It has been known for many years that sensitivity and tolerance to ethanol's pharmacological effects differs between adolescents and adults, across species (reviewed in Spear, 2000
). However, the neurobiological mechanisms behind these differences are largely unknown, and most of the research examining differences in ethanol consumption in adolescents and adults use rodent models. Our studies sought to further corroborate the work done in rodents, by examining the effect of “binge drinking” experience during the adolescent period or adulthood on subsequent ethanol intake. Consistent with recent data in adolescent mice (Chester et al., 2008
; Tambour et al., 2008
), we found that ethanol naïve adolescent mice consumed significantly higher doses of ethanol, when compared to adult naïve mice, and that intake was higher in female versus male mice of both age groups. Furthermore, initial “binge drinking” exposure to ethanol during the adolescent phase produced high ethanol consumption as an adult under a two bottle choice, 24 hour access procedure, with adolescent female mice more susceptible to this effect than male mice.
In the initial phase of the experiment, male and female adolescent and adult mice received “binge drinking” ethanol experience with the SHAC procedure. Notably, the adolescent mice consumed significantly more ethanol than their adult counterparts, and this difference was further corroborated by a significant elevation in BECs over the adult mice. In both age groups, the dose of ethanol consumed not only met, but greatly exceeded the NIAAA's criteria for “binge drinking,” with BECs above 80 mg% in all cases. Thus, both age groups received initial ethanol experience consistent with “binge drinking”. Even in light of the modest evidence for age-related differences in ethanol-induced ataxia (Hefner and Holmes, 2007
; Linsenbardt et al., 2009
), the BECs achieved with the SHAC procedure would be expected to produce visible signs of intoxication (Cronise et al., 2005
). It is possible that since the SHAC procedure was performed in C57BL/6J mice, an alcohol preferring strain, we saw a more robust effect than would be observed in other strains of mice. However, we believe that this is unlikely, given that the SHAC procedure has been shown to produce high levels of intake (BECs > 100mg%) in genetically heterogeneous mice (Finn et al., 2005
In fact, our studies revealed that this early exposure to an intoxicating dose of ethanol during adolescence had a profound effect on later consumption in the same animals. This pre-exposure to ethanol in adolescent mice elevated the subsequent consumption of ethanol in an unlimited access procedure when the animals were tested as adults, when compared to ethanol intake in naïve adolescent mice as well as ethanol intake in naïve and ethanol pre-exposed adult mice. Notably, female adolescent mice were more vulnerable to this effect of prior “binge drinking” experience on subsequent ethanol intake, with the ethanol dose consumed from the 20% solution approaching 30 g/kg (mean = 29.8 g/kg; range in take 23.3 – 38.5 g/kg) in a 24-hr period. In general, female adolescent mice with “binge drinking” experience exhibited elevated drinking over all other experimental groups and following consumption of all three ethanol solutions. In male adolescent mice, “binge drinking” experience only increased consumption of the 5% ethanol solution, but intake remained lower than in similarly treated female mice. Conversely, there was no effect of prior ethanol experience on subsequent ethanol intake in the adult mice, with analyses revealing an overall main effect for intake to be higher in female than in male adult mice, regardless of ethanol experience. To our knowledge, this is the first demonstration that initial “binge drinking” experience selectively increases subsequent ethanol intake during adulthood when the initial exposure occurred during the adolescent period of development.
The mechanism(s) underlying the effect of “binge drinking” experience during adolescence to increase subsequent ethanol intake during adulthood are not known. It is unlikely that the age-related differences in ethanol sensitivity are contributing to the differences in consumption during the 24-hr access period, since at that point in testing all mice were adults and age differences were negligible. Even with the proviso that there was a difference in the range of ages upon initiation of testing for preference drinking during adulthood (adolescent experienced mice = postnatal days 58 – 60; adult experienced mice = postnatal 96 – 109 days; adult naïve mice = postnatal 58 – 82 days), the age range of the adult mice with adolescent ethanol “binge drinking” experience is compatible with reports measuring sensitivity to ethanol's behavioral effects (e.g., Linsenbardt et al., 2009
). This would suggest that the behavioral sensitivity of the adolescent experienced mice (tested during adulthood) would be comparable to that of adult rather than adolescent animals.
We did not measure BEC in animals during the unlimited access procedure, since it is difficult to pinpoint a specific time for assessment. Given that recent work provide some modest evidence for age-related differences in ethanol metabolism in C57BL/6 mice (Hefner and Holmes, 2007
; Linsenbardt et al., 2009
), it is possible that pharmacokinetic factors contributed to the enhanced 24-hr ethanol intake in the mice with prior “binge drinking” experience as adolescents. Based on the time points examined (Hefner and Holmes, 2007
; Linsenbardt et al., 2009
), there appeared to be a slight enhancement in ethanol metabolism in adolescent mice. However, the BEC measurements during the SHAC procedure would suggest that ethanol metabolism was not altered by age in the initial portion of the study, given that the age-related differences in BEC corresponded to the differences in ethanol intake. Additionally, ethanol metabolism studies conducted in our laboratory in male and female adult C57BL/6 mice following a high dose of ethanol indicated that there was no sex difference in ethanol clearance rate (Gorin-Meyer et al., 2007
), suggesting that the enhanced ethanol intake in female mice was not due to a sex difference in ethanol metabolism. Nonetheless, future studies will need to determine whether the enhanced ethanol intake during adulthood in the adolescent mice with “binge drinking” experience corresponded to enhanced metabolism.
Another possible explanation for the elevated consumption in mice subsequent to adolescent “binge drinking” experience is that age-related differences in the neurobiological effects of ethanol differ in adolescent versus adult animals. For example, rats who exhibit an early adolescent preference to alcohol displayed higher densities of serotonin receptors in cerebral-cortical and hippocampal regions, and lower densities of dopamine receptors in the ventral tegmental area (McBride et. al, 2005b
), suggesting that the neurochemistry of the brain changed as the rodents matured. Stamford (1989)
suggested that adolescent rodents have greater storage capacity for dopamine, possibly resulting in larger amounts of dopamine release during novel or risk-taking behavior in adolescent versus adult mice. In fact, it may not be the novel taste of ethanol that is rewarding to adolescents, but the pharmacodynamic properties of ethanol in motivational circuits. There is some evidence to suggest that there is an ontogenetic difference in the expression of type 1 endocannabinoid receptors, which are directly implicated in a hierarchial system involving opioid, cannabinoid, and some of the GABA systems that are thought to mediate the hedonic properties of rewarding stimuli (Peciña et al., 2006
). Endocannabinoid receptors peak during adolescence in striatum, ventral mesencephalon, and limbic systems, but are lower in the nucleus accumbens and hippocampus compared to adults (Rodríguez de Fonseca et al., 1993
; Romero et al., 1997
). Adolescents may be more sensitive to endocannabid receptor agonists due to lower levels in certain areas, which results in compensatory upregulation of these receptors in “hedonic hot spots”. Subsequently this could contribute to the higher hedonic sensitivity seen in adolescents (see Wilmouth & Spear 2009
). Related to this point, a recent review of the rodent literature suggests that the balance of rewarding and aversive effects of drugs of abuse is tipped toward reward in adolescence (Schramm-Sapyta et al., 2009
Evidence shows that adolescents across a wide range of species are known to participate more frequently in “risky” behaviors such as substance abuse, sexual promiscuity and other novelty seeking behavior (Arnett 1992
; Wilson and Daly, 1985
), which could predispose some individuals to later addiction (see Introduction). This has recently been characterized in mice (for a review, see Laviola et al., 2003
) with regard to substance abuse. Specifically, Adriani et. al (2002)
reported greater nicotine intake by adolescents, without a concomitant increase in anxiety (measured by circulating levels of corticosterone) that could lead to stress related increases in nicotine intake. Adolescent rodents have also shown an increase in sensitivity to the effects of cocaine (Zakharova et al., 2009
). This propensity to engage in risk-taking and novelty-seeking behavior without an apparent concomitant increase in anxiety is common in adolescents, and may be related to the increase in ethanol intake reported in this study.
In addition to the effect of age on ethanol consumption, a significant sex difference emerged. In both adolescent and adult mice, females showed a marked increase in 24-hr ethanol intake versus respective male mice. As already noted, pre-exposure to “binge drinking” experience only increased ethanol intake in the adolescent females. Sex differences in ethanol intake has been well-documented in rodents, with intake in females higher than that in males (e.g., Belknap et al., 1993
; Chester et al., 2008
; Finn et al., 2004
; Lancaster et al., 1996
; Yoneyama et al., 2008
). Sex-dependent sensitivity to other drugs has been shown, with females generally being more sensitive to the rewarding effects of drug than males (see reviews by Becker & Hu, 2008
; Carroll et al., 2004
; Fattore et al., 2009
). For example, female rats acquired intravenous self-administration of cocaine, methamphetamine and nicotine faster than males. Female rats also self-administered more cocaine with longer duration of “binges” and greater loss of circadian control over drug intake in an escalation model, and showed greater extinction responding on the drug associated lever after drug removal and greater reinstatement after drug priming than males (see Becker & Hu, 2008
; Carroll et al., 2004
; Fattore et al., 2009
and references therein). Along these same lines, women tend to increase rates of consumption of alcohol, marijuana, opioids and cocaine more rapidly than men do, and also have a harder time quitting after addiction has been established than do men (Brady et al., 1999
). In contrast, women abuse drugs at lower rates when compared with men, but this could reflect differences in opportunity, rather than vulnerability to drug use (Becker & Hu, 2008
). So it is possible that the mechanisms of sex differences in alcohol abuse in human subjects has not been fully characterized and that rodent studies conducted under more controlled circumstances can point clinical researchers in a direction not yet fully explored.
Certainly, the endogenous milieu as well as the endogenous steroid responses to alcohol could be contributing to the differences in intake that we have observed. Laviola et al. (2002)
reported that corticosterone (CORT) levels in adolescent mice were higher than in adults, an effect that was particularly robust in the males, and that the CORT response to acute stress was reduced in females when compared with males. One possibility is that this sex and age-dependent stress contributes to the present findings. Catherine Rivier has shown that the hypothalamic-pituitary-adrenal axis is activated by ethanol consumption, but differentially between the sexes. Females generally have higher circulating levels of CORT, adrenocorticotrophic hormone (ACTH; an upstream regulator of CORT) and estrogen and secrete more CORT and ACTH than males in response to stress and alcohol administration (Ogilvie & Rivier, 1997; Handa et al., 1994; Galluci et al., 1993). Levels of testosterone are also altered by ethanol administration. In addition to sex differences, there seems to be a variable effect. Both acute and chronic ethanol administration was reported to lower plasma testosterone levels in males, mainly due to the inhibition of luteinizing hormone in the testes (Ylikahri et al., 1980
; Rivier, 1999
). Apter & Eriksson (2003)
reported a decrease in plasma testosterone levels after acute ethanol treatment in alcohol-preferring and non-alcohol-preferring rats, but mean testosterone levels were higher in the alcohol-preferring rats. Conversely, Alomary et al., (2003)
, one of the only studies to also investigate brain levels of testosterone, reported increases in both plasma and brain testosterone after acute ethanol administration. Along these same lines, Sarkola et al. (2000)
reported an increase in plasma testosterone in premenopausal women after an acute ethanol administration. Consistent with the idea that there are sex differences in the effect of ethanol on testosterone levels, recent preliminary findings indicate that chronic ethanol exposure and withdrawal significantly increase testosterone levels in female mice, while decreasing levels in male mice (Hashimoto et al., 2009
With regard to female sex hormones, alcohol has been shown to precipitate a rise in estrogen levels in premenopausal women (reviewed in Gill, 2000
). Acute administration of ethanol also increased levels of pregnenolone (a precursor to progesterone) and progesterone (O'Dell et al., 2004
; Korneyev et al., 1993
). In light of the present studies, sex hormones could be partially responsible for the differences in alcohol consumption we report, but since the literature is inconclusive at the present time, it is difficult to make any direct interpretations.
We did not monitor estrus cycles in the present study, but there was no change in drinking patterns that could be related to hormonal changes associated with stage of the estrous cycle. Results by Tambour et al. (2008)
indicated that ethanol intake was significantly higher in female adult and adolescent mice versus respective male counterparts after approximately 3 weeks of alcohol experience. This finding suggests that alcohol history (experience) was more important than the onset of puberty to the sex differences in ethanol intake. Additionally, the higher ethanol intake in female mice during adolescence, which likely exhibited inconsistent estrous or non-existent estrous cycles, would suggest that the overall higher ethanol intake in female mice was not related to a specific stage of the estrous cycle.
Interestingly, no significant effect of age or sex on ethanol intake was observed in the limited access two-bottle choice paradigm. This result was surprising, considering the highly significant divergence seen in the continuous access 24-hr ethanol preference procedure. One possible explanation could be that the pattern of ethanol intake differed in the adolescent versus adult animals so that an examination of 2-hr ethanol intake might not capture the true differences in consumption that would be evident over a 24-hr period. For instance, we have used lickometers to examine the microarchitecture of ethanol drinking behavior in male and female C57BL/6 mice and observed sex differences in bout frequency versus bout size when animals were on a 2-hr limited access schedule (Ford et al., 2005
). Perhaps more relevant to the present findings, we recently examined ethanol intake patterns in male and female wildtype and knockout mice with a null mutation in the 5α-reductase-1 gene and found that the effect of sex and genotype on ethanol intake differed, when assessed with 24-hr versus 2-hr periods of ethanol preference drinking (Nickel et al., 2006
). Thus, an alteration in the pattern of ethanol intake across time might not reveal a sex or age-related difference in ethanol intake, when the analysis was limited to a 2-hr limited access session. This possibility could be examined in future studies utilizing lickometers to assess the pattern of ethanol intake across 24-hr ethanol sessions.
Taken together, “binge drinking” ethanol experience during the adolescent period significantly increased subsequent ethanol intake during adulthood, with female mice more susceptible to this effect. The enhanced vulnerability of female mice is consistent with evidence indicating that female rodents are more sensitive to the rewarding effects of drugs than males during all phases/models of drug abuse (see Becker & Hu, 2008
; Carroll et al., 2004
; Fattore et al., 2009
). While the potential contribution of pharmacokinetic mechanisms to these differences in ethanol intake are unclear, it is possible that fundamental differences in neurobiology, as well as an increased propensity for novelty-seeking and “risky” adolescent behavior could be contributing to the differences seen in adolescent versus adult mice. Based in the important interaction between sex, age, and “binge drinking” ethanol experience on subsequent ethanol intake, research aimed at understanding the mechanisms contributing to the effects of these factors (alone or in combination) on ethanol drinking behavior will be important in the development of pharmacotherapies for the treatment of high alcohol intake.