Figure 3. Optical inhibition of BLA neurons during stimulus-outcome pairing attenuates the encoding of stimulus-outcome memories.
(a). Procedure schematic. CS, conditional stimulus (white noise or tone); O, outcome (sucrose solution or food pellet); A, action (left or right lever press). (b) Schematic of optogenetic strategy for bilateral inhibition of BLA neurons. (c) Representative fluorescent image of ArchT-eYFP expression and fiber placement in the BLA. (d) Schematic representation of ArchT-eYFP expression and placement of optical fiber tips in BLA for all subjects. (e) Elevation [(CS probe entry rate)/(CS probe entry rate + preCS entry rate)] in food-port entries during the CS probe period (after CS onset, before first reward delivery), averaged across trials and CSs for each day of Pavlovian conditioning. Thin light lines represent individual subjects. (f) Elevation in lever presses on the lever that earned the same outcome as the presented CS (Same; [(presses on Same lever during CS)/(presses on Same lever during CS + Same presses during preCS)], averaged across trials and across CSs), relative to the elevation in responding on the alternate lever (Different; [(presses on Different lever during CS)/(presses on Different lever during CS + Different presses during preCS)], averaged across trials and across CSs) during the PIT test. Lines represent individual subjects. (g) Elevation in food-port entries to CS presentation (averaged across trials and CSs) during the PIT test. Circles represent individual subjects. ArchT, N = 9; eYFP, N = 10. *p<0.05. See Figure 3—source data 1.
Figure 3—source data 1. Source data for Figure 3 and Figure 3—figure supplement 1–3.
elife-68617-fig3-data1.xlsx (26.8KB, xlsx)
Figure 3—figure supplement 1. Green light activation of ArchT hyperpolarizes and attenuates the firing of BLA cells.
(a) Confocal image of biocytin-filled BLA cell (red) expressing ArchT-eYFP. (b) Current-clamp recording of an ArchT-expressing BLA cell responding to hyperpolarizing and depolarizing current injections. When illuminated with green light (535 nm, 100 ms pulse, 0.5 mW), activation of ArchT hyperpolarizes the cell membrane resulting in the absence of action potential firing at suprathreshold membrane potentials. This hyperpolarization of the cell membrane occurs only during green light luminescence. (c) Representative recordings from 2 ArchT-expressing BLA cells when injected with a suprathreshold pulse of current (165 or 375 pA 1 s; bottom) with green light off (top) or on (middle). (d) Summary of the number of action potentials recorded in ArchT-expressing BLA cells (N = 12 cells/5 subjects) injected with a suprathreshold amount of current before (Off) and during (On) green light illumination (median = 1 mW, range = 0.25–1). Current injection intensities that resulted in 8–15 action potentials were selected for recordings (median = 275 pA, range 100–800 pA, duration = 1 s). Number of action potentials was averaged across three sweeps/condition. Green light activation of ArchT in BLA cells reduced action potential firing in all cells and abolished (>97% reduction) it in most cells. The average number of action potentials recorded during green light exposure was significantly lower than the control no-light period (t11 = 9.25, p<0.0001). Lines represent individual cells. ***p<0.001.
Figure 3—figure supplement 2. Food-port entry and press rates during Pavlovian conditioning and PIT test for BLA optical inhibition experiment.
(a) Food-port entry rate (entries/min) during CS probe period (after CS onset, before first reward delivery), averaged across trials and across CSs for each day of Pavlovian conditioning. There was no effect of BLA inhibition during reward retrieval on the development of this Pavlovian conditional goal-approach response (CS x Training: F(3.4,57.8) = 16.44, p<0.0001; CS: F(1,17) = 46.73, p<0.0001; Virus: F(1,17) = 0.17, p=0.68; Training: F(2.3,38.5) = 2.37, p=0.10; Virus x Training: F(7,119) = 1.55, p=0.16; Virus x CS: F(1, 17)=0.0009, p=0.98; Virus x Training x CS: F(7,119) = 1.63, p=0.13). *p<0.05, **p<0.01 relative to pre-CS. (b) Lever press rate (presses/min) averaged across levers and across the final 2 days of instrumental conditioning. There was no significant difference in press rate between the control group and the group that received BLA inhibition during Pavlovian conditioning (t17 = 1.44, p=0.17). Circles represent individual subjects. (c). Lever press rate (presses/min) on the lever earning the same outcome as the presented CS (averaged across trials and CSs), relative to the press rate on the alternate lever (Different) during the PIT test. Planned comparisons (Levin et al., 1994), based on the significant interaction and post hoc effect detected in Figure 3f, showed that for the eYFP control group CS presentation significantly increased responding on the lever that earned the same reward as that predicted by the presented CS relative to the preCS baseline period (t9 = 3.11, p=0.01). The CSs did not significantly alter responses on the different lever in the control group (t9 = 1.35, p=0.21). For the ArchT group, the CSs were not capable of significantly altering lever pressing relative to the baseline period (Same: t8 = 2.13, p=0.07; Different: t8 = 0.77, p=0.46). Lines represent individual subjects. (d) Food-port entry rate during CS presentation (averaged across trials and CSs) during the PIT test. For both groups CS presentation triggered a similar elevation in this goal-approach behavior (CS: F(1,17) = 59.41, p<0.0001; Virus: F(1,17) = 0.63, p=0.44; Virus x CS: F(1,17) = 3.42, p=0.08). Lines represent individual subjects. *p<0.05, ***p<0.001.
Figure 3—figure supplement 3. Inhibition of BLA neurons unpaired with reward delivery does not disrupt the encoding of stimulus-outcome memories.
We found that inhibition of BLA neurons specifically at the time of outcome experience during each CS during Pavlovian conditioning attenuated subjects’ encoding of the sensory-specific stimulus-outcome memories, as evidenced by their inability to later use those memories to guide choice behavior during a PIT test. To control for the total amount of BLA inhibition during Pavlovian conditioning, we repeated the BLA inhibition experiment in a separate group of subjects matching the frequency and duration of inhibition to the experimental group (Figure 3), but delivering it during the baseline, 2 min pre-CS periods. We selected this period for control inhibition to maintain proximity to the CS period but avoid inhibition during the CS at periods in which the rat might be expecting, checking for, and/or retrieving reward, events that were not possible for us to time. (a) Procedure schematic. CS, conditional stimulus; O, outcome (sucrose solution or food pellet); A, action (left or right lever press). (b) Schematic of optogenetic strategy for inhibition of BLA neurons. (c) Representative fluorescent image of ArchT-eYFP expression and fiber placement in the BLA. (d) Schematic representation of ArchT-eYFP expression and placement of optical fiber tips in BLA for all subjects. (e) Elevation [(CS probe entry rate)/(CS probe entry rate + preCS entry rate)] in food-port entries during CS probe period (after CS onset, before first reward delivery), averaged across trials and CSs for each day of Pavlovian conditioning. Optical inhibition of BLA neurons unpaired with reward delivery did not affect development of the Pavlovian conditional goal-approach response (Training: F(3.4,20.6) = 16.83, p<0.0001). Thin light lines represent individual subjects. (f) Elevation in lever presses on the lever that earned the same outcome as the presented CS (Same; [(presses on Same lever during CS)/(presses on Same lever during CS + Same presses during preCS)], averaged across trials and across CSs), relative to the elevation in responding on the alternate lever (Different; [(presses on Different lever during CS)/(presses on Different lever during CS + Different presses during preCS)], averaged across trials and CSs) during the PIT test. Inhibition of BLA neurons unpaired with reward delivery during the Pavlovian conditioning sessions did not affect the subsequent ability of the CSs to bias instrumental choice behavior during the PIT test (t6 = 2.88, p=0.03). Lines represent individual subjects. (g) Elevation in food-port entries to CS presentation (averaged across trials and CSs) during the PIT test. The CSs were also capable of elevating food-port entries above baseline during the PIT test. Circles represent individual subjects. N = 7. *p<0.05, corrected post hoc comparison.