Two hundred and forty-nine FCG mice (PND 60+) were bred from breeding pairs consisting of Sry + male and C57BL/6 J (wild type) female mice at the Laboratory of Animal Resources at Miami University. Four groups were bred: XX/Sry–, XY/Sry–, XX/Sry+, and XY/Sry+. For genotyping, mice were lightly anesthetized with isoflurane and then ear tissue samples were collected and sent to Transnetyx (Cordova, TN).

In adulthood, mice were transferred to a temperature-controlled room. For the home cage drinking experiment, they were kept in a room that was on a 12:12 h reverse light:dark cycle (i.e., lights off at 7 AM). For the operant experiments, mice were kept in a room that was on a 12:12 light:dark cycle (i.e., lights on at 7 AM). Prior to experimentation, mice were group housed by gonad type (ovaries or testes). For the home cage drinking experiment, at least 3 days before experimentation, mice were individually housed in standard shoe box udel polysufone rectangular mouse cage (18.4 × 29.2 × 12.7 cm) with 5.08 × 5.08 cm nestlets and Bed-O-Cobb 0.634 cm bedding (Cincinnati Lap Supply, Cincinnati, OH). All mice received standard care and had access to Rodent Diet 5,001 chow (Cincinnati Lap Supply, Cincinnati, OH, United States) and reverse-osmosis (RO) drinking water ad libitum unless specified otherwise. All mice were cared for in agreement with the guidelines set by the National Institutes of Health and all procedures were approved by the Institutional Animal Care Use Committee (IACUC) at Miami University.

EtOH was prepared volume/volume in RO water at concentrations of 5, 10, 15, and 20%. Sucrose was prepared weight/volume in RO water at concentrations of 5 and 10%. Quinine hemisulfate (Q1250-50G, Millipore-Sigma, ST. Louis, MO, United States) solutions were prepared as 100, 250, and 500 μM in EtOH or RO water. All solutions were made fresh prior to each testing session.

Mice (XX/Sry–: n = 14, XY/Sry–: n = 14, XX/Sry+: n = 14, and XY/Sry+: n = 14) were given access to water and 15% EtOH 3 h into the dark:light cycle (i.e., 10 AM) in a two-bottle, limited access DID paradigm for 4 h and for a total of 15 drinking sessions (Figure 1A). Bottles were weighed at 30 min, 2, and 4 h. Aversion-resistant drinking was tested by adding increasing quinine concentrations from 100, 250 to 500 μM for 5 drinking sessions per concentration (i.e., 15 quinine sessions total). Mice were weighed prior to testing and bottles were alternated daily to equate side biases. Mice with ovaries (Sry–) were scruffed two times a week following drinking to habituate them to the scruffing process (28, 29). Sry– mice and Sry + mice were tested in separate cohorts.

The estrous cycle was monitored on the last 4 sessions for each quinine concentration in the home-cage drinking experiment. Vaginal cells were collected in mice with ovaries (Sry–) 10 min before bottles were placed on cages. The sample air-dried for 24 h and was then stained with a DipQuick staining kit (30-s/solution; Jorgensen Laboratories, Loveland, CO). Determination of the phase of the cycle was based on the pattern of cell types of vaginal smear samples (Figure 2E) using a Nikon Labophot Phase Contrast light microscope (Diagnostic Instruments, Inc., Sterling Heights, MI) (30).

The operant procedures used were similar to those used in Sneddon et al. (2020). The conditioning chambers measured 6″ × 5.25″ × 5″ (Med Associates, Fairfax, VT, ENV-307A). The grid floor was made of 19 metal rods separated by 0.31″ (Med Associates, ENV-307A-GFW). Reward and activity responses were recorded through 2 nose poke holes located on one wall of the chamber. The left hole was considered active and resulted in reward delivery, while the right was inactive. A reward receptacle located on the same wall between the nose poke holes dispensed grain pellets (Bio-Serv, Flemington, NJ, F05684) from a pedestal-mounted pellet dispenser (Med Associates, ENV-203-20) and liquid solutions from a 20-ml syringe. Following a nose poke on the active side, 50 μl of solution was pushed into a well inside the reward receptacle via a single speed syringe pump (Med Associates, PHM-100) over 1.5 s. The wells were checked at the end of every session to confirm consumption. A light fixed in the reward receptacle turned on at the start of each session and remained on for the entire 30 min session (Med Associates, ENV-303RL). The rest of the chamber remained dark.

Food restriction, to increase stable engagement with the operant task (31), began 2 weeks prior to test and continued during operant testing to maintain body weights at 85% of each mouse’s free-feeding weight. Mice were given measured quantities of food once daily after sessions. Training and testing sessions took place once daily from Monday to Friday 3 h into the light cycle (i.e., 10 AM). All training sessions lasted for 30 min and testing chambers were cleaned with 70% EtOH in between each session. The well was checked after each session to ensure there was no liquid remaining.

The first 3 days of training consisted of the mice responding for a grain pellet on a fixed ratio (FR) 1 schedule. A sucrose fading procedure was implemented on session 4. This procedure facilitates responding and encourages the intake of EtOH over the course of operant conditioning by first introducing a high-value reward (32–36). Mice responded for 10% sucrose on an FR1 schedule for 3 sessions followed by FR3 for 3 sessions. For all following sessions, mice responded on an FR3 schedule. Then, 10% EtOH was added to the solution and sucrose was faded out. For the experiment assessing responding for escalating concentrations of EtOH, mice were presented with the following solutions during response training: 10% sucrose, 10% sucrose + 10% EtOH, and 5% sucrose + 5% EtOH. For the experiment assessing responding for quinine-adulterated EtOH, mice were presented with the following solutions during response training: 10% sucrose, 10% sucrose +10% EtOH, and 5% sucrose +10% EtOH. Each solution was administered over a minimum of 3, but no more than 5, sessions. If the responses produced a coefficient of variation <20% after 3 consecutive sessions, the mice were moved on to the next solution.

Responses for escalating concentrations of EtOH in the operant chamber were assessed in FCG mice (XX/Sry–: n = 10, XY/Sry–: n = 10, XX/Sry+: n = 10, and XY/Sry+: n = 10). Following response training, responses for escalating concentrations of EtOH were assessed as follows: 5% EtOH, 10% EtOH, 15% EtOH, and 20% EtOH (each for 5 sessions). Two mice (XX/Sry + = 1, XY/Sry– = 1) accidentally received 10% EtOH for five sessions before the 5% concentration was introduced (Figure 3A), but as their responses were similar to other mice of their genotype they were included in the analysis.

Responses for 10% EtOH adulterated with escalating concentrations of quinine were assessed in a separate cohort of FCG mice (XX/Sry–: n = 12, XY/Sry–: n = 19, XX/Sry+: n = 13, and XY/Sry+: n = 14). Following response training, mice responded for 10% EtOH for a minimum of 3 and a maximum of 5 sessions, depending on the previously mentioned coefficient of variation threshold. The mice were then given increasing solutions of 10% EtOH + quinine for 3 sessions each as follows: 10% EtOH +100 μM quinine, 10% EtOH +250 μM quinine, and 10% EtOH + 500 μM quinine (Figure 4A). Partway through the experiment, once it became apparent that FCG mice were potentially less sensitive to quinine, a final footshock session was added to the protocol. Thus, in a subset of mice (XX/Sry–: n = 3, XY/Sry–: n = 5, XX/Sry+: n = 3, and XY/Sry+: n = 2), responding for 10% EtOH was paired with a 0.35 mA footshock on one final test session.

Quinine sensitivity was assessed in a naive cohort of FCG mice (XX/Sry–: n = 14, XY/Sry–: n = 14, XX/Sry+: n = 14, and XY/Sry+: n = 14). Mice had access to two bottles of RO water for 4 h, similar to the home-cage EtOH drinking experiment outlined above. On sessions 2–4, quinine hemisulfate was added to the water in increasing concentrations (100, 250, and 500 μM; Figure 5A); as previously described in (24, 27–29, 37). Bottles were alternated daily to equate side bias.

To assess the sedative effects of EtOH in the FCG line, 39 EtOH-naïve mice (XX/Sry–: n = 9, XY/Sry–: n = 10, XX/Sry+: n = 10, and XY/Sry+: n = 10) were tested for the loss of righting reflex (LORR) (38). Mice were transferred to the experimental room 1 h before testing and were then weighed and given a 3.5 g/kg i.p. injection of 20% EtOH (v/v). A timer was started at the time of injection and the mice were placed in the supine position. When a mouse was no longer able to right itself onto all four paws, the latency to LORR was logged. When the mouse righted itself twice in 30 s, the time was logged as the LORR duration.

Immediately after a mouse regained the righting reflex, as described above, blood was taken using the tail clip technique (20). Blood samples were kept on ice and then were centrifuged using an accuSpinMicro17 microcentrifuge (Fisher Scientific) at 3,000 rpm for 8 min. Serum was pipetted into a sterile centrifuge tube and was then stored at −80°C until testing. An AM1 alcohol analyzer (Analox Technologies) was used to assess blood EtOH concentrations (BECs).

In the home-cage drinking experiments, consumption was calculated as (Initial Bottle Weight – Post Bottle Weight) – Average of Dummy Bottles and was expressed as grams of EtOH or mL of water per kilogram of body weight. To assess front-loading behavior for EtOH, each timepoint was normalized to match consumption in 30 min using the following calculations: 2-h timepoint as ((2-h timepoint – 30-min timepoint)/3) and 4-h timepoint as ((4-h timepoint – 2-h timepoint)/4). Preference was calculated as ((Volume of EtOH)/(Volume of EtOH + Water Consumption))*100. The 0 μM baseline consisted of an average of the last 5 EtOH drinking sessions (11–15) prior to quinine-exposure. Water consumption was averaged across drinking sessions. Initially, data were analyzed for all timepoints, 30 min, 2 h, and 4 h. Since the statistical trends were the same across all timepoints, only data from the 4 h timepoint are reported.

Consumption and preference for EtOH and water data were analyzed separately by gonad type since the mice with ovaries (Sry–) received an additional manipulation (vaginal lavage) to collect estrous cycle data and were run separately from mice with testes (Sry+). Data from bottles spilling or measurement error were excluded from all analyses. For the first 15 sessions of home-cage drinking, repeated measures (RM) Two-Way analysis of variance (ANOVA) or a Mixed Effects ANOVAs with the chromosome complement as the between-subjects factor and session as the within-subject factor were used. Mixed Effects ANOVA was used in the case of missing values as it gives the same results as a RM ANOVA and can be interpreted as such. Data for the quinine sessions were first analyzed at each concentration with session as the within-subjects factor and chromosome complement as the between-subjects factor. Since no interactions with session were observed, analyses of average consumption/preference are presented with quinine concentration as the within-subjects factor and chromosome complement as the between-subjects factor.

For estrous cycle data, a Mixed Effects ANOVA with estrous phase as the within-subjects factor and chromosome complement as the between-subjects factor was used. Mice were excluded if the phase of the cycle could not be determined for a particular quinine session.

For operant response experiments, data are expressed as the number of entries made into the active nose poke hole, inactive nose poke hole, or reward receptacle per session. For each mouse, responses were averaged over the last 3 sessions of each solution presented. Mice that did not make a minimum of 10 responses for the lowest concentration of EtOH presented (averaged over the final 3 sessions) were excluded from all analyses. Two data points are missing from the head entry data in the quinine experiment due to technical malfunction. For each experiment, data were analyzed using RM Three-Way ANOVA with concentration (EtOH or quinine) as the within-subject factor and chromosome complement and Sry as the between-subject factors. For the final test of responding with footshock, shock amplitude was used as the within-subjects factor.

For the EtOH sensitivity test, latency to LORR was measured as the time for the subject to sedate and lose the righting reflex. Duration of LORR was measured as how long it took the subject to right itself following LORR and was measured independently of LORR. Two-Way ANOVA with chromosome complement and Sry as the between-subject factors was used for the latency to LORR, duration of LORR, and BEC analyses.

For cases where sphericity was violated (ε < 0.75), the Greenhouse–Geisser correction was applied to the results of the ANOVA test. Post hoc tests and planned comparison’s corrected for multiple comparisons were used as appropriate (39–41). Specifically, a Holm Sidak’s test was used to compare groups and a Dunnett’s test was used to assess frontloading and quinine-resistant consumption/responding. All data were expressed as mean ± standard error of the mean and was analyzed using GraphPad Prism v. 9.0 (La Jolla, CA).

To assess quinine sensitivity, naïve FCG mice underwent a quinine sensitivity test where escalating concentrations of quinine were added to water. A RM Three Way ANOVA on consumption found a main effect of quinine concentration (F_{(1.992, 103.605)} = 6.061, p = 0.003) and no main effect of Sry (F_{(1, 52)} = 0.305, p = 0.583) or chromosome complement (F_{(1, 52)} = 0.0580, p = 0.811). The interaction between Sry and quinine concentration almost met the threshold of significance (F_{(3, 156)} = 2.491, p = 0.062). All other interactions were not significant (all p > 0.1559). To further explore these data, for Sry– mice, when assessing consumption of water adulterated with quinine in 4 h, a RM Two Way ANOVA found a main effect of quinine concentration (F_{(1.450, 37.71)} = 6.929, p = 0.006) but no main effect of chromosome complement (F_{(1, 26)} = 0.001, p = 0.973) and no interaction between these factors (F_{(3, 78)} = 1.029, p = 0.385). A post hoc Dunnett’s test found that both XX and XY mice in the Sry– group decreased consumption for 500 μM quinine compared to 0 μM quinine (p = 0.034; Figure 5B). For Sry + mice, when assessing consumption of water adulterated with quinine in 4 h, a RM Two Way ANOVA found no main effects of quinine concentration (F_{(1.631, 42.405)} = 0.8755, p = 0.404) or chromosome complement (F_{(1, 26)} = 0.082, p = 0.777) and no interaction between these factors (F_{(3, 78)} = 0.763, p = 0.518; Figure 5C).

A RM Three Way ANOVA on preference found a main effect of quinine concentration (F_{(2.835, 147.423)} = 5.531, p = 0.002) and Sry (F_{(1, 52)} = 17.817, p < 0.0001) and no main effect of chromosome complement (F_{(1, 52)} = 0.040, p = 0.841). All interactions were not significant (all p > 0.1425). To further explore these data, for Sry– mice a RM Two Way ANOVA revealed a main effect of quinine concentration (F_{(2.438, 63.395)} = 8.196, p = 0.0003) but no main effect of chromosome complement (F_{(1, 26)} = 0.031, p = 0.862) and no interaction between these factors (F_{(3, 78)} = 0.429, p = 0.733). A post hoc Dunnett’s test identified that mice with the XX (p = 0.034) and XY (p = 0.002) chromosome complement in the Sry– group showed a decline in preference for the 500 μM quinine compared to 0 μM quinine (Figure 5D). For Sry + mice a RM Two Way ANOVA identified no main effects of quinine concentration (F_{(2.575, 66.949)} = 0.871, p = 0.447) or chromosome complement (F_{(1, 26)} = 0.094, p = 0.762) and no interaction between these factors (F_{(3, 78)} = 0.0803, p = 0.971; Figure 5E).

When evaluating latency to LORR, a Two Way ANOVA found no main effects of Sry (F_{(1, 35)} = 1.481, p = 0.232) or chromosome complement (F_{(1, 35)} = 2.036, p = 0.163) and no interaction between these factors (F_{(1, 35)} = 0.033, p = 0.858). For duration of LORR, a Two Way ANOVA identified no main effects of Sry (F_{(1, 35)} = 0.636, p = 0.431) or chromosome complement (F_{(1, 35)} = 1.330, p = 0.267) and no interaction between these factors (F_{(1, 35)} = 0.627, p = 0.434; Table 1).

Blood EtOH concentrations measured after mice regained the righting reflex did not differ among FCG mice. A Two Way ANOVA revealed no main effects of Sry (F_{(1, 36)} = 0.228, p = 0.636) or chromosome complement (F_{(1, 36)} = 2.544, p = 0.119) and no interaction between these factors (F_{(1, 36)} = 0.730, p = 0.399; Table 4).

The role of sex chromosomes in binge-like drinking was assessed separately in Sry– and Sry + mice using a two-bottle choice DID paradigm. For the Sry– mice, no chromosomal influences were observed across the 15 drinking sessions and drinking increased across sessions. Frontloading behavior did not differ between the genotypes. For the Sry + group, the mice with XX chromosomes consumed less 15% EtOH across the 15 drinking sessions and drinking varied across session for both groups, but no differences in frontloading behavior were observed. These results differ from previous studies of EtOH consumption in the FCG mice. In a study using FCG mice drinking 10% EtOH in a single 30-min session, no effect of sex chromosomes was observed (19). Also at odds with the current results is our previous report of greater consumption of 20% EtOH under continuous access conditions in XX mice regardless of gonadal sex (20).

When assessing preference for 15% EtOH vs. water, XY mice had greater preference across sessions in mice of both gonad type. In the Sry– mice, this effect is due to lower water consumption in the XY mice, as EtOH consumption did not differ by chromosomal sex. For the Sry + group, no differences in water consumption were observed between the genotypes, therefore preference differences can be explained by XY/Sry + mice drinking more 15% EtOH across sessions than their XX counterparts. Although differences in water consumption occurred independently of differences in EtOH intake in the current study, we have previously observed that gonadal sex increases consumption of both fluids in mice with ovaries (20, 42). This observation raises the question of whether commonly observed sex differences in EtOH drinking might sometimes be the result of more general increases in fluid consumption in female animals. Considering these effects, we think it is paramount for future investigations of sex differences in alcohol drinking behaviors to include appropriate controls for nonspecific effects on fluid consumption.

As noted above, both the consumption and preference findings are inconsistent with our prior work demonstrating greater consumption and preference in XX mice (20). A major difference between that study and the current one is the drinking schedule employed. Here, we studied binge-like drinking using a limited-access DID task while our previous study examined consumption during continuous access. Because the DID promotes greater and more rapid consumption than continuous access, this model produces BECs that are more pharmacologically relevant than standard continuous exposure models (43). Intermittent exposure to drugs of abuse is also thought to produce neuroadaptations that are critical for the development of addictive behavior, including in rodent models of alcohol drinking (44–49). Thus, the opposing results in the FCG mice highlight the fact that different EtOH drinking paradigms likely tap into separate but overlapping neurobiological mechanisms, such that they can be regulated by the same manipulation but in different directions. Additional studies with other exposure regimens and drugs of abuse may further clarify this point.

For the present experiment we ran our Sry– and Sry + groups separately and, therefore, could not statistically compare the differences between these groups based on gonad type. It is therefore unclear if there were any gonadal effects on EtOH or water consumption and preference in the DID paradigm, as have been previously observed using other drinking tasks (19, 20, 50–52). However, the finding that sex chromosome complement influenced binge-like consumption in Sry + mice only suggests that chromosomal sex likely interacts with gonadal status to control this behavior.

Chromosomal sex also influenced quinine-resistant consumption of 15% EtOH in the DID paradigm, although a major caveat of this study is that Sry + mice demonstrated insensitivity to quinine when it was presented in water. It is possible that Sry + mice are innately insensitive to the concentrations of quinine used here, as has been observed in lines of mice selectively bred for alcohol preference (53), and that higher concentrations are necessary to reduce drinking in these animals. Thus, although it may appear that XX/Sry + mice were more aversion-resistant than XY/Sry + mice, this result should be interpreted with caution. Because quinine sensitivity was preserved in Sry– mice, the finding of increased aversion-resistance in mice with the XY chromosome complement (vs. XX/Sry– mice), can be made with more confidence. These results agree with prior findings demonstrating that XY mice developed habitual responding for EtOH while XX mice remained sensitive to outcome devaluation (19). The results seen here in FCG mice also align with previous studies of quinine-resistant binge-like EtOH consumption in male and female C57BL/6 J mice (26, 27). These studies have shown that females are aversion-resistant at the 100 μM concentration but not the 250 μM or 500 μM quinine concentrations, similar to XX/Sry– mice in the current study. In contrast, XY/Sry– mice were aversion-resistant at all quinine concentrations.

Although consumption remained aversion-resistant in XY/Sry– mice, preference for quinine + EtOH decreased in both Sry– groups. These results with preference may be due to water consumption as average intake of water increased from 8.240 ± 2.221 ml/kg (100 μM) to 17.633 ± 3.607 ml/kg (500 μM) in the XY mice. Therefore, it seems that mice with the XY chromosome complement consumed more fluids only when quinine was added to the EtOH solution.

Although the estrous cycle has not been found to greatly influence most EtOH drinking behaviors (4, 54), we thought it was important to determine whether this was true in the FCG mice. In FCG mice, others have found that XX and XY mice with ovaries spent the same number of days in each cycle phase (55), which suggests that ovarian function is similar between these two groups. Here we found no differences in EtOH consumption, preference for quinine + EtOH, or water consumption at any quinine concentration across the different phases of the estrous cycle. In addition, the mice spent 1 day in each phase of the cycle. These results suggest that the day-to-day fluctuation of hormones does not influence aversion-resistant drinking in FCG mice with ovaries. This finding is consistent with the literature where others have not seen an influence of the estrous cycle on binge-like drinking in a DID paradigm (10) or quinine-resistant drinking in a continuous access paradigm (23). Although we did monitor the estrous cycle during aversion-resistant sessions, we did not monitor it during any of the initial DID drinking sessions. Therefore, it is unknown if there are any cycling differences between these two groups following EtOH exposure alone.

To complement the DID paradigm, FCG mice were also trained to respond for EtOH in an operant chamber. When testing mice with escalating concentrations of EtOH, there were no effect of Sry or sex chromosome complement, although the main effect of Sry approached the threshold of significance (p = 0.055). Additionally, XX/Sry–, XY/Sry–, and XY/Sry + mice reduced responding when the concentration of EtOH delivered was increased. Only XX/Sry + mice failed to alter their rate of nose poking in response to a change in concentration. The lack of concentration-dependent responding in XX/Sry + mice may reflect differences in their ability to flexibly update their behavior or differences in EtOH reward. Although it is not clear at this time why XX/Sry + mice differed from the other genotypes on this task, this particular result highlights the fact that sex chromosome effects can be dependent on gonad type.

In the second operant experiment, FCG mice did not reduce responding when quinine was added to the EtOH solution and no differences were seen between any of the genotypes. Although this result was unexpected considering that we have previously reported that concentrations of quinine above 250 μM reduce responding for EtOH in male C57BL/6 J mice (24), it can be explained by the quinine insensitivity observed in the Sry + mice. The Sry– mice performed similarly to what we have previously observed in C57BL/6 J mice (24) and the lack of a difference between XX and XY animals of this gonad type suggests that chromosomal sex is not a major determinant of aversion-resistant responding under these experimental conditions. While this result differs from what was observed in the DID paradigm, where the XY complement increased aversion-resistance in Sry– mice, it is important to consider that female mice demonstrate quinine-resistance at higher quinine concentrations in the operant task than in the limited access DID model. When testing C57BL/6 J females, we have demonstrated that they decrease consumption at 250 and 500 μM quinine concentrations when drinking in the home cage (27) but continue responding for EtOH adulterated with these same concentrations in the operant chamber (24). Thus, the lack of an effect of sex chromosomes in Sry– mice in this experiment is likely due to a ceiling effect.

Because the results with quinine were inconclusive in Sry + mice, and to determine whether the Sry + mice are resistant to other punishments, we tested a subset of mice with a 0.35 mA footshock. We have previously found that this amplitude of footshock reduces responding for EtOH in male but not female C57BL/6 J mice (56) as well in both sexes in the selectively bred crossed High Alcohol Preferring line of mice (57). As all FCG mice reduced their responding on the footshock session, we conclude that any insensitivity to punishment in this line may be specific to quinine, although this portion of the study should be considered exploratory considering that only a subset of mice were tested. Future studies could use footshock to more fully explore whether aversion-resistance is regulated by sex chromosomes in the operant self-administration task.

In addition to responding for EtOH in the operant task, we also assessed head entries into the reward receptacle, which could be considered a measure of seeking behavior. While the number of head entries was similar between the two operant experiments, the findings were unexpectedly different. In the concentration experiment, Sry + mice made more head entries than Sry– mice (p = 0.0142). This effect of gonadal sex was not observed in the quinine experiment and instead XX mice made more head entries than XY mice (p = 0.049). Similar to the nose poke data, head entries did not decrease with quinine concentration. Persistent head entries, along with our observation that the drinking cups were empty at the end of the experiment, demonstrate that mice continued to consume quinine-adulterated EtOH at even the highest concentrations. Head entry data also followed nose poke responses in the subset of mice administered footshock, decreasing on the footshock session when compared to baseline. Thus, head entry data further support the conclusion that FCG mice are insensitive to quinine, but not footshock, punishment.

For the operant experiments, it is important to note that many mice, particularly those with the XY chromosome complement, failed to reach the training criterion and were dropped from the study. Of the 18 mice that did not reach criterion in the quinine-resistance experiment, 50% were XY/Sry– and 28% were XY/Sry+. This observation may point to differences in the genetic background of the FCG mice vs. lines that readily learn the operant response task (i.e., C57BL/6 J and crossed High Alcohol Preferring mice; see 23, 39, 40) or changes in the Y chromosome resulting from deletion of the Sry gene in XY mice specifically. Because of the high dropout rate, it is possible that this experiment is underpowered and indeed, there were some interesting trends in the data that did not reach significance (e.g., higher responding in XX/Sry + vs. XY/Sry + mice). However, between the difficulty in training XY mice and the demonstrated quinine insensitivity of the Sry + animals, it would be difficult to draw additional meaningful conclusions from this experiment even if more subjects were tested.