Elsevier

Behavioural Brain Research

Research report

A comparison of Lewis and Fischer rat strains on autoshaping (sign-tracking), discrimination reversal learning and negative automaintenance

Abstract

Lewis (LEW) and Fischer (F344) rat strains differ on a number of physiological characteristics, such as hypothalamic–pituitary–adrenal (HPA) axis activity, as well as on behavioral tasks, including those that measure impulsivity and drug reward. Since autoshaping, the phenomenon where animals approach and contact reward-paired conditioned stimuli, has been linked to HPA axis functioning, impulsivity and drug taking, the present study compared LEW and F344 rats on the rate of acquisition and performance of the autoshaping response. Rats were trained on an autoshaping procedure where insertions of one retractable lever (CS +) were paired response-independently with food, while insertions of another lever (CS−) were not paired with food. LEW rats acquired the autoshaping response more rapidly and also performed the autoshaping response at a higher rate than F344 rats. No differences between the strains were observed when rats were trained on a discrimination reversal where the CS+ and CS− levers were reversed or during a negative automaintenance phase where CS+ lever contacts cancelled food delivery. Potential physiological mechanisms that might mediate the present results, including strain differences in HPA axis and monoamine neurotransmitter activity, are discussed. The finding that LEW (as compared to F344 rats) more readily acquire autoshaping and perform more responses is consistent with research indicating that LEW rats behave more impulsively and more readily self-administer drugs of abuse.

Introduction

Lewis (LEW) and Fischer (F344) rat strains differ on a variety of behavioral and physiological characteristics (for review see [19]). Much interest has focused on how these two inbred strains differ in terms of responsivity to drugs of abuse. In general, LEW rats appear to be more sensitive than F344 rats to the rewarding effects of drugs (see also [12], [22], [32]). For example, LEW rats acquire cocaine self-administration more rapidly and at lower doses than F344 rats [15], [21]. Similarly, LEW rats more readily acquire morphine self-administration and achieve higher breakpoints on a progressive ratio schedule of morphine self-administration than F344 rats [1], [25]. LEW rats consume more orally self-administered ethanol than F344 rats [37]. LEW rats also show greater conditioned place preferences for locations previously paired with cocaine [20], morphine [14] or nicotine [17]. Because of these differences, it has been suggested that the LEW and F344 rat strains may provide a model of human genetic variables that may predispose individuals to drug abuse [19] (though see [13]).

Recently, Anderson and Woolverton [2] have shown that LEW and F344 rats differ in delay discounting, an operant choice task that has been considered to be a measure of impulsivity. LEW rats display higher rates of delay discounting (i.e., more impulsivity) than F344 rats. Specifically, LEW rats choose a small immediate reward (one food pellet) over a larger delayed reward (three food pellets presented after a 10-, 20-, 40- or 60-s delay) significantly more often than do F344 rats. The finding that LEW rats, who more readily self-administer abused drugs than F344 rats, display higher rates of delay discounting than F344 rats is consistent with research from out-bred rat strains (e.g., Long–Evans) showing that individual rats with relatively high delay discounting rates subsequently acquire cocaine self-administration more rapidly and at lower doses [31] and also self-administer more alcohol [33] than do rats with lower delay discounting rates. The difference between LEW and F344 rats in delay discounting also parallels the results of human research showing that drug abusers display higher rates of delay discounting (more impulsivity) than non-abusing controls (for review, see [4]).

The goal of the present study was to investigate further any potential differences in impulsivity between LEW and F344 rats by comparing them on the acquisition and performance of autoshaping (for reviews, see [16], [34], [43]. Autoshaping [5], or "sign-tracking" [16], is the phenomenon where animals approach and contact discrete and localized conditioned stimuli (CSs) that have been paired with appetitive unconditioned stimuli (USs), such as food or water (cf. [18]). In autoshaping, there is no contingency that requires subjects to make a response in order to receive the appetitive US, yet animals often persist in contacting the CS even when doing so cancels US delivery ("negative automaintenance"; [48]). Thus, this directed approach and contact behavior is generally considered to be a Pavlovian conditioned response (CR) elicited by the CS, although it is likely that operant factors are also involved [47].

Autoshaping has been considered to be a form of impulsive behavior [27], [41], [50]. Evidence for this hypothesis comes from a study by Tomie et al. [41] who found that rats with higher rates of delay discounting displayed faster acquisition (as well as higher asymptotic performance) of autoshaping than rats with lower rates of delay discounting. Further support comes from a study by Winstanley et al. [50] who found that depletions of forebrain serotonin (5-HT) in rats: (1) increased the number of autoshaping approach responses to a food-paired CS (as well as to a control stimulus) and (2) also increased impulsive responding on other behavioral tasks. Previous studies found that selective lesions of the 5-HT system increased impulsivity as measured by differential-reinforcement-of-low rates (DRL) procedures [10] as well as delay discounting [51] procedures (though it should be noted that Winstanley et al. [50] found no effect of their 5-HT manipulation on delay discounting). Finally, some have postulated that the fact that autoshaping often persists despite negative consequences (i.e., under negative automaintenance conditions) suggests that there is an impulsive aspect to autoshaping [27].1

Based in part on the relation between autoshaping and impulsivity, Tomie [38], [39], [40] has argued that autoshaping and drug abuse may be closely related phenomena. According to Tomie's autoshaping model of drug abuse, the human drug-taking situation and autoshaping have important environmental features in common. In both situations, a discrete, localized CS (e.g., a keylight versus a heroin syringe) reliably precedes a rewarding US (e.g., food versus heroin) and through such CS–US pairings, the CS may come to elicit CS-directed CRs. The model predicts that because of these similarities, individuals that more readily learn and perform autoshaping CRs should also more readily consume abused drugs. This prediction has been confirmed in rats in a study that found that autoshaping performance was a predictor of subsequent oral alcohol self-administration [40]. Further support for a link between autoshaping and drug taking comes from a study showing that rats that readily engage in autoshaping also display a number of physiological characteristics associated with increased propensity to self-administer drugs, including heightened mesolimbic dopamine system activity [42].

As outlined above, there is evidence that (1) LEW rats more readily self-administer drugs of abuse and (2) LEW rats behave more impulsively than F344 rats. These results, taken together with evidence indicating that (1) autoshaping is related to drug taking and (2) autoshaping is a form of impulsive behavior, suggest that LEW rats should more readily acquire and perform the autoshaping response than F344 rats. The present experiment tested this prediction by training rats on a two-lever autoshaping procedure where insertions of one lever (CS+) were paired with a food pellet and a second control lever (CS−) was presented as frequently as the CS+ lever, but was never paired with food. To explore further potential differences between the strains in autoshaping, following autoshaping acquisition, a discrimination reversal phase was conducted where the former CS− was paired with food and the former CS+ was no longer paired with food. In a final phase, rats were trained on a negative automaintenance procedure where responses on the CS+ lever cancelled food delivery.

Section snippets

Animals

Eight naïve adult male LEW rats and eight naïve adult male F344 rats served as subjects. Half of the rats in each strain were 84–85 days old at the start of training and the other half of rats in each strain were 155–156 days old at the start of training. Because of a limited number of training chambers, half the rats in each strain (four LEW and four F344) completed all phases of training and then the second half of the rats in each strain were trained. Rats were maintained at 85% of their

Autoshaping

It took LEW rats a mean of 44.0 (S.E.M.   =   5.1) trials to achieve the autoshaping acquisition criterion, while F344 rats required a mean of 99.3 trials (S.E.M.   =   26.4) to reach this criterion. One F344 rat did not reach the autoshaping criterion by the end of the 20 autoshaping sessions and therefore this rat's data are not represented in the F344 mean and are not included in the statistical analysis of number of trials to reach criterion. A t-test indicated that the difference between the strains

Discussion

In summary, LEW rats more rapidly acquired the autoshaping response than F344 rats. LEW rats also contacted a lever (CS+) paired response-independently with food on a significantly greater percentage of trials and at a higher response rate than F344 rats when collapsed across the 20 autoshaping sessions. In contrast, there was no significant main effect of strain on responding (in terms of either the percentage of trials or response rate measure) on a control lever (CS−) that was not paired

Cited by (36)

View full text

Copyright © 2006 Elsevier B.V. All rights reserved.