Abstract

Oscillatory neural activity in the cortico-basal ganglia-thalamocortical (CBGTC) loop is associated with the motor state of a subject, but also with the availability of modulatory neurotransmitters. For example, increased low-frequency oscillations in Parkinson’s disease (PD) are related to decreased levels of dopamine and have been proposed as biomarkers to adapt and optimize therapeutic interventions, such as deep brain stimulation. Using neural oscillations as biomarkers require differentiating between changes in oscillatory patterns associated with parkinsonism vs. those related to a subject’s motor state. To address this point, we studied the correlation between neural oscillatory activity in the motor cortex and striatum and varying degrees of motor activity under normal and parkinsonian conditions. Using rats with bilateral or unilateral 6-hydroxydopamine lesions as PD models, we correlated the motion index (MI)—a measure based on the physical acceleration of the head of rats—to the local field potential (LFP) oscillatory power in the 1–80 Hz range. In motor cortices and striata, we observed a robust correlation between the motion index and the oscillatory power in two main broad frequency ranges: a low-frequency range [5.0–26.5 Hz] was negatively correlated to motor activity, whereas a high-frequency range [35.0–79.9 Hz] was positively correlated. We observed these correlations in both normal and parkinsonian conditions. In addition to these general changes in broad-band power, we observed a more restricted narrow-band oscillation [25–40 Hz] in dopamine-denervated hemispheres. This oscillation, which seems to be selective to the parkinsonian state, showed a linear frequency dependence on the concurrent motor activity level. We conclude that, independently of the parkinsonian condition, changes in broad-band oscillatory activities of cortico-basal ganglia networks (including changes in the relative power of low- and high-frequency bands) are closely correlated to ongoing motions, most likely reflecting he operations of these neural circuits to control motor activity. Hence, biomarkers based on neural oscillations should focus on specific features, such as narrow frequency bands, to allow differentiation between parkinsonian states and physiological movement-dependent circuit modulation.