Impaired timing precision produced by striatal D2 receptor overexpression is mediated by cognitive and motivational deficits

Abstract

Increased striatal dopamine D2 receptor activity is thought to contribute to the pathophysiology of schizophrenia. To model this condition in mice, Kellendonk et al. (2006) generated transgenic mice that selectively overexpress the D2 receptor in striatum (D2OE). Drew et al. (2007) reported that D2OE mice display deficits in interval timing and motivation. The present study further explored the impaired timing in D2OE mice. Experiment 1 assessed the role of motivation in producing timing deficits in the peak procedure and found that performance in D2OE mice was improved by increasing motivation. In addition, performance was impaired in control mice when motivation was decreased. In Experiment 2, we found that D2OE mice have no timing impairment when tested using the bisection task, a procedure in which the measure of timing performance is less influenced by motivation to respond. In Experiment 3, we also used the bisection task and found selective impairment in timing of long durations in D2OE mice. These results suggest that striatal D2 overexpression impairs timing by decreasing motivation and through its impact on working memory and/or sustained attention.

Figure 1

Mean responses per second as a function of time during peak trials during all reinforcer percentage conditions. The top panel shows data from control mice and the bottom panel shows data from D2OE mice. The dotted vertical line in each panel indicates the time at which reinforcement was available during fixed interval trials (12 s).

Figure 2

Mean best fitting parameters derived from the fitting of the Gaussian function (see text for details) to the response rate data from the control (open bars) and D2OE (filled bars) mice. The top left panels show estimates of peak height (peak response rate), the top right panels show estimates of peak spread (timing variability), and the bottom left panels show estimates of peak location (timing accuracy). The bottom right panels show the proportion of variance accounted for by the Gaussian function. Error bars indicate one standard error above the mean.

Figure 3

The top panel shows the mean proportion of responses to the lever corresponding to a “long” sample duration on both short and long sample trials across blocks of training sessions in Experiment 2 (2 and 8 s sample durations). Circles and triangles indicate data from control and D2OE mice, respectively. Open data points indicate data from long sample trials, while closed data points indicate data from short sample trials. The bottom panel shows the mean proportion of responses to the lever corresponding to the “long” sample duration as a function of sample duration during Experiment 2 for control (filled circles) and D2OE (open circles) mice.

Figure 4

Mean best fitting parameters derived from the fitting of the cumulative normal function to the proportion long data from control and D2OE mice in Experiment 2. The top panel shows estimates of timing variability, while the bottom panel shows estimates of timing accuracy. Error bars indicate one standard error above the mean.

Figure 5

The top panel shows the mean latency to make a choice response (collapsed across sample duration) for control and D2OE mice in Experiment 2. The bottom panel shows the mean choice response latency as a function of sample duration for both control and D2OE mice. Other details as in Figure 4.

Figure 6

The top panel shows the mean proportion of responses to the lever corresponding to a “long” sample duration on both short and long sample trials across blocks of training sessions in Experiment 3 (6 and 24 s sample durations). The bottom panel shows the mean proportion of responses to the lever corresponding to the “long” sample duration as a function of sample duration during Experiment 3 for control and D2OE mice. Other details as in Figure 3. Mean proportion of responses to the lever corresponding to a “long” sample duration on both short and long sample trials across blocks of training sessions in Experiment 3 (6 and 24 s sample durations). Other details as in Figure 3.

Figure 7

Mean best fitting parameters derived from the fitting of the cumulative normal function to the proportion long data from control and D2OE mice from Experiment 3. Other details as in Figure 5.

Figure 8

Mean latency to make a choice response for control and D2OE mice in Experiment 3. Other details as in Figure 5.