JOURNALOF NEUROPHYSIOLOGY Vol. 68, No. 2, August 1992. Printcld
in U.S.il.
Serial and Parallel Processing of Tactual Information in Somatosensory Cortex of Rhesus Monkeys T. P. PONS, P. E. GARRAGHTY, AND M. MISHKIN Laboratory ofNeuropsychology, National Institute ofMental SUMMARY
AND
CONCLUSIONS
1. Selective ablations of the hand representations in postcentral cortical areas 3a, 3b, 1, and 2 were made in different combinations to determine each area’s contribution to the responsivity and modality properties of neurons in the hand representation in SII. 2. Ablations that left intact only the postcentral areas that process predominantly cutaneous inputs (i.e., areas 3b and 1) yielded SII recording sites responsive to cutaneous stimulation and none driven exclusively by high-intensity or “deep” stimulation. Conversely, ablations that left intact only the postcentral areas that process predominantly deep receptor inputs (i.e., areas 3a and 2) yielded mostly SII recording sites that responded exclusively to deep stimulation. 3. Ablations that left intact only area 3a or only area 2 yielded substantial and roughly equal reductions in the number of deep receptive fields in SII. By contrast, ablations that left intact only area 3b or only area 1 yielded unequal reductions in the number of cutaneous receptive fields in SII: a small reduction when area 3b alone was intact but a somewhat larger one when only area 1 was intact. 4. Finally, when the hand representation in area 3b was ablated, leaving areas 3a, 1, and 2 fully intact, there was again a substantial reduction in the encounter rate of cutaneous receptive fields. 5. The partial ablations often led to unresponsive sites in the SII hand representation. In SII representations other than of the hand no such unresponsive sites were found and there were no substantial changes in the ratio of cutaneous to deep receptive fields, indicating that the foregoing results were not due to longlasting postsurgical depression or effects of anesthesia. 6. The findings indicate that modality-specific information is relayed from postcentral cortical areas to SII along parallel channels, with cutaneous inputs transmitted via areas 3b and 1, and deep inputs via areas 3a and 2. Further, area 3b provides the major source of cutaneous input to SII, directly and perhaps also via area 1. 7. The results are in line with accumulating anatomic and electrophysiological evidence pointing to an evolutionary shift in the organization of the somatosensory system from the general mammalian plan, in which tactile information is processed in parallel in SI and SII, to a new organization in higher primates in which the processing of tactile information proceeds serially from SI to SII. The presumed functional advantages of this evolutionary shift are unknown.
Health, Bethesda, Maryland
20892
to collectively as the postcentral strip. Another region of cortex that processes somatic information is located in the depths of the lateral sulcus and is known as the second somatosensory area or SII (Friedman et al. 1986; Robinson and Burton 1980; Woolsey and Fairman 1946). Until recently, the postcentral somatosensory strip and SII were both thought to receive dense projections directly from the ventroposterior nucleus (VP) of the thalamus in all mammalian species (Burton 1984; Jones and Powell 1970; Mountcastle 1984). Because of this putative dual projection pattern, it had long been assumed that the processing of tactile information proceeded in parallel across these two cortical regions. That assumption was brought into question, however, by evidence demonstrating that VP in macaques provides only a sparse projection onto neurons in SII (Friedman and Murray 1986; Krubitzer and Kaas 1990; Manzoni et al. 1984). Direct evidence that the postcentral strip and SII of macaques do not process somatic inputs in parallel came from combined ablation and recording experiments, which demonstrated that elimination of all four hand representations in the postcentral strip (i.e., areas 3a, 3b, 1, and 2) eliminated the hand representation in SII, whereas removal of SII had no detectable effect on the activation of neurons in the postcentral strip (Pons et al. 1987a). The results of these experiments in monkeys thus provided compelling evidence that SII, whose middle layers receive a direct input from all four areas in the postcentral strip, depends on one or more of those inputs for its activation. Of the four cytoarchitectonic areas that comprise the postcentral strip, two, areas 3b and 1, process predominantly cutaneous inputs, whereas the other two process inputs from receptors that are located primarily in “deep” body tissues such as muscle afferents (area 3a) and joints (area 2) (Krishnamurti et al. 1976; Merzenich et al. 1978; Mountcastle and Powell 1959; Pons et al. 1985; Powell and Mountcastle 1959a,b). The purpose of the present study was to determine the differential contribution, if any, of the four areas comprising postcentral cortex to the responsivity and modality properties of neurons in SII of macaques. A preliminary report of the findings has been presented elsewhere (Pons et al. 1988a).
INTRODUCTION
Primary somatosensory cortex in macaques, located anteriorly in the parietal lobe, consists of four cytoarchitectonitally distinct divisions, areas 3a, 3b, 1, and 2. These areas, each of which processes different components of somatic sensation (see Kaas and Pons 1988 for review), are referred
METHODS
Subjects Ten adult female monkeys (Macaca kg were used in the study.
mdatta)
weighing
3.0-6.0
518
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS IN MACAQUES
Surgery Monkeys were anesthetized with an initial dose of ketamine hydrochloride ( 15 mg/ kg) followed by pentobarbital sodium, which was administered intravenously as required (20-30 mg/ kg). Throughout surgery, which was performed using aseptic precautions, the animals received an intravenous drip of a solution of 5% dextrose-0.45% sodium chloride, and their heart rate, respiratory rate, and temperature were monitored and maintained within normal limits. Ablations of cortical areas were made by aspiration. All ablations were limited to the portions of the cortical area(s) representing the hand as defined by previous studies (Nelson et al. 1980; Pons et al. 1985, 1987b), although to insure the complete removal of the hand zone in any given area the ablations were extended > 1 mm past that zone’s estimated mediolateral and rostrocaudal borders. There are 14 possible combinations of subtotal ablations of the four areas comprising the postcentral strip, of which we performed 8, as summarized in Table 1. The combinations not used, particularly lesions of single areas, were deemed to be unnecessary or relatively uninformative. When areas 3a or 3b, both of which are entirely buried in the posterior bank of the central sulcus, were to be ablated without involvement of other areas in the postcentral bank (i.e., areas 3b or 1 ), a part of area 4 in the anterior bank of the central sulcus was removed to gain access to those buried areas. As a control for this procedure, one hemisphere was prepared with a removal limited to area 4. One unoperated hemisphere served as an additional control. A recording chamber and head fixation device were attached to the skull in a separate procedure. The chamber was positioned to provide maximum access to both SII and the hand representations in the postcentral strip, and the bone under the chamber was removed to expose the dura. All animals received an antibiotic as a prophylactic measure after each surgical procedure.
Recording procedures Six to eight weeks after the cortical ablations, the animals were anesthetized with ketamine hydrochloride ( 15 mg/ kg), intubated with an endotracheal tube coated with lidocaine, and then anesthetized with 2.0% halothane in a mixture of NZ02 and O2 gas. After attachment of the animal’s head bolt to a stereotaxic frame, the top of the previously implanted recording chamber was removed and electrodes were inserted through the dura into the cortex. The mapping and recording procedures were similar to those used in previous studies (Pons et al. 1985, 1987a,b, 1988b,c). Briefly, electrode penetrations were placed at 0.5~mm intervals across the mediolateral extent of SII and at LO-mm intervals across its rostrocaudal extent. Typically, 40-60 electrode penetrations were made through SII in each hemisphere studied. Microelectrodes were hydraulically advanced through parietal cortex, white matter, and then SII or surrounding cortex until single- or
TABLE
1.
Areas ablated
Postcentral
Precentral
No. of Hemispheres
None None 3a, 3b, 1 3a, 3b, 2 3a, 1, 2 3b, 1, 2 3a, 3b 3b, 1 1, 2 3b
None 4 None 4 4 None 4 None None 4
1 1 1 1 2 2 2 2 2 2
519
multiple-unit responses to mechanical stimulation of the body were isolated. Small marker lesions ( 10 PA for 10 s) were placed in cortex to assist in locating the recordings sites postmortem. If hair movement was insufficient to elicit a neuronal response, or if the receptive field was on or beneath the glabrous skin, calibrated Von Frey hairs were used as stimuli to allow quantification of the force required to elicit a consistent neuronal response. If a stimulus force of ~0.5 g was sufficient to evoke a vigorous neuronal response, the receptive field at that site was defined as cutaneous. If a stimulus force of 22.0 g was required, or if a high-intensity stimulation produced by squeezing fingers, bending joints, or kneading muscles was necessary to activate neurons, the receptive field at that site was defined as deep. Although recording sites with deep receptive fields were, by definition, unresponsive to cutaneous stimulation, the converse was not true, i.e., recording sites with cutaneous receptive fields may also have had inputs from deep receptors, but this could not be ascertained using our methodology. If any unit in a multiunit response responded to a cutaneous stimulus, that site was defined as “cutaneous.” Only 4 out of nearly 3,000 recording sites in SII required a stimulus force of 0.5-2.0 g to elicit a response; these four sites were dropped from the data analysis.
Histological
analysis
On completion of the recording, animals were given a lethal dose of pentobarbital and then perfused intracardially with 0.9% saline followed by either 10% formalin or 2% paraformaldehyde and 0.2% glutaraldehyde. The latter perfusate was used in brains reacted for cytochrome oxidase. Brains were cut in the coronal or parasagittal plane on a freezing microtome, and every fourth section was mounted and stained for Nissl substance or cytochrome oxidase to assess the placement of recording tracks, marker lesions, and cortical ablations. Recording sites in SII were identified by reconstructing electrode tracks and marker lesions as detailed elsewhere (Pons et al. 1988b,c). Analysis of cortical sections confirmed that all ablations were as intended (Figs. 1 and 2). The pattern of thalamic degeneration caused by the cortical ablations was restricted in each case to the portion of ventroposterolateral nucleus (VPL) immediately adjacent to ventroposteromedial nucleus (VPM), the portion known to contain the representation of the hand (Dykes et al. 198 1; Kaas et al. 1984; Mountcastle and Henneman 1952). RESULTS
Recordings in SII were made at 2,834 sites in 16 hemispheres of the 10 monkeys (see Tables 1 and 2). Control hemispheres The distribution of SII recording sites responding to cutaneous versus deep stimulation in the normal control hemisphere is shown in Fig. 3. A special effort was made in this case to sample extensively and equally from the hand representation and the representation of all other body parts combined. Both subdivisions contained more than twice as many cutaneous as deep receptive fields (although, as indicated above, some cutaneous receptive fields may also have had deep inputs). These results are in accord with those from previous studies of SII in macaques (Friedman et al. 1986; Pons et al. 1987a; Robinson and Burton 1980). The findings from the case with the removal of a portion of the anterior bank of the central sulcus (caudal area 4) directly across from the postcentral hand representations
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
T. P. PONS, P. E. GARRAGHTY,
AND M. MISHKIN
FIG. 1. Top left: lateral brain view with cytoarchitectonic areas 1, 2, and 5 numbered. Diagonally oriented dashed lines: location of borders between areas. Horizontal dashed lines: approximate borders of the postcentral hand representation. The 3 solid vertical lines numbered 1, 2, and 3 represent the approximate anteroposterior location of the coronal sections shown in the corresponding numbered columns below. The 2 vertical arrows along the lateral sulcus indicate the location of the section shown at the top right. The coronal section illustrated in the top right indicates the location of SII buried in the lateral fissure. The 2 solid straight lines show the angle of electrode penetrations into SII. A: reconstructions of 3 sections showing the locations of areas 4,3a, 3b, 1, and 2 from an unoperated hemisphere. B: 3 coronal sections across the anteroposterior extent of a lesion of the hand representations in areas 3a, 3b, and 2, that included some of area 4 (see METHODS) but that spared the area 1 hand representation. C: 3 coronal sections across the anteroposterior extent of a lesion of the area 3a, 1, and 2 hand representations, which included additional removal of some of area 4 (see METHODS), but spared the area 3b hand representation. D: 3 sections across the anteroposterior extent of a lesion of the area 3b, 1, and 2 hand representations that spared area 3a. In B, C, and D some tissue from areas 3b, 1, and 2 is shown from the representations of other body parts such as the face or arm. ce, central sulcus, pc, postcentral sulcus, ip, intraparietal sulcus, la, lateral sulcus.
law
1
iP
la I
la
a
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS IN MACAQUES
521
focus all electrode penetrations within the SII hand representation, the zone of SII most likely to be affected by the ablation. Again, the results revealed over twice as many cutaneous as deep receptive fields in the SII hand representation. In all other respects as well, the results in this case were essentially the same as those in the unoperated hemisphere. Hemispheres
with selective removals
Selective ablations of the areas comprising the postcentral strip were of three types. In the first, we removed two areas, one processing cutaneous inputs and another processing deep inputs, i.e., areas 1 and 2 were removed and areas 3a and 3b were spared, or the reverse. Also, we removed both areas processing predominantly cutaneous inputs, areas 3b and 1, and left both areas 3a and 2 intact. In the second type of ablation, we removed three cortical areas, leaving a single area intact to process either cutaneous (3b or 1) or deep (3a or 2) stimulation. Finally, in the third type, we removed a single area (3b) to compare the effect of removing it alone with that of preserving it alone. Ablations of two areas
FIG. 2. A: photomicrograph of a coronal section taken through the postcentral hand representations in an unoperated animal, stained for cytochrome oxidase, and shows the location of areas 3a, 3b, 1, and 2. B: photomicrograph of a coronal section through the postcentral hand representations ofa hemisphere that received a lesion ofthe hand representation in areas 3b, 1, and 2. Note the sparing ofarea 3a at the fundus ofthe central sulcus as indicated by the arrow. C: photomicrograph of a section taken through the normal location of the postcentral hand representations in a hemisphere in which only the area 3b hand representation was spared and the area 3a, 1, and 2 hand representations were removed. Note that despite ablating the areas immediately surrounding area 3b, thus creating an island of area 3b, the tissue remained healthy and did not suffer from infarcts due to an altered blood supply. Scale = 0.75 mm. Other conventions as in Fig. 1.
are shown in Fig. 4. As indicated earlier, this hemisphere was a control for those in which this same region had to be removed to gain access to ablate areas 3a, 3b, or both, without damaging area 1. In this case, an effort was made to
Combined ablation of cortical areas 1 and 2, which left intact area 3b for processing cutaneous inputs and area 3a for deep inputs, altered the ratio of cutaneous to deep receptive fields recorded across the SII hand representation (Fig. 5, Table 2). As in the control hemispheres, recording sites responsive to cutaneous stimulation far outnumbered those responsive to deep stimulation. In contrast to the control hemispheres, however, the ratio of cutaneous to deep recording sites increased and a number of recording sites in the SII hand representation were unresponsive to somatic stimulation. The results indicate that, in the absence of areas 1 and 2, areas 3a and 3b are capable of relaying substantial cutaneous and deep receptor information directly to SII, though not the entire amount it normally receives. Qualitatively, also, the deep responses in the SII hand representation differed from normal in that most were of the slowly adapting type, with receptive fields located directly over muscle bellies rather than on hairy skin or on glabrous pads. This feature is suggestive of selective input from area 3a, though quantification of this observation was beyond the scope of the present study. Recordings in SII body part representations other than of the hand, like the recordings in the SII hand representation, revealed essentially the same ratio of cutaneous to deep receptive fields as in the control hemispheres. Ablation of areas 3a and 3b (and a portion of area 4), again leaving one postcentral cortical area intact for processing cutaneous inputs (area 1) and another for deep inputs (area 2), yielded a result different from the one just described (Fig. 6, compare with Fig. 5 ) . First, there were somewhat fewer recording sites in the expected location of the SII hand representation that were driven by any type of somatic stimulation delivered to the hand. And second, although both cutaneous and deep receptive fields were found, recording sites with deep receptor input actually
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
522
T. P. PONS,
TABLE
2.
P. E. GARRAGHTY,
AND
M.
MISHKIN
Frequency ofcutaneous and deep recording sites in SII Other
Hand Figure
Ablated
3 4 5 6 7 8 9 10 11 12
Spared
None 4 1, 2 3a, 3b 3b, 1 3b, 1, 2 3a, 3b, 1 3a, 3b, 2 3a, 1, 2 3b
3a, 3b, 1, 2 3a, 3b, 1, 2 3a, 3b 1, 2 3a, 2 3a 2 1 3b 3a, 1, 2
Cutaneous
Deep
C/D
125 71 63 20 25 25 0 15 61 19
46 33 12 37 164 210 30 2 0 87
2.72 2.15 5.25 0.54 0.15 0.12 n/a 7.5 n/a 0.22
Ablated, postcentral cytoarchitectonic areas removed; Spared, postcentral sites; Other, number of recording sites for body parts other than the hand.
outnumbered those with cutaneous input. Thus, whereas cortical areas 1 and 2 receive from VPL and relay to SII at least some cutaneous and deep receptor information, respectively, area 1 (in the absence of area 3b) is much less effective in providing SII with cutaneous information than is area 2 (in the absence of area 3a) in providing SII with deep receptor information (see Fig. 5). In SII representations of body parts other than the hand, the encounter rate of cutaneous and deep receptive fields was normal, and cutaneous receptive fields were again the rule. Ablation of areas 3b and 1, leaving intact the two areas processing predominantly deep inputs ( areas 3a and 2)) also led to a dramatic reversal in the ratio of cutaneous to deep receptive fields in the SII hand representation (Fig. 7 and Table 2). Indeed, very few recording sites in the SII hand representation were responsive to cutaneous stimuli in these cases. Recordings in SII outside of the hand representation, by contrast, revealed the normal ratio of cutaneous to deep receptive fields.
cytoarchitectonic
Unresponsive 0 0 42 41 71 42 35 76 47 51 areas left intact;
C/D,
Cutaneous
Deep
C/D
117 8 62 130 139 128 62 81 112 73
58
2.02 8 2 1.48 2.62 2.56 2.95 1.53 1.49 1.74
31 88 53 50 21 53 75 42
ratio of cutaneous
-
to deep recording
Ablations of’three areas Having established that either area 3a or 2 in conjunction with at least one other postcentral area was capable of activating a substantial number of SII recording sites with deep receptor input, we sought to determine whether either of these areas alone would be sufficient for such activation. Ablation of areas 3b, 1, and 2, leaving intact only area 3a, still allowed activation of a large number of recording sites in SII by high-intensity stimulation (Fig. 8). Although there were some SII recording sites with cutaneous receptive fields, these fields were generally located at or very near a joint and responded in a slowly adapting fashion to the flexing of a finger, much like cutaneous sites that activate area 3a of normal animals. Thus area 3a alone is capable of supplying substantial deep receptor input to SII, as well as a limited amount of slowly adapting cutaneous input. Recordings in SII representations of body parts other than the hand revealed the usual high ratio of cutaneous to deep receptive fields.
140
80
120 8 co
70
.= 100 F
e
80
5 2
60
z 5 #
40 20 0
Cutaneous
“Deep”
10
Hand 0
FIG. 3.
Distribution of cutaneous and deep receptive fields for SII recording sites in a normal hemisphere. All recording sites in SII were found to be responsive to somatic stimulation in this case. Hand, receptive fields located on the dorsal or glabrous surface of the hand; Other, receptive fields on all body parts other than the hand (face, arm, trunk, leg, foot, tail, etc.) For definition of cutaneous and deep receptive fields see METHODS. See Table 2 for number of unresponsive SII recording sites for each type of lesion and for the ratio of cutaneous to deep responses. Data in this case were collected in 2 recording sessions over a 4-day period.
“Deep”
Cutaneous
Other FIG. 4. Distribution of cutaneous and deep receptive fields for SII recording sites in a hemisphere in which a portion of the rostra1 bank of the central sulcus (caudal area 4) immediately across from the hand representations in the postcentral strip was removed. All SII recording sites were responsive to somatic stimulation of the body. No effort was made to record from body part representations outside of the hand in this case. Other conventions as in Fig. 3.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS IN MACAQUES
523
180160-
60
3 1405 120P ‘5 loo8 _
z s 20 # 10
g pe
80-
z 6
60
20
n Cutaneous
“Deep”
Hand
Cutaneous
“Deep”
Other
FIG. 5. Distribution of cutaneous and deep receptive fields for SII recording sites after removal of the hand representation in areas 1 and 2. Within the expected location of the SII hand representation, most, but not all, recording sites were responsive to somatic stimuli (cutaneous or deep). Other conventions as in Fig. 3.
1
0L
Cutaneous
“Deep”
“Deep”
Cutaneous
Hand
Other
FIG. 7. Distribution of cutaneous and deep receptive fields for SII recording sites after removal of hand representation in areas 3b and 1, both of which process predominantly cutaneous information. Within the SII hand representation, the normal ratio of cutaneous to deep receptive fields was reversed dramatically. In this case, the SII hand representation was sampled extensively relative to the representation of other body parts. All recording sites outside of the SII hand representation were responsive to somatic stimulation. Other conventions as in Fig. 3.
Removal of areas 3a, 3b, and 1, leaving intact only area 2, again allowed activation of a large number of SII recording sites with deep receptive fields on the hand, and in this case Again, recordings from body part representations outside none was responsive to cutaneous stimulation (Fig. 9). Par- that of the hand revealed the normal ratio of cutaneous to adoxically, limited sampling within area 2 of this hemideep receptive fields. sphere did reveal the presence of cutaneous receptive fields Having established that area 3a or 2 alone could relay on the hand. To determine whether we had simply missed deep receptor information to SII, we sought to determine small pockets of cutaneous inputs to SII in this case, we whether area 1 or 3b alone could relay cutaneous informarecorded from many additional SII sites but still found no tion. Figure 10 shows that ablation of areas 3a, 3b, and 2 cutaneous receptive fields on the hand. Thus area 2 alone, (and part of area 4)) leaving intact only area 1, did yield like area 3a alone, appears capable of relaying deep receptor recording sites in the Sll hand representation responsive to information and activating neurons across much of the SII cutaneous stimulation, although their number was sharply hand representation. Unlike area 3a, however, area 2 alone reduced compared with cutaneous receptive fields in body may be incapable of relaying cutaneous information. part representations other than of the hand or in the SII hand representation of normal animals. Virtually no SII recording sites were responsive exclusively to high-intensity stimulation of the hand. Indeed, much of SII in the ex-
FIG. 6. Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of hand representation in areas 3a and 3b. In this case, some of area 4 on the rostra1 bank of the central fissure was removed to gain access to areas 3a and 3b without damaging areas 1 or 2. Note the increase in deep relative to cutaneous receptive fields in the SII hand representation. The boundaries of the hand representation were difficult to define in this case, which resulted in relatively more recording sites in body part representations other than the hand. Note that the effects of the ablation were confined to the SII hand representation. Other conventions as in Fig. 3.
0
“Deep”
Cutaneous
Hand
Other
FIG. 8. Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of hand representation of areas 3b, 1, and 2, leaving intact only area 3a. Again note the dramatic shift in the number of cutaneous to deep receptive fields in the SII hand representation, which was sampled extensively relative to other body part representations. Other conventions as in Fig. 3.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
524
T. P. PONS,
P. E. GARRAGHTY,
AND
M.
MISHKIN
60
Cutaneous
“Deep”
Other FIG. 9. Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of hand representations in areas 3a, 3b, and 1, leaving intact only area 2. In this case, there was a complete lack of cutaneous activation of the hand representation in SII, though a number of deep receptive fields could be readily defined. Body part representations other than the hand were unaffected. Other conventions as in Fig. 3.
Distribution of cutaneous and deep receptive fields for SII FIG. 11. recording sites after ablation of the hand representation in areas 3a, 1, and 2, leaving intact only area 3b. Although recording sites with cutaneous receptive fields could be found easily in SII, no sites were found that responded exclusively to deep stimulation. Other conventions as in Fig. 3.
Ablation petted location of the hand representation in this case did not respond to any somatic stimulation. Thus area 1 alone appears capable of relaying relatively limited amounts of cutaneous input and little or no deep receptor information to SII. Figure 11 shows that ablation of areas 3a, 1, and 2 (and part of area 4), with only area 3b remaining intact, left a relatively greater number of cutaneous recording sites in the SII hand representation than in the cases in which areas 3a, 1 or 2 alone were spared, with many fewer nonresponsive sites than when just areas 1 or 3a remained intact. As in the case described above, no SII recording sites responded exclusively to high-intensity or deep stimulation. Recording sites in SII representations of body parts other than the hand showed a normal ratio of cutaneous to deep receptive fields. Thus area 3b alone appears to contribute relatively more cutaneous input to SII than any of the other postcentral cortical areas alone.
of’one area
Although our results suggested that either area 3a or 1 alone provides relatively little cutaneous input to SII, and area 2 provides little or none, we investigated whether the inputs to SII from all three areas might nevertheless combine to equal or even exceed those contributed by area 3b alone. In fact, ablation of the area 3b hand representation (and part of area 4) resulted in the elimination of the majority of cutaneous receptive fields in the SII hand representation (Fig. 12). Recordings in other parts of SII again indicated no effect of the ablation on representations of body parts other than the hand. Thus, although postcentral areas outside area 3b are capable of supplying limited cutaneous input to SII in macaques, the bulk of such input appears to be provided by area 3b, either directly or indirectly via areas 1 and 3a. 90
1
80 1 80 8 70 %
60
F 5 50 B aQ, 40 z 30 5 0 -i--
Cutaneous
“Deep”
Hand 0
Cutaneous
“Deep”
Hand
Cutaneous
“Deep”
Other
10. Hand representation in areas 3a, 3b, and 2 ablated. Distribution of cutaneous and deep receptive fields for SII recording sites after removal of areas 3a, 3b, and 2, leaving intact only area 1. Other conventions as in Fig. 3. FIG.
Other
FIG. 12.
Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of the hand representation in area 3b. Note the shift in the ratio of cutaneous to deep recording sites in the SII hand representation. The SII hand representation was not as extensively sampled as in most cases because of misplacement of the recording chamber in one hemisphere. Other conventions as in Fig. 3.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS
DISCUSSION
Effect of selective removals of postcentral cortical areas on modality properties of SII neurons Recording studies of the postcentral strip have demonstrated that 1) areas 3b and 1 process cutaneous inputs almost exclusively, 2) area 3a processes predominantly deep muscle-afferent information but also limited cutaneous inputs, and 3) area 2 processes predominantly joint and other deep inputs but, again, limited cutaneous input (Krishnamurti et al. 1976; Merzenich et al. 1978; Mountcastle and Powell 1959; Pons et al. 1985; Powell and Mountcastle 1959a,b). As already indicated, combined ablation of a body part representation in all four of these cortical areas eliminates the corresponding representation in SII. A central question raised by this discovery is: what is the contribution of each of the four postcentral somatosensory areas to the modality properties of SII neurons? Anatomic studies have established that each postcentral area independently projects to layers IV and lower III of SII (Friedman et al. 1986; Pons and Kaas 1986) indicating that information could be relayed along these direct, parallel, cortical routes. The present study has demonstrated that this is indeed the case. Each area alone is capable of supplying somatosensory information directly to SII, cutaneous information separately from areas 3b and 1, and deep receptor information separately from areas 3a and 2. But processing of somatic information also occurs sequentially among postcentral cortical areas, and some information is likely to be relayed to SII only after such processing. For example, some inputs reaching area 3b are then transmitted to area 1 (Pons and Kaas 1986)‘) and similarly, some inputs reaching area 3a are transmitted to area 2 and conversely. Thus somatic information reaching SII is relayed serially from VPL to each of the four areas in the postcentral strip, and is then transmitted to SII either directly or indirectly through another of those areas along parallel modality selective channels. Effect of the ablations on SII somatotopy We had previously demonstrated that after complete removal of a body part representation in the postcentral strip, i.e., after removal of that representation in all four areas, the corresponding representation in SII is deactivated (Pons et al. 1987a), i.e., is rendered “silent” or unresponsive to somatic stimulation of any body part. This was the result immediately after such ablations. Subsequently, however, a very different result was obtained. Six to eight weeks after ablation of the postcentral hand representations, the SII cortex normally receiving inputs from the hand was found to respond to cutaneous stimulation of the foot; that is, the SII foot representation had expanded into and almost fully occupied the region of SII that normally represents the hand (Pons et al. 1988b,c). No such evidence for an expanded representation of the foot or any other body part was obtained in the present study in which recordings were likewise made 6-8 wk after the ablations. This was so even when the hand representation was removed in three of the four areas of the postcen-
IN MACAQUES
525
tral strip, leaving only a portion of a single area intact. This result is in good accord with that of Burton et al. ( 1990) who reported almost total loss of cutaneous activation of neurons in the SII hand zone after incomplete ablations of the postcentral hand representations, yet no expansion into the SII hand zone of any other body part representation. These investigators also found, as we did, that after the incomplete postcentral lesions, in which part of area 3a was left intact, some recording sites in the SII hand representation continued to be activated by high-intensity or deep stimulation of the hand. Apparently, expansion of one body part representation in SII at the expense of another occurs only if the latter is totally vacated as a result of the complete ablation of that representation in the postcentral strip. l Whether there was an expansion of cutaneous inputs within the SII hand representation after cortical lesions of areas 3a and 2 or an expansion of deep inputs after lesions of 3b and 1 is not clear. Portions of the SII hand representation remained unresponsive to somatic stimulation after such removals, and intensive efforts were sometimes required to find responsive recording sites. Thus it did not appear that either cutaneous or deep inputs expanded across partially deafferented portions of SII, though this possibility cannot be ruled out. Anatomic studies have demonstrated that projections from individual cytoarchitectonic areas in postcentral cortex terminate in SII in “columns” or modules -0.5- 1 mm wide (Friedman et al. 1980; Pons and Kaas 1986). The interdigitation of responsive and unresponsive recording sites in SII after subtotal postcentral ablations is consistent with the anatomic finding of modular projections and suggests that different postcentral areas activate different modules in SII. Because connections between modules are common in primary visual and somatosensory cortex (Pons and Kaas 1986; Rockland and Lund 1982) such connections between modules may also exist in SII. If so, perhaps modules in SII that remain activated after partial postcentral ablations continue to transmit impulses, presumably inhibitory, to the deactivated modules. The maintenance of such interaction between modules may somehow prevent the takeover of the deactivated modules by other inputs. Whatever the mechanism, the presence of a projection from any one of the four areas in the postcentral strip to the SII hand representation appears to be sufficient to block the expansion of other body part representations as well as other somatic modalities. Alternative
sources for activation
of SII neurons
SII receives projections from a number of nonpostcentral sources, which could contribute to the somatosensory processing functions of SII. The ventroposterior inferior nucleus (VPI) appears to provide the major source of thalamic input to SII, with additional inputs arising in the ante’ Burton et al. ( 1990) failed to see an expansion of the foot representation in SII in an animal that did receive a complete ablation of the hand representation in the postcentral strip. This animal, however, received the lesion as an infant, suggesting that the rules for postinjury reorganization may change ontogenetically.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
526
T. P. PONS,
P. E. GARRAGHTY,
rior pulvinar and posterior nucleus (Burton et al. 1990; Friedman et al. 1986; Krubitzer and Kaas 1990; Manzoni et al. 1984). In addition, SII is reported to receive a sparse projection from the ventroposterior nucleus (VPL and VPM) (Burton 1984; Burton et al. 1990; Friedman and Murray 1986). But whatever the density of such inputs from the thalamus, or from any other nonpostcentral source, it is clear that, in the absence of the appropriate postcentral cortical area(s), thalamic inputs are incapable of mediating activation of SII neurons to either cutaneous or deep stimulation, at least under our experimental conditions. Although the postcentral strip is thus necessary for somatic activation of neurons in SII, it is conceivable that these postcentral inputs simply lower the threshold of SII neurons and that their actual activation is due to the thalamic inputs. Yet this interpretation of our findings seems highly improbable. First, VPI, which provides the major source of thalamic projections to SII, receives neither cutaneous nor deep somatosensory inputs, but instead has been reported to process vibrotactile stimuli from pacinian receptors (Dykes 198 1; Herron and Dykes 1986; cf. Garraghty et al. 1990). Second, neither the anterior pulvinar nor the posterior nucleus receives ascending cutaneous or deep receptor information (Apkarian and Hodge 1989; Berkely 1980), so these too are unlikely to be the source of somatic activation to neurons in SII. Finally, the sparse input from VPL is also an unlikely candidate because this projection too is incapable of activating SII neurons in the absence of postcentral cortex. To rule out these possibilities entirely, however, one would need to deactivate the candidate thalamic nuclei and thereby test whether postcentral cortex is not only necessary to activate SII neurons but also sufficient. Comparisons
with other species
In contrast to the present study on the macaque, an Old World monkey, a recent study on the marmoset, a New World monkey, demonstrated that combined removal of areas 3a and 3b was sufficient to abolish SII responsivity to both cutaneous and deep somatic stimulation, at least acutely (Garraghty et al. 1990). Comparing results in the two species is problematic, however, for little is known about the organization of parietal cortex caudal to area 3b in marmosets. A field immediately caudal to area 3b in marmosets is inconsistently responsive to somatic stimulation, and the somatotopic organization and receptive field characteristics of the responsive neurons differ from those in area 1 of macaques (Carlson et al. 1986; Krubitzer and Kaas 1990). These latter studies thus raise questions as to whether the field immediately caudal to area 3b in New World primates is homologous to area 1, area 2, both, or neither (Krubitzer and Kaas 1990). As a result, it is unclear what constitutes equivalent somatosensory cortical ablations in the two species. Nevertheless, in both marmosets and macaques, tactile information appears to be relayed serially from VP to both areas 3a and 3b, and from these two areas to SII, with additional cutaneous and deep information being relayed via areas 1 and 2 in macaques only.
AND
M.
MISHKIN
These results in macaques and marmosets contrast with earlier demonstrations that deactivation of SI in rabbits, cats, tree shrews, and galagos has little effect on the responsivity of SII neurons (Burton and Robinson 1987; Garraghty et al. 1992; Manzoni et al. 1979; Woolsey and Wang 1945 ) . Consistent with these contrasting effects of SI lesions across species, SII in cats, tree shrews, and galagos, unlike SII in higher primates, receives substantial projections from VP (Jones and Powell 1970; Kaas 1982; Weller et al. 1987). Thus both anatomic and electrophysiological findings are in accord that in these latter species, unlike in higher primates, SI and SII process somatosensory information in parallel. Major questions raised by this cross-species comparison is how the evolutionary shift from parallel to serial cortical processing occurred, and what, if any, are the functional advantages? We thank Michael Hall for help with the figures. Present address of P. E. Garraghty: Dept. of Psychology, sity, Bloomington, IN 47401. Address for repri nt requests: T. P. Pons. Received
2 1 October
199 1; accepted
in final
form
1 April
Indiana
Univer-
1992.
REFERENCES
ALLOWAY&. D., SINCLAIR, R.J., ANDBURTON, H.Responsesofneurons in somatosensory cortical area II of cats to high-frequency vibratory stimuli during iontophoresis of a GABA antagonist and glutamate. Somatosens. Mot. Rex 6: 109-140, 1988. APKARIAN, A. V. AND HODGE, C. J. Primate spinothalamic pathways. III. Thalamic terminations of the dorsolateral and ventral spinothalamic pathways. J. Comp. Neural. 288: 493-5 11, 1989. BERKELY, K. J. Spatial relationships between the terminations of somatic sensory and motor pathways in the rostra1 brainstem of cats and monkeys. I. Ascending somatic sensory inputs to lateral diencephalon. J. Corny. Neural. 193: 283-3 17, 1980. BURTON, H. Corticothalamic connections from the second somatosensory area and neighboring regions in the lateral sulcus of macaque monkeys. Brain Rex 309: 367-372, 1984. BURTON, H. AND ROBINSON, C. J. Responses in the first or second somatosensory cortical areas in cats during transient inactivation of the other ipsilateral area with lidocaine hydrochloride. Somatosens. Rex 4: 2 15236, 1987. BURTON, H., SATHIAN,K.,ANDDIAN-HUA, S.Alteredresponsestocutaneous stimuli in the second somatosensory cortex following lesions of the postcentral gyrus in infant and juvenile monkeys. J. Clomp. Neural. 29 1: 395-414. 1990. CARLSON, M., HUERTA, M. F., CUSICK, C. G., AND KAAS, J. H. Studies on the evolution of multiple somatosensory representations in primates: the organization of anterior parietal cortex in the New World callitrichid, Saguinis. J. Comp. Neural. 246: 409-426, 1986. DYKES, R. W., LANDRY, P., METHERATE, R., AND HICKS, T. P. Functional role of GABA in cat primary somatosensory cortex: shaping receptive fields of cortical neurons. J. Neurophysiol. 52: 1066- 1093, 1984. DYKES, R. W., SUR, M., MERZENICH, M. M., KAAS, J. H., AND NELSON, R. J. Regional segregation of neurons responding to quickly adapting, slowly adapting, deep and pacinian receptors within thalamic ventroposterior lateral and ventroposterior inferior nuclei in the squirrel monkey (Saimiri sciureus) . Neuroscience 6: 1687- 1692, 198 1. FRIEDMAN, D. P., JONES,E. G., AND BURTON, H. Representation pattern in the second somatosensory area of the monkey cerebral cortex. J. Comp. Neural. 192: 2 l-42, 1980. FRIEDMAN, D. P. AND MURRAY, E. A. Thalamic connectivity of the second somatosensory area and neighboring somatosensory fields of the lateral sulcus of the macaque. J. Comp. Neural. 252: 348-373, 1986. FRIEDMAN, D. P., MURRAY, E. A., O’NEILL, J. B., AND MISHKIN, M. Cortical connections of the somatosensory fields of the lateral sulcus of macaques: evidence for a corticolimbic pathway for touch. J. Camp. Neural. 252: 323-347, 1986. GARRAGHTY, P.E., FLORENCE,~. L., TENHULA, W.N., ANDKAAS, J.H.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW
OF
SENSORY
Parallel thalamic activation of the first and second somatosensory areas in prosimian primates and tree shrews. J. Camp. N~rol. In press. GARRAGHTY, P. E., PONS, T. P., AND KAAS, J. H. Ablations ofareas 3b (SI proper) and 3a of somatosensory cortex in marmosets deactivate the second and parietal ventral somatosensory areas. Somatosens. A4ot. Res. 7: 125-135,199O. HERRON, P. X. AND DYKES, R. W. The ventroposterior inferior nucleus in the thalamus of cats: a relay nucleus in the pacinian pathway to somatosensory cortex. ,I. Nczwphykol. 56: 1475- 1497, 1986. HICKS, T. P. AND DYKES, R. W. Receptive field size for certain neurons is determined by GABA-mediated intracortical inhibition. Brain Rcs. 274:160-164,1983. JONES, E. G. AND POWELL, T. P. S. Connexions of the somatic sensory cortex of the rhesus monkey. III. Thalamic connexions. Brain 93: 3756, 1970. KAAS, J. H. The somatosensory cortex and thalamus. In: Galagn the Lcssc~- Bnsh Buhy jGalago) as an Animal Model, selected topics edited by D. N. Haines. Boca Raton, FL: CRC, 1982, p. 169- 18 1. KAAS, J. H. What, if anything, is SI? Organization of the first somatosensory area of cortex. Physiol. Rev. 63: 206-23 1, 1983. KAAS, J. H., NELSON, R. J., DYKES, R. W., AND MERZENICH, M. M. The somatotopic organization of the ventroposterior thalamus of the squirrel monkey, Saimiri sc*iwcns. .I. Camp. Ncwol. 226: 1 1 I- 140, 1984. KAAS, J. H. AND PONS, T. P. The somatosensory system of primates. In: Comparutiw Primate Biology: NCZIro,sCit’nC~,s‘, edited by H. Steklis. New York: Liss, 1988, vol. 4, p. 42 l-468. KRISHNAMURTI, A., SANIDES, F., AND WELKER, W. I. Microelectrode mapping of modality-specific somatic sensory cerebral neocortex in slow loris. Bruin Bcjhav. Evol. 13: 267-283, 1976. KRUBITZER, L. A. AND KAAS, J. H. The organization and connections of the somatosensory cortex in marmosets. J. Ncznwcci. 3: 952-974, 1990. MANZONI, T. P., BARBARESI, R., AND CONTI, F. Callosal mechanism for the interhemispheric transfer of hand somatosensory information in the monkey. B&av. L?ruin Rcs. 1 1: 155-l 70, 1984. MERZENICH, M. M., KAAS, J. H., SUR, M., AND LIN, C. S. Double representation of the body surface within cytoarchitectonic areas 3b and 1 in “S-I” in the owl monkey (/lotzrs trivirgatns). J. Camp. Nwrol. 18 1: 47-74, 1978. MOUNTCASTLE, V. B. Central nervous mechanisms in mechanoreceptive sensibility. In: I-land~~owk c!f‘Physiology. The NUWUS System. Sensory Yrocwsw. Bethesda, MD: Am. Physiol. Sot., 1984, vol. III, p. 789-878. MOUNTCASTLE, V. B. AND HENNEMAN, E. The representation of tactile sensibility in the thalamus of the monkey. J. Corny. Ncwol. 97: 409440, 1952. MOUNTCASTLE, V. B. AND POWELL, T. P. S. Neural mechanisms subserving cutaneous sensibility, with special reference to the role of afferent inhibition in sensory perception and discrimination. Bnll. Johns Nopkins Ilosp. 105: 20 l-232, 1959. NELSON, R. J., SUR, M., FELLEMAN, D. J., AND KAAS, J. H. The representations of the body surface in postcentral somatosensory cortex of Macac>a jusc~icw1uri.s. J. Camp. Ncwroi. 192: 6 1 l-643, 1980.
INPUTS
IN
MACAQUES
527
T. P., GARRAGHTY, P. E., CUSICK, C. G., AND KAAS, J. H. The somatotopic organization of area 2 in macaque monkeys. J. C~~rnp. Neural. 24 1: 445-466, 1985. PONS, T. P., GARRAGHTY, P. E., FRIEDMAN, D. P., AND MISHKIN, M. Physiological evidence for serial processing in somatosensory cortex. Science Wash. DC 237: 417-420, 1987a. PONS, T. P., GARRAGHTY, P. E., AND MISHKIN, M. Elects of selective removals of the postcentral hand representations on the responsivity of neurons in the second somatosensory area (SII) of the rhesus monkey. SW. Nwws~i. Ahstr. 14: 759, 1988a. PONS, T. P., GARRAGHTY, P. E., AND MISHKIN, M. Lesion-induced plasticity in the second somatosensory cortex of adult macaques. Proc. Natl. Acad. Sci. USA 85: 5279-528 1, 1988b. PONS, T. P., GARRAGHTY, P. E., AND MISHKIN, M. Plasticity in nonprimar-y somatosensory cortex of adult monkeys. In: Post Lesion N~zwal Plasticity, edited by H. Flohr. Berlin: Springer-Verlag, 1988c, p. 5 1 l517. PONS, T. P. AND KAAS, J. H. Corticocortical connections of area 2 of somatosensory cortex in macaque monkeys: a correlative anatomical and electrophysiological study. J. C’omp. Ncwrol. 248: 3 13-335, 1986. PONS, T. P., WALL, J. H., GARRAGHTY, P. E., CUSICK, C. G., AND KAAS, J. H. Consistent features of the representation of the hand in area 3b of macaque monkeys. Somutoswzs. lk.s. 4: 309-33 1, 1987b. POWELL, T. P. S. AND MOUNT~ASTLE, V. B. The cytoarchitecture of the postcentral gyrus of the monkey ~W~cucu mnlutta. Bzrll. .Johns ilopkins Ilosp. 105: 108-131, 1959a. POWELL, T. P. S. AND MOUNTCASTLE, V. B. Some aspects of the functional organization of the cortex of the postcentral gyrus of the monkey: a correlation of findings obtained in a single unit analysis with cytoarchitecture. Bnll. Johns L1opkin.s llosp. 105: 133- 162, 1959b. ROBINSON, C. J. AND BURTON, H. Somatotopographic organization in the second somatosensory area of M. ,J1-rscic~zkwis. .J. Camp. Nwrol. 192: 43-67, 1980. ROCKLAND, K. S. AND LUND. J. S. Widespread periodic intrinsic connections in tree shrew visual cortex (Area 17) Sc%w Wush. DC’ 2 15: 1532-1534, 1982. SCHILLER, P. H. AND MALPELI, J. G. The eflect of striate cortex cooling on area 18 cells in the monkey. Brain Rcjs. 126: 366-369, 1977. TANJI, J. Activity of neurons in cortical area 3a during maintenance of steady state postures by] the monkey. Brain Rcjs. 88: 549-553, 1975. WELLER, R. E., SUR, M., ANL) KAAS, J. H. Callosal and ipsilateral cortical. connections of the body surface representations in SI and SII of tree shrews. Somatowns. RH. 5: 107- 133, 1987. WOOLSEY, C. N. “Second” somatic receiving areas in the cerebral cortex of cat, dog and monkey (Abstract). Fcdcwtion Proc’. 2: 55, 1943. WOOLSEY, C. N. AND FAIRMAN, D. Contralateral, ipsilateral and bilateral representation of cutaneous receptors in somatic areas I and II of the cerebral cortex of pig, sheep and other mammals. S’I~/;~I’/*~~ 19: 684-702, 1946. WOOLSEY, C. N. AND WANG, G. H. Somatic areas 1 and 2 of the cerebral cortex of the rabbit (Abstract). F&rution ProI’. 4: 79, 1945. PONS,
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
in U.S.il.
Serial and Parallel Processing of Tactual Information in Somatosensory Cortex of Rhesus Monkeys T. P. PONS, P. E. GARRAGHTY, AND M. MISHKIN Laboratory ofNeuropsychology, National Institute ofMental SUMMARY
AND
CONCLUSIONS
1. Selective ablations of the hand representations in postcentral cortical areas 3a, 3b, 1, and 2 were made in different combinations to determine each area’s contribution to the responsivity and modality properties of neurons in the hand representation in SII. 2. Ablations that left intact only the postcentral areas that process predominantly cutaneous inputs (i.e., areas 3b and 1) yielded SII recording sites responsive to cutaneous stimulation and none driven exclusively by high-intensity or “deep” stimulation. Conversely, ablations that left intact only the postcentral areas that process predominantly deep receptor inputs (i.e., areas 3a and 2) yielded mostly SII recording sites that responded exclusively to deep stimulation. 3. Ablations that left intact only area 3a or only area 2 yielded substantial and roughly equal reductions in the number of deep receptive fields in SII. By contrast, ablations that left intact only area 3b or only area 1 yielded unequal reductions in the number of cutaneous receptive fields in SII: a small reduction when area 3b alone was intact but a somewhat larger one when only area 1 was intact. 4. Finally, when the hand representation in area 3b was ablated, leaving areas 3a, 1, and 2 fully intact, there was again a substantial reduction in the encounter rate of cutaneous receptive fields. 5. The partial ablations often led to unresponsive sites in the SII hand representation. In SII representations other than of the hand no such unresponsive sites were found and there were no substantial changes in the ratio of cutaneous to deep receptive fields, indicating that the foregoing results were not due to longlasting postsurgical depression or effects of anesthesia. 6. The findings indicate that modality-specific information is relayed from postcentral cortical areas to SII along parallel channels, with cutaneous inputs transmitted via areas 3b and 1, and deep inputs via areas 3a and 2. Further, area 3b provides the major source of cutaneous input to SII, directly and perhaps also via area 1. 7. The results are in line with accumulating anatomic and electrophysiological evidence pointing to an evolutionary shift in the organization of the somatosensory system from the general mammalian plan, in which tactile information is processed in parallel in SI and SII, to a new organization in higher primates in which the processing of tactile information proceeds serially from SI to SII. The presumed functional advantages of this evolutionary shift are unknown.
Health, Bethesda, Maryland
20892
to collectively as the postcentral strip. Another region of cortex that processes somatic information is located in the depths of the lateral sulcus and is known as the second somatosensory area or SII (Friedman et al. 1986; Robinson and Burton 1980; Woolsey and Fairman 1946). Until recently, the postcentral somatosensory strip and SII were both thought to receive dense projections directly from the ventroposterior nucleus (VP) of the thalamus in all mammalian species (Burton 1984; Jones and Powell 1970; Mountcastle 1984). Because of this putative dual projection pattern, it had long been assumed that the processing of tactile information proceeded in parallel across these two cortical regions. That assumption was brought into question, however, by evidence demonstrating that VP in macaques provides only a sparse projection onto neurons in SII (Friedman and Murray 1986; Krubitzer and Kaas 1990; Manzoni et al. 1984). Direct evidence that the postcentral strip and SII of macaques do not process somatic inputs in parallel came from combined ablation and recording experiments, which demonstrated that elimination of all four hand representations in the postcentral strip (i.e., areas 3a, 3b, 1, and 2) eliminated the hand representation in SII, whereas removal of SII had no detectable effect on the activation of neurons in the postcentral strip (Pons et al. 1987a). The results of these experiments in monkeys thus provided compelling evidence that SII, whose middle layers receive a direct input from all four areas in the postcentral strip, depends on one or more of those inputs for its activation. Of the four cytoarchitectonic areas that comprise the postcentral strip, two, areas 3b and 1, process predominantly cutaneous inputs, whereas the other two process inputs from receptors that are located primarily in “deep” body tissues such as muscle afferents (area 3a) and joints (area 2) (Krishnamurti et al. 1976; Merzenich et al. 1978; Mountcastle and Powell 1959; Pons et al. 1985; Powell and Mountcastle 1959a,b). The purpose of the present study was to determine the differential contribution, if any, of the four areas comprising postcentral cortex to the responsivity and modality properties of neurons in SII of macaques. A preliminary report of the findings has been presented elsewhere (Pons et al. 1988a).
INTRODUCTION
Primary somatosensory cortex in macaques, located anteriorly in the parietal lobe, consists of four cytoarchitectonitally distinct divisions, areas 3a, 3b, 1, and 2. These areas, each of which processes different components of somatic sensation (see Kaas and Pons 1988 for review), are referred
METHODS
Subjects Ten adult female monkeys (Macaca kg were used in the study.
mdatta)
weighing
3.0-6.0
518
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS IN MACAQUES
Surgery Monkeys were anesthetized with an initial dose of ketamine hydrochloride ( 15 mg/ kg) followed by pentobarbital sodium, which was administered intravenously as required (20-30 mg/ kg). Throughout surgery, which was performed using aseptic precautions, the animals received an intravenous drip of a solution of 5% dextrose-0.45% sodium chloride, and their heart rate, respiratory rate, and temperature were monitored and maintained within normal limits. Ablations of cortical areas were made by aspiration. All ablations were limited to the portions of the cortical area(s) representing the hand as defined by previous studies (Nelson et al. 1980; Pons et al. 1985, 1987b), although to insure the complete removal of the hand zone in any given area the ablations were extended > 1 mm past that zone’s estimated mediolateral and rostrocaudal borders. There are 14 possible combinations of subtotal ablations of the four areas comprising the postcentral strip, of which we performed 8, as summarized in Table 1. The combinations not used, particularly lesions of single areas, were deemed to be unnecessary or relatively uninformative. When areas 3a or 3b, both of which are entirely buried in the posterior bank of the central sulcus, were to be ablated without involvement of other areas in the postcentral bank (i.e., areas 3b or 1 ), a part of area 4 in the anterior bank of the central sulcus was removed to gain access to those buried areas. As a control for this procedure, one hemisphere was prepared with a removal limited to area 4. One unoperated hemisphere served as an additional control. A recording chamber and head fixation device were attached to the skull in a separate procedure. The chamber was positioned to provide maximum access to both SII and the hand representations in the postcentral strip, and the bone under the chamber was removed to expose the dura. All animals received an antibiotic as a prophylactic measure after each surgical procedure.
Recording procedures Six to eight weeks after the cortical ablations, the animals were anesthetized with ketamine hydrochloride ( 15 mg/ kg), intubated with an endotracheal tube coated with lidocaine, and then anesthetized with 2.0% halothane in a mixture of NZ02 and O2 gas. After attachment of the animal’s head bolt to a stereotaxic frame, the top of the previously implanted recording chamber was removed and electrodes were inserted through the dura into the cortex. The mapping and recording procedures were similar to those used in previous studies (Pons et al. 1985, 1987a,b, 1988b,c). Briefly, electrode penetrations were placed at 0.5~mm intervals across the mediolateral extent of SII and at LO-mm intervals across its rostrocaudal extent. Typically, 40-60 electrode penetrations were made through SII in each hemisphere studied. Microelectrodes were hydraulically advanced through parietal cortex, white matter, and then SII or surrounding cortex until single- or
TABLE
1.
Areas ablated
Postcentral
Precentral
No. of Hemispheres
None None 3a, 3b, 1 3a, 3b, 2 3a, 1, 2 3b, 1, 2 3a, 3b 3b, 1 1, 2 3b
None 4 None 4 4 None 4 None None 4
1 1 1 1 2 2 2 2 2 2
519
multiple-unit responses to mechanical stimulation of the body were isolated. Small marker lesions ( 10 PA for 10 s) were placed in cortex to assist in locating the recordings sites postmortem. If hair movement was insufficient to elicit a neuronal response, or if the receptive field was on or beneath the glabrous skin, calibrated Von Frey hairs were used as stimuli to allow quantification of the force required to elicit a consistent neuronal response. If a stimulus force of ~0.5 g was sufficient to evoke a vigorous neuronal response, the receptive field at that site was defined as cutaneous. If a stimulus force of 22.0 g was required, or if a high-intensity stimulation produced by squeezing fingers, bending joints, or kneading muscles was necessary to activate neurons, the receptive field at that site was defined as deep. Although recording sites with deep receptive fields were, by definition, unresponsive to cutaneous stimulation, the converse was not true, i.e., recording sites with cutaneous receptive fields may also have had inputs from deep receptors, but this could not be ascertained using our methodology. If any unit in a multiunit response responded to a cutaneous stimulus, that site was defined as “cutaneous.” Only 4 out of nearly 3,000 recording sites in SII required a stimulus force of 0.5-2.0 g to elicit a response; these four sites were dropped from the data analysis.
Histological
analysis
On completion of the recording, animals were given a lethal dose of pentobarbital and then perfused intracardially with 0.9% saline followed by either 10% formalin or 2% paraformaldehyde and 0.2% glutaraldehyde. The latter perfusate was used in brains reacted for cytochrome oxidase. Brains were cut in the coronal or parasagittal plane on a freezing microtome, and every fourth section was mounted and stained for Nissl substance or cytochrome oxidase to assess the placement of recording tracks, marker lesions, and cortical ablations. Recording sites in SII were identified by reconstructing electrode tracks and marker lesions as detailed elsewhere (Pons et al. 1988b,c). Analysis of cortical sections confirmed that all ablations were as intended (Figs. 1 and 2). The pattern of thalamic degeneration caused by the cortical ablations was restricted in each case to the portion of ventroposterolateral nucleus (VPL) immediately adjacent to ventroposteromedial nucleus (VPM), the portion known to contain the representation of the hand (Dykes et al. 198 1; Kaas et al. 1984; Mountcastle and Henneman 1952). RESULTS
Recordings in SII were made at 2,834 sites in 16 hemispheres of the 10 monkeys (see Tables 1 and 2). Control hemispheres The distribution of SII recording sites responding to cutaneous versus deep stimulation in the normal control hemisphere is shown in Fig. 3. A special effort was made in this case to sample extensively and equally from the hand representation and the representation of all other body parts combined. Both subdivisions contained more than twice as many cutaneous as deep receptive fields (although, as indicated above, some cutaneous receptive fields may also have had deep inputs). These results are in accord with those from previous studies of SII in macaques (Friedman et al. 1986; Pons et al. 1987a; Robinson and Burton 1980). The findings from the case with the removal of a portion of the anterior bank of the central sulcus (caudal area 4) directly across from the postcentral hand representations
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
T. P. PONS, P. E. GARRAGHTY,
AND M. MISHKIN
FIG. 1. Top left: lateral brain view with cytoarchitectonic areas 1, 2, and 5 numbered. Diagonally oriented dashed lines: location of borders between areas. Horizontal dashed lines: approximate borders of the postcentral hand representation. The 3 solid vertical lines numbered 1, 2, and 3 represent the approximate anteroposterior location of the coronal sections shown in the corresponding numbered columns below. The 2 vertical arrows along the lateral sulcus indicate the location of the section shown at the top right. The coronal section illustrated in the top right indicates the location of SII buried in the lateral fissure. The 2 solid straight lines show the angle of electrode penetrations into SII. A: reconstructions of 3 sections showing the locations of areas 4,3a, 3b, 1, and 2 from an unoperated hemisphere. B: 3 coronal sections across the anteroposterior extent of a lesion of the hand representations in areas 3a, 3b, and 2, that included some of area 4 (see METHODS) but that spared the area 1 hand representation. C: 3 coronal sections across the anteroposterior extent of a lesion of the area 3a, 1, and 2 hand representations, which included additional removal of some of area 4 (see METHODS), but spared the area 3b hand representation. D: 3 sections across the anteroposterior extent of a lesion of the area 3b, 1, and 2 hand representations that spared area 3a. In B, C, and D some tissue from areas 3b, 1, and 2 is shown from the representations of other body parts such as the face or arm. ce, central sulcus, pc, postcentral sulcus, ip, intraparietal sulcus, la, lateral sulcus.
law
1
iP
la I
la
a
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS IN MACAQUES
521
focus all electrode penetrations within the SII hand representation, the zone of SII most likely to be affected by the ablation. Again, the results revealed over twice as many cutaneous as deep receptive fields in the SII hand representation. In all other respects as well, the results in this case were essentially the same as those in the unoperated hemisphere. Hemispheres
with selective removals
Selective ablations of the areas comprising the postcentral strip were of three types. In the first, we removed two areas, one processing cutaneous inputs and another processing deep inputs, i.e., areas 1 and 2 were removed and areas 3a and 3b were spared, or the reverse. Also, we removed both areas processing predominantly cutaneous inputs, areas 3b and 1, and left both areas 3a and 2 intact. In the second type of ablation, we removed three cortical areas, leaving a single area intact to process either cutaneous (3b or 1) or deep (3a or 2) stimulation. Finally, in the third type, we removed a single area (3b) to compare the effect of removing it alone with that of preserving it alone. Ablations of two areas
FIG. 2. A: photomicrograph of a coronal section taken through the postcentral hand representations in an unoperated animal, stained for cytochrome oxidase, and shows the location of areas 3a, 3b, 1, and 2. B: photomicrograph of a coronal section through the postcentral hand representations ofa hemisphere that received a lesion ofthe hand representation in areas 3b, 1, and 2. Note the sparing ofarea 3a at the fundus ofthe central sulcus as indicated by the arrow. C: photomicrograph of a section taken through the normal location of the postcentral hand representations in a hemisphere in which only the area 3b hand representation was spared and the area 3a, 1, and 2 hand representations were removed. Note that despite ablating the areas immediately surrounding area 3b, thus creating an island of area 3b, the tissue remained healthy and did not suffer from infarcts due to an altered blood supply. Scale = 0.75 mm. Other conventions as in Fig. 1.
are shown in Fig. 4. As indicated earlier, this hemisphere was a control for those in which this same region had to be removed to gain access to ablate areas 3a, 3b, or both, without damaging area 1. In this case, an effort was made to
Combined ablation of cortical areas 1 and 2, which left intact area 3b for processing cutaneous inputs and area 3a for deep inputs, altered the ratio of cutaneous to deep receptive fields recorded across the SII hand representation (Fig. 5, Table 2). As in the control hemispheres, recording sites responsive to cutaneous stimulation far outnumbered those responsive to deep stimulation. In contrast to the control hemispheres, however, the ratio of cutaneous to deep recording sites increased and a number of recording sites in the SII hand representation were unresponsive to somatic stimulation. The results indicate that, in the absence of areas 1 and 2, areas 3a and 3b are capable of relaying substantial cutaneous and deep receptor information directly to SII, though not the entire amount it normally receives. Qualitatively, also, the deep responses in the SII hand representation differed from normal in that most were of the slowly adapting type, with receptive fields located directly over muscle bellies rather than on hairy skin or on glabrous pads. This feature is suggestive of selective input from area 3a, though quantification of this observation was beyond the scope of the present study. Recordings in SII body part representations other than of the hand, like the recordings in the SII hand representation, revealed essentially the same ratio of cutaneous to deep receptive fields as in the control hemispheres. Ablation of areas 3a and 3b (and a portion of area 4), again leaving one postcentral cortical area intact for processing cutaneous inputs (area 1) and another for deep inputs (area 2), yielded a result different from the one just described (Fig. 6, compare with Fig. 5 ) . First, there were somewhat fewer recording sites in the expected location of the SII hand representation that were driven by any type of somatic stimulation delivered to the hand. And second, although both cutaneous and deep receptive fields were found, recording sites with deep receptor input actually
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
522
T. P. PONS,
TABLE
2.
P. E. GARRAGHTY,
AND
M.
MISHKIN
Frequency ofcutaneous and deep recording sites in SII Other
Hand Figure
Ablated
3 4 5 6 7 8 9 10 11 12
Spared
None 4 1, 2 3a, 3b 3b, 1 3b, 1, 2 3a, 3b, 1 3a, 3b, 2 3a, 1, 2 3b
3a, 3b, 1, 2 3a, 3b, 1, 2 3a, 3b 1, 2 3a, 2 3a 2 1 3b 3a, 1, 2
Cutaneous
Deep
C/D
125 71 63 20 25 25 0 15 61 19
46 33 12 37 164 210 30 2 0 87
2.72 2.15 5.25 0.54 0.15 0.12 n/a 7.5 n/a 0.22
Ablated, postcentral cytoarchitectonic areas removed; Spared, postcentral sites; Other, number of recording sites for body parts other than the hand.
outnumbered those with cutaneous input. Thus, whereas cortical areas 1 and 2 receive from VPL and relay to SII at least some cutaneous and deep receptor information, respectively, area 1 (in the absence of area 3b) is much less effective in providing SII with cutaneous information than is area 2 (in the absence of area 3a) in providing SII with deep receptor information (see Fig. 5). In SII representations of body parts other than the hand, the encounter rate of cutaneous and deep receptive fields was normal, and cutaneous receptive fields were again the rule. Ablation of areas 3b and 1, leaving intact the two areas processing predominantly deep inputs ( areas 3a and 2)) also led to a dramatic reversal in the ratio of cutaneous to deep receptive fields in the SII hand representation (Fig. 7 and Table 2). Indeed, very few recording sites in the SII hand representation were responsive to cutaneous stimuli in these cases. Recordings in SII outside of the hand representation, by contrast, revealed the normal ratio of cutaneous to deep receptive fields.
cytoarchitectonic
Unresponsive 0 0 42 41 71 42 35 76 47 51 areas left intact;
C/D,
Cutaneous
Deep
C/D
117 8 62 130 139 128 62 81 112 73
58
2.02 8 2 1.48 2.62 2.56 2.95 1.53 1.49 1.74
31 88 53 50 21 53 75 42
ratio of cutaneous
-
to deep recording
Ablations of’three areas Having established that either area 3a or 2 in conjunction with at least one other postcentral area was capable of activating a substantial number of SII recording sites with deep receptor input, we sought to determine whether either of these areas alone would be sufficient for such activation. Ablation of areas 3b, 1, and 2, leaving intact only area 3a, still allowed activation of a large number of recording sites in SII by high-intensity stimulation (Fig. 8). Although there were some SII recording sites with cutaneous receptive fields, these fields were generally located at or very near a joint and responded in a slowly adapting fashion to the flexing of a finger, much like cutaneous sites that activate area 3a of normal animals. Thus area 3a alone is capable of supplying substantial deep receptor input to SII, as well as a limited amount of slowly adapting cutaneous input. Recordings in SII representations of body parts other than the hand revealed the usual high ratio of cutaneous to deep receptive fields.
140
80
120 8 co
70
.= 100 F
e
80
5 2
60
z 5 #
40 20 0
Cutaneous
“Deep”
10
Hand 0
FIG. 3.
Distribution of cutaneous and deep receptive fields for SII recording sites in a normal hemisphere. All recording sites in SII were found to be responsive to somatic stimulation in this case. Hand, receptive fields located on the dorsal or glabrous surface of the hand; Other, receptive fields on all body parts other than the hand (face, arm, trunk, leg, foot, tail, etc.) For definition of cutaneous and deep receptive fields see METHODS. See Table 2 for number of unresponsive SII recording sites for each type of lesion and for the ratio of cutaneous to deep responses. Data in this case were collected in 2 recording sessions over a 4-day period.
“Deep”
Cutaneous
Other FIG. 4. Distribution of cutaneous and deep receptive fields for SII recording sites in a hemisphere in which a portion of the rostra1 bank of the central sulcus (caudal area 4) immediately across from the hand representations in the postcentral strip was removed. All SII recording sites were responsive to somatic stimulation of the body. No effort was made to record from body part representations outside of the hand in this case. Other conventions as in Fig. 3.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS IN MACAQUES
523
180160-
60
3 1405 120P ‘5 loo8 _
z s 20 # 10
g pe
80-
z 6
60
20
n Cutaneous
“Deep”
Hand
Cutaneous
“Deep”
Other
FIG. 5. Distribution of cutaneous and deep receptive fields for SII recording sites after removal of the hand representation in areas 1 and 2. Within the expected location of the SII hand representation, most, but not all, recording sites were responsive to somatic stimuli (cutaneous or deep). Other conventions as in Fig. 3.
1
0L
Cutaneous
“Deep”
“Deep”
Cutaneous
Hand
Other
FIG. 7. Distribution of cutaneous and deep receptive fields for SII recording sites after removal of hand representation in areas 3b and 1, both of which process predominantly cutaneous information. Within the SII hand representation, the normal ratio of cutaneous to deep receptive fields was reversed dramatically. In this case, the SII hand representation was sampled extensively relative to the representation of other body parts. All recording sites outside of the SII hand representation were responsive to somatic stimulation. Other conventions as in Fig. 3.
Removal of areas 3a, 3b, and 1, leaving intact only area 2, again allowed activation of a large number of SII recording sites with deep receptive fields on the hand, and in this case Again, recordings from body part representations outside none was responsive to cutaneous stimulation (Fig. 9). Par- that of the hand revealed the normal ratio of cutaneous to adoxically, limited sampling within area 2 of this hemideep receptive fields. sphere did reveal the presence of cutaneous receptive fields Having established that area 3a or 2 alone could relay on the hand. To determine whether we had simply missed deep receptor information to SII, we sought to determine small pockets of cutaneous inputs to SII in this case, we whether area 1 or 3b alone could relay cutaneous informarecorded from many additional SII sites but still found no tion. Figure 10 shows that ablation of areas 3a, 3b, and 2 cutaneous receptive fields on the hand. Thus area 2 alone, (and part of area 4)) leaving intact only area 1, did yield like area 3a alone, appears capable of relaying deep receptor recording sites in the Sll hand representation responsive to information and activating neurons across much of the SII cutaneous stimulation, although their number was sharply hand representation. Unlike area 3a, however, area 2 alone reduced compared with cutaneous receptive fields in body may be incapable of relaying cutaneous information. part representations other than of the hand or in the SII hand representation of normal animals. Virtually no SII recording sites were responsive exclusively to high-intensity stimulation of the hand. Indeed, much of SII in the ex-
FIG. 6. Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of hand representation in areas 3a and 3b. In this case, some of area 4 on the rostra1 bank of the central fissure was removed to gain access to areas 3a and 3b without damaging areas 1 or 2. Note the increase in deep relative to cutaneous receptive fields in the SII hand representation. The boundaries of the hand representation were difficult to define in this case, which resulted in relatively more recording sites in body part representations other than the hand. Note that the effects of the ablation were confined to the SII hand representation. Other conventions as in Fig. 3.
0
“Deep”
Cutaneous
Hand
Other
FIG. 8. Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of hand representation of areas 3b, 1, and 2, leaving intact only area 3a. Again note the dramatic shift in the number of cutaneous to deep receptive fields in the SII hand representation, which was sampled extensively relative to other body part representations. Other conventions as in Fig. 3.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
524
T. P. PONS,
P. E. GARRAGHTY,
AND
M.
MISHKIN
60
Cutaneous
“Deep”
Other FIG. 9. Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of hand representations in areas 3a, 3b, and 1, leaving intact only area 2. In this case, there was a complete lack of cutaneous activation of the hand representation in SII, though a number of deep receptive fields could be readily defined. Body part representations other than the hand were unaffected. Other conventions as in Fig. 3.
Distribution of cutaneous and deep receptive fields for SII FIG. 11. recording sites after ablation of the hand representation in areas 3a, 1, and 2, leaving intact only area 3b. Although recording sites with cutaneous receptive fields could be found easily in SII, no sites were found that responded exclusively to deep stimulation. Other conventions as in Fig. 3.
Ablation petted location of the hand representation in this case did not respond to any somatic stimulation. Thus area 1 alone appears capable of relaying relatively limited amounts of cutaneous input and little or no deep receptor information to SII. Figure 11 shows that ablation of areas 3a, 1, and 2 (and part of area 4), with only area 3b remaining intact, left a relatively greater number of cutaneous recording sites in the SII hand representation than in the cases in which areas 3a, 1 or 2 alone were spared, with many fewer nonresponsive sites than when just areas 1 or 3a remained intact. As in the case described above, no SII recording sites responded exclusively to high-intensity or deep stimulation. Recording sites in SII representations of body parts other than the hand showed a normal ratio of cutaneous to deep receptive fields. Thus area 3b alone appears to contribute relatively more cutaneous input to SII than any of the other postcentral cortical areas alone.
of’one area
Although our results suggested that either area 3a or 1 alone provides relatively little cutaneous input to SII, and area 2 provides little or none, we investigated whether the inputs to SII from all three areas might nevertheless combine to equal or even exceed those contributed by area 3b alone. In fact, ablation of the area 3b hand representation (and part of area 4) resulted in the elimination of the majority of cutaneous receptive fields in the SII hand representation (Fig. 12). Recordings in other parts of SII again indicated no effect of the ablation on representations of body parts other than the hand. Thus, although postcentral areas outside area 3b are capable of supplying limited cutaneous input to SII in macaques, the bulk of such input appears to be provided by area 3b, either directly or indirectly via areas 1 and 3a. 90
1
80 1 80 8 70 %
60
F 5 50 B aQ, 40 z 30 5 0 -i--
Cutaneous
“Deep”
Hand 0
Cutaneous
“Deep”
Hand
Cutaneous
“Deep”
Other
10. Hand representation in areas 3a, 3b, and 2 ablated. Distribution of cutaneous and deep receptive fields for SII recording sites after removal of areas 3a, 3b, and 2, leaving intact only area 1. Other conventions as in Fig. 3. FIG.
Other
FIG. 12.
Distribution of cutaneous and deep receptive fields for SII recording sites after ablation of the hand representation in area 3b. Note the shift in the ratio of cutaneous to deep recording sites in the SII hand representation. The SII hand representation was not as extensively sampled as in most cases because of misplacement of the recording chamber in one hemisphere. Other conventions as in Fig. 3.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW OF SENSORY INPUTS
DISCUSSION
Effect of selective removals of postcentral cortical areas on modality properties of SII neurons Recording studies of the postcentral strip have demonstrated that 1) areas 3b and 1 process cutaneous inputs almost exclusively, 2) area 3a processes predominantly deep muscle-afferent information but also limited cutaneous inputs, and 3) area 2 processes predominantly joint and other deep inputs but, again, limited cutaneous input (Krishnamurti et al. 1976; Merzenich et al. 1978; Mountcastle and Powell 1959; Pons et al. 1985; Powell and Mountcastle 1959a,b). As already indicated, combined ablation of a body part representation in all four of these cortical areas eliminates the corresponding representation in SII. A central question raised by this discovery is: what is the contribution of each of the four postcentral somatosensory areas to the modality properties of SII neurons? Anatomic studies have established that each postcentral area independently projects to layers IV and lower III of SII (Friedman et al. 1986; Pons and Kaas 1986) indicating that information could be relayed along these direct, parallel, cortical routes. The present study has demonstrated that this is indeed the case. Each area alone is capable of supplying somatosensory information directly to SII, cutaneous information separately from areas 3b and 1, and deep receptor information separately from areas 3a and 2. But processing of somatic information also occurs sequentially among postcentral cortical areas, and some information is likely to be relayed to SII only after such processing. For example, some inputs reaching area 3b are then transmitted to area 1 (Pons and Kaas 1986)‘) and similarly, some inputs reaching area 3a are transmitted to area 2 and conversely. Thus somatic information reaching SII is relayed serially from VPL to each of the four areas in the postcentral strip, and is then transmitted to SII either directly or indirectly through another of those areas along parallel modality selective channels. Effect of the ablations on SII somatotopy We had previously demonstrated that after complete removal of a body part representation in the postcentral strip, i.e., after removal of that representation in all four areas, the corresponding representation in SII is deactivated (Pons et al. 1987a), i.e., is rendered “silent” or unresponsive to somatic stimulation of any body part. This was the result immediately after such ablations. Subsequently, however, a very different result was obtained. Six to eight weeks after ablation of the postcentral hand representations, the SII cortex normally receiving inputs from the hand was found to respond to cutaneous stimulation of the foot; that is, the SII foot representation had expanded into and almost fully occupied the region of SII that normally represents the hand (Pons et al. 1988b,c). No such evidence for an expanded representation of the foot or any other body part was obtained in the present study in which recordings were likewise made 6-8 wk after the ablations. This was so even when the hand representation was removed in three of the four areas of the postcen-
IN MACAQUES
525
tral strip, leaving only a portion of a single area intact. This result is in good accord with that of Burton et al. ( 1990) who reported almost total loss of cutaneous activation of neurons in the SII hand zone after incomplete ablations of the postcentral hand representations, yet no expansion into the SII hand zone of any other body part representation. These investigators also found, as we did, that after the incomplete postcentral lesions, in which part of area 3a was left intact, some recording sites in the SII hand representation continued to be activated by high-intensity or deep stimulation of the hand. Apparently, expansion of one body part representation in SII at the expense of another occurs only if the latter is totally vacated as a result of the complete ablation of that representation in the postcentral strip. l Whether there was an expansion of cutaneous inputs within the SII hand representation after cortical lesions of areas 3a and 2 or an expansion of deep inputs after lesions of 3b and 1 is not clear. Portions of the SII hand representation remained unresponsive to somatic stimulation after such removals, and intensive efforts were sometimes required to find responsive recording sites. Thus it did not appear that either cutaneous or deep inputs expanded across partially deafferented portions of SII, though this possibility cannot be ruled out. Anatomic studies have demonstrated that projections from individual cytoarchitectonic areas in postcentral cortex terminate in SII in “columns” or modules -0.5- 1 mm wide (Friedman et al. 1980; Pons and Kaas 1986). The interdigitation of responsive and unresponsive recording sites in SII after subtotal postcentral ablations is consistent with the anatomic finding of modular projections and suggests that different postcentral areas activate different modules in SII. Because connections between modules are common in primary visual and somatosensory cortex (Pons and Kaas 1986; Rockland and Lund 1982) such connections between modules may also exist in SII. If so, perhaps modules in SII that remain activated after partial postcentral ablations continue to transmit impulses, presumably inhibitory, to the deactivated modules. The maintenance of such interaction between modules may somehow prevent the takeover of the deactivated modules by other inputs. Whatever the mechanism, the presence of a projection from any one of the four areas in the postcentral strip to the SII hand representation appears to be sufficient to block the expansion of other body part representations as well as other somatic modalities. Alternative
sources for activation
of SII neurons
SII receives projections from a number of nonpostcentral sources, which could contribute to the somatosensory processing functions of SII. The ventroposterior inferior nucleus (VPI) appears to provide the major source of thalamic input to SII, with additional inputs arising in the ante’ Burton et al. ( 1990) failed to see an expansion of the foot representation in SII in an animal that did receive a complete ablation of the hand representation in the postcentral strip. This animal, however, received the lesion as an infant, suggesting that the rules for postinjury reorganization may change ontogenetically.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
526
T. P. PONS,
P. E. GARRAGHTY,
rior pulvinar and posterior nucleus (Burton et al. 1990; Friedman et al. 1986; Krubitzer and Kaas 1990; Manzoni et al. 1984). In addition, SII is reported to receive a sparse projection from the ventroposterior nucleus (VPL and VPM) (Burton 1984; Burton et al. 1990; Friedman and Murray 1986). But whatever the density of such inputs from the thalamus, or from any other nonpostcentral source, it is clear that, in the absence of the appropriate postcentral cortical area(s), thalamic inputs are incapable of mediating activation of SII neurons to either cutaneous or deep stimulation, at least under our experimental conditions. Although the postcentral strip is thus necessary for somatic activation of neurons in SII, it is conceivable that these postcentral inputs simply lower the threshold of SII neurons and that their actual activation is due to the thalamic inputs. Yet this interpretation of our findings seems highly improbable. First, VPI, which provides the major source of thalamic projections to SII, receives neither cutaneous nor deep somatosensory inputs, but instead has been reported to process vibrotactile stimuli from pacinian receptors (Dykes 198 1; Herron and Dykes 1986; cf. Garraghty et al. 1990). Second, neither the anterior pulvinar nor the posterior nucleus receives ascending cutaneous or deep receptor information (Apkarian and Hodge 1989; Berkely 1980), so these too are unlikely to be the source of somatic activation to neurons in SII. Finally, the sparse input from VPL is also an unlikely candidate because this projection too is incapable of activating SII neurons in the absence of postcentral cortex. To rule out these possibilities entirely, however, one would need to deactivate the candidate thalamic nuclei and thereby test whether postcentral cortex is not only necessary to activate SII neurons but also sufficient. Comparisons
with other species
In contrast to the present study on the macaque, an Old World monkey, a recent study on the marmoset, a New World monkey, demonstrated that combined removal of areas 3a and 3b was sufficient to abolish SII responsivity to both cutaneous and deep somatic stimulation, at least acutely (Garraghty et al. 1990). Comparing results in the two species is problematic, however, for little is known about the organization of parietal cortex caudal to area 3b in marmosets. A field immediately caudal to area 3b in marmosets is inconsistently responsive to somatic stimulation, and the somatotopic organization and receptive field characteristics of the responsive neurons differ from those in area 1 of macaques (Carlson et al. 1986; Krubitzer and Kaas 1990). These latter studies thus raise questions as to whether the field immediately caudal to area 3b in New World primates is homologous to area 1, area 2, both, or neither (Krubitzer and Kaas 1990). As a result, it is unclear what constitutes equivalent somatosensory cortical ablations in the two species. Nevertheless, in both marmosets and macaques, tactile information appears to be relayed serially from VP to both areas 3a and 3b, and from these two areas to SII, with additional cutaneous and deep information being relayed via areas 1 and 2 in macaques only.
AND
M.
MISHKIN
These results in macaques and marmosets contrast with earlier demonstrations that deactivation of SI in rabbits, cats, tree shrews, and galagos has little effect on the responsivity of SII neurons (Burton and Robinson 1987; Garraghty et al. 1992; Manzoni et al. 1979; Woolsey and Wang 1945 ) . Consistent with these contrasting effects of SI lesions across species, SII in cats, tree shrews, and galagos, unlike SII in higher primates, receives substantial projections from VP (Jones and Powell 1970; Kaas 1982; Weller et al. 1987). Thus both anatomic and electrophysiological findings are in accord that in these latter species, unlike in higher primates, SI and SII process somatosensory information in parallel. Major questions raised by this cross-species comparison is how the evolutionary shift from parallel to serial cortical processing occurred, and what, if any, are the functional advantages? We thank Michael Hall for help with the figures. Present address of P. E. Garraghty: Dept. of Psychology, sity, Bloomington, IN 47401. Address for repri nt requests: T. P. Pons. Received
2 1 October
199 1; accepted
in final
form
1 April
Indiana
Univer-
1992.
REFERENCES
ALLOWAY&. D., SINCLAIR, R.J., ANDBURTON, H.Responsesofneurons in somatosensory cortical area II of cats to high-frequency vibratory stimuli during iontophoresis of a GABA antagonist and glutamate. Somatosens. Mot. Rex 6: 109-140, 1988. APKARIAN, A. V. AND HODGE, C. J. Primate spinothalamic pathways. III. Thalamic terminations of the dorsolateral and ventral spinothalamic pathways. J. Comp. Neural. 288: 493-5 11, 1989. BERKELY, K. J. Spatial relationships between the terminations of somatic sensory and motor pathways in the rostra1 brainstem of cats and monkeys. I. Ascending somatic sensory inputs to lateral diencephalon. J. Corny. Neural. 193: 283-3 17, 1980. BURTON, H. Corticothalamic connections from the second somatosensory area and neighboring regions in the lateral sulcus of macaque monkeys. Brain Rex 309: 367-372, 1984. BURTON, H. AND ROBINSON, C. J. Responses in the first or second somatosensory cortical areas in cats during transient inactivation of the other ipsilateral area with lidocaine hydrochloride. Somatosens. Rex 4: 2 15236, 1987. BURTON, H., SATHIAN,K.,ANDDIAN-HUA, S.Alteredresponsestocutaneous stimuli in the second somatosensory cortex following lesions of the postcentral gyrus in infant and juvenile monkeys. J. Clomp. Neural. 29 1: 395-414. 1990. CARLSON, M., HUERTA, M. F., CUSICK, C. G., AND KAAS, J. H. Studies on the evolution of multiple somatosensory representations in primates: the organization of anterior parietal cortex in the New World callitrichid, Saguinis. J. Comp. Neural. 246: 409-426, 1986. DYKES, R. W., LANDRY, P., METHERATE, R., AND HICKS, T. P. Functional role of GABA in cat primary somatosensory cortex: shaping receptive fields of cortical neurons. J. Neurophysiol. 52: 1066- 1093, 1984. DYKES, R. W., SUR, M., MERZENICH, M. M., KAAS, J. H., AND NELSON, R. J. Regional segregation of neurons responding to quickly adapting, slowly adapting, deep and pacinian receptors within thalamic ventroposterior lateral and ventroposterior inferior nuclei in the squirrel monkey (Saimiri sciureus) . Neuroscience 6: 1687- 1692, 198 1. FRIEDMAN, D. P., JONES,E. G., AND BURTON, H. Representation pattern in the second somatosensory area of the monkey cerebral cortex. J. Comp. Neural. 192: 2 l-42, 1980. FRIEDMAN, D. P. AND MURRAY, E. A. Thalamic connectivity of the second somatosensory area and neighboring somatosensory fields of the lateral sulcus of the macaque. J. Comp. Neural. 252: 348-373, 1986. FRIEDMAN, D. P., MURRAY, E. A., O’NEILL, J. B., AND MISHKIN, M. Cortical connections of the somatosensory fields of the lateral sulcus of macaques: evidence for a corticolimbic pathway for touch. J. Camp. Neural. 252: 323-347, 1986. GARRAGHTY, P.E., FLORENCE,~. L., TENHULA, W.N., ANDKAAS, J.H.
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
CORTICAL
FLOW
OF
SENSORY
Parallel thalamic activation of the first and second somatosensory areas in prosimian primates and tree shrews. J. Camp. N~rol. In press. GARRAGHTY, P. E., PONS, T. P., AND KAAS, J. H. Ablations ofareas 3b (SI proper) and 3a of somatosensory cortex in marmosets deactivate the second and parietal ventral somatosensory areas. Somatosens. A4ot. Res. 7: 125-135,199O. HERRON, P. X. AND DYKES, R. W. The ventroposterior inferior nucleus in the thalamus of cats: a relay nucleus in the pacinian pathway to somatosensory cortex. ,I. Nczwphykol. 56: 1475- 1497, 1986. HICKS, T. P. AND DYKES, R. W. Receptive field size for certain neurons is determined by GABA-mediated intracortical inhibition. Brain Rcs. 274:160-164,1983. JONES, E. G. AND POWELL, T. P. S. Connexions of the somatic sensory cortex of the rhesus monkey. III. Thalamic connexions. Brain 93: 3756, 1970. KAAS, J. H. The somatosensory cortex and thalamus. In: Galagn the Lcssc~- Bnsh Buhy jGalago) as an Animal Model, selected topics edited by D. N. Haines. Boca Raton, FL: CRC, 1982, p. 169- 18 1. KAAS, J. H. What, if anything, is SI? Organization of the first somatosensory area of cortex. Physiol. Rev. 63: 206-23 1, 1983. KAAS, J. H., NELSON, R. J., DYKES, R. W., AND MERZENICH, M. M. The somatotopic organization of the ventroposterior thalamus of the squirrel monkey, Saimiri sc*iwcns. .I. Camp. Ncwol. 226: 1 1 I- 140, 1984. KAAS, J. H. AND PONS, T. P. The somatosensory system of primates. In: Comparutiw Primate Biology: NCZIro,sCit’nC~,s‘, edited by H. Steklis. New York: Liss, 1988, vol. 4, p. 42 l-468. KRISHNAMURTI, A., SANIDES, F., AND WELKER, W. I. Microelectrode mapping of modality-specific somatic sensory cerebral neocortex in slow loris. Bruin Bcjhav. Evol. 13: 267-283, 1976. KRUBITZER, L. A. AND KAAS, J. H. The organization and connections of the somatosensory cortex in marmosets. J. Ncznwcci. 3: 952-974, 1990. MANZONI, T. P., BARBARESI, R., AND CONTI, F. Callosal mechanism for the interhemispheric transfer of hand somatosensory information in the monkey. B&av. L?ruin Rcs. 1 1: 155-l 70, 1984. MERZENICH, M. M., KAAS, J. H., SUR, M., AND LIN, C. S. Double representation of the body surface within cytoarchitectonic areas 3b and 1 in “S-I” in the owl monkey (/lotzrs trivirgatns). J. Camp. Nwrol. 18 1: 47-74, 1978. MOUNTCASTLE, V. B. Central nervous mechanisms in mechanoreceptive sensibility. In: I-land~~owk c!f‘Physiology. The NUWUS System. Sensory Yrocwsw. Bethesda, MD: Am. Physiol. Sot., 1984, vol. III, p. 789-878. MOUNTCASTLE, V. B. AND HENNEMAN, E. The representation of tactile sensibility in the thalamus of the monkey. J. Corny. Ncwol. 97: 409440, 1952. MOUNTCASTLE, V. B. AND POWELL, T. P. S. Neural mechanisms subserving cutaneous sensibility, with special reference to the role of afferent inhibition in sensory perception and discrimination. Bnll. Johns Nopkins Ilosp. 105: 20 l-232, 1959. NELSON, R. J., SUR, M., FELLEMAN, D. J., AND KAAS, J. H. The representations of the body surface in postcentral somatosensory cortex of Macac>a jusc~icw1uri.s. J. Camp. Ncwroi. 192: 6 1 l-643, 1980.
INPUTS
IN
MACAQUES
527
T. P., GARRAGHTY, P. E., CUSICK, C. G., AND KAAS, J. H. The somatotopic organization of area 2 in macaque monkeys. J. C~~rnp. Neural. 24 1: 445-466, 1985. PONS, T. P., GARRAGHTY, P. E., FRIEDMAN, D. P., AND MISHKIN, M. Physiological evidence for serial processing in somatosensory cortex. Science Wash. DC 237: 417-420, 1987a. PONS, T. P., GARRAGHTY, P. E., AND MISHKIN, M. Elects of selective removals of the postcentral hand representations on the responsivity of neurons in the second somatosensory area (SII) of the rhesus monkey. SW. Nwws~i. Ahstr. 14: 759, 1988a. PONS, T. P., GARRAGHTY, P. E., AND MISHKIN, M. Lesion-induced plasticity in the second somatosensory cortex of adult macaques. Proc. Natl. Acad. Sci. USA 85: 5279-528 1, 1988b. PONS, T. P., GARRAGHTY, P. E., AND MISHKIN, M. Plasticity in nonprimar-y somatosensory cortex of adult monkeys. In: Post Lesion N~zwal Plasticity, edited by H. Flohr. Berlin: Springer-Verlag, 1988c, p. 5 1 l517. PONS, T. P. AND KAAS, J. H. Corticocortical connections of area 2 of somatosensory cortex in macaque monkeys: a correlative anatomical and electrophysiological study. J. C’omp. Ncwrol. 248: 3 13-335, 1986. PONS, T. P., WALL, J. H., GARRAGHTY, P. E., CUSICK, C. G., AND KAAS, J. H. Consistent features of the representation of the hand in area 3b of macaque monkeys. Somutoswzs. lk.s. 4: 309-33 1, 1987b. POWELL, T. P. S. AND MOUNT~ASTLE, V. B. The cytoarchitecture of the postcentral gyrus of the monkey ~W~cucu mnlutta. Bzrll. .Johns ilopkins Ilosp. 105: 108-131, 1959a. POWELL, T. P. S. AND MOUNTCASTLE, V. B. Some aspects of the functional organization of the cortex of the postcentral gyrus of the monkey: a correlation of findings obtained in a single unit analysis with cytoarchitecture. Bnll. Johns L1opkin.s llosp. 105: 133- 162, 1959b. ROBINSON, C. J. AND BURTON, H. Somatotopographic organization in the second somatosensory area of M. ,J1-rscic~zkwis. .J. Camp. Nwrol. 192: 43-67, 1980. ROCKLAND, K. S. AND LUND. J. S. Widespread periodic intrinsic connections in tree shrew visual cortex (Area 17) Sc%w Wush. DC’ 2 15: 1532-1534, 1982. SCHILLER, P. H. AND MALPELI, J. G. The eflect of striate cortex cooling on area 18 cells in the monkey. Brain Rcjs. 126: 366-369, 1977. TANJI, J. Activity of neurons in cortical area 3a during maintenance of steady state postures by] the monkey. Brain Rcjs. 88: 549-553, 1975. WELLER, R. E., SUR, M., ANL) KAAS, J. H. Callosal and ipsilateral cortical. connections of the body surface representations in SI and SII of tree shrews. Somatowns. RH. 5: 107- 133, 1987. WOOLSEY, C. N. “Second” somatic receiving areas in the cerebral cortex of cat, dog and monkey (Abstract). Fcdcwtion Proc’. 2: 55, 1943. WOOLSEY, C. N. AND FAIRMAN, D. Contralateral, ipsilateral and bilateral representation of cutaneous receptors in somatic areas I and II of the cerebral cortex of pig, sheep and other mammals. S’I~/;~I’/*~~ 19: 684-702, 1946. WOOLSEY, C. N. AND WANG, G. H. Somatic areas 1 and 2 of the cerebral cortex of the rabbit (Abstract). F&rution ProI’. 4: 79, 1945. PONS,
Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on August 8, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
