Dopamine And Attention Sharing

In contrast to the above neuropharmacological mapping, some reports suggest that dopaminergic drugs affect attention to temporal signals without selectively altering the speed of an internal clock (e.g., Santi et al., 1995; Stanford and Santi, 1998). These results are congruent with a separate line of evidence that tends to suggest a different role for the dopaminergic synapse: an attention-getting device used to track significant events (see Gray et al., 1997; Lee et al., 1998; Schultz et al., 1997). These data suggest that dopamine involvement in the speed of an internal clock needs to be dissociated from the possible confounder with attentional effects.

To this end, one could examine the effect of dopaminergic drugs in a behavioral task in which both the timing and the attentional components are well understood. For example, behavioral data from dual-task performance in humans support the cognitive implementation of the time processor that acknowledges attentional processing (more specifically, attention sharing). Indeed, a behavioral study by Lejeune et al. (1999), using a paradigm in which pigeons were exposed to an analog of the dual-task procedure used to test attention sharing in humans, found data similar to those collected in humans, thus giving support to attention sharing between the time processor and the general processor (Block and Zakay, 1996; Lejeune, 1998; Zakay, 2000) in pigeons. Therefore, one could examine the effect of dopaminergic drugs on dual-task performance in animals. Examination of the effect of dopaminergic drugs in such procedures might allow the differentiation of the effect of the drug on clock speed and attention.

In our work, we pursued a different approach involving a single-task procedure previously used to study timing and memory for time in animals (Catania, 1970; Meck et al., 1984; Roberts, 1981; Roberts and Church, 1978). We examined a variation of the PI procedure in which subjects (rats or pigeons) have to filter out the gaps that (sometimes) interrupt timing (Catania, 1970; Roberts, 1981). In the gap procedure, the subjects are exposed to three types of trials: FI trials, PI trials, and gap trials. Gap trials are similar to PI trials, but the signal is interrupted for a brief duration called a gap. Typically, in gap trials the mean response rate increases in the pregap interval, declines during the gap, and then increases again after the gap and reaches a peak that is delayed relative to the peak time during PI trials. Evidence discussed below suggests that in this paradigm the delay in peak time is controlled by an attention-sharing process that can be differentiated from the pacemaker stage of the internal clock both behaviorally (Buhusi and Meck, 2000; Buhusi et al., 2002) and pharmacologically (Buhusi and Meck, 2002). This differentiation is critical, because both processes are at least partly dependent on dopaminergic function (Buhusi and Meck, 2002). Before we discuss the differentiation of pacemaker and attention-sharing components, it is important to discuss alternative interpretations of the data in the gap procedure.

12.3.1 The Switch Hypothesis

In rats, the peak time in gap trials was found to be delayed relative to PI trials for approximately the duration of the gap (Meck et al., 1984; Roberts, 1981; Roberts and Church, 1978). These data were taken to suggest that rats "stop" their timing process during the gap and resume it where they left off after the gap. To address these data, Gibbon et al. (1984) proposed an on-off switch mechanism controlled by the presence of the to-be-timed stimulus. In PI trials, the switch is closed, pulses from the pacemaker reach the accumulator, and the response rate reaches a peak near the time of reinforcement. During the gap, the switch is open, so that pulses from the pacemaker fail to reach the accumulator. The pulses accumulated during the pregap interval are not lost, however, due to their proposed maintenance in working memory. Therefore, in accord with experimental data in rats (Meck et al., 1984; Roberts, 1981; Roberts and Church, 1978), the peak time in gap trials is delayed by the duration of the gap (stop rule).

12.3.2 The Decay Hypothesis

On the other hand, in gap procedures that use gap durations similar to those used with rats, pigeons' mean response rate after the gap is delayed relative to the time of reinforcement with approximately the duration of the gap plus the duration of the fixed interval (e.g., Roberts et al., 1989). These data suggest that during the gap pigeons "reset" the interval timing process and restart timing after the gap from the beginning of the interval (reset rule). Importantly, a parametric study in pigeons suggests that the delay in peak time during gap trials increases with the duration of the gap (Cabeza de Vaca et al., 1994). Based on this observation, Cabeza de Vaca et al. (1994) proposed a rather different mechanism by which the gap might affect interval timing, namely, that the subjective time — stored in the accumulator — "decays" passively during the gap. This passive decay process is presumably minimal for short gaps, but rather large for longer gaps, thus accounting for both the stop and reset rules. In summary, previous studies in rats and pigeons interpret the results obtained in a gap procedure in terms of two mechanisms, a switch and a passive decay process, which are an integral part of the time processor and are exclusively controlled by the temporal parameters of the procedure.

12.3.3 Dopamine and the Attention-Sharing Pattern

We recently examined the involvement of dopamine in putative processes like the pacemaker, the switch, and the passive decay of accumulated time in the gap procedure (Buhusi and Meck, 2002). The predicted results of acute administration of the dopaminergic agonist MAP in the gap procedure, given the assumption that dopaminergic drugs affect solely the pacemaker level of the internal clock, are shown in Figure 12.2. The upper panels of Figure 12.2 show the predicted effect of the gap under vehicle: according to the switch hypothesis, the presentation of the gap should delay the peak time for approximately the duration of the gap (left upper panel). In contrast, according to the decay hypothesis, the presentation of the gap should delay the peak time more than the duration of the gap (right upper panel). The lower panels of Figure 12.2 show the predicted effect of the acute administration of MAP in PI and gap trials. Under the assumption that dopaminergic drugs affect solely the pacemaker level of the internal clock, the lower panel of Figure 12.2 shows that during PI trials, as well as before and after a gap in gap trials, the rate of accumulation is predicted to increase under MAP. Indeed, while the accumulation reaches the 30-sec criterion after 30 sec under vehicle (upper panels), accumulation is predicted to be faster under MAP and to reach the temporal criterion after, for example, 25 sec. Critically for the logic of the experiment, irrespective of the operational hypothesis (switch or decay) the peak time is predicted to be delayed in gap trials relative to PI trials with about the same duration under MAP (lower panels) than under vehicle (upper panels).

To evaluate the latter prediction, we estimated the shift in peak time, computed as follows:

Shift = PTg - PTPI - Gap where PTGap is the estimated peak time in gap trials, PTPI is the estimated peak time in PI trials, and Gap is the duration of the gap. A null shift would suggest that rats use a stop rule, while a shift equal to the pregap interval (15 sec in this experiment) would suggest that rats use a reset rule. The results of this experiment reported by Buhusi and Meck (2002) are shown in Figure 12.3. In PI trials (left panel), acute administration of MAP decreased the peak time, and acute administration of HAL increased the peak time, supporting the proposal that dopaminergic drugs affect the rate of the pacemaker of an internal clock. However, in contrast to the predictions shown in Figure 12.2, acute administration of both MAP and HAL was found to affect the shift in response peak time in gap trials relative to PI trials. Predictions shown in Figure 12.2 suggest that the difference in peak time between PI and gap trials should be about the same, irrespective of drug manipulation. In contrast, we found that MAP determined a leftward displacement in peak time in PI trials (left panel of Figure 12.3), but it also increased the shift in peak time in gap trials relative to PI trials, toward the reset (right panel of Figure 12.3). In similarity to the predictions shown in Figure 12.2, the slower accumulation of time under HAL

Vehicle

0 0

Post a comment