Compound Short (Sensory Control Task)

Compound Short

Compound Short (Sensory Control Task)

-15 0 15 30 45 60 75 -15 0 15 30 45 60 75 Time (s) Time (s)

FIGURE 14.5 Another example of a type 1 neuron from the lateral agranular cortex. For this rat, the short fixed interval was 10 sec and the long fixed interval was 40 sec. This cell was recorded during the simultaneous temporal processing procedure (A to C) and during a sensory control task (D). Although the cell responded on compound trials during the STP task, the cell did not respond during the same stimulus presentation (compound short stimulus) in the sensory control task. The overall lower firing rate in the sensory control task (D) than in the STP task (C) suggests that motor responses contribute partly to the overall activity of these cells. The time on the abscissa represents time from stimulus onset. The two horizontal lines in each graph represent the 95% confidence limits of the baseline firing rate. Bin width = 1 sec.

alternatives are sensory and motor interpretations. A sensory explanation proposes that type 1 cells are sensory neurons that fire to the combination of visual and auditory stimulation (the two modalities of stimuli used in the study). Two results make the sensory interpretation unlikely. First, as mentioned above, neuronal recordings were obtained from probe trials. If type 1 cells behaved as conjunctive sensory neurons, we would expect these cells to respond to the simultaneous presentation of both stimuli, which on probe trials lasted the duration of the trial. Instead, type 1 neurons had responses that had transient rises or falls that returned to baseline prior to the termination of the trial. Second, on the day following recording in the STP procedure, cells were recorded in a sensory control task (Figure 14.5D). During this task, response levers were unavailable to the rats and the sensory stimuli were presented as in the STP procedure. During the sensory control task, type 1 neurons did not respond to the simultaneous presentation of both stimuli on "compound" trials, as they did in the STP procedure. Thus, type 1 neurons do not appear to have purely sensory correlates for the conjunction of short and long stimuli.

A motor interpretation may be more likely. The lateral agranular cortex is thought to be the rat analog of the primary motor cortex. This conclusion is based on microstimulation studies and anatomical connections with thalamic motor areas

(Donoghue et al., 1979; Donoghue and Wise, 1982; Krettek and Price, 1977; Neafsey et al., 1986; Wang and Kurata, 1998). It is therefore possible that type 1 neurons might be responding to some motor aspect of the task. Although the patterns of neuronal activity and of lever responses were similar on compound trials, similar response patterns were not observed for simple trials. The rate of lever pressing increased during simple trials, with a peak near the fixed interval associated with each of the stimuli (Figure 14.3). In contrast, neuronal activity of type 1 cells was fairly constant throughout the simple trials (Figure 14.4A and B). In our study, motor activity may be reflected in the basal firing rate of the neuron. This can be seen as a lower overall firing rate in the sensory control task (Figure 14.5D) than in the STP task (Figure 14.5C). As mentioned earlier, response levers were inaccessible to the rats during the sensory control task, although all other aspects of the STP procedure were present. Future studies will have to explore the motor interpretation more carefully, possibly by using nontemporal tasks that require lever responses.

One interpretation that was not addressed in our study and will need to be examined is that the firing of type 1 cells reflects a general increase in cognitive demand, rather than a specific role in divided attention. Compound trials require more cognitive effort than simple trials, possibly leading to the differential firing pattern that we observed. One way that divided attention and cognitive demand might be separated is to increase the difficulty in performing a single-task procedure, such as making stimuli less distinguishable in a discrimination task or increasing the retention interval in a task with a memory component. Although it may be difficult to distinguish between cognitive demand and divided attention, this distinction is an important one for models that attempt to explain the interaction of timing and attention.

As mentioned earlier, proposed timing functions of the frontal cortex include involvement in the representation of duration and its storage in memory (Meck, 1996; Niki and Watanabe, 1979; Olton et al., 1988). Consistent with this view, type 2 cells responded on all types of trials in the STP procedure, supporting their general involvement in timing processes. Type 2 cells, however, constituted only 10% of the cells that responded in the task, suggesting that only a small proportion of the cells in this brain region may serve a direct clock or memory function. The most commonly observed cells were type 1 cells, which responded only on compound trials. A somewhat provocative interpretation of this finding is that type 1 cells are divided attention neurons with the primary function of coordinating the processing of multiple sources of information.

We would argue, however, that it is more natural to propose that a cognitive function such as divided attention resides in secondary motor areas than the primary motor cortex. This view is consistent with electrophysiological and imaging studies of normal participants and with studies in brain-damaged individuals, which support the involvement of the premotor and supplementary motor cortex in divided attention and timing (Casini and Ivry, 1999; Corbetta et al., 1991; Godefroy et al., 1996; Godefrey and Rousseaux, 1996; Pardo et al., 1990; Vanderberghe et al., 1997). One possibility is that the divided attention neuronal correlates we observed in the primary motor cortex are merely reflecting a response that is originating from a region that projects to the primary motor cortex. To resolve this issue, future studies are needed that record from the premotor and supplementary motor cortex to determine if, indeed, response patterns similar to those observed in the primary motor cortex are found.

Finally, it is interesting that many of the critical components of short-interval timing have been proposed to lie within the motor system (Ivry, 1993, 1996; Meck, 1996). Key motor areas that are directly involved in short-interval timing are the basal ganglia, cerebellum, and frontal motor cortex. Frontal motor areas may also serve a coordinating role because of their direct reciprocal connections with both the basal ganglia and cerebellum (Diedrichsen et al., this volume). One possibility is that simple oscillatory neural networks between the frontal cortex, basal ganglia, and cerebellum are responsible for the perception and production of durations under a second that comprise directly perceivable rhythms. These oscillatory networks may form the basis of a system that is capable of encoding and remembering longer durations. A recent approach along these lines is the striatal beat frequency model (Matell and Meck, 2000; Matell et al., this volume; Miall, 1989).

0 0

Post a comment