Reversible Deficit in Haptic Delay Tasks from Cooling Prefrontal Cortex

Transcription

1 Reversible Deficit in Haptic Delay Tasks from Cooling Prefrontal Cortex Waleed W. Shindy, Keith A. Posley, and Joaquin M. Fuster Brain Research Institute and Department of Psychiatry, School of Medicine, University of California, Los Angeles, California The main purpose of this study was to explore the role of dorsolateral prefrontal cortex in skilled and sequential haptic performance. Monkeys were trained to perform a delayed matching-to-sample task that required the memorization of three-dimensional objects perceived either by palpation (haptically) or by tight At the start of a trial the animal was allowed to touch or view an object, the sample; after a period of delay, during which the object remained out of touch and out of sight, the animal was presented with two side-by-side objects one of them the sample for either tactile or visual recognition, and the choice of the sample (correct match) was rewarded. Three variants of the task were used: (1) visual sample, haptic match; (2) haptic sample, visual match; and (3) haptic sample, haptic match. The temporary bilateral cooling of dorsolateral prefrontal cortex to 15 C induced a reversible deficit in performance of all three tasks. Cooling to the same degree a portion of posterior parietal cortex of equivalent size did not significantly alter either performance or reaction time. These findings indicate that the functional integrity of the dorsolateral prefrontal cortex is important for performance of sequential behavior dependent on haptic skill. Further, the results suggest that the role of this cortex in active memory, already well documented for spatially and nonspatially defined visual information, extends also to tactile information and associated motor acts. Ablations of the dorsolateral prefrontal cortex in the monkey induce deficits in performance of a number of delay tasks (see Fuster, 1989, for review). Because most of those deficits have been shown in spatial tasks, such as delayed response and delayed alternation, the role of the prefrontal cortex in short-term memory has long been thought to apply, mainly if not exclusively, to the retention of spatially defined cues. This presumption, supported by a long series of studies of prefrontal neuron activity in spatial delay tasks (e.g., Fuster, 1973; Funahashi et al., 1989), has become firmly entrenched in the literature despite several electrophysiological and neuropsychological challenges. One such challenge is the evidence that reversible lesions of the dorsolateral prefrontal cortex cause an impairment in visual delayed matching to sample (DMS), a task in which the memorandum is not spatially defined (Bauer and Fuster, 1976). The selective ablation of the cortex of the inferior prefrontal convexity, namely, the portion of dorsolateral prefrontal cortex that lies below the sulcus principalis, induces a similar deficit (Mishkin and Manning, 1978). Another challenge is the evidence of prefrontal cells involved in color memory (Fuster et al., 1982; Quintana et al., 1988; Quintana and Fuster, 1992). Apart from the spatial versus nonspatial controversy, there is the issue that all reported memory deficits from prefrontal lesion in monkeys are of the visual modality, whether the visual memorandum is spatial or not. Yet there is human neuropsychological evidence, albeit sparse and subject to the topographic uncertainty of clinical lesions, suggesting the sensory supramodality of dorsolateral prefrontal deficits (Lewinsohn et al., 1972). It is important, therefore, to test the hypothesis that the memory role of the dorsolateral prefrontal cortex transcends vision and applies to other sensory modalities as well. The purpose of this work was to examine this hypothesis with respect to touch and, more generally, the involvement of that cortex in behavior guided by the orderly use of tactile stimuli. This was accomplished by testing the effects of the reversible cryogenic lesion of a large portion of the dorsolateral prefrontal cortex (areas 9, 46) on performance of a DMS task with stimuli perceived by active touch (i.e., haptically). The results were contrasted with those obtained by cooling posterior parietal cortex, which is involved in the retention and use of spatial information (Quintana and Fuster, 1993). Cerebral Cortex July/Aug 1994;4:443-45O; /94/14.00

2 SAMPLE DELAY CHOICE V - H H - V H - H Figure 1. Diapam of the three vanaras of the delayed maichmg-to-sanpie task used for the expenmenr visual-haptic \V-H). haptic-visual \H-V), and haptic-haptic \H-H). See Materials and Methods for details. Two of the task variants used for this study required the cross-modal and cross-temporal transfer of information about touch and vision. A preliminary report of some of the data has been previously published (Fuster et al., 1990). 444 Prefrontal Cortex and Hapncs Shindy el al Materials and Methods Subjects and Apparatus Three male, adult, rhesus monkeys (8-10 kg) were used for this study. The animals (subjects J, R, and

3 E) were experimentally naive at the start of the experiment. They were housed in large individual cages and maintained on a 12 hr light/dark cycle Since liquid reinforcement was used for the experiment, water availability was restricted for about 20 hr before any given experimental session. During testing (Fig. 1), the experimental animal sat in a primate restraining chair inside a sound-attenuated chamber with background masking noise. A metal plate attached to one side of the chair obliged the animal to use only one hand for the task (subjects J and R the left, subject E the right). The monkey faced a panel with a Lucite window equipped with a vision-occluding, vertically sliding shutter that, when open, allowed visualization of the test objects Below the visual shutter was a haptic shutter that, when open, provided manual but not visual access to the test objects. Located immediately below the haptic opening and shutter was a hand mold on which the animal was trained to rest its operating hand between trials. Fruit juice reinforcement was administered through a spigot by an electrically operated fluid dispenser with a solenoid valve. Two aluminum test objects, a sphere (20 mm diameter) and a cube (15 mm sides), both of smooth surface, were used throughout this experiment. The two objects were attached to a mechanized fork and could be presented one at a time or in pairs. The fork could be moved into two positions: in the "sample position," only one object could be viewed or touched by the animal; in the "choice position," both objects could be viewed and/or touched. The hand mold and objects were included in electronic sensing circuits, allowing for the measurement of tactile contacts and responses. The shutters, objects, and feeder were operated automatically by a personal computer. Details of training have been described elsewhere (DiMattia et al., 1990). Behavioral Testing In the fully trained monkey, three variants or modes of the delayed matching task were used (Fig. 1). (1) In visual-baptic (V-H) mode, the animal was first visually presented the sample object, cube or sphere, for 5 sec; the shutter was then closed and, after a delay, the two test objects were presented for haptic choice. A forceful electronically detectable pull of the object matching the sample dispensed juice reward. Pulling the other object ended the trial without reinforcement (2) In baptic-visual (H-V) mode, the sample was presented for touch only; after the delay, both the visual and haptic shutters were opened, allowing the animal to choose the correct (i.e., sample) object, for which he was rewarded. If the first object touched was the incorrect one, the shutters were closed and the trial ended without reinforcement. (3) In hapticbaptic (H-H) mode, only the haptic shutter was opened at both sample and choice, and thus the sampling and matching were done only by touch. In all three modes of testing, a discrete sound signaled the accessibility of test objects for haptic sampling or matching. Reaction time was measured from the start of choice-objects presentation (sound signal for hap- Hgera 2. Schematic drawing ol two ceding probes above a frontal view of the brain A Peltier thermode is shown attached to one of them The thermode consists of a stack of dissvnilar metal plates. Passage of direct current through the stack induces a temperature gradient in it (Pettier effect), whereby heat can be extracted from the cortex underlying the probe and dissipated cnto a circulating water system here simplified as a U-shaped tuba tic choice) to the pull of the chosen object. The sample for each trial and its position at the time of choice were changed randomly. An animal was considered fully trained on a task when its performance consistently equaled or exceeded 75% correctness with 10 sec delays. Surgical Preparation After training was completed on all three tasks, epidural cooling probes were surgically implanted under general Nembutal anesthesia over prefrontal and parietal conex on both hemispheres (Quintana and Fuster, 1993). The circular cooling surface of each probe was 2.77 cm 2. The prefrontal probes covered a region bisected by the sulcus principalis, including major portions of Brodmann's areas 9 and 46. The parietal probes covered an area of the same size that included portions of the superior and inferior parietal lobules (areas 5 and 7) Thermistors were implanted subdurally under each probe for temperature monitoring and control. In addition, head fixation sockets were attached to the skull with acrylic cement. Dexamethasone was administered for prevention of edema. Antibiotics were topically and systemically administered to prevent infection after surgery. Experimental Procedures The animals were allowed a minimum of 10 days to recover from surgery before experiments. At the be- Cerebral Cortex Julv/Aug 1994, V 4 N 4 446

4 Tibia 1 MANOVA to prefromal cooling 15 C) Perfonrencs Reaction ume df F P F P Monkey Task Delay Condition (tetnpennurel Task x condhnn Delay x condition B <0001 <0001 <0.001 < > <0.001 <0.0Ol <0.001 <001 ginning of each testing session the animal's head was immobilized and two thermoelectric coolers, operating on the Peltier principle, were attached to either the frontal or parietal probes (Fig. 2). Daily sessions were run, alternating cooling sessions with control normal temperature sessions (coolers attached but inoperative). In cooling sessions, frontal or parietal cortex was cooled bilaterally to either 15 C or 25 C and maintained at that temperature throughout the session. Each session consisted of a total of approximately 100 DMS trials of one task mode with delays of 5, 10, 20, or 40 sec (subjects R and E were also administered 60 sec delay trials). The various delays were given in blocks of five trials each, and the blocks were randomly ordered. Typically, an animal would be tested for several sessions on a given task mode (V-H, H-V, or H-H) until completing a minimum of 96 trials with each delay for each condition (cooling to 25 C or 15 C, and noncooling). The animal would then be given several days to acclimate to the next task mode before proceeding with cooling on that mode. Parietal cooling was tested only on one task mode: H-H. Data Analysts Performance and choice reaction time values were averaged for each delay and session. A multiple analysis of variance (MANOVA) for repeated measures was conducted on performance and reaction time values after, respectively, arcsine and square root transformation. Separate analyses were done for 25 C and 15 C tests, each with its separate set of control sessions. Thus, the main variables used in each analysis were monkey (3), task (3), delay (5), and condition or temperature (2). Post hoc Tukey tests were conducted to make pairwise comparisons. Results General Effects of Cortical Cooling Previous data from acute monkey preparations indicate that under and around a cortical cooling probe temperature gradients are relatively sharp (Fuster and Bauer, 1974). With cortical surface temperature at 20 C, the temperature approximately 10 mm away from the probe is about 35 C, that is, almost normal. Thus, prefrontal cooling at 15 C or 25 C, as used in the present study, could be assumed to inactivate or depress only the cortical region in and around the sulcus principalis. Parietal cooling, on the other hand, affected portions of areas 5 and 7, including some of the cortex buried in the intraparietal and superior temporal sulci. Bilateral cooling of either prefrontal or parietal cortex failed to induce any observable effects on the behavior of the animals outside the testing environment. Neither ocular motility nor manual dexterity appeared to be impaired at 15 C or 25 C. Nor did cortical cooling seem to affect the animals' motivation to perform the tasks. Cortical Cooling on Task Performance Individually, the three monkeys differed somewhat in their proficiency at the tasks. On average, and under normal conditions (control), the haptic-visual (H-V) task seemed more difficult for the animals to perform than the other two tasks, V-H and H-H. Also at normal temperature, performance of all three tasks generally showed a decrement as a function of delay. Choice reaction time was generally longer as a function of delay; that delay-related increment was most obvious between 5 and 10 sec delays in H-V and H-H. Bilateral prefrontal cooling to 15 C induced a reversible deficit in performance across all tasks and delays (Table 1, Fig. 3). In each of the three tasks, this effect of cooling attained high statistical significance (p < 0.001). Interactions between prefrontal cooling and task or delay were insignificant. As Figure 3 illustrates, the cooling performance curves for the three tasks followed the same general downward trend as the control curves. Choice reaction time was slightly but significantly shorter under prefrontal cooling than under control condition in the H-V task (p < 0.001). Reaction time was not affected in the other two tasks. Prefrontal cooling to 25 C elicited performance deficits of lesser magnitude than at 15 C on Figure 3. Effects of tnlaieial prefrontal codmg to 15 C on performance of the three ttsks [V-H. H-V, and H-H] Left atom Chocs reaction tune The values ptoned are averages from the three animals Percentage of correct responses. Right column. 448 Prefrontal Cortex and Haptlcs Shindy et al

5 CONTROL. NORMAL TEMPERATURE PREFRONTAL COOL 1 5 C n b O u H D. 100 V-H V-H H-V Q z o u H z o i H-V H-H DELAY (SECONDS) Cerebral Cortex July/Aug 1994, V 4 N 4 447

6 Tibia 2 MANOVA to parietal coding 15 C Perfonnance Reaction time F Monkey Detay Condnton (temperature) Delay x condition <0.05 <O0Ol B <0001 <001 >0.05 the three tasks, without modifying reaction time on any of them. Bilateral parietal cooling to 15 C induced no significant change in H-H performance or choice reaction time (Table 2, Fig. 4). A performance difference at 60 sec delay (see Fig. 4) was not significant. The delay x condition interactions were not significant either. Discussion It is reasonable to infer that the observed effects of prefrontal cooling were caused by the reversible functional depression of the cortex underlying the cooling probes (Moseley et al., 1972; Fuster and Bauer, 1974; Brooks, 1983). Therefore, as far as we know this is the first demonstration of deficit in haptic delay-task performance from lesion or dysfunction of dorsolateral prefrontal cortex. The deficit is unlikely to have G 100 A CONTROL. NORMAL TEMPERATURE PARIETAL COOL 1 5 C 90 - resulted from cooling-induced abnormalities of the sensory or motor mechanisms of haptic behavior. Prefrontal cooling did not elicit any somatosensory deficit that we could observe out of the test situation, nor did it prolong reaction time (on the contrary, it shortened it in one task). Moreover, we know of no evidence of sensory or motor deficit from lesions of the prefrontal area cooled. Our findings show that the dorsolateral prefrontal cortex is in some manner involved in performance of tasks that depend on the temporary retention of tactile information. They also confirm previous evidence from reversible cryogenic lesion implicating this cortex in visual delay tasks with nonspatially defined memoranda (Bauer and Fuster, 1976; Quintana and Fuster, 1993). In any case, the present results are compatible with the functional specialization of discrete areas of dorsolateral prefrontal cortex. Our cooling probes 0 u H LU U Q. 80 " H-H DELAY (SECONDS) Figure 4. Bilamaf parietal cooling 15 C) on H-H performance (averages from three annuals). 448 Prefrontal Cortex and Haptics Shindy el al

7 overlaid parts of the cortex of both the sulcus principalis and the inferior prefrontal convexity, which ablation studies have implicated in, respectively, spatial (Goldman and Rosvold, 1970) and nonspatial (Mishkin and Manning, 1978) visual memory. We note, however, that the delay x cooling interaction was not statistically significant. This outcome, in the absence of a simultaneous matching condition, which was impractical in our paradigms, prevents us from concluding that the prefrontal cortex plays a critical role in working memory of haptic information. This remains nonetheless a plausible interpretation of the data from the two tasks with tactile memoranda. Thus, our specific hypothesis (see introductory remarks) remains unproven but viable. Two of the tasks utilized (V-H and H-V) required a cross-modal transfer of information. Therefore, by operational definition and judging from the effects of cooling, the dorsolateral prefrontal cortex seems involved in behavior that depends on association of visual and haptic information. This does not mean, however, that such associations are made or held in this cortex a claim no more questionable than for any other neural structure with well-substantiated polymodal convergence like the prefrontal cortex. Rather, what it probably means is that the prefrontal cortex is involved in those tasks insofar as the crossmodal transfer is also cross-temporal. The absence of H-H deficit from parietal cooling can be explained on two grounds. First, the placement of our probes was such as to spare primary somatosensory cortex from hypothermia. In our experience, neurons in this cortex, particularly its hand region, are more implicated in haptic tasks than are the neurons of the posterior parietal cortex covered by our probes (Koch and Fuster, 1989; Zhou and Fuster, 1992). Second, posterior parietal cortex has been shown to be critical for spatial (i.e., place-defined) memory, but not for nonspatial memory (Quintana and Fuster, 1993). The present results are in harmony with this conclusion, in that parietal cooling induced no apparent deficit in memorization of a tactile cue The results of prefrontal cooling on V-H performance complement single-unit data from prefrontal cortex (Fuster, 1973; Fuster et al., 1982; Funahashi et al., 1989; Quintana and Fuster, 1992; Wilson et al., 1993) and inferotemporal cortex (Fuster and Jervey, 1982; Miyashita and Chang, 1988) during performance of visual delay tasks. The unit data show that cells in those two cortical regions have visual memory properties. Anatomical connections between the two are probably essential for visual cognitive functions (Pandya and Yeterian, 1985; Goldman-Rakic, 1988). Regarding visual short-term memory, their role is supported by evidence of functional interactions between inferotemporal and prefrontal cortex in a visual delay task (Fuster et al., 1985). The inferotemporal contribution, however, appears largely restricted to what has been termed "object vision": whereas inferotemporal cooling markedly affects a nonspatial visual memory task, it has little effect on delayed response, a spatial memory task (Fuster et al., 1981) The V-H deficit from prefrontal cooling may be attributed to depression of inferior-convexity cortex, the frontal area most tightly linked to inferotemporal cortex. In conclusion, both cooling and unit studies support the concept of a role of dorsolateral prefrontal cortex in active memory. This role is probably based on functional interactions between prefrontal cortex and cortices of postcentral regions. The role of these other cortices appears conditional on the particular nature of the memorandum, whereas that of the prefrontal cortex is predicated on the need to activate and retain a memorandum, whatever its nature, for prospective behavior. The active memory of a recent sensory item and of the consequent motor act is one of the cognitive functions by which the prefrontal cortex mediates the cross-temporal transfer of information for integrating behavior in the time domain (Fuster, 1985, 1989). The present results are consistent with this notion as it pertains to haptic behavior. Notes This research was supported by grants from the National Science Foundation (BNS ) and the Office of Naval Research (N14-89-J-1805), and an NIMH Research Scientist Award toj.m.f (MH-25082). We thank Mr. William Bergerson for excellent technical assistance. We also thank Javier Quintana and Yongdi Zhou for critical reading of the manuscript Correspondence should be addressed to J. M. Fuster, UCLA Neuropsychiatric Institute, 760 Wesrwood Plaza, Los Angeles, CA References Bauer R, Fuster J (1976) Delayed-matching and delayedresponse deficit from cooling dorsolateral prefrontal cortex in monkeys J Comp Physiol Psychol 90: Brooks VB (1983) Study of brain function by local, reversible cooling Rev Physiol Biochem Pharmacol 95: DiMattia BV, Posley KA, Fuster JM (1990) Crossmodal shortterm memory of haptic and visual information. Neuropsychologia 28: Funahashi S, Bruce CJ, Goldman-Rakic PS (1989) Mnemonic coding of visual space in the monkey's dorsolateral prefrontal cortex J Neurophysiol M9. Fuster JM (1973) Unit activity in prefrontal cortex during delayed-response performance- neuronal correlates of transient memory. J Neurophysiol 36: FusterJM (1985) The prefrontal cortex, mediator of crosstemporal contingencies Hum Neurobiol 4: FusterJM (1989) The prefrontal cortex, anatomy, physiology, and neuropsychology of the frontal lobe. New York- Raven. FusterJM, Bauer RH (1974) Visual short-term memory deficit from hypothermia of frontal cortex. Brain Res Fuster JM, Jervey JP (1982) Neuronal firing in the inferotemporal cortex of the monkey in a visual memory task. J Neurosci 2: Fuster JM, Bauer RH, Jervey JP (1981) Effects of cooling inferotemporal cortex on performance of visual memory tasks. Exp Neurol 71: Fuster JM, Bauer RH, Jervey JP (1982) Cellular discharge in the dorsolateral prefrontal cortex of the monkey in cognitive tasks Exp Neurol 77: FusterJM, Bauer RH, Jervey JP (1985) Functional interactions between inferotemporal and prefrontal cortex in a cognitive task. Brain Res Cerebral Cortex July/Aug 1994, V 4 N 4 449

8 FusterJM, DiMattiaBV, PosleyKA, Shindy WW (1990) Deficit in unimodal (tactile) and crossmodal delayed matching from cooling prefrontal cortex. Soc Neurosci Abstr Goldman PS, Rosvold HE (1970) Localization of function within the dorsolateral prefrontal cortex of the rhesus monkey. Exp Neurol 27: Goldman-Rakic PS (1988) Topography of cognition- parallel distributed networks in primate association cortex Annu Rev Neurosci 11: Koch KW, FusterJM (1989) Unit activity in monkey parietal cortex related to haptic perception and temporary memory. Exp Brain Res 76: Lewinsohn PM, Zieler RE, Libet J, Eyeberg S, Nielson G (1972) Short-term memory: a comparison between frontal and nonfrontal right- and left-hemisphere brain damaged patients. J Comp Physiol Psychol Mishkin M, Manning FJ (1978) Nonspatial memory after selective prefrontal lesions in monkeys. Brain Res Miyashita Y, Chang HS (1988) Neuronal correlate of pictorial short-term memory in the primate temporal cortex. Nature Moseleyjl, Ojemann GA, Ward AA (1972) Unit activity during focal cortical hypothermia in the normal cortex. Exp Neurol 37: Pandya DN, Yetenan EH (1985) Architecture and connections of cortical association areas. In: Association and auditory cortices (Peters A, Jones EG,eds),pp 3-61 New York: Plenum. Quintana J, FusterJM (1992) Mnemonic and predictive functions of cortical neurons in a memory task. Neuro- Report Quintana J, FusterJM (1993) Spatial and temporal factors in the role of prefrontal and parietal cortex in visuomotor integration. Cereb Cortex Quintana J, Yajeya J, Fuster JM (1988) Prefrontal representation of stimulus attributes during delay tasks. I Unit activity in cross-temporal integration of sensory and sensory-motor information. Brain Res 474: Wilson FAW, Scalaidhe SPO, Goldman-Rakic PS (1993) Dissociation of object and spatial processing domains in primate prefrontal cortex. Science Zhou Y, Fuster JM (1992) Unit discharge in monkey's parietal cortex during perception and mnemonic retention of tactile features. Soc Neurosci Abstr 18: Prefrontal Conex and Haptics Shindy et al.