0131 Theta oscillations during REM sleep synchronize behavior and neural activity in the developing motor system

Abstract

Abstract Introduction Myoclonic twitches are abundantly produced during REM sleep in skeletal muscles across the body. In infant rats, movements are produced by the red nucleus (RN), with the RN both sending motor commands and receiving sensory feedback from twitches. The RN’s role in producing twitches contrasts with that of primary motor cortex (M1), which does not generate motor commands at early postnatal ages. Instead, M1 functions as a sensory structure, processing sensory feedback from self-generated movements, including twitches. By postnatal day (P) 12, the RN (but not M1) also begins to exhibit a continuous theta rhythm (~6 Hz) during REM sleep that promotes sensorimotor integration with other brain areas. Given that the RN and M1 collaborate to control movement in adult rats, we hypothesized that theta emerges in M1 after P12, at which time theta synchronizes M1 and RN activity. Methods To determine if and when theta synchronizes activity in the RN and M1, we recorded local field potentials and unit activity in the RN and the forelimb region of M1 in unanesthetized preweanling rats at P12 and P20. Rats were head-fixed but were able to locomote and cycle freely between sleep and wake. Results Neurons in the RN and M1 continued to respond to twitches through P20. Further, as predicted, we observed the developmental emergence of REM-associated theta oscillations in M1 by P20 that were coherent with theta in the RN. Additionally, neural activity was phase-locked to theta; surprisingly, twitches were also phase-locked to theta, with twitches being more likely during the troughs of the oscillation. Finally, the temporal relationship between twitch-related activity in the two structures depended on the phase of theta, with twitch-related activity in M1 lagging behind twitch-related activity in the RN in the rising phase of theta. However, in the falling phase of theta, twitch-related activity in the RN and M1 showed similar time courses. Conclusion These results show how theta during REM sleep promotes the developmental integration of behavior with neural activity in the RN and M1. Because synchronous activity strengthens synaptic connectivity, and theta synchronized twitch-related activity in the RN and M1, these results also implicate twitches and twitch-related activity in the development of somatotopically precise functional connectivity between the RN and M1. Support (If Any) R37-HD081168 to M.S.B. and SRSF Career Development Award to J.C.D.