All animal care and euthanasia procedures were performed in accordance with the requirements and recommendations presented by the United States Department of
Agriculture Animal Welfare Act and Regulations, Public Health Services Policy on Humane Care and Use of Laboratory Animals, and American Association for Accreditation of
Laboratory Animal Care**. Humane endpoints were established by the same standards.
Animals
All mice were obtained from Jackson Laboratories at 8 weeks of age and were initially habituated to the animal vivarium for 1-2 weeks in group housing (n=4) before then being habituated to single housing for 1 week prior to experimentation. Mice were housed with 12-hour light-dark cycles and ad libitum access to water and standard chow (#7912, Harlan Teklad, Madison, WI, United States), in humidity-and temperature-controlled rooms with shaved cedar bedding. Cages were changed and body weights were recorded weekly.
Due to availability of strains from the supplier, two separate experiments were run for complete analysis of all progenitor strains. Experiment 1 contained 10 mice from each of the following strains: 1) C57BL/6J, A/J, 129S1/SvlmJ, NOD/ShiLtJ, NZO/HiLtJ, and CAST/EiJ. For Experiment 2 mice from the two remaining progenitor strains (WSB/EiJ and PWK/PhJ) were obtained from Jackson Laboratories on a later date. C57BL/6J mice were also included in this group, as a standard comparison.
Experimental Design
Experiment 1
Five mice from each strain were exposed to a Chronic Ethanol Access (2) paradigm, and five were exposed to an Intermittent Ethanol Access (IEA) paradigm, each with three- bottle choice (15% and 30% ethanol v/v, and water), for 8 weeks. Ethanol mixtures were composed of anhydrous ethanol and tap water. Mice undergoing CEA had constant access to the three bottles, whereas mice undergoing IEA had 3 periods of 24-hour access (12 hours in the dark, then 12 hours in the light) each week (Monday, Wednesday and Friday), alternating with periods of access to three bottles of water only, as seen in Hwa et al. (36) (Fig. 1). Ethanol and water consumption were measured for each 24-hour period of access under IEA, for both paradigms, resulting in 3 recordings per week and 23 recordings total for each paradigm. On IEA ethanol access days, placement order of bottles was randomized for both paradigms to avoid place preference confounds. Bottles of each fluid were placed in
uninhabited cages to measure evaporation and spillage. Measurements from these bottles were taken concurrently with consumption measurements, and were subtracted from that day’s consumption values.
Figure 1: Experimental paradigms for Chronic Ethanol Access (2) and Intermittent Ethanol Access (IEA). Each block represents a day of the week, for which horizontal stripes represent the dark phases of the light-dark cycle, and solid areas represent light phases. White space represents periods during which mice had access to water only, and blue space represents period of ethanol access (1 bottle 15% EtOH, 1 bottle 30% EtOH, and 1 bottle water). These paradigms were repeated 8 times, resulting in 8 weeks of IEA or CEA.
On the 24th ethanol access day of IEA, ethanol bottles were removed for all mice to
begin a 7-10 day period of forced abstinence. At the end of this period, mice received
intraperitoneal injections of 4 g/kgof 20% v/v ethanol, for pharmacokinetic analysis. Trunk
blood was collected during euthanasia via cervical dislocation and decapitation, which was staggered across the following time points: 30 mins, 60 mins, and 240 minutes. One to two mice per strain from each treatment group were euthanized at each time point. Blood Ethanol Concentration (BEC) was quantified via Analox AM1 Analyzer (Analox Instruments, North
Yorkshire, UK), and recorded in mgEtOH/dLblood (or %BEC).
Experiment 2
Because there were no significant differences between paradigms in Exp. 1, all mice in Exp. 2 were exposed to CEA, to increase power. Mice were exposed to 5 weeks of the same CEA paradigm described in Exp. 1. Additionally, to gather information about possible binge-like ethanol consumption in the PWK and WSB wild-derived strains, we measured ethanol and water consumption at 3 hrs after bottle placement, in addition to 24 hrs (as in
Exp. 1), and the additional time point of 3 hrs, after bottle order was changed for 5
consecutive days per week, to characterize the pace at which they consumed ethanol on the final day of CEA. After 3 weeks of CEA access, bottles were removed to begin a one-week period of forced abstinence.
As done in Exp. 1, following the week of forced abstinence, mice received
intraperitoneal injections of 4 g/kgof 20% v/v ethanol. Trunk blood was collected during
euthanasia via cervical dislocation and decapitation, which was staggered across the
following time points, with 2 mice per time point: 15 mins, 30 mins, 60 mins, 120 mins, and 180 mins. These time points, which differ from those used in Exp. 1, were chosen to achieve a better estimation of the shape of the metabolism curve. Due to deaths by unknown causes during the abstinence period, data was collected from only one mouse for C57/BL6J at 15mins and PWK at 30mins. BEC was quantified using the same methods as Exp. 1.
Statistical Analysis
Experiment 1
Ethanol consumption was measured as gEtOH/kgmouse, to account for weight
differences. Total ethanol consumption was calculated by summing the 15% and 30% ethanol consumption values. Total ethanol preference was calculated as
mLTotalEtOH/mLTotalFluid consumed. For heritability analysis and paradigm effect testing, each
mouse’s total ethanol consumption and preference were averaged across the entire study. Because these mice were ethanol-naïve on the first day of ethanol access, this day was excluded from these calculations. All statistical analyses were run using the “stats” and “DescTools” R packages (http://www.R-project.org/).
Heritability was calculated using one-way ANOVA, such that broad-sense heritability
(h2) = SSB/(SSB+SSW), where SSB and SSW represent the sum of squared error between and
within strains, respectively ((149, 150)). Heritability was calculated for consumption and preference for each paradigm separately, and significance was determined by the p-value for mouse strain effect on the trait. Tukey Post-Hoc Test was performed to determine which strain differences contributed to significant heritability estimates.
Because we were interested in the distinct effect of paradigm on consumption and preference within each strain, we tested this effect via two-tailed Welch’s t-test for each individual strain. We calculated False Discovery Rates to account for multiple testing within each trait.
To test for an increase in ethanol consumption or preference over time for each strain, every mouse’s daily values were averaged over each the first and last week of ethanol access. We ran repeated-measures two-tailed Welch’s T-tests for consumption and preference within each strain, under CEA and IEA separately. False Discovery Rates were calculated to
account for multiple-testing within each paradigm for each trait. We were further interested in the difference in this effect across strains, also known as a gene x environment interaction (151). Where at least one strain showed significant or suggestive time (i.e. ethanol exposure duration) effects, we tested for gene x environment interaction (i.e. time x strain interaction) via two-way ANOVA. Although the main effects of strain and time period were not of interest here, they were included in each model to preclude false inflation of the interaction effect.
Pearson correlations were used to assess the relationship between mean total ethanol consumption and BEC, at each post-injection time point. Because there were < 3 mice per
strain at some time points, the contribution of metabolic variation to strain differences was assessed qualitatively. This was done by calculating the cumulative BEC across all post- injection time points and the ethanol elimination rate for each strain. Cumulative BEC was estimated by the area under the curve, based on strain mean BEC at each time point. Elimination rate was estimated by the slope of the same curve from point of highest
measured BEC for that strain, to the final time point, calculated by linear regression. Because paradigm did not significantly effect ethanol consumption or preference (see Results), and all mice had been abstinent for ≥1 week, data were collapsed across treatment group for this analysis.
Experiment 2
Whole-study mean total ethanol consumption was calculated as seen in Exp. 1, for the 3hr and 24hr time points. The proportion of daily ethanol consumed during the first 3hrs of access was calculated for each day. These values were averaged across the entire study, again excluding the first day of ethanol access.
The same algorithms presented in Exp. 1 were used for the following analyses: heritability of mean total ethanol consumption and preference, and the associated post-hoc comparisons; change in consumption over time (first v. last week of ethanol access); and strain-specific ethanol kinetics and the relationship between BEC and total ethanol consumption. One-Way ANOVA was used to test for between-strain differences in proportion of daily ethanol consumed in the first 3hrs of access.