Home : Data and Research : Abstracts/Posters
These are posters presented by members of the Neurogenetics and Behavior Center at various conferences over the past few years. In order to view these posters, you must have Adobe Acrobat installed on your computer.
Please click on the abstract title to open a PDF version of the poster.
Examining the role of effort-related functions on reward palatability and incentive learning
Most reinforcement theories suggest that animals typically prefer to exert less effort rather than more in order to obtain food reward (e.g., Hull, 1943). Alternatively, it has been suggested that the degree of effort exerted for a particular commodity enhances the value attributed to it. The effects of effort on preference for a primary reinforcer were examined using a novel paradigm that required mice to respond on two separate levers for two different rewards, whereby each response results in delivery of a specific reward on a fixed-ratio-1 (FR-1 schedule). After acquisition, the response-reinforcement schedule was made gradually more strict for one lever (i.e., increasing to FR15; high-effort schedule), while the schedule for the other lever remains fixed at FR-1 (low-effort schedule). At test, mice showed preferential consumption of the reward that was associated with the high-effort schedule (Experiment 1). In Experiment 2, auditory cues were presented during reward delivery on the high- and low-effort schedules. Subsequently, the effectiveness of these cues to reinforce new nose-poke learning (i.e., conditioned reinforcement) was assessed. Under these conditions, mice showed a preference for the nose-poke associated with presentations of the high-effort cue. In Experiment 3, baseline consumption tests were initially conducted using either higher-caloric (e.g., 6% polycose solution) or lower-caloric (e.g., 2% polycose solution) solutions. Typically mice showed greater consumption of the high- vs. low-caloric reward. Mice were then trained to respond under high- or low-effort conditions for either the high- or low-caloric reward. At test, mice consumed significantly more of the low-caloric reward when it was associated with the high-effort schedule, but not when this reward was associated with the low-effort schedule. Finally, the dopamine antagonist Haloperidol was used to evaluate the pharmacology associated with this effect. Our results are consistent with the proposition that reward value is functionally related to the amount of effort required to obtain it.
A necessary role for Darpp-32 in incentive salience attribution
Dopamine and cyclic-AMP regulated phosphoprotein-32 (Darpp-32) is selectively enriched in dopamine-receptor expressing forebrain neurons, including medium spiny neurons of the nucleus accumbens and neurons of the central nucleus of the amygdala. When phosphorylated at Thr34, Darpp-32 acts as a potent inhibitor of protein phosphatase 1 (PP1) activity and subsequently as an amplifier of PKA-mediated signal transduction. This amplifying property of DARPP-32 is critical for dopamine-receptor mediated signal transduction, as well as for the actions of other neurotransmitters (e.g. glutamate) on neuronal function. In light of these anatomical and functional characteristics of Darpp-32, we hypothesized that Darpp-32 plays an important role in a form of incentive learning known to be dependent on dopamine and glutamate receptor activation, and the integrity of subregions of the accumbens and amygdala. Mice lacking Darpp-32 (KO) and wildtype (WT) controls were first trained on a Pavlovian conditioning task to associate an auditory cue with reward delivery (CS+) and then the ability of the CS+ to subsequently energize or potentiate ongoing instrumental responding was assessed – a phenomenon known as Pavlovian-to-instrumental transfer or PIT. Whilst Darpp-32 KO mice were capable of acquiring the simple cue-reward association, when subsequently tested for PIT, Darpp-32 KO mice failed to show the typical CS+ induced potentiation of instrumental responding that was seen in WT control mice. Interestingly, a parallel experiment revealed that the Darpp-32 associated deficit was highly selective to PIT as other aspects of reward learning (conditioned reinforcement) were unaffected by the Darpp-32 mutation. Our findings provide the first evidence for a role of the Darpp-32 signaling pathway, presumably within the accumbens and central nucleus of the amygdala, in a specific form of incentive learning involved in motivated behavior. Moreover, as PIT is thought to model aspects of psychological processes altered in a number of psychopathological conditions, such as schizophrenia and drug addiction, our findings provide new insights into the proposed role of Darpp-32 in these conditions.
Assessment of consumption and operant behavior in 3 strains of mice.
Consummatory reward-based learning protocols rely on the subject’s willingness to consume the solution/substance being offered. In the current study, we examined the preference for three types of solutions (sweet, starch, and fat) and the effectiveness of each solution as a reinforcer. We used three strains of mice commonly used to generate genetically modified mice (C57BL/6; C57, 129/SvImJ; 129, and an F1 hybrid strain derived from the crossing of the two; F1). Understanding such strain differences is crucial in the field of mouse genetics because a wide array of background strains are used to generate mutant mice. All training and testing occurred in 3 phases and all strains of mice were trained and tested with each taste solution (reward) in the order sucrose, polycose, lecithin. Firstly, in phase 1, to determine the palatability of each reward we measured the number of licks made and the number of entries into the food cup during 30 minutes free access to 1 of 5 concentrations of reward (0, 2.5, 5, 10, and 20% for sucrose and polycose; 0, 1.25, 2.5, 3.75, and 5% for lecithin) for a total of 5 sessions. Each session mice received one concentration of a particular reward, such that by the 5th session mice had received access to all reward concentrations. During the second phase the mice learned to press a lever to dispense the mid-range concentration of the reward. The second phase was carried out over 5 sessions. On completion, mice that failed to reach a criterion of 50 rewards per session were subsequently excluded from the final phase of testing. During the final phase each reward was tested for its effectiveness as a reinforcer under progressive ratio conditions. Our results are consistent with previous reports that there are strain differences in sucrose preference, however we did not observe any strain differences in preference for polycose or lecithin. Results from the progressive ratio procedure revealed strain differences for sucrose and lecithin, but not for polycose. The implications for neurogenetics of using specific taste solutions in consumption-based and instrumental-response-based measures across a range of genetic backgrounds are discussed.
Deficits in Outcome-Specific Devaluation of Instrumental Responding Following Cannabinoid CB1 Receptor Knockout
Previous studies, including our own, suggest that CB1 receptor activation plays an important modulatory role in appetitive and consummatory behaviors for food and non-food rewards. Thus, reward sensitivity is altered by pharmacological and genetic targeting of CB1 receptor, as indicated by changes in instrumental responding for reward and consumption once reward is available. Additionally, we have found that the normal ability of Pavlovian reward cues to modulate appetitive behavior (conditioned reinforcement and Pavlovian-to-instrumental transfer) is disrupted in cannabinoid receptor knockout mice (Crombag et al. SFN 2007). These results raise questions about the nature of the deficits in reward representation in CB1 knockout mice. Representations of a reward involve both sensory and motivational properties of reinforcement, and cues may form associations with both of these features of a reward. Here we used an outcome-specific devaluation procedure to examine whether the deficits previously observed after CB1 receptor deletion may be due to an inability to represent and/or use sensory-specific incentive motivational properties of reward. Homozygote (KO) and heterozygote (HET) CB1 mutant mice, together with wildtype (WT) controls, were trained to associate responding on one lever with one outcome (orange flavored sucrose) and responding on a second lever with a different, sensory specific outcome (grape flavored sucrose). Subsequently, the value of one of these outcomes was attenuated by selectively pre-feeding mice during a 2-hr period and the propensity of the mice to respond for the devalued and non-devalued outcome was assessed under extinction conditions. As expected, in WT mice, responding on the lever associated with the pre-fed outcome (i.e., the devalued outcome) was significantly lowered compared to responding on the lever associated with the non-devalued outcome. By contrast, CB1 KO and HET mice failed to suppress responding on the lever associated with the devalued outcome. Importantly, during a subsequent choice test we established that mice were able to discriminate between the devalued and non-devalued flavors directly, irrespective of genotype. These deficits mirror those seen after lesions of the basolateral amygdala or prefrontal cortex and suggest that CB1-mediated neurotransmission in this region is necessary for encoding relationships between sensory-specific features of food rewards and their incentive value.
Inducible inactivation of BDNF receptor TrkB impairs rate of discriminative eyeblink conditioning in mice
Eyeblink conditioning is a powerful paradigm for studying the effects of specific genetic mutations on cerebellar function in mice (Chen, et al., 1996; Kato, et al., 2005; Weiss, et al., 2002; Woodruff-Pak, 2005). Qiao et al. (1998) reported a deficit in the acquisition of eyeblink conditioning in the stargazer mouse, a spontaneous mutant characterized by cerebellar anomalies, ataxia and generalized spike wave seizures, with reportedly low levels of BDNF expression in the cerebellum (Qiao, et al., 1996). To directly examine the contribution of BDNF and its receptor TrkB, the present study examined discriminative eyeblink conditioning in inducible knockin mice susceptible to TrkB inhibition by systemically injected 1NMPP1 (Chen et al., 2005). Furthermore, as most mouse eyeblink studies involve single-cue conditioning, the present study examines discrimination learning (two-cue conditioning) because this procedure offers a nonassociative control condition within the same subjects. TrkB-inhibited and non-inhibited control mice received 8 acquisition sessions of tone-light discrimination training (Paczkowski et al., 1999) using the following parameters: 1 session/day, 50 CS+ and 50 CS- trials/session, 380 msec 70 dB tone or 15 W light presented against dark background, US = 100 msec, 0.75 mA perioccular-shock, ISI=280 msec, alpha-response period = 0-40 ms after CS onset. Alpha- and unconditioned responses did not differ across groups. Although inhibited mice reached the same level of asymptotic conditioned responding as the non-inhibited controls by the end of training, the rate of acquisition for the inhibited mice was slower than controls, particularly in the CR amplitude measure. Our findings support a role of BDNF-TrkB signaling during the acquisition of eyeblink conditioning.
Comparison of three mouse strains in discrimination and reversal learning of the conditioned eyeblink response
In discrimination and reversal of eyeblink conditioning, conditioned responses (CRs) are acquired to a conditioned stimulus (CS+) that is paired with the unconditioned stimulus (US), but not a second conditioned stimulus (CS-) that is presented alone. During reversal learning, in which the CS+ becomes the CS- and vice versa, CRs are acquired to the (new) CS+ and extinguish to the (new) CS-. Initial discriminative learning is cerebellar dependent (Kim & Thompson, 1997) whereas extinction of CRs during subsequent reversal learning also engages forebrain areas (Berger & Orr, 1983; Chachich & Powell, 1998). Here we present the results for three strains of mice tested on a discrimination and reversal eyeblink conditioning protocol. C57/BL6, 129/SvImJ and a hybrid of these two strains (F1) received 4 acquisition sessions of tone-light discrimination training followed by 8 reversal sessions (Stanton et al., 2007) using the following parameters: 1 session/day, 50 CS+ and 50 CS- trials/session, 380 msec 70 dB tone or 15 W light presented against dark background, US = 100 msec, 0.75 mA perioccular-shock, ISI=280 msec, alpha-response period = 0-40 ms after CS onset. All three strains showed robust discriminative eyeblink conditioning, with CRs to the CS+ reaching 80-90% by the 4th session, and levels of responding to the CS- remaining at ~40% throughout. During reversals only the C57 and F1 mice acquired CRs to the new CS+ (former CS-) reaching asymptotic levels of 90% CRs, vs. ~60% for the 129 mice. Extinction to the new CS- (former CS+) was modest in the C57 and F1 mice and somewhat greater in the 129 miceIt is likely that with additional reversal trials the C57 and F1 mice will show further extinction to the CS-, and this will be tested in future studies. Our results show that discrimination reversal can be used in studies of eyeblink conditioning in genetic mouse models to examine functional interactions between cerebellar circuitry and forebrain regions.
Normal Pavlovian modulation of appetitive responding is disrupted in cannabinoid receptor (CB1) knockout mice
Overeating is rapidly becoming a major health concern in our society. Much benefit will be gained by identifying the psychological and neurobiological processes involved in normal and abnormal eating. Increasingly, learning and memory processes are being considered as studies have shown that learned cues can override adaptive regulatory mechanisms and alter food-procuring (appetitive) and consuming behavior. For instance, Pavlovian cues associated with food delivery can sustain or reinforce food-seeking behavior (conditioned reinforcement; CRf) and/or potentiate ongoing instrumental responding for food (Pavlovian-to-instrumental transfer; PIT). These effects of learned cues on appetitive behaviors depend on a relatively well characterized neural circuitry that includes regions of the amygdala and prefrontal cortex. The current experiments studied the role of cannabinoid (CB1) receptor in the effects of Pavlovian cues in CRf and PIT. Mice with a full (KO) or partial (HET) genetic knockout of the CB1 receptor, and littermate control mice (WT), were trained on a Pavlovian discrimination task to associate an auditory cue (CS+) with reward delivery (10% sucrose) and a second cue with the absence of reward (CS-). During the subsequent test for CRf, the ability of response-contingent presentations of the CS+ (and CS-) to reinforce acquisition of a new instrumental response was assessed. In the PIT study, mice underwent Pavlovian discrimination training as before and, on separate sessions, training to instrumentally respond (lever press) for the same reward. PIT was tested by presenting the CS+ during performance of the instrumental behavior. The following results are reported: In both experiments, CB1 receptor mutation did not alter Pavlovian discrimination learning. Performance during instrumental training was attenuated in CB1 KO and HET mice but this genotype difference was seen only when the behavioral demand (i.e., schedule of reinforcement) for reward delivery was increased. More important, both CRf and PIT were affected by CB1 mutation in that, 1) KO mice failed to respond more for the CS+ during the test for CRf, and, 2) both KO and HET mice failed to show potentiated responding during non-contingent CS+ presentations on the test for PIT. Our findings add to a growing literature identifying cannabinoid CB1 receptor as a potential therapeutic target for the treatment of eating disorders, as well as for non-food maladaptive disorders such as nicotine and cocaine addiction.
Evaluating the effects of cocaine exposure on goal-directed and habitual responding
According to the DSM-IV a defining characteristic of drug addiction is the individual’s failure to abstain from using the substance despite having evidence of the difficulty it is causing. Although initial drug use can be characterized as a goal-directed process, after a prolonged drug-taking history the addict’s behavior becomes dominated by a maladaptive habitual process in which drug-taking continues despite knowledge of the adverse consequences associated with those actions. It is possible to model the transition from goal-directed to goal-independent behavior using tests of instrumental reinforcer devaluation. In this paradigm, subjects are trained to perform two separate instrumental responses, each leading to a specific outcome. After training, one of the outcomes is devalued by prefeeding the reinforcer prior to conducting the test under extinction conditions. Typically, subjects who respond in a goal-directed manner prefer the response whose reinforcer was not previously devalued. By contrast, subjects responding in a goal-independent habitual manner fail to show this sensitivity to post-conditioning changes in reinforcer value. In the present experiments, mice were pre-treated with cocaine (30 mg/kg) or saline for 10 sessions either prior to (Experiment 1) or following (Experiment 2) instrumental training. Subsequently, mice were pre-fed with one of the two reinforcers prior to an assessment of responding. As expected, mice treated with saline responded in a goal-directed manner, favoring the response trained with the non-devalued reinforcer. However, cocaine treatment resulted in habitual responding, whereby mice were unable to shift responding based on changes in reinforcer value. Importantly, both groups of mice consumed more of the non-devalued reinforcer in a subsequent reinforcer choice test, indicating that performance in the instrumental response test could not be accounted for by a failure to discriminate between the reinforcers themselves or a failure of the devaluation treatment. Collectively, these results mimic that seen following orbitofrontal-amygdalar damage and indicate that cocaine exposure results in habitual responding, paralleling the persistent drug-taking despite adverse consequences seen in cocaine addicts.
The role of neuronal activity regulated pentraxin (narp) in motivational learning
Neuronal activity regulated pentraxin (NARP) is a secreted immediate early gene product regulated by synaptic activity. Recent studies have shown it plays a key role in excitatory synaptogenesis by regulating AMPA receptors at the synapse. As AMPA receptors mediate synaptic plasticity and learning, and as Narp is selectively expressed in amygdala and lateral hypothalamus, we investigated Narp’s role in motivational learning. Specifically, we examined whether mice lacking NARP were capable of learning a simple cue-reward discrimination, and whether such cues were able to influence motivated behavior. Mice were trained to discriminate between presentations of a reward-paired auditory cue (CS+) and an unpaired auditory CS-. Both NARP knock-out (KO) and littermate control mice (WT) acquired the Pavlovian discrimination. During subsequent testing phases the reward-paired CS+ was capable of augmenting ongoing instrumental performance (Pavlovian-to-instrumental transfer) and was capable of supporting the acquisition of novel instrumental nose-poke responding (conditioned reinforcement). These results suggest that NARP is not involved in mediating general motivational effects of rewarded cues on learning and performance. However, the results of an additional experiment implicate NARP in using sensory-specific features of expected reward value to direct instrumental performance. Using an outcome-specific devaluation paradigm, mice were trained to respond on one lever for one reward (e.g., grape-flavored sucrose) and respond on a second lever for a different reward (e.g., orange-flavored sucrose). Following training, mice were pre-fed with one of the two rewards for a 2-hr period. This sensory-specific satiety treatment served to devalue one reward, while maintaining the value of the other reward. Responding to both levers was then assessed in an extinction test. As expected, WT control mice showed a significant preference in responding to the lever associated with the non-devalued reward. In contrast, NARP KO mice responded equally to both levers, failing to suppress responding to the lever associated with the devalued reward. Both groups of mice consumed more of the non-devalued reward in a subsequent reward choice test, indicating that performance in the instrumental response test could not be accounted for by a failure to discriminate between the rewards themselves or a failure of the devaluation treatment. Collectively, these results suggest that NARP-mediated neurotransmission has a selective role in encoding sensory-specific motivational value, but not the general motivational effects of rewarded cues on performance.
>Narp mediates retention of emotional memories
The lateral hypothalamus is a key brain region implicated in reward and motivation. In recent years the orexin neurons have been identified in this region as important mediators of both food and drug rewarding behaviors, for example, conditioned place preference to morphine. In prior studies we have shown that orexin neurons selectively express Narp, a secreted protein which regulates AMPA receptors at synapses and which plays a role in maintaining memory traces induced by aversive stimuli. To test if Narp also mediates rewarding behavior we have assessed if Narp KO mice display morphine place preference using a single-chamber, unbiased design. We found that Narp KO mice acquire morphine place preference at least as well as WT control mice. To check how well this memory is retained, we attempted to extinguish preference for the conditioned side by daily administering a saline injection and then exposing mice to the conditioning chamber. Whilst control mice showed no preference for the morphine-paired side after 13 extinction trials, the Narp KO mice extinguish at a slower rate, failing to extinguish preference for the morphine paired side after an additional 10 trials. Together with our previous findings, these results suggest Narp plays a key role in regulating retention of rewarding as well as aversive memories. Future studies are aimed at determining whether this phenotype is caused by Narp deletion from the orexin cells or from another brain region and what the molecular mechanism are underlying the phenotype.
Effects of cues associated with surprising meal interruption on food consumption in mice
Food consumption is controlled by both internal and external factors. Environmental signals associated with food may prepare an animal to forage, consume and digest more effectively. Furthermore, environmental cues that provide information about food availability enable animals to make predictions about future food resources and act upon that knowledge in appropriate fashion. For example, when exposed to a cue that signals the presence of food, an animal experiencing, or having experienced, hunger can eat beyond its present needs to cope with predicted future famine. Interestingly, cues paired with meal interruption have a similar effect. That is, cues previously paired with surprising meal terminations can increase food consumption later. Here, using a novel Pavlovian conditioning procedure, food-deprived mice learned to associate a conditioned stimulus with an unexpected termination of a food trial. Subsequently, when experimental food was made available, sated mice consumed more food in the presence of these cues than when exposed to cues that were paired with food, or in the absence of any discrete food-related signals.
ABSTRACT ONLY Narp Mediates Aversive Effects of Opiate Withdrawal
As Narp is an immediate early gene implicated in regulating AMPA receptor trafficking, it is thought to play a key role in behavioral plasticity. In previous studies, we have found that Narp is induced in the central nucleus of the amygdala, the BNST and the nucleus accumbens by opiate withdrawal. As these regions mediate the long-lasting aversive responses to opiate withdrawal, we have examined whether Narp KO mice display abnormalities in this behavioral paradigm. Narp KO and wild type mice were made opiate-dependent by implantation of morphine pumps (4mg/kg/hr). Naltrexone 250 micrograms/kg was used to trigger opiate withdrawal. Scoring of multiple somatic signs of opiate withdrawal did not reveal an effect of genotype during the acute withdrawal phase. Conditioned place aversion was tested on day 7 after conditioning and acquisition of this behavior was similar in Narp KO and WT mice. However, when mice underwent active extinction training, we noted that Narp KO mice displayed more rapid extinction of this behavior. In the absence of active extinction, the aversive response was sustained for four weeks, indicating that Narp is not needed for maintaining this behavior. Thus, these studies suggest that Narp plays a key role in aspects of synaptic plasticity that are critical for generating stable behavioral responses that are resistant to extinction. As the persistence of aversive responses to opiate withdrawal have been postulated to contribute to opiate addiction, these findings indicate that suppression of Narp induction or function may decrease craving for opiates or may enhance the efficacy of extinction based behavioral interventions.
Pavlovian Influences on Goal-Directed Behavior in Mice: The Role of Cue-Reinforcer Contiguity
Many contemporary theories of appetitive motivation emphasize roles for Pavlovian conditioned cues (reward cues) in guiding and energizing goal-directed behavior. Whilst a number of models for studying Pavlovian-instrumental interactions have been developed in rats and monkeys, and certain critical conditions and variables have been well characterized, few of these procedures have been adapted for use with mice. Given the promise of genetically altered mice for providing new insights into a variety of psychopathologies, such as addiction, eating disorders and depression, further development of mouse models of these Pavlovian-instrumental interactions is valuable. We present a series of studies that were aimed at developing a mouse model of two types of Pavlovian influences on goal-directed behavior (1) the ability of Pavlovian reward cues to support (reinforce) acquisition of a novel instrumental response (conditioned reinforcement) and (2) the ability of Pavlovian reward cues to energize or potentiate ongoing instrumental responding (Pavlovian-to-instrumental transfer or PIT). For conditioned reinforcement, independent groups of C57Bl/6 mice were trained on a Pavlovian task to associate either a 10 sec or a 2 min auditory stimulus with reward delivery (30% condensed milk). In the 10 sec condition, reward was presented 5 sec after cue onset and terminated with cue offset. In the 2 min condition, reward was delivered every 30 sec on average (RT30) during the 2 min CS period. Next, the ability of the 10 sec versus 2 min training CSs to reinforce novel instrumental responding (nose-poking producing a 3 sec CS presentation) was assessed. Separate mice in a PIT experiment were similarly Pavlovian trained using either the 2 min or 10 sec CS and, in separate sessions, to instrumentally respond for milk reward. On the test for PIT, we tested the ability of non-contingent presentations of the CS to enhance instrumental lever responding. We report that, whilst a 10 sec CS training condition produced robust conditioned reinforcement and a small PIT effect, the 2 min training condition produced a strong PIT effect but little conditioned reinforcement. These results are in line with earlier suggestions (Holland, 1980; Lovibond, 1981) regarding the content of learning and the role of temporal relationships of Pavlovian cue-reward pairings.
Normal Cocaine Cue Preference Conditioning and Psychomotor Sensitization in EGR1 Knockout Mice
EGR1, also called Zif268, is a ubiquitously expressed transcription regulatory protein that has been linked to neuronal activation and experience-dependent neuroplasticity. One interesting form of experience-dependent plasticity is produced by repeated exposure to psychomotor stimulant drugs of abuse. For example, repeated intermittent administration of cocaine produces a persistent hypersensitivity to the psychomotor activating and rewarding effects of the drug (sensitization). Additionally, environmental cues consistently paired with cocaine injections acquire motivational qualities and come to be preferred over neutral (non-paired) stimuli (conditioned place preference). In the present study we tested the hypothesis that EGR1 is necessary for the acquisition and/or maintenance of cocaine-induced psychomotor sensitization and conditioned cue preference. To this end, EGR1 knockout mice (obtained from J. Milbrandt, Wash U. ) and WT controls on a C57BL/6N background were tested with an unbiased, single arena cue preference procedure. This procedure consisted of a single open field in which interchangeable floor textures provided the environmental cues. This procedure allowed us to simultaneously quantify cocaine (floor) cue preferences and changes in psychomotor activation. No genotype differences were evident in the ability of an acute (first) injection of cocaine to increase psychomotor activation across the effective dose range of cocaine. More critically, repeated injections of cocaine (4 x 10 mg/kg) produced identical sensitization to the psychomotor stimulant effects of cocaine in EGR1 and WT mice. Furthermore, cocaine injections produced a robust preference for the drug-associated floor cues, and the development and long-term maintenance (over 4 weeks) of this preference was completely unaltered in EGR1 mice. Thus, contrary to our hypothesis, our findings do not suggest a critical role of EGR1 in cocaine-induced neuroplasticity, as indicated by the development and maintenance of psychomotor sensitization and cue preference conditioning. These results differ from those obtained in similar studies using a different line of EGR1 KO mice (Valjent et al., J. Neurosci., 2006).
Genetic Background Differences and Nonassociative Effects in Mouse Trace Fear Conditioning
In trace fear conditioning the CS and US are separated by a stimulus-free “trace interval,” a variant of fear conditioning that engages the hippocampus and prefrontal cortex. There is surprisingly little data concerning nonassociative influences on mouse strain differences in trace fear conditioning. We developed a trace protocol that yields strong conditioned freezing and low levels of nonassociative freezing. We trained C57/BL6 (C57) and 129/SvlmJ (129) mice, and a strain derived from crossing the C57 and 129 strains (F1 Hybrids) in the task using paired or unpaired presentations of a tone CS and footshock US , separated by an 18 sec trace interval. Tests for learning (conditioned freezing) were done in the training context vs. a novel test context, and to the trace-CS during CS-alone trials in the test context. The protocol yielded strong context conditioning (?=45%) and low levels of nonassociative freezing (< 5%) in all mouse strains. This contrasts with what we have obtained in more standard procedures in which higher levels of non-associative freezing were observed. During the trace-CS test, freezing in unpaired controls remained low (~5-15%) in all strains, and both the C57s and F1 Hybrids showed reliable associative trace fear conditioning. Trace conditioning, however, was not obtained in the 129/SvlmJ (129) mice. In addition to showing strain differences across context and trace conditioning, development of our protocol identified critical parameters that are important for minimizing non-associative effects in mouse contextual and trace paradigms for mice.
A Unique and Selective Role for GLUR1 S831 Phosphorylation in Incentive Learning
Phosphorylation of AMPA receptor GluR1 subunit is important for the expression of synaptic and behavioral plasticity. Thus, mice with targeted mutations of the serine 831 and 845 sites show deficits in LTD/LTP and spatial learning and memory (Lee et al 2003). Recently, we have reported that these mice also show deficits in the ability of reward-associated cues to (1) potentiate or energize ongoing instrumental behavior (Pavlovian-to-instrumental transfer or PIT) and (2) to serve as reinforcers themselves (conditioned reinforcement or CRf). In contrast, mutation of the s845 phosphorylation site alone did not affect these two aspect of incentive learning. Together, these results suggested that GluR1 s831 phosphorylation by CaMKII/PKC provides a critical neurobiological step in the ability of reward cues to influence goal-directed behaviors. The present experiment tested this hypothesis. Mice with targeted mutations of the s831 site and WT mice were trained to associate an auditory cue (CS+) with sucrose reward and, in separate sessions, to lever press for sucrose reward. PIT was tested by presenting the CS+ during performance of the instrumental behavior under extinction conditions. After mice were retrained to associate a different CS with reward, we determined the ability of that cue to reinforce acquisition of a new instrumental response (CRf). Both s831 mutant and WT mice acquired reliable discriminated responding to the Pavlovian CS+ (vs. an unpaired CS-) and instrumental responding for the reward, although instrumental response rates were attenuated in s831 mice. In subsequent tests, WT mice showed robust PIT and conditioned reinforcement: CS+ presentations significantly elevated instrumental responding relative to baseline/CS- response levels, and cue-reinforced instrumental responding was significantly elevated over baseline. Although s831 mice showed a comparable PIT effect as WT mice, CRf was significantly attenuated in these mice. Our results suggest a unique and selective role of GluR1 S831 phosphorylation in one type of incentive learning.
Innovation in the Assessment of Olfactory-Based Behavior in Mice
Behavioral analysis utilizing olfaction in mice can provide a versatile and valuable research tool for the neurogeneticist, but is relatively under utilized in rodent research. Here, we present data from two different settings that exploit the olfactory modality for testing mice. The first is a novelty odor test that can be used to broadly screen for functions such as olfactory perception, olfactory driven exploration and habituation, novelty preference, and memory. As an illustrative example, we assessed mice lacking the pentraxin Narp over two sessions in an open arena where odors were located. In the 1st session both odors were identical (odor A vs. odor A) and in the 2nd session one of the odors was replaced with a novel odor (odor A vs. odor B). Our data show that during the 1st session both the (+/+) and the (-/-) mice equally explored both ‘A’ odor stimuli, but in the 2nd session only the (+/+) mice showed a preference for the novel odor (B). These data show that the (-/-) mice have a deficit, the nature of which is under further study. The 2nd setting uses a state-of-the-art olfactory behavioral apparatus for extended testing and flexible use of different learning and memory protocols. Here we provide the results of the performance of C57BL/6 mice in a simple go-no-go odor discrimination in which mice were presented with one of two odors at an odor port on each trial, followed by delivery of either sucrose or quinine solution at a liquid delivery port. The mice learned to withhold responding at the liquid port after the odor cue for quinine (90% correct criterion) and then again achieved that criterion when the contingencies were changed in reversal training. These data demonstrate that mice can be readily tested in this odor discrimination and reversal task. In addition, the test system can be readily modified to assess spatial discrimination (multiple wells, multiple locations), working memory (multiple odors; match and non-match to sample), and spatial working memory (multiple odors and wells). Together, these variants can provide a test battery for a range of assessments of mice in a setting that takes advantage of their natural olfactory capabilities.
Innovation in the Assessment of Gustatory-Based Behavior in Mice
We have developed several gustatory paradigms to assess consummatory behavior, memory for gustatory events, and motivational responses in mice. We first designed a new type of gustometer to measure licking at a liquid delivery port. This system uses photobeam detection of licks in a sunken well to which a liquid solution can be delivered. Using this technology we collected a wide range of information including the amount of solution consumed, the time spent in the well, latency to lick, and aspects of the microstructure of licking. A second set of procedures assessed neophobia and memory as determined by changes in the consumption of a novel flavor solution. Mice (C57/BL6, 129/SvlmJ, and a hybrid strain derived from both) were acclimated to a liquid diet until they reached stable consumption values. Mice were then given a novel flavored diet with 3 presentations of the flavored diet spaced over days. The results showed decreased consumption for all 3 strains during the first flavor presentation, indicative of neophobia and intact gustatory processing. Mice in all strains also similarly increased consumption on the next 2 presentations, indicating maintained retention for exposure to the flavored diet. Finally, we designed an outcome-selective devaluation protocol to assess the impact of changes in the value of a reward. C57BL/6 mice were trained to respond on specific levers for delivery of specific gustatory reward outcomes (Grape- or Orange-flavor Kool-Aid). Following acquisition, we assessed the impact of outcome devaluation (sensory-specific satiety), induced by prefeeding one of the two outcomes for 2 hours, on subsequent instrumental performance assessed in extinction. Responding on the lever associated with the devalued flavor was suppressed compared with responding on the lever associated with the non-devalued flavor. These protocols provide settings to assess a range of ingestive and emotional/motivational behaviors, including functions under cognitive control as exemplied by adaptive behavior guided by changes in outcome value.
The Role of the BDNF Receptor Trk-B on the Acquisition and Expression of Conditioned Incentive Value
The neurotrophin brain-derived neurotrophic factor (BDNF) and its tyrosine receptor kinase B (TrkB) play a key role in regulating neuronal structure in both the developing and adult CNS. Additionally, these molecules have been implicated in synaptic plasticity and memory. Taking a chemical-genetic approach using a mouse model harboring TrkB knock-in alleles allows for specific stable and reversible in-vivo inhibition of expression of TrkB in the adult mouse when the inhibitor 1NMPP1 (INH) is administered (Chen et al., 2005). The use of such mice combines the benefits of pharmacology and mouse genetics to assess the influence of TrkB at specific time points of a test procedure corresponding to different phases of learning. Using a conditioned reinforcement procedure (CRf), whereby a previously conditioned stimulus (CS) later acts as the reinforcer for an instrumental action, we assessed whether TrkB is necessary for the acquisition or expression of learned incentive value. In four different groups of mice, the inhibitor was given to prevent TrkB expression during either acquisition or test, both of these phases, or neither phase (control condition). During training, all mice, regardless of TrkB expression, learned the Pavlovian discrimination, spending significantly more time in the magazine recess during presentations of the CS+ compared to the CS-. However, during the CRf test for acquired incentive value, mice that had inhibited TrkB during acquisition failed to show any preference of responding for presentations of the previously rewarded CS. In contrast, the inhibition of TrkB during that test session alone was without effect; those mice showed the normal expression of incentive value by responding selectively for presentation of the CS+. These data demonstrate that BDNF signaling through TrkB receptors is essential for the acquisition, but not expression of learned incentive value.
The Role of Serum Response Factor in Incentive Learning
The transcription factor, serum response factor (SRF) regulates the activity of a wide range of immediate early genes (IEGs) that are implicated in neural plasticity underlying learning and memory. Here we used two lines of SRF knockout mice generated with the Cre/lox P system to induce targeted deletions of SRF. In one line of mice Cre is under the control of a CamKII promoter, and SRF deletion occurs in forebrain limbic/cortical circuitry (SRF f/f;CamKIICre); in the other line of mice Cre is under the control of a synapsin promoter and the SRF deletion is restricted to the CA3 and DG of the hippocampus (SRF f/f;SynCre) (Ramanan et al., 2005). These mice were evaluated for incentive motivational learning in a conditioned reinforcement paradigm. Mice were first given conditioning trials in which reward followed presentation of one cue (CS+) but not a second (CS-). All groups of mice learned the Pavlovian discrimination, spending significantly more time in the magazine recess during presentations of the CS+ compared to the CS-. We then tested for acquired incentive value of the CS+ in a setting where one response led to CS+ presentations while a different response led to presentations of the CS-. Results showed that the SRF f/f and SRF f/f;synCre mice maintained selective responding for presentation of the CS+, indicative of incentive motivational learning, but the mutant SRF f/f;CamKIICre mice did not. These results are consistent with the hippocampal-independent nature of acquired value and suggest a role for SRF in other forebrain circuits that are known to be essential for this form of learning.
The Role of GLUR1 in Reward Learning I: Phosphorylation at S845 Does Not Modulate Pavlovian Influences on Instrumental Behavior.
Appetitive Pavlovian conditioned stimuli can exert motivational influences on goal-directed behavior. For example, stimuli associated with primary rewards can enhance instrumental responding for that reward (Pavlovian-to-instrumental transfer [PIT]), and support acquisition of novel instrumental responding (conditioned reinforcement [CRf]). These two manifestations of Pavlovian-instrumental interactions are dissociable neurobiologically. For instance, mice lacking the GluR1 or GluR2 subunit of AMPA receptor show selective deficits in CRf and PIT, respectively (Mead & Stephens 2003). To further explore the role of GluR1 in reward learning we examined PIT and CRf in mice lacking the S845 phosphorylation site of the AMPA-GluR1 subunit. First, S845 mutant and WT mice were trained to associate an auditory cue (CS+) with milk reward and in separate sessions to lever press for the same reward. Once accurate and stable responding was established mice were tested for PIT by presenting the CS+ during instrumental responding, but under extinction conditions. To assess CRf, mice were trained to associate a 10 sec auditory cue with reward and a single session was conducted to determine the ability of the cue to support acquisition of a new instrumental response. We report that 1) no genotype differences were seen during Pavlovian or instrumental training sessions; 2) CS+ presentation significantly elevated (by ~200%) instrumental responding over baseline levels but no genotype differences were evident in the magnitude of PIT; 3) CS+ reinforced instrumental responding was significantly elevated relative to baseline (non-reinforced) levels but, again, no genotype differences were seen in CRf. We conclude that the AMPA receptor GluR1-S845 site, phosphorylated by cAMP-dependent protein kinase (PKA), is not necessary for normal reward learning to occur and for Pavlovian cues to influence instrumental behavior.
Effects of Dopamine Receptor Manipulations on Stimulus-Reward Learning
Knowing how reward-associated cues influence instrumental behaviors is important for understanding many aspects of motivated behavior, including pathological states of motivation such as addiction. A prime example of this is the Pavlovian-to-instrumental transfer phenomenon (or PIT). In this procedure, after animals learn to associate a Pavlovian cue with reward (CS+) and, on separate occasions, learn to instrumentally respond for this reward, presentation of the CS+ during instrumental performance typically potentiates responding. Although little is known about the neurobiology of PIT, studies have clearly implicated glutamatergic (AMPA) and dopaminergic mechanisms. For instance, pretreatment with D1/D2-like (a-flupenthixol or Pimozide) antagonists was reported to disrupt PIT in rats (Dickinson et al, 2000). Here we further explore the role of dopamine receptor activation in PIT by pretreating C57/b6 mice, 10 min prior to PIT testing, with different IP doses of D1 or D2 receptor antagonists. We report that: 1) Lower doses of the D1-like receptor antagonist SCH23390 (0.015-0.03 mg/kg), although producing a trend towards a selective effect on CS+ potentiated responding, did not significantly affect PIT whilst higher doses (0.06-0.12 mg/kg) produced a non-selective reduction in both CS+ and baseline response levels; 2) All doses of the D2-like receptor antagonist Raclopride (0.075-0.6 mg/kg) non-selectively attenuated instrumental responding; 3) Doses of SCH23390 and Raclopride that non-selectively reduced levels of instrumental responding also depressed locomotor activity. Based on these results we conclude that, at least within the dose range used here, D1 or D2 antagonists do not selectively modulate the enhancing effects of Pavlovian cues on instrumental behavior. Although different doses of D1 or D2 antagonists, or a combination of these, may produce more selective effects, our study supports the view that dopamine receptors mediate a wide array of locomotor, activational and motivational processes.
The Role of GLUR1 in Reward Learning II: Phosphorylation at S845 and S831 Modulates Pavlovian Influences on Instrumental Behavior.
Phosphorylation of AMPA receptor GluR1 subunit is required for the expression of synaptic and behavioral plasticity. Mutant mice lacking both serine 831 and 845 sites (phosphorylated by CaMKII/PKC and PKA, respectively) show deficits in LTD/LTP and concomitant deficits in spatial learning and memory (Lee et al 2003). Little is known, however, about the role of AMPA-GluR1 phosphorylation in other forms of learning. Therefore, we used mice in which both S845 and S831 on GluR1 were mutated to test if GluR1 phosphorylation is necessary for two dissociable aspects of reward learning: the ability of conditioned stimuli to potentiate instrumental behavior (Pavlovian-to-instrumental transfer [PIT]) and the ability to act as reinforcers themselves (conditioned reinforcement [CRf]). Mutant and WT mice were trained to associate an auditory cue (CS+) with milk reward and in separate sessions to lever press for milk reward. PIT was tested by presenting the CS+ during performance of the instrumental behavior under extinction conditions. After mice were trained to associate a CS+ with reward, we determined the ability of that cue to reinforce acquisition of a new instrumental response. Both mutant and WT mice acquired reliable discriminated responding to the Pavlovian CS+ (vs. an unpaired CS-) and instrumental responding for the reward, although instrumental response rates were somewhat lower in mutant mice. In subsequent tests, WT mice showed robust PIT and conditioned reinforcement: CS+ presentations significantly elevated instrumental responding relative to baseline/CS- response levels, and cue-reinforced instrumental responding was significantly elevated over baseline. Both effects were dramatically attenuated in mutant mice. These results suggest an important modulatory role of GluR1 in the expression of reward learning and, together with the results from a parallel study, suggest a critical role of the S831 phosphorylation site.
Advancing Tools and Technology for Research Using Animal Models
We have refined and optimized several widely used behavioral protocols for the evaluation of mice using strains of various genetic backgrounds.
Copyright © 2011 The Johns Hopkins University Department of Psychological and Brain Sciences