- Email: [email protected]
Unilateral periarcuate and posterior parietal lesions impair conditional position discrimination learning in the monkey
,~rurop.,?choluym. Vol 27. No. 9. pp. I1 19. 1127. 1989. Prmted m Great Britam.
t
co28&3932/89 $3.00+0.00 1989 Pergamon Press plc
UNILATERAL PERIARCUATE AND POSTERIOR PARIETAL LESIONS IMPAIR CONDITIONAL POSITION DISCRIMINATION LEARNING IN THE MONKEY DOUGLAS P. CROWNE, KIM A. DAWSON and CLAUDETTE M. RICHARDSON Departmentof Psychology, Universityof Waterloo, Waterloo, Ontario, Canada, N2L 3Gl (Receired
26 July
1988:
accepted
7 April
1989)
Abstract-Monkeys were trained on a conditional position discrimination in which the conditional cue was a light blinking at two distinct rates and the discriminanda (illuminated buttons) appeared in varying symmetrical positions of eccentricity. Unilateral arcuate, posterior parietal, or principal sulcus lesions were performed at criterion. The monkeys were tested to recovery when a homologous lesion was made contralaterally. The first of two analyses examined a period of4 weeks following each lesion; the unit of analysis was lesions. The arcuate and parietal lesions produced impairments on both widely eccentric and central discriminanda locations; initially. virtually all responses were deflected to the ipsilateral side. There was significant improvement after the arcuate and parietal lesions from weeks 1 to 4. An analysis of total trials to criterion showed major deficits from the second arcuate and parietal lesions, with the arcuate lesion impairment being particularly severe. These results establish that a spatial concept of left vs right is seriously deranged by unilateral lesions of cortical association areas involved in spatial orientation and discrimination.
INTRODUCTION DISORDERS of spatial
localization, orientation, and perception have been recognized in humans and in monkeys for a century [l, 93 and have been associated with parietal lobe damage since the studies of Holmes in 1918-1919 [ 11, 121. The involvement of the frontal lobes in spatial behaviour was not appreciated for another 40 yr. Semmes et al. [23] reported that frontal lesions produce a spatial defect that is clearly dissociated from the spatial semeiology of parietal injury. Their distinction between impairments of personal spatial (prefrontal cortex) and extrapersonal (parietal cortex) orientation was confirmed in studies of humans [3,4,23, 261 and monkeys [ZO] with frontal and parietal damage. BRODY and PRIBRAM [2] then showed that while posterior parietal lesions in the monkey impair performance only on tasks imposing extrapersonal orientation demands, frontal lesions degrade the ability to perform both personal and extrapersonal tasks when the stimulus context varies. Despite this important evidence distinguishing the spatial cognitive processes of frontal and parietal cortex, there are functional aspects of these regions whose implications are yet to be fully explored. It is now firmly established, for example, that there are several highly important behavioral specializations within prefrontal cortex of the monkey [22]. A spatial problem of particular interest, the conditional position discrimination, is impaired by arcuate cortex (frontal eye field) lesions. GOLDMAN and ROSVOLD [lo] presented monkeys with an auditory cue through speakers above or below the test chamber. The upper cue signified choice of the left foodwell, the lower cue the right. Periarcuate lesions severely degraded performance on this discrimination, while principal sulcus damage did not. 1119
MILNERct d. [ 171 showed that the deficit is not specific to auditory cues and is clearly thus a spatial impairment. Conditional position discrimination certainly imposes egocentric (“frontal”) demands on left--right orientation; would those features, appearing in a difficult problem, also make it vulnerable to parietal damage? In this experiment we studied the performance of monkeys with serial periarcuate. posterior parietal or principal sulcus resections on a particularly complex version of the conditional position discrimination problem. Characteristically, studies of spatial learning and orientation have placed the discriminanda in two specific locations to the left and right. Animals may thus respond to a particular locus and not on the basis of a spatial concept“to the left” or “to the right”. We sought to teach monkeys the spatial concept of left vs right, quite independent of specific place. Accordingly, the positions of the spatial stimuli varied randomly in their eccentricity from trial to trial. Our conditional position discrimination could not be solved as a place-learning problem. The evidence of spatial impairments in animals is based on bilateral lesions. To reveal more clearly the processes involved in spatial learning and performance and to enable comparisons with humans, in whom unilateral lesions produce severe and often unremitting deficits, we studied spatial discrimination after unilateral lesions in monkeys.
Six Macaque monkeys (live juvenile malt irus. one adult female assamese) received unilateral cortical leaions of pcriarcuate cortex (two irus). posterior parietal cortex (two irus), or principal sulcus (one irus. one assamese) after training to criterion on a conditional position discrimination. The initial arzuate and parietal lesions were ba!anccd as to side; both principal sulcur lesions were In the left hemisphere. A second homologous lesion was performed on recovery from discrimination delicits or approx 6 weeks after the initial resection in the case of the animals given prmcipal sulcus lesions. Surgery -as carried out aseptically under metofane and nitrous oxide inhalation anesthesia. with fluid and electrolq te replacement ~,a an intravenous drip. The cortex for each lesion was exposed by a trephinc hole expanded with rongeurs. Tissue was removed by aspiration with a finely-drawn glass pipette. and bleeding was controlled by packing with cottonoid patties. The operations were completed by closing the dura with individual sutures and closing muscle, subcutaneous tissues. and skin In anatomical layers. To conclude, long-acting bicillin (300 000 U, i.m.) was admInistered. The intended arcas ofthc lesions were as follows. The pcriarcuate lesion was directed at area 8. the anterlor hank and the depth of the arcuate sulcus from the tip of the superior ramus to the end of the inferior branch. extending approx 3 mm rostrally. Posterior parietal lesions were aimed at an area slightly more extended than 7a. including part ofthc polysensory region inferior to it [?7], from the parieto-occipital notch ventrally along the anterior banks of the superior temporal sulcu~ and lateral sulcus to a horizontal line orginating at the bottom of the intraparietal sulcus. The anterior and superior reach of this lesion was the inferior bank of the lntraparictal sulcus. Resections of the principal sulcus were to extend through its entiret). including both banks and its depth.
At the end ofthc cxpcriment, the animala were adm~nixtercd a lethal barbiturate dose and perfuhed intracardiall> with saline and 10% formalin. The brains were removed. photographed on each side, and then hardened in formalin and blocked coronally. Sections of40 /cm through the IeSlons were taken; every tenth section v,ab saved and stained by metachromatic thionin for microscopic examination The lateral view of each lesion wah reconstructed from the photographs and from the coronal sections. and the coronal view5 \*ere traced from sections cvcry I .h (arcuatc and principal sulcus) to 2 mm (parietal). Figure I shows the reconstructions ofeach lesion. The arcuatc lesions ucrc as specified In monkey ARC-J. The lesions ofARCivere moreextensivc rostrally on both Gdes and in\adcd arc;, 6 on the right. There was ~IDO some incursion into white matter on both sides. ARC-4 did not differ from .4RC-2 except 111 the somewhat prcatcr scvcrity of deficit from ARC-~‘S second (right) lesion in the lirst 3 weeks. The right (lirst) parletal lesion of monkey PAR-5 wab as planned. but the left leston was smaller. On each side, cortical tlbsuc to rhc depth ofthc inferior bank ofthe intraparietal sulcus was removed. PAR-6 austalncd much larger Icslom on each Gdc. extending posteriorly through middle superior temporal cortex to the lunate sulcu\ on the right. and lnt,) somatoscnsory cortex anterior and Inferior to the intraparietal sulcu~ on the left. The Icslon\ Ncrc Jrq’ in each
CONDITIONAL
POSITION DlSCRlMlNATlON
ARC-4 @a
PAR-5
PAR-6 @_gi&
PRIN-3
PRIN - 303
FIG. I. Lateral views and representative coronal sections for each lesion. The lateral views were reconstructed from photographs of the brains and from coronal sections. ARC=arcuate; PAR = parietal; PRIN = principal sulcus. The dark lines on the coronal sections denote damaged or destroyed tissue.
1122
DOUGI AS P. ~‘KOWNF
YI U/
hemisphere, with obvious damage to white matter. These differences in the lesions of PAR-5 and PAR-6 were reflected in their deficits and recovery. Monkey PAR-6 was more impaired over the first 4 weeks following its second (right) lesion, and this monkey also required a greater number of trials to reach criterion after both first and second lesions than PAR-5 (see Table I). The principal sulcus lesions were as intended.
panel Training was conducted in a housing-case six compartment on the far wall of which was a 90 semicircular containing 19 recessed buttons illuminated from the rear. Thirteen of the buttons, six on each side of center. were used in this experiment. The monkeys were first required to press the lit center button for banana pellet reward (Noyes, 190 mg). delivered to a foodwell beneath that button. In the next step ofthe shaping procedure, pressing the lit center button illuminated a second button in a randomly-appearing position elsewhere in the array. The response to this second button was rewarded. Further training and testing on this task. which has been previously dcscrihed 171. was concurrent with the spatial problem. The conditional position discrimination was introduced as soon as the monkeys were proficient in responding to the second button. The center light now blinked at one of two rates 0.5 or IO HZ. Pressmg it brought on a pair ofsteadily-lit buttons in varying symmetrical locations on each side ofcenter. The slow rate signalled the right button. the fast rate the button to the left. There was a 3 set ITI. Forty-two trials a day were prescntcd. A correction procedure was used in thcearly training with continuous repetition until a correct response was made. succeeded by non-correction when the performance approximated 75%~. Conditional cue and stimulus presentation, recording of responses. and delivery of reinforcement were controlled by computer. As soon as the animals were close to criteria1 performance. they wcrc switched to a second, similar apparatus. with the button array subtending I80 Immediately in front of the button array was a flat vertical panel of heavy aluminum sheet with a slot 5 cm high extending 55 cm from one side of the button array to the other through which the monkeys could reach to press the buttons. A centrally-placed oval cutout of 8 x 7 cm above the slot gave visual data access to the entire array. Training was carried to 90%) correct on 2 successive days. Ten days of preoperattve were then collected, followed by the initial unilateral lesions. Testing was resumed 4 IO days postoperatively. continuing at 5 days per week. 4’2 trials per day, until the preoperative criterion was reattained. The second lesion was produced. and the same regimen was again taken up. continuing to the 2-day criterion or 5460 trials (130days).
RESULTS The six spatial stimuli on each side ofthe center button were combined into four groups for analysis: the three outermost buttons on the side ipsilateral to the lesion (wide ipsilateral); the three buttons adjacent to the center button ipsilateral to the lesion (central ipsilateral); the three outer buttons contralateral to the lesion (wide contralateral); and the three buttons adjacent to the center button contralateral to the lesion (central contralateral). Per cent correct responses for each of 4 weeks were calculated for these stimulus groups and transformed by the arcsin transformation for proportions [24]. The first analysis was of these 4 weeks (840 trials) following each lesion, and the unit of analysis. as in the work of EIDELBERG and SCHWARTZ 181, was lesions. The stimulus position means of the three lesion groups over 4 weeks are displayed in Figs 24. Figures 2 and 3 show nearly perfect ipsilateral performance and striking contralateral impairments following the arcuate and parietal lesions, with evidence of recovery in the third and fourth weeks. The arcuate group appears to have the more severe initial deficit (but see below). The data of Fig. 4 do not suggest a decrement from the principal sulcus lesions. An analysis of variance (lesion groups x stimulus location x weeks) resulted in a three-way interaction (F [lg, 81]=3.18. P
CONDITIONAL
POSITION
DlSCRlMlNATlON
1123 0 u x A
WEEKI WEEK2 WEEK3 WEEK4
-___ v’ilDE(l) POSlTlON
CENTRAL(l) OF CORRECT
CENTRAL(C)
WIDE(C)
STIMULUS(l=IPSl,C:CONTRA)
FIG. 2. Mean per cent correct responses (arcsin transformed) at each stimulus position in the first 4 weeks following the arcuate lesions. Bars above/below each data point are SEMs.
0 hEEY’ 0 WEEK2 x WEEK3 b WEEK4
GI
1 ‘NIDE(II
POSITIONOF FIG.
CENTRALIll CORRECT
CENTRAL(C)
WDEICI
STlMULUSil=lPSI,C=CONTRA)
3. Mean per cent correct responses (arcsin transformed) at each stimulus position in the first 4 weeks following the parietal lesions. Bars above/below each data point are SEMs.
WIDE(I) POSITIONOF
CENTRAL(I) CORRECT
CENTRAL(C) STIMULUS
(I:IPSI,
WIDE(C) C=CONTRA)
FIG. 4. Mean percent correct responses (arcsin transformed) at each stimulus position in the first 4 weeks following the principal sulcus lesions. Bars above/below each data point are SEMs.
DOIJGLAS P. CKOWI\;F et al
1124
contralateral comparisons, the arcuate and parietal groups had marked deficits relative to the principal sulcus group, as Figs 2-4 reveal. By week 4, only the arcuate lesion group differed from the principal sulcus group on the wide contralateral stimuli. There were no significant differences between the two impaired groups in any comparison over the 4 weeks. Contrasts between the ipsilateral and contralateral stimulus positions showed that in both arcuate and parietal lesion groups the proportion of correct responses to the contralateral stimulus positions was significantly lower than to the corresponding ipsilateral positions through week 3, except for central stimuli in the parietal group. By week 4, only the arcuate lesions resulted in poorer contralateral than ipsilateral performance on wide stimuli. None of the ipsilateralcontralateral contrasts was significant in the principal sulcus group. There was significant improvement on both contralateral central and contralateral wide stimuli from week 1 to week 4 in the arcuate lesion group, but only on wide contralateral stimuli in the parietal group. For the parietals, the weeks 14 comparison of central contralateral stimuli was marginally significant (PcO.10). The principal sulcus group did not change. A second analysis, based on groups of animals rather than lesions, examined the total number of trials to reattain the preoperative criterion or to a maximum of 5460 trials. The data are shown in Table 1. They reveal the impairments of the arcuate and parietal groups after each lesion, but particularly the more severe deficit after the second lesion in the arcuate group, both in comparison to the first lesion and to the second lesion of the parietal and principal sulcus animals. A groups x first/second lesion analysis of variance confirmed the Table
I. Number of trmla to criterion or 5460 trials after the first and second lesions
Monkeys SP3 SP303 A2 A4 P5 P6
First lesion 622 367 1851 2708 I214 3648
*SP. monkeys with principal monkeys with parietal lesions.
Second
lesion
Total
416 344 5238 5178 I649 2655 sulcua lesions;
A, monkeys
IO38 71 I 70x9 7886 2863 6303 with arcuate
lesions;
P.
second lesion effects. The interaction was significant (F 12, 3]= 13.44, P< 0.05). Tests of simple main effects showed the second arcuate lesion to produce a more severe deficit than the first (F[l, 3]=35.32, P
(P
CONDITIONAL
POSITION
DISCRIMINATION
1125
lesions, the animals’ choices were mainly directed to the side of the lesion, which appears to be an effect of the hemifield neglect that these lesions produce [b, 131. That neglect, as Figs 2 and 3 show, strikingly separated responses to the discriminanda at the midline, suggesting a central division of the visual hemifields that distinguishes this spatial task from experimental studies of neglect in which responsiveness to stimuli decreases with displacement from the meridian [7, IS]. The conditional discrimination of spatial stimuli imposes far more strict demands than simple responses to stimulus presence, and it may be that the cognitive requirements exacerbated neglect and extended it to stimuli close to the midline. Substantial recovery occurred over 840 trials, although the arcuate lesion group continued to show a deficit. While the monkeys with principal sulcus lesions, never significantly impaired, returned to the preoperative criterion in 2-3 weeks after each of their lesions, it required l&12 weeks for the parietal lesion animals to reattain criterion after their lesions. Recovery from the deficit after the first arcuate lesion took no longer than it did for the parietals, but the second lesion impaired performance for 25 weeks. The arcuate animals were only marginally worse than the parietals in the level of their performance after either lesion in the first 4 weeks; in week 2 they had fewer (but not significantly) correct responses to central contralateral stimuli than the parietal monkeys, and in week 4 only the arcuate animals continued to differ from the monkeys with principal sulcus lesions on wide contralateral stimuli. When the effects of the first and second lesions were examined, however, the second arcuate lesion produced a very persistent deficit. A spatial concept of left vs right is seriously deranged by unilateral lesions of cortical areas involved in spatial orientation and discrimination. The procedures to establish that concept-varying the location of the spatial stimuli-may have especially contributed to the impairments of both the arcuate and parietal lesion groups. As BRODY and PRIBRAM [2] noted, a “peripheral inattention hypothesis” is insufficient to account for the spatial deficits of monkeys with frontal lesions. Rather, frontal monkeys are particularly disadvantaged by “shifting of the spatial location of the stimuli”, a conclusion affirmed in an experiment by LATTO [ 141. The monkeys with parietal damage were probably also disadvantaged by the variable position of the spatial stimuli, as were parietal lesion monkeys in LATTO’S [14] shifting-landmark discrimination problem. Hemifield neglect appears to be an important feature of the conditional position discrimination impairment, however, after both frontal and parietal damage, in view of the dramatic ipsilateral response bias in the first post-lesion weeks. This marked deflection of choices strongly suggests that more than a breakdown in spatial discrimination occurred. As a general point, our results imply that other spatial deficits after lesions to these cortical areas might be considerably affected by hemifield inattention. A discrimination deficit per se-that is, a spatial impairment independent of motor or intentional difficulties-is implicated by the observation that performance was not immediately restored with the return of contralateral responses. Instead, it required many hundreds of trials subsequent to recovery from unilateral neglect to reach the criteria1 level. For the arcuate and parietal lesion animals, then, three levels of difficulty seem to be implicated in the deficit: a transient initial neglect of stimuli in the hemifield contralateral to the lesion; a defect in the ability to locate places designated by the conditional cue (the wellestablished impairment on this problem previously shown in place-learning versions of it [ 10, 171; and an inability to associate relatiue left and right stimulus positions with the cue, a true spatial deficit. The strong commonality in the effects of posterior parietal and periarcuate lesions could be expected on the basis of their reciprocal connections [b, 16, 18, 193, but there is another
1126
DOUGLAS P. CK~WNE er a/
anatomical reason why these two areas, so far removed from each other, would have functional similarities. Thalamocortical projections to both premotor/prefrontal regions and posterior cortex in the parietal lobe have thalamic inputs originating medially to laterally in distinction to the reversed thalamic terminations of the middle peri-Rolandic band receiving thalamic projections from the ventrolateral nuclei [.5, 211. As Pribram [21] points out, this thalamic resemblance implies a functional one, and in the cases of apraxia, neglect, and the combination of neglect and disordered spatial performance we show here, the implication is clearly borne out. To summarize, our findings show that the contributions of frontal and parietal association cortex to conditional spatial discrimination have a common element in directing attention to hemispace. Both these cortical regions also take part in the representation of egocentric space. It is significant for the latter conclusion that humans with left parietal damage show both egocentric and allocentric impairments 14. 23, 25). Acknowledyements-We are greatly Kathy Blom who did the histology, apparatus.
indebted to Dr Bruce Bolster for his thoughful review of the manuscript. to and to Gerry Blom and Chris Wingelaar who designed and built the testing
REFERENCES 1. BAIIAL, J. Contribution 2. 3. 4. 5. 6. 7.
8. 9. IO. II. 12. 13. 14. 15. 16.
17.
18. 19.
ii l’ktude des c&it&s psychiques: alexie, agraphie, hkmianopsie infkrieure, trouble du sens de I’espace. Archs. Uphral. (Paris) 8, 97 117. 1888. BKODY. B. A. and PKIBRAM, K. H. The role of frontal and parietal cortex m cognitive processing tests of spatial and sequence functions. Brain 101, 607 633, 1978. BUTTERS, N. and BARTON, M. Effects of parietal lobe damage on the performance of reversible operations in space. Neuropsyholoyia 8, 205~-214, 1970. BUTTERS. N., SOELUNEK. C. and FEIIIO, P. Comparison of parietal and frontal lobe spatial deficits in man: extrapersonal vs personal (egocentric) space. Percept. mof Skills 34, 27-34. 1972. CHOW, K. L. and PRIBRAM, K. H. Cortical projection of the thalamic ventrolateral nuclear group in monkeys. J. camp. New-o/. 104, 57 75, 1956. CKOWNE, D. The frontal eye field and attention. Psyc,hol. Bull. 93, 232--260. 19X3. CKOWNE, D. P.. YEO, C. H. and STEELE RUSSELL. 1. The effects of unilateral frontal eye field lesions in the monkey: visual-motor guidance and avoidance behaviour. Behat:. Brain Res. 2, 165-187, 1981. EXXLBEKG, E. and SCHWARTZ, A. S. Experimental analysis of the extinction phenomenon in monkeys. Brain 94, 91~ 108, 1971. FERRIER, D. The Func~tions ofthr Bruin (2nd ed). Smith. Elder, London, 1886. GOLDMAN, P. S. and ROSVOLD, H. E. Localization of function within the dorsolateral prefrontal cortex of the rhesus monkey. E-up. Neural. 27, 291-304. 1970. HOLMES, G. Disturbances of visual orientation. Br. J. Ophrhal. 2, 449 468, 506-516, 1918. HOLMES, G. and HOKKAX. G. Disturbances of spatial orientation and visual attention. with loss of stereoscopic vision. Archs Nuurol. Psyhiar. (Chicago) 1, 385 407, 1919. HYV~INEN, J. The Purietul Cortex of Monkey and Mm. Springer-Verlag, Berlin, 1982. LATTO, R. The role of inferior parietal cortex and the frontal eye-fields in visuospatial discriminations in the macaque monkey. Behar~. Brain Res. 22, 41 52, 1986. LATTO, R. and COWEY, A. Visual field defects after frontal eye-field lesions in monkeys. Brain Res. 30. I 24. 1971. MESULAM, M.-M.. VAN HOESEX. G. W.. PANVYA. D. N. and GESCHWIND. N. Limbic and sensory connections of the inferior parietal lobule (area PC) in the rhesus monkey: a study with a new method for horseradish peroxidase histochemistry. Brain Res. 136, 393-414. 1977. MILNER, A. D., FOKEMAX, N. P. and GOODAL~. M. A. Go-left go-right discrimmation performance and distractibility following lesions of prefrontal cortex or superior colliculus in stumptail macaques. Neuropsychologia 16, 381 ~390, 1978. PANIIYA. D. N. and VIGIZOLO. L. A. Interhemispheric projections of the parietnl lobe in the rhesus monkey. Brain Res. 15, 49 65. 1969. PANI)YA. D. N. and VIONOLO, L. A. Intra- and inter-hemispheric projections of the precentral. prcmotor. and arcuate areas in the rhesus monkey. Brtrin Res. 26, 217 233, 1971.
CONDITIONALPOSITION DISCKIMINATION
1127
20. POHL, W. Dissociation of spatial discrimination deficits following frontal and parietal lesions in monkeys. J. camp. physiol. Psychol. 82, 227-239, 1973. 21. PRIBRAM, K. H. Subdivisions of the frontal cortex revisited. In The Frontal Lobes Reoisited, BROWS, E. and PERECMAN E. (Editors). IRBN Press, New York, 1987. 22. ROSENKILDE,C. E. Functional heterogeneity of the prefrontal cortex in the monkey: a review. Beha. neural Biol. 25,301&345, 1979. 23. SEMMES,J., WEINSTEIN, S., GHENT, L. and TEUBEK, H.-L. Correlates of impaired orientation in personal and extrapersonal space. Brain 86, 747-772, 1963. 24. SNEDECOR,G. W. and COCHRAN, W. G. Statistical Merhods (6th ed). Iowa State University Press, Ames, Iowa. 1967. 25. TEUBEK, H.-L. The riddle of frontal lobe function in man. In The Frontal Granular Cortex and Behariour, WARREN, J. M. and AKERT, K. (Editors). McGraw-Hill, New York, 1964. 26. TEUBER, H.-L. and MISHKIN, M. Judgment of visual and postural vertical after brain injury. J. Psycho/. 38, 161 27.
175, 1954.
VAN ESSEN, D. C. and MAUNSELL, J. H. R. Hierarchical Trends
Neurosci.
6, 37C375,
1983.
organization
and functional
streams in the visual cortex.
t
co28&3932/89 $3.00+0.00 1989 Pergamon Press plc
UNILATERAL PERIARCUATE AND POSTERIOR PARIETAL LESIONS IMPAIR CONDITIONAL POSITION DISCRIMINATION LEARNING IN THE MONKEY DOUGLAS P. CROWNE, KIM A. DAWSON and CLAUDETTE M. RICHARDSON Departmentof Psychology, Universityof Waterloo, Waterloo, Ontario, Canada, N2L 3Gl (Receired
26 July
1988:
accepted
7 April
1989)
Abstract-Monkeys were trained on a conditional position discrimination in which the conditional cue was a light blinking at two distinct rates and the discriminanda (illuminated buttons) appeared in varying symmetrical positions of eccentricity. Unilateral arcuate, posterior parietal, or principal sulcus lesions were performed at criterion. The monkeys were tested to recovery when a homologous lesion was made contralaterally. The first of two analyses examined a period of4 weeks following each lesion; the unit of analysis was lesions. The arcuate and parietal lesions produced impairments on both widely eccentric and central discriminanda locations; initially. virtually all responses were deflected to the ipsilateral side. There was significant improvement after the arcuate and parietal lesions from weeks 1 to 4. An analysis of total trials to criterion showed major deficits from the second arcuate and parietal lesions, with the arcuate lesion impairment being particularly severe. These results establish that a spatial concept of left vs right is seriously deranged by unilateral lesions of cortical association areas involved in spatial orientation and discrimination.
INTRODUCTION DISORDERS of spatial
localization, orientation, and perception have been recognized in humans and in monkeys for a century [l, 93 and have been associated with parietal lobe damage since the studies of Holmes in 1918-1919 [ 11, 121. The involvement of the frontal lobes in spatial behaviour was not appreciated for another 40 yr. Semmes et al. [23] reported that frontal lesions produce a spatial defect that is clearly dissociated from the spatial semeiology of parietal injury. Their distinction between impairments of personal spatial (prefrontal cortex) and extrapersonal (parietal cortex) orientation was confirmed in studies of humans [3,4,23, 261 and monkeys [ZO] with frontal and parietal damage. BRODY and PRIBRAM [2] then showed that while posterior parietal lesions in the monkey impair performance only on tasks imposing extrapersonal orientation demands, frontal lesions degrade the ability to perform both personal and extrapersonal tasks when the stimulus context varies. Despite this important evidence distinguishing the spatial cognitive processes of frontal and parietal cortex, there are functional aspects of these regions whose implications are yet to be fully explored. It is now firmly established, for example, that there are several highly important behavioral specializations within prefrontal cortex of the monkey [22]. A spatial problem of particular interest, the conditional position discrimination, is impaired by arcuate cortex (frontal eye field) lesions. GOLDMAN and ROSVOLD [lo] presented monkeys with an auditory cue through speakers above or below the test chamber. The upper cue signified choice of the left foodwell, the lower cue the right. Periarcuate lesions severely degraded performance on this discrimination, while principal sulcus damage did not. 1119
MILNERct d. [ 171 showed that the deficit is not specific to auditory cues and is clearly thus a spatial impairment. Conditional position discrimination certainly imposes egocentric (“frontal”) demands on left--right orientation; would those features, appearing in a difficult problem, also make it vulnerable to parietal damage? In this experiment we studied the performance of monkeys with serial periarcuate. posterior parietal or principal sulcus resections on a particularly complex version of the conditional position discrimination problem. Characteristically, studies of spatial learning and orientation have placed the discriminanda in two specific locations to the left and right. Animals may thus respond to a particular locus and not on the basis of a spatial concept“to the left” or “to the right”. We sought to teach monkeys the spatial concept of left vs right, quite independent of specific place. Accordingly, the positions of the spatial stimuli varied randomly in their eccentricity from trial to trial. Our conditional position discrimination could not be solved as a place-learning problem. The evidence of spatial impairments in animals is based on bilateral lesions. To reveal more clearly the processes involved in spatial learning and performance and to enable comparisons with humans, in whom unilateral lesions produce severe and often unremitting deficits, we studied spatial discrimination after unilateral lesions in monkeys.
Six Macaque monkeys (live juvenile malt irus. one adult female assamese) received unilateral cortical leaions of pcriarcuate cortex (two irus). posterior parietal cortex (two irus), or principal sulcus (one irus. one assamese) after training to criterion on a conditional position discrimination. The initial arzuate and parietal lesions were ba!anccd as to side; both principal sulcur lesions were In the left hemisphere. A second homologous lesion was performed on recovery from discrimination delicits or approx 6 weeks after the initial resection in the case of the animals given prmcipal sulcus lesions. Surgery -as carried out aseptically under metofane and nitrous oxide inhalation anesthesia. with fluid and electrolq te replacement ~,a an intravenous drip. The cortex for each lesion was exposed by a trephinc hole expanded with rongeurs. Tissue was removed by aspiration with a finely-drawn glass pipette. and bleeding was controlled by packing with cottonoid patties. The operations were completed by closing the dura with individual sutures and closing muscle, subcutaneous tissues. and skin In anatomical layers. To conclude, long-acting bicillin (300 000 U, i.m.) was admInistered. The intended arcas ofthc lesions were as follows. The pcriarcuate lesion was directed at area 8. the anterlor hank and the depth of the arcuate sulcus from the tip of the superior ramus to the end of the inferior branch. extending approx 3 mm rostrally. Posterior parietal lesions were aimed at an area slightly more extended than 7a. including part ofthc polysensory region inferior to it [?7], from the parieto-occipital notch ventrally along the anterior banks of the superior temporal sulcu~ and lateral sulcus to a horizontal line orginating at the bottom of the intraparietal sulcus. The anterior and superior reach of this lesion was the inferior bank of the lntraparictal sulcus. Resections of the principal sulcus were to extend through its entiret). including both banks and its depth.
At the end ofthc cxpcriment, the animala were adm~nixtercd a lethal barbiturate dose and perfuhed intracardiall> with saline and 10% formalin. The brains were removed. photographed on each side, and then hardened in formalin and blocked coronally. Sections of40 /cm through the IeSlons were taken; every tenth section v,ab saved and stained by metachromatic thionin for microscopic examination The lateral view of each lesion wah reconstructed from the photographs and from the coronal sections. and the coronal view5 \*ere traced from sections cvcry I .h (arcuatc and principal sulcus) to 2 mm (parietal). Figure I shows the reconstructions ofeach lesion. The arcuatc lesions ucrc as specified In monkey ARC-J. The lesions ofARCivere moreextensivc rostrally on both Gdes and in\adcd arc;, 6 on the right. There was ~IDO some incursion into white matter on both sides. ARC-4 did not differ from .4RC-2 except 111 the somewhat prcatcr scvcrity of deficit from ARC-~‘S second (right) lesion in the lirst 3 weeks. The right (lirst) parletal lesion of monkey PAR-5 wab as planned. but the left leston was smaller. On each side, cortical tlbsuc to rhc depth ofthc inferior bank ofthe intraparietal sulcus was removed. PAR-6 austalncd much larger Icslom on each Gdc. extending posteriorly through middle superior temporal cortex to the lunate sulcu\ on the right. and lnt,) somatoscnsory cortex anterior and Inferior to the intraparietal sulcu~ on the left. The Icslon\ Ncrc Jrq’ in each
CONDITIONAL
POSITION DlSCRlMlNATlON
ARC-4 @a
PAR-5
PAR-6 @_gi&
PRIN-3
PRIN - 303
FIG. I. Lateral views and representative coronal sections for each lesion. The lateral views were reconstructed from photographs of the brains and from coronal sections. ARC=arcuate; PAR = parietal; PRIN = principal sulcus. The dark lines on the coronal sections denote damaged or destroyed tissue.
1122
DOUGI AS P. ~‘KOWNF
YI U/
hemisphere, with obvious damage to white matter. These differences in the lesions of PAR-5 and PAR-6 were reflected in their deficits and recovery. Monkey PAR-6 was more impaired over the first 4 weeks following its second (right) lesion, and this monkey also required a greater number of trials to reach criterion after both first and second lesions than PAR-5 (see Table I). The principal sulcus lesions were as intended.
panel Training was conducted in a housing-case six compartment on the far wall of which was a 90 semicircular containing 19 recessed buttons illuminated from the rear. Thirteen of the buttons, six on each side of center. were used in this experiment. The monkeys were first required to press the lit center button for banana pellet reward (Noyes, 190 mg). delivered to a foodwell beneath that button. In the next step ofthe shaping procedure, pressing the lit center button illuminated a second button in a randomly-appearing position elsewhere in the array. The response to this second button was rewarded. Further training and testing on this task. which has been previously dcscrihed 171. was concurrent with the spatial problem. The conditional position discrimination was introduced as soon as the monkeys were proficient in responding to the second button. The center light now blinked at one of two rates 0.5 or IO HZ. Pressmg it brought on a pair ofsteadily-lit buttons in varying symmetrical locations on each side ofcenter. The slow rate signalled the right button. the fast rate the button to the left. There was a 3 set ITI. Forty-two trials a day were prescntcd. A correction procedure was used in thcearly training with continuous repetition until a correct response was made. succeeded by non-correction when the performance approximated 75%~. Conditional cue and stimulus presentation, recording of responses. and delivery of reinforcement were controlled by computer. As soon as the animals were close to criteria1 performance. they wcrc switched to a second, similar apparatus. with the button array subtending I80 Immediately in front of the button array was a flat vertical panel of heavy aluminum sheet with a slot 5 cm high extending 55 cm from one side of the button array to the other through which the monkeys could reach to press the buttons. A centrally-placed oval cutout of 8 x 7 cm above the slot gave visual data access to the entire array. Training was carried to 90%) correct on 2 successive days. Ten days of preoperattve were then collected, followed by the initial unilateral lesions. Testing was resumed 4 IO days postoperatively. continuing at 5 days per week. 4’2 trials per day, until the preoperative criterion was reattained. The second lesion was produced. and the same regimen was again taken up. continuing to the 2-day criterion or 5460 trials (130days).
RESULTS The six spatial stimuli on each side ofthe center button were combined into four groups for analysis: the three outermost buttons on the side ipsilateral to the lesion (wide ipsilateral); the three buttons adjacent to the center button ipsilateral to the lesion (central ipsilateral); the three outer buttons contralateral to the lesion (wide contralateral); and the three buttons adjacent to the center button contralateral to the lesion (central contralateral). Per cent correct responses for each of 4 weeks were calculated for these stimulus groups and transformed by the arcsin transformation for proportions [24]. The first analysis was of these 4 weeks (840 trials) following each lesion, and the unit of analysis. as in the work of EIDELBERG and SCHWARTZ 181, was lesions. The stimulus position means of the three lesion groups over 4 weeks are displayed in Figs 24. Figures 2 and 3 show nearly perfect ipsilateral performance and striking contralateral impairments following the arcuate and parietal lesions, with evidence of recovery in the third and fourth weeks. The arcuate group appears to have the more severe initial deficit (but see below). The data of Fig. 4 do not suggest a decrement from the principal sulcus lesions. An analysis of variance (lesion groups x stimulus location x weeks) resulted in a three-way interaction (F [lg, 81]=3.18. P
CONDITIONAL
POSITION
DlSCRlMlNATlON
1123 0 u x A
WEEKI WEEK2 WEEK3 WEEK4
-___ v’ilDE(l) POSlTlON
CENTRAL(l) OF CORRECT
CENTRAL(C)
WIDE(C)
STIMULUS(l=IPSl,C:CONTRA)
FIG. 2. Mean per cent correct responses (arcsin transformed) at each stimulus position in the first 4 weeks following the arcuate lesions. Bars above/below each data point are SEMs.
0 hEEY’ 0 WEEK2 x WEEK3 b WEEK4
GI
1 ‘NIDE(II
POSITIONOF FIG.
CENTRALIll CORRECT
CENTRAL(C)
WDEICI
STlMULUSil=lPSI,C=CONTRA)
3. Mean per cent correct responses (arcsin transformed) at each stimulus position in the first 4 weeks following the parietal lesions. Bars above/below each data point are SEMs.
WIDE(I) POSITIONOF
CENTRAL(I) CORRECT
CENTRAL(C) STIMULUS
(I:IPSI,
WIDE(C) C=CONTRA)
FIG. 4. Mean percent correct responses (arcsin transformed) at each stimulus position in the first 4 weeks following the principal sulcus lesions. Bars above/below each data point are SEMs.
DOIJGLAS P. CKOWI\;F et al
1124
contralateral comparisons, the arcuate and parietal groups had marked deficits relative to the principal sulcus group, as Figs 2-4 reveal. By week 4, only the arcuate lesion group differed from the principal sulcus group on the wide contralateral stimuli. There were no significant differences between the two impaired groups in any comparison over the 4 weeks. Contrasts between the ipsilateral and contralateral stimulus positions showed that in both arcuate and parietal lesion groups the proportion of correct responses to the contralateral stimulus positions was significantly lower than to the corresponding ipsilateral positions through week 3, except for central stimuli in the parietal group. By week 4, only the arcuate lesions resulted in poorer contralateral than ipsilateral performance on wide stimuli. None of the ipsilateralcontralateral contrasts was significant in the principal sulcus group. There was significant improvement on both contralateral central and contralateral wide stimuli from week 1 to week 4 in the arcuate lesion group, but only on wide contralateral stimuli in the parietal group. For the parietals, the weeks 14 comparison of central contralateral stimuli was marginally significant (PcO.10). The principal sulcus group did not change. A second analysis, based on groups of animals rather than lesions, examined the total number of trials to reattain the preoperative criterion or to a maximum of 5460 trials. The data are shown in Table 1. They reveal the impairments of the arcuate and parietal groups after each lesion, but particularly the more severe deficit after the second lesion in the arcuate group, both in comparison to the first lesion and to the second lesion of the parietal and principal sulcus animals. A groups x first/second lesion analysis of variance confirmed the Table
I. Number of trmla to criterion or 5460 trials after the first and second lesions
Monkeys SP3 SP303 A2 A4 P5 P6
First lesion 622 367 1851 2708 I214 3648
*SP. monkeys with principal monkeys with parietal lesions.
Second
lesion
Total
416 344 5238 5178 I649 2655 sulcua lesions;
A, monkeys
IO38 71 I 70x9 7886 2863 6303 with arcuate
lesions;
P.
second lesion effects. The interaction was significant (F 12, 3]= 13.44, P< 0.05). Tests of simple main effects showed the second arcuate lesion to produce a more severe deficit than the first (F[l, 3]=35.32, P
(P
CONDITIONAL
POSITION
DISCRIMINATION
1125
lesions, the animals’ choices were mainly directed to the side of the lesion, which appears to be an effect of the hemifield neglect that these lesions produce [b, 131. That neglect, as Figs 2 and 3 show, strikingly separated responses to the discriminanda at the midline, suggesting a central division of the visual hemifields that distinguishes this spatial task from experimental studies of neglect in which responsiveness to stimuli decreases with displacement from the meridian [7, IS]. The conditional discrimination of spatial stimuli imposes far more strict demands than simple responses to stimulus presence, and it may be that the cognitive requirements exacerbated neglect and extended it to stimuli close to the midline. Substantial recovery occurred over 840 trials, although the arcuate lesion group continued to show a deficit. While the monkeys with principal sulcus lesions, never significantly impaired, returned to the preoperative criterion in 2-3 weeks after each of their lesions, it required l&12 weeks for the parietal lesion animals to reattain criterion after their lesions. Recovery from the deficit after the first arcuate lesion took no longer than it did for the parietals, but the second lesion impaired performance for 25 weeks. The arcuate animals were only marginally worse than the parietals in the level of their performance after either lesion in the first 4 weeks; in week 2 they had fewer (but not significantly) correct responses to central contralateral stimuli than the parietal monkeys, and in week 4 only the arcuate animals continued to differ from the monkeys with principal sulcus lesions on wide contralateral stimuli. When the effects of the first and second lesions were examined, however, the second arcuate lesion produced a very persistent deficit. A spatial concept of left vs right is seriously deranged by unilateral lesions of cortical areas involved in spatial orientation and discrimination. The procedures to establish that concept-varying the location of the spatial stimuli-may have especially contributed to the impairments of both the arcuate and parietal lesion groups. As BRODY and PRIBRAM [2] noted, a “peripheral inattention hypothesis” is insufficient to account for the spatial deficits of monkeys with frontal lesions. Rather, frontal monkeys are particularly disadvantaged by “shifting of the spatial location of the stimuli”, a conclusion affirmed in an experiment by LATTO [ 141. The monkeys with parietal damage were probably also disadvantaged by the variable position of the spatial stimuli, as were parietal lesion monkeys in LATTO’S [14] shifting-landmark discrimination problem. Hemifield neglect appears to be an important feature of the conditional position discrimination impairment, however, after both frontal and parietal damage, in view of the dramatic ipsilateral response bias in the first post-lesion weeks. This marked deflection of choices strongly suggests that more than a breakdown in spatial discrimination occurred. As a general point, our results imply that other spatial deficits after lesions to these cortical areas might be considerably affected by hemifield inattention. A discrimination deficit per se-that is, a spatial impairment independent of motor or intentional difficulties-is implicated by the observation that performance was not immediately restored with the return of contralateral responses. Instead, it required many hundreds of trials subsequent to recovery from unilateral neglect to reach the criteria1 level. For the arcuate and parietal lesion animals, then, three levels of difficulty seem to be implicated in the deficit: a transient initial neglect of stimuli in the hemifield contralateral to the lesion; a defect in the ability to locate places designated by the conditional cue (the wellestablished impairment on this problem previously shown in place-learning versions of it [ 10, 171; and an inability to associate relatiue left and right stimulus positions with the cue, a true spatial deficit. The strong commonality in the effects of posterior parietal and periarcuate lesions could be expected on the basis of their reciprocal connections [b, 16, 18, 193, but there is another
1126
DOUGLAS P. CK~WNE er a/
anatomical reason why these two areas, so far removed from each other, would have functional similarities. Thalamocortical projections to both premotor/prefrontal regions and posterior cortex in the parietal lobe have thalamic inputs originating medially to laterally in distinction to the reversed thalamic terminations of the middle peri-Rolandic band receiving thalamic projections from the ventrolateral nuclei [.5, 211. As Pribram [21] points out, this thalamic resemblance implies a functional one, and in the cases of apraxia, neglect, and the combination of neglect and disordered spatial performance we show here, the implication is clearly borne out. To summarize, our findings show that the contributions of frontal and parietal association cortex to conditional spatial discrimination have a common element in directing attention to hemispace. Both these cortical regions also take part in the representation of egocentric space. It is significant for the latter conclusion that humans with left parietal damage show both egocentric and allocentric impairments 14. 23, 25). Acknowledyements-We are greatly Kathy Blom who did the histology, apparatus.
indebted to Dr Bruce Bolster for his thoughful review of the manuscript. to and to Gerry Blom and Chris Wingelaar who designed and built the testing
REFERENCES 1. BAIIAL, J. Contribution 2. 3. 4. 5. 6. 7.
8. 9. IO. II. 12. 13. 14. 15. 16.
17.
18. 19.
ii l’ktude des c&it&s psychiques: alexie, agraphie, hkmianopsie infkrieure, trouble du sens de I’espace. Archs. Uphral. (Paris) 8, 97 117. 1888. BKODY. B. A. and PKIBRAM, K. H. The role of frontal and parietal cortex m cognitive processing tests of spatial and sequence functions. Brain 101, 607 633, 1978. BUTTERS, N. and BARTON, M. Effects of parietal lobe damage on the performance of reversible operations in space. Neuropsyholoyia 8, 205~-214, 1970. BUTTERS. N., SOELUNEK. C. and FEIIIO, P. Comparison of parietal and frontal lobe spatial deficits in man: extrapersonal vs personal (egocentric) space. Percept. mof Skills 34, 27-34. 1972. CHOW, K. L. and PRIBRAM, K. H. Cortical projection of the thalamic ventrolateral nuclear group in monkeys. J. camp. New-o/. 104, 57 75, 1956. CKOWNE, D. The frontal eye field and attention. Psyc,hol. Bull. 93, 232--260. 19X3. CKOWNE, D. P.. YEO, C. H. and STEELE RUSSELL. 1. The effects of unilateral frontal eye field lesions in the monkey: visual-motor guidance and avoidance behaviour. Behat:. Brain Res. 2, 165-187, 1981. EXXLBEKG, E. and SCHWARTZ, A. S. Experimental analysis of the extinction phenomenon in monkeys. Brain 94, 91~ 108, 1971. FERRIER, D. The Func~tions ofthr Bruin (2nd ed). Smith. Elder, London, 1886. GOLDMAN, P. S. and ROSVOLD, H. E. Localization of function within the dorsolateral prefrontal cortex of the rhesus monkey. E-up. Neural. 27, 291-304. 1970. HOLMES, G. Disturbances of visual orientation. Br. J. Ophrhal. 2, 449 468, 506-516, 1918. HOLMES, G. and HOKKAX. G. Disturbances of spatial orientation and visual attention. with loss of stereoscopic vision. Archs Nuurol. Psyhiar. (Chicago) 1, 385 407, 1919. HYV~INEN, J. The Purietul Cortex of Monkey and Mm. Springer-Verlag, Berlin, 1982. LATTO, R. The role of inferior parietal cortex and the frontal eye-fields in visuospatial discriminations in the macaque monkey. Behar~. Brain Res. 22, 41 52, 1986. LATTO, R. and COWEY, A. Visual field defects after frontal eye-field lesions in monkeys. Brain Res. 30. I 24. 1971. MESULAM, M.-M.. VAN HOESEX. G. W.. PANVYA. D. N. and GESCHWIND. N. Limbic and sensory connections of the inferior parietal lobule (area PC) in the rhesus monkey: a study with a new method for horseradish peroxidase histochemistry. Brain Res. 136, 393-414. 1977. MILNER, A. D., FOKEMAX, N. P. and GOODAL~. M. A. Go-left go-right discrimmation performance and distractibility following lesions of prefrontal cortex or superior colliculus in stumptail macaques. Neuropsychologia 16, 381 ~390, 1978. PANIIYA. D. N. and VIGIZOLO. L. A. Interhemispheric projections of the parietnl lobe in the rhesus monkey. Brain Res. 15, 49 65. 1969. PANI)YA. D. N. and VIONOLO, L. A. Intra- and inter-hemispheric projections of the precentral. prcmotor. and arcuate areas in the rhesus monkey. Brtrin Res. 26, 217 233, 1971.
CONDITIONALPOSITION DISCKIMINATION
1127
20. POHL, W. Dissociation of spatial discrimination deficits following frontal and parietal lesions in monkeys. J. camp. physiol. Psychol. 82, 227-239, 1973. 21. PRIBRAM, K. H. Subdivisions of the frontal cortex revisited. In The Frontal Lobes Reoisited, BROWS, E. and PERECMAN E. (Editors). IRBN Press, New York, 1987. 22. ROSENKILDE,C. E. Functional heterogeneity of the prefrontal cortex in the monkey: a review. Beha. neural Biol. 25,301&345, 1979. 23. SEMMES,J., WEINSTEIN, S., GHENT, L. and TEUBEK, H.-L. Correlates of impaired orientation in personal and extrapersonal space. Brain 86, 747-772, 1963. 24. SNEDECOR,G. W. and COCHRAN, W. G. Statistical Merhods (6th ed). Iowa State University Press, Ames, Iowa. 1967. 25. TEUBEK, H.-L. The riddle of frontal lobe function in man. In The Frontal Granular Cortex and Behariour, WARREN, J. M. and AKERT, K. (Editors). McGraw-Hill, New York, 1964. 26. TEUBER, H.-L. and MISHKIN, M. Judgment of visual and postural vertical after brain injury. J. Psycho/. 38, 161 27.
175, 1954.
VAN ESSEN, D. C. and MAUNSELL, J. H. R. Hierarchical Trends
Neurosci.
6, 37C375,
1983.
organization
and functional
streams in the visual cortex.