An examination of d-amphetamine self-administration in pedunculopontine tegmental nucleus-lesioned rats

Neuroscience 125 (2004) 349 –358

AN EXAMINATION OF d-AMPHETAMINE SELF-ADMINISTRATION IN PEDUNCULOPONTINE TEGMENTAL NUCLEUS-LESIONED RATS H. L. ALDERSON,a* M. P. LATIMER,a C. D. BLAHA,b A. G. PHILLIPSc AND P. WINNa

The pedunculopontine tegmental nucleus (PPTg) is a structure at the junction of the pons and midbrain that has been suggested to have a role in reward-related behaviours, especially with regard to the integration of rewardrelated input with behavioural outputs (Inglis and Winn, 1995; Winn et al., 1997). Neurons in the PPTg respond to both reward and stimuli predictive of it (Dormont et al., 1998) and, although all PPTg cholinergic neurons project to the thalamus (Oakman et al., 1995), it is the connexions that the PPTg has with both dorsal and ventral striatum, that are most implicated in reward-related behaviour. The PPTg sends efferent projections containing acetylcholine (ACh), at least some of which also contain and release glutamate (Lavoie and Parent, 1994), to the midbrain dopamine (DA) neurons of the substantia nigra pars compacta and, from its caudal aspects, to those of the ventral tegmental area. It also sends efferents to the subthalamic nucleus and the entopeduncular nucleus, the internal segment of rodent globus pallidus, and the lateral hypothalamus (Jackson and Crossman, 1983; Rye et al., 1987; Gould et al., 1989; Bevan and Bolam, 1995; Oakman et al., 1995). In turn, the PPTg is innervated by many elements of corticostriatal circuitry, including subthalamic nucleus, substantia nigra pars reticulata, ventrolateral caudate putamen and globus pallidus (Spann and Grofova, 1991; Moriizumi and Hattori, 1992; Semba and Fibiger, 1992; Grofova and Zhou, 1998). In terms of ventral striatal connexions, the nucleus accumbens shell and rostral pole project to the PPTg via the ventral pallidum and lateral hypothalamus, whilst the core can influence the PPTg through connexions with basal ganglia structures; PPTg also receives afferent projections from structures functionally associated with the ventral striatum, such as the central nucleus of the amygdala and bed nucleus of the stria terminalis (Heimer at al., 1991; Groenewegen et al., 1993; Semba and Fibiger, 1992; Steininger et al., 1992; Zahm et al., 1999, 2001). Several studies have supported the idea of a functional role for the PPTg in reward-related behaviour, and there is particular interest in its potential role in drug reward. A number of studies have examined the involvement of specific neurotransmitter receptors within the PPTg in drug reward using i.v. self-administration (IVSA) techniques. Microinjection of the ␮-opioid agonist DAMGO into the PPTg reduces cocaine IVSA (Corrigall et al., 1999, 2002) and nicotine IVSA is attenuated by intra-PPTg microinjection of the nicotinic ACh receptor antagonist dihydro-␤-erythrodine (Corrigall et al., 2002), and both nicotine and cocaine IVSA under a fixed ratio (FR) schedule are reduced by intra-PPTg microinjection

a School of Psychology, University of St. Andrews, St Mary’s Quad, St. Andrews, Fife, KY16 9JP, UK b Department of Psychology, University of Memphis, Memphis, TN 38152, USA c

Department of Psychiatry, University of British Columbia, 2C1 Detwiller Pavilion, 2255 Westbrook Mall, Vancouver, BC, Canada V6T 2A1

Abstract—The pedunculopontine tegmental nucleus (PPTg) has long been suggested to have a role in reward-related behaviour, and there is particular interest in its possible role in drug reward systems. Previous work found increased i.v. self-administration (IVSA) of d-amphetamine following PPTg lesions when training had included both operant pre-training and priming injections. The present study examined the effect of excitotoxin lesions of the PPTg on d-amphetamine IVSA under three training conditions. Naive: no previous experience of d-amphetamine or operant responding. Pretrained: given operant training with food before lesion surgery took place. Primed: given single non-contingent d-amphetamine infusion (0.1 mg/0.l ml) at the start of each session. Rats in all conditions were given either ibotenate or phosphate buffer control lesions of the PPTg before d-amphetamine (0.1 mg/0.1 ml infusion) IVSA training took place. Rats received eight sessions of training under a fixed ratio (FR2) schedule of d-amphetamine IVSA, followed by four sessions under a progressive ratio (PR5) schedule. In the naive condition, PPTg-lesioned rats were attenuated in their responding under FR2, and took significantly fewer infusions under PR5 than the control group. Under FR2 in the pretrained condition, there was no difference between PPTg excitotoxin and control lesioned rats; however, PPTglesioned rats took significantly fewer infusions under the PR5 schedule. In the primed condition, there were no differences between PPTg-lesioned and control rats under either FR2 or PR5 schedules. These data demonstrate that operant training prior to PPTg lesion surgery corrects some, but not all, of the deficits seen in the naive condition. PPTg-lesioned rats in both naive and pre-trained conditions showed reduced responding for d-amphetamine under a PR5 schedule. These deficits are overcome by priming with d-amphetamine. We suggest that alterations in striatal dopamine activity following PPTg lesions underlie these effects. © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: acetylcholine, dopamine, mesopontine, striatum, reward, associative learning. *Corresponding author. Tel: ⫹44-1334-462085; fax: ⫹44-1334-463042. E-mail address: [email protected] (H. L. Alderson). Abbreviations: ACh, acetylcholine; DA, dopamine; FR, fixed ratio; IVSA, i.v. self-administration; NADPH, nicotinamide adenine dinucleotide phosphate; NOS, nitric oxide synthase; PPTg, pedunculopontine tegmental nucleus; PR, progressive ratio.

0306-4522/04$30.00⫹0.00 © 2004 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2004.02.015

349

350

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358

of GABA agonists (Corrigall et al., 2001). However, ibotenic acid lesions of the dorsal portion of PPTg have been reported not to affect nicotine IVSA (Corrigall et al., 1994) though the acquisition of heroin IVSA is impaired following excitotoxin lesions of the PPTg (Olmstead et al., 1998). PPTg lesions also impair conditioned place preference reinforced by d-amphetamine (Bechara and van der Kooy, 1989; Olmstead and Franklin, 1994), but not cocaine (Parker and van der Kooy, 1995). These conflicting data could be taken to suggest that the possible involvement of the PPTg in drug reward might be limited to specific types of drugs. Alternatively, previous experience with drugs or pre-lesion training might also influence the effects that PPTg lesions have on drugrelated behaviours. For example, conditioned place preference to opiates was impaired by excitotoxin lesions of the PPTg in rats which had no prior exposure to opiates, but not in opiate-dependent rats (Bechara and van der Kooy, 1989). In the case of heroin IVSA, if PPTg lesions were made after IVSA training, post-surgery IVSA responding for heroin was not affected (Olmstead et al., 1998). It is also worth noting that PPTg lesions failed to change nicotine IVSA in rats trained before lesion surgery to lever press for nicotine (Corrigall et al., 1994). These results indicate that the PPTg may only play a role in drug-related learning under specific conditions. The PPTg exerts a strong influence over midbrain DA neurons (Blaha and Winn, 1993; Blaha et al., 1996) and therefore an effect of PPTg lesions on d-amphetamine IVSA is predictable: given that the reinforcing effect of d-amphetamine is linked to release of DA. Rats bearing excitotoxin lesions of the PPTg showed an increased break point, relative to that shown by sham-lesioned controls, when undergoing d-amphetamine IVSA under a PR schedule (Keating et al., 1997). However, these rats were pre-trained with food (establishing a lever-reinforcement association) before being given the opportunity to selfadminister the drug, and were also given non-contingent experimenter-administered priming infusions of d-amphetamine at the start of each self-administration session. In the present study, we examined d-amphetamine selfadministration in PPTg-lesioned rats which had received no previous operant training, and investigated the possible contribution of pre-training and priming to these previous effects. In experiment 1, experimentally naive rats, bearing either excitotoxin or sham lesions of the PPTg, were trained in d-amphetamine IVSA to examine the acquisition of operant responding in such rats. In experiment 2, rats were given operant training for food prior to lesion surgery, after which they were given the opportunity to selfadminister d-amphetamine. Experiment 3 investigated the effect of a non-contingent priming injection of damphetamine on the acquisition of d-amphetamine IVSA in PPTg excitotoxin and sham-lesioned rats which had not received any prior operant training for food.

EXPERIMENTAL PROCEDURES Male Lister hooded rats (Charles River Ltd., Margate, UK) were used and were housed individually in cages at 20 –22 °C, with

lights on 7:00 a.m.–7:00 p.m. During behavioural testing, rats were fed 18 –20 g laboratory chow daily, in order to maintain body weight. Water was available ad libitum in the home cage at all times. All testing was conducted during the light phase. Compliance was ensured with national (Animals [Scientific Procedures] Act, 1986) and international (European Communities Council Directive of 24 November 1986 [86/609/EEC]) legislation governing the maintenance of laboratory animals and their use in scientific experiments. All efforts were made to minimize the number of animals used in this study and to avoid unnecessary suffering.

Surgery Rats were anaesthetized with 1.0 ml/kg “Sagatal” (May and Baker; sodium pentobarbitone, 60 mg/ml) diluted 50:50 with sterile water. Infusions of ibotenate (Tocris-Cookson Ltd, Bristol, UK; made up as 0.12 M solution in phosphate buffer [pH 7.4]; final pH adjusted to pH 7.0 using 2 M NaOH) were delivered in a volume of 0.2 ␮l (24 nmol) to each site in 0.01 ␮l steps at 10-s intervals (total injection time: 200 s) using a 0.5 ␮l syringe (Scientific Glass Engineering, Milton Keynes, UK). The injection needles were left in situ for 300 s after the infusion to allow for diffusion away from the tip. Control rats received the same volume of phosphate buffer only, delivered at the same rate as ibotenate. Two injections were made in each hemisphere at the following stereotaxic coordinates: ⫺0.8 mm anterior to the interaural line, ⫾1.6 mm from midline and 7.0 mm below skull surface (posterior PPTg); and 1.5 mm anterior to the interaural line, ⫾1.7 mm from midline and 7.8 mm below skull surface (anterior PPTg): all with level skull (determined by the DV co-ordinates at bregma and lambda). Bilateral lesions were made and rats received two separate unilateral operations separated by a minimum of 7 days. A minimum of 2 weeks was given for recovery following the second PPTg surgery before i.v. catheterisation surgery. Under “Sagatal” anesthesia, rats were implanted with an i.v. jugular catheter, which was anchored dorsally between the scapulae (see Caine et al., 1993 for detailed methods). For the first 5 days post-surgery, catheters were flushed once a day with 0.1 ml of antibiotic solution (0.66 g Timentin/1 ml saline; Beecham Research, Welwyn, Herts, UK) and then daily with 0.1 ml heparinised saline solution (30 units/ml; CP Pharmaceuticals Ltd, Wrexham, UK) to maintain catheter patency.

Apparatus Behavioural training was carried out in four Med-Associates operant boxes (Med-Associates, St Albans, VT, USA) which contained testing chambers measuring 30 cm (L)⫻22.5 cm (W)⫻28 cm (D), equipped with both an IVSA arm attached to an external pump (Med-Associates) and a dispenser for food pellets. The operant boxes contained two levers 11 cm apart on one wall of the testing chamber, between which was the food hopper. The levers were 5 cm wide, and protruded 3 cm into the chamber. Seven centimeters above each lever was a light (1.5 W, 28 V) which could be illuminated when appropriate, and the chambers were illuminated by a house-light (1.5 W, 28 V) on the opposite wall. The IVSA arm was removed from the chamber during foodreinforced testing. The operant boxes were controlled by a PC running the Med-PC for Windows operating system.

Behavioural procedures Experiment 1: acquisition of d-amphetamine self-administration by naive rats. A group of 20 rats, with an average body weight of 313 g (290 –348 g), received either ibotenate lesions of the PPTg (n⫽12) or phosphate buffer control lesions (n⫽8), followed by implantation of an i.v. catheter, as described above. Following 1 week of recovery from i.v. catheterisation surgery, rats began d-amphetamine IVSA training. The dose of

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358 d-amphetamine (Sigma-Aldrich Ltd; weight given as sulphate salt in 0.9% physiological saline) used for IVSA was 0.1 mg/ 0.1 ml infusion delivered over 5.46 s. Responding on the active lever, according to the schedule conditions, resulted in an infusion of d-amphetamine, accompanied by illumination of a stimulus light (20 s) above the active lever and retraction of both active and inactive levers for 20 s. Responses on the inactive lever were recorded, but had no programmed consequences. The left or right lever was designated as being active according to rat number. Rats received 8 days of training under a schedule, in which an infusion was given for every two responses on the active lever (FR2). Following this, rats were trained for 4 days under a progressive ratio schedule (PR5). Under this schedule, the first infusion was delivered after two responses on the active lever, and from then onward, the schedule requirement was incremented by five with every infusion. The session length for both FR2 and PR5 training was 3 h, and the maximum number of infusions which could be received during this time was limited to 20. Experiment 2: acquisition of d-amphetamine self-administration following lever-training for food reward. Thirty-three rats, with an average body weight of 327 g (298 –365 g), received lever-training for food reward prior to surgery. On the first day of training, 20 food pellets (45 mg; Noyes A/1 formula; PJ Noyes Company Inc., Lancaster, NH, USA) were placed in the food hopper, and rats were given a 30 min session in the operant chamber in which to eat them. The house light was illuminated, and the levers present, but responding on them had no consequences. For the next 2 days, rats were trained to respond for food pellets under an FR1 schedule. Delivery of the pellet was contingent on a single response on the active lever, and was accompanied by 20 s illumination of the stimulus light above the active lever; both levers then retracted for 20 s. Next, rats underwent 3 days of training under an FR2 schedule. All sessions were 30 min in length, and a maximum of 20 food pellets was available. Between 1 and 3 days after the completion of training, rats underwent surgery and received either ibotenate lesions of the PPTg (n⫽23) or phosphate buffer control lesions (n⫽10), followed by implantation of an i.v. catheter. Approximately 1 week after i.v. catheterisation surgery, training was resumed, beginning with a single 30 min session of food training under an FR1 schedule, with a maximum of 20 food pellets being available. Following this, rats proceeded with d-amphetamine IVSA training under FR2 then PR5, as described for experiment 1. The lever designated as active was constant through all phases of the experiment. Experiment 3: acquisition of primed d-amphetamine self-administration. A total of 24 rats, with a mean body weight of 372 g (330 – 413 g) began this experiment. Sixteen received excitotoxin lesions of the PPTg, and eight received sham lesions. IVSA training was as previously described in experiment 1, with the exception that on entering the self-administration chamber, the houselight was illuminated but no levers were present. Rats received a single experimenter administered “priming” injection of d-amphetamine (0.1 mg/infusion) before the start of each IVSA session. No CS light accompanied the priming injection. Following the priming injection, there was a 20-s time-out before the levers extended into the box, and the session proceeded as detailed in experiment 1. Rats were tested under both FR2 and PR5 as described above.

Histological analysis At the end of behavioural testing, rats were humanely killed by an overdose of barbiturate (“Dolethal”; Univet Ltd., Bicester, UK; sodium pentobarbitone, 200 mg/ml) and perfused transcardially with phosphate-buffered saline followed by 4% paraformaldehyde in 0.1 M phosphate buffer. Brains were removed and stored in paraformaldehyde prior to histological analysis. Sections of 50 ␮m

351

were cut on a freezing microtome. One in four sections was stained with Cresyl Violet and one in four was processed for nicotinamide adenine dinucleotide phosphate (NADPH) diaphorase histochemistry, using a modification of the method of Vincent and Kimura (1992). NADPH diaphorase is a NO synthase (NOS); in this part of the CNS, NOS marks almost exclusively cholinergic neurons. Previous studies have reported a strong correlation between counts of cholinergic neuronal number made using sections processed for NADPH diaphorase histochemistry and those processed to show choline acetyltransferase immunohistochemistry (Inglis et al., 1994; Vincent and Kimura, 1992). Individual sections were examined using a Leitz “Diaplan” microscope fitted with a Sony DXC-3000P video camera for visualisation of sections on a high resolution colour monitor.

Statistical analysis The measures taken during all experiments were the number of responses made on both the active and inactive levers and the number of infusions received. Break point under the PR schedule was defined as the final ratio requirement completed, that is, the final ratio under which an infusion was received. In the PR analyses for each experiment, both break point and number of infusions are reported: infusion number is reported to maintain comparability across phases of these experiments and break point is reported because this is the most common output measure in PR studies. Note that within each experiment, ANOVA outcomes are the same for break point and number of infusions analyses. Inactive lever responses made during responding under the PR schedule were expressed as a percentage of the total number of lever responses made (i.e. active and inactive responses together). Analysis was by repeated-measures ANOVA (Winer, 1971) using Statistica 5.5 (StatSoft Inc., Tulsa, OK, USA). Where a significant interaction was present, it was analysed further using restricted ANOVA or by planned comparisons, as appropriate.

RESULTS Histological analysis Fig. 1 illustrates a representative acceptable PPTg excitotoxin lesion, compared with a sham-lesioned control. Tissue from a sham-lesioned control is shown in panel A, stained for Cresyl Violet, and C, processed for NADPH diaphorase. Panels B and D show tissue from an excitotoxin-lesioned brain stained for Cresyl Violet (B) and NADPH diaphorase (D). Only data from rats that were found to have complete bilateral lesions of the PPTg were included in statistical analyses. Acceptable lesions were those involving almost complete bilateral loss of PPTg neurons, without damage to the laterodorsal tegmental nucleus. Lesions in which there was extensive bilateral damage outside the PPTg (in the deep mesencephalic or cuneiform nuclei, or the pontine reticular formation for example) were excluded, as were those in which the whole PPTg was not lesioned. A total of five rats were discounted from statistical analysis because of either incomplete lesions or damage extending bilaterally outside the PPTg. Of these, two rats were excluded from experiment 1, two from experiment 2, and one from experiment 3. Behavioural analysis Experiment 1: Acquisition of d-amphetamine IVSA by naive rats. Fig. 2A shows the number of d-amphetamine infusions received by PPTg-lesioned and control rats un-

352

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358

Fig. 1. Representative PPTg control and excitotoxin lesions. Panels A and C show tissue from a sham lesioned control rat, stained with Cresyl Violet and NADPH diaphorase, respectively. Panels B (Cresyl Violet) and D show a representative excitotoxin lesion of the PPTg; the edge of the lesioned area is indicated with arrows. LDTg, laterodorsal tegmental nucleus; scp, superior cerebellar peduncle.

der an FR2 schedule of reinforcement, over eight sessions. There was a significant main effect of session [F(7,70)⫽13.02, P⬍0.01] and a significant lesion⫻session interaction [F(7,70)⫽3.73, P⬍0.01] but no main effect of lesion. Restricted analysis revealed that there was no significant effect of session for the PPTg excitotoxin lesion group, whilst the sham lesion group showed a significant increase in the number of infusions received over sessions [F(7,70)⫽13.90, P⬍0.01]. Temporal patterns of responding did not differ between PPTg excitotoxin and shamlesioned groups. Both groups showed an initial “loading” phase during which a number of infusions were taken over a short time-interval, whilst subsequent infusions were generally evenly spaced over the session. Table 1A shows the number of responses made on the non-reinforced lever by PPTg excitotoxin and sham-lesioned rats: there were no significant main effects of lesion or session, and no interaction. Responding for d-amphetamine under a PR schedule of reinforcement by PPTg excitotoxin and sham-lesioned rats is shown in Fig. 3A: this plots the number of infusions received during four PR sessions. There were significant main effects of lesion [F(1,10)⫽34.49, P⬍0.01] and of session [F(3,30)⫽3.00, P⬍0.05] but no lesion⫻session interaction. Table 2A shows the mean break point reached under this schedule. Again, there were significant main

effects of lesion [F(1,10)⫽34.49, P⬍0.01] and session [F(3,30)⫽3.00, P⬍0.05] but no interaction. Table 3A shows the percentage of total responses that were made on the inactive lever during self-administration under the PR schedule. There was a significant main effect of lesion [F(1,10)⫽5.35, P⬍0.05], with the PPTg excitotoxin lesion group making a significantly higher proportion of their total responses on the inactive control lever compared with controls. There was no significant effect of session, or lesion⫻session interaction. Experiment 2: acquisition of d-amphetamine IVSA following lever-training for food reward There was no significant difference between PPTg excitotoxin and sham-lesioned controls during the single postoperative food training session (data not shown). The mean number of d-amphetamine infusions received under an FR2 schedule of reinforcement over 8 days of acquisition by PPTg excitotoxin and sham-lesioned rats is shown in Fig. 2B. There was no effect of lesion, and no lesion⫻session interaction, but there was a significant main effect of session [F(7,91)⫽2.29, P⬍0.05]. Planned comparisons showed that rats received significantly more infusions on day 1 than on days 2 (P⬍0.05) and 3 (P⬍0.01). Two sham-lesioned control rats and four PPTg excitotoxin-lesioned rats finished sessions before the time

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358

353

made on the non-reinforced lever during responding for d-amphetamine under the FR2 schedule. There was no significant effect of lesion or session on this measure, and neither was there a significant interaction between these. Fig. 3B shows the mean number of d-amphetamine infusions received by PPTg excitotoxin- and sham-lesioned rats under the PR5 schedule. There was a significant main effect of lesion [F(1,13)⫽9.53, P⬍0.01], with PPTg-lesioned rats clearly taking significantly fewer infusions than controls, as was the case in experiment 1. There was no significant effect of session, and no session by lesion interaction. Table 2B shows the mean breaking point reached over 4 days responding under the PR5 schedule. There was a significant main effect of lesion [F(1,13)⫽9.53, P⬍0.01], but no significant effect of session or interaction. Table 3B shows the percentage of total responses made on the inactive lever during self-administration under the PR schedule. There was no significant effect of lesion or session, and no significant lesion⫻session interaction. Experiment 3: acquisition of primed d-amphetamine IVSA

Fig. 2. Mean (⫾S.E.M.) number of d-amphetamine infusions taken under FR2 schedule. The number of sessions is marked along the x axis, and the number of infusions taken up the y axis. The sham lesion control group is represented by the closed circles, whilst the PPTg excitotoxin lesion group is represented by the open circles. (A) Naive condition: PPTg excitotoxin-lesioned rats showed no significant increase in the number of infusions taken over the eight sessions, whilst sham-lesioned controls showed a significant increase in infusions taken; for statistical analysis, see text. (B) Food-trained condition: There was no significant difference in the mean number of infusions taken by PPTg excitotoxin- and sham-lesioned rats. Both groups took significantly more infusions on day 1 than on days 2 (P⬍0.05) and 3 (P⬍0.01). (C) Primed condition: There was no significant difference in the mean number of infusions taken by PPTg excitotoxin- and shamlesioned rats. There was a significant increase over sessions in the number of infusions taken.

allowed had elapsed as a result of having reached the maximum number of available infusions. This took place on between two and five sessions. Response patterns seen were similar to those reported for experiment 1, with no difference in patterns between lesion and control groups. Table 1B shows the mean number of responses

Fig. 2C shows the mean number of d-amphetamine infusions received by PPTg excitotoxin- and control-lesioned rats during eight sessions of acquisition of primed IVSA under an FR2 schedule. There was no significant main effect of lesion, but there was a significant main effect of session [F(7,77)⫽9.29, P⬍0.01]. There was no significant lesion⫻session interaction. The patterns of responding seen did not differ between PPTg excitotoxin-lesioned rats and controls, and were similar to those described for experiment 1. Table 1C shows the mean number of responses on the inactive lever during the eight FR2 sessions. There was no significant effect of lesion, and no lesion by session interaction, but there was a significant main effect of session [F(7,84)⫽4.85, P⬍0.01]. The mean number of infusions received during four sessions of PR d-amphetamine IVSA is shown in Fig. 3C. There were no significant main effects of lesion or session, and there was no significant lesion⫻session interaction. Table 2C shows the mean breaking point reached during PR self-administration. Again, there was no significant main effect of lesion or session, and no significant lesion⫻session interaction. Table 3C shows the percentage of total responses made on the inactive lever during self-administration under the PR schedule. There was no significant effect of lesion or session, and no significant lesion⫻session interaction.

DISCUSSION In studies examining drug IVSA in rats, response changes following brain lesions can, in principle, be accounted for in many different ways. A decrease in responding could reflect a change in the drug’s reinforcing value (its ability to support operant responding) or its rewarding value (a problem of altered incentive salience); alternatively, lesioned rats might have a learning or attentional difficulty, such that

354

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358

Table 1. Mean (⫾S.E.M.) responses on the inactive lever during FR2 acquisitiona

A: Exp. 1 B: Exp. 2 C: Exp. 3

Control PPTg Control PPTg Control PPTg

Session 1

Session 2

Session 3

Session 4

Session 5

Session 6

Session 7

Session 8

28.7⫾26.5 7.0⫾1.3 2.9⫾1.1 15.5⫾6.9 6.4⫾1.8 11.4⫾4.8

9.1⫾4.5 6.8⫾2.0 1.9⫾0.8 12.7⫾5.6 3.6⫾1.0 3.4⫾1.9

1.8⫾0.8 2.3⫾0.8 3.7⫾1.7 12.8⫾7.2 4.4⫾2.7 3.2⫾1.0

7.0⫾5.8 2.2⫾1.3 5.3⫾3.7 9.7⫾4.8 1.3⫾1.1 1.4⫾0.7

0.7⫾0.3 2.8⫾1.0 3.3⫾3.7 12.0⫾6.4 0.3⫾0.4 2.8⫾1.6

1.0⫾0.4 2.3⫾1.3 3.2⫾1.8 11.8⫾6.9 0.1⫾0.1 5.0⫾4.7

0.2⫾0.2 1.2⫾0.5 5.0⫾3.8 8.3⫾7.3 0.6⫾0.4 0.8⫾0.6

0.2⫾0.2 2.2⫾0.8 1.7⫾0.7 8.0⫾6.0 0.3⫾0.2 0.2⫾0.2

a There were no significant differences between PPTg sham-lesioned control and excitotoxin-lesioned groups in responding on the non-reinforced lever during any FR2 d-amphetamine self-administration sessions in experiments (Exp.) 1, 2 or 3.

associative learning processes (the ability to learn that lever pressing leads to drug delivery) are impaired; alternatively again, the rats might appreciate the value of the reward, but no longer be either willing (a failure of motivation) or able to work the levers (a response selection problem-involving, for example, perseveration or impulsivity). The present experiments examining the effects of PPTg lesions on d-amphetamine IVSA show a complex pattern of deficits: in experiment 1, PPTg-lesioned rats naive with respect to the operant chambers showed impairments under both FR2 and PR5 schedules. In experiment 2, learning an association between levers and reinforcement via prior training abolished the deficit under the FR2 but not PR5 schedule. In contrast again, in experiment 3, presentation of a priming dose of 0.1 mg d-amphetamine ameliorated the deficits shown by PPTg-lesioned rats under both the FR2 and PR5 schedules. Methodological considerations Animal models of drug self-administration use a number of different methodologies, and interactions between training method and lesion effect are not often discussed. However, the present data demonstrate that the same lesion can have different effects on acquisition of both FR and PR schedules when different methodologies are used. It is clearly important to consider that in any IVSA study, the outcome can be influenced by training regimen. Many studies make use of operant pre-training using another reinforcer, typically food, in order to train animals to lever press and to discriminate between active and inactive levers, before being given the opportunity to engage in IVSA. It is assumed that this will facilitate acquisition, with a stable baseline level of responding reached more quickly. This can be of importance in extended experiments in which loss of catheter patency is a potential problem. In relation to lesion studies, it is clear from the present data that whether the lesion is made before or after operant pre-training significantly affects outcome. If the lesion is made after pre-training but before IVSA training, as in experiment 2, its effect on operant acquisition per se cannot be tested. This might be advantageous or not, depending on the hypothesis being investigated, but the present data show that pre-training can radically alter the effect of neural manipulations, such as lesions, on IVSA. Another training method is to give a non-contingent priming injection before the start of the session, as in experiment 3, but our data indicate that such a procedure also

can change outcome. Differences in methodology presumably lie behind the failure of any of the experiments reported here to replicate our previous effects (Keating et al., 1997). PPTg and reward perception It is thought that the PPTg might have a role in reward perception. Responses to natural reinforcers by PPTglesioned rats are unchanged in the home cage—they maintain themselves on laboratory chow and water satisfactorily (Inglis et al., 1994)— but there are problems when they are challenged. For example, PPTg-lesioned rats over-consume high concentration sucrose solutions (⬎12%) compared with controls (Olmstead et al., 1999; Alderson et al., 2001; Keating et al., 2002). However, this effect is not accompanied by a change in sucrose-reinforced conditioned place preference, regardless of the concentration of sucrose used (Alderson et al., 2001). This observation of normal place preference suggests strongly that sucrose over-consumption is not caused by changed reward perception, that is, the ability to perceive the appetitive or incentive nature of a stimulus. Moreover, PPTglesioned rats adjust their energy intake when allowed 20% sucrose solution, reducing laboratory chow intake to compensate for their increased energy intake from sucrose (Keating et al., 2002). Under a PR schedule of food reward with 45 mg food pellets as the reinforcer, PPTg-lesioned rats showed a reduction in breaking point, which could suggest a drop in the perceived reward value of the food, as it has been suggested that breaking point is a measure of reward strength (Hodos, 1961). But as the demands of the PR schedule increased and the responding of lesioned rats on the reinforced lever dropped, there was a corresponding increase in responding on the non-reinforced lever, suggesting that the lesioned rats problem was in response control, not in perception of the appetitive nature of a stimulus (Alderson et al., 2002). Taken together, these studies suggest that the perception of the reward value of natural stimuli—food—in PPTg-lesioned rats is normal. In IVSA studies, PPTg-lesioned rats showed a reduction in the break point reached under a PR schedule of heroin IVSA. However, when the lesions were made after heroin IVSA training there was no effect on either FR or PR responding (Olmstead et al., 1998). Likewise, ibotenic acid lesions of the dorsal portion of PPTg were found not to affect nicotine IVSA when rats were trained to respond for nicotine prior to lesion surgery (Corrigall et al., 1994). As

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358 16

Number of infusions

14

A: naive

12 10 8 6 4 2 0 1

2

3

4

Session 16

Number of infusions

14

B: pre-trained

12

8 6 4 2 0 1

2

3

4

Session 16

Number of Infusions

acquire operant behaviour. This is supported by the increased responding on the inactive lever by the lesion group under the PR schedule in experiment 1, but not in experiment 2. The association between inactive lever and absence of reward had been learned by the lesioned rats prior to surgery in experiment 2, but not in experiment 1. Although there was no effect of lesion on inactive lever responses under the FR2 schedule, once responding was under the more complex PR schedule, this learning impairment became apparent. As such, had PPTg lesions simply reduced reward value, a deficit in FR2 responding would have been expected in experiment 2, as well as in experiment 1. Combined with the evidence of the studies above however, the present d-amphetamine IVSA data strongly suggest that the PPTg is not involved in reward perception per se. Rather, the deficits must represent deficiencies in some other process or processes. PPTg and d-amphetamine IVSA

10

14

355

C: primed

Control PPTg

12 10 8 6 4 2 0 1

2

3

4

Session Fig. 3. Mean (⫾S.E.M.) number of d-amphetamine infusions taken under PR schedule. Session number is marked along the x axis, and the number of infusions taken up the y axis. The sham lesion control group is represented by the closed circles, whilst the PPTg excitotoxin lesion group is represented by the open circles. (A) Naive condition: PPTg-lesioned rats took significantly fewer d-amphetamine infusions than sham-lesioned controls during all four sessions. (B) Food-trained condition: PPTg-lesioned rats took significantly fewer d-amphetamine infusions than sham-lesioned controls during all four sessions. (C) Primed condition: There was no significant difference between PPTg excitotoxin- and sham-lesioned rats in the number of infusions taken. Note that baseline levels were lower than in other conditions; this reflects the fact that rats’ drug levels had been artificially raised before self-administration started (Yokel and Pickens, 1974).

with the data from studies using food, these IVSA studies suggest that PPTg lesions did not alter the rewarding effect of either heroin or nicotine, but impaired rats’ ability to

The FR2 data presented in experiments 1 and 2 are relatively straightforward: PPTg-lesioned rats acquired FR2 d-amphetamine IVSA poorly unless they had already made an association between levers and reinforcement. One way to account for these data are to suggest that the critical variable is the degree to which DA neurons are appropriately activated. PPTg neurons make a monosynaptic, excitatory, innervation of DA-containing cells in the substantia nigra pars compacta and, to a lesser extent, the VTA (Oakman et al., 1995); PPTg lesions remove this control over DA neurons (Blaha and Winn, 1993; Blaha et al., 1996). In the naive condition (experiment 1), it can be argued that lack of proper control over DA neurons by the PPTg leads to poor FR acquisition because PPTg-lesioned rats have difficulty associating lever pressing with reinforcement. In experiment 2, lever training prior to IVSA established an association between levers and reward delivery. It is recognised that DA is important for the establishment of associations between action and reinforcement but, once established, the importance of DA is diminished (Schultz, 2002). Whilst this simple “DA and associative learning” hypothesis like this may account for the FR data in experiments 1 and 2, it is perhaps more difficult to account for the deficit under the PR5 schedule. However, recent thinking about DA neuron activity suggests that the relationship between these cells and associative learning reflects their role as probability detectors, firing most in conditions of highest reinforcement uncertainty (Fiorillo et al., 2003). The FR2 condition is relatively predictable: a reinforcement occurs with every two lever presses and so, once the relationship between levers and reinforcement has been learned, DA would have little role to play. Under a PR5 schedule however, the contingencies continuously change, with a different lever pressing requirement for every reinforcement delivery: 5, then 10, then 15 lever presses and so on. Under these conditions, DA neuron activity must be maintained because uncertainty about action-reinforcement relationships persists. Under the PR5 schedule in experiment 2, PPTg-lesioned rats continued to

356

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358

Table 2. Mean (⫾S.E.M.) breaking point under a PR schedulea

A: Exp. 1 B: Exp. 2 C: Exp. 3

Control PPTg Control PPTg Control PPTg

Session 1

Session 2

Session 3

Session 4

40.3⫾3.1 21.2⫾2.4 45.9⫾3.1 28.7⫾2.1 34.5⫾3.9 30.0⫾4.1

42.0⫾1.8 22.0⫾2.6 43.7⫾4.2 28.7⫾2.1 38.3⫾2.8 30.0⫾4.1

45.3⫾3.3 22.8⫾2.4 46.4⫾4.4 30.3⫾2.1 38.3⫾2.5 32.0⫾3.5

47.0⫾4.5 22.0⫾2.2 47.0⫾4.2 33.7⫾2.8 38.3⫾2.1 31.0⫾4.3

a In both experiments (Exp.) 1 and 2, the PPTg excitotoxin lesion group achieved a significantly lower breaking point than the sham-lesioned control group [F(1,10)⫽34.490, P⬍0.01, and F(1,13)⫽9.526, P⬍0.01 respectively]. In experiment 3, there was no significant difference between excitotoxin lesion and control groups. Breaking point was defined as the final ratio requirement completed.

show deficits even though they had learned the associative relationship between levers and reinforcement, as demonstrated by their lack of impairment under FR2; however, they may have been impaired under PR due to an inability to learn the changing requirements with each lever press. DA depleting lesions of the nucleus accumbens produce similar effects to the present results, with responding for food reward unaffected by such lesions at low ratio requirements, but impaired when ratio requirements were increased (Salamone et al., 2001). This supports the suggestion that the present results were due to disruption of normal midbrain DA activity. In the final experiment, rats were naive with regard to lever-training. This would normally lead to impoverished acquisition of responding by the PPTg lesion group as in experiment 1, but the presence of non-contingent d-amphetamine effectively overcame this, enabling adequate performance under both FR2 and PR5 schedules. The lack of impairment in IVSA performance under both schedules following non-contingent d-amphetamine priming points strongly to a relationship between DA systems and the effects of PPTg lesions, but the question as to how a priming dose of 0.1 mg d-amphetamine ameliorates both FR and PR deficits requires resolution. The particular difficulty is that, in strict pharmacological terms at least, the priming dose is exactly what the rats in experiments 1 and 2 received when they earned their first reinforcement. The difference lies in the contingencies: in experiments 1 and 2 d-amphetamine delivery followed lever pressing, but in experiment 3 d-amphetamine preceded it. Presumably, the priming dose of d-amphetamine may have reinstated a degree of striatal DA function in the lesioned rats, normally maintained by PPTg input to the somatodendritic portion of

DA neurons in the midbrain. Without this normal baseline activity, effective action-reinforcement learning could not occur. Consequently, in those conditions in which prior learning had not occurred, and in which learning was still required for effective performance to occur—the FR2 condition in experiment 1 (but not experiment 2) and the PR5 conditions in experiments 1 and 2—PPTg-lesioned rats failed to perform properly. These priming data imply that the PPTg is required to drive DA neurons initially, but once the process is initiated— or substituted for by the priming dose of d-amphetamine—the PPTg is not required unless contingencies change. Such an hypothesis is consistent with a variety of neurochemical and behavioural data (Di Ciano et al., 1998; Kuczenski et al., 1989) indicating that DA activity is more closely associated with sensory events rather than motor output. In addition, the lack of a continuing requirement for PPTg activity is consistent with observations of a potent GABA-mediated inhibition of PPTg neurons derived from the ventral midbrain and triggered by cholinergic activation there (Laviolette et al., 2002; Saitoh et al., 2003). It has been hypothesised, specifically in relation to opiates, by van der Kooy and his colleagues (Bechara et al., 1992, 1998) that the PPTg forms one of two separate motivational systems. They argued that the PPTg mediated motivation when rats were in a drug-naive state but in morphine-dependent rats motivation for opiate reward was mediated not by the PPTg but by a DA-dependent system (Bechara et al., 1998). This was based on their finding that PPTg lesions affected place preference to morphine in drug naive but not morphine-dependent animals, whilst manipulations of DA were without effect in the drug naive condition, but impaired place prefer-

Table 3. Mean percent total responses on inactive lever under PR schedulea

A: Exp. 1 B: Exp. 2 C: Exp. 3 a

Control PPTg Control PPTg Control PPTg

Session 1

Session 2

Session 3

Session 4

3.3⫾1.4 9.1⫾3.8 7.4⫾2.7 11.0⫾6.1 5.3⫾4.8 11.4⫾10.0

5.9⫾3.6 9.7⫾3.2 5.6⫾1.7 17.9⫾8.9 0.4⫾0.4 10.2⫾8.2

2.4⫾1.0 11.9⫾3.2 5.7⫾2.0 14.5⫾8.8 1.8⫾1.7 9.2⫾7.5

3.0⫾1.2 7.3⫾2.4 5.5⫾2.0 12.9⫾8.0 0.6⫾0.4 6.1⫾4.8

In experiment (Exp.) 1, the PPTg excitotoxin lesion group made a significantly higher proportion of their total (active plus inactive) responses on the inactive control lever than did the sham-lesioned controls (P⬍0.05). In experiments 2 and 3, there were no significant differences between the excitotoxin and sham lesion groups in their level of responding on the inactive lever.

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358

ence in dependent rats (Bechara et al., 1992). However, it could be argued that in these place preference experiments PPTg lesions were without effect in morphinedependent rats because DA was already in an activated state, matching closely the priming condition reported here. The suggestion that PPTg and DA systems are independent of each other has little power to explain the present results. Rather, the experiments presented here suggest that there is actually a close inter-relationship between PPTg neurons and midbrain DA systems in the acquisition of d-amphetamine self-administration behaviour.

CONCLUSIONS (i) An important conclusion of this study is that care must be taken to avoid experimental confounds through interactions between behavioural methods and the effects of lesions. Our data show clearly that the behavioural conditions under which drug reinforcement studies are run have significant effects on outcome of lesion studies. (ii) PPTglesioned rats show impairments in the acquisition of damphetamine IVSA in conditions requiring learning of new relationships between lever pressing and reinforcement outcomes, but not in conditions where relevant learning has already taken place. (iii) These data provide further, albeit indirect, evidence supporting an important role for the PPTg in the regulation of midbrain DA neurons. Acknowledgements—Supported by Wellcome Trust Project Grant 054927/Z/98/Z. We wish also to acknowledge the support of the School of Psychology, University of St. Andrews.

REFERENCES Alderson HL, Jenkins TA, Kozak R, Latimer MP, Winn P (2001) The effects of excitotoxic lesions of the pedunculopontine tegmental nucleus on conditioned place preference to 4%, 12% and 20% sucrose solutions. Brain Res Bull 56:599 –605. Alderson HL, Brown VJ, Latimer MP, Brasted PJ, Robertson AH, Winn P (2002) The effect of excitotoxic lesions of the pedunculopontine tegmental nucleus on performance of a progressive ratio schedule of reinforcement. Neuroscience 112:417–425. Bechara A, van der Kooy D (1989) The tegmental pedunculopontine nucleus: a brain-stem output of the limbic system critical for the conditioned place preferences produced by morphine and amphetamine. J Neurosci 9:3400 –3409. Bechara A, Harrington F, Nader K, van der Kooy D (1992) Neurobiology of motivation: double dissociation of two motivational mechanisms mediating opiate reward in drug-naive versus drug-dependent animals. Behav Neurosci 106:798 –807. Bechara A, Nader K, van der Kooy D (1998) A two-separate motivational systems hypothesis of opiate addiction. Pharm Biochem Behav 59:1–17. Bevan MD, Bolam JP (1995) Cholinergic, GABAergic, and glutamateenriched inputs from the mesopontine tegmentum to the subthalamic nucleus in the rat. J Neurosci 15:7105–7120. Blaha CD, Winn P (1993) Modulation of dopamine efflux in the striatum following cholinergic stimulation of the substantia nigra in intact and pedunculopontine tegmental nucleus-lesioned rats. J Neurosci 13: 1035–1044. Blaha CD, Allen LF, Das S, Inglis WL, Latimer MP, Vincent SR, Winn P (1996) Modulation of dopamine efflux in the nucleus accumbens after cholinergic stimulation of the ventral tegmental area in intact,

357

pedunculopontine tegmental nucleus-lesioned, and laterodorsal tegmental nucleus-lesioned rats. J Neurosci 16:714 –722. Caine SB, Lintz R, Koob GF (1993) Intravenous drug self-administration techniques in animals. In: Behavioural neuroscience: a practical approach (Saghal A, ed), pp 117–143. Oxford: Oxford University Press. Corrigall WA, Coen KM, Adamson KL (1994) Self-administered nicotine activates the mesolimbic dopamine systems through the ventral tegmental area. Brain Res 653:278 –284. Corrigall WA, Coen KM, Adamson KL, Chow BLC (1999) Manipulations of mu-opioid and nicotinic cholinergic receptors in the pontine tegmental region alter cocaine self-administration in rats. Psychopharmacology 145:412–417. Corrigall WA, Coen KM, Zhang J, Adamson KL (2001) GABA mechanisms in the pedunculopontine tegmental nucleus influence particular aspects of nicotine self-administration selectively in the rat. Psychopharmacology 158:190 –197. Corrigall WA, Coen KM, Zhang J, Adamson KL (2002) Pharmacological manipulations of the pedunculopontine tegmental nucleus in the rat reduce self-administration of both nicotine and cocaine. Psychopharmacology 160:198 –205. Di Ciano P, Blaha CD, Phillips AG (1998) The relation between dopamine oxidation currents in the nucleus accumbens and conditioned increases in motor activity in rats following repeated administration of d-amphetamine or cocaine. Eur J Neurosci 10:1113–1120. Dormont JF, Conde H, Farin D (1998) The role of the pedunculopontine tegmental nucleus in relation to conditioned motor performance in the cat: I. Context-dependent and reinforcement-related single unit activity. Exp Brain Res 121:401–410. Fiorillo CD, Tobler PN, Schultz W (2003) Discrete coding of reward probability and uncertainty by dopamine neurons. Science 299: 1898 –1902. Gould E, Woolf NJ, Butcher LL (1989) Cholinergic projections to the substantia nigra from the pedunculopontine and laterodorsal tegmental nuclei. Neuroscience 28:611–623. Groenewegen HJ, Berendse HW, Haber SN (1993) Organization of the output of the ventral striatopallidal system in the rat: ventral pallidal efferents. Neuroscience 57:113–142. Grofova I, Zhou M (1998) Nigral innervation of cholinergic and glutamatergic cells in the rat mesopontine tegmentum: light and electron microscopic anterograde tracing and immunohistochemical studies. J Comp Neurol 395:359 –379. Heimer L, Zahm DS, Churchill L, Kalivas PW, Wohltmann C (1991) Specificity in the projection patterns of accumbal core and shell in the rat. Neuroscience 41:89 –125. Hodos W (1961) Progressive ratio as a measure of reward strength. Science 134:943–944. Inglis WL, Allen LF, Whitelaw RB, Latimer MP, Brace HM, Winn P (1994) An investigation into the role of the pedunculopontine tegmental nucleus in the mediation of locomotion and orofacial stereotypy induced by d-amphetamine and apomorphine in the rat. Neuroscience 58:817–833. Inglis WL, Winn P (1995) The pedunculopontine tegmental nucleus: where the striatum meets the reticular formation. Prog Neurobiol 47:1–29. Jackson A, Crossman AR (1983) Nucleus tegmenti pedunculopontinus: efferent connections with special reference to the basal ganglia, studied in the rat by anterograde and retrograde transport of horseradish peroxidase. Neuroscience 10:725–765. Keating GL, Blaha CD, Winn P, Di Ciano P, Latimer MP, Phillips AG (1997) Amphetamine self-administration is enhanced by excitotoxic lesions of the pedunculopontine tegmental nucleus. Soc Neurosci Abstr 23:2145. Keating GL, Walker SC, Winn P (2002) An examination of the effects of bilateral excitotoxic lesions of the pedunculopontine tegmental nucleus on responding to sucrose reward. Behav Brain Res 134: 217–228. Kuczenski R, Segal D (1989) Concomitant characterization of behav-

358

H. L. Alderson et al. / Neuroscience 125 (2004) 349 –358

ioral and striatal neurotransmitter response to amphetamine using in vitro microdialysis. J Neurosci 9:1051–1065. Laviolette SR, Alexson TO, van der Kooy D (2002) Lesions of the tegmental pedunculopontine nucleus block the rewarding effects and reveal the aversive effects of nicotine in the ventral tegmental area. J Neurosci 22:8653–8660. Lavoie B, Parent A (1994) Pedunculopontine nucleus in the squirrel monkey: distribution of cholinergic and monoaminergic neurons with direct evidence for the presence of glutamate in cholinergic neurons. J Comp Neurol 344:190 –290. Moriizumi T, Hattori T (1992) Separate neuronal populations of the rat globus pallidus projecting to the subthalamic nucleus, auditory cortex and pedunculopontine tegmental nucleus. Neuroscience 46: 701–710. Oakman SA, Faris PL, Kerr PE, Cozzari C, Hartman BK (1995) Distribution of pontomesencephalic cholinergic neurons projecting to substantia nigra differs significantly from those projecting to ventral tegmental area. J Neurosci 15:5859 –5869. Olmstead MC, Franklin KBJ (1994) Lesions of the pedunculopontine tegmental nucleus block drug-induced reinforcement but not amphetamine-induced locomotion. Brain Res 638:29 –35. Olmstead MC, Munn EM, Franklin KBJ, Wise RA (1998) Effects of pedunculopontine tegmental nucleus lesions on responding for intravenous heroin under different schedules of reinforcement. J Neurosci 18:5035–5044. Olmstead MC, Inglis WL, Bordeaux CP, Clarke EJ, Wallum NP, Everitt BJ, Robbins TW (1999) Lesions of the pedunculopontine tegmental nucleus increase sucrose consumption but do not affect discrimination or contrast effects. Behav Neurosci 113:732–743. Parker JL, van der Kooy D (1995) Tegmental pedunculopontine nucleus lesions do not block cocaine reward. Pharm Biochem Behav 52:77–83. Rye DB, Saper CB, Lee HJ, Wainer BH (1987) Pedunculopontine tegmental nucleus of the rat: cytoarchitecture, cytochemistry and some extrapyramidal connections of the mesopontine tegmentum. J Comp Neurol 259:483–528. Saitoh K, Hattori S, Song W-J, Isa T, Takakusaki K (2003) Nigral

GABAergic inhibition upon cholinergic neurons in the rat pedunculopontine tegmental nucleus. E J Neurosci 18:879 –886. Salamone JD, Wisniecki A, Carlson BB, Correa M (2001) Nucleus accumbens dopamine depletions make animals highly sensitive to high fixed ratio requirements but do not impair primary food reinforcement. Neuroscience 105:863–870. Schultz W (2002) Getting formal with dopamine and reward. Neuron 36:241–263. Semba K, Fibiger HC (1992) Afferent connections of the laterodorsal and the pedunculopontine tegmental nuclei in the rat: a retro- and antero-grade transport and immunohistochemical study. J Comp Neurol 323:387–410. Spann BM, Grofova I (1991) Nigropedunculopontine projection in the rat: an anterograde tracing study with Phaseolus vulgaris-leucoagglutinin (PHA-L). J Comp Neurol 311:375–388. Steininger TL, Rye DB, Wainer BH (1992) Afferent-projections to the cholinergic pedunculopontine tegmental nucleus and adjacent midbrain extrapyramidal area in the albino-rat. 1. Retrograde tracing studies. J Comp Neurol 321:515–543. Vincent SR, Kimura H (1992) Histochemical mapping of nitric oxide synthase in the rat brain. Neuroscience 46:755–784. Winer BJ (1971) Statistical principals in experimental design. London: McGraw-Hill. Winn P, Brown VJ, Inglis WL (1997) On the relationships between the striatum and the pedunculopontine tegmental nucleus. Crit Rev Neurobiol 11:241–261. Yokel RA, Pickens R (1974) Drug level of d- and l-amphetamine during intravenous self-administration. Psychopharmacology 34:255– 264. Zahm DS, Jensen SL, Williams ES, Martin IIIJR (1999) Direct comparison of projections from the central amygdaloid region and nucleus accumbens shell. Eur J Neurosci 11:1119 –1126. Zahm DS, Williams EA, Latimer MP, Winn P (2001) Ventral mesopontine projections of the caudomedial shell of the nucleus accumbens and extended amygdala in the rat: double dissociation by organization and dissociation. J Comp Neurol 436:111–125.

(Accepted 11 February 2004)