but remains constant for the RA population (data not shown).
Initial firing rates of SA and RA neurons, measured as the number of spikes discharged during the first 5 ms of the on response, increase between P14 and P65 (Fig. 3B, P < 0.001), becoming adult-like for both neuronal populations by P28 (t-test, P21 vs. P65, P < 0.002; P28 vs. P65, P > 0.05). This increase could reflect a developmental redistribution of spikes toward the early component of the on response or the age-dependent increase in total on response magnitudes (Fig. 2, A and B). We found that the number of spikes discharged during the first 5 ms comprises an invariant 32–34% of the total response in all age groups (data not shown, P = 0.93). Thus the developmental increase in initial firing rates reflects the overall increase in on response magnitudes and not spike redistribution. on responses are, however, more prolonged in older animals as the time to 50th percentile spike in the response increases with age for SA and RA neurons (Fig. 3C, P < 10–4 and P = 0.027, respectively).
Development of angular tuning
Whisker-responsive NV neurons in adult rats are tuned for deflection angle (Lichtenstein et al. 1990). We examined development of angular tuning using two different measures, a TI and the onmax/onmean ratio (see methods). Distributions of TI values for SA neurons are similar at all ages (not shown, χ2, P = 0.9), with the large majority (>80% in all age groups) being well-tuned. TI values for RA neurons are also age-invariant (χ2, P = 0.8). Well-tuned cells constitute a smaller proportion of the RA than the SA population in young animals as well as in adults (<50%). In contrast to a previous report from this laboratory (Lichtenstein et al. 1990), no bias toward upward deflections was detected in RA neurons in any age group perhaps due to small sample sizes.
onmax/onmean ratios decrease with age for RA but not SA neurons (Fig. 4, B and C, P < 10–4 and P = 0.24, respectively). In RA neurons, onmax increases approximately twofold, whereas responses at minimally effective angles (onmin) increase ∼10-fold (see Fig. 4B). In SA neurons, increases in onmax and onmin are proportionate (Fig. 4C). SA neurons thus maintain an adult-like degree of angular tuning between P14 and P65, whereas RA neurons become less tuned with age.
We recorded responses of NV neurons to controlled whisker deflections in 2-, 3-, and 4-wk-old and adult rats under essentially identical experimental conditions. NV neurons respond to the deflection of one, and only one, whisker at all ages studied. This is consistent with previous anatomic and physiologic evidence that individual NV axons innervate a single whisker follicle even in prenatal animals (Chiaia et al. 1993; Rhoades et al. 1990).
Primary afferent neurons can be classified as SA or RA at the earliest time examined. The whisker-responsive population of NV neurons in all age groups consists of 75% SA and 25% RA cells; these values are identical to those obtained in several earlier studies in adult rats (Lichtenstein et al. 1990; Shoykhet et al. 2000). SA and RA neurons are thought to differ in their receptor endings within the whisker follicle with SA neurons having Merkel cell endings (MC) and RA neurons having lanceolate endings (Gottschaldt et al. 1973, Gottschaldt and Vahle-Hinz 1981; Lichtenstein et al. 1990; Rice et al. 1986). The age-invariant proportions of SA and RA populations are consistent with the presence of distinct cutaneous receptor subtypes at birth in rodents (Fitzgerald 1987; Woodbury et al. 2001, Woodbury and Koerber 2003).
Development of response magnitudes
Evoked response magnitudes of SA and RA neurons in NV increase substantially between P14 and P65. Similarly, responses of cutaneous primary afferent neurons to mechanical stimulation increase in magnitude between P0 and P14 (Fitzgerald 1987) and between P14–16 and adulthood (Koltzenburg et al. 1997). This suggests that common mechanisms may be responsible for changes in responses of NV and cutaneous primary afferent neurons prior and subsequent to P14. Primary afferent neurons have largely mature action potential properties and ion channel composition by the third postnatal week in rodents (Benn et al. 2001; Cabanes et al. 2002; Felts et al. 1997), suggesting that developmental changes in neuronal excitability cannot account for the age-dependent increase in response magnitudes. It is likely, however, that mechanical properties of the whisker/follicle complex change during development as whisker hairs, follicles, and surrounding tissue become stiffer and less compliant with age. In addition, whisker follicles prominently contain blood sinuses (Ebara et al. 2002; Rice et al. 1986). Systemic blood pressure increases approximately threefold in the rat between P14 and P65 (see Supplementary information), perhaps elevating turgor within the whisker/follicle complex. Taken together, stiffer whisker hairs, less compliant tissues, and increased turgor probably result in more efficient mechanical transmission of deflection energy to receptor endings in older animals. Hence, evoked response magnitudes increase with age.
Greater mechanical stiffness also likely accounts for the age-dependent increase in spontaneous firing rates of SA neurons. The lack of muscle tone under conditions of neuromuscular blockade often leads to a failure of the whisker/follicle complex to return to a “neutral” position (Minnery and Simons 2003). This mechanical hysteresis can result in high rates of “spontaneous” firing among SA neurons, which can be abolished by appropriate repositioning of the whisker (see Minnery and Simons 2003). Stiffer and heavier whiskers in older animals are likely to exert more pressure onto the MC endings than less rigid and lighter whiskers in younger rats, thus resulting in an apparent increase in spontaneous firing rates during development.
Development of response timing
We found that response latencies of NV neurons decrease between P14 and P28. This decrease almost certainly reflects ongoing myelination of primary afferent axons. A recent study in mouse NV neurons showed that conduction velocities of myelinated fibers increase approximately fourfold between P0–7 and P21–50 (Cabanes et al. 2002). We also found that the variance of observed latency values and the proportion of evoked spikes discharged during the first 5 ms of the on response remain constant in NV neurons between P14 and P65. In contrast, response latencies of thalamic “barreloid” neurons become more narrowly distributed during this time, and spikes evoked by deflection onset redistribute toward the early phase of the thalamic on response (Shoykhet and Simons 2002). Together, the findings indicate that, both peripherally and centrally, the whisker-to-barrel pathway processes information differently throughout an extended period of postnatal development.
Development of angular tuning
RA neurons become less selective for deflection angle between P14 and P65, whereas SA neurons are as tuned in 2-wk-old animals as in adults. This conclusion is based on the finding that responses at minimally effective angles increase disproportionately in RA but not SA neurons. RA neurons, likely associated with lanceolate endings, are thought to fire when the mesenchymal sheath within the follicle is stretched (Gottschaldt et al. 1973; Rice et al. 1986). With the deflection amplitudes used in our studies, some amount of stretch is probably transmitted to the lanceolate ending even when the whisker is deflected in nonoptimal directions, resulting in a weak but detectable neural response (Lichtenstein et al. 1990). If, as suggested in the preceding text, the whisker/follicle complex becomes stiffer with age, small changes in force produced by deflections at nonoptimal directions would be more likely to be transmitted to, and transduced by, the lanceolate ending in older animals.
SA neurons, associated with MC endings, are thought to be activated by compression of receptor endings between the hair shaft and the rigid glassy membrane (Gottschaldt and Vahle-Hinz 1981). Thus whisker deflections that move the whisker shaft away from the nerve ending, i.e., in nonoptimal directions, result in a “null” rather than a small but detectable response. Indeed, SA neurons are more likely to have null deflection angles than RA neurons (Lichtenstein et al. 1990). Null responses are likely to remain null regardless of developmental changes in tissue stiffness.
Implications for higher-order circuits
Peripheral input early in life impacts functional development of thalamocortical circuitry in the rat somatosensory system (Simons and Land 1987). We unexpectedly found that whisker movement-evoked responses increase twofold for SA neurons and threefold for RA neurons between P14 and P65 and become adult-like only during the fourth postnatal week. Central circuits in the rat somatosensory system thus receive immature and continuously changing neural inputs for most of the animal's prepubertal life. A comprehensive understanding of the postnatal development of the whisker/barrel system must therefore include the changing nature of the peripheral input and the extended period during which the system presumably adapts to it.
This work was supported by National Institute of Neurological Disorders and Stroke Grant NS-19950.
We are grateful to Dr. C. Jeffrey Woodbury for helpful comments on the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 by the American Physiological Society