| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Psychology, University of Colorado, Boulder, Colorado
Submitted 30 September 2005; accepted in final form 29 December 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
). The vibrissa system may be separated into two subdivisions, with 25 (five vertical arcs of five vibrissae) large and caudally positioned macrovibrissae that range in length from one to several centimeters, and more numerous and rostrally positioned microvibrissae with lengths on the order of several millimeters. Central representation of the macro- and microvibrissae occupies a majority of somatosensory cortex and is characterized by an orderly arrangement of discrete vertically oriented cellular aggregates or cortical columns, often referred to as "barrels" (Woolsey and Van der Loos 1970). The posterior medial barrel subfield (PMBSF), the most thoroughly studied, reveals a clear somatotopic organization in approximate register with the macrovibrissae on the contralateral mystacial pad (Chapin and Lin 1990). A group of microbarrels, rostral to the PMBSF, receive information from the microvibrissae, although their somatotopy has not been explored.
Based on behavioral evidence, it has been suggested that the macro- and microvibrissa systems may also be distinguished functionally, with the macrovibrissae serving as distance detectors, providing head-centered spatial information, and the microvibrissae reserved as an object-recognition sense organ (Brecht et al. 1997). Although the function of microvibrissae has not been further documented, the hypothesized role of macrovibrissae in spatial sampling is supported by a number of studies indicating that this system is critical for tasks such as gap detection, distance discrimination, and head-centered orientation (Harris et al. 1999; Hutson and Masterson 1986; Krupa et al. 2001; Sachdev et al. 2000; Schiffman et al. 1970; Shuler et al. 2002; Vincent 1912).
Yet, other behavioral and physiological studies suggest that the macrovibrissae may also provide information about object features. During exploration, rats move ("whisk") the macrovibrissae in a caudorostral direction at approximately 10 Hz, making repeated object contact on each rostral extension (Carvell and Simons 1990). It has been proposed that sensory capacities provided by whisking the macrovibrissae as a unified sensory array may be analogous to active touch of the primate fingertip (Carvell and Simons 1990; Simons 1995). There is compelling evidence that movement of the macrovibrissae across an object’s surface provides essential information about fine texture (Andermann et al. 2004; Arabzadeh et al. 2004, 2005; Carvell and Simons 1990; Guilmage-Robles et al. 1989; Neimark et al. 2003). However, whisking also establishes a rapid and asynchronous vibrissa contact (Sachdev et al. 2001), presumably producing momentary and complex spatiotemporal patterns of afferent input to the PMBSF that may uniquely reflect spatial features of an object. Single-unit studies of single and paired whisker stimulation indicate a temporal consistency between vibrissa contact and cortical response (Ego-Stengel et al. 2005; Sachdev et al. 2001; Simons 1985; Simons and Carvell 1989), introducing the possibility that more complex spatiotemporal afferent patterns evoked by multiple vibrissa contact in the behaving animal could be accurately represented by patterns of electrical activity in larger neural networks of PMBSF. If true, then temporal dynamics of the population response could provide information about not just distance and texture, but also object orientation and shape. However, it is not clear whether the temporal fidelity between vibrissa contact and cortical response demonstrated in unit studies is preserved in the large populations of cells constituting the PMBSF, particularly in the presence of inhibitory and excitatory interactions between barrels when large groups of vibrissae are engaged.
To explore this issue, we used high-resolution arrays of epipially placed electrodes to measure population field potentials from the entire PMBSF while stimulating either one or five arcs of macrovibrissae using straight and curved edges moved at rostrocaudal velocities mimicking those expected during natural whisking. Our objectives were to: 1) determine what parameters of the somatosensory-evoked potential (SEP) complex (timing, amplitude, and spatial distribution) are influenced by changing object features; 2) determine whether SEP patterns measured from the entire PMBSF are sufficiently stable and unique to accurately classify single presentations of a given stimulus; 3) develop a general method for modeling spatiotemporal SEP patterns that might permit not just classification, but realistic reconstruction, of object features; and 4) determine what the limits this feature estimation may imply about the capacity of the macrovibrissae/barrel system for object recognition.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
All procedures were performed in accordance with University of Colorado Institutional Animal Care and Use Committee guidelines for the humane use of laboratory animals in biological research. Seven adult male Sprague–Dawley rats (300–400 g) were anesthetized to surgical levels using an intraperitoneal injection of urethane, placed on a regulated heating pad, and maintained with subsequent injections of a mixture of xylazine (13 mg/kg) and acepromazine (2 mg/kg) throughout the experiment so that the eye-blink reflex could be barely elicited. A unilateral craniectomy was performed over the right hemisphere extending from bregma to lambda and from the midsagittal sinus lateral to the temporal bone, exposing a wide region of parietotemporal cortex where the dura was reflected. Animals were killed by anesthesia overdose without regaining consciousness at the conclusion of the experiment.
Stimulation
The stimulating apparatus consisted of a smooth 140-mm-diameter cylinder (45 mm high) fitted with a straight wire (70 mm length; 0.8 mm diameter) stimulus mounted to surface of the cylinder at its vertical midpoint and attached at the center with a thumbscrew (Fig. 1A). The cylinder was attached to a laboratory-built vibration-free mount with a play out <10 µm. The stimulus was manually positioned at various angles in relation to upright (0° in the sagittal plane) using calibrated visual markers. Angles typically included ±45, 30, 20, 10, and 5° (Fig. 1B). In two animals, a vertically oriented curved-wire (same thickness) hemicircle (25 mm diameter) stimulus was also used (Fig. 1B, dashed red line). Rotation speeds were controlled with a programmable stepping motor (ServoDyne mixer head; 3–180 rpm; controller model 50003–00), and adjusted so that the speed with which the stimulus passed through the vibrissae was either 0.4 or 0.2 mm/ms in the rostrocaudal direction. These speeds were chosen to best simulate the 10- to 20-ms delays between sequential contact of vibrissae in a row observed in behaving animals during rostral whisking motions (Sachdev et al. 2001), assuming an average 4-mm distance between vibrissae in a row. In three animals, the 25 macrovibrissae on the left mystacial pad (see Fig. 2A) were trimmed to about 2 cm (the remaining whiskers were clipped). Because the angle at which the vibrissa emerges from the face differs from one whisker to the next, the exact length of each whisker was adjusted so that it made light (causing a caudal displacement of about 5°) and constant contact with the smooth continuously rotating drum. The stimulus was repeatedly swept through the vibrissae in the rostrocaudal direction and each trial of data sampling was triggered by a tab that interrupted an infrared emitter/detector pair. The beam interrupter was positioned on the inside of the rotating drum at a location that triggered data sampling about 50–100 ms before the stimulus made contact with the most rostral vibrissae (Fig. 1A). A total of 85–100 trials of evoked responses were collected in this way before the stimulus was manually repositioned to a new angle and another run performed. In the four remaining animals, procedures were the same except that all vibrissae were clipped except for those of the middle arc (consisting of vibrissae B3, C3, D3, and E3). Vibrissa A3 in this arc was also clipped because it typically was difficult to position adequately to make good contact with the stimulus.
|
|
Evoked potential recording
Epipial maps of the vibrissa-evoked SEP complex were recorded using a flat multichannel electrode array consisting of 64 silver wires arranged in a 8 x 8 grid (tip diam: 100 µm; interelectrode spacing: 500 µm) covering a 3.5 x 3.5-mm area of the cortical surface in a single placement. Light pressure was applied to this flat array so that all of the electrodes made contact with the pial surface. After array placement, surrounding exposed regions of cortex were covered with cotton soaked in artificial cerebrospinal fluid and periodically moistened throughout the experiment to prevent desiccation. Surface field potentials were referenced to a silver ball electrode secured over the contralateral frontal bone, amplified (x1,000), analog filtered (band-pass cutoff = –6 dB at 1 to 3,000 Hz, roll-off = 5 dB/octave), and digitized at 10 kHz.
Data collection and analysis
Samples of 200-ms whisker-evoked responses were recorded, with data from individual trials (n = 85–100) stored digitally for subsequent analysis. Averaged responses were plotted on a template of the PMBSF in approximate register with the surface recording sites. The template was derived from previous histology and was used here for illustrative purposes only. Histological verification of precise electrode positions was not performed in the present study because this was not required for interpretation of results. The recording array was aligned to the PMBSF using single vibrissae-evoked responses (see Fig. 4A).
|
0.01. Results are reported as mean ± SD unless otherwise noted. In animals with a single arc of four vibrissae intact, classification was based on a more general model that reconstructed the actual orientation and shape of a stimulus based on spatiotemporal patterns of the SEP complex it produced. Averaged SEPs (n = 100) from transient (0.1 ms; 0.5-mm deflection) stimulation (using a laboratory-built solenoid and 4-cm stick) of each individual vibrissa were first computed. Normalized maps (8 x 8 electrodes) of root-mean-squared (RMS) amplitude of the averaged SEP were used to represent the spatial distribution of each vibrissa’s contribution to SEPs evoked when all four vibrissae were subsequently contacted by the stimulus (Fig. 4A). RMS was calculated at each electrode as the average sum of squared values for sample points covering the P1/N1 sharp wave. The multivibrissa SEP complex for a given stimulation condition was modeled as a weighted combination of the four single vibrissa response patterns for each of 200 sampled time points. This resulted in four time series consisting of 200 regression weights used to model the time course of each vibrissa’s activation during the SEP complex. Regression weights computed in this way for 0° stimulus were compared with similar regression weights modeling SEPs from other conditions, again by recording time lags of the maximum cross-correlation function. Thus SEPs (either averaged or single trial) evoked by each orientation of the stimulus were represented by four time lags, reflecting the timing of when each vibrissa contacted the stimulus relative to the timing pattern evoked at 0°. Because the velocity of the stimulus was known, and the dorsoventral positions of each vibrissa where they contacted the stimulus were recorded for each animal, timing differences between each vibrissa response could be converted to relative distances along the drum and thus used to directly reconstruct the orientation (and shape) of the stimulus. Modeling accuracy was quantified as average error (in degrees) and tested for significant differences between means using multiple comparisons with significance set to P < 0.01.
Single-trial SEPs from animals with four vibrissa stimulation were also classified with a template-matching procedure that used spatial patterns of response amplitude as opposed to time lags as the classification criteria. SEPs were analyzed separately for low-, middle-, and high-frequency bandwidths (1–10, 10–200, and 200–1,000 Hz, respectively). Within each bandwidth, amplitude template maps were computed from the normalized RMS averaged across trials for each stimulus orientation. RMS was calculated at each electrode as the average sum of squared values for sample points covering the filtered SEP waveform. Similarly computed RMS amplitude maps of single trials were assigned to a given stimulus condition based on their minimum least-squared fit to the templates. Classification accuracy was quantified as percentage correct compared with random classification, and evaluated using unpaired t-test with significance set to P 0.01.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
). In this and subsequent illustrations, a standard template of the PMBSF was scaled and appropriately oriented to represent the approximate relationship between the barrel field and recording electrodes in a given recording position (Fig. 2, C and D, light traces). However, this template was used only for graphical purposes because histology was not performed in the present study. Averaged SEPs, evoked by repeatedly brushing an upright stimulus oriented perpendicular to the vibrissa rows (0° orientation) at a velocity of 0.4 mm/ms, formed a distribution of responses centered on and largely constrained to the PMBSF (Fig. 2C). SEP morphology was similar at all recording sites and consisted of a single positive/negative sharp wave whose amplitude peaks are labeled P1 and N1 to reflect their polarity and sequence of occurrence (Fig. 2E). This sweep speed established a nearly 10-ms delay between sequential contact of vibrissae within a row on the rostrocaudal axis. At a slower sweep speed of 0.2 mm/ms, establishing a 20-ms delay between sequential vibrissa contact, the SEP morphology was substantially altered. At rostral electrode sites (Fig. 2, Da and Fa), the SEP began with a large P1/N1 sharp wave reflecting the response to initial contact of the stimulus with the rostral vibrissae. This was followed by smaller waves at 20-ms intervals, coincident with contact of progressively more caudal vibrissae. Thus the rostral barrels appeared to respond both to contact of their principal vibrissae and to more caudal vibrissae. SEPs in the middle of the PMBSF reflected a similar horizontal integration, with similar responses to their principal vibrissae and to those at rostral and caudal sites (Fig. 2, Db-c and Fb-c). SEPs at the most caudal sites (Fig. 2, Dd and Fd) consisted of a large P1/N1 sharp wave that was delayed by 50–60 ms to the initial rostral response and was dominated by contact with the principal caudal vibrissae with only little responsiveness to previously contacted rostral vibrissae.
SEPs evoked by 0° stimulus were used as reference for comparison of relative latency shifts in SEPs corresponding to other orientations. For example, when the stimulus was oriented at +10°, it struck ventral vibrissae in row E slightly earlier and dorsal vibrissae in row A slightly later than when at 0°. This difference in timing was reflected in the relative latencies of the SEPs. Responses in row A of the PMBSF were later when the stimulus was oriented at +10° (Fig. 3B, left blue trace; single-electrode site) as opposed to 0° (Fig. 3B, left black trace; single-electrode site). In contrast, SEPs evoked by +10° stimulus were earlier in row E (Fig. 3B, right blue trace) compared with the 0° response (Fig. 3B, right black trace). An opposite pattern of latency shifts was produced by stimulus at –10° (Fig. 3B, red traces). To quantify relative latency shifts at each electrode site, maximum normalized cross-correlation functions were computed (±50-ms lags) between 0° responses and responses to other orientations (shown in Fig. 3C for just –10, 0, and +10° at an electrode site in row A and one in row E). The absolute values of relative latency shifts at each electrode were then averaged across all stimulus orientations (in this example, –45, –30, –20, –10, –5, 0, +5, +10, +20, +30, and +45°) to obtain a composite map reflecting their spatial distribution. At sweep speeds of 0.4 mm/ms, the composite map indicated maximum latency shifts across all orientations in the dorsal and ventral regions of the PMBSF (Fig. 3A, left map). Latency shifts were minimal at intervening regions of the PMBSF corresponding to vibrissa rows C and D. These vibrissae were positioned closest to the axis of rotation of the stimulus and thus would be expected to have a minimal change in the timing of vibrissa contact at different orientations. Although the waveform morphology of SEPs evoked by slower sweep speeds (0.2 mm/ms) was more complex than that of the faster speed (see Fig. 2F), the composite latency map based on maximum cross-correlation functions was nearly identical (Fig. 3A, right map).
|
Results presented thus far were from three animals with 25 macrovibrissae contacting the stimulus. Yet, they indicated that temporal patterns of the SEP were sufficient for classifying stimulus orientation and that this classification was dominated by latency differences between the rows (thus within the arcs) of barrels in the PMBSF (i.e., latency differences between rows A–E). To evaluate how well a single arc of vibrissae could perform, all vibrissae except B3–E3 were trimmed for subsequent experiments in four remaining animals (A3 was typically oriented too dorsal to make secure contact with the stimulus). In this preparation, a sweep speed of 0.4 mm/ms with the stimulus at 0° produced SEPs of largest amplitude over the principal barrels (Fig. 4 B, darkened barrels), with some rostral and caudal spread within the PMBSF (Fig. 4B, black traces). It was assumed that this pattern was dominated by activity of barrels B3–E3. Therefore to simplify the analysis, a model was constructed in which the spatial distribution of SEP amplitude (calculated at each electrode as the RMS for sample points covering the P1/N1 sharp wave) evoked by single vibrissa stimulation of B3–E3, was used to represent the contribution of each vibrissa throughout the recording array during multivibrissa stimulation. Thus each time point of the multivibrissa SEP complex was modeled as a weighted combination of only four single vibrissa patterns (Fig. 4A). The model (Fig. 4B, green traces) closely fit the multivibrissa response complex and accounted for >90% of the system variance across all stimulus conditions and animals (92 ± 6.5%). Regression weights associated with each individual vibrissa map were used to represent the time course of activity in their respective barrels during a given stimulation condition (Fig. 4C, left traces). For example, regression weights associated with the 0° stimulus (Fig. 4C, left black traces) when multiplied times their respective single-vibrissa RMS potential maps (Fig. 4A) and summed, produced a model SEP complex for this condition (Fig. 4B, green traces). Regression weights computed in the same way for +10 and –10° stimuli (Fig. 4C, left blue and red traces, respectively) had latency shifts of the P1/N1 wave preceding and following the 0° weights, with most pronounced shifts for the E3 vibrissa positioned at the greatest distance below the stimulus rotation axis, almost no shifts for the C3 vibrissa nearest the rotation axis, and a reversal of latency shift for the B3 vibrissa positioned just above the rotation axis. Latency shifts for the four respective regression weights were quantified by computing their normalized cross-correlation functions with those from 0° stimulation (Fig. 4 C, right traces).
Given the known velocity of the stimulus passing through the vibrissa in the rostrocaudal direction, latency shifts for each vibrissa were converted to relative positions along the rostrocaudal axis where the vibrissa were contacted by stimuli of different orientations. This information, combined with measurements of the locations of the vibrissae on the dorsoventral axis, permitted a direct estimate of the orientation of the stimulus during a particular stimulation condition. Figure 4D depicts actual (dashed lines) and reconstructed (solid lines) orientations for +10, 0, and –10° stimuli (blue, black, and red lines, respectively) and results for the other stimulus orientations (solid gray lines). Stimulus orientations predicted from temporal SEP responses differed significantly between conditions (multiple comparison test of means; P < 0.01) during sweep speeds of both 0.4 mm/ms (Fig. 5A) and 0.2 mm/ms (Fig. 5B), with an average error of 2.4 ± 2.6 and 2.8 ± 2.7°, respectively. An advantage of this modeling method was that it allowed us to directly reconstruct any stimulus, without matching templates derived from known stimuli (except that of the 0° stimulus used to calibrate all relative latency shifts). Thus not only the orientation of linear edges of different orientations could be reconstructed directly from the data, but the curvature of convex and concave stimuli could be estimated as well, as shown in two animals for sweep speeds of 0.4 mm/ms (Fig. 6A) and 0.2 mm/ms (Fig. 6B).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Studies of whisking dynamics in behaving animals indicate that on each forward extension, the vibrissae make sequential contact in the caudorostral direction with delays between each vibrissa of a row ranging from 10 to 20 ms (Sachdev et al. 2001). In this study, we assumed that with an average intervibrissa distance along a given row of 4 mm, sweep speeds of 0.4–0.2 mm/ms would approximate these naturally occurring delays. However, it is conceivable that during exploratory whisking, rats may occasionally group their vibrissae even closer together during forward extension. This possibility, combined with the fact that the rat and/or the object may be moving, suggests even shorter intervibrissa delays may occur. Although faster sweep speeds were not tested here, stimulus classification should remain accurate even with delays in the millisecond or submillisecond range. This conclusion is based on the fact that discriminations were accurate even for stimulus angles of ±5° from upright, a circumstance that produced delays between barrels of an arc that were <1 ms (see Fig. 4D, vibrissae D3–B3).
These data indicate that rapid temporal patterns of SEPs in the PMBSF constitute the essential parameter in our model for identifying object features. One might expect that synaptic interactions between the barrels of an arc would systematically influence the spatial pattern of SEP amplitude as well. Yet, amplitude maps of the lowest-frequency components (1–10 Hz) of the SEP appear unchanged by stimulus orientation (Fig. 7B, bottom row) and perform no better than chance when used as templates for single-trial classification. Similar maps of SEP amplitude in the 10- to 200-Hz frequency range, emphasizing the P1/N1 sharp wave, are clearly influenced by stimulus orientation (Fig. 7B, middle row) but also perform poorly for single-trial classification. We recently demonstrated that high-frequency (200–1,000 Hz) FO are distinctly influenced by phase-sensitive interactions between barrels in the submillisecond range and may provide a mechanism for high-speed coincidence detection in the PMBSF (Barth 2003). Indeed, spatial maps of FO amplitude measured here are changed considerably by stimulus orientation and typically reveal pairs of amplitude maxima with intervening minima suggestive of phase-sensitive interactions between barrels (Fig. 7B, top row). When these amplitude patterns are used as templates for classification of single trials, however, their performance was no better than that of the sharp waves.
The failure of sharp-wave, and particularly FO, amplitude patterns to accurately classify single trials may indicate that they do not directly reflect the mechanism for rapid spatiotemporal integration in the PMBSF. Alternatively, our negative results may indicate a sensitivity problem in single-trial epipial SEP recordings, particularly for very low amplitude FO, and require further examination with microelectrode unit recording. Laminar recordings have shown that the SEP recorded at the cortical surface is produced by synchronized postsynaptic currents in the aligned apical dendrites of supragranular pyramidal cells (generating the P1 arising from proximal depolarization), followed by distal depolarization of apical dendrites of both supra- and infragranular pyramidal cells extending near the cortical surface (generating the N1) (Di et al. 1990; Kulics and Cauller 1986, 1989). The laminar potential pattern conforms to a vertical current dipole for the supra- and infragranular pyramidal cell groups, with polarity reversals in the upper and middle cortical layers, respectively. Although field potentials are volume conducted, the strength of the field potential from a current dipole declines with the inverse square of distance from the source (Nuñez 1981). With epipial electrodes, the recording distance from the distal apical dendrites is small (the thickness of the pia), and thus potentials from any distance greater than this are negligible. Evidence for this is the fact that the laminar potential pattern reverses polarity (as expected from a current dipole) at depths as little as 200 µm from the surface for the P1 and at a depth of about 500 µm for the N1 (Di et al. 1990).
An advantage of field potential measures is that they show the central tendency of the response of populations of neurons. A disadvantage is that epipial field potentials do not discriminate between supra- and subthreshold postsynaptic potentials, and their spatial resolution is not sufficient to discriminate between activation of subpopulations of cells within a barrel. Because they reflect central tendency, it is reasonable to anticipate from field potentials the general response properties of units within a population. For example, the P1 and rising phase of the N1 are closely associated with excitation and cell firing, and the falling phase of the N1 and subsequent slow waves are associated with fast and slow inhibition and with cessation of firing in many units (Purpura 1959; Steriade 1984). Conversely, because unit recording is much more sensitive to distinct responses of individual cells, it is not easy to anticipate what field potentials they will be associated with. Unit studies have mainly examined multiwhisker integration between barrels along the rows of the PMBSF (Carvell and Simons 1988; Shimegi et al. 1999, 2000; Simons 1985). However, there have been several reports of supralinear unit responses when barrels along an arc are sequentially activated with interstimulus intervals in the millisecond range similar to timing of barrel activation reported here (Ego-Stengel et al. 2005; Ghazanfar and Nicolelis 1997). Supralinear responses within the barrel arcs may well reflect integration of object features at the cellular level that cannot be reliably recorded in the epipial field potential.
A single arc of four vibrissae is sufficient to reconstruct not only the orientation of straight edges, but also the curvature of large contours. In this context, an analogy between the macrovibrissa system and peripheral vision may be appropriate. The macrovibrissae provide a coarse analysis of an objects’ features (Harvey et al. 2001; Polley et al. 2005), perhaps for more detailed examination by the rostral microvibrissae (Brecht et al. 1997), and because of their length, they provide an early warning of object presence for head-centered orientation. Nonetheless, it is conceivable that during active "discriminative whisking," the macrovibrissae provide even more detailed information than this (Harvey et al. 2001). Sensitivity to orientation changes as small as 5° may actually represent a lower bound on the potential precision of the barrel system given that our model reconstructions were based on epipial field potentials that are no doubt insensitive to more precise temporal patterns recordable at the unit level. We have also ignored the fact that the vibrissae possess a direction sensitivity that could increase their capacity to discriminate form, not based on timing but on directionally sensitive input from each vibrissa (Bruno et al. 2003; Lee and Simons 2004; Simons and Carvell 1989; Wilent and Contreras 2005). Furthermore, it should be noted that we held constant the distance between the vibrissae on the dorsoventral axis, a limitation not imposed on the behaving animal. The ability to resolve curves of differing spatial frequency (i.e., number of curves per millimeter or curvature sharpness) must depend on the spatial sampling frequency of the vibrissae. This is directly analogous to time series analysis, where time-varying signals must be sampled at half their shortest period to be resolved. The present experiment held the vibrissae at average dorsoventral distances of about 5 mm, imposing a spatial Nyquist frequency of 0.1 curves/mm, for lack of a better term. This spatial resolution could be improved by decreasing the distance between the vibrissae, an act that may be within the unanesthetized rat’s behavioral repertoire. Rats have been shown to alter their whisking movement strategies during discriminative task acquisition. Parameters changed include frequency, velocity, amplitude, duration, and the amount of whisking (Harvey et al. 2001). The present results suggest that rats may also modify the vertical distance between vibrissae, depending on the required spatial resolution of a task.
Our data indicate that rapid temporal interactions between barrels in a single arc are useful for feature extraction in our model and may represent a distinct processing module in the PMBSF. This result is consistent with the expectation that changes in stimulus orientation (and shape) should produce maximum timing differences between vibrissae in an arc versus row during caudorostral whisking. Given that a single vibrissa arc can be used for stimulus reconstructions described here, one might wonder why there are five such arcs, with corresponding barrels preferentially interconnected along the rows of the PMBSF (Bernardo et al. 1990; Hoeflinger et al. 1995). Indeed, SEPs evoked by 25 vibrissa stimulation suggest substantial intrarow integration, with a majority of barrels sensitive to sequential activation of the arcs during a rostrocaudal sweep (Fig. 2F). One possibility is that the aggregate barrels in the PMBSF may function as a spatial frequency analyzer. It has been noted that vibrissa length increases exponentially from the rostral to caudal arcs (Brecht et al. 1997). If the entire vibrissa array were adjusted to a common intervibrissa angle emerging from the mystacial pad on the vertical axis, systematic differences in length would establish different spatial sampling frequencies for the arcs, with the longest caudal vibrissae sensitive to the lowest frequencies (gradual curvatures), and progressively more rostral and shorter vibrissae sensing higher spatial frequencies (finer detail). In this way, intrarow processing within the PMBSF may integrate a complementary series of spatial frequencies in a single whisk.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: D. S. Barth, Department of Psychology, University of Colorado, Campus Box 345, Boulder, CO 80309-0345 (E-mail: dbarth{at}psych.colorado.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Arabzadeh E, Panzeri S, and Diamond ME. Whisker vibration information carried by rat barrel cortex neurons. J Neurosci 24: 6011–6020, 2004.
Arabzadeh E, Zorzin E, and Diamond ME. Neuronal encoding of texture in the whisker sensory pathway. PLoS Biol 3: e17, 2005.[CrossRef][Medline]
Barth DS. Submillisecond synchronization of fast electrical oscillations in neocortex. J Neurosci 23: 2502–2510, 2003.
Bernardo KL, McCasland JS, Woolsey TA, and Strominger RN. Local intra- and interlaminar connections in mouse barrel cortex. J Comp Neurol 291: 231–255, 1990.[CrossRef][Web of Science][Medline]
Brecht M, Preilowski B, and Merzenich MM. Functional architecture of the mystacial vibrissae. Behav Brain Res 84: 81–97, 1997.[CrossRef][Web of Science][Medline]
Bruno RM, Khatri V, Land PW, and Simons DJ. Thalamocortical angular tuning domains within individual barrels of rat somatosensory cortex. J Neurosci 23: 9565–9574, 2003.
Carvell GE and Simons DJ. Membrane potential changes in rat SmI cortical neurons evoked by controlled stimulation of mystacial vibrissae. Brain Res 10: 186–191, 1988.
Carvell GE and Simons DJ. Biometric analyses of vibrissal tactile discrimination in the rat. J Neurosci 10: 2638–2648, 1990.[Abstract]
Chapin JK and Lin CS. The somatic sensory cortex of the rat. In: The Cerebral Cortex of the Rat, edited by Kolb B and Tees RC. Cambridge, MA: MIT Press, 1990, p. 341–380.
Di S, Baumgartner C, and Barth DS. Laminar analysis of extracellular field potentials in rat vibrissa/barrel cortex. J Neurophysiol 63: 832–840, 1990.
Ego-Stengel V, Mello e Souza T, Jacob V, and Shulz DE. Spatiotemporal characteristics of neuronal sensory integration in the barrel cortex of the rat. J Neurophysiol 93: 1450–1467, 2005.
Ghazanfar AA and Nicolelis MAL. Nonlinear processing of tactile information in the thalamocortical loop. J Neurophysiol 78: 506–510, 1997.
Guilmage-Robles E, Valdivieso C, and Guajardo G. Rats can learn a roughness discrimination using only their vibrissal system. Behav Brain Res 31: 285–289, 1989.[CrossRef][Web of Science][Medline]
Harris JA, Petersen RS, and Diamond ME. Distribution of tactile learning and its neural basis. Proc Natl Acad Sci USA 96: 7587–7591, 1999.
Harvey MA, Bermejo R, and Zeigler HP. Discriminative whisking in the head-fixed rat: optoelectronic monitoring during tactile detection and discrimination tasks. Somatosens Mot Res 18: 211–222, 2001.[CrossRef][Web of Science][Medline]
Hoeflinger BF, Bennett-Clarke CA, Chiaia NL, Killackey HP, and Rhoades RW. Patterning of local intracortical projections within the vibrissae representation of rat primary somatosensory cortex. J Comp Neurol 354: 551–563, 1995.[CrossRef][Web of Science][Medline]
Hutson KA and Masterson RB. The sensory contribution of a single vibrissa’s cortical barrel. J Neurophysiol 56: 1196–1223, 1986.
Krupa DJ, Matell MS, Brisben AJ, Oliveira LM, and Nicolelis MA. Behavioral properties of the trigeminal somatosensory system in rats performing whisker-dependent tactile discriminations. J Neurosci 21: 5752–5763, 2001.
Kulics AT and Cauller LJ. Cerebral cortical somatosensory evoked responses, multiple unit activity and current source-densities: their interrelationships and significance to somatic sensation as revealed by stimulation of the awake monkey’s hand. Exp Brain Res 62: 46–60, 1986.[Web of Science][Medline]
Kulics AT and Cauller LJ. Multielectrode exploration of somatosensory cortex function in the awake monkey. In: Sensory Processing in the Mammalian Brain: Neural Substrates and Experimental Strategies, CNUP Neuroscience Reviews, edited by Lind JS. Oxford, UK: Oxford Univ. Press, 1989, p. 85–115.
Lee SH and Simons DJ. Angular tuning and velocity sensitivity in different neuron classes within layer 4 of rat barrel cortex. J Neurophysiol 91: 223–229, 2004.
Neimark MA, Andermann ML, Hopfield JJ, and Moore CI. Vibrissa resonance as a transduction mechanism for tactile encoding. J Neurosci 23: 6499–6509, 2003.
Nuñez PL. Electric Fields of the Brain: The Neurophysics of EEG. New York: Oxford Univ. Press, 1981.
Polley DB, Rickerta JL, and Frostig RD. Whisker-based discrimination of object orientation determined with a rapid training paradigm. Neurobiol Learn Mem 83: 134–142, 2005.[CrossRef][Web of Science][Medline]
Purpura DP. Nature of electrocortical potentials and synaptic organizations in cerebral and cerebellar cortex. Int Rev Neurobiol 1: 47–163, 1959.[Web of Science][Medline]
Sachdev RN, Sellien H, and Ebner F. Temporal organization of multi-whisker contact in rats. Somatosens Mot Res 18: 91–100, 2001.[CrossRef][Web of Science][Medline]
Sachdev RNS, Egli M, Stonecypher M, Wiley RG, and Ebner FF. Enhancement of cortical plasticity by behavioral training in acetylcholine-depleted adult rats. J Neurophysiol 84: 1971–1981, 2000.
Schiffman HR, Lore R, Passafiume J, and Neeb R. Role of vibrissae for depth perception in the rat (Rattus norvegicus). Anim Behav 18: 290–292, 1970.[CrossRef][Web of Science][Medline]
Shimegi S, Akasaki T, Ichikawa T, and Sato H. Physiological and anatomical organization of multiwhisker response interactions in the barrel cortex of rats. J Neurosci 20: 6241–6248, 2000.
Shimegi S, Ichikawa T, Akasaki T, and Sato H. Temporal characteristics of response integration evoked by multiple whisker stimulations in the barrel cortex of rats. J Neurosci 19: 10164–10175, 1999.
Shuler MG, Krupa DJ, and Nicolelis MAL. Integration of bilateral whisker stimuli in rats: role of the whisker barrel cortices. Cereb Cortex 12: 86–97, 2002.
Simons DJ. Temporal and spatial integration in the rat SI vibrissa cortex. J Neurophysiol 54: 615–635, 1985.
Simons DJ. Neuronal integration in the somatosensory whisker/barrel cortex. In: The Barrel Cortex of Rodents (1st ed.), edited by Jones EG and Diamond IT. New York: Plenum Press, 1995.
Simons DJ and Carvell GE. Thalamocortical response transformation in the rat vibrissa/barrel system. J Neurophysiol 61: 311–330, 1989.
Staba RJ, Ard T, Benison A, and Barth DS. Intracortical pathways mediate nonlinear fast oscillation (>200 Hz) interactions within rat barrel cortex. J Neurophysiol 93: 2934–2939, 2005.
Steriade M. The excitatory–inhibitory response sequence in thalamic and neocortical cells: state-related changes and regulatory systems. In: Dynamic Aspects of Neocortical Function, edited by Edelman GM, Gall WE, and Cowan WM. New York: Wiley, 1984, p. 107–157.
Vincent SB. The function of the vibrissae in the behavior of the white rat. Behav Monogr 1: 1–85, 1912.
Wilent WB and Contreras D. Stimulus-dependent changes in spike threshold enhance feature selectivity in rat barrel cortex neurons. J Neurosci 25: 2983–2991, 2005.
Woolsey TA and Van der Loos H. The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex. Brain Res 17: 205–242, 1970.[CrossRef][Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  R. D. Frostig, Y. Xiong, C. H. Chen-Bee, E. Kvasnak, and J. Stehberg Large-Scale Organization of Rat Sensorimotor Cortex Based on a Motif of Large Activation Spreads J. Neurosci., December 3, 2008; 28(49): 13274 - 13284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Drew and D. E. Feldman Representation of Moving Wavefronts of Whisker Deflection in Rat Somatosensory Cortex J Neurophysiol, September 1, 2007; 98(3): 1566 - 1580. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Rodgers, A. M. Benison, and D. S. Barth Two-Dimensional Coincidence Detection in the Vibrissa/Barrel Field J Neurophysiol, October 1, 2006; 96(4): 1981 - 1990. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
