After genetic analysis, there is substantial agreement here that populations which self-administer ethanol in a free-choice access paradigm also work for ethanol in operant studies. This parallel indicates that free-choice alcohol consumption in the home cage may indicate greater ethanol reinforcement when measured by operant oral self-administration, both in selected lines and in populations for which other methods for creating genetic change were used, including targeted mutations. We observed this association despite important procedural differences between home cage drinking and operant self-administration, including the requirement for a response (other than drinking) to obtain ethanol in an operant setting, the fact that assessment of operant behavior is typically conducted during a defined time period each day in an operant chamber (instead of assessing behavior for 24 hours a day in the home cage), and the fact that operant studies often break up free-choice drinking behavior into small sips that typically serve as a reinforcer. Agreement in findings between these divergent behaviors suggests that they share genetic and neurobiological mechanisms.
One way to think about the similar genetic mechanisms underlying home cage drinking and operant oral self-administration is that they are both, in fact, consummatory behaviors. As suggested by Samson and colleagues based upon a fine-grained temporal analysis of operant responding for ethanol (Samson et al., 2000
), there is a great degree of similarity between intake of ethanol from a sipper tube (i.e., a completely consummatory response) and intake of animals responding for small quantities of ethanol, each interspersed by a relatively small string of operant responses. A procedure in which animals emit all of their responses at the beginning of a session, followed by uninterrupted access to ethanol, may yield better understanding of behavior which is not interchangeable with consummatory responses (Czachowski and Samson, 1999
; Grahame and Grose, 2004
We were also able to observe a consistent relationship between ethanol CTA and home cage ethanol consumption, both in the meta-analyses depicted in and in the inbred strain panels at the 2 g/kg dose (see ). We found that higher home cage alcohol drinking associates with lower CTA, agreeing with previous findings that strains of mice which show stronger taste aversion also tend to show lower ethanol preference and higher withdrawal severity (Broadbent et al., 2002
). Such results suggest that with respect to ethanol, CTA reflects the likelihood of animals to acquire an aversion to flavors (including ethanol itself) paired with ethanol’s actions, an effect that interferes with drinking behavior.
In spite of the fact that CTA produces taste avoidance, there are several theories that interpret this behavior as indicative of a reinforcing effect of the drug US. Most self-administered drugs produce CTA (Grigson, 1997
; Hunt and Amit, 1987
), and this finding is further supported with this review. CTA has been hypothesized previously to be positively correlated with sensitivity to drug reinforcement (Hunt and Amit, 1987
). According to this hypothesis, taste aversion is caused by the same appetitive drug effects that mediate self-administration. It is this positive rewarding drug state, rather than some aversive effect of drugs that are self-administered, that will lead to suppression of intake due to a conditioned “taste shyness.” In a similar direction, the resulting taste avoidance may be due to a comparison between the less valued flavor stimulus and the anticipation of a highly valued drug effect when saccharin serves as the CS, as was often (but not always) the case in the reviewed experiments (Grigson and Freet, 2000
). Such an interpretation is hard to maintain, however, when one considers that populations such as DBA/2J mice, and selectively bred NP rats and LAP mice, which drink almost no alcohol in the home cage, show robust conditioned taste aversion to ethanol. One would have to speculate that these populations are so highly reinforced by ethanol that they drink almost none of it, an interpretation that seems implausible. Additionally, the Broadbent et al. (2002)
study explicitly avoided a saccharin CS (opting for a salt flavor) to discourage any successive contrast interpretation of their findings. All in all, the more parsimonious explanation of the present findings is that CTA with an ethanol US is, in fact, mediated by aversive properties of that US, which also tend to discourage free-choice alcohol consumption.
Notably, both the inbred strain panel in as well as the genetic analysis we offered in support an association between CTA and ethanol intake, while the recombinant strain panel examined in does not. While it is possible that this difference could reflect a true weakness of this association, a perhaps more likely reason that association was not observed in the recombinant inbred strain panel is that these panels are made of populations that are descended from just two inbred parent strains. As such, they can only have two alleles at any locus at which the parent strains differ. This may limit the ability to generalize results from these populations (Crabbe et al., 1990
). Inbred strain panels, as well as the genetic analysis technique we used in , do not suffer from this shortcoming because they use genetically very diverse populations. Therefore, these techniques may have more power to detect genetic associations that truly exist, providing that there are enough differing populations in the analysis.
We observed a link between home cage drinking and CPP, albeit one with somewhat more heterogeneity throughout the literature for this phenotype than for the others, both currently and historically. There was support from the quantitative analysis for this connection, but in the qualitative analysis and the recombinant inbred strain panel, we observed a near-normal distribution of findings. In CPP, a change in the amount of time spent in the environment paired with the drug is thought to indicate acquisition of an association between that environment and the effects of the drug. If a preference for the environment is observed, the drug is presumed to be rewarding, while in contrast, if avoidance of the environment associated with the drug is seen, the drug is thought to produce dysphoric effects. Overall results suggest that CPP is greater when animals voluntarily consume more alcohol, in line with the idea that both of these behaviors are related to the reinforcing effects of ethanol.
However, this result is complicated by the fact that although mice typically show a preference for ethanol-paired environments, rats typically show an aversion. When CPP was initially used to test for reinforcing or aversive effects of ethanol in rats, Black et al. (1973)
concluded that the central effects of ethanol were reinforcing. This report proved over time to be controversial. Cunningham (1981)
using the same parameters and dose later reported conditioned place aversion, suggesting that rats find ethanol aversive in this situation. Conditioned place aversion to ethanol for rats has subsequently been the most common finding, including the one rat place conditioning paper fitting the criterion for inclusion in the meta-analysis (Stewart et al., 1996
). While this review integrates findings across both rats and mice, there may be some intrinsic biological differences in response to the reinforcing and aversive effects of ethanol existing between rats and mice. Ethanol-induced CPP is fairly well established among mouse literature (though notably lacking in the C57BL/6J mouse, the prototypical high-drinking strain; see ), but the condition under which rats will exhibit an ethanol CPP remains unclear (Cunningham and Henderson, 2000
; Fidler et al., 2004
; Tzschentke, 1998
). Fidler et al. (2004)
discusses the role route of administration plays on the development of CPP. Through the implementation of an intragastric route of administration, they support the conclusion that it does not play a role, in contrast with previous work that suggested that the route of administration did play a role in selected lines of rats (Ciccocioppo et al., 1999
). Furthermore, it has also been proposed (Cunningham et al., 2002
; Cunningham and Henderson, 2000
) that both CPP, conditioned place aversion, and CTA can be produced in the same animal by the same dose of ethanol simply by varying the timing of administration of the ethanol and the nature of the cues paired with ethanol delivery (gustatory or tactile/environmental). One recent study indicated a genetic correlation between CTA and place aversion conditioning wrought by injecting ethanol after exposure to the CS, suggesting that these two phenotypes can be related when measuring the aversive effects of ethanol (Cunningham and Ignatoff, 2000
Although there are relatively few papers assessing intravenous self-administration, one of the targets of this review, these data seem to suggest that it, like CPP, may not be closely related to home cage drinking. For example, C57BL/6J mice, which self-administer more in free-choice, as well as operant paradigms, than a common comparison inbred strain, the DBA/2J mice, seem to not differ from that strain in intravenous self-administration (Grahame and Cunningham, 1997
). However, beta endorphin knockout mice, which do not drink more ethanol than their wild-type counterparts, self-administer ethanol to a greater extent intravenously than wild-types (Grahame et al., 1998
). While these are interesting findings, more work is called for in this area of research. Perhaps the lack of IVSA literature may not be surprising, considering that this method is much more time consuming and technically-challenging compared to others mentioned here, and given the fact that ethanol is typically orally self-administered in humans. However, IVSA has long been used in the study of other drugs of abuse, and has numerous advantages compared to other methods of assessing the reinforcing properties of drugs. The most important of these with respect to ethanol is the ability of this procedure to easily bypass preingestive factors such as taste and smell, which may interfere with ethanol self-administration in populations that otherwise find the pharmacological effects of ethanol rewarding (for reviews of this literature, see Thomsen and Caine, 2007 and Yokel, 1987).
When new genetically modified populations come into play on an almost daily basis, many of which are potentially related to ethanol’s rewarding effects, researchers tend to choose behavioral assays that are both easy to perform and are interpretable. The current review suggests that home cage drinking is genetically correlated with other behavioral assays - OSA, CTA, and CPP - that rely on operant and classical conditioning. Although these other measures are important, one might argue based on the present data that they do not contribute greatly, genetically speaking, to what one already knows after assessing home cage drinking alone. However, several sets of data that converge on the same conclusion increase confidence in genetic differences, and a genetic difference seen both in drinking and a more formal behavioral test such as OSA, CTA, or CPP should be treated as more robust than a difference seen in drinking alone. Notably, after detecting a difference in home cage drinking, it may be somewhat easier to detect similar differences in OSA and CTA as opposed to CPP, which showed a more modest association with home cage drinking.
Finally, these animal studies may be further extended to compare with the human literature on human family history-positive studies on alcoholism. Research on genetic variations such as dopamine D2 and D3 receptors, as well as the cannabinoid and opioid receptor systems (Blum et al., 1995
; Bowirrat and Oscar-Berman, 2005
; Manzanares et al., 2005
; Oswald and Wand, 2004
) have begun to uncover genetic influences in alcoholism in humans. Through the use of animal research and animal models such as selective breeding, transgenic and knockout models, as well as inbred strains, we may begin to gain insight into the human condition and develop treatments or prevention strategies to help control, treat, and manage alcoholism.