EXPERIMENTALNEUROLOGY
108,162-175
(1990)
Physiological Changes in the Somatosensory Forepaw Cerebral Cortex of Adult Raccoons following Lesions of a Single Cortical Digit Representation GERNOT *Department Georgia
S.DOETSCH,*~~ KIM W. JOHNSTON,*ANDCHARLESJ.HANNAN,JR.$,~
of Surgery (Section of Neurosurgery), 30912; and SDepartment of Clinical
and TDepartment of Physiology and Endocrinology, Investigation, Dwight D. Eisenhower Army Medical
Medical College Center, Augusta,
of Georgia, Augusta, Georgia 30905
INTRODUCTION The purpose of this study was to determine whether restricted lesions within primary somatosensory (SmI) cortex cause changes in the functional organization of cortical areas bordering on the site of injury. Focal ablations of cortical tissue were made in the representational area for digit 3 within the SmI forepaw cortex of adult raccoons. Electrophysiological mapping experiments done 15-17 weeks later showed that significant alterations had occurred in the response properties of clusters of neurons within those representational zones adjoining the lesion -the zones for digit 2, digit 4, and the palmar pads. These three cortical areas were modified by the appearance of new, usually weaker secondary inputs and changes in some properties of the normal primary inputs from the forepaw. (i) Many neurons responded to stimulation of previously ineffective skin regions; the new inputs often originated from digit 3 but frequently involved other digits or the pads as well. (ii) Neuronal receptive fields (RFs), mapped at a standard suprathreshold stimulus intensity, were larger than normal. (iii) Skin type and submodality sensitivity typically were less specific than normal; more neurons had RFs that included both glabrous and hairy skin or claws and displayed mixtures of responsiveness to skin touch, hair deflection, or claw touch. (iv) The representation of RF location, skin type, and submodality sensitivity was more variable as a function of horizontal and vertical distance through the cortex. In general, the physiological changes were found to degrade the somatotopic order and response specificity of the intact cortical areas adjoining the lesion. The relatively mixed nature of the novel inputs following cortical damage was qualitatively similar to that produced by digital nerve transection described previously by Kelahan and Doetsch (Somatosens. Res., 1984, 2: 49-81). The data suggest that both cortical and peripheral somatosensory injuries may produce their effects by unmasking synaptically weak neural connections. Synaptic unmasking may involve facilitation or disinhibition of neurons with convergent properties-especially those located within the heterogeneous subdivisions of SmI cortex. 0 1990 Academic Press, Inc.
0014-4&x%/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
The somatic afferent system of the raccoon provides a unique model for studying the normal functional organization of somatosensory (SmI) cerebral cortex and for determining possible organizational changes following experimental manipulation. The representation of the forepaw within SmI cortex of this species is enormous relative to that of other mammals. The forepaw cortex is anatomically delineated by a pattern of sulci whereby the focal representation of each digit and the palmar pads occupies a separate and relatively distinct cortical subgyrus (20). Recent microelectrode mapping experiments (9, 10, 11) have shown that each forepaw digit zone appears to contain at least two major subdivisions, one representing the glabrous skin and the other representing the hairy skin and claws. The glabrous skin subdivision of each digit area is large, centrally located, and bounded on three sidesby the smaller hairy skin and claw subdivision. Neurons within the glabrous skin sector typically respond to low-intensity tactile stimulation of the skin and have small receptive fields (RFs) that are somatotopically organized. In contrast, neurons within the hairy skin and claw subdivision respond to movement of hairs, to touch of the hairy skin or claw, and sometimes also to tactile stimulation of glabrous skin; their RFs seem to be larger and are distributed with less somatotopic precision than those of cells in the glabrous skin subdivision. Because of its mixed response properties, the hairy skin and claw sector has been referred to as the “heterogeneous” region by Johnson et al. (10). The results of recent anatomical tracer studies have shown that the two cortical subdivisions have different thalamic (3) and intracortical (2) connections that may largely account for the differences in their physiological properties. The highly developed SmI cortex of the raccoon has been used as a model system to examine the effects of 1 Present address: Clinical Investigation Army Medical Center, Tacoma, WA 98431. 162
Inc. reserved.
Department,
Madigan
EFFECTS
OF
LESIONS
IN
peripheral nerve lesions on the organization of central somatosensory maps (1, 11,12,16). The results of these studies have shown that digital nerve transection-by surgical removal of a digit-causes neurons within the deafferented cortical digit area to develop responsiveness to “new” peripheral inputs. Novel RFs were found primarily on the digits adjoining the site of injury and to a lesser extent on the neighboring palmar pads. The emergence of altered neuronal response properties and new RFs was interpreted as being due primarily to unmasking of preexisting but synaptically ineffective inputs to the deprived cortical zone. This plasticity suggested that similar physiological changes might occur following damage to the cortex itself. Any significant alterations in neuronal responsiveness could provide insight into the mechanisms of functional recovery following cerebral insult. To test this notion-and to develop a physiological model of recovery from focal brain damage-lesions were made in the SmI cortical subgyrus representing the third forepaw digit of adult raccoons. About 4 months later, portions of the SmI forepaw cortex of these animals were studied electrophysiologically to determine whether any functional changes had occurred within cortical regions bordering on the lesion. Comparison of electrophysiological data obtained pre- and postoperatively from lesioned animals with data from normal animals showed that significant functional alterations can occur within cortical zones adjoining the lesion. METHODS
Electrophysiological experiments were performed on 14 adult raccoons (Procyon lotor) of both sexes. Cortical lesions were made in seven of these animals using the technique described below. Possible long-term effects were examined by studying five lesioned animals at about 4 months (15-17 weeks) following surgery; shortterm effects were assessed in two animals at 2 and 4 weeks after the lesion. The results were compared with data obtained from the same animals before the operation and with corresponding data taken from seven normal raccoons. Surgical Procedures Before making the cortical lesion, each animal was sedated with ketamine HCl (Ketaset, 15 mg/kg, im) and anesthetized with a-chloralose (35 mg/kg, ip) or pentobarbital sodium (Nembutal, 30 mg/kg, ip). An endotracheal tube was inserted and the animal was artificially ventilated with a respirator pump. The head was secured in a stereotaxic frame and a craniotomy was performed under antiseptic conditions to expose the SmI forepaw cortex of the right hemisphere. The dura mater was in-
SOMATOSENSORY
163
CORTEX
cised and reflected to provide access to the digit 3 cortical zone and the surrounding areas representing digit 2, digit 4, and the palmar pads. Photographs were taken of the exposed cortex before and after making the lesion. In five of these animals, the boundaries of the digit 3 zone were mapped electrophysiologically using a silver ball electrode to record primary tactile-evoked responses; in the other two animals, mapping was done with a tungsten microelectrode inserted into cortical layers III-IV to record the responses of small clusters of neurons. After the somatotopic borders had been determined, small blood vessels penetrating the cortical tissue of the digit 3 zone were electrocoagulated using bipolar forceps. The crown of the digit 3 subgyrus was then removed by subpial aspiration. The lesion extended anteriorly to the caudal lip of the ascending coronal (lateral central) sulcus and posteriorly to the rostra1 lip of the triradiate sulcus (see Fig. 1A). The lesion was confined to cortical tissue, sparing the underlying white matter as much as possible. After the ablation was completed and all bleeding had been stopped, the cranial defect was repaired by suture of the reflected tissues. The raccoon was given an injection of ampicillin sodium (Omnipen-N, 250 mg, im), returned to the animal care facility, and allowed to recover. Electrophysiological
Recording Procedures
After survival periods of 15-17 weeks (five animals) or 2 and 5 weeks (one animal each), the operated raccoon was studied electrophysiologically. Systematic mapping experiments were done to determine whether any physiological changes had occurred in those cortical regions that border on the lesioned digit 3 area-namely, the zones representing digit 2, digit 4, and the palmar pads. The unoperated raccoons were studied electrophysiologically in the same manner to permit appropriate statistical comparisons to be made. Conventional surgical and electrophysiological recording procedures were used in all experiments. Before each recording session, the raccoon was sedated with ketamine HCl (Ketaset, 15 mg/kg, im) and then anesthetized with a-chloralose (35 mg/kg, ip). The animal was intubated with a tracheal cannula and artificially ventilated with room air. The right femoral vein was cannulated to administer supplementary doses of chloralose and to infuse dextrose-saline to replace lost body fluids; the femoral artery was cannulated to monitor arterial blood pressure. The animal’s head was placed in a stereotaxic frame, and the right SmI forepaw cortex was surgically exposed and photographed. An agar-gauze dam was constructed around the cranial defect and was filled with warm mineral oil to prevent drying of cortical tissue. Bilateral pneumothorax and drainage of cerebro-
164
A
DOETSCH,
JOHNSTON,
B
AND
HANNAN
C
FIG. 1. Photographs of a portion of the exposed SmI forepaw cortex before (A), immediately after (B), and about 16 weeks after (C) a lesion was made in the digit 3 subgyrus. (A) The cortical representational zones for digits 2,3, and 4 (2-4) and the palmar pads (P) with respect to the triradiate (T) sulcus and the ascending branch of the coronal (C) sulcus.
spinal fluid from the cisterna magna were used to minimize movements of the brain due to respiration or arterial pulse. Core body temperature was maintained at 3738°C with an abdominal dc heating pad, servoregulated with a rectal thermometer. Tungsten microelectrodes were used to record the extracellular responses of small clusters of neurons to tactile stimulation of the digits, palmar pads, and other forepaw regions. The microelectrode was advanced into the cortex using a depth-calibrated hydraulic microdrive. All penetrations were made as normal to the cortical surface as possible, to a depth of at least 2.0 mm. The neural signals were fed into a preamplifier and audiomonitor and displayed on an oscilloscope. All recording sites were marked on the photographs of the cortex. The composite RFs of small clusters of neurons were mapped at a standard suprathreshold stimulus intensity of 5 g with a calibrated nylon filament probe. The RFs were drawn on pictures of the forepaw; the area of each field was later measured with a planimeter and expressed in absolute planimetric units. The submodality sensitivity of each neuronal cluster was determined using conventional criteria: (i) skin touch-response to light touch of glabrous or hairy skin with a nylon filament probe; (ii) hair movement-activation by deflection of small groups of hairs with light air puffs or gentle stroking; (iii) claw touch-response to touch or light tap of a claw with a nylon probe; and (iv) deep pressure/joint movement-activation by steady pressure on skin over-
lying deep tissues/response to changes in joint position. While recording in each electrode track, care was taken to note clear changes in RF location, size, skin type, or submodality sensitivity of neuronal clusters with cortical depth. At the end of each experiment, the animal was sacrificed with an intravenous injection of sodium pentobarbital and perfused through the heart with 0.9% saline followed by 10% formalin-saline. The brain was then removed and stored in 10% formalin-saline. Blocks of tissue containing the lesioned area or the recording sites were later embedded in paraffin, serially sectioned, and stained with hematoxylin-eosin or 1~x01 fast blue. RESULTS
The results obtained by comparing pre- and postlesion data from the operated raccoons and the results of comparing data from different groups of normal and lesioned animals were consistent with one another. The findings indicated that damage restricted to the digit 3 cortical zone produced changes in the physiological characteristics of the zones bordering on the site of injury. Clusters of neurons were found to be activated from more widespread regions of the forepaw digits and palmar pads; the expansion of composite RFs was accompanied by greater mixtures of skin type and by decreased somatotopic and submodality specificity.
EFFECTS
OF
LESIONS
IN
Histology The surface morphology of the forepaw cortex was documented photographically before and immediately following the ablation, and again at the time of electrophysiological study, 2, 5, or 15-17 weeks later (Fig. 1). Just after the surgery, the edges of the lesion were sharp and easily distinguished by the margins of intact cortical tissue. After 15-17 weeks, the border regions had become smooth and partially revascularized, making the edges of the lesion more difficult to identify. Microscopic examination of the histological material revealed a lesioned area with extensive loss of cortical tissue that was surrounded by a region of gliosis containing few neurons; the latter region bordered directly on cortical zones exhibiting normal cytoarchitectonic features. Comparison of the histological data with previously defined electrophysiological cortical maps showed that the ablations were confined to the digit 3 zone and were subtotal in extent. In all cases, the cortical tissue that was removed constituted the representation of approximately the distal one-half of the glabrous skin of digit 3. Small portions of the anterior heterogeneous subdivision, representing the hairy skin and claw of digit 3, were also included in the lesion. The cortical ablation spared the more proximal glabrous digit 3 representation located within the banks of the triradiate sulcus. Comparison of Pre- and Postlesion Data from Individual Animals The most direct method for assessing the lesion effects was to compare the response properties of neuronal clusters within a cortical zone before and after surgery in the same animal. The results of such analyses are illustrated in Figs. 2 and 3 for two raccoons studied before and about 16 weeks after the digit 3 area had been lesioned. Recordings taken from nearly identical or closely adjacent electrode sites at comparable cortical depths (7501250 pm) were selected for comparison. The results indicated that the functional characteristics of neurons in a few tracks had remained relatively stable, with only minor changes in RF location, size, skin type, and submodality sensitivity (Fig. 2, track 3A, Fig. 3, tracks 7A and lOA). However, the vast majority of electrode penetrations revealed definite alterations in neuronal responsiveness. The most prominent change was the development of responses to stimulation of previously ineffective skin regions of the forepaw. This was reflected in the expansion of composite RFs to include portions of one or several digits or palmar pads that were outside the borders of the prelesion RF. Of the 22 pairs of RFs shown in Figs. 2 and 3, fully 17 increased in size, four decreased in size, and one remained unchanged after the lesion. In most cases, stimulation of the original RF still elicited the strongest “multiunit” response,
SOMATOSENSORY
CORTEX
165
whereas stimulation of the newly acquired, postlesion component of the RF produced a weaker response (fewer multiunit spikes). The novel inputs often originated from digit 3, but frequently involved other digits or the pads, depending on the site of recording. Furthermore, the postlesion RFs typically included greater mixtures of skin type, with varying combinations of glabrous or hairy skin and claws forming part of the RF. Consistent with this heterogeneity, many clusters of neurons also exhibited increased mixtures of submodality sensitivity to skin touch, hair deflection, or claw touch. Finally, the functional alterations observed within the digit and pad zones adjoining the lesion tended to degrade the normal somatotopic order and response specificity of these zones. Comparison
of Data from Normal and Lesioned Animals
It was not feasible to obtain extensive electrophysiological data from any one animal prior to making a lesion, due to the time required for mapping the large cortical zones under study and the need for restricting the duration of anesthesia and surgery. Consequently, the pre- and postlesion analysis of individual raccoons was supplemented by statistical comparison of data obtained from approximately matched electrode tracks in separate groups of normal and operated animals. Definitions
and Criteria
To ensure the validity of the statistical procedures, strict definitions of the cortical zones and rigorous criteria for measuring various neuronal response properties were adopted. Cortical representational zones. The borders of the three cortical zones adjoining the digit 3 area were defined by previously determined electrophysiological maps and their relationship to the surface pattern of the triradiate sulcus in normal animals. The digit 2/digit 3 border was found to be demarcated consistently by the anterolateral branch of the descending triradiate sulcus. The pad/digit 3 border corresponded to the major central portion of the triradiate sulcus. In contrast, the boundary between the digit 4 and digit 3 zones was less well delineated, typically related only to a small anteriorly oriented spur of the triradiate sulcus. Consequently, this border was defined solely by electrophysiological criteria in both groups of animals-recording sites were considered to be within the digit 4 or the digit 3 zone depending on whether the strongest responses first recorded in a penetration were elicited by stimulation of digit 4 or digit 3, respectively. This electrophysiologitally defined border usually fell along an imaginary line extending anteromedially from the tip of the triradiate sulcal spur located between the two digit areas.
DOETSCH,
JOHNSTON,
AND
HANNAN
B DIGIT
2
DIGIT
PAD
4
12
FIG. 2. Comparison of the receptive fields (RFs) of neurons recorded before and about 16 weeks after a lesion was made in the digit 3 cortical zone of the same animal. (A) The cortical surface map shows the locations of recording sites 1-12 (filled circles) before the lesion (shaded area) and sites lA-12A (open circles) after the lesion. (B) The RFs of neurons recorded from pairs of closely adjacent sites shown in A are compared. Prelesion RFs are indicated by numbers and are represented on the paws by stippling; comparable postlesion RFs are indicated by corresponding numbers followed by letter A and are represented by hatched lines. Note the considerable increase in size of most RFs after the lesion.
Responsive tracks and sites. A responsive electrode track was defined simply as a track containing at least one neuronal cluster that could be activated by periphera1 stimulation. A responsive site was designated as a
region with a specified depth interval within a responsive track containing neuronal clusters with very similar RFs and physiological properties; responsive sites were defined as different when regions of nonoverlapping
EFFECTS
DIGIT
2
DIGIT
4
OF
LESIONS
IN
DIGIT
SOMATOSENSORY
4
167
CORTEX
PAD
6
FIG. Symbols
3. Comparison of the receptive fields (RFs) of neurons recorded before and about 16 weeks after a lesion in the digit in A and B have the same meaning as those in Figs. 2A and 2B. Note the increase in RF size following the lesion.
depths contained neurons with obviously different characteristics. Primary and secondary responses. To quantify the strength of neuronal activation by peripheral stimulation, the responses were divided into two major categories. The strongest neuronal response produced by stimulation of a particular digit or the palmar pads was defined as primary. Weaker responses elicited by stimulation of other digits or the pads were designated as secondary.
3 cortical
zone.
On-focus and off-focus responses. The responses were subdivided also in terms of the locations of neuronal RFs. Responses recorded from a predesignated cortical zone (e.g., the digit 2 zone) were defined as on-focus if they were elicited by stimulation of the digit or pads (e.g., digit 2) “represented” within that zone. Responses were designated as off-focus if they were produced by stimulation of a digit or the pads (e.g., digit 3) not “represented” within that zone. Thus, the dominant pattern of neuronal activation found within a normal cortical zone
168
DOETSCH,
TABLE
JOHNSTON,
1
Number of Electrode Tracks and ResponsiveSites in Three SmI Cortical ZonesStudied
Cortical zone
Experimental animal”
Digit
2
Digit
4
Pad
Number of tracks 71 56 84 82 91 75
Normal Lesioned Normal Lesioned Normal Lesioned
“Data are from seven 15-17 weeks after surgery.
normal
and
five
Number/ percentage of responsive tracks
Number of responsive sites
71/100 56/100 84/100 77/93.9 91/100 74/98.7
85 86 101 108 128 132
lesioned
animals
studied
consisted of a primary on-focus response that was occasionally associated with a secondary off-focus response. Responsive Neuronal Clusters within Cortical Zones In both normal and lesioned raccoons, the vast majority of electrode tracks made within the cortical zones for digit 2, digit 4, and the palmar pads contained neurons that responded to cutaneous stimulation of the forepaw (Table 1). The RF locations and other response properties of neuronal clusters in any one track generally were similar as a function of depth below the pial surface. The number of responsive sites within a track showing obvious changes in RF location, size, or other characteristics averaged less than 2.0, but was consistently greater in the cortical zones of the lesioned animals. Functional Organization of Cortical Zones in Normal and Lesioned Animals Cortical surface maps were constructed for each animal to show the location of each electrode track and to summarize the major RF properties of the neuronal clusters encountered at each distinct responsive site within a given track. In Figs. 4-6, the maps obtained from three lesioned raccoons are compared with corresponding maps from three normal raccoons. The data shown by these maps are consistent with the results obtained before and after ablation of the digit 3 zone of individual animals. In normal raccoons, the composite RFs of neuronal clusters within the cortical zones for digit 2, digit 4, and the palmar pads were relatively small, typically confined to the appropriate digit or pad, and usually involved only one skin type. Each of the cortical areas representing a digit consisted of two major subdivisions-a posterior sector containing neurons with RFs on glabrous skin and a more anterior heterogeneous sector containing neurons with RFs on hairy skin and claws or combinations of skin type (Figs. 4A, 5A,
AND
HANNAN
and 6A). In the lesioned raccoons studied 15-17 weeks following surgery, neurons within the cortical digit and pad zones were found to receive novel inputs from the forepaw in addition to their normal afferent drive from the appropriate digits or pads. Neuronal RFs were larger, often including several digits or pads, and involved greater mixtures of different skin types (Figs. 4B, 5B, and 6B). Within the two digit zones studied, the border between the glabrous skin subdivision and the heterogeneous subdivision was considerably less distinct. Furthermore, neuronal clusters within the pad zone responded to stimulation of digits more often than normal. Finally, the lesioned animals showed more variability in RF location, size, and skin type as a function of horizontal and vertical distance through the cortex. This was evident from the more abrupt changes in RF properties of neuronal clusters within adjacent electrode tracks and from the more numerous changes at different depths within the same track. In short, neurons within the cortical areas adjoining the lesion had larger RFs with novel components; i.e., they were activated by new inputs from the forepaw. Most of the novel inputs evoked relatively weak secondary responses and originated from “inappropriate” offfocus digits or pads. These secondary off-focus responses typically coexisted with vigorous primary responses produced by stimulation of the appropriate on-focus digit or palmar pads. Statistical Comparisons of Data Obtained from Normal and Lesioned Animals The specific response characteristics measured in the normal and lesioned animals were compared statistically and are summarized below (see Table 2). Unless stated otherwise, all statistics were based on the absolute number of responsive sites exhibiting a particular functional neuronal property. Primary and secondary responses. By definition, 100% of the responsive sites within a given cortical zone in both groups of raccoons contained neuronal clusters that gave strong primary responses to stimulation of a specific digit or the palmar pads. Sites with weaker secondary responses to peripheral stimulation were relatively rare in normal animals, especially within the cortical zones for digit 2 (3.5%) and digit 4 (14.9%); secondary responses were more common in the pad zone (35.2%), indicative of greater variability in the normal inputs to this zone (Fig. 7). In contrast, each of the three cortical zones in the lesioned animals had significantly more sites with secondary responses (x2 tests, P < O.OOl), by factors ranging from about 2 to 8, than the same zones in normal animals (Fig. 7, Table 2). This finding dramatically reflects the appearance of novel physiological inputs following ablation of the digit 3 area.
EFFECTS
’
2mm
OF
LESIONS
IN
SOMATOSENSORY
CORTEX
169
’
FIG. 4. Cortical surface maps comparing the representation of the receptive fields (RFs) of neurons recorded in three cortical zones of a normal animal (A) and an animal with a digit 3 cortex lesion made about 16 weeks earlier (B). The shaded area in B shows the approximate extent of the lesion. Each circle in A and B represents an electrode penetration in which responsive neurons were recorded, each X represents a track containing unresponsive neurons. The labels in each circle designate the RF location of neurons within the penetration in order from left to right of their sensitivity to cutaneous stimulation: l-5, digits l-5; P, palmar pads; F, entire forepaw. Various shadings refer to different types of skin included within a RF: white, glabrous skin; stippled, hairy skin and/or claws; black, mixtures of glabrous and hairy skin and/or claws. The subdivided circles indicate tracks in which neurons at different depths showed shifts in the location of RFs to include different digits and pads or different skin types. The sequence of changes with increasing depth is shown within a circle from top to bottom or clockwise from upper left to lower right. Note that neurons in the lesioned animal had considerably larger RFs but greater variability in peripheral location and skin type involved, indicative of new inputs but less somatotopic precision and submodality specificity than in the normal animal.
On-focus and off-focus responses. In both groups of raccoons, the primary inputs typically originated from the appropriate on-focus digit or pads (SO-100%) and the secondary inputs arose from neighboring off-focus digits or pads (90-100%). In the lesioned animals, significant changes in the distribution of the foci-particularly increases in the number of primary off-focus responseswere observed in the digit 4 zone and the pad zone (x2 tests, P < 0.03), but not in the digit 2 area (Table 2). No significant differences were found in the foci of secondary responses within any zone, partly because of the low number of secondary inputs present in normal animals. Receptive field location. In general, the distributions of the particular digits or pads on which the primary and secondary RFs were located were similar in both sets of raccoons. The vast majority of RFs associated with primary responses involved the appropriate digits (94100%) or pads (76-80%). Only the primary responses within the digit 4 area of the lesioned animals were associated with significantly more inappropriate RF locations (x2 test, P < 0.005), mainly on digit 3 or the pads (Table 2). No significant differences in secondary RF locations were found.
Receptive field size. The composite RFs of neuronal clusters in the lesioned raccoons typically were much larger than those in normal raccoons. Significant increases were found in the size of the primary RFs of neurons within the digit 2 zone (t test, P < 0.001) and the digit 4 zone (t test, P < 0.003), but not in those of neurons within the pad zone (Fig. 8A, Table 2). The secondary RFs were significantly larger for the digit 4 area (t test, P -C 0.04) but not for the other two zones (Fig. 8B, Table 2). When the neuronal RFs were not subdivided into primary and secondary components, but were treated together as whole units, all three cortical zones were found to have larger RF areas in the lesioned raccoons (t tests, P < 0.003). The distributions of these “combined” RF sizes and their mean values are shown for each cortical area in Fig. 9. The primary Skin type and submodality sensitivity. responses within the digit 2 and digit 4 zones involved greater mixtures of skin type (x2 tests, P < 0.001) and associated submodality sensitivity (x2 tests, P < 0.001) in the lesioned animals than in the normal animals (Table 2). Neuronal responsiveness shifted from purely glabrous skin regions of a digit to various combinations of glabrous skin, hairy skin, or claws. Likewise, submodal-
170
DOETSCH,
JOHNSTON,
AND
HANNAN
A
FIG. 5. Cortical surface maps normal animal (A) and an animal as those in Figs. 4A and 4B. Note and submodality precision than in
comparing the representation of the receptive fields (RFs) of neurons recorded in three cortical zones of a with a digit 3 cortex lesion made about 16 weeks earlier (B). All symbols in A and B have the same meaning that neurons in the lesioned animal had larger RFs, indicative of new responsiveness but less somatotopic the normal animal.
ity sensitivity changed from predominantly skin touch to mixtures of skin touch, hair deflection, or claw touch. As expected, the pad zone showed no comparable differences between the two groups of animals since this zone represents the glabrous surface of the palm and borders
on the glabrous subdivisions of adjacent digit zones. In contrast, the secondary responses were associated with increased mixtures of skin type (x2 test, P < 0.04) and submodality sensitivity (x2 test, P < 0.03) only within the digit 4 area (Table 2).
A
FIG. 6. Cortical surface maps normal animal (A) and an animal as those in Figs. 4A and 4B. Note and submodality precision than in
comparing the representation of the receptive fields (RFs) of neurons recorded in three cortical zones of a with a digit 3 cortex lesion made about 16 weeks earlier (B). All symbols in A and B have the same meaning that neurons in the lesioned animal had larger RFs, indicative of new responsiveness but less somatotopic the normal animal.
EFFECTS
OF
LESIONS
IN
SOMATOSENSORY
TABLE
171
CORTEX
2
Statistical Comparisons(P Values”) of Data Obtained from Normal and LesionedAnimals Digit
Response
Primary responses
property
Number of primary and secondary responses On versus off focus responses Receptive field location Receptive field area Skin type Submodality sensitivity
2 zone
Digit
Secondary responses
Primary responses
1.0
108,162-175
(1990)
Physiological Changes in the Somatosensory Forepaw Cerebral Cortex of Adult Raccoons following Lesions of a Single Cortical Digit Representation GERNOT *Department Georgia
S.DOETSCH,*~~ KIM W. JOHNSTON,*ANDCHARLESJ.HANNAN,JR.$,~
of Surgery (Section of Neurosurgery), 30912; and SDepartment of Clinical
and TDepartment of Physiology and Endocrinology, Investigation, Dwight D. Eisenhower Army Medical
Medical College Center, Augusta,
of Georgia, Augusta, Georgia 30905
INTRODUCTION The purpose of this study was to determine whether restricted lesions within primary somatosensory (SmI) cortex cause changes in the functional organization of cortical areas bordering on the site of injury. Focal ablations of cortical tissue were made in the representational area for digit 3 within the SmI forepaw cortex of adult raccoons. Electrophysiological mapping experiments done 15-17 weeks later showed that significant alterations had occurred in the response properties of clusters of neurons within those representational zones adjoining the lesion -the zones for digit 2, digit 4, and the palmar pads. These three cortical areas were modified by the appearance of new, usually weaker secondary inputs and changes in some properties of the normal primary inputs from the forepaw. (i) Many neurons responded to stimulation of previously ineffective skin regions; the new inputs often originated from digit 3 but frequently involved other digits or the pads as well. (ii) Neuronal receptive fields (RFs), mapped at a standard suprathreshold stimulus intensity, were larger than normal. (iii) Skin type and submodality sensitivity typically were less specific than normal; more neurons had RFs that included both glabrous and hairy skin or claws and displayed mixtures of responsiveness to skin touch, hair deflection, or claw touch. (iv) The representation of RF location, skin type, and submodality sensitivity was more variable as a function of horizontal and vertical distance through the cortex. In general, the physiological changes were found to degrade the somatotopic order and response specificity of the intact cortical areas adjoining the lesion. The relatively mixed nature of the novel inputs following cortical damage was qualitatively similar to that produced by digital nerve transection described previously by Kelahan and Doetsch (Somatosens. Res., 1984, 2: 49-81). The data suggest that both cortical and peripheral somatosensory injuries may produce their effects by unmasking synaptically weak neural connections. Synaptic unmasking may involve facilitation or disinhibition of neurons with convergent properties-especially those located within the heterogeneous subdivisions of SmI cortex. 0 1990 Academic Press, Inc.
0014-4&x%/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form
The somatic afferent system of the raccoon provides a unique model for studying the normal functional organization of somatosensory (SmI) cerebral cortex and for determining possible organizational changes following experimental manipulation. The representation of the forepaw within SmI cortex of this species is enormous relative to that of other mammals. The forepaw cortex is anatomically delineated by a pattern of sulci whereby the focal representation of each digit and the palmar pads occupies a separate and relatively distinct cortical subgyrus (20). Recent microelectrode mapping experiments (9, 10, 11) have shown that each forepaw digit zone appears to contain at least two major subdivisions, one representing the glabrous skin and the other representing the hairy skin and claws. The glabrous skin subdivision of each digit area is large, centrally located, and bounded on three sidesby the smaller hairy skin and claw subdivision. Neurons within the glabrous skin sector typically respond to low-intensity tactile stimulation of the skin and have small receptive fields (RFs) that are somatotopically organized. In contrast, neurons within the hairy skin and claw subdivision respond to movement of hairs, to touch of the hairy skin or claw, and sometimes also to tactile stimulation of glabrous skin; their RFs seem to be larger and are distributed with less somatotopic precision than those of cells in the glabrous skin subdivision. Because of its mixed response properties, the hairy skin and claw sector has been referred to as the “heterogeneous” region by Johnson et al. (10). The results of recent anatomical tracer studies have shown that the two cortical subdivisions have different thalamic (3) and intracortical (2) connections that may largely account for the differences in their physiological properties. The highly developed SmI cortex of the raccoon has been used as a model system to examine the effects of 1 Present address: Clinical Investigation Army Medical Center, Tacoma, WA 98431. 162
Inc. reserved.
Department,
Madigan
EFFECTS
OF
LESIONS
IN
peripheral nerve lesions on the organization of central somatosensory maps (1, 11,12,16). The results of these studies have shown that digital nerve transection-by surgical removal of a digit-causes neurons within the deafferented cortical digit area to develop responsiveness to “new” peripheral inputs. Novel RFs were found primarily on the digits adjoining the site of injury and to a lesser extent on the neighboring palmar pads. The emergence of altered neuronal response properties and new RFs was interpreted as being due primarily to unmasking of preexisting but synaptically ineffective inputs to the deprived cortical zone. This plasticity suggested that similar physiological changes might occur following damage to the cortex itself. Any significant alterations in neuronal responsiveness could provide insight into the mechanisms of functional recovery following cerebral insult. To test this notion-and to develop a physiological model of recovery from focal brain damage-lesions were made in the SmI cortical subgyrus representing the third forepaw digit of adult raccoons. About 4 months later, portions of the SmI forepaw cortex of these animals were studied electrophysiologically to determine whether any functional changes had occurred within cortical regions bordering on the lesion. Comparison of electrophysiological data obtained pre- and postoperatively from lesioned animals with data from normal animals showed that significant functional alterations can occur within cortical zones adjoining the lesion. METHODS
Electrophysiological experiments were performed on 14 adult raccoons (Procyon lotor) of both sexes. Cortical lesions were made in seven of these animals using the technique described below. Possible long-term effects were examined by studying five lesioned animals at about 4 months (15-17 weeks) following surgery; shortterm effects were assessed in two animals at 2 and 4 weeks after the lesion. The results were compared with data obtained from the same animals before the operation and with corresponding data taken from seven normal raccoons. Surgical Procedures Before making the cortical lesion, each animal was sedated with ketamine HCl (Ketaset, 15 mg/kg, im) and anesthetized with a-chloralose (35 mg/kg, ip) or pentobarbital sodium (Nembutal, 30 mg/kg, ip). An endotracheal tube was inserted and the animal was artificially ventilated with a respirator pump. The head was secured in a stereotaxic frame and a craniotomy was performed under antiseptic conditions to expose the SmI forepaw cortex of the right hemisphere. The dura mater was in-
SOMATOSENSORY
163
CORTEX
cised and reflected to provide access to the digit 3 cortical zone and the surrounding areas representing digit 2, digit 4, and the palmar pads. Photographs were taken of the exposed cortex before and after making the lesion. In five of these animals, the boundaries of the digit 3 zone were mapped electrophysiologically using a silver ball electrode to record primary tactile-evoked responses; in the other two animals, mapping was done with a tungsten microelectrode inserted into cortical layers III-IV to record the responses of small clusters of neurons. After the somatotopic borders had been determined, small blood vessels penetrating the cortical tissue of the digit 3 zone were electrocoagulated using bipolar forceps. The crown of the digit 3 subgyrus was then removed by subpial aspiration. The lesion extended anteriorly to the caudal lip of the ascending coronal (lateral central) sulcus and posteriorly to the rostra1 lip of the triradiate sulcus (see Fig. 1A). The lesion was confined to cortical tissue, sparing the underlying white matter as much as possible. After the ablation was completed and all bleeding had been stopped, the cranial defect was repaired by suture of the reflected tissues. The raccoon was given an injection of ampicillin sodium (Omnipen-N, 250 mg, im), returned to the animal care facility, and allowed to recover. Electrophysiological
Recording Procedures
After survival periods of 15-17 weeks (five animals) or 2 and 5 weeks (one animal each), the operated raccoon was studied electrophysiologically. Systematic mapping experiments were done to determine whether any physiological changes had occurred in those cortical regions that border on the lesioned digit 3 area-namely, the zones representing digit 2, digit 4, and the palmar pads. The unoperated raccoons were studied electrophysiologically in the same manner to permit appropriate statistical comparisons to be made. Conventional surgical and electrophysiological recording procedures were used in all experiments. Before each recording session, the raccoon was sedated with ketamine HCl (Ketaset, 15 mg/kg, im) and then anesthetized with a-chloralose (35 mg/kg, ip). The animal was intubated with a tracheal cannula and artificially ventilated with room air. The right femoral vein was cannulated to administer supplementary doses of chloralose and to infuse dextrose-saline to replace lost body fluids; the femoral artery was cannulated to monitor arterial blood pressure. The animal’s head was placed in a stereotaxic frame, and the right SmI forepaw cortex was surgically exposed and photographed. An agar-gauze dam was constructed around the cranial defect and was filled with warm mineral oil to prevent drying of cortical tissue. Bilateral pneumothorax and drainage of cerebro-
164
A
DOETSCH,
JOHNSTON,
B
AND
HANNAN
C
FIG. 1. Photographs of a portion of the exposed SmI forepaw cortex before (A), immediately after (B), and about 16 weeks after (C) a lesion was made in the digit 3 subgyrus. (A) The cortical representational zones for digits 2,3, and 4 (2-4) and the palmar pads (P) with respect to the triradiate (T) sulcus and the ascending branch of the coronal (C) sulcus.
spinal fluid from the cisterna magna were used to minimize movements of the brain due to respiration or arterial pulse. Core body temperature was maintained at 3738°C with an abdominal dc heating pad, servoregulated with a rectal thermometer. Tungsten microelectrodes were used to record the extracellular responses of small clusters of neurons to tactile stimulation of the digits, palmar pads, and other forepaw regions. The microelectrode was advanced into the cortex using a depth-calibrated hydraulic microdrive. All penetrations were made as normal to the cortical surface as possible, to a depth of at least 2.0 mm. The neural signals were fed into a preamplifier and audiomonitor and displayed on an oscilloscope. All recording sites were marked on the photographs of the cortex. The composite RFs of small clusters of neurons were mapped at a standard suprathreshold stimulus intensity of 5 g with a calibrated nylon filament probe. The RFs were drawn on pictures of the forepaw; the area of each field was later measured with a planimeter and expressed in absolute planimetric units. The submodality sensitivity of each neuronal cluster was determined using conventional criteria: (i) skin touch-response to light touch of glabrous or hairy skin with a nylon filament probe; (ii) hair movement-activation by deflection of small groups of hairs with light air puffs or gentle stroking; (iii) claw touch-response to touch or light tap of a claw with a nylon probe; and (iv) deep pressure/joint movement-activation by steady pressure on skin over-
lying deep tissues/response to changes in joint position. While recording in each electrode track, care was taken to note clear changes in RF location, size, skin type, or submodality sensitivity of neuronal clusters with cortical depth. At the end of each experiment, the animal was sacrificed with an intravenous injection of sodium pentobarbital and perfused through the heart with 0.9% saline followed by 10% formalin-saline. The brain was then removed and stored in 10% formalin-saline. Blocks of tissue containing the lesioned area or the recording sites were later embedded in paraffin, serially sectioned, and stained with hematoxylin-eosin or 1~x01 fast blue. RESULTS
The results obtained by comparing pre- and postlesion data from the operated raccoons and the results of comparing data from different groups of normal and lesioned animals were consistent with one another. The findings indicated that damage restricted to the digit 3 cortical zone produced changes in the physiological characteristics of the zones bordering on the site of injury. Clusters of neurons were found to be activated from more widespread regions of the forepaw digits and palmar pads; the expansion of composite RFs was accompanied by greater mixtures of skin type and by decreased somatotopic and submodality specificity.
EFFECTS
OF
LESIONS
IN
Histology The surface morphology of the forepaw cortex was documented photographically before and immediately following the ablation, and again at the time of electrophysiological study, 2, 5, or 15-17 weeks later (Fig. 1). Just after the surgery, the edges of the lesion were sharp and easily distinguished by the margins of intact cortical tissue. After 15-17 weeks, the border regions had become smooth and partially revascularized, making the edges of the lesion more difficult to identify. Microscopic examination of the histological material revealed a lesioned area with extensive loss of cortical tissue that was surrounded by a region of gliosis containing few neurons; the latter region bordered directly on cortical zones exhibiting normal cytoarchitectonic features. Comparison of the histological data with previously defined electrophysiological cortical maps showed that the ablations were confined to the digit 3 zone and were subtotal in extent. In all cases, the cortical tissue that was removed constituted the representation of approximately the distal one-half of the glabrous skin of digit 3. Small portions of the anterior heterogeneous subdivision, representing the hairy skin and claw of digit 3, were also included in the lesion. The cortical ablation spared the more proximal glabrous digit 3 representation located within the banks of the triradiate sulcus. Comparison of Pre- and Postlesion Data from Individual Animals The most direct method for assessing the lesion effects was to compare the response properties of neuronal clusters within a cortical zone before and after surgery in the same animal. The results of such analyses are illustrated in Figs. 2 and 3 for two raccoons studied before and about 16 weeks after the digit 3 area had been lesioned. Recordings taken from nearly identical or closely adjacent electrode sites at comparable cortical depths (7501250 pm) were selected for comparison. The results indicated that the functional characteristics of neurons in a few tracks had remained relatively stable, with only minor changes in RF location, size, skin type, and submodality sensitivity (Fig. 2, track 3A, Fig. 3, tracks 7A and lOA). However, the vast majority of electrode penetrations revealed definite alterations in neuronal responsiveness. The most prominent change was the development of responses to stimulation of previously ineffective skin regions of the forepaw. This was reflected in the expansion of composite RFs to include portions of one or several digits or palmar pads that were outside the borders of the prelesion RF. Of the 22 pairs of RFs shown in Figs. 2 and 3, fully 17 increased in size, four decreased in size, and one remained unchanged after the lesion. In most cases, stimulation of the original RF still elicited the strongest “multiunit” response,
SOMATOSENSORY
CORTEX
165
whereas stimulation of the newly acquired, postlesion component of the RF produced a weaker response (fewer multiunit spikes). The novel inputs often originated from digit 3, but frequently involved other digits or the pads, depending on the site of recording. Furthermore, the postlesion RFs typically included greater mixtures of skin type, with varying combinations of glabrous or hairy skin and claws forming part of the RF. Consistent with this heterogeneity, many clusters of neurons also exhibited increased mixtures of submodality sensitivity to skin touch, hair deflection, or claw touch. Finally, the functional alterations observed within the digit and pad zones adjoining the lesion tended to degrade the normal somatotopic order and response specificity of these zones. Comparison
of Data from Normal and Lesioned Animals
It was not feasible to obtain extensive electrophysiological data from any one animal prior to making a lesion, due to the time required for mapping the large cortical zones under study and the need for restricting the duration of anesthesia and surgery. Consequently, the pre- and postlesion analysis of individual raccoons was supplemented by statistical comparison of data obtained from approximately matched electrode tracks in separate groups of normal and operated animals. Definitions
and Criteria
To ensure the validity of the statistical procedures, strict definitions of the cortical zones and rigorous criteria for measuring various neuronal response properties were adopted. Cortical representational zones. The borders of the three cortical zones adjoining the digit 3 area were defined by previously determined electrophysiological maps and their relationship to the surface pattern of the triradiate sulcus in normal animals. The digit 2/digit 3 border was found to be demarcated consistently by the anterolateral branch of the descending triradiate sulcus. The pad/digit 3 border corresponded to the major central portion of the triradiate sulcus. In contrast, the boundary between the digit 4 and digit 3 zones was less well delineated, typically related only to a small anteriorly oriented spur of the triradiate sulcus. Consequently, this border was defined solely by electrophysiological criteria in both groups of animals-recording sites were considered to be within the digit 4 or the digit 3 zone depending on whether the strongest responses first recorded in a penetration were elicited by stimulation of digit 4 or digit 3, respectively. This electrophysiologitally defined border usually fell along an imaginary line extending anteromedially from the tip of the triradiate sulcal spur located between the two digit areas.
DOETSCH,
JOHNSTON,
AND
HANNAN
B DIGIT
2
DIGIT
PAD
4
12
FIG. 2. Comparison of the receptive fields (RFs) of neurons recorded before and about 16 weeks after a lesion was made in the digit 3 cortical zone of the same animal. (A) The cortical surface map shows the locations of recording sites 1-12 (filled circles) before the lesion (shaded area) and sites lA-12A (open circles) after the lesion. (B) The RFs of neurons recorded from pairs of closely adjacent sites shown in A are compared. Prelesion RFs are indicated by numbers and are represented on the paws by stippling; comparable postlesion RFs are indicated by corresponding numbers followed by letter A and are represented by hatched lines. Note the considerable increase in size of most RFs after the lesion.
Responsive tracks and sites. A responsive electrode track was defined simply as a track containing at least one neuronal cluster that could be activated by periphera1 stimulation. A responsive site was designated as a
region with a specified depth interval within a responsive track containing neuronal clusters with very similar RFs and physiological properties; responsive sites were defined as different when regions of nonoverlapping
EFFECTS
DIGIT
2
DIGIT
4
OF
LESIONS
IN
DIGIT
SOMATOSENSORY
4
167
CORTEX
PAD
6
FIG. Symbols
3. Comparison of the receptive fields (RFs) of neurons recorded before and about 16 weeks after a lesion in the digit in A and B have the same meaning as those in Figs. 2A and 2B. Note the increase in RF size following the lesion.
depths contained neurons with obviously different characteristics. Primary and secondary responses. To quantify the strength of neuronal activation by peripheral stimulation, the responses were divided into two major categories. The strongest neuronal response produced by stimulation of a particular digit or the palmar pads was defined as primary. Weaker responses elicited by stimulation of other digits or the pads were designated as secondary.
3 cortical
zone.
On-focus and off-focus responses. The responses were subdivided also in terms of the locations of neuronal RFs. Responses recorded from a predesignated cortical zone (e.g., the digit 2 zone) were defined as on-focus if they were elicited by stimulation of the digit or pads (e.g., digit 2) “represented” within that zone. Responses were designated as off-focus if they were produced by stimulation of a digit or the pads (e.g., digit 3) not “represented” within that zone. Thus, the dominant pattern of neuronal activation found within a normal cortical zone
168
DOETSCH,
TABLE
JOHNSTON,
1
Number of Electrode Tracks and ResponsiveSites in Three SmI Cortical ZonesStudied
Cortical zone
Experimental animal”
Digit
2
Digit
4
Pad
Number of tracks 71 56 84 82 91 75
Normal Lesioned Normal Lesioned Normal Lesioned
“Data are from seven 15-17 weeks after surgery.
normal
and
five
Number/ percentage of responsive tracks
Number of responsive sites
71/100 56/100 84/100 77/93.9 91/100 74/98.7
85 86 101 108 128 132
lesioned
animals
studied
consisted of a primary on-focus response that was occasionally associated with a secondary off-focus response. Responsive Neuronal Clusters within Cortical Zones In both normal and lesioned raccoons, the vast majority of electrode tracks made within the cortical zones for digit 2, digit 4, and the palmar pads contained neurons that responded to cutaneous stimulation of the forepaw (Table 1). The RF locations and other response properties of neuronal clusters in any one track generally were similar as a function of depth below the pial surface. The number of responsive sites within a track showing obvious changes in RF location, size, or other characteristics averaged less than 2.0, but was consistently greater in the cortical zones of the lesioned animals. Functional Organization of Cortical Zones in Normal and Lesioned Animals Cortical surface maps were constructed for each animal to show the location of each electrode track and to summarize the major RF properties of the neuronal clusters encountered at each distinct responsive site within a given track. In Figs. 4-6, the maps obtained from three lesioned raccoons are compared with corresponding maps from three normal raccoons. The data shown by these maps are consistent with the results obtained before and after ablation of the digit 3 zone of individual animals. In normal raccoons, the composite RFs of neuronal clusters within the cortical zones for digit 2, digit 4, and the palmar pads were relatively small, typically confined to the appropriate digit or pad, and usually involved only one skin type. Each of the cortical areas representing a digit consisted of two major subdivisions-a posterior sector containing neurons with RFs on glabrous skin and a more anterior heterogeneous sector containing neurons with RFs on hairy skin and claws or combinations of skin type (Figs. 4A, 5A,
AND
HANNAN
and 6A). In the lesioned raccoons studied 15-17 weeks following surgery, neurons within the cortical digit and pad zones were found to receive novel inputs from the forepaw in addition to their normal afferent drive from the appropriate digits or pads. Neuronal RFs were larger, often including several digits or pads, and involved greater mixtures of different skin types (Figs. 4B, 5B, and 6B). Within the two digit zones studied, the border between the glabrous skin subdivision and the heterogeneous subdivision was considerably less distinct. Furthermore, neuronal clusters within the pad zone responded to stimulation of digits more often than normal. Finally, the lesioned animals showed more variability in RF location, size, and skin type as a function of horizontal and vertical distance through the cortex. This was evident from the more abrupt changes in RF properties of neuronal clusters within adjacent electrode tracks and from the more numerous changes at different depths within the same track. In short, neurons within the cortical areas adjoining the lesion had larger RFs with novel components; i.e., they were activated by new inputs from the forepaw. Most of the novel inputs evoked relatively weak secondary responses and originated from “inappropriate” offfocus digits or pads. These secondary off-focus responses typically coexisted with vigorous primary responses produced by stimulation of the appropriate on-focus digit or palmar pads. Statistical Comparisons of Data Obtained from Normal and Lesioned Animals The specific response characteristics measured in the normal and lesioned animals were compared statistically and are summarized below (see Table 2). Unless stated otherwise, all statistics were based on the absolute number of responsive sites exhibiting a particular functional neuronal property. Primary and secondary responses. By definition, 100% of the responsive sites within a given cortical zone in both groups of raccoons contained neuronal clusters that gave strong primary responses to stimulation of a specific digit or the palmar pads. Sites with weaker secondary responses to peripheral stimulation were relatively rare in normal animals, especially within the cortical zones for digit 2 (3.5%) and digit 4 (14.9%); secondary responses were more common in the pad zone (35.2%), indicative of greater variability in the normal inputs to this zone (Fig. 7). In contrast, each of the three cortical zones in the lesioned animals had significantly more sites with secondary responses (x2 tests, P < O.OOl), by factors ranging from about 2 to 8, than the same zones in normal animals (Fig. 7, Table 2). This finding dramatically reflects the appearance of novel physiological inputs following ablation of the digit 3 area.
EFFECTS
’
2mm
OF
LESIONS
IN
SOMATOSENSORY
CORTEX
169
’
FIG. 4. Cortical surface maps comparing the representation of the receptive fields (RFs) of neurons recorded in three cortical zones of a normal animal (A) and an animal with a digit 3 cortex lesion made about 16 weeks earlier (B). The shaded area in B shows the approximate extent of the lesion. Each circle in A and B represents an electrode penetration in which responsive neurons were recorded, each X represents a track containing unresponsive neurons. The labels in each circle designate the RF location of neurons within the penetration in order from left to right of their sensitivity to cutaneous stimulation: l-5, digits l-5; P, palmar pads; F, entire forepaw. Various shadings refer to different types of skin included within a RF: white, glabrous skin; stippled, hairy skin and/or claws; black, mixtures of glabrous and hairy skin and/or claws. The subdivided circles indicate tracks in which neurons at different depths showed shifts in the location of RFs to include different digits and pads or different skin types. The sequence of changes with increasing depth is shown within a circle from top to bottom or clockwise from upper left to lower right. Note that neurons in the lesioned animal had considerably larger RFs but greater variability in peripheral location and skin type involved, indicative of new inputs but less somatotopic precision and submodality specificity than in the normal animal.
On-focus and off-focus responses. In both groups of raccoons, the primary inputs typically originated from the appropriate on-focus digit or pads (SO-100%) and the secondary inputs arose from neighboring off-focus digits or pads (90-100%). In the lesioned animals, significant changes in the distribution of the foci-particularly increases in the number of primary off-focus responseswere observed in the digit 4 zone and the pad zone (x2 tests, P < 0.03), but not in the digit 2 area (Table 2). No significant differences were found in the foci of secondary responses within any zone, partly because of the low number of secondary inputs present in normal animals. Receptive field location. In general, the distributions of the particular digits or pads on which the primary and secondary RFs were located were similar in both sets of raccoons. The vast majority of RFs associated with primary responses involved the appropriate digits (94100%) or pads (76-80%). Only the primary responses within the digit 4 area of the lesioned animals were associated with significantly more inappropriate RF locations (x2 test, P < 0.005), mainly on digit 3 or the pads (Table 2). No significant differences in secondary RF locations were found.
Receptive field size. The composite RFs of neuronal clusters in the lesioned raccoons typically were much larger than those in normal raccoons. Significant increases were found in the size of the primary RFs of neurons within the digit 2 zone (t test, P < 0.001) and the digit 4 zone (t test, P < 0.003), but not in those of neurons within the pad zone (Fig. 8A, Table 2). The secondary RFs were significantly larger for the digit 4 area (t test, P -C 0.04) but not for the other two zones (Fig. 8B, Table 2). When the neuronal RFs were not subdivided into primary and secondary components, but were treated together as whole units, all three cortical zones were found to have larger RF areas in the lesioned raccoons (t tests, P < 0.003). The distributions of these “combined” RF sizes and their mean values are shown for each cortical area in Fig. 9. The primary Skin type and submodality sensitivity. responses within the digit 2 and digit 4 zones involved greater mixtures of skin type (x2 tests, P < 0.001) and associated submodality sensitivity (x2 tests, P < 0.001) in the lesioned animals than in the normal animals (Table 2). Neuronal responsiveness shifted from purely glabrous skin regions of a digit to various combinations of glabrous skin, hairy skin, or claws. Likewise, submodal-
170
DOETSCH,
JOHNSTON,
AND
HANNAN
A
FIG. 5. Cortical surface maps normal animal (A) and an animal as those in Figs. 4A and 4B. Note and submodality precision than in
comparing the representation of the receptive fields (RFs) of neurons recorded in three cortical zones of a with a digit 3 cortex lesion made about 16 weeks earlier (B). All symbols in A and B have the same meaning that neurons in the lesioned animal had larger RFs, indicative of new responsiveness but less somatotopic the normal animal.
ity sensitivity changed from predominantly skin touch to mixtures of skin touch, hair deflection, or claw touch. As expected, the pad zone showed no comparable differences between the two groups of animals since this zone represents the glabrous surface of the palm and borders
on the glabrous subdivisions of adjacent digit zones. In contrast, the secondary responses were associated with increased mixtures of skin type (x2 test, P < 0.04) and submodality sensitivity (x2 test, P < 0.03) only within the digit 4 area (Table 2).
A
FIG. 6. Cortical surface maps normal animal (A) and an animal as those in Figs. 4A and 4B. Note and submodality precision than in
comparing the representation of the receptive fields (RFs) of neurons recorded in three cortical zones of a with a digit 3 cortex lesion made about 16 weeks earlier (B). All symbols in A and B have the same meaning that neurons in the lesioned animal had larger RFs, indicative of new responsiveness but less somatotopic the normal animal.
EFFECTS
OF
LESIONS
IN
SOMATOSENSORY
TABLE
171
CORTEX
2
Statistical Comparisons(P Values”) of Data Obtained from Normal and LesionedAnimals Digit
Response
Primary responses
property
Number of primary and secondary responses On versus off focus responses Receptive field location Receptive field area Skin type Submodality sensitivity
2 zone
Digit
Secondary responses
Primary responses
1.0
