Voltage-sensitive dye imaging of intervibrissal fur-evoked activity in the rat somatosensory cortex

Neuroscience Letters 381 (2005) 258–263

Voltage-sensitive dye imaging of intervibrissal fur-evoked activity in the rat somatosensory cortex Ichiro Takashima a,∗ , Riichi Kajiwara a , Toshio Iijima b a

Neuroscience Research Institute, National Institute of Advanced Industrial Science and Technology, 1-1-1 Umezono, Tsukuba 305-8568, Japan b Graduate School of Life Sciences, Tohoku University, 2-1-1 Katahira, Aoba, Sendai 980-8577, Japan Received 4 January 2005; received in revised form 18 February 2005; accepted 22 February 2005

Abstract The intervibrissal fur-evoked activity in the rat somatosensory cortex was investigated using high-resolution optical imaging with a voltagesensitive dye. The optical imaging revealed that the intervibrissal fur representation forms a U-shaped band around the borders of the posteromedial barrel subfield (PMBSF), and that this representation is characterized by a rostral-to-caudal somatotopic organization. When GABAA -mediated inhibition was partially suppressed by treatment with bicuculline, stimulation of the intervibrissal fur elicited spreading of an excitation wave in an area outside the PMBSF. The spreading wave propagated in both directions along the aforementioned U-shaped band of cortex, but barely invaded the center of the PMBSF. These imaging results suggest a distinct subdivision of cortex adjacent to, but outside, the PMBSF in the rat somatosensory cortex; this region receives input from intervibrissal fur, and seems to process its sensory information through well-developed local horizontal connections. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Vibrissae; Intervibrissal fur; Barrel cortex; PMBSF; Optical imaging; Voltage-sensitive dye

The anatomical representations of the large mystacial vibrissae are topographically organized in the posteromedial barrel subfield (PMBSF) of layer IV primary somatosensory (SI) cortex in rat. Neurons above, below, and within individual barrels are most responsive to input from a particular vibrissa [8]. Conversely, little is known about the cortical projections of intervibrissal fur in rodents. It had been thought that projections from intervibrissal fur would be somatotopically interspersed among those of the mystacial vibrissae within the PMBSF. However, most electrophysiological studies have failed to locate fur-evoked activity within the PMBSF, and the barrel field of the rat SI cortex is remarkably unresponsive to stimulation of intervibrissal fur [2,5,16]. In two electrophysiological studies, intervibrissal fur-evoked responses were elicited close to the anterior border of the PMBSF in rat [11], and along the lateral and medial borders of the PMBSF in mouse [9]. Studies in which 2-deoxyglucose (2DG) was

∗

Corresponding author. Tel.: +81 29 861 5563; fax: +81 29 861 5559. E-mail address: [email protected] (I. Takashima).

0304-3940/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2005.02.062

used to map cortical response locations in rat produced similar results [12,13]. Each of these studies suggested that the representation of intervibrissal fur lies outside the PMBSF and appears instead to be localized to the borders of the PMBSF. In the present study, we used optical imaging of a voltage-sensitive dye to map the representation of intervibrissal fur in rat SI cortex with a greater degree of spatial resolution than has been the case in previous studies. Eight male Wistar rats (180–220 g) were anesthetized with ketamine (80 mg/kg i.p.) and xylazine (8.8 mg/kg i.p.). Supplemental injections were used to maintain a constant level of anesthesia as assessed by respiration rate, heart rate, rectal temperature, corneal reflex, and the color of the extremities. Each animal was positioned in a stereotaxic frame and a craniotomy (5 mm × 5 mm) was performed over the left somatosensory cortex. The dura was removed after a well of dental acrylic had been built around the exposed cortex. The surface of the cortex was exposed for ∼1 h to 0.6 mg/ml RH795 (Molecular Probes, Eugene, OR) that had been dissolved in artificial cerebrospinal fluid (ACSF). RH-795 is a fast response probe that transforms changes in excitable membrane

I. Takashima et al. / Neuroscience Letters 381 (2005) 258–263

potential into optical signals with a submillisecond delay [3]. After the staining, the cortex was rinsed thoroughly and the well was filled with ACSF. The composition of ACSF was (mM): NaCl 125, KCl 5, CaCl2 2, MgSO4 1.25, NaH2 PO4 1.25, NaHCO3 22, glucose 10. All of the mystacial whiskers on the right side of the snout were carefully plucked under a microscope. We did this time consuming procedure 1 day before the experiment, assuming that no reorganization of the barrel cortex would occur in the short time before the optical experiments. An electromechanical actuator (P1105m, TDK) was used to stimulate the intervibrissal fur. A fine cosmetic brush (3 mm wide) was attached to the actuator probe and a single driving pulse to the actuator displaced the brush by ∼2 mm for 10 ms, which swept the intervibrissal fur on the snout over an area of ∼6 mm2 . We carefully positioned the brush not to touch the rat’s skin; therefore the brushing stimulation was effective for the intervibrissal fur, but not for the whisker pads. Optical signals from the exposed cortex were recorded using a tandem-type epifluorescence microscope and a MOSbased monolithic array camera (64 × 64 pixels), both of which were developed in our laboratory [4,14]. Light from a tungsten–halogen lamp was filtered (535 nm) before being reflected down onto the brain surface by a dichroic mirror (580 nm), and fluorescence from the cortex was projected to a sensor through a long-pass filter (600 nm). Fluorescence images were acquired at a rate of 1.2 ms/frame and covered an area of cortex of ∼4.2 mm × 4.2 mm. The camera was positioned to focus 250 ␮m below the surface of the brain. Image acquisition was triggered by an electrocardiogram using a stimulus/non-stimulus subtraction method [10,15]. Data for eight trials were averaged. In the first experiment, intervibrissal fur-evoked responses were mapped and somatotopy was assessed in the normal condition. Then, GABAergic inhibition was partially suppressed using bicuculline. Bicuculline methiodide (5 ␮M) was dissolved in ACSF and applied to the SI cortex by placing a cotton swab that had been soaked with the bicuculline solution onto the surface of the cortex. After 5 min, the cotton swab was removed and the cortex was rinsed thoroughly with ACSF. In the second experiment, the bicuculline-enhanced spread of fur-evoked activity was monitored. During the experiments, we recorded epicortical potentials with a silver ball electrode that was positioned in close proximity to the area of cortex from which we recorded optical signals. No spontaneous epileptiform activity was observed in any animal. After the experiments, the cortex was lesioned and processed for cytochrome oxidase (CO) histochemistry. Lesions (at least 3) were made at distinct points on the cortical surface vasculature, such as bifurcations, and a bipolar stimulating electrode was used to produce a small lesion by passing dc current through it. The animals were killed with an overdose of pentobarbital (Nembutal). A block of brain that included the area of cortex that had been imaged was removed, fixed overnight at 4 ◦ C in a 4% paraformaldehyde solution, and

259

then sectioned tangentially at a thickness of 80 ␮m. Tissue sections were exposed to CO to assess oxidative metabolic activity, and the pattern of the barrels in the SI cortex was reconstructed using two or three CO-stained sections from layer IV. The barrel pattern was superimposed on the optical imaging results by comparing the lesions in CO sections with the vasculature of the cortex surface. Fig. 1 shows the optically mapped cortical representation of intervibrissal fur. We stimulated the intervibrissal fur in five areas on the rat snout: (1) dorsocaudal, (2) dorsorostral, (3) rostral, (4) ventrorostral, and (5) ventrocaudal. Panels 1–5 show the cortical maps obtained after stimulating the fur in each of these regions, respectively. Optical responses in the cortex were always observed around the borders of the PMBSF. Examination of such maps revealed that the intervibrissal fur in the dorsal (ventral) region of the snout projected ventrally (dorsally) to the PMBSF, while the caudal (rostral) fur projected caudally (rostrally). In the next experiment, we examined the role of GABAergic inhibition in shaping the intervibrissal fur-evoked responses in the surrounding cortex of the PMBSF. Fig. 2 shows spreading waves of excitation that were elicited by stimulating the intervibrissal fur, after GABAA -mediated inhibition was partially suppressed. The spreading wave propagated along a U-shaped band around the PMBSF, and the U-shaped band included the receptive fields of intervibrissal fur that were mapped in Fig. 1. When the fur in the dorsocaudal area of the snout was stimulated (Fig. 2A), the resultant cortical activity was initiated from the region ventrocaudal to the PMBSF, propagated in a band lateral to the row A barrels from caudal to rostral, via the anterior border of the PMBSF, and then in a band medial to the row E barrels from rostral to caudal; a clockwise direction in the figure. By contrast, activity was propagated in the opposite direction (i.e., counterclockwise in the figure) when the fur in the ventrocaudal area was stimulated (Fig. 2B). Some activity was also observed in the anterolateral barrel subfield. The optical signals in response to stimulation of the intervibrissal fur before and after the application of bicuculline are shown in Fig. 2C. The traces on the left and right correspond to Fig. 2A and B, respectively. The onset of the fluorescence signals was ∼10 ms after stimulation in the area where the cortical response was initiated. After the application of bicuculline, the propagation of the spreading wave shown in Fig. 2 was monitored in every animal. The velocity of propagation was essentially constant along the U-shaped pathway in both directions and was estimated to be 0.28–0.33 m/s. We next analyzed the extent of fur-evoked activity across the borders of the PMBSF (Fig. 3). The borders were defined by the circumscribed curves placed on the medial edges of the row E barrels (as shown in Fig. 3A) and the lateral edges of the row A barrels. Pixels were chosen at positions along the line perpendicular to the PMBSF border, and the maximum amplitude of the optical signal was evaluated for each pixel. Example records in response to fur stimulation in the ventrocaudal area before and after bicuculline are indicated

260

I. Takashima et al. / Neuroscience Letters 381 (2005) 258–263

Fig. 1. Localization of the somatosensory representation of intervibrissal fur. The lateral surface of a rat snout is illustrated in the center panel, showing the five areas that were stimulated. As whiskers were removed, dots in the rat snout indicate whisker pads. The optical signal was measured as the fractional change in fluorescence intensity (F/F). In each panel (panels 1–5), cortical activity was encoded according to a pseudocolor scale and was superimposed on the image of the cortex surface. The anatomical whisker barrel map is also illustrated. These functional maps depict the maximum spatial propagation of optical responses that were measured 24 ms after the stimulation of intervibrissal fur. Note that ventral is up in all panels (1–5). These panels are flipped for easy comparison between the optical maps and the stimulation sites. The time course of the optical signals corresponding to panels 1 and 5 is shown in Fig. 2C as the control response. The A1 and E1 barrels are outlined in yellow and labeled in panel 1. dc, dorsocaudal; dr, dorsorostral; r, rostral; vr, ventrorostral; vc, ventrocaudal. R, rostral; V, ventral.

in Fig. 3A, where the five pixels were located at intervals of ∼200 ␮m. The maximum amplitude of the optical signal is denoted by S1 (control) and S2 (bicuculline). Fig. 3B shows the profile of the maximum signal amplitude across the PMBSF border. The signal amplitude was normalized at the outer edge of row A or row E. The fur-evoked activity peaked around 200 ␮m from the PMBSF border and spread about 700–800 ␮m (control) or 1200–1400 ␮m (bicuculline). The activity extended to the border barrels (row A or E) and slightly to the barrel septa; however, it did not extend to the central barrels (rows B, C, and D), even after bicuculline. We have demonstrated that intervibrissal fur is represented outside the PMBSF in rat SI cortex. Specifically, projections from intervibrissal fur were located ventral to the barrels of row A, dorsal to the barrels of row E, and immediately ante-

rior to the PMBSF. These areas of cortex formed a U-shaped band that was located adjacent to the PMBSF, and closely resemble the cortical fur projection zones described by Nussbaumer and Van der Loos [9]. The superior resolution that was made possible by using optical imaging in the present study revealed additional detail about the cortical representation of intervibrissal fur. Specifically, there appears to be a rostralto-caudal somatotopic organization within the U-shaped cortical band. Nussbaumer and Van der Loos [9] reported that the ventral half of the snout projected to an area of cortex located immediately medial to row E, and that the dorsal half projected to an area that was lateral to row A. We confirmed this observation in the present study. In addition, our results indicate that rostral fur projected rostrally and that caudal fur projected caudally to narrow strips of cortex that were located

I. Takashima et al. / Neuroscience Letters 381 (2005) 258–263

261

Fig. 2. Bicuculline-enhanced spreading of fur-evoked activity. (A) The spatiotemporal pattern of the excitation wave that was evoked by stimulating intervibrissal fur. The imaging data are from the same animal used for Fig. 1. The optical signal was encoded according to the same pseudocolor scale that was used in Fig. 1. The stimulus onset was at 0 ms and the time (in ms) after stimulation is indicated below each image. The anatomical whisker barrel map is illustrated in every other image. The area of intervibrissal fur that was stimulated is indicated in the leftmost frame. dc, dorsocaudal. (B) Similar to (A), except that a different area of intervibrissal fur was stimulated. vc, ventrocaudal. (C) Time course of the optical responses. The traces on the left show records from selected pixels when the dc-area of intervibrissal fur was stimulated, while the traces on the right are for stimulation of the vc-area. The dotted and continuous lines correspond to control and bicuculline-enhanced responses, respectively. The 15 selected pixels are indicated in the center panel of a CO-stained section. These pixels were chosen on the U-shaped curve surrounding the PMBSF at regular ∼400 ␮m intervals. The estimated propagation velocity was 0.28–0.33 m/s.

immediately adjacent to the barrels of rows A and E (see panels 1, 2, 4, 5 in Fig. 1). The area of cortex located anterior to the PMBSF (including the anterolateral barrel subfield) was activated when the rostral snout was stimulated (see panel 3 in Fig. 1). Using a more localized stimulation of fur, we explored whether there might be medial-to-lateral somatotopy within each strip of cortex adjacent to row A or E, but were unable to determine this (data not shown). The spatiotemporal pattern of the spreading excitation wave that was elicited by stimulating intervibrissal fur was highly distinctive (see Fig. 2A and B). This pattern of activity was observed in all animals whenever GABAA -mediated inhibition had been suppressed by bicuculline. It is difficult to determine to what degree our method of applying bicuculline suppressed GABAA receptor activation. We showed previously that when GABAA -mediated inhibition was suppressed completely by the application of a high concentration of bicuculline, spontaneous epileptiform activity in rat SI cortex was propagated from the PMBSF into the surrounding cortex and vice versa [14]. There was no boundary between the PMBSF and the surrounding cortex in the

case of spontaneous epileptiform activity. Therefore, in the present study, we monitored epicortical field potentials in the exposed cortex, and confirmed that there was no spontaneous epileptiform activity throughout the imaging experiments. From this observation, we conclude that GABAA mediated inhibition was not suppressed completely by the bicuculline treatment used in the present study. The band of cortex that was activated in response to the stimulation of intervibrissal fur included the dysgranular zone (DZ), which encompasses the PMBSF. (The DZ refers to the region in and around the barrel field in layer IV that has low reactivity to CO, which is used to assess oxidative metabolic activity [2].) Using immunocytochemistry, Land et al. [7] showed that the density of GABAA receptors in rat SI cortex is correlated with the intensity of CO reactivity, and that the area surrounding the PMBSF contains fewer GABAA receptors and a lower density of GABAA -containing neurons. These observations would explain the spatiotemporal characteristics of the spreading wave of activity that was evoked in this subdivision of cortex by the stimulation of intervibrissal fur in the present study.

262

I. Takashima et al. / Neuroscience Letters 381 (2005) 258–263

Fig. 3. Fur-evoked activity extended to the barrels at the border of the PMBSF. (A) The left panel shows a magnified photo of a CO-stained section including barrels E1 and E2. A dotted line indicates the border that was circumscribed to the medial edges of the row E barrels. One pixel on the border was selected (pixel 3) and the other selected pixels were on a line normal to the border curve at distances of 200 ␮m (pixels 2 and 4) and 400 ␮m (pixels 1 and 5) from the border. The traces on the right show the signals recorded by these pixels before (dotted) and after (solid lines) bicuculline, when the intervibrissal fur in the ventrocaudal area was stimulated. The maximum amplitude of the optical signal was assessed for each pixel as S1 (control) and S2 (bicuculline). (B) The profile of the maximum signal amplitude (S1 and S2 ) evaluated across the PMBSF border. The dotted and solid lines correspond to before (S1 ) and after (S2 ) bicuculline, respectively. The optical responses to the dorsocaudal fur stimulation were analyzed around barrels A1–A3 (the result is shown on the left), while the responses to ventrocaudal fur stimulation were analyzed around barrels E1–E3 (the result is on the right). Data from 71 to 96 pixels (eight animals) were used to calculate the mean ± S.D. The signal amplitude was normalized at the outer edge of the row A or row E barrels. The hatched bars indicate the averaged size of the row A (284 ± 72 ␮m) and row E (313 ± 48 ␮m) barrels measured orthogonal to the PMBSF border.

Detailed examination of the optical signals revealed that fur-evoked activity extended to the barrels at the border of the PMBSF (see Fig. 3B). The dye-related optical signals represent the integral of membrane potential changes in neurons, as well as a possible contribution from the depolarization of neighboring glial cells [6]. It is not easy to determine from which cortical element the optical signals mostly originate, but if we assume that the signals mainly reflect the postsynaptic potentials in the fine dendrites of cortical neurons [3], then our imaging results suggest that neurons receiving intervibrissal fur projections extend their dendritic arborization to the border barrels, presumably, in the supragranular layers. However, regarding the functional interaction between the border barrels (row A or E) and their adjacent cortex, we can say nothing from our experiments. Using intrinsic signal optical imaging, Brett-Green et al. [1] showed that the activity in response to border whisker (A2 or E2) stimulation spread beyond the PMBSF. Their results and our data may suggest the possible integration of sensory information arising from vibrissae and the intervibrissal fur. This study visualized neural activity related to the stimulation of intervibrissal fur in the rodent SI cortex using a

voltage-sensitive dye. The results not only verified the notion that the cortical representation of fur lies outside the PMBSF, but also illustrated somatotopy of the cortex bordering the PMBSF. After partial suppression of GABAA -mediated inhibition, a distinctive pattern of spreading excitation waves was monitored outside the PMBSF, which suggests that the neural circuits for the PMBSF and surroundings are separated and different. Further studies should verify the functional connectivity between the PMBSF and its surrounding cortex; however, we propose that the U-shaped band of cortex bordering the PMBSF is a relatively segregated subdivision of cortex, in which the intracortical horizontal connections are highly developed for processing intervibrissal fur information.

References [1] B.A. Brett-Green, C.H. Chen-Bee, R.D. Frostig, Comparing the functional representations of central and border whiskers in rat primary somatosensory cortex, J. Neurosci. 21 (2001) 9944–9954. [2] J.K. Chapin, C.S. Lin, Mapping the body representation in the SI cortex of anesthetized and awake rats, J. Comp. Neurol. 229 (1984) 199–213.

I. Takashima et al. / Neuroscience Letters 381 (2005) 258–263 [3] A. Grinvald, R.D. Frostig, E. Lieke, R. Hildesheim, Optical imaging of neuronal activity, Physiol. Rev. 68 (1988) 1285–1366. [4] M. Ichikawa, T. Iijima, G. Matsumoto, Real-time optical imaging of neural activities in the brain, in: T. Ono, L.R. Squire, M.E. Raichle (Eds.), Brain Mechanisms of Perception and Memory, Oxford University Press, Oxford, 1993, pp. 638–648. [5] H.P. Killackey, G.O. Ivy, T.J. Cunningham, Anomalous organization of SmI somatotopic map consequent to vibrissae removal in the newborn rat, Brain Res. 155 (1978) 136–140. [6] A. Konnerth, A.L. Obaid, B.M. Salzberg, Optical recording of electrical activity from parallel fibers and other cell types in skate cerebellar slices in vitro, J. Physiol. 393 (1987) 681–702. [7] P.W. Land, A.L. de Blas, N. Reddy, Immunocytochemical localization of GABAA receptors in rat somatosensory cortex and effects of tactile deprivation, Somatosens. Mot. Res. 12 (1995) 127–141. [8] C.I. Moore, S.B. Nelson, M. Sur, Dynamics of neuronal processing in rat somatosensory cortex, Trends Neurosci. 22 (1999) 513–520. [9] J.C. Nussbaumer, H. Van der Loos, An electrophysiological and anatomical study of projections to the mouse cortical barrelfield and its surroundings, J. Neurophysiol. 53 (1985) 686–697. [10] H.S. Orbach, L.B. Cohen, A. Grinvald, Optical mapping of electrical activity in rat somatosensory and visual cortex, J. Neurosci. 5 (1985) 1886–1895.

263

[11] B. Pidoux, R. Verley, E. Farkas, J. Scherrer, Projections of the common fur of the muzzle upon the cortical area for mystacial vibrissae in rats dewhiskered since birth, Neurosci. Lett. 11 (1979) 301– 306. [12] F.R. Sharp, M.F. Gonzalez, C.W. Morgan, M.T. Morton, J.W. Sharp, Common fur and mystacial vibrissae parallel sensory pathways: 14C2-deoxyglucose and WGA-HRP studies in the rat, J. Comp. Neurol. 270 (1988) 446–469. [13] E. Siucinska, M. Kossut, Plasticity of mystacial fur representation in SI cortex of adult vibrissectomized rats—a 2DG study, Neuroreport 5 (1994) 1605–1608. [14] I. Takashima, M. Ichikawa, T. Iijima, High-speed CCD imaging system for monitoring neural activity in vivo and in vitro, using a voltage-sensitive dye, J. Neurosci. Methods 91 (1999) 147– 159. [15] I. Takashima, R. Kajiwara, T. Iijima, Voltage-sensitive dye versus intrinsic signal optical imaging: comparison of optically determined functional maps from rat barrel cortex, Neuroreport 12 (2001) 2889–2894. [16] C. Welker, Microelectrode delineation of fine grain somatotopic organization of SmI cerebral neocortex in albino rat, Brain Res. 26 (1971) 259–275.