until its performance in 10 consecutive trials was equal to its performance of the last 10 trials ofPhase 1. When this criterion was reached, Sadvanced to Phase 4,.
Hippocampal lesions and delayed alternation in the rat RONALD J. RACINE AND DANIEL P. KIMBLE ! UNIVER SI TY OF OREGON
Abstract Ra ts we re trained preope ratively on a de layed a l te rnation task in which th e delay period wa s in creased un til S accumulated four e r rors at a gi ven del a y inte r val. When th e temporal limit fo r a n S wa s thus determ ined , S was given eithe r a bilateral hippocampal les ion or a bilateral cor tica l lesion . Po stoperatively, the hippocampectomized Ss were totally un able to perform th e delayed a lte rnati on task , ev en a t extremely brief delays . Th e co r ti call y damaged Ss showed no deficit, a ll improving on their preoperative temporal limit. Th e s e r esults confirmed a nd extended previous results with hippocampectomized monkeys. Problem Th e hypothesis that th e hippocampus i s involved in "re cen t memory" (Penfield & Milner, 195 8) has not r e ceived consis te nt suppor t fr om laboratory studie s with a nimals (Kimble, 1963; Mahut & Cor deau , 196 3; Madsen & Kimble, 1965) . On the other hand, laboratory s tu dies of monkeys with med ial te mpo ral lob e-hippocampal r emovals (Orbach, Mil ne r & Rasmu s sen, 196 0; Mahut & Cordeau, 1963; Rosvold & Szwarcbart , 196 4; Pribram, Wil son & Con ners, 1962) hav e pointed to dela yed a lternati on de ficits as particularly cons pic uous po stoperative r es ul t s . Th e present stu dy investiga ted th e effects of hippocampal le sions on the temporal limits of r etention of a del ayed a lte rnat ion task in r ats, using P etrinovich & Bolles method (1961). If th e hippocampus is involved in r e cent m emory, th e lesion effec t should be come in creasingly appar e nt as the delay period is ext ended.
r eg ardle s s of the side chosen. Th e second a nd th ird choi ces were rewarded onl y if S a lter na ted its last r e spons e. Thi s tra ining wa s conti nue d for 28 da ys with a n intertrial interval of 10 min . Phase 2 . Th e 10 Ss whi ch a lte r nated most cons i s tentl y conti nued into Pha se 2. Th e sam e procedure was followed except that the intertrial interval was extended, fir st to 20 min., then to 40 min ., then to 1 hr., a nd then in 1/2 hr. inte r vals . If S a lte rnate d without e r r o r for two consecuti ve days, it was advanced to th e next longer delay interval. When an animal accumulated four errors a t a n interval, it was cons ide r ed to have failed and surgery wa s then performed on that S. The surgical procedure was similar to that r eported previously (Kimble, 1963). Phase 3. Ph a se 3 was initiated 10 days following surgery. The procedure was identical to that of Phase 1. Each S wa s tested with a 10 min. intertrial interval until its performance in 10 consecuti ve trials was equ al to its performan ce of the la st 10 trials of Phase 1. When this c r ite r ion was reach ed, S adv anc ed to Phase 4 , which wa s iden tical to Phas e 2. An e xamination of the secti oned brains revealed exte ns ive (50-90%) a nd bilate r al hippo campal lesions in a ll the hippocampectomized Ss , Th e brains of the co r tically lesioned Ss r evealed bilateral damage to neo cortex, and no hippocampal damage except for a slight surfa ce invasion in one S (Fig. 1).
Jlc~thod
Th e Ss were 30 experimen tally naive, 60 day old, male Sprague-Dawley r ats . The appa ratus consisted of a T-maze with a n 48 x 4-1/2 x 6 in stem, 24 x 4-1/2 x 6 in a r ms , and two detach able 12 x 6 x 6 in goal boxes . Ss were placed on a 23 hr . water depriv ation sche dule a nd gentled for two days . Th ey were then given four trials/day for two days in a s traight alley alternate ly a ttache d to each goal box. S was all owed 1 mi n. of drinking on each trial. Ss were then given two free a nd two forced trials each day for three more da ys to insure that they would hav e e ntere d each goal a r m 50% of the time prior to the beginning of alte rna ti on t rainin g . Maz e traini ng. Phase 1. Ss were run three trials/day using a non correction method. Th ey were a ll owed 20 sec . drinking following ea ch co r rect r espons e and were detained for 20 sec . in th e goal bo x afte r each in correct r e sponse. Th e first choice of each day was rewarded P s y ch on . Sci. , 1965 , Vol. 3
Cd)
(b)
Fig. 1. Reconstruction of hippo campal (a) a nd cortical (b) les io ns . Black a rea indi ca tes minimal les ion ex te nt , lin ed a rea indi cat es ma ximal les ion ex te nt.
285
TABL E I Delayed Alt ern ation Tempora l Limits and Total % Correc t
Rat No. Cortical Le sion Ss 3 5 7 8 9
Temporal L imit in min. and (total % correct ) ~~fl~r~ a t_iv_e,--
Postoperat ive ..:..::..:..c..c.,--_-,----_
90 (73.5) 120 (76.3) 120 (75.0) 180 (91.6) 210 (82. 1)
240 (82. 1)
40 (68.7) 60 (77 . 2) 120 (83.3) 120 (76.4) 330 (88.7)**
fail fail foil foil foil
150 (82.3) 180 (85.4)
Hippocampal Lesion
5s 1
2 4
6
10
(30.2) (30.0) ( 6.2) (25.0) (46.2)
died during surgery did not fail , test ing dis continued at 5).2 hr. interval
Results The 10 Ss with the highest alte rnati on sc ores during Phase 1 were chos en to continue training. Table 1 shows the pre- and postoperative performance of ea ch animal. Postoperatively, all of the cor ticall y lesioned Ss immediately reached their pr eoperative performanc e level of Phase 2 within the minimum number of trial s ne cessary . However, none of the hippo campectomized Ss eve r appr oached their preoperative level, during 20 days (60 tr ials ) at delay inte r val s of 10 min. The delay inter va l was then lowered to appr oximately 5- 10 sec . (exc luding drinking time, Ss being replaced in the star t box immediately after each trial) in order to determine if a lternati on behavior could be obt ained under these conditi ons . lt could not. Ss were run an addition al 20 days (60 trials) with this short delay. The level of alter na tion behavior r anged from 5-37% a t the 10 min. delay, and 0-55% a t the 10 sec . delay among th e hippoca mpe cto mized Ss , The number of days oflOO%alternation behavior , i.e., two consecuti ve alter nati ons, did not increase for any of these Ss , r emaining a t les s than 5% for 4/ 5 Ss eve n at th e 10 sec . delay. Th is result indicates that even at < 10 s ec. del ay s, the hippocampectomized Ss were incap able of performing the a lte rnati on task. For example, S No. 10, who had never failed preoperatively, failed postoperati vely , even at the 10 sec . interval. Ea ch cor ticall y le sioned S extende d its preop erative temporal limit follow ing su r gery. Discussion The complete failure of the hippo campectomized Ss to perform the delayed a lternati on task cont rasts with the total la ck of effec t seen in the cor tically les ion ed Ss, The present s tudy thus confirms for the r at th e results of Mahut &Cordeau (1963), Orbach, Milner &Rasmus sen
286
(1960), Pribr am et a l (1962) and Ro svold & Szwarcba r t (1964) with monkey s, that medi al temporallobe-hippoca mpa l ablations impa ir delayed alternation be havior . The exact reason for thi s r esult, howev er, i s not co mpl etely clear . While "recent memory lo s s" ca nnot be di smissed as a po s sible interpretation , the hippoca m pectomized animals showed no consiste nt improvement at eve n very short delay pe r iods , a s might be pr edicted from thi s hypothesis . Anothe r possibl e inter pr etati on is th at of "response perseveration" (Ellen &Wilson,1963; Mahut &Cordeau , 1963). Eviden ce for thi s effec t was s een at th e 10 min . delay; three of the five temporal lob e-hippocampal Ss fixating on a position habit . Howev er, onl y one S remained fix ated at th e shor te r delay period, while the other Ss di splayed a random r e sponse pattern . Kimble & Pribram (1963) hav e shown that hippocampal lesions in the monkey interfere with " sequential res ponding , " and that res ponse per-severation may onl y be one manife station of suc h a defi cit. Since the delayed a lternation task cons titut es a simple behavio r se quence , our r esults a re compatib le with th is interpretation , although conclusions r egarding the functional role of the hippocampus mu st r emain tentative . References E llen , P, & Wilson , A. S P er s everati on in t he rat foll owin g hippoc a mpa l le s ions E xp . Nc uro l ., 1963,8 ,310-317 . Kimbl e . D. P . Th e e ffec ts of bilat era l hipp oca mpa l l e s ions in rats. J . com po phusio l, P ssich ol ., 1963 . 56 , 273 -283 . Kimble , D P . , & Pri bra m, K. II . Hipp oc a mpe ct omy and behav ior s equ e nces . Sci en c e. 1963 , 139 , 824-825. Ma dsen. M. C. , & Kimbl e, D. P . The maze behavior of hi ppc campe ct omized rat s unde r massed a nd di s t ributed tri al s . Psu c hon, Sci. . 1965 , 3 . 'I ahu t, lI el en , & Cordea u , ,J. P . Spa tial re ver s al defi c it in monkeys wit h amygdalohippoca mpal ablati on s . E .l'P. Neu rol . • 1963 . 7, 426-434 . Orbach , .J ., Miln er , Brend a . & Rasmu ssen , T . L e arni ng and retentio n in monke ys aft er amygdala -hipp ocampus re s ec ti on . A rch . Ne uro l. , 1960. 3 . 230-251. P enfield . W.. & Miln er , Brend a . Me mory defi cit produ c ed by bilat eral l e si ons in th e hipp ocampa l z on e . Arch . Ne ura l. P suchia t . . 1958 , 79 . 475 -49 7. P e trinovi ch , L . , & Boll e s , R. Delayed alternati on: ev ide nce for sy mbo li c pro c e s s e s in t he rat. J . c omp. phy si ol. P such ol . . 1957 , 50, 363-365. Pribram, K. II . , Wilson , W. A . , & Conne rs , .I ane , E ffe cts of le s ions of th e medi al for ebrain on alternation behavi or of rhesu s monkeys . Ex» . Neurol . , 1962 , 6, 36-47. Rosvold . II . E., & Sz warc ba rt, Ma ri a K. Neu rai st ruc t ures inv ol ved in de la yed-re sp on se performan c e . In J . M. warr en, an d K. Ake rt, (E ds.), Th e fro nt al . gr anu l ar corte x an d behav io r . Ne w York: McGraw-lIilI , 1964 . Pp . 1- 15.
Note
I. Th is resea rc h was s uppo rte d by USP IIS Gra nt No. 08545-02 , D. P. Kimbl e, prin cipal i nve stiga tor , and by an und ergraduat e NSF grant to R. J. Ra ci ne .
Psycho n . Sci. , 1965 , Vol. 3
