- Email: [email protected]
A role for hippocampus in the utilization of hunger signals
BEHAVIORAL AND NEURAL BIOLOGY 59, 167--171 (1993)
BRIEF REPORT A Role for Hippocampus in the Utilization of Hunger Signals T. L. DAVIDSON Purdue University AND LEONARD E . JARRARD 1
Washington and Lee University
attention to the hippocampus as a substrate for learning and memory in humans and other species. However, more recent evidence indicates that H.M. also suffered deficits which do not appear to rely on memory processes. These deficits were similar to what N a u t a (1971) described in patients with frontal-lobe damage as "interoceptive agnosia." Similar to some frontal lobe damaged patients, H.M. appeared unable to detect or process certain types of information arising from his internal milieu. For example, H.M. had great difficulty identifying his state of food deprivation/satiation (Hebben, Corkin, Eichenbaum, & Shedlack, 1985). When asked to rate his degree of hunger or satiety on a numeric scale H.M. was as likely to rate himself as hungry after a meal as he was immediately before eating. Since H.M. could not remember the meal itself (due to his severe amnesia), his difficulty in this regard indicates that he was also unable to use internal hunger cues as signals for whether or not he had recently eaten. The importance of this finding with respect to hippocampal function is difficult to assess given that H.M. also had damage to medial temporal structures in addition to hippocampus. The purpose of the present experiments was to examine the utilization of hunger state cue information by rats with hippocampal damage that was much more selective than the damage sustained by H.M. Selective damage was achieved through the use of an ibotenate lesion technique described previously by J a r r a r d (e.g., 1989). Briefly, eight naive male S p r a g u e - D a w l e y rats, about 90 days old and weighing about 300 g
The hippocampus is generally regarded as an important anatomical substrate for learning and memory (e.g., Eichenbaum, Otto, & Cohen, Behavioral and Neural Biology, 57, 2-36, 1992; Squire, Psychological Review, 99, 195-231, 1992). In the present research, we provide evidence that the hippocampus is also involved with another function--utilization of hunger state signals. Rats with selective ibotenate lesions of the hippocampus were found to be impaired in their ability to discriminate between the interoceptive sensory consequences of food deprivation and satiation. At the same time the ability of these rats to discriminate between different exteroceptive cues was unaffected. These results suggest that deficits in discriminative performance were specific to interoceptive state stimuli. In addition, hippocampal-damaged rats also seemed unable to use their food deprivation stimuli as signals to engage in normal feeding behavior. Our results argue that although the hippocampus may be important for learning and memory processes, it also deserves consideration as a neural substrate for the regulation of food intake and perhaps other functions which involve interoceptive signals. © 1993 Academic Press, Inc. Perhaps the best known example of a memory deficit following damage to the hippocampal formation is the now classic case of H.M., a patient who suffered severe anterograde amnesia following bilateral damage to hippocampus and adjacent medial temporal lobe structures (Scoville & Milner, 1957). The discovery of H.M.'s impairment attracted 1 This research was supported by g r a n t s from NIH (to T.L.D.) and from NSF (to L.E.J.). Address correspondence and r e p r i n t requests to Terry L. Davidson, D e p a r t m e n t of Psychological Sciences, Purdue University, West Lafayette, IN 47907. 167
0163-1047/93 $5.00 Copyright © 1993 by Academic Press, Inc. All rights of reproduction in any form reserved.
168
DAV1DSON AND JARRARD
FIG. 1. Photomicrographs of horizontal, cresyl violet-stained sections for an unoperated animal (top) and a rat with ibotenate hippocampal lesions (bottom). Note that in the ibotenate-lesioned rat the pyramidal cells in the hippocampus and most of the granule cells in dentate gyrtm have been removed and are replaced by glial cells. Scale bar = 1 mm.
at the beginning of the experiment, were anesthetized with injections of a mixture of chloral hydrate and sodium pentobarbital. The rats were placed in a stereotaxic instrument and the bone overlying the hippocampus was removed. Rats in Group IBO received focal injections ofibotenic acid (concentration of 10 mg/ml; volumes of .05 and .10/~1) at 26 different sites (13 on each side of hippocampus). With this technique, damage was limited to hippocampus (i.e., dentate gyrus, pyramidal cells), while fibers of passage and adjacent hippocampal formation structures (e.g., entorhinal cortex, subiculum) were intact. The nature and extent of the resulting brain damage obtained in the present research can be seen in Fig. 1. Additionally, eight rats served as controls (five underwent operations similar to that of hippocampals except that no ibotenic acid (IBO) was injected, and three served as unoperated controls). The effects of these lesions on the utilization of state cue information was assessed by requiring the
rats to solve a discrimination problem based on cues arising from different degrees of food deprivation. Training involved placing each rat in a conditioning box for six, 4-min sessions. Half the rats in each group (IBO and controls) received shock (0.5 s, 0.9 mA) at the end of each session when they were 24h food deprived but not when they were 0-h food deprived. The remaining rats received the reversed hours of food deprivation-shock contingency. Following three training sessions with each food deprivation cue, the rats were tested for discriminative responding during 12 extinction sessions (i.e., no shocks were delivered), six under each deprivation level. All training and test sessions took place during the first hour of the light period of the possessive rats 12:12 h light: dark cycle. Conditioned freezing (skeletal muscle immobility) was the index of discrimination learning. Incidence of freezing was obtained from videotapes of behavior during test sessions. The tapes were scored using a time-sampling technique (each rat observed once every 10 s during each session), by observers who were unaware of the treatment histories of the rats. Previous research showed that rats exhibit more conditioned freezing to a food deprivation cue that signals shock than to one signaling the absence of shock (e.g., Davidson, 1987), and that this discrimination develops following as few as three reinforced trials (Davidson, Flynn, & Jarrard, 1992). Figure 2A presents discrimination ratios for IBO lesioned (IBO) and control rats (CON). No differences between operated and unoperated controls were obtained, thus their results were combined. Discrimination ratios for Group IBO were near .50 throughout testing indicating no discrimination. In contrast, much better deprivation discrimination performance was shown by Group CON as their discrimination ratios exceeded .70 by the completion of testing. Differences between the groups were statistically significant over the last two blocks of testing (throughout this report, the reliability of differences between treatments was evaluated using analysis of variance and a priori tests with alpha level was set a t p < .05). Thus, discrimination based on food deprivation intensity stimuli was impaired, relative to controls, following ibotenate lesions of hippocampns. What was the basis of Group IBO's impairment? The impairment might have been a consequence of a general (a) deficit in learning ability, (b) change in fear motivation, or (c) reduction in the effectiveness of the shock reinforcer. We evaluated these hypotheses in a second experiment that compared discrimination based on food deprivation cues with
169
HIPPOCAMPUS AND HUNGER CUES
r.
0.7
m
/
-I-
< /
0.6
o
0.5
i
CON
A
IBO
i
Deprivotion Cue c
•
III-___AjA i
c,-
in/i
Auditory Cue
B
0.7
A
"to
. -G9 E3
0.6
0.5 i
i
i
i
1 2 3 1 Blocks of Two '% ond Two " - " Sessions
FIG. 2. Food deprivation intensity discrimination performance for rats with ibotenate lesions of hippocampus (IBO) and their controls (CON) is shown. 2A shows performance for Experiment 1, whereas 2B shows performance for Experiment 2. The leftmost panel of 2B depicts discrimination based on food deprivation intensity cues, whereas the rightmost panel depicts performance based on auditory cues. Discrimination ratios were obtained using the formula A/(A + B), where A represents percentage observations of freezing under the shocked deprivation level and B represents percentage observation of freezing under the nonshocked deprivation level. Ratio values near .50 indicate no discrimination, whereas values increasingly higher than .50 indicate greater discriminative performance.
that based on auditory stimuli. Rats (eight IBO lesioned, eight operated, and seven unoperated controls) of the same description as those in Experiment 1 were trained as described for that experiment except that auditory cues (1500 Hz tone or 3 Hz clicker) were presented in compound with food deprivation cues throughout each session. The identity of the auditory cues presented with the shocked and nonshocked food deprivation cues was counterbalanced within each group. Learning about each type
of cue (i.e., deprivation or auditory) was tested separately in extinction. Figure 2B compares the groups with respect to discriminative control by deprivation cues and by auditory cues. By the end of deprivation cue testing, Group CON (again combining operated and unopcrated controls) showed clear discriminative responding, whereas Group IBO did not. Discrimination ratios for deprivation cue testing differed significantly between the two groups on the last test block. In contrast, there was a nonsignificant difference between the groups during the auditory cue test. Thus, the results of Experiment 2 not only replicated the findings of impaired deprivation intensity discrimination reported for hippocampal rats in Experiment 1, but extended those findings by showing that hippocampals were not impaired relative to controls in utilizing auditory cues. This latter finding indicates that Group IBO's impaired deprivation cue discrimination was not the result of a general deficit in learning ability, a change in fear motivation, or a reduction in the effectiveness of the shock reinforcer. The hippocampal rats learned about auditory cues and this learning was based on the same motivational system and reinforcer as was learning about deprivation intensity cues.
The above results are consistent with the view that rats without a hippocampus are impaired in using information provided by their hunger state. This interpretation is also supported by the data presented in Table 1, which shows homecage ad libitum feeding behavior and general activity for both groups in Experiment 2 on Days 11 and 12 postsurgery, i.e., prior to the beginning of discrimination training. These data were obtained from six rats each in Groups IBO and CON (three operated and three unoperated controls). The behavior of each rat was monitored once every 5 s, 23 h a day by an
TABLE 1 C o m p a r i s o n of Eating, Activity, a n d Freezing: E x p e r i m e n t 2 Homecage eating and activity postsurgery days 11 and 12
Group CON HIP
Mean percentage freezing last block of testing
Mean total food contactsa
Mean amount consumed (g)
Mean total activity countsb
Shocked dep level
Nonshocked dep level
124.3 205.3*
25.1 27.3
3205.8 4460.3*
56.2 55.5
32.5 47.4*
a Food hopper contacts were indexed by the number of changes in the electrical conductance of a wire mesh grid which the rats must touch to gain access to powdered laboratory chow. b Activity was monitored with an ultrasonic recording device. * Reliably different from Group CON (ps < .05).
170
DAVIDSON AND JARRARD
automated system. For all rats, feeding during this phase of the experiment was mostly confined to the dark period of the light:dark cycle. Table 1 shows that while rats in Group IBO ate about the same amount of food as controls, the hippocampals had significantly more contacts with the food hopper. In addition, general activity in the homecage during this period was also greater for Group IBO than for Group CON. However, it is important to note that increased activity was not observed when food was absent, i.e., in the apparatus during discrimination testing. Table 1 also shows that Group IBO froze more (i.e., was less active) than Group CON under the nonshocked level of food deprivation during the last block of deprivation cue testing. This pattern of results indicates that Groups IBO's impaired food deprivation intensity discrimination performance was not due to impairment in freezing behavior (a potential consequence of heightened general activity). It is not surprising that hippocampal lesions resulted in increased food contacts without increased food consumption if one assumes that hippocampal rats are unable to use information provided by their internal food deprivation cues to anticipate when feeding will be followed by rewarding postingestive consequences. Without these signals, rats may be less able to anticipate the consequences of their appetitive behavior prior to actually engaging in that behavior. Increased contact with food without increased intake suggests that hippocampal rats may be forced to rely more on stimuli associated directly with food (e.g., the taste or smell of food) than on stimuli arising from their deprivation state to determine the consequences of feeding. They sample food more often to obtain this information, yet since their energy needs are unchanged the amount they actually consume remains about the same. From this perspective, just as hippocampal rats are impaired in using food deprivation signals to anticipate the occurrence of shock, they are also impaired in using these state cues to anticipate when feeding will result in positive reinforcement. How the hippocampus might receive information pertaining to metabolic state is not clear. It is possible that the structure is a processing site for metabolic or hormonal signals detected elsewhere; however, the neural pathways that would bring this information to the hippocampus are unknown. It is also possible that signals of energy need are detected more directly by the hippocampus. For example, insulin receptors exist in rat brain and potential routes for transport of peripheral insulin to deep brain structures (e.g., cerebrospinal fluid, and via
endothelial cells of brain microvessels) have been identified (see Baskin, Figlewicz, Woods, Porte, & Dorsa, 1987). Furthermore, the concentration of insulin receptors in hippocampus is high relative to other brain structures--possibly in the brain second only to that of the olfactory bulb (Zahniser, Goens, Hanaway, & Vinych, 1984). One could speculate that insulin plays a role in signaling metabolic need and the hippocampus plays a role in detecting those signals. The impairments shown by our hippocampal rats could also be part of a larger class of deficits. Recent evidence suggests that the hippocampus mediates context learning (e.g., Selden, Everitt, Jarrard, & Robbins, 1991). If food deprivation states can be regarded as contextual cues (e.g., Hirsch, Leber, & Gillman, 1978), then our results might be based on a general deficit in learning about contextual events (also see Hsaio & Isaacson, 1971). However, this interpretation of our data is complicated by the finding that our rats learned about auditory cues, which were similar to contextual cues in that they were present throughout each training session. It is also possible that food deprivation cues are but one example of a class of interoceptive stimuli which hippocampal rats are unable to process. However, our findings also indicate that the hippocampus is not involved with the processing of all types of state information, since our hippocampal rats appeared to acquire conditioned fear as well as controls. In summary, our results suggest that the hippocampus may be important not only for learning and memory but also for the regulation of feeding. With respect to feeding, our data indicate that an intact hippocampus is required if rats are to utilize state cues arising from their condition of food deprivation. REFERENCES Baskin, D. G., Figlewicz, D. P., Woods, S. C., Porte, D., & Dorsa, D. M. (1987). Insulin in the brain. Annual Review of Physiology, 49, 335-347. Davidson, T. L. (1987). Learning about deprivation intensity stimuli. Behavioral Neuroscience, 101, 198-208. Davidson, T. L., Flynn, F. W., & Jarrard, L. E. (1992). Potency of food deprivation intensity cues as discriminative stimuli. Journal of Experimental Psychology: Animal Behavior Processes, 18, 174-181. Eichenbaum, H., Otto, T., & Cohen, N. J. (1992). The hippocampus--What does it do? Behavioral and Neural Biology, 57, 2-36. Hebben, N., Corkin, S., Eichenbaum, H., & Shedlack, K. (1985). Diminished ability to interpret and report internal states after bilateral medial temporal resection: Case H.M. Behavioral Neuroscience, 99, 1031-1039. Hirsch, R., Leber, B., & Gillman, K. (1978). Fornix fibers and
HIPPOCAMPUS AND HUNGER CUES motivational states as controllers of behavior: A study stimulated by contextual retrieval theory. Behavioral Biology, 22, 463-478. Hsiao, S., & Isaacson, R. L. (1971). Learning of food and water positions by hippocampus damaged rats. Physiology and Behavior, 6, 81-83. Jarrard, L. E. (1989). On the use of ibotenic acid to lesion selectively different components of the hippocampal formation. Journal of Neuroscience Methods, 29, 251-259. Nauta, W. J. H. (1971). The problem of the frontal lobe: A reinterpretation. Journal of Psychiatric Research, 8, 167-187. Scoville, W. B, & Milner, B. (1957). Loss of recent memory after
171
bilateral hippocampal lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 20, 11-12. Selden, N. R. W., Everitt, B. J., Jarrard, L. E., & Robbins, T. W. (1991). Complementary roles for the amygdala and hippocampus in aversive conditioning to explicit and contextual cues. Neuroscience, 42, 335-350. Squire, L. R. (1992). Memory and the hippocampus: A synthesis from findings with rats, monkeys, and humans. Psychological Review, 99, 195-231. Zahniser, N. H., Goens, M. B., Hanaway, P. J., & Vinych, J. V. (1984). Characterization and regulation of insulin receptors in rat brain. Journal of Neurochemistry, 42, 13541362.
BRIEF REPORT A Role for Hippocampus in the Utilization of Hunger Signals T. L. DAVIDSON Purdue University AND LEONARD E . JARRARD 1
Washington and Lee University
attention to the hippocampus as a substrate for learning and memory in humans and other species. However, more recent evidence indicates that H.M. also suffered deficits which do not appear to rely on memory processes. These deficits were similar to what N a u t a (1971) described in patients with frontal-lobe damage as "interoceptive agnosia." Similar to some frontal lobe damaged patients, H.M. appeared unable to detect or process certain types of information arising from his internal milieu. For example, H.M. had great difficulty identifying his state of food deprivation/satiation (Hebben, Corkin, Eichenbaum, & Shedlack, 1985). When asked to rate his degree of hunger or satiety on a numeric scale H.M. was as likely to rate himself as hungry after a meal as he was immediately before eating. Since H.M. could not remember the meal itself (due to his severe amnesia), his difficulty in this regard indicates that he was also unable to use internal hunger cues as signals for whether or not he had recently eaten. The importance of this finding with respect to hippocampal function is difficult to assess given that H.M. also had damage to medial temporal structures in addition to hippocampus. The purpose of the present experiments was to examine the utilization of hunger state cue information by rats with hippocampal damage that was much more selective than the damage sustained by H.M. Selective damage was achieved through the use of an ibotenate lesion technique described previously by J a r r a r d (e.g., 1989). Briefly, eight naive male S p r a g u e - D a w l e y rats, about 90 days old and weighing about 300 g
The hippocampus is generally regarded as an important anatomical substrate for learning and memory (e.g., Eichenbaum, Otto, & Cohen, Behavioral and Neural Biology, 57, 2-36, 1992; Squire, Psychological Review, 99, 195-231, 1992). In the present research, we provide evidence that the hippocampus is also involved with another function--utilization of hunger state signals. Rats with selective ibotenate lesions of the hippocampus were found to be impaired in their ability to discriminate between the interoceptive sensory consequences of food deprivation and satiation. At the same time the ability of these rats to discriminate between different exteroceptive cues was unaffected. These results suggest that deficits in discriminative performance were specific to interoceptive state stimuli. In addition, hippocampal-damaged rats also seemed unable to use their food deprivation stimuli as signals to engage in normal feeding behavior. Our results argue that although the hippocampus may be important for learning and memory processes, it also deserves consideration as a neural substrate for the regulation of food intake and perhaps other functions which involve interoceptive signals. © 1993 Academic Press, Inc. Perhaps the best known example of a memory deficit following damage to the hippocampal formation is the now classic case of H.M., a patient who suffered severe anterograde amnesia following bilateral damage to hippocampus and adjacent medial temporal lobe structures (Scoville & Milner, 1957). The discovery of H.M.'s impairment attracted 1 This research was supported by g r a n t s from NIH (to T.L.D.) and from NSF (to L.E.J.). Address correspondence and r e p r i n t requests to Terry L. Davidson, D e p a r t m e n t of Psychological Sciences, Purdue University, West Lafayette, IN 47907. 167
0163-1047/93 $5.00 Copyright © 1993 by Academic Press, Inc. All rights of reproduction in any form reserved.
168
DAV1DSON AND JARRARD
FIG. 1. Photomicrographs of horizontal, cresyl violet-stained sections for an unoperated animal (top) and a rat with ibotenate hippocampal lesions (bottom). Note that in the ibotenate-lesioned rat the pyramidal cells in the hippocampus and most of the granule cells in dentate gyrtm have been removed and are replaced by glial cells. Scale bar = 1 mm.
at the beginning of the experiment, were anesthetized with injections of a mixture of chloral hydrate and sodium pentobarbital. The rats were placed in a stereotaxic instrument and the bone overlying the hippocampus was removed. Rats in Group IBO received focal injections ofibotenic acid (concentration of 10 mg/ml; volumes of .05 and .10/~1) at 26 different sites (13 on each side of hippocampus). With this technique, damage was limited to hippocampus (i.e., dentate gyrus, pyramidal cells), while fibers of passage and adjacent hippocampal formation structures (e.g., entorhinal cortex, subiculum) were intact. The nature and extent of the resulting brain damage obtained in the present research can be seen in Fig. 1. Additionally, eight rats served as controls (five underwent operations similar to that of hippocampals except that no ibotenic acid (IBO) was injected, and three served as unoperated controls). The effects of these lesions on the utilization of state cue information was assessed by requiring the
rats to solve a discrimination problem based on cues arising from different degrees of food deprivation. Training involved placing each rat in a conditioning box for six, 4-min sessions. Half the rats in each group (IBO and controls) received shock (0.5 s, 0.9 mA) at the end of each session when they were 24h food deprived but not when they were 0-h food deprived. The remaining rats received the reversed hours of food deprivation-shock contingency. Following three training sessions with each food deprivation cue, the rats were tested for discriminative responding during 12 extinction sessions (i.e., no shocks were delivered), six under each deprivation level. All training and test sessions took place during the first hour of the light period of the possessive rats 12:12 h light: dark cycle. Conditioned freezing (skeletal muscle immobility) was the index of discrimination learning. Incidence of freezing was obtained from videotapes of behavior during test sessions. The tapes were scored using a time-sampling technique (each rat observed once every 10 s during each session), by observers who were unaware of the treatment histories of the rats. Previous research showed that rats exhibit more conditioned freezing to a food deprivation cue that signals shock than to one signaling the absence of shock (e.g., Davidson, 1987), and that this discrimination develops following as few as three reinforced trials (Davidson, Flynn, & Jarrard, 1992). Figure 2A presents discrimination ratios for IBO lesioned (IBO) and control rats (CON). No differences between operated and unoperated controls were obtained, thus their results were combined. Discrimination ratios for Group IBO were near .50 throughout testing indicating no discrimination. In contrast, much better deprivation discrimination performance was shown by Group CON as their discrimination ratios exceeded .70 by the completion of testing. Differences between the groups were statistically significant over the last two blocks of testing (throughout this report, the reliability of differences between treatments was evaluated using analysis of variance and a priori tests with alpha level was set a t p < .05). Thus, discrimination based on food deprivation intensity stimuli was impaired, relative to controls, following ibotenate lesions of hippocampns. What was the basis of Group IBO's impairment? The impairment might have been a consequence of a general (a) deficit in learning ability, (b) change in fear motivation, or (c) reduction in the effectiveness of the shock reinforcer. We evaluated these hypotheses in a second experiment that compared discrimination based on food deprivation cues with
169
HIPPOCAMPUS AND HUNGER CUES
r.
0.7
m
/
-I-
< /
0.6
o
0.5
i
CON
A
IBO
i
Deprivotion Cue c
•
III-___AjA i
c,-
in/i
Auditory Cue
B
0.7
A
"to
. -G9 E3
0.6
0.5 i
i
i
i
1 2 3 1 Blocks of Two '% ond Two " - " Sessions
FIG. 2. Food deprivation intensity discrimination performance for rats with ibotenate lesions of hippocampus (IBO) and their controls (CON) is shown. 2A shows performance for Experiment 1, whereas 2B shows performance for Experiment 2. The leftmost panel of 2B depicts discrimination based on food deprivation intensity cues, whereas the rightmost panel depicts performance based on auditory cues. Discrimination ratios were obtained using the formula A/(A + B), where A represents percentage observations of freezing under the shocked deprivation level and B represents percentage observation of freezing under the nonshocked deprivation level. Ratio values near .50 indicate no discrimination, whereas values increasingly higher than .50 indicate greater discriminative performance.
that based on auditory stimuli. Rats (eight IBO lesioned, eight operated, and seven unoperated controls) of the same description as those in Experiment 1 were trained as described for that experiment except that auditory cues (1500 Hz tone or 3 Hz clicker) were presented in compound with food deprivation cues throughout each session. The identity of the auditory cues presented with the shocked and nonshocked food deprivation cues was counterbalanced within each group. Learning about each type
of cue (i.e., deprivation or auditory) was tested separately in extinction. Figure 2B compares the groups with respect to discriminative control by deprivation cues and by auditory cues. By the end of deprivation cue testing, Group CON (again combining operated and unopcrated controls) showed clear discriminative responding, whereas Group IBO did not. Discrimination ratios for deprivation cue testing differed significantly between the two groups on the last test block. In contrast, there was a nonsignificant difference between the groups during the auditory cue test. Thus, the results of Experiment 2 not only replicated the findings of impaired deprivation intensity discrimination reported for hippocampal rats in Experiment 1, but extended those findings by showing that hippocampals were not impaired relative to controls in utilizing auditory cues. This latter finding indicates that Group IBO's impaired deprivation cue discrimination was not the result of a general deficit in learning ability, a change in fear motivation, or a reduction in the effectiveness of the shock reinforcer. The hippocampal rats learned about auditory cues and this learning was based on the same motivational system and reinforcer as was learning about deprivation intensity cues.
The above results are consistent with the view that rats without a hippocampus are impaired in using information provided by their hunger state. This interpretation is also supported by the data presented in Table 1, which shows homecage ad libitum feeding behavior and general activity for both groups in Experiment 2 on Days 11 and 12 postsurgery, i.e., prior to the beginning of discrimination training. These data were obtained from six rats each in Groups IBO and CON (three operated and three unoperated controls). The behavior of each rat was monitored once every 5 s, 23 h a day by an
TABLE 1 C o m p a r i s o n of Eating, Activity, a n d Freezing: E x p e r i m e n t 2 Homecage eating and activity postsurgery days 11 and 12
Group CON HIP
Mean percentage freezing last block of testing
Mean total food contactsa
Mean amount consumed (g)
Mean total activity countsb
Shocked dep level
Nonshocked dep level
124.3 205.3*
25.1 27.3
3205.8 4460.3*
56.2 55.5
32.5 47.4*
a Food hopper contacts were indexed by the number of changes in the electrical conductance of a wire mesh grid which the rats must touch to gain access to powdered laboratory chow. b Activity was monitored with an ultrasonic recording device. * Reliably different from Group CON (ps < .05).
170
DAVIDSON AND JARRARD
automated system. For all rats, feeding during this phase of the experiment was mostly confined to the dark period of the light:dark cycle. Table 1 shows that while rats in Group IBO ate about the same amount of food as controls, the hippocampals had significantly more contacts with the food hopper. In addition, general activity in the homecage during this period was also greater for Group IBO than for Group CON. However, it is important to note that increased activity was not observed when food was absent, i.e., in the apparatus during discrimination testing. Table 1 also shows that Group IBO froze more (i.e., was less active) than Group CON under the nonshocked level of food deprivation during the last block of deprivation cue testing. This pattern of results indicates that Groups IBO's impaired food deprivation intensity discrimination performance was not due to impairment in freezing behavior (a potential consequence of heightened general activity). It is not surprising that hippocampal lesions resulted in increased food contacts without increased food consumption if one assumes that hippocampal rats are unable to use information provided by their internal food deprivation cues to anticipate when feeding will be followed by rewarding postingestive consequences. Without these signals, rats may be less able to anticipate the consequences of their appetitive behavior prior to actually engaging in that behavior. Increased contact with food without increased intake suggests that hippocampal rats may be forced to rely more on stimuli associated directly with food (e.g., the taste or smell of food) than on stimuli arising from their deprivation state to determine the consequences of feeding. They sample food more often to obtain this information, yet since their energy needs are unchanged the amount they actually consume remains about the same. From this perspective, just as hippocampal rats are impaired in using food deprivation signals to anticipate the occurrence of shock, they are also impaired in using these state cues to anticipate when feeding will result in positive reinforcement. How the hippocampus might receive information pertaining to metabolic state is not clear. It is possible that the structure is a processing site for metabolic or hormonal signals detected elsewhere; however, the neural pathways that would bring this information to the hippocampus are unknown. It is also possible that signals of energy need are detected more directly by the hippocampus. For example, insulin receptors exist in rat brain and potential routes for transport of peripheral insulin to deep brain structures (e.g., cerebrospinal fluid, and via
endothelial cells of brain microvessels) have been identified (see Baskin, Figlewicz, Woods, Porte, & Dorsa, 1987). Furthermore, the concentration of insulin receptors in hippocampus is high relative to other brain structures--possibly in the brain second only to that of the olfactory bulb (Zahniser, Goens, Hanaway, & Vinych, 1984). One could speculate that insulin plays a role in signaling metabolic need and the hippocampus plays a role in detecting those signals. The impairments shown by our hippocampal rats could also be part of a larger class of deficits. Recent evidence suggests that the hippocampus mediates context learning (e.g., Selden, Everitt, Jarrard, & Robbins, 1991). If food deprivation states can be regarded as contextual cues (e.g., Hirsch, Leber, & Gillman, 1978), then our results might be based on a general deficit in learning about contextual events (also see Hsaio & Isaacson, 1971). However, this interpretation of our data is complicated by the finding that our rats learned about auditory cues, which were similar to contextual cues in that they were present throughout each training session. It is also possible that food deprivation cues are but one example of a class of interoceptive stimuli which hippocampal rats are unable to process. However, our findings also indicate that the hippocampus is not involved with the processing of all types of state information, since our hippocampal rats appeared to acquire conditioned fear as well as controls. In summary, our results suggest that the hippocampus may be important not only for learning and memory but also for the regulation of feeding. With respect to feeding, our data indicate that an intact hippocampus is required if rats are to utilize state cues arising from their condition of food deprivation. REFERENCES Baskin, D. G., Figlewicz, D. P., Woods, S. C., Porte, D., & Dorsa, D. M. (1987). Insulin in the brain. Annual Review of Physiology, 49, 335-347. Davidson, T. L. (1987). Learning about deprivation intensity stimuli. Behavioral Neuroscience, 101, 198-208. Davidson, T. L., Flynn, F. W., & Jarrard, L. E. (1992). Potency of food deprivation intensity cues as discriminative stimuli. Journal of Experimental Psychology: Animal Behavior Processes, 18, 174-181. Eichenbaum, H., Otto, T., & Cohen, N. J. (1992). The hippocampus--What does it do? Behavioral and Neural Biology, 57, 2-36. Hebben, N., Corkin, S., Eichenbaum, H., & Shedlack, K. (1985). Diminished ability to interpret and report internal states after bilateral medial temporal resection: Case H.M. Behavioral Neuroscience, 99, 1031-1039. Hirsch, R., Leber, B., & Gillman, K. (1978). Fornix fibers and
HIPPOCAMPUS AND HUNGER CUES motivational states as controllers of behavior: A study stimulated by contextual retrieval theory. Behavioral Biology, 22, 463-478. Hsiao, S., & Isaacson, R. L. (1971). Learning of food and water positions by hippocampus damaged rats. Physiology and Behavior, 6, 81-83. Jarrard, L. E. (1989). On the use of ibotenic acid to lesion selectively different components of the hippocampal formation. Journal of Neuroscience Methods, 29, 251-259. Nauta, W. J. H. (1971). The problem of the frontal lobe: A reinterpretation. Journal of Psychiatric Research, 8, 167-187. Scoville, W. B, & Milner, B. (1957). Loss of recent memory after
171
bilateral hippocampal lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 20, 11-12. Selden, N. R. W., Everitt, B. J., Jarrard, L. E., & Robbins, T. W. (1991). Complementary roles for the amygdala and hippocampus in aversive conditioning to explicit and contextual cues. Neuroscience, 42, 335-350. Squire, L. R. (1992). Memory and the hippocampus: A synthesis from findings with rats, monkeys, and humans. Psychological Review, 99, 195-231. Zahniser, N. H., Goens, M. B., Hanaway, P. J., & Vinych, J. V. (1984). Characterization and regulation of insulin receptors in rat brain. Journal of Neurochemistry, 42, 13541362.