Engagement for alcohol escalates in the 5-choice serial reaction time task after intermittent access

Abstract

Excessive intake plays a significant role in the development of alcohol use disorder and impacts 15 million Americans annually, with approximately 88 000 dying from alcohol related deaths. Several facets we contribute to alcohol use disorder include impulsivity, motivation, and attention. Previous studies have used the 5-Choice Serial Reaction Time Task (5-Choice) to analyze these types of behaviors using sugar, but recently we have published using 10% alcohol as the reward. This study analyzed 48 mice that were trained to respond for alcohol in the 5-Choice. All mice distributed and analyzed first by alcohol preference and then by consumption. Here, we became interested in a new classification called "engagement". High-engaged and low-engaged mice were determined by the number of correct responses during final Late-Stage training sessions. Interestingly, during Early-Stage training, the mice began to separate themselves into two groups based on their interaction with the task. Throughout both training stages, high-engaged mice displayed a greater number of trials and correct responses, as well as a lower percentage of omissions compared to low-engaged mice. Following three weeks of intermittent access homecage drinking, low-engaged mice showed greater increase in perseverative responding relative to high-engaged. Additionally, low-engaged mice decreased their reward and correct latencies compared to high-engaged mice suggesting an increase in motivation for alcohol. Overall, engagement analysis presents two clearly different groups, with only one being motivated to work for alcohol. These two distinct phenotypes in the 5-Choice could be used to model alcohol motivated behavior, which could help us further understand alcohol use disorder.

Keywords: 5-Choice, 5-choice serial reaction time task; HE, high-engagement; LE, low-engagement.

Figure 1.. Experimental schematic.

Timeline of the behavioral and drinking schedule. Phase 1: water restricted mice were trained in the 5-choice 4-5 days per week. Between each training week, DID sessions occurred (green bars). Mice were switched to weight restriction for sweet milk testing. Phase 2: Mice stopped 5-Choice and DID drinking to undergo 3 weeks of intermittent access 2-bottle choice (IA2BC) with 20% alcohol (red). Intermittent access continued when mice returned to 5-Choice testing with the process described in Phase 2 Testing Day Example (A). Example cartoon of the 5-Choice waiting period (5 seconds) and a correct response that is used to determine engagement in the task (B).

Figure 2.. Animal classification based on engagement.

High-engaged (HE) and low-engaged (LE) had similar alcohol consumption (A) and preference (B) during homecage drinking sessions. Previously, these animals were classified as either high/low- preference or consumption for analysis. The amount of HE and LE mice previously considered high-preference or low-preference with further separation based on range of preference (C). The amount of HE and LE mice previously considered high-consumption or low-consumption with further separation based on range of consumption (D). The range of preference and consumption was divided into 3 groups: highest 1/3, middle 1/3, and lowest 1/3.

Figure 3.. High-engaged mice perform greater during all phases of training.

Early-Stage performance was much greater in HE mice with increased accuracy, correct responding, and decreased omissions. Additionally, HE mice correct responding correlated with homecage alcohol preference and consumption (A). HE continued to have greater behavioral performance during Late-Stage training with a strong correlative trend of HE correct responding to homecage alcohol preference (B). When the reward was switched to strawberry milk, both groups greatly increased performance with HE mice performing greater. LE mice, if tested longer, are hypothesized to match HE performance as described by the significant decrease in omission percentage (C). #p<0.05 for main effect. *p<0.05 post-hoc significance.

Figure 4.. HE mice show high motivation for an alcohol reward primarily during the first half of the session.

HE mice displayed significantly lower reward latency during Early (A) training, but both HE and LE were similar during Late-Stage (B) and sweet milk training (C). HE mice had significantly lower reward latency during the final five sessions of Late-Stage and sweet milk training (D). The number of reward latencies greater than 2.5s occurred more frequently during the last half of the session for HE mice for an alcohol reward (E) unlike HE for sweet milk (F), LE for alcohol (G), and LE for sweet milk (H). #p<0.05 for main effect. *p<0.05 post-hoc significance.

Figure 5.. Intermittent access increases perseverative responding for alcohol in low-engaged mice.

The percentage of perseverative responding prior to intermittent access was greater in HE mice and became similar between the groups after intermittent access (A). LE mice faced a significantly greater change in perseveration compared to HE mice (B). No changes were seen in sweet milk pre- or post-intermittent access (C) and the overall change remained similar (D). Perseverative percentage analyzed by high and low preference showed increases in perseveration after intermittent access in both groups (E) but no difference in overall change in perseveration between the groups (F). No changes were seen in sweet milk pre- or post-intermittent access (G) and the overall change remained similar (H). Perseverative percentage analyzed by high and low consumption showed increases in perseveration after intermittent access in both groups (I) but no difference in overall change in perseveration between the groups (J). No changes were seen in sweet milk pre- or post-intermittent access (K) and the overall change remained similar (L). #p<0.05 for main effect. *p<0.05 post-hoc significance.