In home-cage self-administration procedures, animals simply approach and consume ethanol. In contrast, operant procedures require animals to engage in an explicit "seeking" response (e.g., bar press or nose poke) to gain access to ethanol (13,14). Although ethanol is usually consumed orally, operant models allow for the possibility of administering ethanol directly into the stomach, blood, or brain via surgically implanted can-nulae (15). Thus, in contrast to home-cage procedures, operant models allow one to separate ethanol "consumption" from ethanol "seeking" and, if desired, to eliminate the oral route entirely. The value of the latter feature is well illustrated by a recent study showing that two mouse strains known to differ dramatically in home-cage etha-nol intake (DBA/2 and C57BL/6) showed little difference when nose poking produced intravenous ethanol injections in an operant procedure (16). This finding supports the suggestion that aversive orosensory (preabsorptive) effects of orally administered ethanol contribute to the normally low home-cage intakes of DBA/2 mice (17).
Interpretive issues raised in the study of home-cage oral self-administration generally also apply to operant studies using the oral route (e.g., concern over the temporal pattern of ethanol intake). In both models, investigators will sometimes add a sweetener or other flavor to the ethanol solution in an effort to increase overall intakes. Although these flavor additives may be "faded out" over time (18), use of this strategy in genetic studies raises the possibility that strain differences are caused by differences in sensitivity to the added flavor rather than to postingestive effects of ethanol. Again, as suggested earlier, this hypothesis can be addressed by examining responding for the flavor additive in the absence of ethanol.
One difficulty in the interpretation of both home-cage and operant self-administration procedures is that increases (or decreases) in the target behavior produced by a genetic manipulation do not unambiguously reflect increases (or decreases) in ethanol reward. This ambiguity arises due to the inverted U-shaped relationship between ethanol intake and variables that presumably affect ethanol's reinforcing efficacy, such as dose or concentration (19,20). Thus, a genetic manipulation that reduced ethanol reward (e.g., a null mutation) could either increase or decrease ethanol intake, depending on where control intakes fell on the concentration/dose-response function. Strategies for addressing this problem include examining a range of ethanol doses/concentrations and using conditioning procedures to provide converging evidence on effects of the genetic manipulation (21-23). A unique feature of the operant model is its potential to separate genetic differences in appetitive processes underlying ethanol-seeking behavior from consummatory processes involved in the regulation of ethanol intake (14). This feature could be important for determining whether different genes influence appetitive and consummatory processes. However, as recently noted (24,25), much of the literature on operant self-administration of ethanol has not allowed a clear separation of these processes, due to frequent alternation between "seeking" and "consuming" within self-administration sessions. Although promising alternative procedures involving chain schedules have recently been introduced (24,25), these schedules have not yet been used in the study of genetic differences.
Was this article helpful?