Our over-arching goal is to understand how subordinate habitual motor sequences are organized by higher-level task sequences through research that integrates human and animal experimental models. These complementary levels of analysis allow us to construct an understanding of sequential control spanning cells, networks, and behavior.
Recent work in humans using fMRI and TMS has shown that the rostrolateral prefrontal cortex (RLPFC) shows distinct activity across a task sequence, and is necessary for performance of repeated sequences of simple feature judgment tasks (e.g. Color, Shape, Shape, Color). We found that RLPFC activity increases (“ramps”) from the first position in the sequence to the last, then resets (Desrochers et al., 2015b, Neuron). Neural activity with similar dynamics has been recorded electrophysiologically from rodents and monkeys. There is active debate both as to the neural activity underlying this imaging signal (e.g., single cells vs. population), and process (e.g., evidence accumulation vs. progress towards a goal). We have hypothesized that ramping activation in RLPFC aids in tracking and resolving positional uncertainty in the sequence, but the neural basis of this novel sequential control dynamic determined with fMRI remains unknown.
- Are monitoring, evidence accumulation, and uncertainty resolution the computational functions associated with ramping in RLPFC during human sequential task control? Using three separate, complementary fMRI experiments with human participants we will investigate the nature of the ramping signal in the frontal cortex previously found to be necessary for sequential task performance (Desrochers et al., 2015b, Neuron).
- What are the specific frontal cortical areas responsible for sequential task control in the animal? Animals will perform the same simple sequence monitoring task performed by humans. Through examining the time course of task-relevant activations, we will determine how the network of areas active in the animal relates to the network of areas active in the human.
- What are the specific neural mechanisms underlying sequential control in frontal cortex? Answering this central question requires linking cellular and meso-level insights from animal models to my findings in the human. To achieve this goal, we will use animal fMRI to bridge the two systems and allow insights to move beyond behavioral analogy and into functional homology. Parallel animal/human fMRI datasets will enable future recordings in specific, functionally homologous regions in animals and the direct comparison between animals and humans.
- Does the novel representation of cost we observed in the striatum (Desrochers et al., 2015a, Neuron) during habitual motor sequences generalize to more abstract sequential control, i.e. task sequences? Our working hypothesis is yes, and that these more abstract representations will be more anterior in striatum, paralleling caudal-to-rostral gradients in frontal cortex and its connections with striatum (Desrochers & Badre, 2012, TICS). This work will employ more abstract tasks in animals where invasive neural recordings and manipulation are possible, and the study of parallel tasks in humans.
- Despite the commonality of OCD, which affects ~2.3% (3.3 million) of the US population, we know relatively little about the specific brain circuits responsible. A specific outstanding question is: Do patients with OCD have deficits in sequential control? This intuitive prediction has, surprisingly, not received detailed investigation. Do patients perform sequences of tasks, in a controlled lab setting, the same way as healthy controls? What brain areas are recruited? Is there a deficit in striatal end signaling or the cost/reward signals we previously delineated? We will apply our sequential control tasks in OCD populations to investigate these questions. Such studies could have clinical importance and lead to novel understandings of this disorder.