Disruption of acquisition and performance of operant response-duration differentiation by unilateral nigrostriatal lesions

Behavioural Brain Research 114 (2000) 65–77

Research report

Disruption of acquisition and performance of operantresponse-duration differentiation by unilateral nigrostriatal lesions

T.J. Hudzik *, A. Howell, M. Georger, A.J. CrossDepartment of Neuroscience, Astrazeneca R & D Wilmington, 1800 Concord Pike, P.O. Box 5437, Wilmington, DE 19850-5437, USA

Received 22 December 1999; received in revised form 14 March 2000; accepted 14 March 2000

Abstract

Response duration differentiation (RDD), an operant schedule requiring fine motor timing and control, was assessed as apossible baseline for study of the long-term consequences of nigrostriatal lesions and as a possible baseline to test the therapeuticefficacy of candidate palliative, neuroprotective and neurorestorative drugs. Rats were subjected to unilateral 6-hydroxydopamine(6-OHDA) lesions of striatum, medial forebrain bundle (mfb), or were sham lesioned, and their ability to acquire the operant taskwas studied in a single overnight session. In a second set of studies, rats that had been well trained in the RDD task were shamlesioned or were given unilateral 6-OHDA lesions of the mfb, and behavior under this baseline was studied for more than 30weeks. Lesions of both striatum and of mfb resulted in impaired acquisition of RDD responding, with the relatively greater effectby the mfb lesion. In rats previously trained under the RDD schedule, mfb lesions produced marked disruptions in RDDperformance, which did not fully recover. L-DOPA administration decreased the variability of the response durations, primarilyby decreasing the proportion of short-duration lever presses. © 2000 Elsevier Science B.V. All rights reserved.

Keywords: Nigrostriatal lesion; Rats; Operant behavior; Long-term deficits; Timing

www.elsevier.com/locate/bbr

1. Introduction

Parkinson disease, a result of progressive deteriora-tion of the nigrostriatal dopaminergic tract, is charac-terized by cognitive deficits in addition to thewell-defined motor deficits. One such cognitive deficitdescribed is in employment of appropriate strategiesduring new learning [4,6,23,25]. Another deficit de-scribed is in temporal perception, thought to be due todisruption of an hypothesized ‘internal clock’ localizedin basal ganglia and under the control of dopamine[3,20]. While motor function has received much atten-tion in experimental parkinsonism, relatively less atten-tion has been paid to disruption in cognitive function.

There have been a number of indices of damage thathave been studied in animals following unilateral de-struction of dopaminergic motor tracts with toxins suchas 6-hydroxydopamine (6-OHDA). One of the mostcommon, drug-induced rotation in rats [30] has proven

useful in mapping the circuitry and neurochemistry ofthe nigrostriatal tract, as well as provided a baseline forstudying the efficacy of palliative and neuroprotectiveantiparkinson drugs (e.g. [2]). However, because drugadministration is required to elicit the deficit, which canconceivably impact the magnitude of the deficit overtime, it was of interest to develop a behavioral proce-dure in rodents which would not require stimulantadministration, and perhaps have greater relevance tohuman clinical deficits. Additionally, in order to assessneurorestorative approaches to therapy, it was alsonecessary to demonstrate a relatively stable deficit.

Acquisition of operant responding is one of manydifferent behavioral baselines which may provide anindex of learning performance. The requirements of theprocedure include exploratory activity, manipulation ofa lever within the operant chamber, and formation ofthe contingency between lever pressing and food deliv-ery. This baseline has been previously shown to beextremely sensitive to the disruptive effects of a numberof pharmacologic agents [15]. Given the deficits in newlearning ability shown in Parkinson’s patients, and thefact that the task requires acquisition of a novel behav-

* Corresponding author. Tel.: +1-302-8863000; fax: +1-302-8864983.

E-mail address: [email protected] (T.J. Hudzik).

0166-4328/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 6 -4328 (00 )00216 -3

T.J. Hudzik et al. / Beha6ioural Brain Research 114 (2000) 65–7766

ior, it was considered possible that operant acquisitionwould provide a baseline that would be sensitive todamage in the nigrostriatal tract.

Response duration differentiation (RDD) schedules,in which subjects are trained to emit responses of a veryspecific duration in the absence of external cues, havebeen shown to be useful in characterizing broad classesof pharmacologic activity [11], identifying antidepres-sant activity [9,13], and in providing a baseline forstudying the consequences of administration of neuro-toxins, such as trimethyltin [14]. We have generallytrained very discreet lever press durations in the proce-dure (e.g. 1–1.3 s), and our hypothesis was that thesetypes of responses would be sensitive to manipulationsin the dopaminergic: system, for two reasons. The firstwas that given the well-known role of dopamine incontrol of motor function, disruption of this transmit-ter system would likely result in impairment in perfor-mance of the task. The second reason was because ofthe temporal perception component of the task. Differ-entiated motor responses, because they are trained andmaintained without any external cue, may require useof an ‘internal clock’ [3] suggested to be under controlby dopamine [20]. That there are temporal aspects toRDD responses is supported by pharmacologic studies.For example, drugs that are well known to alter tempo-ral perception, such as D9-tetrahydrocannabinol, or theneurotoxin trimethyltin, produce a discrete, leftwardshift in the relative frequency distribution of responsedurations (underestimation of duration) without alter-ing the shapes of those distributions [12]. This suggeststhat there may be temporal aspects of the response,which can be dissected from the motor aspects.

2. Methods

2.1. Subjects

All subjects were male, Long–Evans rats weighingbetween 250–400 g. Animals were individually housedunder a 12-h light/dark cycle.

2.2. Apparatus

Standard 2-lever operant chambers (Gerbrands), withdimensions 29 cm length×24 width ×21 cm in height,were used. The chambers were fitted with stimuluslamps over each lever in the chamber, and a houselightmounted at the top of the chamber. Between the levers,4 cm above the grid floor was a food cup, to which wasdelivered 45-mg food pellets (Noyes) at scheduledtimes. The levers in the chambers required a force of30–35 g to register a response. A microprocessor, lo-cated in an adjacent room, recorded and controlledexperimental events in the chambers.

2.3. Stereotaxic surgery

All animals were pretreated with desipramine (20mg/kg, i.p.) prior to anesthesia. Under sodium pento-barbital (64 mg/kg i.p., RDD acquisition) or isoflurane(RDD performance) anesthesia, male Long–Evans ratsweighing 250–400 g where stereotaxically lesioned with6-OHDA (2 mg/ml in 0.2% ascorbic acid in saline). Atotal volume of 4 ml of 6-OHDA was injected througha cannula over the course of 15 min. The coordinates(from bregma) for striatal lesions were anterior/poste-rior −1.4 mm, lateral/medial −4 mm and dorsal/ven-tral −7.1 mm with the incisor set at +5 mm. Thecoordinates for medial forebrain bundle (mfb) lesionswere a/p −3.5, lat −1.6, d/v −8.8. All animals werelesioned in the right hemisphere. Surgical controls(shams) were submitted to the identical procedure aslesioned rats, but only 2% ascorbic acid in saline wasadministered through the cannula.

2.4. RDD acquisition

Eight rats following striatal lesions, ten followingmedial forebrain lesions, and ten unlesioned controlswere studied. Two weeks after surgery, and following a24-h fast, rats were placed into standard 2-lever operantchambers for 15 h, in order to assess ability to learn toemit a response on the lever, then adapt the response tochanging schedule requirements. Initially, each leverpress of a duration of 0.03 s or longer resulted indelivery of a food pellet (each lever press was reinforcedin order to establish lever pressing behavior). Followingdelivery of each 15th food pellet, the minimum durationfor which the lever was required to be depressed forfood pellet delivery was increased in the followingsuccession: 0.1, 0.3, 0.5, 0.7, and 1 s thereafter for theremainder of the session. Total reinforcers delivered,total responses emitted, and the average duration oflever holds were recorded. Cumulative reinforcersearned during the course of the session were also calcu-lated. Data are reported only for those lesioned ratswhich rotated in response to amphetamine (\50 turnsin 90 min) on the day following the operant acquisitionsession, which was considered indicative of correct le-sion placement (see below). Data were analyzed by2-way ANOVA.

2.5. RDD performance

Rats were maintained at 85% of free feeding weight,and trained to respond under an RDD schedule inwhich responses of a duration within the range of 1–1.3sec. were reinforced by delivery of a food pellet (RDD1–1.3%% schedule). Once stable responding occurred, ani-mals were assigned to one of two surgical groups: (1)mfb lesion as described above (N=8); or (2) sham

T.J. Hudzik et al. / Beha6ioural Brain Research 114 (2000) 65–77 67

lesioned (N=6). Animals were allowed 4–5 weeks torecover from surgery (until able to maintain bodyweight) then returned to the operant procedure for 3–5days per week. Daily sessions were 30-min in duration.Animals’ behavior was studied under this baseline fromthe 5th week post-lesion until the 30 week post lesion.

The number of responses, the number of reinforcedresponses (accuracy) mean and variance of responsedurations were recorded. Where three or more meanswere compared, ANOVA were used, followed by Dun-nett’s post hoc analysis to compare vehicle points tothose following drug treatment. Additionally, relativefrequency distributions of response durations were con-structed by plotting the duration of responses (by suc-cessive 0.1 s intervals) against the % of total responsesemitted. Ninety-five percent confidence limits aboutcontrol (prelesion) distributions were drawn aboutthose following lesion in order to facilitate visual in-spection of changes in these distributions, and statisticalcomparisons to pre-lesion distributions were made byx2 for relative frequency distributions on each of threeseparate parts of the distributions: response durationsless than 1 s (d f=9) those within the range that werereinforced (d f=2) and those greater than the rein-forced range (d f=13).

Beginning 35 weeks following surgery, the effects ofL-DOPA administration upon the RDD baseline wereassessed in both lesioned rats and in sham controls.Carbidopa (30 mg/kg) was administered 30 min prior toL-DOPA (0.01–30 mg/kg), which was administered 15min prior to operant sessions. Both drugs were dis-solved in sterile saline and administered i.p. in a volumeof 1 cc/kg body weight. Drug tests were conducted onceper week in order to minimize their potential influenceupon the behavioral baseline.

2.6. Amphetamine rotation

In order to study the relationship between deficits inoperant behavior, lesion size, and rotation, RDD-trained rats were injected with 3 mg/kg s.c.d-amphetamine and placed into automated rotometers(Med Associates, St Albans, VT) for a 90-min period.Rotational testing was conducted at several time pointsfollowing lesion: 3 weeks, 3 months and 9 monthsfollowing lesion. RDD acquisition rats were tested foramphetamine-induced rotation a single time 2 daysfollowing operant acquisition training, with more than50 turns considered indicative of proper lesionplacement.

2.7. Histology

2.7.1. Tissue processingThe animals were euthanized utilizing sodium pento-

barbital and transcardially perfused using heparinized-

sodium nitrate saline solution followed by a 10%neutral buffered formalin. The brains were removedand post fixed in fresh 10% neutral buffered formalinand left for 24 h at 4°C. The brains were blocked usinga precision brain matrix from Harvard Apparatus. Pre-determined coordinates assured consistent optimal sec-tions through striatum. The blocked pieces of tissuewere processed in paraffin utilizing a 6 h embeddingscheme. The sections were cut at 4 mm utilizing aMicrom microtome (Zeiss, HM315) and Accu-Edge lowprofile disposable microtome blades (Sakura, 4689).The sections were floated onto untreated slides andstored at room temperature until staining.

2.7.2. Slide stainingThe slides to be stained for tyrosine hydroxylase

(TH) were heated to 56°C for 1 h. To ensure completeremoval of paraffin the slides were placed warm in onechange of Xylene for 5 min and then three changes ofPropar (Anatech Ltd., 511) at 5 min each. The slideswere then placed in 100% EtOH for two changes, 5 mineach, and then 70% EtOH for two changes, 5 min each.To improve the intensity of staining the slides weresubjected to heat induced antigen retrieval using Bio-genex Antigen Retrieval Citra (Biogenex, HK086-9K)at the manufacturers recommended protocol. The slideswere heated in the microwave oven at power level 9 for2 min 45 s and then power level 2 for 10 min, with closeobservation to ensure the level of fluid covered theslides. The slides were then cooled at room temperaturefor 20 min and then washed for 5 min in deionizedwater. The slides were then transferred to TBS for fourchanges, 5 min each. Endogenous peroxidase wasquenched by placing the slides in 0.3% H2O2 in TBS for15 min at room temperature. The slides were rinsed inTBS for three changes, 1 min each. To block secondarystaining and to increase membrane permeability theslides were incubated in 10% Normal Goat Serum(Chemicon, S26) in TBS with 0.5% Triton X (Sigma,T9484) for 20 min at room temperature. The slides wererinsed in TBS with 5% Triton for three changes, 1 mineach and then incubated with anti-TH 1:477 (Cherni-con, �318) in 1% NGS in TBS overnight at roomtemperature. The slides were placed in a humid cham-ber to ensure that the tissue remained moist. On thesecond day the slides were rinsed with TBS for threechanges, 1 min each, and then incubated with BiogenexSuper Sensitive Mouse I-ink (Biogenex, BK391-9n for 2h 30 min at room temperature. The slides were rinsed inTBS for three changes, 1 min each, and incubated withBiogenex Super Sensitive Label for Animal DetectionKit (Biogenex, BK399-9K) for 2 h at room tempera-ture. The slides were rinsed in TBS for three changes, 1min each and covered with liquid DAB (Biogenex,HK086-9K) for 10 min at room temperature. The slideswere rinsed for a final time in TBS for four changes, 1


min each, and dehydrated through ascending alcohols,Propar and coverslipped with Permount (Fisher, SP15-500).

2.7.3. Histologic analysisOptical intensity of the TH-stained sections was de-

termined by digitally imaging representative coronalsections from each animal at the level of mid striatum.(−0.1–0.6 from bregma). Images were processed withImage Pro + software (Media Cybernetics) by redefin-ing TH stain (appearing as brown on the slides) aswhite, and redefining other wavelengths in the sectionas black. The sum of the average optical intensity(mean white area, proportional to the density of THstaining) within each of 256 overlaid grids is reported.Dorsal and ventral striatum were imaged separately onboth the lesioned and unlesioned sides of each section.

3. Results

3.1. RDD acquisition

The effects of striatal and mfb lesions on the acquisi-tion of RDD responding are shown in Fig. 1. Therewas a main effect of lesion (ANOVA F(2, 62)=95,PB0.001). Dunnett’s post-hoc analysis revealed thatrats with both striatal and mfb lesions differed fromsham-lesioned rats’ performance. Control rats re-sponded steadily throughout the course of the session,and reached the 1 s minimum hold within 9 h in thechamber. Rats with striatal lesions began respondingearly in the session, like controls, but responding gener-ally subsided within 9 h and only two of the rats in thatgroup achieved the 1 s minimum hold. Rats with mfblesions generally emitted no responses for the first 6 hin the session, then responded only very little thereafter,earning an average of only 20% of the number ofreinforcers earned by controls (Table 1).

The lower panel of Fig. 1 compares the accuracy ateach minimum hold during the course of the overnightacquisition session between sham and striatal-lesionedgroups. Because of the low level of responding in themfb-lesioned group, accuracy for this group is notshown. At shorter minimum holds (0.03 and 0.1 s) thegroups did not differ. However, as the response require-ment was increased to longer lever holds, the groupsdiverged, with less accurate responding observed in thelesioned group relative to controls.

3.2. RDD performance

The effects of mfb 6-OHDA lesions on weekly per-formance of RDD responding for 5 weeks prior to, andup to 34 weeks following lesions are shown in Fig. 2and Fig. 3. Means are calculated from the performance

Fig. 1. Upper panel: acquisition of operant RDD responding in asingle overnight session following lesions of mfb (open squares)striaturn (filled circles) or sham lesion (open circles). Left vertical axisshows cumulative reinforcers earned and right vertical axis the mini-mum reinforced lever hold achieved in successive 30min intervals(horizontal axis). Lower panel: Accuracy (% of reinforced responses)at each required minimum hold during the overnight session in shamlesioned (open circles) and striatal lesioned (filled circles) animals. *indicate significant difference from sham controls (repeated measures2-way ANOVA, Dunnett’s post hoc: comparison to control, PB0.05).

Table 1Mean9SEM responses and reinforcers earned in the 14-h operantacquisition session*

Lesion group N Responses Reinforcers

Controls 10 128.7913216.9928.8Striatum 8 131.6924.8* 67.7199.3*

10 59.9922.3*mfb 26.999.3*

* Indicates PB0.05 (ANOVA, Dunn’s post hoc analysis).


Fig. 2. Effects of 6-OHDA lesions of mfb on average RDD performance 5 weeks prior to, and up to 34 weeks following lesioning. Panel A:mean9SEM responses emitted in single sessions (Thursdays). Panel B: Simplified presentation of data for responses emitted, collapsed across thepre-lesion baseline, then in 10-week intervals following lesioning. indicates significant difference (PB0.05) from pre-lesion baseline (Dunnett’s posthoc comparison, following significant omnibus ANOVA). Panel C: mean9SEM accuracy (% of reinforced responses) prior to and followingsurgery in single sessions. Panel D: Accuracy data collapsed across the pre-lesion interval, and across 10-week intervals following surgery. * as inpanel B.

of five lesioned rats (those rotating in response toamphetamine) and of six shams. Performance of threeadditional lesioned rats was also recorded (Table 2),but these data are excluded from analysis on the basisnegative rotational response to amphetamine (Table 2),suggesting an incomplete lesion. Following recoveryfrom surgery, the number of responses emitted by thelesioned group were reduced with respect to pre-lesionbaseline for the entire observation period (Fig. 2A). Ifdata are simplified, collapsing the pre-lesioned period,

in comparison to 10-week blocks of the post lesionperiod, responses remain reduced in each 10-week seg-ment with respect to control (ANOVA, F(3, 30)=29,PB0.001). Similarly, accuracy (the % of responsesreinforced), was also reduced throughout the entireperiod in which responding was measured (Fig. 2C, andin each 10-week segment with respect to control(ANOVA, F(3, 30)=15.3, PB0.001). The mean re-sponse duration remained quite steady in the shamlesioned rats, but in the lesioned rats varied both above

T.J.

Hudzik

etal./

Beha6ioural

Brain

Research

114(2000)

65–

7770

Table 2Optical intensity of TH stain in each hemisphere of brain, conversion to% decrease in staining intensity on the lesioned side, ispilateral rotations over 90 mins post d-amphetamine, and averageaccuracy in RDD responding in the rats in the lesioned group in which histology could be performed, unexpected attrition prevented histologic assessment in all subjects

Average optical intensity leftIdentifier 3 WeeksAverage optical intensity right Rotations Mean accuracy in% Decrease onlesioned side RDD(lesioned side)(unlesioned side)

9 Months3 Months

6-OHDA-lesioned group862 1190.710.5 717 898333 113

n.t. 26334 172 4.1 97.6 1236 863n.t. 35.459336

337 2216 15.61962 1226n.t. 26.7340342 1108

2229.7 0 50176 23.3 23451316248 1468 2136.6 85.2 1261346

1 5400348 24.021528364.2916.6 29.995.5Group mean 209.2930.0 88.4944.7

(9SEM)

Sham lesioned groupa

5494.9233.697.3 227.6948.5 2.52912.2Group mean(9SEM)

a Corresponding group mean values for the sham-lesioned group (bottom of table).


and below pre-lesion baseline (Fig. 3A). Because of thevariability of this endpoint, average mean responsedurations in 10-week blocks remained unchanged (Fig.3B, ANOVA, F(3, 30)=1.13, P=0.35). The varianceabout the mean response duration tended to be higherin the lesioned rats for much of the post-lesion period(Fig. 3C, but was significantly elevated only during thefirst 10-week block (Fig. 3D; ANOVA, F(3, 30)=5.8,PB0.003).

The distributions of response-durations of lesionedrats at 5-week intervals throughout the observationperiod are shown in Fig. 4. Unlesioned rats producedgreater proportions of both longer and shorter leverpresses with respect to pre-lesion baseline. By 30–34weeks post lesion, the distribution of response dura-tions had recovered considerably, but still was signifi-cantly different from the pre-lesion baseline in terms ofheightened proportions of both short- and long-dura-

Fig. 3. Effects of 6-OHDA lesions of mfb on average RDD performance 5 weeks prior to, and up to 34 weeks following lesioning. Panel A:mean9SEM mean response duration in single sessions (Thursdays). Panel B: Simplified presentation of data for mean response duration,collapsed across the pre-lesion baseline, then in 10-week intervals following lesioning. * indicates significant difference (PB0.05) from pre-lesionbaseline (Dunnett’s post hoc comparison, following significant omnibus ANOVA). Panel C: mean9SEM response duration variance prior to andfollowing surgery in single sessions. Panel D: variance data collapsed across the pre-lesion interval, and across 10-week intervals following surgery.* as in panel B.


Fig. 4. Effects of mfb lesion on the distribution of response durations at 5 week intervals following lesion. Each distribution represents groupsmean, S.E. on a single day (Thursdays). Horizontal axis-successive 0.1-sec interval, vertical axis, % of total responses (relative frequency).Response durations in the final interval (2.3 s) represent all response durations which were 2.3 s or greater. Dashed lines represent 95% confidencelimits about the prelesion distributions. *, PB0.05; **, PB0.01; ***, PB0.001, x2 test for relative frequency distributions versus prelesioncontrol distribution.

tion lever presses with respect to pre-lesion baseline.Sham lesioned animals, on the other hand, exhibitedremarkable stability in the pattern of response dura-tions throughout the study (data not shown). Thisstable pattern of responding in the shams is consistentwith that observed in previous studies [11–13].

Dose-effect functions for L-DOPA on performanceof the RDD schedule in both lesioned and sham-le-sioned rats are shown in Fig. 5. In both lesioned ratsand shams, L-DOPA evoked a dose-related suppressionof both responses and reinforcers earned, and the

highest dose (30 mg/kg) reduced accuracy of respond-ing in the shams (Dunnett’s post hoc, PB0.05), buthad more variable effects in the lesioned rats (P=0.34).Because of the already low baseline number of rein-forcers earned in the lesioned group, the highest dose ofL-DOPA did not significantly decrease reinforcers.Mean response duration was not consistently altered byL-DOPA in either group. In lesioned rats, 3 mg/kgL-DOPA tended to increase both the number of rein-forcers earned and to elevate accuracy of respondingalthough these effects were not significantly different


from control points. However, baseline accuracy postsurgically averaged 35.998.5%, which was significantlylower (T=2.58, P=0.021) than presurgically (52.892.5%). After either 3 or 10 mg/kg L-DOPA, accuracywas 50.896.2 and 50.792.6%, respectively, neither ofwhich differed significantly from pre-lesion control (T,P\0.05). These data argue for a definite trend ofefficacy of L-DOPA upon accuracy. In lesioned rats,L-DOPA also reduced the variance about mean responsedurations (F(7, 39)=2.32, PB0.04). Dunnett’s post hoccomparisons to control revealed that only the 3 mg/kgdose reduced variance significantly (PB0.05 vs. control).

The effects of L-DOPA on the pattern of responsedurations are shown in Fig. 6 and Fig. 7. Under controlconditions, lesioned rats had elevated proportions ofshort-duration responses and reduced proportions ofresponse durations that were reinforced (Fig. 6). Acrossmuch of the dose range of L-DOPA, consistent with thetrend for reduction of variance by L-DOPA, distribu-

tions of response durations sharpened following low tomoderate doses of L-DOPA. Following the 0.3, 1 mg/kg(not shown) and 3 mg/kg doses of L-DOPA, the distri-butions did not differ significantly from pre-lesion con-trol distributions. Although the 10 mg/kg dose ofL-DOPA also resulted in a distribution which did notdiffer from pre-lesion, that distribution was flatter thanthose following lower doses. After 30 mg/kg L-DOPA,a greater degree of flattening of the distribution wasobserved, which affected all three aspects, but mostsignificantly, increased the proportions of response dura-tions that were greater than 1.3 s. In sham lesioned rats(Fig. 7), distributions remained largely unaffected untilafter administration of 10 and 30 mg/kg L-DOPA,which, as in the lesioned rats after 30 mg/kg, increasedthe proportion of longer-duration responses (\1.3 s). Arepresentative stained section is shown in Fig. 8, and asummary of the TH intensities for each of the subjectsare shown in Table 2. There were no differences be-

Fig. 5. Effects of L-DOPA/carbidopa on performance of the RDD in sham lesioned (open circles) and rats with mFB lesions (filled circles). *indicates significant difference from control (c) performance (ANOVA, Dunnett’s post hoc). Control points are averages of performance insessions during which vehicle was administered (Thursdays).


Fig. 6. Effects of selected doses of L-DOPA on the RDD distributions in lesioned rats. Other details are as described in the legend for Fig. 4.

tween lesioned and sham groups in terms of TH inten-sities in the unlesioned hemispheres. However, on thelesioned side there was an average 64% decrease in THstaining relative to the unlesioned side, or 55% de-crease relative to the sham-operated group. Three ofthe five lesioned rats had relatively complete lesions(\85% reduction in TH density with respect to theunlesioned side) whereas the remaining two has verymodest lesions (−24% reduction). Three of the eightlesioned animals did not rotate in response to D-am-phetamine at any point following surgery (Table 2).However, because generally very high levels of DAdepletion is required to elicit rotation, and the relation-ship between the level of DA depletion and disruptionof RDD responding was not clear, the performance ofthese animals in the RDD was nevertheless assessed. Incomparing the size of the lesion to animals’ perfor-mance in the operant task, (Table 2) it is clear thatthose animals with greater lesions performed morepoorly on the task (reduced accuracy) than those withpartial lesions, which appeared to be relatively unaf-fected in the RDD.

4. Discussion

There were two main findings in the present study.The first was that acquisition of operant responding isdramatically impaired in rats given unilateral 6-OHDAlesions of the mfb, and somewhat less so in rats givenlesions in striatum. This behavioral deficit may providean additional, rapidly acquired baseline for studyingthe efficacy of putative neuroprotective agents. Thesecond finding was that in rats well-trained under theprocedure, unilateral mfb lesions produce very longlasting impairment in the performance of the response,providing an easily measured baseline, in addition torotation, for studying the effects of neurorestorativeagents, as well as of palliative treatments. This impair-ment in RDD responding was at least partially re-versed by administration of the antiparkinson drugL-DOPA.

A number of studies have assessed the effects ofbilateral doparninergic depletion on schedule-con-trolled behaviors. For example, in rats which had beenneonatally depleted of DA by intracisternal injection of


6-OHDA, deficits in adulthood, such as acquisition ofoperant responding [11,21] have been shown. In adultanimals, early bilateral DA depletion has been shownto alter operant differential reinforcement of low rates(DRL) [22], fixed ratio (FR) performance [5,24,31] andperformance in reaction-time tasks [1]. However, withneonatal DA depletion, compensatory changes in DAand related motor circuitry may tend to limit the utilityof this approach to gauge the potential efficacy of

antiparkinson drugs. The bilateral lesion model in adultrats possesses more face validity in terms of modelingParkinson disease, but can result in reasonably highmortality, at least partly due to deleterious effects uponfeeding behavior. If animals survive, the impact uponfeeding or motivation may in turn alter any foodreinforced behavior to be measured. This was the pri-mary reason that we set out to assess motor deficitsfollowing unilateral lesion.

Fig. 7. Effects of doses of L-DOPA on the RDD distributions in sham lesioned rats. Other details are as described in the legend for Fig. 4.

Fig. 8. Comparison of tyrosine hydroxylase staining in unlesioned (left) and lesioned (right) hemispheres in a single section through dorsal striatumin a representative subject. Dark stain corresponds to TH.


The deficits in acquisition of operant respondingshown in the present study contrast with observationsfollowing similar lesions engendered by different means.For example, following experimental stroke, by occlu-sion of the middle cerebral artery or by unilateralcarotid ligation coupled to hypoxia, both of whichresult in extensive striatal damage in addition to overly-ing cortex, operant acquisition is completely unaffected[16,17], despite appearance of other deficits [16,26]. Thismay be due the relative areas of striaturn which areaffected. In middle cerebral artery occlusion, dorsal andmedial striatum, is primarily damaged enough to pro-duce rotation [8,16,19] but lateral striatum, specificallyventrolateral striatum, is often spared. Ventrolateralstriatum was affected by both our striatal and medialforebrain 6-OHDA lesions, in agreement with literaturesupporting the involvement of this region in operantresponding [5]. It remains to be distinguished whetherthe disruption in operant responding was due to adisruption in motivation, in motor function, in newlearning ability, or to a combination of these. However,the fact that animals with experimental stroke havemotor deficits, with accompanying difficulty maintain-ing body weight [14,15], and yet can acquire operantresponding as well as controls, suggests that motorfunction and motivation may to some extent be separa-ble from the learning component. Additionally, itshould be noted that in striatally-lesioned rats, responseoutput was generally similar to controls, and yet accu-racy of responding was less than in controls. Thisargues against bradykinesia as an explanation of theacquisition deficit noted. Rather, it seems more likelythat subjects had difficulty adapting their behavior tothe changing schedule conditions.

However, this can be more directly addressed byinitially studying acquisition behavior at time pointsmore distant from the surgical insult than was done inthe present study.

Relatively few studies have examined the effects ofunilateral dopaminergic lesion upon behaviors otherthan rotation. Paw reaching tasks have shown sensitiv-ity in the contralateral side, and in some cases in theipsilateral side, which have been shown to persist for upto at least 3–6 months following surgery [7,32]. Severaloperant studies, which cleverly restricted the use of theipsilateral paw for emitting lever presses, have alsoshown relatively long-term sensitivity [27–29]. Whydeficits occur in the lever holding task employed in thepresent study following unilateral lesion without re-stricting paw use may be due to several factors. Bothnormal and lesioned subjects most often use both pawson the operant manipulanda in the RDD procedure,and deficits in release time of the affected, contralateralpaw may have been primarily responsible. Second,there is precedent for bilateral deficits following unilat-

eral lesion [32], (e.g. due to overactivity in the intacthemisphere), and the possibility that both paws wereaffected cannot be ruled out.

It is possible that, in addition to motor disturbances,there is a disruption in temporal perception followinglesion. Parkinson’s patients underestimate the passageof time as demonstrated in a number of tasks [4,23].Temporal underestimation would be reflected in theRDD procedure as an increased proportion of longerduration responses (\1.3 s). While this was observedin the present study, increases in short-duration re-sponses (B1 s), were also, and perhaps more consis-tently observed. It is, however, unclear how unilaterallesion and consequent interhemispheric dopamine im-balance might interact with timing as controlled bystriatum. It would be of interest to assess in futurestudies the impact of bilateral depletion, as well as ofmore focal lesions, on these timed responses.

There is literature to suggest that mediation of timingbehavior may be lateralized primarily to the right hemi-sphere [10,18]. In the present studies, lesions were re-stricted to the right hemisphere, which then necessarilyshould have altered motor timing. If such lateralizationexists in the rat, then similar lesions in the left hemi-sphere should either leave responding either relativelyintact, or should produce qualitatively different deficitsif overactivity in the right hemisphere impacts perfor-mance. This, of course, is readily testable.

In trained rats subsequently treated with 6-OHDA,the duration of the deficits in the present study (morethan 34 weeks post-lesion) despite repeated daily train-ing was noteworthy. Both the number of responses andaccuracy of responding remained poorer in 6-OHDAtreated rats than in controls throughout the entireobservation period. Although it is clear that repeatedexperience can alter the magnitude of lesion-induceddeficits over time, as well as partially reverse the do-pamine depletion (e.g. [31]) deficits in the present studypersisted, perhaps, due to the very fine motor andtemporal requirements of the task.

The RDD schedule can provide useful measures ofmotor performance, which could be used to test preclin-ical drug efficacy. For example, accuracy, which wasaffected throughout the entire post-lesion period, meanand variance about the mean response duration, andthe distribution of response durations all provide addi-tional indices of performance. Although, in the presentstudy, the effects of L-DOPA upon the RDD baselinein lesioned rats were modest, the antiparkinson drugdid clearly improve the quality of responding by de-creasing the variance of the distributions of responsedurations, and tending toward increasing accuracy. Ourprimary goal, however, was to assess the duration ofthe sensitivity of the baseline. In subsequent studies, forthe purpose of assessing drug efficacy, it may be more


informative to begin drug testing at earlier time periods,for example at less than 20 weeks post lesion, wherevariance was highest and accuracy lowest.

References

[1] Amaleric M, Koob GF. Depletion of dopamine in the caudatenucleus but not in nucleus accumbens impairs reaction timeperformance in rats. J Neurosci 1987;7:21292134.

[2] Barneoud P, Mazadier M, Miquet JM, et al. Neuroprotectiveeffects of riluzole on a model of Parkinson’s disease in the rat.Neuroscience 1996;74:971–83.

[3] Church RM. Properties of the internal clock. Ann NY Acad Sci1984;423:566–82.

[4] Cooper JA, Sagar HJ, Doherty SM, Jordan N, Tidewell P,Sullivan EV. Different effects of dopaminergic and anticholiner-gic therapies on cognitive and motor function in parkinsonsdisease: a follow-up study of untreated patients. Brain1992;115:1701–23.

[5] Cousins MS, Sokolowski JD, Salamone JD. Different effects ofnucleus accumbens and ventrolateral striatal dopamine deple-tions on instrumental response selection in the rat. PharmacolBiochem Behav 1993;46:943–51.

[6] Dominey PF, Jeanneroid M. Contribution of frontostriatal func-tion to sequence learning in Parkinson’s disease: evidence fordissociable systems. Neuroreport 1997;8:8.

[7] Evenden JL, Robbins TW. Effects of unilateral 6-OHDA lesionsof the caudate putamen on skilled forepaw use in the rat. BehavBrain Res 1984;14:61–8.

[8] Grabowski M, Brundin P, Johansson BB. Paw-reaching, sensori-motor, and rotational behavior after brain infarction in rats.Stroke 1993;24(6):889–95.

[9] Kinney GG, Griffilths JC, Hudzik TL. Effects of 5-hydrox-ytryptaminel 1A receptor agonists on operant responding underthe response duration differentiation schedule. Behav Pharmacol1998;9:309–18.

[10] Harrington DL, Haaland KY, Knight RT. Cortical networksunderlying mechanisms of time perception. J Neurosci1998;18:1085–95.

[11] Heffner TG, Seiden LS. Impaired acquisition of operant re-sponse in young rats depleted of brain dopamine in neonatal life.Psychopharmacology 1983;79:115–9.

[12] Hudzik TJ, McMillan DE. Drug effects on response durationdifferentiation 1: differential effects of drugs of abuse. Psy-chopharmacology 1994;114:620–7.

[13] Hudzik U, McMillan DE. Drug effects on response durationdifferentiation 11: selective effects of antidepressant drugs. JPharmacol Exp Ther 1994;268:1335–42.

[14] Hudzik U, McMillan DE. Drug effects on response durationdifferentiation IV: effects of trimethyltin. Neurotoxicol Teratol1995;17:665–7.

[15] Hudzik U, Palmer GC. Effects of anticonvulsants on operantlearning in rats: comparison of remacemide hydrochloride and

FPL 15896AR to other anticonvulsant agents. Epilepsy Res1995;21:183–93.

[16] Hudzik TJ, Borrelli A, Bialobok P, et al. Long-term functionalend points following middle cerebral artery occlusion in the rat.Pharmacol Biochem Behav 277;65:553–62.

[17] Hudzik U, Sydserff S, Bialobok P, et al. Short and long termeffects of hypoxia-ischemia in young brain 11: behavioral assess-ment, Soc Neurosci Abs, 1999, 321.

[18] Maquet P, Lejeurte H, Pouthas V, et al. Brain activation inducedby estimation of duration: a PET study, NeuroImage 3:119–126.

[19] Markgraf CG, Green EJ, Hurwitz BE, et al. Sensorimotor andcognitive consequences of middle cerebral artery occlusion inrats. Brain Res 1992;575(2):238–46.

[20] Meck WH. Selective adjustment of the speed of the internalclock and memory processes. J Exp Psychol Anim Behav1983;9:171–201.

[21] Moy SS. Impaired acquisition and operant responding afterneonatal dopamine depletion in rats. Pharmacol Biochem Behav1995;52:433–41.

[22] O’Donnel M, Seiden LS. Altered effects of desipramine onoperant performance after 6-hydroxydopamine-induced deple-tion of brain dopamine or norepinepherine. J Pharmacol ExpTher 1984;229:629–35.

[23] Pastor MA, Artieda J, Jahanshahi M, Obeso J. Time estimationand reproduction is abnormal in parkinson’s disease. Brain1992;115:211–25.

[24] Peterson DW, Laverty R. Operant behavioural and neurochemi-cal effects after 6hydroxydopamine treatment. Psychopharmacol-ogy 1976;50:55–60.

[25] Sommer M, Graftnan J, Clark K, Hallett M. Learning inParkinson’s disease: eyeblink conditioning, declarative learning,and procedural learning. J Neurol Neurosurg Psychiat1999;67:27–34.

[26] Smith SE, Hodges H, Sowinski P, et al. Long-term beneficialeffects of BW619C89 on neurological deficit, cognitive deficitand brain damage after middle cerebral artery occlusion in therat. Neuroscience 1997;77(4):1123–35.

[27] Uguru-Okerie DC. Nigrostriatal dopaminergic control of oper-ant and spatial behavior in the rat. Physiol Psychol 1982;10:97–107.

[28] Uguru-Okerie DC. Postsynaptic dopamine receptor stimulationand sensorimotor behavior in rats with a unilateral medialforebrain lesion. Physiol Psychol 1984;12:153–5.

[29] Uguru-Okerie DC. Feeding behavior and sensorimotor function-ing in rats with medial forebrain bundle damage. Physiol Psychol1984;12:311–4.

[30] Ungerstedt U. 6-hydroxydopamine-induced degeneration of thenigrostriatal dopamine pathway: the turning syndrome. Pharma-col Ther 1976;2(1):37–40.

[31] VanKeuren KR, Stodgell CJ, Schroeder SR, Tessel RE. Fixedratio discrimination training as replacement therapy in Parkin-son’s disease: studies in a 6-hydroxydopamine treated rat model.Brain Res 1998;780:56–66.

[32] Whishaw IQ, O’Connor WT, Dunnett SB. The contributions ofmotor cortex, nigrostriatal dopamine and caudate-putamen toskilled forelimb use in the rat. Brain 1986;109:805–43.

.

Documents

Disruption of acquisition and performance of operant response-duration differentiation by unilateral nigrostriatal lesions