| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Neuroscience Graduate Program, 2Ernest Gallo Clinic and Research Center, Departments of Neurology and Physiology, and Wheeler Center for the Neurobiology of Addiction, University of California, San Francisco, California
Submitted 21 March 2006; accepted in final form 1 November 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
). This general pattern also holds true for exploration that is not maintained by response-contingent rewards. Although more time is spent exploring previously rewarded locations, animals continue to explore unrewarded locations. This phenomenon has been extensively studied in the conditioned place preference (CPP) paradigm.
CPP is widely used in studies of reward and drug addiction (Tzschentke 1998). The conditioning procedure involves confining the animal to one room while it experiences a reward. On alternate days, the animal is confined to a distinctively different room after administering a control, such as saline. After conditioning, the animal is allowed to move freely between the rooms and the preference is measured as the increased time spent in the reward-paired room. Because CPP requires little training and is relatively simple to administer and score, it is by far the most common test of drug reward in rodents (Bardo and Bevins 2000; Bardo et al. 1995; Hoffman 1989; Tzschentke 1998).
Despite its widespread use and acceptance as a measure of reward, almost nothing is known about the discrete behaviors that mediate the increased time spent in the preferred location. The total time spent in a location is a consequence of the discrete actions affecting entry and departure from that location. During expression of CPP, animals engage in a variety of behaviors, including exploration of each room, grooming, rearing, and sniffing. Isolation of the relevant behaviors is required before one can begin to explain CPP at the level of neural activity.
To pinpoint the mode of expression, we divided each test session into a series of room visits. A room visit was defined as the period of time bounded by one entry and the following exit from a room. Summing the durations of all visits to a room gives the total time in that room. A preference could be expressed by either increasing the number of visits to the preferred room, or by increasing the visit duration in the preferred room, or both.
One hypothesis for the behavioral strategy used in CPP expression is visit duration timing. For example, increasing the duration of certain behaviors such as grooming bouts, in the preferred room, would increase the time spent in that room. Alternatively, visit duration could be controlled by a clock circuit in the brain that sets an optimal time to spend in one area before moving on (Matell and Meck 2000). Although this explanation could be realized in different ways, they all require that visit durations to each room have a Gaussian distribution, whose mean reflects the animal's intended visit duration for that location.
Animals may behave probabilistically when allocating time between reward sources (Glimcher 2002). Therefore an alternative hypothesis is that the rats alter their probability of making spontaneous location-directed transitions between specific rooms. This hypothesis predicts that the initiation of such transitions will be independent of other actions performed by the animal and of the time elapsed from room entry. This would result in an exponential distribution of visit durations. Although a change in the average visit duration could not distinguish whether the animal is using a visit duration timing strategy or a location-directed transition strategy, the shape of the distribution of visit durations will distinguish between a timed room visit duration (Gaussian) and a constant room transition probability (exponential).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The conditioned place preference (CPP) chamber was a rectangular Plexiglas box 84 x 46 cm with 45-cm-high walls. The long axis of the chamber was divided into three equal-size rooms (28 x 46 cm) using removable partitions. These partitions could be used to completely separate each room, or they could be left partly open, with 10-cm-wide doorways connecting each of the two end rooms with the center room. All the walls were painted in semigloss medium gray. Each room possessed a distinctive floor color and texture. For two groups of animals (n = 15), the central room floor was smooth, semitransparent acrylic and one side room floor was covered with a white plastic grid of 1-cm squares. The other side room floor was covered with a black plastic grid of 2-cm squares. A third group (n = 7) had a different floor arrangement with the white 1-cm grid and smooth acrylic for the side rooms and a semitransparent crystal-patterned acrylic for the center room floor. All groups had similar preferences. The entire CPP chamber was enclosed in a sound-insulated cabinet that was lit from above by a 100-W incandescent light. Within this cabinet, the CPP chamber was mounted on a sheet of semitransparent acrylic, allowing the rat's shadow to be recorded by a CCD camera mounted below the chamber. The animal's movements were digitized (5-Hz sampling rate) using Ethovision software (Noldus Information Technology, Leesburg, VA).
Experimental subjects
In all, 22 male rats were used: 15 Long–Evans and seven Wistar (there was no effect of strain in this study). Subjects weighed 250–300 g on arrival from the vendor (Harlan, Indianapolis, IN). The animals were housed individually and were handled and habituated to the lab environment for 1 wk before experiments began.
Behavioral paradigm
Each rat was randomly assigned to one of the two side rooms for morphine pairing. After assignment, on day 1, each animal was placed in the chamber for a 15-min pretest. During the pretest, both doors were open and the rats were free to explore all three rooms. The rats initially showed no bias for either side. One rat was removed from the study after it spent >80% of the pretest time in one room. On 4 of the next 8 days, the animals received 10 mg/kg subcutaneous injections of morphine and were immediately confined to one of the side rooms for 40 min. On alternate days, they received pairings of saline vehicle injections with the opposite side room for 40 min. After conditioning the animals were again tested in the open chamber for 15 min. To remove odors, the floors were wiped with 70% ethanol between sessions.
Data analysis
Off-line analysis of the behavior was conducted in Matlab (The MathWorks, Natick, MA). Much of our analysis focused on behavioral epochs we termed visits, defined as the time between an entry and the following exit of a room. Each time an animal entered a room, it was counted as a separate visit. The rooms that the animal occupied at the start and end of a session were not counted as visits because there was no entry or exit, respectively. Visit analysis began after the animal made its first room transition in a session and ended when it completed its last voluntary transition before the end of the session. Visit frequency and duration were calculated for each of the three rooms. We also divided these results into 3-min bins to observe the time course of the behavior. Goodness of fit between center room visit duration distributions and predicted summed exponential curves was quantified as the correlation coefficient between the log transform of the data and the log transform of the calculated exponential curves.
We constructed two models to examine constant probability room transitions as a description of CPP. Constants for the two models were calculated using the number of transitions between rooms i and j (denoted by nij) and the total duration of all the visits to a room (Ti). Each second was treated as a discrete time bin and the probability of moving from room i at time t – 1 (denoted by rt–1 = i) to room j at time t (denoted by rt = j) was calculated as
 | (1) |
 | (2) |
Computer simulations of CPP
The model simulations consisted of a room variable that indicated which of the three rooms the animal was located in. For the physical model simulation, the room variable could change only to an adjacent room. At each time point (corresponding to 1 s in the CPP box), the room variable was updated according to the probability that it would move to another room as given by the transition constants of the model. These transition constants were the same values calculated from the animal experiments. A random number between zero and one was generated for each time interval. If the random number was less than or equal to the probability of moving to the adjacent room, then the room variable was changed to indicate that the animal had moved into that room. If the random number was greater than the transition probability, then the room variable remained unchanged.
When there were two possible transitions from a room, such as from the center room, or from the side rooms in the location-directed model simulation, the random number was generated and if it was less than the combined probability of moving to either room, then the animal was moved depending on the ratio of the probability for each room. For example, if the simulation rat was in room one and it had a probability of moving to room two of P12 and a probability of moving to room three of P13, then if the random number was between zero and P12, it moved to room two. If the random number was between P12 and (P12 + P13) then the rat moved to room three; and if the random number was between (P12 + P13) and one, then the rat did not move.
In the location-directed model, whenever the rat made a simple transition directly between the morphine and saline rooms, the room variable was set to the center room for 5 s and then set to the destination value. The 5 s was added to the center room time to represent the mean time of a simple center room visit. Simulation outputs were formatted exactly like the real animal data, allowing subsequent analysis on the simulation data to be conducted using the same programs.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
Visit analysis confirmed that there was no pretest difference between rooms for either frequency or duration of visits. After conditioning, visit frequency to the morphine room increased significantly [Fig. 2 A; F(3,63) = 4.32, P = 0.008, RM ANOVA, P < 0.05 for all pairwise comparisons to test morphine]. Meanwhile, visit duration to the saline room significantly decreased after conditioning [Fig. 2D; F(3,63) = 5.59, P = 0.002, RM ANOVA; P < 0.05 for all pairwise comparisons to test saline]. This large decrease in saline room visit duration recurred in many of the results of our analysis, as we will show below.
|
To examine the stability of the preference over time we divided the test session into five bins of 3-min duration. The percentage of time in each room during the pretest was not affected by elapsed time. During the test, there was a significant main effect of room [Fig. 3B; F(1,42) = 114.31, P < 0.0001]. However, there was no significant effect of time [F(4,210) = 0.3, P = 0.88] and no interaction of room x time [F(4,210) = 1.06, P = 0.38] on the percentage of time the animals spent in each room.
|
Room visit durations were stable, showing no significant effect of elapsed time during either pretest or test (Fig. 3, E and F). During the test, morphine and saline room visit durations differed significantly [F(1,42) = 12.35, P = 0.0006]. Although there appears to be a sudden drop in visit duration during the last 3 min of the session (Fig. 3F), this trend is not significant, as revealed by a lack of interaction between the visit duration and elapsed time [room x elapsed time F(4,210) = 2.41, P = 0.051]. Therefore the morphine versus saline room differences in visit frequency and duration were present at the earliest times (within the first 3 min) of the trial and continued for the entire 15-min duration of the test session.
We next examined the distribution of visit durations for each of the three rooms during the pretest and test sessions (Fig. 4). For comparison, visit duration distributions are plotted along with an exponential or a Gaussian curve, using the mean and SD of the actual visit durations for each room as parameters. The distribution of visit durations had a clearly exponential shape. The exponential distribution is commonly found in physical systems that trigger events at random intervals with a fixed probability. This finding strongly suggests that the observed room durations are a function of a constant room exit probability. Therefore room duration reflects the instantaneous probability of exiting a room per unit time in that room, rather than a desired visit duration determined by an internal or behavioral clock. This exit probability determines the average time between events, but is independent of elapsed time since the immediately preceding event (i.e., room entry). This indicates that during the expression of CPP, the rats do not use the memory of how long they have been in a given room after entry. They are just as likely to exit immediately on entering as they are several minutes after entering.
|
We formalized this observation by creating a probabilistic model of CPP. We converted the two variables of visit duration and visit frequency into the more general variable of transition probability. The probabilistic nature of the rat's transitions means that visit duration and visit frequency can be recast in terms of a probabilistic model of CPP transitions. This model consists of three states, representing the three rooms (Fig. 5A). The rat occupies one state at any time. During each unit of time, the rat has a certain probability of moving into another state. These transition probabilities (or transition coefficients) were directly calculated from the visit frequencies and visit durations for each room (see METHODS).
|
|
A useful analogy for the physical model is that of the equilibrium equation of a chemical reaction. The significantly increased probability of the animal moving from the saline room to the center room results in the animal spending more time in the center room. As a consequence of being in the center room for a longer period of time with a constant probability for exit, the animal will transition more often into the morphine room, thereby increasing time there as well. This is similar to adding reagent to an intermediate chemical state, thus increasing secondary products. Nevertheless, this result seemed odd because it suggested that the animals did not develop a preference for the morphine-paired room, but only an aversion to the saline-paired room. This apparent paradox was resolved when the rat's movements were analyzed in greater detail.
Simple and complex center room visits
To further explore this issue, we plotted the actual path the animals took during each center room visit. Visual inspection suggested two distinct types of movement through the center room. Center room visits were classified as "simple" or "complex" in reference to the path the animal navigated through the room. Simple center room visits were movements that took the animal directly through the center room without stopping to explore it (Fig. 6A). In contrast, complex center room visits had a more circuitous, indirect path and more stops within the center room (Fig. 6B). We further analyzed these movements to test the hypothesis that complex visits resulted when the animal's intended destination was the center room, whereas simple visits occurred when the opposite side room was the animal's intended destination and the center room was the route between the side rooms.
|
Simple and complex center room visits also differ in duration. The mean time for simple center room "visits" (crossings) was 4.5 s and for complex visits it was 24.7 s. When we made a scatterplot of the duration of a center room visit with the distance from the central corridor, the two types appear as distinct clusters (Fig. 6, C and D). This bimodal visit duration is also visible in the plot of the visit duration distribution for the center room (Fig. 4). Here, the experimental data were well fit by an exponential curve that is the weighted sum of the two exponential curves for simple and complex visit durations (pretest r = 0.9697; test r = 0.9812). In contrast, if there was only one type of movement through the center room, we would expect the data to be better fit by a single exponential defined by the average duration of all visits together (mean = 14.7 s; pretest r = 0.9444; test r = 0.9612). The single-exponential curve does not fit the data as well as the summed exponential predicted by the two visit types. These data support the hypothesis that the brief simple visits represent the time it takes the rat to cross the center room when the opposite room is the intended destination, whereas the long complex visits reflect the time spent exploring the center room as a destination.
A third critical distinction between visit types was found by comparing the entry door and exit door of a center room visit. Simple visits usually resulted in the animal leaving the center room through the door opposite to the one through which it entered. For example, if the animal entered from the morphine-paired side, it was most likely to exit into the saline side. The opposite door was chosen on 79% of simple visits to the center room during the pretest. In contrast, complex visit exits were evenly split between the two end rooms; 52% of exits were to the room opposite the entry room on the pretest. A chi-square analysis of exit directions (entry door or opposite door) finds a significant difference between the simple and complex exits (P < 0.0005), but no difference in the exits on the pretest and test sessions, indicating that these two visit types represent different motor programs that retain their distinctive character after conditioning. This is consistent with the interpretation that simple visits are routes to the opposite room, whereas complex visits are intended to end in the center room. During the pretest, once the animal was in the center room on a complex visit, it was equally likely to go to either side room because it was not heading toward one side room at the time that it entered the center room.
In summary, simple and complex center room visits are distinctly different room transitions distinguished by path complexity, deviation from the central corridor, visit duration, and opposite door exit probability. One possibility is that the goal of each transition type is distinct. The center room itself is the intended destination for complex transitions, whereas the opposite room is the goal in simple transitions. We explored this hypothesis by explicitly modeling these two types of transitions as independent variables in a revised version of our room transition model.
Location-directed model transition coefficients
The probabilistic physical model accounted for movements only between physically adjacent rooms. This is equivalent to counting all visits to the center room as the same type of movement. Based on the distinction between simple and complex visits, we modified the model by adding two more transition constants to represent the simple transitions directly from the morphine to the saline-paired room and vice versa (Fig. 7A). This new location-directed model divides the rat's transitions based on the profile of its movement into visits that are directed toward the center room or toward the opposite side room. When the animal's center room visits were divided and the transition coefficients recalculated (see METHODS), we found that the largest effect of conditioning was a marked increase in the simple movements directly from the saline-paired room to the morphine-paired room (Fig. 7B; P = 0.004, paired t-test of pretest and test coefficients; see Table 2 for all P values). The only other significantly altered coefficient was a decrease in transitions from the center room into the saline room (P = 0.007). The location-directed model helps us to better understand the result of the first, physical model. The counterintuitive result that only transitions out of the saline room changed after conditioning can now be explained. The physical model did not distinguish between the two types of movements out of the saline room. The location-directed model, however, shows that it is primarily the movements destined to end in the morphine-paired room that were enhanced. The rats developed a preferential increase in their probability of entering the morphine-paired room from the saline room; however, once in the morphine-paired room, their exit probability was unchanged from preconditioning levels. This change accounts for the effect of conditioning on both the visit frequency and visit duration measures. When in the morphine-paired room, the animal's visit duration and destination choice are not altered by conditioning. However, after conditioning, whenever the animal is in the saline-paired room, it is more likely to transition directly into the morphine room. This leads to an increase in morphine room visit frequency relative to that of the other rooms. Similarly, the increased probability of saline room exits leads to a decrease in the average visit duration for that room.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
CPP is an interesting behavior from the perspective of learning theory (Carr et al. 1989). Although it requires an approach behavior during testing, such a behavior is never emitted nor reinforced during conditioning when reward is present. In fact, CPP displays features of classical conditioning, i.e., learning from temporally contiguous pairing of an unconditioned stimulus (US = reward) with a conditioned stimulus (CS = room cues). It is also subject to latent inhibition, extinction, and recovery (Bardo and Bevins 2000; Hoffman 1989; Tzschentke 1998). However, CPP falls outside of the standard model of classical conditioning, which specifies that a US will elicit an unconditioned response (UR). Because the morphine is present while the animal is confined to a single room, the unconditioned approach (UR) cannot be emitted during training. Indeed, motoric URs are not even required within the conditioning room because restrained rats can still learn a preference (Carr et al. 1988). Therefore the conditioned response (CR = place preference) is not necessarily predicted as a consequence of pairing a reward with room confinement.
This leaves open the possibility that the cues in the room become conditioned (secondary) reinforcers (Robbins and Everitt 2002). In our apparatus, the most salient difference between rooms was the floor texture. It is possible that contact with the previously morphine-paired floor becomes rewarding to the rats, even in the absence of morphine. A key prediction of the conditioned reinforcer proposal is that the animals will acquire preference behavior during testing because the reinforcing consequences of entering the room should elicit an increasing frequency and duration of visits. On the other hand, if as we found, the animals show a full expression of CPP in the first moments of the test, and no further development of a preference at later times, there is no evidence of a conditioned reinforcement process at work. An alternative hypothesis is that a stimulus–reward association, established during training, directly induces CPP expression without instrumental learning required during testing. In this model, stimuli acquire value through classical conditioning and, subsequently, these conditioned stimuli evoke innate approach responses as the CR. If approach behavior is an innate response to reward-associated stimuli, animals should, as we found, show a full preference at the beginning of the test session.
Three other studies have examined the time course of a preference during a single test session (Bardo et al. 1984; Bozarth 1987a; Mueller and Stewart 2000) and all found a full preference during the first minutes of testing. We also found no interaction between conditioning room and elapsed time, whether the effect was measured as percentage time in room, visit frequency, or visit duration (Fig. 3, A–F). Therefore CPP does not increase during the test session and there is no evidence that conditioned reinforcement of approach behavior contributes to CPP.
Reward learning during acquisition of CPP is independent of movement within or between rooms and can influence approach behavior without any further behavioral reinforcement. Thus expression of location (or context) preference appears to involve an innate approach behavior that is biased by previous stimulus–reward associations. This is an important perspective because we can divide CPP into two processes. The first is the formation of a classically conditioned reward–stimulus association. The second process appears to be an innate approach behavior that makes use of these learned associations. Ethologically, this may represent a class of approach behaviors, similar to innate defense behaviors (Dielenberg and McGregor 2001) or autoshaping (Robbins and Everitt 2002). Such an innate approach behavior may be a mechanism that contributes to the drives and cravings that sustain drug and alcohol addiction.
Transition initiation is independent of elapsed visit time
We found that the visit duration distribution is best fit by an exponential decay curve. A previous study (Krauth 1992) found that the assumption of an exponential distribution for visit durations passed a test for statistical validity. This allows us to reject one possible behavioral strategy used by the rat to increase its time in the preferred room. If the rat intended to stay in the room for a certain amount of time (preferred duration) or if the duration of the visit was determined by a highly stereotyped behavior, such as a grooming bout, then we would expect each visit to be roughly of the same mean duration. This type of behavior would have generated a Gaussian distribution of visit durations around a mean preferred duration. The exponential distribution that we observed is what we would expect from a random process with a time constant. Therefore the decision to change rooms is not influenced by how long the animal has been in the room, i.e., by a timing mechanism. Although room exits are random, mean visit duration is determined by instantaneous exit probabilities and can be quantified as a time constant. This time constant is different for each type of room transition.
Simple and complex visits reflect location-directed transitions
Further analysis showed that the rats made two types of movements in the center room: either a rapid crossing from one end room to the other or a much longer visit in which the center room was thoroughly explored. We classified these center room visits as simple or complex by the sole criterion of deviation from the central corridor between rooms. However, a number of differences emerged from this split. First, the mean duration of complex visits was fivefold longer than the mean simple duration. Second, the exit direction choice was different for the two movement types. Finally, the location-directed model of CPP dynamics showed that the major effect of conditioning was on simple transitions from saline to morphine rooms. To reconcile the random occurrence of the transitions with the location-directed nature of the transition itself, we propose that a transition is a brief, all-or-none location-directed motor act. A transition is not part of some general random meandering between rooms, nor is it dependent on any other behaviors occurring aside from those that constitute the discrete act of moving between rooms.
CPP measures probability of initiating location-directed transitions
We postulated two types of location-directed exits from the end rooms and calculated the transition probability constants to discover that the major change after CPP training was a marked increase in direct (simple) saline- to morphine-room transitions. The other significant change was a decrease in probability of entering the saline room from the center room. Interestingly, conditioning did not affect the morphine room visit duration or exit probability. There may be other effects of conditioning on behavior, such as conditioned locomotion or conditioned tolerance, although we propose that the major behavioral effect of CPP conditioning is a change in the probability of initiating a transition from the saline room to the morphine room.
Although the probability of initiating these movements is increased 76% after conditioning, the mean number of these events is still fewer than five transitions per 15 min for both the pretest and test sessions. Therefore the biggest behavioral contribution to CPP is hardly visible in terms of physical movements, but does appear as a robust increase in the instantaneous probability of initiating this class of movements when the rat is in the saline room. Furthermore, for most of the time that the animal is being scored during the test session, the animal's active behavior does not influence the learned preference times. Our model demonstrates that increasing the probability of a subset of relatively infrequent and very brief motor acts is sufficient to dramatically alter the aggregate time spent in each room.
A role for the saline room in CPP
The increased exit probability in the saline-paired room leads to a decrease in the average visit duration for that room. In particular, the rat is biased to make location-directed transitions directly to the morphine room. After conditioning the animals also make fewer entries into the saline room from the center room, although center to morphine room transitions are unchanged. Interestingly, there is no bias for entering either the saline or center room when the animal is in the morphine room, and the probability of leaving the morphine room is not altered by conditioning, perhaps because morphine is not experienced during the test session. Morphine and saline conditioning increased the rat's bias (incentive) to initiate movements to the morphine room but had no effect on its behavior once in that room. The animals have learned a reason to go there, but there is no reason to stay.
This robust change in transition probabilities while the animal is in the saline room is open to interpretation. For example, it suggests that an addict could become more behaviorally active in a nondrug-related context. However, if this were the case, we would expect behavior to be altered in the center room as well as the saline room. This raises the possibility that the saline room has become associated with the "absence of reward," whereas the center room has remained novel. If this were true, one prediction is that addicts may have a heightened craving when encountering cues or contexts explicitly associated with the absence of reward.
A completely different explanation is that the animals were responding more to the discomfort of the injection in the saline room than to the reward in the morphine room. We think this is unlikely. Evidence that the morphine-paired room was rewarding includes the animals’ preference for the morphine room as a transition destination and the increased total time in the morphine room. However, this ambiguity of interpretation is not unique to the present study and could be applied to most morphine CPP studies. A similar confound of anxiety reduction in the drug-paired room during CPP was previously raised and explored (Bozarth 1987b). Although this issue could be raised for most CPP studies, it is highlighted in the present study because of the transition probability results. This suggests that more widespread use of transition probability analysis could aid with the interpretation of CPP behavioral results.
Transition probability model and simulations
The biggest advance of our model is to provide the first description of CPP behavior on an instantaneous timescale. Traditional descriptions of CPP are in terms of complete sessions (
10 min) and have not provided any description of what the animal was doing at a given instant of time. As a result the animal's behavior can now be compared with short-duration neural events, such as neuronal firing. The computer simulations using our model and the subsequent analysis demonstrate that the model description, although simple, is able to re-create many features of the actual behavior, including aggregate time, visit frequency, and duration.
Currently, the model consists of six free parameters that are derived directly from the animal's observed behavior. A model with fewer variables may ultimately be sufficient to describe CPP expression and learning, although the experimental evidence does not support a particular abstraction for reducing the number of variables in our model. Further work may reveal clues as to what abstractions would reflect real behavioral or neural processes.
Implications for neural encoding of CPP expression
These behavioral findings have implications for understanding the neural circuit that is the proximal cause of biased exploratory behavior. Brain regions involved in CPP include the amygdala (Everitt et al. 1991; White and McDonald 1993), dorsal hippocampus (Ferbinteanu and McDonald 2001; Meyers et al. 2003), medial prefrontal cortex (Tzschentke and Schmidt 1999), and nucleus accumbens (Layer et al. 1993; Popik and Kolasiewicz 1999). Neurons in these regions encode information relevant to exploration guided by previous reward experience. For example, neurons in the amygdala encode reward anticipation (Pratt and Mizumori 1998; Schoenbaum et al. 1998). Dorsal hippocampal place cells encode location information (Jung et al. 1994). Medial prefrontal cortex neurons encode preference for expected outcomes (Schoenbaum and Roesch 2005; Tremblay and Schultz 1999). Nucleus accumbens neurons encode both relative reward value (Roitman et al. 2005; Taha and Fields 2005) and spatial cues (Lavoie and Mizumori 1994; Shibata et al. 2001). Further work is needed to elucidate how each of these structures contributes to biased exploration during CPP.
The present behavioral observations lead to the prediction that the neurons in this circuit encode the initiation of transitions to particular destinations. However, they do not need to encode the elapsed time in any room because the initiation of room transitions is stochastic. Furthermore, the persistent neural changes that result from conditioning may not be restricted to neuronal activations in and around the morphine room. The large effect on direct transitions from the saline room suggest the encoding of locations or movements relative to the saline room, but not in relation to the center room.
In conclusion, the appeal of CPP as a test for reward learning may lie partly in the intuitive understanding that a subject will spend more time in a location previously paired with reward. However, the details of the behavior and their implications for how the brain encodes reward-associated cues are unclear. Our analysis reveals that the effect of morphine conditioning is to increase the probability of a preexisting class of self-initiated directed movements. The identification of these discrete location-directed movements as the foundation of CPP provides a basis to more directly compare the extensive CPP literature with studies of other motivated behaviors and allows new predictions about its underlying neural mechanism.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: P. German, Ernest Gallo Clinic and Research Center, 5858 Horton Street, Suite 200, Emeryville, CA 94608 (E-mail: german{at}phy.ucsf.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Bardo MT, Miller JS, Neisewander JL. Conditioned place preference with morphine: the effect of extinction training on the reinforcing CR. Pharmacol Biochem Behav 21: 545–549, 1984.[ISI][Medline]
Bardo MT, Rowlett JK, Harris MJ. Conditioned place preference using opiate and stimulant drugs: a meta-analysis. Neurosci Biobehav Rev 19: 39–51, 1995.[CrossRef][ISI][Medline]
Bozarth MA. Conditioned place preference: a parametric analysis using systemic heroin injections. In: Methods of Assessing the Reinforcing Properties of Abused Drugs, edited by Bozarth MA. New York: Springer-Verlag, 1987a, p. xiv, 658.
Bozarth MA. (Editor). Methods of Assessing the Reinforcing Properties of Abused Drugs. New York: Springer-Verlag, 1987b.
Carr GD, Fibiger HC, Phillips AG. Conditioned place preference as a measure of drug reward. In: The Neuropharmacological Basis of Reward, edited by Liebman JM, Cooper SJ. New York: Clarendon Press, 1989, p. x
Carr GD, Phillips AG, Fibiger HC. Independence of amphetamine reward from locomotor stimulation demonstrated by conditioned place preference. Psychopharmacology (Berl) 94: 221–226, 1988.[Medline]
Dayan P, Balleine BW. Reward, motivation, and reinforcement learning. Neuron 36: 285–298, 2002.[CrossRef][ISI][Medline]
Dielenberg RA, McGregor IS. Defensive behavior in rats towards predatory odors: a review. Neurosci Biobehav Rev 25: 597–609, 2001.[CrossRef][ISI][Medline]
Everitt BJ, Morris KA, O'Brien A, Robbins TW. The basolateral amygdala-ventral striatal system and conditioned place preference: further evidence of limbic-striatal interactions underlying reward-related processes. Neuroscience 42: 1–18, 1991.[CrossRef][ISI][Medline]
Ferbinteanu J, McDonald RJ. Dorsal/ventral hippocampus, fornix, and conditioned place preference. Hippocampus 11: 187–200, 2001.[CrossRef][ISI][Medline]
Glimcher P. Decisions, decisions, decisions: choosing a biological science of choice. Neuron 36: 323–332, 2002.[CrossRef][ISI][Medline]
Hoffman DC. The use of place conditioning in studying the neuropharmacology of drug reinforcement. Brain Res Bull 23: 373–387, 1989.[CrossRef][ISI][Medline]
Jung MW, Wiener SI, McNaughton BL. Comparison of spatial firing characteristics of units in dorsal and ventral hippocampus of the rat. J Neurosci 14: 7347–7356, 1994.[Abstract]
Krauth J. Parametric analysis of sojourn times in conditioned place preference experiments. J Neurosci Methods 42: 219–227, 1992.[CrossRef][ISI][Medline]
Lavoie AM, Mizumori SJ. Spatial, movement- and reward-sensitive discharge by medial ventral striatum neurons of rats. Brain Res 638: 157–168, 1994.[CrossRef][ISI][Medline]
Layer RT, Uretsky NJ, Wallace LJ. Effects of the AMPA/kainate receptor antagonist DNQX in the nucleus accumbens on drug-induced conditioned place preference. Brain Res 617: 267–273, 1993.[CrossRef][ISI][Medline]
Matell MS, Meck WH. Neuropsychological mechanisms of interval timing behavior. Bioessays 22: 94–103, 2000.[CrossRef][ISI][Medline]
McClure SM, Daw ND, Montague PR. A computational substrate for incentive salience. Trends Neurosci 26: 423–428, 2003.[CrossRef][ISI][Medline]
Meyers RA, Zavala AR, Neisewander JL. Dorsal, but not ventral, hippocampal lesions disrupt cocaine place conditioning. Neuroreport 14: 2127–2131, 2003.[CrossRef][ISI][Medline]
Mueller D, Stewart J. Cocaine-induced conditioned place preference: reinstatement by priming injections of cocaine after extinction. Behav Brain Res 115: 39–47, 2000.[CrossRef][ISI][Medline]
Popik P, Kolasiewicz W. Mesolimbic NMDA receptors are implicated in the expression of conditioned morphine reward. Naunyn Schmiedebergs Arch Pharmacol 359: 288–294, 1999.[CrossRef][ISI][Medline]
Pratt WE, Mizumori SJ. Characteristics of basolateral amygdala neuronal firing on a spatial memory task involving differential reward. Behav Neurosci 112: 554–570, 1998.[CrossRef][ISI][Medline]
Robbins TW, Everitt BJ. Limbic-striatal memory systems and drug addiction. Neurobiol Learn Mem 78: 625–636, 2002.[CrossRef][ISI][Medline]
Roitman MF, Wheeler RA, Carelli RM. Nucleus accumbens neurons are innately tuned for rewarding and aversive taste stimuli, encode their predictors, and are linked to motor output. Neuron 45: 587–597, 2005.[CrossRef][ISI][Medline]
Schoenbaum G, Chiba AA, Gallagher M. Orbitofrontal cortex and basolateral amygdala encode expected outcomes during learning. Nat Neurosci 1: 155–159, 1998.[CrossRef][ISI][Medline]
Schoenbaum G, Roesch M. Orbitofrontal cortex, associative learning, and expectancies. Neuron 47: 633–636, 2005.[CrossRef][ISI][Medline]
Shibata R, Mulder AB, Trullier O, Wiener SI. Position sensitivity in phasically discharging nucleus accumbens neurons of rats alternating between tasks requiring complementary types of spatial cues. Neuroscience 108: 391–411, 2001.[CrossRef][ISI][Medline]
Taha SA, Fields HL. Encoding of palatability and appetitive behaviors by distinct neuronal populations in the nucleus accumbens. J Neurosci 25: 1193–1202, 2005.
Tremblay L, Schultz W. Relative reward preference in primate orbitofrontal cortex. Nature 398: 704–708, 1999.[CrossRef][Medline]
Tzschentke TM. Measuring reward with the conditioned place preference paradigm: a comprehensive review of drug effects, recent progress and new issues. Prog Neurobiol 56: 613–672, 1998.[CrossRef][ISI][Medline]
Tzschentke TM, Schmidt WJ. Functional heterogeneity of the rat medial prefrontal cortex: effects of discrete subarea-specific lesions on drug-induced conditioned place preference and behavioural sensitization. Eur J Neurosci 11: 4099–4109, 1999.[CrossRef][ISI][Medline]
White NM, McDonald RJ. Acquisition of a spatial conditioned place preference is impaired by amygdala lesions and improved by fornix lesions. Behav Brain Res 55: 269–281, 1993.[CrossRef][ISI][Medline]
This article has been cited by other articles:
|
|
 |
|
  P. W. German and H. L. Fields Rat Nucleus Accumbens Neurons Persistently Encode Locations Associated With Morphine Reward J Neurophysiol, March 1, 2007; 97(3): 2094 - 2106. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Beeler Should I Stay or Should I Go? Neural Substrates of Being in the Right Place at the Right Time. Focus on "How Prior Reward Experience Biases Exploratory Movements: A Probabilistic Model" J Neurophysiol, March 1, 2007; 97(3): 1878 - 1879. [Full Text] [PDF]
|
|
| ||||
