Behavioral correlates of activity of optogenetically identified locus coeruleus noradrenergic neurons in rats performing T-maze tasks

Abstract

The nucleusLocus Coeruleus (LC) is the major source of forebrain norepinephrine. LC is implicated in arousal, response to novelty, and cognitive functions, including decision-making and behavioral flexibility. One hypothesis is that LC activation promotes rapid shifts in cortical attentional networks following changes in environmental contingencies. Recent recordings further suggest LC is critical for mobilizing resources to deal with challenging situations. In the present study optogenetically identified LC neuronal activity was recorded in rats in a self-paced T-maze. Rats were trained on visual discrimination; then place-reward contingencies were instated. In the session where the animal shifted tasks the first time, the LC firing rate after visual cue onset increased significantly, even as the animal adhered to the previous rule. Firing rate also increased prior to crossing photodetectors that controlled stimulus onset and offset, and this was positively correlated with accelerations, consistent with a role in mobilizing effort. The results contribute to the growing evidence that the noradrenergic LC is essential for behavioral adaptation by promoting cognitive flexibility and mobilizing effort in face of changing environmental contingencies.

Figure 1

The behavioral task. (a) Schematic of the automated T-maze with return arms. When the rat crosses the visual cue (VC) onset photodetector (PD), this triggers one of the two screens behind the reward arms to be lit in pseudo-random sequence. Crossing the appropriate reward arm photodetector triggers the delivery of a drop of liquid reward at the corresponding reward site. Crossing the return arm (‘VC OFF’) photodetector triggers the lit screen to be turned off. (b) The two task rules. Under the ‘VC rule’, rats were only rewarded by choosing the reward arm in front of the lit screen regardless of whether it was to the left or the right. Under the ‘Turn rule’, rats were only rewarded on the same (non-preferred) side, regardless of which screen was lit.

Figure 2

Characterization of unit recordings. (a) Optical responses to 2 Hz, 100 ms, 10 mW pulse train laser stimulation in a unit in the virus injected rat (R328). (b) Immuno-histochemical preparations showing electrode track and neurons transduced with CAV2-PRS-ChR2-mCherry. (c) Distribution of average firing rates of units recorded. Optogenetically identified noradrenergic units (rat R328) are shaded black. (d) Distribution of spike widths of units recorded. Optogenetically identified noradrenergic units are shaded black. (e) Left) Overdrawn waveforms of a single unit recorded in LC in a behaving rat. (f) Nissl stained section showing the placement of the LC electrode in one of the recorded rats. Red arrow indicates the electrode tip position. TH, Tyrosine Hydroxylase; 4V, fourth ventricle; Me5, mesencephalic trigeminal nucleus.

Figure 3

LC response to VC onset before and after first response contingency changes. (a) Example rasters of LC unit activity as a function of VC onset before (bottom, trials 1–24) and after (top, trials 25–51) the first contingencency changes from VC to the Turn rule. Corresponding PETHs are shown to the right of each raster display. (b) Mean and SEM of firing rate of the shaded area in (a) of LC units in three individual rats before and after the first rule shift. *p < 0.005 Wilcoxon test.

Figure 4

(a) LC activity around task events. Each row shows activity from a different recording. The third row is from an optogenetically identified noradrenergic neuron (rat R328). Left column, increased firing in the 500 ms before crossing the photodetector on the central arm, triggering VC onset. Middle column, activity increases prior to return arm (Rtn) PD crossing, triggering VC OFF. Right column, baseline activity prior to reward arm photodetector crossing (Rwd PD), followed by reduced activity at the reward sites. Black circles to the left indicate previous VC onset. Dashed vertical lines at 0 indicates the PD crossing for the respective task events for these trials. (b) Response ratios of the task events. A response ratio of 1 means no change, and greater (or less) than 1 indicates an increase (or decrease) in firing relative to baseline. Error bars show SEM. *Significant differences from 1 (see text). (c) Significant correlation between Rtn and VC onset response ratios for each recording (n = 30, r₍₂₉₎ = 0.4, p = 0.03); VC onset responses tended to be greater than Rtn responses for individual neurons.

Figure 5

Relation between LC firing rate and acceleration around maze photodetector events. (a) In this typical session, speed increased in the 0.5 s period after photodetector crossings. White horizontal lines indicate when a rule shift was imposed and black horizontal lines indicate when the rat changed its behavioral strategy. (b) Acceleration increased in the 1 s prior to the respective photodetector crossings in the same session. (c) Corresponding raster and PETH plots of LC firing around the three task events. (d) LC firing rate correlation with acceleration around task photodetector events. The p values are 4.83E-07 for VC onset, 0.0268 for the Rtn photodetector, and 0.0324 for the Rwd photodetector.

Figure 6

Cross-correlation acceleration onset relative to onset of LC firing increases during a representative session. The arrowheads indicate the mean lags of acceleration after firing onset.