Brain Research Bulletin, Vol. 2, pp. 31-39, 1977. Copyright 0 ANKHO All rights of reproduction in any form reserved. Printed in the U.S.A.
Behavioral Correlates of Denervation and Reinnervation of the Hippocampal Formation of the Rat: Recovery of Alternation Performance Following Unilateral Entorhinal Cortex Lesions JACQUELYN Departments
LOESCHE AND OSWALD STEWARD’
Surgery and Physiology, Charlottesville,
University of Virginia School
(Received 6 January
LOESCHE, J. AND 0. STEWARD. Behavioral correlates of denervation and reinnervation of the hippocampal formation of the rat: recovery of alternation performance following unilateral entorhinal cortex lesions. BRAIN RES. BULL. 2(l) 31-39, 1977. - Following unilateral lesions of the entorhinal cortex (EC.) of the rat, cells in the dentate gyrus which have been deprived of their normal ipsilateral input are reinnervated in part by axons from the contralateral E.C. The proliferation of this crossed projection occurs largely between 8 and 12 days postlesion. The present experiments analyze changes in alternation behavior which occur during this period of afferent proliferation. Rats were trained to alternate responses (L-R) in a T-maze for food reward. Bilateral E.C. lesions resulted in a persistent deficit in alternation performance which did not recover after over 50 days of postoperative testing. Unilateral E.C. lesions, however, resulted in a transient deficit in alternation which recovered over time to preoperative levels. For example, animals permitted a lo-day recovery before the initiation of postlesion testing exhibited no more of a performance deficit than following a lOday no-training period alone. However, animals permitted only a 3-day postoperative recovery were impaired in the alternation task until 1 O-l 2 days postlesion, despite daily training. Thus, recovery of performance following unilateral lesions was dependent on postlesion time rather than the amount of testing/retraining. Since bilateral lesions resulted in a persistent performance deficit while unilateral lesions resulted in a deficit with recovery, we hypothesize that behavioral recovery might be related to the reinnervation of the dentate gyrus by the contralateral E.C. To test this hypothesis, secondary lesions were placed in operated-recovered animals. Secondary lesions of the surviving E.C. resulted in a deficit in alternation performance similar to that following one stage bilateral lesions. In addition, secondary lesions of the dorsal psalterium (the fiber tract which carries the crossed E.C.-dentate projections) also disrupted performance in operated-recovered animals. Primary lesions of the dorsal psalterium alone had only slight and transient effects on alternation performance, however. Thus, the time course of the recovery, the results following bilateral lesions, and the results of secondary lesions are all consistent with the hypothesis that recovery of alternation performance following unilateral E.C. lesions may depend upon the reinnervation of the dentate gyrus by the contralateral E.C. Postlesion
FOLLOWING unilateral lesions of the entorhinal cortex of the adult rat (a major extrinsic afferent to the ipsilateral dentate gyrus of the hippocampal formation) several surviving afferents proliferate to form synaptic connections which reinnervate the dentate granule cells [ 12, 13, 22, 241. These include the acetylcholinesterase containing septo-hippocampal fibers [ 121, the dentate commissural system which originates in the contralateral hippocampal formation [ 131, and the normally minute projection from the contralateral entorhinal cortex which proliferates by approximately 500% within the denervated zones estab-
lishing extensive new connections with the dentate granule cells [21-25) The reinnervation of the dentate gyrus by afferents from the contralateral entorhinal area provides a rather unusual example of postlesion plasticity, since the reinnervation is accomplished by afferents which are in many ways quite comparable to the normal projections which have been destroyed by the lesion. For example, the reinnervating fibers originate from the contralateral homologue of the damaged structure [ 2 11, terminate on the granule cells in a manner which is highly reminiscent of the normal mode of
’ Reprint requests to be addressed to 0. Steward. 31
32 termination [ 251 and activate the granule cells in a manner which compares well with the normal mode of activation by ipsilateral entorhinal afferents [ 22, 261. In addition, electrophysiological [ 261 and anatomical analyses [ 25 I indicate that the reinnervation occurs largely between 8 and 12 days postlesion. The present studies were designed to investigate the behavioral significance of the postlesion reinnervation of the dentate granule cells by the contralatera1 entorhinal cortex. If the reinnervation of the dentate gyrus by the contralateral entorhinal cortex has behavioral consequences, then one would expect: (1) that the behavioral change which resulted from the postlesion growth should have a time course similar to the reinnervation process; (2) that bilateral lesions of the entorhinal cortex should prevent the behavioral change; and, (3) that secondary lesions of the contralateral entorhinal area or the decussating fiber systems should result in the disappearance of whatever behavioral change resulted from the growth of these contralateral entorhinal fibers. Smith, Steward, Cotman and Lynch [ 191 have previously analyzed spontaneous alternation in a Y-maze following both unilateral and bilateral entorhinal lesions. The results were difficult to interpret, however, since spontaneous alternation behavior was not an optimal behavioral measure for assessing postlesion deficits. First, very few trials ’ can be run on a given day, since the behavior depends on the spontaneous exploratory activity of the rat. Since our goal was to analyze postlesion performance on a day-to-day basis, this limitation was quite severe. Second, since performance is unrewarded, slight changes in motivation can easily disrupt the performance. Therefore, in the present experimental series, our goal was to develop a behavioral task which would require the same sorts of performance as spontaneous alternation, but which would also provide consistent measures of daily performance. The tasks to be described utilize food reinforced alternation in a T-maze to analyze the behavioral consequences of entorhinal lesions, and subsequent afferent reorganization. METHOD
Animals Experimental animals were 45 male Sprague-Dawley derived rats obtained from Flow Labs. Prior to training, the animals were food-deprived to 80% of their initial weight (initial weight ranged from 250-350 g), and were maintained at the 80% level for the duration of the experiments. Animals were housed in groups of three or four and were provided with water ad lib. Apparatus A plywood T-maze with alleys measuring 36x 6x 6 in. was used. The start box (8 in. in length) was separated from the approach alley by a sliding wooden door. Plastic food cups were placed at either end of the goal arms. Procedure,
placed in the arm opposite the one chosen on the preceding trial. This procedure was repeated such that the correct arm was always the one which had not been chosen on the preceding trial. In well-trained animals, the inter-trial interval averaged approximately lo-15 sec. Animals were trained for a total of 1 1 massed trials/day (10 L-R alternations) until criterion performance had been attained (at least 8 alternations (80%) for three consecutive days). After criterion performance had been attained, training was discontinued for either 3 or 10 days, and after this no-training period, the animals were again brought to criterion performance. On the day following this period of retention testing, the operations were performed. A total of 27 animals received unilateral lesions of the entorhinal cortex (E.C.), 5 animals received bilateral lesions, and 4 animals received sham operations. The unilaterally operated animals were divided into four groups, spaced over a period of 10 months. Surgery Anaesthesia was induced by intraperitoneal injections of sodium pentobarbital (Nembutal) at a dosage of 50 mg/kg. The entorhinal area was destroyed by passing a 1 ma current (45 set in duration) at a total of 9 stereotaxically defined sites. Coordinates for the stereotaxic placements were 1.5 mm anterior to the transverse sinus, 3, 4, and approximately 5 mm lateral to the midsaggital sinus (the most lateral placement was as far lateral as the skull permitted); and 2, 4, and 6 mm deep to the cortical surface. For the entire series, the electrode was positioned at a 10” angle away from the midline. After the completion of the lesion series, the skull was filled with gelfoam, the scalp sutured, and the animals were given an intramuscular injection of penicillin (Wycillin, Wyeth Labs). For the sham operations, the same procedure was followed, except that the brain was not penetrated by the electrode. Postlesion
Following surgery, the operated animals were tested in one of three ways: (1) five animals with bilateral lesions were permitted a 3-day postoperative recovery, and were subsequently tested every fourth day until 40 days postlesion and were then tested every day from postlesion Day 40 through Day 58; (2) 8 animals with unilateral lesions were permitted a IO-day postoperative recovery and were subsequently tested every day beginning on the 10th day, until criterion was attained; (3) 22 animals with unilateral lesions, and 4 animals with sham operations were permitted a 3-day postoperative recovery, and were tested every day beginning on postlesion day 3 until criterion performance was attained. Postlesion testing was continued either until the performance of the animals returned to the criterion level (3 consecutive days at 80% alternation or better) or until 58 days postlesion (in the case of the animals with bilateral lesions).
Animals were placed in the start box, and permitted to traverse the approach alley to select either the left or right goal arm. For the initial trial on any given day, both goal arms contained the food reinforcement (three 45 mg, Noyes pellets). After the food was consumed, the animals were returned to the start box, and the reinforcement was
Operated-recovered animals with unilateral E.C. lesions were subjected to secondary lesions 15-26 days following the primary operations. Secondary lesions were of two types: (1) ablation of the surviving entorhinal area, or (2) transection of the dorsal psalterium, the fiber tract which
carries the decussating projections of the entorhinal area. Secondary ablation of the surviving entorhinal area was accomplished in the manner described above for the primary lesions, Transections of the dorsal psalterium were accomplished by lowering a blade as near to the midline as possible to a depth of 4-5 mm. The blade was then moved in the anterior-posterior plane from bregma to a point approximately 1 mm rostra1 to lambda. This operative procedure, henceforth called a dorsal psalterium transection, was placed as a secondary lesion in 8 unilaterally operated-recovered animals, and in 9 intact animals trained in the alternation task, as a control for the behavioral effects of the transection alone. Histology Upon completion of behavioral testing, the animals were sacrificed with an overdose of sodium pentobarbital, and perfused transcardially with a 10% formalin-saline solution. Brains were removed, embedded in egg yolk according to the method of Ebbesson  and sectioned on a freezing microtome at 33pm. Alternate sections were stained using a modification of the Fink-Heimer [ 71 method. Lesions were evaluated without knowledge of the individual animals’ performance.
5 60. ._ P zi 2 ap 40-
acquisition Retention X--X 10 Day Retention 0-O 3Day
$ RI 5 a 20-
Examples of the initial acquisition of the alternation performance, and the performance following both 3-day and 1O-day preoperative no-testing periods are illustrated in Fig. 1, for one group of animals. Animals initially alternated at a rate greater than chance as might be expected from the tendency of rats to spontaneously alternate in a nonreinforced situation [ 2, 3, 4, 181. With subsequent training, performance rapidly improved, reaching criterion in an average of 7.3 days for all 45 animals. As illustrated in Fig. 1, after either 3 or 10 days without training, performance remained near criterion levels. There was no significant difference in performance following the 3-day preoperative no-training interval (t = 0.25, comparing the last day of initial acquisition with the first day of retention testing). Following a IO-day no-training interval, there was a slight deficit between the last day of acquisition and the first day of retention testing (t = 2.08, p 0.05). As illustrated by Fig. 2, bilateral E.C. lesions resulted in an immediate and profound deficit in alternation performance which did not recover over the postoperative testing period. On the first postoperative testing day (3 days postlesion) the average rate of alternation fell to 26%, a value somewhat below chance. However, by the second postlesion testing day (7 days postlesion) animals alternated at 44%, and their performance did not improve further over the 58 day postlesion testing period (alternation on the last 5 testing days averaged 54%). While bilateral E.C. lesions resulted in persistent deficits in alternation, the performance of unilaterally operated animals depended on the postlesion interval. For example, animals which were permitted a IO-day postoperative recovery before the initiation of testing/retraining exhibited little if any deficit. As illustrated in Fig. 3, alternation performance in the retention test following a lo-day no-training interval is comparable with the performance
I 10 Training
FIG. 1. Acquisition of the alternation task, and preoperative retention testing. An example of acquisition of the alternation task is illustrated for the group which subsequently received bilateral lesions. Three day retention illustrates the performance of animals after a 3 day no-training period, while 10 day retention illustrates the performance of animals after a 10 day no-training period. For the statistical evaluation of retention, each group was compared with its own pre-operative performance on the final day of initial acquisition. following a IO-day postlesion recovery period. Indeed, the graphs of alternation performance during the preoperative retention test and the postrecovery period are superimposable. Thus, unilaterally operated animals which are permitted a IO-day recovery before the initiation of postoperative testing exhibit no more deficit than would have resulted from a 1 O-day no-training period alone. Animals with unilateral E.C. lesions did, however, exhibit a significant deficit if testing began 3 days postlesion (see Fig. 4). The performance of these animals was significantly lower than their own preoperative control from postoperative Days 3-9. Using as a basis for comparison the last day of preoperative testing, t values ranged from 1.94 (~~0.05) to 5.11 (p