Physiology and Behavior, Vol. 14, pp. 133-142. Brain Research Publications Inc., 1975. Printed in the U.S.A.
Hypothalamic Influences on Sensory Reinforcement JAMES E. ACKIL 2 , MOSHE J. LEVISON AND GABRIEL P. FROMMER
Department o f Psychology, Indiana University, Bloomington, Indiana 4 7401
(Received 30 January 1974) ACKIL, J. E., M. J. LEVISON AND G. P. FROMMER. Hypothalamic influences on sensory reinforcement. PHYSIOL. BEHAV. 14(2) 133-142, 1975. - Rats were tested in a darkened chamber containing two levers, one of which turned on a dim light when depressed. Rats receiving non-contingent electrical stimulation of the lateral hypothalamus made many more lever presses than did unstimulated controls and made more presses on the light producing lever than on the inactive one. Pretests had shown that the electrical stimulation elicited consummatory behavior in the presence of appropriate goal objects. Dynamic hyperphagi¢ rats maintained on a restricted diet made more total responses early in testing than did their controls which also received the same restricted diet. They also preferred the light producing lever, but not reliably more than did their controls. Static hyperphagic rats made fewer total responses than did their controls, both groups receiving food ad lib. Neither group showed any preference between the two levers. Lateral hypothalamus Semory reinforcement
Ventromedial hypothalamus
Hyperphagia
SENSORY reinforcement refers to the increase in probability or strength of a response that is followed by the presentation or removal of some neutral stimulus of moderate intensity [18]. Fowler [11] has attempted through incentive motivation theory to relate this phenomenon to conventional positive reinforcement based on organic need conditions. He argued that all stimuli have incentive properties, that incentive properties of stimuli can be related to preference as measured by the approaches animals make to them, and that incentive properties of stimuli are varied by specific antedating conditions such as deprivation and habituation. According to incentive motivation theory as developed by a number of workers (e.g., [3,35] ) a drive state such as hunger functions to increase the incentive properties of environmental stimuli. It does so by making the animal more responsive to stimuli, especially to those relevant to the prevailing drive state. Tapp [38] has presented data for sensory reinforcement that are consistent with this interpretation. He showed that at least under some experimental conditions, deprivation produces or enhances sensory reinforcement, apparently through its facilitation of the animal's reactivity to sensory stimuli produced by its own behavior. The purpose of the present experiments was to determine- whether direct interventions in the hypothalamus that increase consummatory behavior would produce effects on sensory reinforcement similar to those produced by depri-
Stimulation induced consummatory behavior
vation. Recent interpretations of hypothalamic function in motivation and reinforcement provide ample reason to suggest they might. These interpretations have emphasized a mode of action for this structure that in many respects parallels the features of incentive motivation theory. Stimulation of the lateral hypothalamus appears to induce integrated consummatory behaviors by making the animal more responsive to environmental stimuli, especially those that lead toward the goal objects appropriate to the central motive state that is induced (e.g., [3, 10, 16, 28, 36, 42]). However, the simple correspondence of these induced states to conventional motivational labels such as hunger or thirst or to particular groups of movement patterns has been challenged [40,41]. Behavioral effects of lateral hypothalamic lesions have been accounted for in terms of decreased responsiveness to environmental stimuli [8, 22, 39]. Many features of the syndrome following ventromedial hypothalamic lesions have been interpreted to result from the animal's increased reactivity to environmental stimuli because the lateral hypothalamus is released from inhibition from the ventromedial nucleus (e.g., [ 15, 19, 26, 27, 32, 34] ). In one of the experiments that follow, non-contingent electrical stimulation in the lateral hypothalamus was found to facilitate responding more on a lever which when depressed turned on a dim light than on a lever that produced no external environmental change. The same stimula-
'Supported by PHS Grants MH-10852 and MH-16046. Experiment 1 was taken from part of a dissertation submitted by J. E. Ackil in partial fuLfillment of the requL,ements for the Ph.D. degree, Indiana University, 1972. Portions of the data were reported at the Annual Meeting of the M i d ~ Psychological Association, Detroit, May, 1971. We thank Virginia Wasser for assistance in Experiment 2. Present address: Dept. of Psychology, Western Illinois University, Macomb, IL 61455. 133
134 tion induced consummatory behavior when tested in the presence of appropriate goal objects. In the other experiment, hungry rats in the dynamic phase of hyperphagia following ventromedial hypothalamic lesions tended to make more lever responses at the start of testing than did hungry control animals irrespective of the light contingency. Both groups preferred the light producing lever. Non-deprived rats in the static phase of hyperphagia consistently responded less on both light producing and inactive levers than did their non-deprived controls, and neither group showed a lever perference. EXPERIMENT 1: EFFECT OF HYPOTHALAMICSTIMULATION ON RESPONDING FOR LIGHT ONSET Method Animals. Eleven male albino rats were used. They were individually housed under continuous illumination with food and water provided ad lib. Electrode implantation. Rats were anesthetized with sodium pentobarbital (50 mg/kg, i.p.). Bipolar electrodes made of stainless steel wire, 175 v in dia. and insulated with Teflon except for the cross section area of the tips, were stereotaxically aimed at the lateral hypothalamus on both sides in all animals. The coordinates referred to bregma (top of skull horizontal, [ 14] ) were 3.0 to 3.5 mm posterior, 1.3 to 1.6 mm lateral, and 8.4 to 9.0 mm ventral. Electrodes were fixed in place with dental cement and terminated in miniature contacts [13]. Nitrofurazone and butacaine sulfate ointment (Furacin) was applied to the closed wound margin, and 80,000 units of benzathin penicillin (Bicillin) were injected intramuscularly. Apparatus. The implanted rats were screened for stimulation induced behavior in a 30 × 15 cm box, one 61 cm high side wall of which was made of clear plastic and the other three of black painted wood. The chamber contained food pellets and wood blocks strewn about the hardware cloth floor and a drinking tube filled with water at one end. Testing for sensory reinforcement was carried out in I of 2 identical chambers housed in ventilated refrigerator shells for light and sound attenuation. Each chamber was 30 cm wide, 30 cm long, and 60 cm high, and had a grid floor and open top. Three of the walls were made of sheet aluminum, while the fourth was made of clear plastic and served as a door. Two identical levers made of sheet aluminum were mounted 3 cm above the floor on the middle of opposite side wails of the chambers. Each lever was 7.5 cm long, 5.0 cm wide, and 0.3 cm thick and required approximately 10 g of force to operate the attached microswitch. A light (28 V, 0.48 W) was mounted in a 2.5 cm dia. green lamp jewel 3 cm above each lever. The lights were wired in series with a 56 ohm resistor and produced approximately 0.4 ft-c as measured by a photometer at a point equidistant from the two levers. The experiment was controlled by conventional solid state programming equipment, and lever presses were recorded on electromechanical counters. Electrical stimulat i o n c o n s i s t e d o f biphasic rectangular pulse pairs ( 1 0 - 1 5 0 uA peak amplitude, 0.5 msec pulse width, 60 Hz repetition rate) delivered from a solid state constant current source and monitored with an oscilloscope across a 1-kohm resistor in series with the animal. Procedure. Prior to testing for sensory reinforcement and one week after electrode implantation each animal was screened for stimulation induced behaviors of eating, drink-
ACKIL, LEVISON AND FROMMER ing, and gnawing by a method similar to that used by Valenstein et al. [41]. Stimulation was alternately turned on for 30 sec and off for 60 sec in the presence of the food pellets, drinking tube, and wood blocks. Stimulation intensity was initially 10 gA and was increased by l0 gA steps on successive stimulation periods until the animal displayed stimulation induced eating, drinking, gnawing, or what was judged to be a disorganized aversive reaction to stimulation. If no stimulation induced consummatory behavior was observed at one electrode site, the other was tested. If consummatory behavior was observed, the animal was given 5 more 30-sec stimulation tests at that site and intensity. Those animals that displayed stimulation induced behaviors were placed in the experimental brain stimulation group on subsequent testing; the others served as unstimulated controls. In all subsequent tests the animals in the stimulation group were stimulated only at one site and at the lowest current (50 or 100~A) which reliably induced consummatory behavior in the pretest for stimulation induced behavior. Animals were tested 1 hr/day for 16 consecutive days. On the first 8 days of testing (Sessions 1 - 8 ; light contingency, stimulation phase), pressing one lever turned on the light above it in the otherwise darkened chamber for the duration of the lever press; pressing the other lever had no effect on chamber illumination. The position of the light producing lever was counterbalanced between animals, but remained constant for each animal. Those rats which exhibited a stimulation bound behavior in the screening procedure received unilateral brain stimulation for alternate 30-sec periods throughout each of these 8 sessions. Conditions for control animals were identical except that they were not stimulated. For the following 4 days (Sessions 9 - 1 2 ; no light, stimulation phase) animals were tested in the same manner, except that neither lever turned on a light. Finally, animals were tested for four days (Sessions 1 3 - 1 6 ; no light , no stimulation phase), with neither lever effective and with neither group receiving brain stimulation. At no time in this experiment did the delivery of electrical stimulation depend on the animals' behavior. At the end of experimentation the locations of the deepest penetration of the electrodes were determined from photographic enlargements made from unstained 40 u thick frozen sections as well as directly from sections stained with cresyl violet. Results Five rats exhibited some form of stimulation induced behvior: 3 ate, 1 drank, and 1 both ate and drank when stimulated at one or at either site during the screening. These animals composed the experimental group which received brain stimulation in later tests. None of the other 6 animals displayed clear stimulation induced consummatory behavior during the screening, and they served as unstimulated controls. Figure 1 shows the daily median number of responses on both levers for both groups over all three test phases. In the first 8 days of testing (light contingency, stimulation phase) brain stimulated rats made many more responses on both levers than did control animals. The experimental animals also responded more on the lever that turned on the light than on the inactive lever. The control animals showed little or no difference in rate on the t w o levers, except on Day 1, when they made more responses on the light lever (t = 2.94,
HYPOTHALAMUS AND SENSORY REINFORCEMENT
135
LIGHT CONTINGENCY, STIMULATION
300-
NO LIGHT, NO LIGHT, STIMULATIONNO STIMULATION
~
I .
/
Stimulated
/~
~ ~
Control
Gr°ul~P Gr°-uP(~ LightLever
.-.....,ooc,v..v.,
1 |
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16
SESSION FIG. 1. Median number of responses made by brain stimulated and control animals on light producing lever and on inactive lever plotted as a function of daily 1-hr sessions. In this and subsequent figures open symbols represent responses on lever that turned on or had previously turned on light; f'flled symbols represent responses on inactive lever; squares connected by solid lines and circles connected by broken lines represent experimental and control groups respectively.
df = 5, p
Hypothalamic Influences on Sensory Reinforcement JAMES E. ACKIL 2 , MOSHE J. LEVISON AND GABRIEL P. FROMMER
Department o f Psychology, Indiana University, Bloomington, Indiana 4 7401
(Received 30 January 1974) ACKIL, J. E., M. J. LEVISON AND G. P. FROMMER. Hypothalamic influences on sensory reinforcement. PHYSIOL. BEHAV. 14(2) 133-142, 1975. - Rats were tested in a darkened chamber containing two levers, one of which turned on a dim light when depressed. Rats receiving non-contingent electrical stimulation of the lateral hypothalamus made many more lever presses than did unstimulated controls and made more presses on the light producing lever than on the inactive one. Pretests had shown that the electrical stimulation elicited consummatory behavior in the presence of appropriate goal objects. Dynamic hyperphagi¢ rats maintained on a restricted diet made more total responses early in testing than did their controls which also received the same restricted diet. They also preferred the light producing lever, but not reliably more than did their controls. Static hyperphagic rats made fewer total responses than did their controls, both groups receiving food ad lib. Neither group showed any preference between the two levers. Lateral hypothalamus Semory reinforcement
Ventromedial hypothalamus
Hyperphagia
SENSORY reinforcement refers to the increase in probability or strength of a response that is followed by the presentation or removal of some neutral stimulus of moderate intensity [18]. Fowler [11] has attempted through incentive motivation theory to relate this phenomenon to conventional positive reinforcement based on organic need conditions. He argued that all stimuli have incentive properties, that incentive properties of stimuli can be related to preference as measured by the approaches animals make to them, and that incentive properties of stimuli are varied by specific antedating conditions such as deprivation and habituation. According to incentive motivation theory as developed by a number of workers (e.g., [3,35] ) a drive state such as hunger functions to increase the incentive properties of environmental stimuli. It does so by making the animal more responsive to stimuli, especially to those relevant to the prevailing drive state. Tapp [38] has presented data for sensory reinforcement that are consistent with this interpretation. He showed that at least under some experimental conditions, deprivation produces or enhances sensory reinforcement, apparently through its facilitation of the animal's reactivity to sensory stimuli produced by its own behavior. The purpose of the present experiments was to determine- whether direct interventions in the hypothalamus that increase consummatory behavior would produce effects on sensory reinforcement similar to those produced by depri-
Stimulation induced consummatory behavior
vation. Recent interpretations of hypothalamic function in motivation and reinforcement provide ample reason to suggest they might. These interpretations have emphasized a mode of action for this structure that in many respects parallels the features of incentive motivation theory. Stimulation of the lateral hypothalamus appears to induce integrated consummatory behaviors by making the animal more responsive to environmental stimuli, especially those that lead toward the goal objects appropriate to the central motive state that is induced (e.g., [3, 10, 16, 28, 36, 42]). However, the simple correspondence of these induced states to conventional motivational labels such as hunger or thirst or to particular groups of movement patterns has been challenged [40,41]. Behavioral effects of lateral hypothalamic lesions have been accounted for in terms of decreased responsiveness to environmental stimuli [8, 22, 39]. Many features of the syndrome following ventromedial hypothalamic lesions have been interpreted to result from the animal's increased reactivity to environmental stimuli because the lateral hypothalamus is released from inhibition from the ventromedial nucleus (e.g., [ 15, 19, 26, 27, 32, 34] ). In one of the experiments that follow, non-contingent electrical stimulation in the lateral hypothalamus was found to facilitate responding more on a lever which when depressed turned on a dim light than on a lever that produced no external environmental change. The same stimula-
'Supported by PHS Grants MH-10852 and MH-16046. Experiment 1 was taken from part of a dissertation submitted by J. E. Ackil in partial fuLfillment of the requL,ements for the Ph.D. degree, Indiana University, 1972. Portions of the data were reported at the Annual Meeting of the M i d ~ Psychological Association, Detroit, May, 1971. We thank Virginia Wasser for assistance in Experiment 2. Present address: Dept. of Psychology, Western Illinois University, Macomb, IL 61455. 133
134 tion induced consummatory behavior when tested in the presence of appropriate goal objects. In the other experiment, hungry rats in the dynamic phase of hyperphagia following ventromedial hypothalamic lesions tended to make more lever responses at the start of testing than did hungry control animals irrespective of the light contingency. Both groups preferred the light producing lever. Non-deprived rats in the static phase of hyperphagia consistently responded less on both light producing and inactive levers than did their non-deprived controls, and neither group showed a lever perference. EXPERIMENT 1: EFFECT OF HYPOTHALAMICSTIMULATION ON RESPONDING FOR LIGHT ONSET Method Animals. Eleven male albino rats were used. They were individually housed under continuous illumination with food and water provided ad lib. Electrode implantation. Rats were anesthetized with sodium pentobarbital (50 mg/kg, i.p.). Bipolar electrodes made of stainless steel wire, 175 v in dia. and insulated with Teflon except for the cross section area of the tips, were stereotaxically aimed at the lateral hypothalamus on both sides in all animals. The coordinates referred to bregma (top of skull horizontal, [ 14] ) were 3.0 to 3.5 mm posterior, 1.3 to 1.6 mm lateral, and 8.4 to 9.0 mm ventral. Electrodes were fixed in place with dental cement and terminated in miniature contacts [13]. Nitrofurazone and butacaine sulfate ointment (Furacin) was applied to the closed wound margin, and 80,000 units of benzathin penicillin (Bicillin) were injected intramuscularly. Apparatus. The implanted rats were screened for stimulation induced behavior in a 30 × 15 cm box, one 61 cm high side wall of which was made of clear plastic and the other three of black painted wood. The chamber contained food pellets and wood blocks strewn about the hardware cloth floor and a drinking tube filled with water at one end. Testing for sensory reinforcement was carried out in I of 2 identical chambers housed in ventilated refrigerator shells for light and sound attenuation. Each chamber was 30 cm wide, 30 cm long, and 60 cm high, and had a grid floor and open top. Three of the walls were made of sheet aluminum, while the fourth was made of clear plastic and served as a door. Two identical levers made of sheet aluminum were mounted 3 cm above the floor on the middle of opposite side wails of the chambers. Each lever was 7.5 cm long, 5.0 cm wide, and 0.3 cm thick and required approximately 10 g of force to operate the attached microswitch. A light (28 V, 0.48 W) was mounted in a 2.5 cm dia. green lamp jewel 3 cm above each lever. The lights were wired in series with a 56 ohm resistor and produced approximately 0.4 ft-c as measured by a photometer at a point equidistant from the two levers. The experiment was controlled by conventional solid state programming equipment, and lever presses were recorded on electromechanical counters. Electrical stimulat i o n c o n s i s t e d o f biphasic rectangular pulse pairs ( 1 0 - 1 5 0 uA peak amplitude, 0.5 msec pulse width, 60 Hz repetition rate) delivered from a solid state constant current source and monitored with an oscilloscope across a 1-kohm resistor in series with the animal. Procedure. Prior to testing for sensory reinforcement and one week after electrode implantation each animal was screened for stimulation induced behaviors of eating, drink-
ACKIL, LEVISON AND FROMMER ing, and gnawing by a method similar to that used by Valenstein et al. [41]. Stimulation was alternately turned on for 30 sec and off for 60 sec in the presence of the food pellets, drinking tube, and wood blocks. Stimulation intensity was initially 10 gA and was increased by l0 gA steps on successive stimulation periods until the animal displayed stimulation induced eating, drinking, gnawing, or what was judged to be a disorganized aversive reaction to stimulation. If no stimulation induced consummatory behavior was observed at one electrode site, the other was tested. If consummatory behavior was observed, the animal was given 5 more 30-sec stimulation tests at that site and intensity. Those animals that displayed stimulation induced behaviors were placed in the experimental brain stimulation group on subsequent testing; the others served as unstimulated controls. In all subsequent tests the animals in the stimulation group were stimulated only at one site and at the lowest current (50 or 100~A) which reliably induced consummatory behavior in the pretest for stimulation induced behavior. Animals were tested 1 hr/day for 16 consecutive days. On the first 8 days of testing (Sessions 1 - 8 ; light contingency, stimulation phase), pressing one lever turned on the light above it in the otherwise darkened chamber for the duration of the lever press; pressing the other lever had no effect on chamber illumination. The position of the light producing lever was counterbalanced between animals, but remained constant for each animal. Those rats which exhibited a stimulation bound behavior in the screening procedure received unilateral brain stimulation for alternate 30-sec periods throughout each of these 8 sessions. Conditions for control animals were identical except that they were not stimulated. For the following 4 days (Sessions 9 - 1 2 ; no light, stimulation phase) animals were tested in the same manner, except that neither lever turned on a light. Finally, animals were tested for four days (Sessions 1 3 - 1 6 ; no light , no stimulation phase), with neither lever effective and with neither group receiving brain stimulation. At no time in this experiment did the delivery of electrical stimulation depend on the animals' behavior. At the end of experimentation the locations of the deepest penetration of the electrodes were determined from photographic enlargements made from unstained 40 u thick frozen sections as well as directly from sections stained with cresyl violet. Results Five rats exhibited some form of stimulation induced behvior: 3 ate, 1 drank, and 1 both ate and drank when stimulated at one or at either site during the screening. These animals composed the experimental group which received brain stimulation in later tests. None of the other 6 animals displayed clear stimulation induced consummatory behavior during the screening, and they served as unstimulated controls. Figure 1 shows the daily median number of responses on both levers for both groups over all three test phases. In the first 8 days of testing (light contingency, stimulation phase) brain stimulated rats made many more responses on both levers than did control animals. The experimental animals also responded more on the lever that turned on the light than on the inactive lever. The control animals showed little or no difference in rate on the t w o levers, except on Day 1, when they made more responses on the light lever (t = 2.94,
HYPOTHALAMUS AND SENSORY REINFORCEMENT
135
LIGHT CONTINGENCY, STIMULATION
300-
NO LIGHT, NO LIGHT, STIMULATIONNO STIMULATION
~
I .
/
Stimulated
/~
~ ~
Control
Gr°ul~P Gr°-uP(~ LightLever
.-.....,ooc,v..v.,
1 |
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16
SESSION FIG. 1. Median number of responses made by brain stimulated and control animals on light producing lever and on inactive lever plotted as a function of daily 1-hr sessions. In this and subsequent figures open symbols represent responses on lever that turned on or had previously turned on light; f'flled symbols represent responses on inactive lever; squares connected by solid lines and circles connected by broken lines represent experimental and control groups respectively.
df = 5, p
