Dopamine denervation of specific striatal subregions differentiallyaffects preparation and execution of a delayed response task in the
rat
T. Florio, A. Capozzo, A. Nisini, A. Lupi, E. Scarnati *Department of Biomedical Technology, Uni6ersity of L’Aquila, Via Vetoio Coppito 2, I-67100 L’Aquila, Italy
Delayed initiation and execution of movements aremajor symptoms observed in parkinsonian patients[25,36,46], and in animals subjected to pharmacological
blockade or destruction of the dopaminergic centralsystems [31,59]. However, the behavioral alterationswhich can be appreciated in these conditions also sug-gest an involvement of the basal ganglia in cognitiveand in motivational aspects of the organization ofgoal-directed behavior.
The dopaminergic innervation of specific regions ofthe striatum is critical for these mechanisms owing toits influence on a wide range of high order functions of
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–6252
the brain, including sensorimotor integration[18,21,26,27,51], attention, memory, and motivationalprocesses [6,15,17,23,27,28,43,45,46,48,51–54,57] thatoperate in behavioral learning.
The possibility that different neuronal populationswithin the striatum could subserve many of these func-tions has been supported by the results of single cellrecordings in monkeys performing rewarded move-ments in response to the presentation of signals havingbehavioral significance. Thus, neurons recorded mainlyin the caudate nucleus have been shown to specificallymodulate their discharge at the presentation of sensorysignals to initiate arm, head or eye movements in abehavioral context [9,33–35,49,50,58]. As expected, as aconsequence of dopamine denervation in behaving ani-mals, some of these responses have been proved to becritically influenced by the nigrostriatal dopamine sys-tem [7]. Other caudate neurons have been found toincrease their discharge up to the presentation of anexpected impending event, including reward delivery[10,35,60], while the majority of neurons of the ventralstriatum have been reported to exhibit a sustainedincrease of activity during expectance of reward [8,60].These responses could provide a basis for the centralrepresentation of predictable environmental events andmay participate in the neuronal processes that underlythe organization of the behavioral output by the basalganglia.
Interestingly, phasic neuronal responses linked to thepresentation of stimuli of behavioral significance, aswell as to reward delivery have also been reported inthe rat and cat striatum [3,19,29,39,64,65]. Thus, it ispossible that at least some of the behavioral propertiesof striatal neurons are common across species.
The contribution of the dopaminergic innervation ofboth the dorsal and ventral striatum in the expectanceof impending reward-related stimuli has not been spe-cifically investigated. Thus, the present study has beenundertaken with two objectives: first, to obtain a morecomprehensive knowledge of the relationship betweenthe dopaminergic innervation of these striatal subre-gions in the preparation and execution of a delayeddual bar pressing task using unpredictable time re-straints, and second, to test the capability of the rat toperform in a paradigm requiring sustained waiting peri-ods comparable to those previously tested in monkeys[8–10].
2. Materials and methods
2.1. Animals
All the experiments have been conducted in accor-dance with Italian law 116/95 governing the use ofexperimental animals, and supervised by the university
veterinary service. Consistent with the national policyon the use of laboratory animals, efforts were made tominimize suffering and the number of animals.
Fifteen male albino Wistar rats weighing 160–180 gat the beginning of experiments were used. Animalswere maintained at 90% of the free-feeding weight ofcontrol animals for the entire duration of the experi-ments by 15 g/day per rat of standard laboratory chow,provided after the daily experimental session. Animalswere housed individually in a temperature-regulated(22°C) environment under an inverted 12 h light–darkcycle.
2.2. Apparatus and stimuli
Two operant chambers (26×30×30 cm, CoulbournInstruments) housed in acoustically insulated boxes andsupplied with an extractor fan providing a constant lowlevel background noise, were used.
Two adjacent 35 mm levers were placed horizontallyin the front panel of each chamber, at a distance of 45mm each other, and at a height of 1.7 cm from thechamber floor. Levers required a pressure force of 0.3N for switch closure. A food pellet dispenser, equippedwith a 2.8 W white light, was located to the right of thesecond lever. Chambers were softly illuminated by a 2.8W white lamp located on the ceiling. A green light-emitting diode (LED), associated to a 2.9 kHz tone andvisible to the animal when its head was directed to thelever, was located 3 cm above the center of the firstlever. A second green LED, associated to a 4.5 kHztone, was placed above the center of the second leverwhile the loudspeakers for tones were mounted on thetop of the front panel. A miniaturized videocameraplaced in each chamber allowed the evaluation of theposition of the animal in order to deliver the stimuliwhen its head was appropriately directed to the LEDpanel.
2.3. Beha6ioral training
Animals were subjected to classical conditioningthough complex training schedules that introduced suc-cessive components of the final paradigm. Correct re-sponses resulted in the delivery of a single pellet (45 mgFormula P, P.J. Noyes Company) with a concomitant 2s illumination of the food tray. Incorrect trials were notrewarded and no illumination occurred. Initially therats were trained to press the second lever at thepresentation of the LED concerned with this leverand/or to the 4.5 kHz tone. If the animal did not pressthe lever within the 2 s in which the signal was pre-sented the trial was terminated. Next, the animals wererequested to press the first lever at the presentation ofthe signal concerned with this lever. When they reliablyresponded to the individual presentation of each of the
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–62 53
behavioral signals, the paradigm was implemented ask-ing the animals to respond to the consecutive presenta-tion of the two signals. In the final paradigm, thestimulus concerning the first lever assumed the value ofan instruction signal initiating the trial while the stimu-lus related to the second lever had the meaning of atrigger signal starting the movement. As the rat pressedthe first lever the instruction signal turned off but theanimal was required to hold the lever until the presen-tation of the trigger signal. The delay between the offsetof the instruction and the onset of the trigger signal wasthen progressively increased across trials until the ratswere able to wait for 2, 3 and 4 s. When a stableperformance of 70–80% of correct trials was reached(approximately after 4–5 weeks) in the delayedparadigm, a session for each animal was dedicated toascertain whether the teleceptive nature of the triggersignal could have any influence in the occurrence ofcorrect responses. This session consisted of 30 trials foreach delay with the trigger signal randomly delivered aseither visual or acoustic or combined whereas the in-struction signal was always presented in form of com-bined LED and 2.9 kHz tone. Thereafter, data werecollected for 10 consecutive days in order to obtainprelesion baseline values. Each daily session included 30trials for each of the three delays randomly andequiprobably generated. Usually, there was a 5–20 sintertrial interval, depending on the proper position ofthe animals facing the LED panel.
The two levers assured to calculate movementparameters with the animal initiating to move alwaysfrom the same position.
Reaction time was measured between the onset of thetrigger signal and release of the first lever, whereasmovement time was calculated between release of thefirst lever and press of the second lever (Fig. 1). Mea-
surements were made on-line by a Basilink data acqui-sition system (Ugo Basile-Comerio), stored andevaluated using appropriate software.
2.4. Surgery
The animals were anesthetized with choral hydrate(400 mg/kg i.p.) and placed in a stereotaxic apparatus.Rupture of the tympanic membrane by the ear bars wasprevented by using protective ear cups. A hole wasdrilled in the theca to insert a thin microcannula (o.d.200 mm) fixed vertically to a micromanipulator andconnected to a Hamilton micrometric microsyringethough a polyethylene PE10 tube. A volume of 1.5 ml ofsterile saline containing 4 mg/ml of 6-hydroxydopamine(6-OHDA) and 0.1% ascorbic acid was injected over 5min, leaving the microcannula in situ for further 5 min.
For the dorsal striatum the stereotaxic coordinateswere: AP+10.0 mm from the interaural line, L92.0mm from the midline and DV−3.5 mm from thecortical surface, while those for ventral striatum were:AP+9.2 mm, L92.0 mm and DV−7.1 mm, accord-ing to the atlas of Paxinos and Watson [44]. The firstinjection was made in the right side, i.e. ipsilaterally tothe side at which the lever to press in response to thetrigger presentation was located.
One week after the first injection, the rats were testedfor 10 consecutive days in the delayed paradigm, and4–5 days later they received a second injection of6-OHDA in the other hemisphere under the same surgi-cal procedures. After recovery for 1 week from thesecond surgery, animals were tested for a further 10consecutive days. Thus, each rat had its own behavioralcontrol before each surgery.
2.5. Immunohistochemistry
After completion of experiments the rats, under deepchloral hydrate anesthesia, were perfused through theheart with 50 ml of cold saline containing 0.2 ml ofheparin (5000 IU/ml), followed by 300 ml of 4%paraformaldehyde in 0.1 M phosphate buffer (pH 7.4).The brains were removed from the skull and slabscontaining the striatum and substantia nigra werepostfixed in the same perfusion solution overnight at4°C. After incubation in two cryoprotective 15 and 30%sucrose solutions, 30 mm-thick coronal sections were cutwith a freezing microtome, placed on gelatin-coatedslides and processed for tyrosine hydroxylase im-munoreactivity using a monoclonal antibody againsttyrosine hydroxylase (Boehringer Mannheim Biochem-ica) diluted 1:1000 according to the method of Abrouset al. [1]. Tyrosine-hydroxylase positive staining wasrevealed by the biotin-avidin technique (ABC kit, Vec-tor) using 3,3%-diaminobenzidine as the chromogen (per-oxidase substrate KIT-DAB, Vector). Sections were
Fig. 1. Schematic representation of the delayed dual bar pressingtask. The animal had to press the first lever starting from thepresentation of the instruction signal until the onset of the triggersignal. Then, it had to release the lever and press the second lever forfood reinforcement. Reaction time (RT) was measured between theonset of the trigger signal and release of the first lever, whilemovement time (MT) corresponded to the interval between release ofthe first lever and press of the second lever.
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–6254
cleared in xylene, mounted in permount and analyzedby a computerized image analyzer system. Two brainstaken from conditioned animals not subjected to dener-vation, were processed with the same procedure as thebrains of the lesioned animals to serve as controls forthe immunohistochemical staining. Results obtainedfrom rats in which the loss of the dopaminergic termi-nals was confined to either the dorsal striatum (n=6)or ventral striatum (n=5) are included in the study.
2.6. Data and statistical analysis
The results are expressed as mean9S.E.M. for reac-tion and movement times and for percentages of correctand incorrect responses. The data were analyzed by atwo-way analysis of variance (ANOVA) by combininginto a single design one repeated measures factor (de-lay) and one between groups factor (lesion), with afurther comparison using the post hoc Newman–Keulstest. The level of statistical significance was taken to bePB0.05. Each group was evaluated according to pre-operative, uni- and bilateral denervation values. Thestatistical package software CSS STATISTICA ver.3.1(Statsoft, USA) was employed.
3. Results
3.1. Histochemical analysis
Reconstruction of the extention of the striatal do-paminergic denervations is reported in Fig. 2. Theinfusion of 6-OHDA caused a discrete denervation ofthe target regions of the striatum. The lesion wassymmetrical in both sides and accompanied by a retro-grade degeneration of dopamine neurons in the sub-stantia nigra, reflecting the topographical organizationof the dopaminergic nigro-striatal projection, i.e. thedorsal striatum receiving fibers from the lateral, and theventral striatum from the medial part of the parscompacta of the substantia nigra, as shown in Fig. 3.
3.2. Motor performance
The animals did not display any particular pawpreference when pressing levers. The majority held thefirst lever with the right paw and pressed the secondwith the same or with both paws simultaneously. Thismotor strategy did not change following the uni- orbilateral denervation of the two striatal regions.
The motor performance appeared to be independentof the teleceptive nature of the trigger signal elicitingmovement. After reaching a 70–80% of successful trialsin response to a composite trigger signal, all animalsshowed no significant response reduction whether thetrigger was presented in the form of either an acousticor visual signal.
Fig. 2. Overall extent of the loss of tyrosine-hydroxylase immunoreac-tive fibers in the dorsal (A) and ventral striatum (B) groups. Thereconstruction is based on coronal plates derived from the atlas ofPaxinos and Watson [44] from 11.2 to 8.7 mm from the interaural linefor the dorsal striatum and from 11.7 to 9.2 mm for the ventralstriatum along with the anteroposterior axis. The solid area repre-sents the core of the lesion while the surrounding hatched areacorresponds to the border of the lesion where a weak immunoreactiv-ity was still present.
3.3. Correct responses
As shown in Fig. 4A, in the dorsal striatum group asignificant effect of the delay factor was observed [PB0.001 each delay, Newman–Keuls, F(2,154)=46.2], i.e.the number of correct responses decreased as the prob-ability of occurrence of the trigger signal at the longestdelays increased. In addition, there was a significantlesion×delay interaction [PB0.001, F(4,154)=6.1] asthe denervations did not reduce the performance exceptfor the 2 s delay [PB0.05 unilateral vs prelesion, andbilateral vs pre- and unilateral lesion, Newman–Keulstest], thus the delay effect was suppressed after thebilateral lesion.
In the ventral striatum group there was a significanteffect of the delay factor [PB0.001 each delay, New-man–Keuls, F(2,152)=43.7], of the lesion factor [PB0.05 bilateral vs pre- and unilateral lesion,Newman–Keuls, F(2,76)=12.4] and a lesion×delayinteraction [PB0.001, F(4,152)=8.2]. The Newman–Keuls test indicated that the unilateral lesion caused areduction only at the 2 s delay (PB0.001 vs prelesion),and that the bilateral denervation caused a sharp reduc-tion of correct responses (PB0.001 vs pre- and unilat-eral lesion, each delay) with a concomitant suppressionof the delay effect. The comparison between the twogroups revealed a significant decrease in the occurrenceof correct responses in the ventral striatum group only
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–62 55
for the 2 s delay after the unilateral lesion [PB0.001,Newman–Keuls, F(2,98)=3.9], and for each delay af-ter the bilateral denervation [PB0.001, Newman–Keuls, F(2,82)=0.2].
3.4. Reaction time
Reaction time (Fig. 4B) decreased preoperatively as afunction of the delay with faster values the longer thedelay. The two-way analysis of variance in the dorsalstriatum group showed a significant reduction for thedelay [PB0.05 each delay, Newman–Keuls,F(2,154)=35.6] but not for the lesion factor. On thecontrary, the analysis in the ventral striatum group wassignificant for the delay as well as for the lesion factor[PB0.05, F(2,152)=5.6 and PB0.001, F(2,76)=9.2,respectively]. Further analysis by the Newman–Keulstest showed that the unilateral denervation increasedthe reaction time at 3 and 4 s (PB0.05 vs prelesion),but the bilateral lesion significantly increased the reac-tion time for all the delays (PB0.001 vs pre- andunilateral lesion) thus suppressing the delay effect. Theincrease of the reaction time following the unilateraldenervation of the ventral striatum group compared tothe dorsal striatum group was significant for the 3 and
4 s delays [PB0.05, Newman–Keuls, F(2,98)=1.2],but reached levels of significance for all the delayscomparing the bilateral lesions of the two groups [PB0.001, Newman–Keuls, F(2,82)=2.6].
3.5. Mo6ement time
Movement time (Fig. 4C) did not significantly changein the dorsal striatum group either as a function of thelesion or delay. As far as the ventral striatum groupwas concerned only the bilateral denervation induced asignificant lesion effect [PB0.001 vs pre- and unilaterallesion, Newman–Keuls, F(2,76)=13.1]. The movementtime following the bilateral denervation of the ventralstriatum compared to that observed following theanalogous lesion of the dorsal striatum was significantlyincreased for each delay [PB0.001, Newman–Keuls,F(2,82)=3.1].
3.6. Incorrect responses
3.6.1. Incorrect responses to the presentation of theinstruction signal
Three types of incorrect responses occurred upon thepresentation of the instruction signal. The first wasrepresented by an unconditioned response to the in-struction (UNI in Fig. 5), i.e. the animals exhibitedspontaneous behavior such as grooming or rearing,instead of pressing and holding the first lever. Thesecond type of incorrect response occurred when theanimals remained in front of the instruction LED,showing no overt motor reaction (no press of first lever,NP1 in Fig. 5). The UNI responses significantly in-creased in the dorsal striatum group following both theuni- and bilateral denervation [PB0.05 both lesions vsthe preoperative value, Newman–Keuls, F(2,77)=9.6],whereas the NP1 responses significantly increased fol-lowing the bilateral denervation [PB0.05 vs the preop-erative value, Newman–Keuls, F(2,77)=4.6]. Incontrast, in the ventral striatum group only the bilateraldenervation significantly increased both UNI and NP1responses [PB0.001 vs pre- and unilateral lesion, New-man–Keuls, F(2,76)=12.2 for the UNI responses, andPB0.001, Newman–Keuls, F(2,76)=24.1 for the NP1ones]. In comparing the dorsal and ventral striatumgroups it was apparent that the former was character-ized by a remarkable number of UNI responses [PB0.05 both lesions, Newman–Keuls, F(1,49)=4.9 andF(1,41)=10.2 in uni- and bilateral lesion, respectively],whereas the latter preferably manifested NP1 responsesfollowing the bilateral lesion [PB0.05, Newman–Keuls, F(1,41)=8.1]. The third type of incorrect re-sponse was to press the lever after the offset of theinstruction signal, thus, it was considered a delayedpress of the first lever (DP1 in Fig. 5). This responsewas not altered by the dorsal striatum denervations
Fig. 3. Representative coronal sections showing the focal dopaminer-gic denervations of dorsal and ventral regions of the striatum withretrograde loss of tyrosine-hydroxylase immunostaining in differentmesencephalic zones. Control sections of the striatum and substantianigra (A and B). The denervation of the dorsal striatum (C) wasaccompanied by a loss of immunoreactivity in the lateral parts of thesubstantia nigra pars compacta (D), while the denervation of theventral striatum (E) caused a loss of immunoreactivity mainly in themedial part of the pars compacta and adjacent ventral tegmental area(F). The lateral striatum was spared by the denervation in both cases.Scale bar, 1 mm.
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–6256
Fig. 4. The effects of dorsal (DS) and ventral striatum (VS) uni- and bilateral dopaminergic denervations (UL and BL, respectively) on motorperformance (A), reaction time (B) and movement time (C) compared to the prelesion values (PL). The effects have been evaluated according tothe delay elapsed between presentations of the instruction and trigger signals. The delay factor significantly affected the percentage of correctresponses (* PB0.001 for both groups, ANOVA) and the reaction time (* PB0.001 for DS group, and * PB0.05 for VS group, ANOVA). Thelesion factor significantly affected the VS group for the motor performance, reaction and movement times only after the bilateral lesion(° PB0.001, ANOVA). The lesion×delay interaction was significant for the correct responses in both groups (c PB0.001, ANOVA). Thedifference between DS and VS groups was significant for the correct responses (@ PB0.001, ANOVA), for the reaction time (@ PB0.001,ANOVA) and for the movement time (@ PB0.05, ANOVA).
whereas it was significantly increased following theventral striatum denervations [PB0.05, Newman–Keuls, F(2,76)=7.2]. Both uni- and bilateral lesionssignificantly increased the frequency of this response inthe ventral striatum [PB0.05, Newman–Keuls,F(1,49)=11.3 and PB0.001, Newman–Keuls,F(1,41)=22.6, respectively] as compared to the dorsalstriatum group.
3.6.2. Incorrect responses on the first le6erDuring the period in which the lever was to be held
prior to the onset of the signal triggering movement ananticipated release (AR in Fig. 6) of the lever couldoccur. This type of incorrect response was significantlycorrelated to the delay in both groups, i.e. the longerwas the delay, the higher was its occurrence [PB0.001each delay, Newman–Keuls, F(2,154)=74.3 andF(2,152)=80.5 for the dorsal and ventral striatumgroups, respectively]. The bilateral denervation of thedorsal striatum significantly reduced these responses[PB0.05 vs pre- and unilateral lesion, Newman–Keuls,F(2,77)=4.5]. As proved by the two-way ANOVAanalysis there was a significant interaction between thelesion and delay factors, as the bilateral lesion reducedthe responses at the delays of 3 and 4 s [PB0.05,Newman–Keuls, F(4,154)=6.3]. The ventral striatumdenervation failed to alter the frequency of the re-
sponses despite a significant lesion×delay interaction[PB0.001, F(4,152)=6.1]. Furthermore, the New-man–Keuls test demonstrated that the bilateral lesionof the ventral striatum abolished the delay effect (PB0.05 at the 2 s and 4 s vs pre- and unilateral lesion),resulting in differences not significant between the vari-
Fig. 5. Incorrect responses to the presentation of the instructionsignal included unconditioned responses (UNI), no press (NP1) anddelayed press of the first lever (DP1). The lesion factor was significantin both dorsal (DS) and ventral striatum (VS) groups for UNI(° PB0.001, ANOVA), and for NP1 (° PB0.05 and ° PB0.001,respectively, ANOVA). The UNI and NP1 responses were bothsignificantly different between the two groups (@ PB0.05, ANOVA).The DP1 occurred at a significant frequency in the VS group (° PB0.05, ANOVA). The difference between the two groups for this kindof response was significant after uni- and bilateral denervation(@ PB0.05 and @ PB0.001, respectively, ANOVA).
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–62 57
Fig. 6. Anticipated release of the first lever (AR). This type ofincorrect response was significant correlated with the delay factor inboth dorsal (DS) and ventral striatum (VS) groups (* PB0.001,ANOVA). The lesion factor was significant for the DS group (° PB0.05, ANOVA). The lesion×delay interaction in both groups wassignificant (c PB0.001, ANOVA). The difference in the occurrenceof this response comparing the two groups was also significant(@ PB0.05, ANOVA).
except for the 2 s delay (PB0.001 vs unilateral lesion,Newman–Keuls). The lesion×delay interaction waspositive since a net delay effect was observed followingthe bilateral denervation [PB0.05 each delay, New-man–Keuls, F(4,152)=4.2]. This indicated that theanimals behaved like those bearing bilateral dorsalstriatum denervation, i.e. they correctly responded ifsufficient time was given for preparing movement. Thefact that UNT responses were most likely to occur inthe ventral striatum group with respect to the dorsalgroup was confirmed by the post hoc comparison [PB0.05 each delay, F(2,98)=4.0, and PB0.05 at 2 and 3s, F(2,82)=0.8, for uni- and bilateral lesion,respectively].
The DP2 responses significantly increased only fol-lowing the bilateral denervation of the ventral striatum[PB0.05 vs pre- and unilateral lesion, Newman–Keuls,F(2,76)=8.7, and PB0.001 vs dorsal striatum, New-man–Keuls, F(1,41)=15.0].
ous delays. A comparison between the dorsal and ven-tral striatum groups demonstrated that the responseshad occurred at higher frequency in the ventral stria-tum group following the bilateral lesion with delays of2 and 3 s [PB0.05, Newman–Keuls, F(2,82)=2.2].
3.6.3. Incorrect responses to the trigger signalIncorrect responses occurred when the animal disre-
garded the trigger signal (unconditioned response to thetrigger, UNT in Fig. 7A), when they moved away fromthe second lever or stood on the first one, or when theypressed the lever after the trigger signal had extin-guished (delayed press of the second lever, DP2 in Fig.7B). The UNT responses were analyzed taking intoaccount the delay elapsed before the onset of the triggersignal. In the dorsal striatum group the frequency ofUNT responses decreased with the delay, reaching lev-els of significance with the delay of 4 s [PB0.05 vs 2 s,Newman–Keuls, F(2,154)=4.7]. These responses weresignificantly more frequent following the bilateral den-ervation [PB0.05 vs pre- and unilateral lesion, New-man–Keuls, F(2,77)=7.5]. A significant lesion×delayinteraction was present following the bilateral denerva-tion for 2 s [PB0.001 vs pre- and unilateral lesion,Newman–Keuls, F(4,154)=2.5], in which a delay ef-fect occurred (PB0.05 vs 3 and 4 s, Newman–Keuls).Hence, it is likely that the animals were still able tomake a correct response if they were given a longerpreparation period.
In the ventral striatum group the frequency of theseresponses decreased with delay [PB0.05 each delay,Newman–Keuls, F(2,152)=20.0]. Unilateral denerva-tion increased their occurrence when compared to thepreoperative period [PB0.001 each delay, Newman–Keuls, F(2,76)=15.3], but no further significant in-crease was noted following the bilateral denervation
Fig. 7. Incorrect responses to the trigger signal included (A) uncondi-tioned responses to this signal (UNT), and (B) delayed press of thesecond lever (DP2). The delay factor was significant for the UNTresponses in both dorsal (DS) and ventral striatum (VS) groups(* PB0.05 and * PB0.001, respectively, ANOVA), as well as thelesion factor (° PB0.05 and ° PB0.001, respectively, ANOVA). Thelesion×delay interaction was also significant for the UNT responses(c PB0.05, ANOVA, for both groups). The difference in the occur-rence of this kind of incorrect response comparing the two groupswas also significant (@ PB0.05 for UL, and @ PB0.001 for BL,ANOVA). The lesion factor was significant for DP2 responses only inthe VS group (° PB0.001, ANOVA). The occurrence of both typesof incorrect responses was significantly greater in the VS compared tothe DS group (@ PB0.001, ANOVA).
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–6258
Fig. 8. Omission of second lever pressing (NP2). The lesion factorwas significant only in the ventral striatum (VS) group (° PB0.001,ANOVA). This kind of incorrect response was more frequent follow-ing the bilateral denervation in the VS compared to the DS group(@ PB0.05, ANOVA).
of the striatal subregions, to avoid denervation of thelateral striatum.
4.1. The influence of dopaminergic inner6ation onmotor performance
The strategy of making a first injection of 6-OHDAipsilaterally to the side that the lever was to be pressedat trigger presentation was adopted to ascertainwhether a deficit in initiating response directed con-tralaterally to the dopamine-denervated side developed[15,16,22]. Such an impairment did not occur in thedorsal striatum group as the percentage of correctresponses was stable following the lesion made in thecontralateral side. The slight decrease of correct re-sponses caused by bilateral ventral striatum denerva-tion, rather than to a spatial deficit, seems to be due toa complex disruption of motor preparation and execu-tion, discussed below. On the other hand, visual stimulipresented to both eyes, and associated with differenttones, further assured that both hemispheres were in-formed of the occurrence of behavioral stimuli. Thus, itis unlikely that sensory deficits in detection of stimuli,as well as in spatial orientation, could have influencedthe data.
The independence of the behavioral response fromthe teleceptive nature of the sensory signal triggeringmovement is persuasive evidence that animals weredriven by the behavioral significance of the signal,rather than by its mere sensory nature. This is inagreement with the polymodal character of striatalneurons which responded to presentation of sensorysignals in monkeys in situations only in which animalswere engaged in a behavioral context [50].
Neither the uni- nor the bilateral lesions causedimpairment of motor strategy since the animals contin-ued to use the same limb to hold and press levers as inthe preoperative period. Denervated animals were stillable to sequentially handle levers and exhibited nodeficit in the retention of sequence and task require-ments. This lack of motor effect is in agreement withthe scanty number of afferents that the dorsal andventral areas of the striatum receive from the sensori-motor cortex [41]. Furthermore, the lateral striatum,where motor representation of body, paws and mouthis located in the rat [21,24,64], was spared followingdenervation. This would explain why deleterious effects,as those reported by others authors, in limbs used forskilled reaching or bar-pressing movements[11,12,30,43,51,61], and in the mouth use [37,54], werenot observed.
The delay-dependent shortening of reaction time,observed in the present experiments, is in accordancewith a previous report by Brown and Robbins [16]using shorter delays in rats performing externally-in-structed head movements, and by Baunez et al. [12] in
3.6.4. Incorrect responses on the second le6erIncorrect responses on the second lever occurred
when the animals neglected the lever by directly goingto the food tray (no press of the second lever, NP2 inFig. 8), despite a correct response to the trigger signal.The analysis of this kind of incorrect responses indi-cated that they were significantly affected only by thebilateral denervation of the ventral striatum [PB0.05vs dorsal striatum, Newman–Keuls, F(1,41)=11.1].There was a sharp increase of NP2 responses followingeither denervations [PB0.05, Newman–Keuls,F(2,76)=10.5].
4. Discussion
The behavioral paradigm specifically designed forthis study was planned to analyze the preparation andexecution of a rewarded motor act following focaldopaminergic denervation of either the dorsal or ven-tral striatum. This paradigm was an analogue of thatpreviously employed in monkeys [10,60] being charac-terized by a sequence of events including an instructionsignal, a trigger signal and a reward, so that each signalcould serve as a predictor for the subsequent event. Theparadigm required a lever to be held while the succes-sive trigger-induced movement was preceded by a longpreparation period, assuring a sustained expectation ofthe forthcoming reward-related signal. In this state ofexpectation, the animals were also required to hold thelever according to the memorized instruction signal.Thus, various functions extending beyond a merepreparation to move were involved.
Since there is compelling evidence of a major involve-ment of regional striatal dopamine in instrumental leverpressing, appetitive behaviors and in sensorimotor inte-gration [4,23,37,38,40,56,57], great attention was paidto restricting the 6-OHDA-induced denervation in each
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–62 59
rats releasing a lever at the extinction of a light condi-tioning stimulus. According to Brown and Robbins[16], the shortening of the reaction time should indicatea progressive motor readiness as the delay of an im-pending signal triggering movement elapses. The aboli-tion of the reaction time shortening, caused by thedenervation of the ventral striatum and the net increaseof the reaction time, as well as of the movement time,observed in the same group point to a general slowingof motor performance. Support for this conclusioncomes from other studies on the ventral striatum, usingdifferent behavioral paradigms, which reported a widerange of deficits including reduction of lever pressingand slowing of interresponse time [53,62], slowing ofinitial response rate [40], attenuation of both speed andimpulsivity of responding in a five-choice serial reactiontime task [22], and attenuation of enhanced respondingproduced by intra-accumbens d-amphetamine [63]. In-terference with appetitively motivated instrumental re-sponding [24,52,55,56] is generally claimed to beresponsible for the general reduction of motor execu-tion in these studies. However, greater insight may beobtained by detailed analysis of incorrect responsespresented in the following section.
4.2. The nature of incorrect responses occurring in thepreparation and execution of mo6ement
The main effect of dorsal striatum lesions was adifficulty to conditionally respond to the instructionsignal initiating the behavioral paradigm. In nearly 20%of trials, the animals failed to react appropriately tothis stimulus and responded with natural behavior,such as grooming or rearing movements. It was appar-ent that these animals had a limited capability toconditionally respond to the behavioral signal, but weredriven by internally-derived rather than externally-cuedbehavior, as suggested by grooming and rearing move-ments. However, if animals responded correctly to thepresentation of the instruction signal, performance wascorrected in the majority of trials. This observation iscompatible with the concept of ‘attention to action’impairment caused by striatal dopaminergic depletionsformulated by Carli et al. [20]. Impairment of otherfunctions, subserved by the associative territory of thestriatum, such as associative processes and workingmemory could also be involved. The rat striatum con-tains neurons that modulate their discharge at thepresentation of behavioral signals [19,39,64,65]. There-fore, it is possible that reduced responsiveness of theseneurons to conditioned stimuli due to dopamine dener-vation, as observed in the monkey striatum [7], couldrepresent a cellular mechanism to explain the lack ofconditioned response.
In contrast, the effects of ventral striatum denerva-tion caused a large repertoire of incorrect responses,
including timeout pressing or release or, even, omissionof lever pressing. Both groups of animals, however,retained a general readiness to respond to the behav-ioral signals. This demonstrates that the nature ofdeficits arising from dopaminergic denervation of thetwo striatal subregions is different. While the denerva-tion of the dorsal striatum, as discussed above, im-paired conditional response to the instruction signalinitiating the paradigm, denervation of the ventral stria-tum disrupted motor preparation and execution, as wellas the ability of the animals to respond appropriately tothe sensory cues signaling each step of the behavioralparadigm. In particular, the high number of trials inwhich the animals refrained from reacting to the triggersignal, which was directly linked to the reinforcer,suggests that, in addition to a general impairment torespond, the ventral striatum denervation affected thecapability of the animals to appropriately wait for animpending behavioral signal while maintaining thestimulus-reward association. The former possibility issubstantiated by the fact that by increasing the delaypreceding the presentation of the trigger signal, theincorrect responses at the presentation of the triggersignal (UNT responses in Fig. 7A) decreased in com-parison to shorter delays, suggesting that attention tothe incoming event could recover if the animals wereleft an appropriate time to react.
The hypothesis that the inner process of waiting foran impending event could be disrupted in animalssubjected to ventral striatum denervation should beconsistent with the presence in this striatal region ofneurons increasing their discharge up to the presenta-tion of incoming signals directly linked to reward deliv-ery [9,10,60]. These neurons could also have a role inaspects of timing and time perception, a function inwhich the dopaminergic innervation of the striatum isinvolved as a guiding force in time-scheduled operantbehaviors [42]. This function, still poorly investigated inthe basal ganglia system, may help to explain prema-ture as well as delayed pressing or release of levers intime-scheduled behaviors. Once the task, albeit tempo-rally complex, has been learned to the extent that theperformance is automatic, lack of dopamine could im-pair the scheduled behavior and predictability ofevents, thus facilitating the incorrect responses. Thus,given the fact that the rat and monkey are able tomaintain a prolonged expectancy of behavioral events,it would be interesting to investigate whether this be-havioral state in the rat is correlated to a prolongedincrease in striatal neuronal activity, and whether thenature of the expected reward influences this activity.
4.3. Comparison with other studies
Contrary to present results, previous studies havereported increased reaction time following 6-OHDA, or
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–6260
excitotoxic lesion of the dorsal striatum, and abolitionof the delay-dependent speeding of reaction time indifferent behavioral paradigms [2,4,5,11,16]. Discrepan-cies still exist about the increase of reaction time afterventral striatum lesions, reported in this and otherpapers [23,53], in comparison to the results which arenot in favour of such an effect [2,14].
A number of factors may help to explain some ofthese discrepancies. Firstly, the extent of denervation inmany of the above studies varied greatly. For instance,following a 60–90% dopamine depletion in the caudatenucleus, a 40–50% depletion of dopamine in the ventralstriatum was also present [16,20], thus suggesting thatthe lesion extended abundantly beyond the targetedarea. In other studies, the denervation reached thelateral striatum thus impairing the use of limbs forskilled reaching or bar-pressing movements[4,5,11,12,26,32,47,61]. Secondly, the different timingmechanisms and task contingencies used could havedifferently engaged the animals attention to move.Thirdly, reaction time depends not only on sensoryintegration, but also on the ability to formulate, orselect, central motor programs to appropriately startand execute the conditioned movements. Hence, it islikely that the lesions could have differently affected thecentral mechanisms leading to the activation of thedifferent muscle groups involved in the variety of move-ments employed, as simple forelimb release movements[4,5,11,12], complex sequential forelimb release-pressmovements (this study), head withdraw movements[15–17,47], and locomotion [31,32].
In addition, as recently pointed out by Brasted et al.[13], the patterns of reaction time deficit after striatallesions are not simply attributable to procedural differ-ences in the lesions, training conditions and taskparameters, but also critically depend on the apparatusused and the precise response requirements for eachtask.
In conclusion, the present data are consistent withthe indications that the correct execution of a reactiontime task depends not only on the striatal dopaminergicinnervation [46], but also on the complexity of the task,and provide further support in favor of different rolesof dopaminergic innervation of the dorsal and ventralstriatum in appropriate planning and adapting goal-di-rected behaviors to the requirements of behavioralparadigms.
Acknowledgements
This study was supported by contracts BMH4-CT95-0608 (Biomed 2), CHRX-CT94-0463 (Human Capitaland Mobility) of the European Community and bygrants from Ministero dell’Universita e della RicercaScientifica e Tecnologica (40–60%) and CNR
(97.04506.CT04). The authors wish to thank Dr An-drew McKay, Dr Howard Casey Cromwell and DrMario Giunta for their helpful comments on themanuscript.
References
[1] Abrous DN. Dopamine-rich grafts in the neostriatum and/ornucleus accumbens effects on drug-induced behaviors and skilledpaw-reaching. Neuroscience 1993;53:187–97.
[2] Amalric M, Koob GF. Depletion of dopamine in the caudatenucleus but not in nucleus accumbens impairs reaction-timeperformance in rats. J Neurosci 1987;7:2129–34.
[3] Amalric M, Conde H, Dormont JF, Farin D, Schmied A.Activity of caudate neurons in cat performing a reaction timetask. Neurosci Lett 1984;49:253–8.
[4] Amalric M, Baunez C, Nieoullon A. Does the blockade ofexcitatory amino acid transmission in the basal ganglia simplyreverse reaction time deficits induced by dopamine inactivation?Behav Pharmacol 1995;6:508–19.
[5] Amalric M, Moukhles H, Nieoullon A, Daszuta A. Complexdeficits on reaction time performance following bilateral intras-triatal 6-OHDA infusion in the rat. Eur J Neurosci 1995;7:972–80.
[6] Annett LE, McGregor A, Robbins TW. The effects of ibotenicacid lesions of the nucleus accumbens on spatial learning andextinction in the rat. Behav Brain Res 1989;31:231–41.
[7] Aosaki T, Graybiel AM, Kimura M. Effect of the nigrostriataldopamine system on acquired neural responses in the striatum ofbehaving monkeys. Science 1994;265:412–5.
[8] Apicella P, Ljungberg T, Scarnati E, Schultz W. Responses toreward in monkey dorsal and ventral striatum. Exp Brain Res1991;85:491–500.
[9] Apicella P, Scarnati E, Schultz W. Tonically discharging neuronsof monkey striatum respond to preparatory and rewarding stim-uli. Exp Brain Res 1991;84:672–5.
[10] Apicella P, Scarnati E, Ljiungberg T, Schultz W. Neuronalactivity in monkey striatum related to expectancy of predictableenvironmental events. J Neurophysiol 1992;68:945–60.
[11] Baunez C, Nieoullon A, Amalric M. Dopamine and complexsensorimotor integration: further studies in a conditioned motortask in the rat. Neuroscience 1995;65:375–84.
[12] Baunez C, Nieoullon A, Amalric M. In a rat model of parkin-sonism, lesions of the subthalamic nucleus reverse increases ofreaction time but induce a dramatic premature respondingdeficit. J Neurosci 1995;15:6531–41.
[13] Brasted PJ, Dobrossy MD, Robbins TW, Dunnett SB. Striatallesions produce distinctive impairments in reaction time perfor-mance in two different operant chambers. Brain Res Bull1998;46:487–93.
[14] Brown VJ, Bowman EM. Discriminative cues indicating rewardmagnitude continue to determine reaction time of rats followinglesions of the nucleus accumbens. Eur J Neurosci 1995;12:2479–85.
[15] Brown VJ, Robbins TW. Deficits in response space followingunilateral striatal dopamine depletion in the rat. J Neurosci1989;9:983–9.
[16] Brown VJ, Robbins TW. Simple and choice reaction time perfor-mance following unilateral striatal dopamine depletion in the rat.Brain 1991;114:513–25.
[17] Cador M, Robbins TW, Everitt BJ. Involvement of the amyg-dala in stimulus-reward associations: interaction with the ventralstriatum. Neuroscience 1989;30:77–86.
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–62 61
[18] Carelli RM, West MO. Representation of the body by singleneurons in the dorsolateral striatum of the awake, unrestrainedrat. J Comp Neurol 1991;309:231–49.
[19] Carelli RM, Wolske M, West MO. Loss of lever press-relatedfiring of rat striatal forelimb neurons after repeated sessions in alever pressing task. J Neurosci 1997;17:1804–14.
[20] Carli M, Jones GH, Robbins TW. Effects of unilateral dorsaland ventral striatal dopamine depletion on visual neglect in therat: a neural and behavioral analysis. Neuroscience 1989;29:309–27.
[21] Cho J, West MO. Distributions of single neurons related to bodyparts in the lateral striatum of the rat. Brain Res 1997;756:241–6.
[22] Cole BJ, Robbins TW. Effects of 6-hydroxydopamine lesions ofthe nucleus accumbens septi on performance of a 5-choice serialreaction time tasks in rats: implication for theories of selectiveattention and arousal. Behav Brain Res 1989;33:165–79.
[23] Cousins MS, Salamone JD. Involvement of ventrolateral striataldopamine in movement initiation and execution: a microdialysisand behavioral investigations. Neuroscience 1996;70:849–59.
[24] 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.
[25] Evarts EV, Teravainen H, Calne H. Reaction time in Parkin-son’s disease. Brain 1981;99:167–86.
[26] Evenden JL, Robbins TW. Effects of unilateral 6-hydroxydo-pamine lesions of the caudate-putamen on skilled forepaw use inthe rat. Behav Brain Res 1984;14:61–8.
[27] Fairley PC, Marshall JF. Dopamine in the lateral caudate-puta-men of the rat is essential for somatosensory orientation. BehavNeurosci 1986;100:652–63.
[28] Furtado JCS, Mazurek MF. Behavioral characterization ofquinolinate-induced lesions of the medial striatum: relevance forHuntington’s disease. Exp Neurol 1996;138:158–68.
[29] Gardiner TW, Kitai ST. Single-unit activity in the globus pal-lidus and neostriatum of the rat during performance of a trainedhead movement. Exp Brain Res 1992;88:517–30.
[30] Hamilton MH, Garcia-Munoz M, Arbuthnott GW. Separationof the motor consequences from other actions of unilateral6-hydroxydopamine lesions in the nigrostriatal neurones of ratbrain. Brain Res 1985;348:220–8.
[31] Hauber W. Impairments of movement initiation and executioninduced by a blockade of dopamine D1 or D2 receptors arereversed by a blockade of N-methyl-d-aspartate receptors. Neu-roscience 1996;73:121–30.
[32] Hauber W, Schmidt WJ. Differential effects of lesions of thedorsomedial and dorsolateral caudate-putamen on reaction timeperformance in rats. Behav Brain Res 1994;60:211–5.
[33] Hikosaka O, Sakamoto M, Usui S. Functional properties ofmonkey caudate neurons. I. Activities related to saccadic eyemovements. J Neurophysiol 1989;61:780–98.
[34] Hikosaka O, Sakamoto M, Usui S. Functional properties ofmonkey caudate neurons. II. Visual and auditory responses. JNeurophysiol 1989;61:799–813.
[35] Hikosaka O, Sakamoto M, Usui S. Functional properties ofmonkey caudate neurons. III. Activities related to expectation oftarget and reward. J Neurophysiol 1989;61:814–32.
[36] Jahanshahi M, Brown RG, Mrsden CD. Simple and choicereaction time and the use of advance information for motorpreparation in Parkinson’s disease. Brain 1992;115:539–64.
[37] Jicha GA, Salamone JD. Vacous jaw movements and feedingdeficits in rats with ventrolateral striatal dopamine depletion:possible relation to parkinsonian symptoms. J Neurosci1991;11:3822–9.
[38] Kiyatkin EA, Gratton A. Elettrochemical monitoring of extra-cellular dopamine in nucleus accumbens of rats lever-pressing forfood. Brain Res 1994;652:225–34.
[39] Lavoie AM, Mizumori SJY. Spatial, movement- and reward-sen-sitive discharge by medial ventral striatum neurons of rats. BrainRes 1994;638:157–68.
[40] McCullough LD, Cousins MS, Salamone JD. The role of nu-cleus accumbens dopamine in responding on a continuous rein-forcement operant schedule: a neurochemical and behavioralstudy. Pharmacol Biochem Behav 1993;46:581–6.
[41] McGeorge AJ, Faull RRM. The organization of the projectionfrom the cerebral cortex to the striatum in the rat. Neuroscience1989;29:503–37.
[42] Meck WH. Neuropharmacology of timing and time perception.Cogn Brain Res 1996;3:227–42.
[43] Miklyaeva E, Castaneda E, Whishaw IQ. Skilled reachingdeficits in unilateral dopamine-depleted rats: impairments inmovement and posture and compensatory adjustments. J Neu-rosci 1994;14:7148–58.
[44] Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordi-nates, 4th edition. San Diego, CA: Academic Press, 1998.
[45] Pisa M, Cyr J. Regionally selective roles of the rat’s striatum inmodality-specific discrimination learning and forelimb reaching.Behav Brain Res 1990;37:281–92.
[46] Pullman SL, Watts RL, Juncos JL, Chase TN, Sanes JN.Dopaminergic effects on simple and choice reaction time perfor-mance in Parkinson’s disease. Neurology 1988;38:249–54.
[47] Reading PJ, Dunnett SB, Robbins TW. Dissociable roles ofventral, medial and lateral striatum on the acquisition andperformance of a complex visual stimulus-response habit. BehavBrain Res 1991;45:147–61.
[48] Robbins TW, Everitt BJ. Neurobehavioral mechanisms of re-ward and motivation. Curr Opin Neurobiol 1996;6:228–36.
[49] Rolls ET, Thorpe SJ, Maddison SP. Responses of striatal neu-rons in the behaving monkey. I. Head of the caudate nucleus.Behav Brain Res 1983;7:179–210.
[50] Romo R, Scarnati E, Schultz W. Role of primate basal gangliaand frontal cortex in the internal generation of movements. II.Movement-related activity in the anterior striatum. Exp BrainRes 1992;91:385–95.
[51] Sabol KE, Neil DB, Wages SA, Church WH, Justice JB. Do-pamine depletion in a striatal subregion disrupts performance ofa skilled motor task in the rat. Brain Res 1985;335:108–20.
[52] Salamone JD. Complex motor and sensorimotor functions ofstriatal and accumbens dopamine: involvement in instrumentalbehavior processes. Psychopharmacology 1992;107:160–74.
[53] Salamone JD, Kurth PA, McCullough D, Sokolowski JD,Cousins MS. The role of brain dopamine in response initiation:effects of haloperidol and regionally specific dopamine depletionson the local rate of instrumental responding. Brain Res1993;628:218–26.
[54] Salamone JD, Mahan K, Rogers S. Ventrolateral striatal do-pamine depletions impair feeding and food handling in rats.Pharmacol Biochem Behav 1993;44:605–10.
[55] Salamone JD, Cousins MS, McCullough LD, Carriero DL,Berkowitz RJ. Nucleus accumbens dopamine release during in-strumental lever pressing for food but not free food consump-tion. Pharmacol Biochem Behav 1994;49:25–31.
[56] Salamone JD, Cousins MS, Snyder BJ. Behavioral functions ofnucleus accumbens dopamine: empirical and conceptual prob-lems with the anhedonia hypothesis. Neurosci Biobehav Rev1997;21:341–59.
[57] Schultz W. Dopamine neurons and their role in reward mecha-nisms. Curr Opin Neurobiol 1997;7:191–7.
[58] Schultz W, Romo R. Role of primate basal ganglia and frontalcortex in the internal generation of movements. I. Preparatoryactivity in the anterior striatum. Exp Brain Res 1992;91:363–84.
[59] Schultz W, Studer A, Romo R, Sundstrom E, Jonsson G,Scarnati E. Deficits in reaction times and movement times ascorrelates of hypokinesia in monkeys with MPTP-induced stri-atal dopamine depletion. J Neurophysiol 1989;61:651–68.
T. Florio et al. / Beha6ioural Brain Research 104 (1999) 51–6262
[60] Schultz W, Apicella P, Scarnati E, Ljungberg T. Neuronalactivity in monkey ventral striatum related to the expectation ofreward. J Neurosci 1992;12:4595–610.
[61] Siegfried B, Bures J. Handedness in rats: blockade of reachingbehavior by unilateral 6-OHDA injections into substantia nigraand caudate nucleus. Physiol Psychol 1980;8:360–8.
[62] Sokolowski JD, Salamone JD. The role of accumbens dopaminein lever pressing and response allocations: effects of 6-OHDAinjected into core and dorsomedial shell. Pharmacol BiochemBehav 1998;59:557–66.
[63] Taylor JR, Robbins TW. 6-Hydroxydopamine lesions of the
nucleus accumbens but not of the caudate nucleus, attenuateenhanced responding with reward-related stimuli produced byintra-accumbens d-amphetamine. Psycopharmacology1986;90:390–7.
[64] West MO, Carelli RM, Pomerantz M, Cohen SM, Gardner JP,Chapin JF, Woodward DJ. A region in the dorsolateral striatumof the rat exhibiting single-unit correlations with specific locomo-tor limb movements. J Neurophysiol 1990;64:1233–46.
[65] White IM, Rebec GV. Responses of rat striatal neurons duringperformance of a lever-release version of the conditioned avoid-ance response task. Brain Res 1993;616:71–82.
