Figure 4.

dCA1 Inputs Drive NAc MSNs and of a Subset of FSIs

(A) dCA1 PYRs were transduced with either the ArchT-GFP or the control GFP-only construct. Dual-site tetrode implant allowed for simultaneous monitoring of dCA1 and NAc ensembles (n = 671 PYRs and 271 NAc neurons, including 169 MSNs and 65 FSIs from five behaving CamKII^dCA1::ArchT-GFP mice; n = 181 NAc neurons, including 107 MSNs and 35 FSIs from two behaving CamKII^dCA1::GFP control mice). Optic fibers inserted above NAc allowed delivering yellow light to dCA1 axons in NAc (ON^dCA1→NAc) in both CamKII^dCA1::ArchT-GFP (“ArchT-GFP”) and CamKII^dCA1::GFP (“control-GFP”) mice.

(B) Examples of a recorded PYR, FSI, and MSN. From left to right shown for each neuron the average spike waveform, example spikes recorded on each channel of corresponding tetrode, spike train auto-correlogram, and inter-spike interval distribution. See also Figure S5.

(C) Firing rate of FSIs and MSNs (mean ± SEM; 50-ms bins) relative to light ON^dCA1→NAc delivery (one 10-min pulse) and recorded from the NAc of CamKII^dCA1::GFP control mice (in gray) or CamKII^dCA1::ArchT-GFP mice (with MSNs in black; FSIs in red). See Figure S5I for the distinct firing responses of light ON^dCA1→NAc responding and non-responding FSIs separately.

(D and E) Corresponding firing rate changes (scores) of FSIs and MSNs (D: mean ± SEM; E: distributions of rate scores using a kernel density approach) (Silverman, 1981). For each cell, light-driven score corresponds to the firing rate in light ON minus that in light OFF divided by the sum; a firing rate score was also computed between the two OFF epochs either side of the ON^dCA1→NAc (as shown in C) for comparison. Note that light ON^dCA1→NAc delivery to dCA1 axons in NAc of CamKII^dCA1::ArchT-GFP mice caused bidirectional rate changes within the MSN population (with some MSNs showing increased firing rate, while others decreased firing). In the same time, light delivery to dCA1 ArchT-GFP axons drastically decreased the rate of a subset of FSIs (i.e., FSIs with a score toward –1). The activity of the other FSIs remained unaffected (i.e., other scores near 0). All FSIs recorded from CamKII^dCA1::ArchT-GFP mice pooled together in Figure 4C (see also Figure S5I). The same light delivery did not cause rate changes of FSIs (E; top gray curve) and MSNs (E, bottom gray curve) in CamKII^dCA1::GFP control mice (i.e., all scores near 0). ^∗∗p < 0.01 unpaired t test FSIs versus corresponding MSNs just below, ^### p < 0.001 compared to zero.

(F and G) Temporal relationship between spike discharge of simultaneously recorded dCA1 and NAc neurons. Time zero equals spike times of PYRs, used as reference. Black histograms: population cross-correlograms computed from the original spike trains of PYR-FSI (F) and PYR-MSN (G) cell pairs. Note the theta-paced spiking of both PYR-FSI and PYR-MSN pairs, with phase opposition. Gray histograms: control cross-correlograms computed from PYR spikes shifted to random theta cycles with preserved (original) theta phase. Note the sharp, short-latency increase of FSIs spike probability following the original, but not the shifted, PYRs discharge (black asterisk; PYR-to-FSI offset: 3.2–8 ms), suggestive of a direct entrainment of some NAc FSIs by dCA1 PYRs.

(H) dCA1 PYRs of CamKII-Cre mice were transduced with ChR2-eYFP. Tetrodes were implanted for NAc ensemble recordings (n = 222 NAc neurons including 128 MSNs and a total of 64 FSIs from two behaving CamKII^dCA1::ChR2-eYFP mice). Optic fibers inserted above NAc allowed delivering blue light to photo-stimulate dCA1 axons (ON^dCA1→NAc, 5-ms pulses).

(I–K) Spike discharge of NAc FSIs and MSNs in response to in vivo photo-stimulation of dCA1 ChR2-eYFP axons in NAc of CamKII^dCA1::ChR2-eYFP mice (H). Top row: examples of single-neuron raster plots for a dCA1-responding FSI (I), a non-responding FSI (J) and a responding MSN (K). Note the short-latency light-driven spiking of the responding FSI compared to the absence of firing response of the other FSI (both FSIs simultaneously recorded from the same mouse, same tetrode). Note also the longer-latency spiking of the MSN. Bottom row: corresponding spike discharge probabilities (mean ± SEM; 0.2-ms bins) for the populations of dCA1-responding FSIs (n = 12), non-responding FSIs (n = 52), and MSNs (n = 128), relative to the light ON^dCA1→NAc onset.