Diazepam modulates lateral hypothalamic self-stimulation but not stimulation-escape in rats

Br~TinResea~'ch,483 (1989) 327-334 Elsevier

327

BRE 14355

Diazepam modulates lateral hypothalamic self-stimulation but not stimulation-escape in rats Susan E. Carden and Edgar E. Coons Department of Psychology, New York University, New York, NY IO003 (U.S.A.) (Accepted 30 August 1988)

Key words: Aversion; Benzodiazepine; Diazepam; Escape; Lateral hypothalamus; Reward; Self-stimulation

Rats electrically stimulated via chronically implanted lateral hypothalamic (LH) electrodes were assessed with and without diazepam (DZ), for thresholds of stimulation-bound feeding (SBF) and for barpressing rates to administer and to escape from the same current. Six pure-reward rats, who self-stimulated but did not escape LH stimulation, exhibited SBF. Their electrode tips lay in medial forebrain bundle (MFB) and zona inserta along the entire rostral-caudal extent of the ventromedial nucleus of the hypothalamus (VMH). Six reward-escape rats, who self-stimulated and escaped from LH stimulation, did not (with one histologically deviant exception) show SBF. Reward-escape electrode tips were anterior to all the pure-reward placements. They lay in MFB rostral to the VMH up to the level of the bed nucleus of the stria terminalis (with the deviant electrode tip located on the zona inserta/ventral thalamic border). After i.p. injections of DZ, self-stimulation (SS) rates increased for both groups of animals and SBF thresholds decreased. Stimulation-escape (SE) rates, however, remained unchanged by the drug. The results are consistent with the existence of dual substrates: a DZ-sensitive reward system, present in both groups of animals, and a simultaneously stimulated, drug-resistant aversion system which is powerfully engaged in reward-escape animals only.

INTRODUCTION The rewarding nature of lateral hypothalamic (LH) brain stimulation has long been established 28,29. Two groups of self-stimulating rats (reward-escape, who will turn off continuous trains of L H current, and pure-reward, who will not) have been identified by some researchers 7'z4 although others have found escape behavior at every reward site 41. Stimulationb o u n d feeding (SBF), where sated rats eat during L H stimulation and stop feeding at current offset 9,1°,26, occurs only in p u r e - r e w a r d animals, not reward-escape 9. Few studies of reward-escape have contrasted the performance characteristics of p u r e - r e w a r d and reward-escape rats. In the following experiments, barpressing rates in the two groups were m e a s u r e d for self-stimulation (SS) and stimulation-escape (SE), with and without diazepam ( D Z ) . D Z has been

shown to increase L H reward (raises SS rates and lowers thresholds) 3°-32, and L H motivation (lowers SBF thresholds) 39'42. H o w e v e r , in shuttlebox studies, D Z effects are often interpreted as reduced aversion rather than increased reward 22. SS and SE are variously p o r t r a y e d as being supp o r t e d by two functions within a single neural substrate 13'17'19 or by two separate systems 35'4°. If SS and SE are m e d i a t e d by the same neural pathway, D Z effects should be a p p a r e n t in both behaviors. If separate neurological substrates exist, then pharmacological discrimination of effect may be possible. Large differences have been r e p o r t e d in the duration preference of p u r e - r e w a r d and reward-escape rats 7. Therefore, train length and magnitude of stimulation were also m a n i p u l a t e d and analyses made of the resultant contrasts between groups with and without D Z . To compile a m o r e complete record of purereward/reward-escape contrasts, animals were classi-

Correspondence: S. Carden, New York State Psychiatric Institute, 722 West 168th Street, Department of Psychiatry, Developmental Psychobiology, Box 40, New York, NY 10032, U.S.A. 0006-8993/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

328 fled as stimulation-bound feeders and non-feeders and the effects of DZ on this behavior were also measured. MATERIALS AND METHODS

Subjects and surgical procedures Monopolar electrodes insulated except for 0.3 mm at the tips were chronically implanted in the MFB of the LH in 12 anesthetized male Sprague-Dawley rats. Implantation coordinates (at skull fiat orientation) were 1.4 mm lateral to the sagittal suture and 8.5 mm ventral to the dorsal surface of the skull. Anterior/posterior placements ranged from 1.5 to 3.0 mm posterior to Bregma. A caudally placed uninsulated jeweler's screw functioned as the indifferent electrode. Animals were individuall S housed with food and water available ad libitum. All testing was performed during the light cycle of a 12-h light/dark schedule.

train lengths as well as the criterion for SE. Practice sessions continued until response rates stabilized. Test sessions consisted of 16 1-min trials, separated.by l-rain intertrial intervals (ITI). The animal's first response reset the clock to begin the timing of the 1-min trial. SBF sessions consisted of 30-s trials alternating with 60-s ITI. The method of limits was employed to establish the lowest stimulation level which would produce at least 5 s of continuous eating, ending abruptly with the offset of current. Baseline data collection was followed by a week during which the animals were acclimated to DZ and allowed to develop a tolerance to the initial ataxia. DZ, 5 mg/kg, was administered i.p. 20 min before each drug session. DZ and saline data were collected on alternate days, with each of the baseline conditions for SS, SE and SBF being replicated with the drug. RESULTS

Apparatus and brain stimulation Solid state logic circuitry delivered trains of 0.1 ms capacitance-coupled, negative-going pulses. Current levels were controlled by a 50 k~2 potentiometer. Frequency, current, pulse width and train duration were monitored continuously on an oscilloscope. A Plexiglas test chamber measuring 34 x 34 x 46 cm, with a retractable lever installed on one wall, was the site of all testing.

Histology After completion of behavioral testing, rats were terminally anesthetized and perfused. Their brains were removed and histologically prepared by standard methods for identification of the location of electrode tips.

Behavioral screening and preliminary training Following a week or more of recovery, animals were screened for SS over a range of parameters. Those animals who were not stable self-stimulators were discarded. Once animals met the SS criteria, shaping began for SE and animals were screened for the presence of SBF. Using the identical 4 levels of stimulation (one current combined with 4 frequencies), animals had to meet the criteria for SS at both the 0.5-s and 3.0-s

Classification Six animals in this study escaped from continuous trains of LH stimulation and 6 did not. Animals who would not work for offset of stimulation were designated pure-reward. Animals who barpressed to turn LH current off as well as on were classified as reward-escape. No animal, once identified as pure-reward, ever showed escape b e h a v i o r - - nor was there ever reason to reclassify a reward-escape rat. SBF, linked to pure-reward characteristics by previous studies, was present in each of 6 pure-reward and initially absent in 6 reward-escape rats. There were no marginal cases. Animals either ate consistently during stimulation or showed no interest in the food. (One reward-escape rat, RE 12, belatedly emerged as a feeder.) A Fisher exact test (two-tailed) performed on the variables pure-reward vs reward-escape and SBF vs non-SBF yielded P = 0.015 that association of these characteristics occurred by chance. The statement that pure-reward and reward-escape animals form two distinct groups based on the presence or absence of SE appears to be valid.

Stimulation levels A standardized method will be used to report lev-

329 A.

2O

PURE-REWARD

5 . O - s e c trains

B.

20

0--0 SS Base 0--0 SS DZ F-l-- [] SE Base • -- • SEDZ

REWARD-ESCAPE 0--0 SS Base 0--0 SS DZ I-l-- [ ] SEBose • --

-~

15

3 . O - s e c trains

•

SE DZ

15

c

E

E

/

if) L.

Q ffl

5

0 /

if) if)

0

Q

m

. Z2

-=~•

2

5

o 0

l--

--

Freq. I

C. 120

--I-Freq. 2

PURE-REWARD 0--0 0--0

--

=lIl:

Freq. 3

Freq. 4

Freq. I

D.

0 . 5 - s e c trains

ss Base SS DZ

120

90

~

C

0

0

----- _ - - i

REWARD-ESCAPE 0--0 0--0

Freq. 3

Freq. 4

0 . 5 - s e c trains

SS Base SS DZ

90

E

/ e -

E O~

• ~0

E (n

8 c~ m

Freq. 2

8 c~ m

30

e/° o

o Freq. 1

0__..___._.---0

30

0~ /

~0

--0

0

o Freq. 2

Freq. 3

Freq. 4

Freq. 1

Stimulation level

Freq. 2

Freq. 3

Freq. 4

Stimulation level

Fig. 1. Barpressing rates of self-stimulation (SS) and stimulation-escape (SE) for pure-reward (n = 6) and reward-escape (n = 6) rats. A and B show response rates for 3.0-s current trains, C and D for 0.5-s trains. Rates are presented for baseline condition and after administration of 5 mg/kg diazepam (DZ).

els of stimulation. Identical behavioral criteria were used to choose currents and frequencies for the 6 p u r e - r e w a r d and 6 reward-escape animals. In addition, the results of testing show linear relationships b e t w e e n the 4 stimulation levels and barpressing rates for every animal for both SS and SE. (The m e a n correlation for SS was q0 = 0.86 and for SE, r 4 = 0.93.) Therefore, the lowest stimulation level for each animal, where he is receiving the fewest pulses p e r second, will be designated frequency 1. The next highest, will be frequency 2, etc. In this way it is possible to combine data across animals.

Self-stimulation Pure-reward rats self-stimulated faster than reward-escape animals. The data were analyzed in two split plot A N O V A s , one for each train length. Significant main effects for group m e m b e r s h i p (pure-reward vs reward-escape) were found for long trains (F1,10 = 23.82, P < 0.001, Figs. 1A vs 1B) and for short (FIA0 = 16.69, P < 0.01, Fig. 1C vs 1D). The stimulation levels required to achieve criterion levels of response were greater for reward-escape rats than for p u r e - r e w a r d (Table I). Therefore, faster response rates in pure-reward rats are not a function of

330 TABLE I

TABLE II

Stimulation parameters for self-stimulation and escape

Electrode placements for pure-reward and reward-escape animals

Subject

Current

Freq. 1

Freq.2

Freq.3

Freq. 4

(/~A)

(p.p.s.)

(p.p.s.)

(p.p.s.)

(p.p.s.)

Animal

Anterior/ posterior

Medial/ lateral

Dorsal/ ventral

108.33 45.55

41.67 12.91

70.83 18.82

112.50 30.62

166.67 40.82

141.67 20.41

48.83 10.21

91.67 20.41

141.67 20.41

200.00 0.00

41.67 12.91

83.33 25.82

129.17 33.23

183.33 40.82

PR 17 PR 1 PR 8 PR 15 PR 18 PR 7 RE9 RE 11 RE 12 RE 2 RE 16 RE 21

4100 4380 4620 4890 4890 5020 5280 5660 5660 5840 6060 6280

1.2 1.4 1.3 1.4 1.2 1.3 2.0 1.4 1.5 1.3 1.4 1.4

-2.7 -2.5 -2.4 -2.0 -2.3 -2.4 -2.6 -2.3 -1.2 -2.0 -2.0 -2.0

PR

Mean S.D. RE Mean S.D. Mean* S.D.

* Frequencies used for escape trials only.

higher levels of stimulation. When D Z was administered, there was a large increment over baseline barpressing rates at every frequency, at both train lengths, for all animals. The main effects of drug administration in a split plot A N O V A were/:1.10 = 34.73, P < 0.001 (Figs. 1A and 1B) for long trains and F1,10 = 149.77, P < 0.0001 (Figs. 1C and 1D) for short. During drug sessions, as in the baseline conditions, pure-reward animals barpressed faster than rewardescape. However, significant 3-way interactions - F3.30 = 5.44, P < 0.01 for short trains and F3,30 = 4.45, P < 0.05 for long - - indicate a DZ-altered relationship of train length and group membership. By comparing Figs. 1A and 1C with 1B and 1D, the difference in drug effects between groups can be seen. A behavioral ceiling appears to make D Z relatively ineffective in boosting pure-reward SS rates at the highest frequencies, while the drug increases reward-escape SS rates across both train lengths at all but one stimulation level.

on SS, there was no change in SE rates (Fig. 1B). A split plot A N O V A on all animals yielded highly significant (P < 0.0001) F-scores for the main effects of group (Fa,10 = 61.85) and level of stimulation (F3,30 = 14.57), as well as for the interaction of these two factors (F3,30 = 14.41). This same A N O V A produced Fscores of less than 1.0 for the main effects of drug and for all drug interactions.

Stimulation-bound feeding SBF was present in every pure-reward and, initially, absent in every reward-escape rat. Later in the experiment, one reward-escape animal ( R E 12) became a feeder, although continuing to escape. Ten animals completed testing for SBF. The 5 feeding animals each showed decreased thresholds to feed following drug administration (correlated t4 = 3.19, P < 0.05, two-tailed). D Z did not induce SBF in any non-feeding rat.

Stimulation-escape All animals were tested for rates of barpressing for 3.0-s escapes from continuous L H stimulation at the same 4 current/frequency combinations used for SS trials. ( D Z made rat R E 12 spin in tight circles on SE trials. Therefore, his SE trials were re-run at lower frequencies in baseline and drug conditions.) Reward-escape rats, by definition, all met or exceeded the 8 bp/min criterion, while the mean SE rate for pure-reward animals was never more than 1 bp/min at any stimulation level in any condition (Fig. 1A). For all reward-escape rats, when D Z was administered at the same dose which had a profound effect

Histology Electrode sites for all animals are listed in Table II. According to the coordinates of K6nig and Klippel 2°, every reward-escape animal had an electrode placed anterior to 5150 pM. Every pure-reward electrode was posterior to this point. Pure-reward placements lay in lateral hypothalamic medial forebrain bundle (MFB) and zona inserta-thalamic border. There were negative correlations between anterior/posterior (A/P) electrode location and SS barpressing rates, at both train lengths, which were significant at frequency 3 (rl0 = -0.63, P < 0.05 for

331 short trains, rl0 = -0.72, P < 0.01 for long). These correlations were essentially replicated at frequency 4 (rl0 = -0.65, P < 0.05 for short and rl0 = -0.65, P < 0.05 for long). Negative correlations indicate that SS was faster at more posterior, pure-reward placements. When performance within each group was correlated with location, there was no continuum of increasing reward value from the most anterior to the most posterior placements. On the contrary, rewardescape rats showed higher SS barpressing rates associated with the more anterior electrodes. Higher SE rates were also associated with the more anterior placements in reward-escape rats. Although not significant, these correlations, ranging f r o m rl0 ----0.65 t o ra0 = 0.81 for the 4 stimulation levels can be interpreted as a trend in the data. Medial/lateral (M/L) correlations with SS barpressing rates were significant for frequencies 3 and 4 on 3.0-s trials (rl0 = -0.74, P < 0.01) and approached significance (rl0 = -0.55) for frequency 4 on 0.5-s trials. Significance at the 0.05 level requires rl0 = 0.5760. The more medial locations were associated with SS higher rates. Correlations between M/L locations and SE rates were strongly negative at frequencies 2, 3, and 4 ( r 4 = -0.91, P < 0.02, r4 = -0.90, P < 0.02, r 4 = -0.99, P < 0.001), with greater SE sensitivity at the more medial placements. However, the range of M/L placements was, with only one exception, very narrow (1.2-1.5 mm lateral). There is, therefore, danger of overinterpreting these data. Dorsal/ventral correlations with SS and SE rates were not significant. Fig. 2 presents SS and SE rate data as a function of A/P location for frequency 1. While not intended to be a formal measure of thresholds, this figure shows clearly that the lowest level of stimulation elicited SE, but not SS in 4 of 6 reward-escape rats. When DZ sensitivity measures were correlated with histology, group membership did not appear to determine the impact of the drug. DISCUSSION

SBF was present in every pure-reward animal but absent in reward-escape animals. One, RE 12, emerged as a feeder late in the experiment. This ani-

20

[3 PR ss • 0

PR SE RE SS

• RESE

Z

15

E co

I0

5

[]

o

•

r-qE3 I

4000

=lil•~

• eo

i

I

I

4500

5000

5500

Posterior

O I

6000

o

I

6500 Anterior

Electrode location

Fig. 2. Self-stimulation (SS) and stimulation-escape (SE) barpressing rates for onset of 3.0-s current trains or 3.0-s escapes from continuous stimulation at frequency 1. The rates for purereward (PR) (n = 6) and reward-escape (RE) (n = 6) rats are presented as a function of anterior/posterior electrode location (according to the coordinates of K6nig and Klippel).

mal also functioned poorly on DZ escape trials, necessitating an adjustment in stimulation parameters. A possible explanation is the atypical electrode placement in this rat. The tip of the electrode, in the same A/P range as other reward-escape rats, was far dorsal to the rest, stimulating the boundary of zona inserta and ventral thalamus. SBF thresholds for every animal decreased in the presence of DZ. Where SBF was not already present, DZ could not induce a reversible feeding effect. (The one animal, RE 12, who emerged as a feeder, did so on saline and posttest days as well as in the presence of the drug.) Reward and SBF, therefore, responsed to DZ, while SE and the absence of SBF did not. It is possible that the same mechanism which causes animals to escape prevents their being stimulation-bound feeders. Pure-reward animals produced higher SS rates than reward-escape rats using lower stimulation levels. The reward of LH stimulation comes from excitation of fibers of passage in the MFB 36"43which is compact in posterior hypothalamus and more diffuse in the anterior portion 16. Therefore, posterior, pure-reward electrodes may stimulate a larger number of reward neurons. Higher SS rates for pure-reward rats could also indicate the absence in these animals of a simultaneously stimulated aversive substrate which is present in the reward-escape animals (who require higher stimulation levels and barpress more slowly). The

332 proximity of the electrode tip to an anterior aversive substrate might be a factor in determining SS response rates. Why do rats escape from rewarding LH stimulation? The adaptation hypothesis states that there is a decrease in reward over time 1M4'~9. Interrupting, a n d then restoring current flow, returns the reward value to its initial high level. In the present study, at the lowest frequency, 4 out of 6 reward-escape animals showed SE but not SS (a relationship reported earlier37). To say that a rat turns current off in order to have it re-initiated at that s a m e stimulation level, as adaptation theory demands, is not credible for these animals. They will not turn current on at all unless the intensity is increased. While adaptation can explain some of the data in this study, it cannot account for it all. The second major explanation of SE suggests progressively increasing aversion 35'4°. In the current study the barpressing rates of both groups increased with DZ, with only reward-escape rats showing significant increases at the higher stimulation levels. It appears that in baseline reward-escape animals are performing well below their behavioral ceiling. This could be consonant with an animal receiving simultaneously rewarding and aversive stimulation. If DZ boosts the reward value while leaving the aversiveness untouched, the result would be a net increase in the positive value of the reinforcement, reflected in more rapid responding. In contrast, the lack of significant drug-induced increases at the highest stimu-

In shuttlebox studies, longer latencies to interrupt LH current in the presence of D Z were attributed to a decrease in aversion 22. There is an apparent discrepancy between this interpretation and the lack of drug effect on escape barpressing rates reported here. (In a complementary threshold study, D Z decreased thresholds for SS but again left those for SE unchangedS.) If higher SS rates are attributable to a drug-related decrease in aversion, this decrease in aversion should have lowered rates for SE as well. Yet SE rates did not change. It is possible, therefore, that higher SS rates on drug trials are a function of an increase in reward rather than a decrease in aversion. (Since DZ increased SS barpressing rates in both groups of animals, the lack of effect of DZ on SE-responding cannot be attributed to drug-insensitivity in reward-escape rats.) This explanation is still problematic. If D Z boosts the reward value of LH stimulation - - why doesn't

lation levels (where pure-reward animals are approaching their behavioral ceiling even without the drug) could be anticipated in a system which has only positive input, the magnitude of which increases with the stimulation level. In a study of SS thresholds, where ceiling effects were not a factor, there was no difference in DZ-induced changes between pure-reward and reward-escape rats 8. An alternate explanation is that SS and SE behaviors are controlled by reward neurons firing at offset as well as onset 17. This would require one of two constructs: (a) a single reward substrate, DZ-sensitive when firing at onset but not when firing at offset - which is present in reward-escape animals only, or (b) two independent reward systems, anatomically and pharmacologically dissociable, one sensitive and the other resistant to the drug. In both cases, one

that increased reward lower SE rates by making the animal more ambivalent about turning the current off? Rather than reward and aversion being additive, aversion may increase after several seconds of stimulation with a step function - - wresting control of the final common pathway whether reward has been augmented by DZ or not. In effect, it is proposed that barpressing for the onset of current is reinforced by the current's rewarding properties, which, in turn, are enhanced by DZ. Barpressing for current offset is reinforced by an aversive property which emerges more slowly than reward and is unaffected by DZ. Another possibility is that DZ has its impact downstream from the decision whether or not to escape. Once the pathway of SE is engaged, it is refractory to fluctuations in concomitant reward. That reward and escape are pharmacologically dis-

must postulate that pure-reward animals have only the substrate for onset reward. In animals who receive both onset and offset reward, the degree of sensitivity to the two forms of reward would have to be independent, since latencies to turn on and to turn off current have been found to be independent 2'5. While there is nothing in the current study to disprove this hypothesis, the data are most readily reconciled with theories which state: (a) that escape is caused by aversion 24,25,35,40 and (b) that aversion is a function of a neural substrate separate and independent from that which supports reward 6'21'27'37'38.

333 sociable here, as in earlier studies 4"21'27, argues strongly for separate neural substrates governing the two behaviors. This leaves unresolved the discrepancies between the results of shuttlebox studies and those of the present barpressing experiment. The populations of rats used in these two types of studies are, in all likelihood, only partially overlapping. The shuttle response is learned more quickly than barpressing and some animals who will not barpress will shuttle vigorously for reward 13. In addition, one of the most common responses to rewarding LH stimulation is forward locomotion 34. It is possible that this forced movement might impel a pure-reward rat to cross a shuttlebox. Since amphetamine, at doses which increase SS by augmenting reward 23, selectively decreases the latency to turn current on (ON latency), DZ-induced increases in the latencies to turn current off (OFF latency) have been interpreted as a manipulation of aversion =. Amphetamine, however, is a psychomotor stimulant and DZ, a muscle relaxant. Therefore, the assumption that if D Z increased reward, the increase would be expressed in exactly the same way as amphetamine-mediated increases in reward does not have to be accepted unquestioningly. In addition, when anxiogenic drugs were administered, anticipating a decrease in the OFF latency, increases in the ON latency were seen instead 15. Therefore, manipulations of aversion do not appear to be immutably wed to changes in OFF latency. A separation was seen in the location of electrodes of the two groups in this study, with all reward-escape placements anterior to all pure-reward. This suggests either that the neural substrate for aversion may have a more anterior location or that it travels in a trajectory which in anterior hypothalamus brings it into closer proximity with reward neurons. Recent research on differences between the reward and aversive substrates 6.37 use stereotaxic coordinates which would place electrodes lateral to all the pure-reward locations in the current experiments. An aversive fiber tract traveling in a posterolateral direction could be present in the medial part of anterior hypothalamus where reward-escape placements were found in this study, be present also in the more REFERENCES 1 Atrens, D.M., Reinforcing and emotional consequences of electrical self-stimulation of the subcortical limbic fore-

lateral parts of the hypothalamus at the level of the ventromedial nucleus (the location of reward-escape placements in the Shizgal and Bielajew studies6'36-38), but miss the more medial placements at this level which proved to be pure-reward in the current study. Not every investigator reports the existence of both pure-reward and reward-escape animals. When both groups of animals are identified in the same study, there are reports of a pure-reward/reward-escape dichotomy in electrode placement - - in the dorsal/ventral plane 24 and in the medial/lateral plane TM, but the earliest and most consistent reports are of a separation of anterior and posterior placements7,29,33. Within the group of reward-escape animals, both SS and SE rates were higher in animals with more anterior placements. If the systems for reward and aversion were integrated, one would anticipate an inverse relationship, with the strongest SS at locations yielding the weakest SE. In a series of experiments with reward-escape rats 2'3"5 latencies for ON and OFF responding in the shuttlebox were found not to be correlated with each other. The authors concluded that reward and aversive substrates were independent, although anatomically coextensive. This conclusion of independence was echoed by experimenters who found differences in rates of temporal summation for reward and escape 37, different trajectories 6, and different rates of local potential summation 38. It was echoed by experimenters who selectively altered ON-time using morphine 2~, or OFF-time using apomorphine 4, or found pimozide attenuated SS but left SE responding unaltered 27. This conclusion that reward and aversion are supported by separate neural systems fits most closely with the results of the current experiment. ACKNOWLEDGEMENTS The gift of diazepam from Hoffmann-La Roche, Inc. is gratefully acknowledged. This article is based upon a dissertation submitted by the first author under the guidance of the second author to the Graduate School of Arts and Science of New York University in partial fulfillment of the requirements for the degree of Doctor of Philosophy. brain, Physiol. Behav., 5 (1970) 1461-1471. 2 Atrens, D.M., A reinforcement analysis of rat hypothalamus, Am. J. Physiol., 224 (1973) 62-65. 3 Atrens. D.M. and Becker, F.T., Assessingthe aversiveness

334

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20 21

22 23

of intracranial stimulation, Psychopharmacologia, 44 (1975) 159-163. Atrens, D.M., Becker, F.T. and Hunt, G.E., Apomorphine: selective inhibition of the aversive component of lateral hypothalamic self-stimulation, Psychopharmacology, 71 (1980) 97-99. Atrens, D.M. and von Vietinghoff-Riesch, F., The motivational properties of electrical stimulation of the medial and paraventricular hypothalamic nuclei, Physiol. Behav., 9 (1972) 229-235. Bielajew, C. and Shizgal, P., Dissociation of the substrates for medial forebrain bundle self-stimulation and stimulation-escape using a two-electrode stimulation technique, Physiol. Behav., 25 (1980) 707-711. Bower, G.H. and Miller, N.E., Rewarding and punishing effects from stimulating the same place in the rat's brain, J. Comp. Physiol. Psychol., 51 (1958) 669-674. Carden, S.E., Diazepam modifies lateral hypothalamic selfstimulation and stimulation-bound feeding but not escape responding: evidence of separate neural substrates for reward and aversion, Doctoral dissertation, New York University, New York, 1988. Coons, E.E., Motivational correlates of eating elicited by electrical stimulation in the hypothalamic feeding area, Doctoral dissertation, Yale University, New Haven, 1964. Coons, E.E., Levak, M. and Miller, N.E., Lateral hypothalamus: learning of food-seeking response motivated by electrical stimulation, Science, 150 (1965) 1320-1321. Deutsch, J.A. and Albertson, T.E., Refractory period and adaptation in prolonged brain reward, Behav. Biol., 11 (1974) 275-279. Deutsch, J.A., Chisholm, D. and Mason, P.A., Adaptation to rewarding brain stimuli of differing amplitude, Behav. Neural Biol., 29 (1980) 359-364. Deutsch, J.A. and Hawkins, R.D., Adaptation as a cause of apparent aversiveness of prolonged rewarding brain stimulation, Behav. Biol., 7 (1972) 285-290. Deutsch, J.A., Roll, P.L. and Wetter, F., Choice between rewarding brain stimuli of differing length, Behav. Biol., 18 (1976) 369-377. Gerhardt, S. and Liebman, J.M., Self-regulation of ICSS duration: effects of anxiogenic substances, benzodiazepine antagonists and antidepressants, Pharmacol. Biochem. Behav., 22 (1985) 71-76. German, D.C. and Holloway, F.A., Behaviorally determined neurophysiological properties of MFB self-stimulation fibers, Physiol. Behav., 9 (1972) 823-829. Hodos, W., Motivational properties of long durations of rewarding brain stimulation, J. Comp. Physiol. Psychol., 59 (1965) 219-224. Hoebel, B.G., Brain-stimulation reward and aversion in relation to behavior. In A. Wauquier and E.T. Rolls (Eds.), Brain-Stimulation Reward, Elsevier, New York, 1976, pp. 335-372. Keesey, R.E., Duration of stimulation and the reward properties of hypothalamic stimulation, J. Comp. Physiol. Psychol., 58 (1964) 201-207. K6nig, J.F.R. and Klippel, R.A., The Rat Brain. A Stereotaxic Atlas of the Forebrain and Lower Parts of the Brainstem, Williams and Wilkins, Baltimore, 1963. Levitt, R.A., Stillwell, D.J. and Evers, T.M., Morphine and shuttlebox self-stimulation in the rat: tolerance studies, Pharmacol. Biochem. Behav., 9 (1978) 567-569. Liebman, J.M., Anxiety, anxiolytics and brain stimulation reinforcement, Neurosci. Biobehav. Rev., 9 (1985) 75-86. Liebman, J.M.. Gerhardt, S. and Prowse, J., Differential

effects of D-amphetamine. pipradrol and buproprion on shuttlebox self-stimulation, Pharrnacol. Biochem. Behav., 16 (1982) 791-794. 24 Margules, D.L., Separation of positive and negative reinforcing systems in the diencephalon of the rat, Amer. Y. Psychol., 79 (1966) 205-216. 25 Mendelson, J. and Freed, W.J., Do rats terminate hypothalamic stimulation only in order to turn it on again? Behav. Biol., 8 (1973) 619-628. 26 Miller, N.E., Experiments on motivation, Science, 126 (1957) 1271-1278. 27 Miller, P.J., An investigation of the neural substrates of reward and aversion in the rat., Doctoral dissertation, Univesity of Pennsylvania, Philadelphia, PA, 1981. 28 Olds, J., Satiation effects in self-stimulation of the brain, Y. Comp. Physiol. Psychol., 51 (1958) 675-678. 29 Olds, J., Differential effects of drives and drugs on selfstimulation at different brain sites. In D,F. Sheer (Ed.), Electrical Stimulation of the Brain, Austin University Press, Austin, 1961, pp. 350-366. 30 Olds, M.E., Facilitory action of diazepam and chlordiazepoxide on hypothalamic reward behavior, Y. Cornp. Physiol. Psychol., 62 (1966) 136-140. 31 Olds, M.E., Effectiveness of morphine and ineffectiveness of diazepam and phenobarbital on the motivational properties of hypothalamic self-stimulation behavior, Neuropharmacology, 15 (1976) 117-131. 32 Olds, M.E. and Gardner, M., Effects of diazepam and phenobarbital on self-stimulation in posterior hypothalamic and preoptic regions and on the thermoregulatory responses to rewarding brain stimulation, Neuropharmacology, 15 (1976) 103-115. 33 Olds. M.E. and Olds, J., Approach-avoidance analysis of rat diencephalon, Y. Comp. Neurol., 120 (1963) 259-283. 34 Panksepp, J., Hypothalamic integration of behavior. In P.J. Morgane and J. Panksepp (Eds.), Handbook of the Hypothalamus, Marcel Dekker, New York, 1981. 35 Roberts, W., Rapid escape learning without avoidance learning motivated by hypothalamic stimulation in cats, Y. Comp. Physiol. Psychol. , 51 (1958) 391-399. 36 Shizgal, P., Bielajew, C., Corbett, D., Skelton, R.W. and Yeomans, J., Behavioral methods for inferring anatomical linkage between rewarding brain stimulation sites, Y. Comp. Physiol. Psychol., 94 (1980) 227-237. 37 Shizgal, P. and Matthews, G., Electrical stimulation of the rat diencephalon: differential effects of interrupted stimulation on ON- and OFF-responding, Brain Research, 129 (1977) 319-333. 38 Skelton, R.W. and Shizgal, P., Parametric analysis of ONand OFF-responding for hypothalamic stimulation, Physiol. Behav., 25 (1980) 699-706. 39 Soper, W.Y. and Wise, R.A., Hypothalamically induced eating: eating from 'non-eaters' with diazepam, T.-L-T. Y. Life Sci., 1 (1971) 79-84. 40 Stein, L., An analysis of stimulus-duration preference in self-stimulation of the brain, J. Comp. Physiol. Psychol., 55 (1962) 405-414. 4l Valenstein, E.S. and Valenstein, T., Interaction of positive and negative reinforcing neural systems, Science, 145 (1964) 1456-1458. 42 Watson, P.J., Short, M.A., Huenink, G.L. and Hartman, D.F., Diazepam effects on hypothalamically elicited drinking and eating, Physiol. Behav., 24 (1980) 39-44. 43 Wise, R.A., Action of drugs of abuse on brain reward systems, Pharmacol. Biochem. Behav., 13, Suppl. 1 (1980) 213-223.