アイテム詳細

要旨

Cognitive theories assume that the locus coeruleus (LC), a brain stem neuromodulatory nucleus, broadcasts a redundant signal to the entire forebrain due to synchronized activity of a homogeneous population of diffusely projecting norepinephrine (NE) neurons. Until recently, technical challenges limited recordings to 1-2 LC single units, which was insufficient for characterizing the diversity of LC cell types and their ensemble activity patterns. We recorded as many as 75 single units simultaneously in the urethane-anesthetized rat using a high-density multi-electrode array and analyzed 11893 unit pairs. To assess input-output specificity of LC units, we electrically stimulated 15 forebrain LC projection sites and analyzed evoked orthodromic and antidromic LC spiking. Using noise and cross correlation analyses, we assessed the heterogeneity of unit activity with respect to input-output circuits. Our results revealed 2 cell populations differing by spike width (Type 1: 461±14μs, Type 2: 1076±9μs) and rate (Type 1: 1.6±0.06Hz, Type 2: 0.84±0.04Hz). NE identity was confirmed for both types by a α2 agonist. Spontaneous noise correlations were weak and did not depend on linear distance, but Type 1 units exhibited higher correlations with one another (pair of Type 1’s: 0.138±0.007, Type 2’s: 0.049±0.001, mixed Type1/2 pair: 0.038±0.002 in 200ms bins). Evoked noise correlations were also weak (Type 1: 0.155±0.009, Type 2: 0.051±0.006, Type 1/2: 0.040±0.008, in 750ms window after five 5mA, 0.5ms, 30Hz foot shocks). Consistent with weak noise correlations, the majority (77) of pairs did not have any significant spike count change at any cross-correlelogram bin (-2 to +2s). Some pairs were correlated (from 0 to +50ms or +1.0 to 1.5ms) or anticorrelated (+1.0 to 2.0ms). We defined projection targets for 65 of 205 units. 53 of those projected to only 1 forebrain site and the remainder projected to multiple (up to 8) sites. Both cell types projected widely to the forebrain. Unit pairs projecting to thalamus had significantly higher noise correlations (0.198±0.015) than pairs with diverging projections (0.132±0.011). PFC input to the LC was area-specific: PL stimulation decreased and IL increased spiking (PL: 9 of units dec., 3 inc.; IL: 1 dec., 14 inc.). PFC stimulation also decorrelated LC activity (200ms before / after stimulation in PL: 0.063±0.005 / 0.050±0.005 and IL: 0.053±0.006 / 0.038±0.007). Thus, we found 2 LC cell types differing in spike width, rate, and noise correlation; however, overall correlations were weak. These findings challenge the prevailing view that LC cells are physiologically similar and are homogenously driven to provide a global signal.