Figure 8. Lateral/ventrolateral periaqueductal gray-synapsin-expressing (l/vlPAG-syn) cells are more active near threat, while l/vlPAG-cholecystokinin-expressing (l/vlPAG-CCK) cells are more active far from threat.

(A) Top: viral schematic for synapsin-specific and CCK-specific GCaMP6 expression in l/vlPAG. Bottom: timeline for in vivo photometry recordings. (B) Fiber photometry recording set-up. (C) Histology of GCaMP6f expression in synapsin-specific (left) and CCK-specific (right) cells in the l/vlPAG. Scale bar, 200 μm. (D) Example heatmaps showing z-scored dF/F in mice expressing synapsin-specific GCaMP6 (left) or CCK-specific GCaMP6 (right) in l/vlPAG in an elevated plus maze assay. Vertical arms of heatmaps represent closed arms. (E, F) Mean dF/F (z-scored) 1 s after arm entry in syn-GCaMP6 (E) and CCK-GCaMP6 mice (F). Mean dF/F 1 s post-entry into the open arms is greater than into closed arms in syn-GCaMP6 mice (E, n = 9; paired t-test, *p=0.0213), whereas mean dF/F 1 s post-entry into the open arms is lower than into the closed arms for CCK-GCaMP6 mice (F, n = 11; paired t-test, *p=0.0283). (G) Example heatmaps showing z-scored dF/F in syn-GCaMP6 (left) or CCK-GCaMP6 (right) in live rat exposure assay. Rat was confined to the right of the map, as indicated by the red bar. (H, I) Mean dF/F (z-scored) during approaches toward the rat (top) or escapes from the rat (bottom) within 10 spatial bins of varying distance from the safe wall of syn-GCaMP6 (H, n = 9) or CCK-GCaMP6 (I, n = 13) mice (syn-approach, n = 6744 samples; syn-escape, n = 2150 samples; CCK-approach, n = 7170 samples; CCK-escape, n = 2088 samples). (H) In syn-GCaMP6f mice, dF/F is positively correlated with distance from the safe wall during escapes from the predator (Pearson’s correlation coefficient r = 0.932, p<0.0001). (I) In CCK-GCaMP6f mice, dF/F is negatively correlated with distance from safe wall during both approaches and escapes (approach, r = −0.792, p=0.006; escape, r = –0.703, p=0.023). (J, K) Mean dF/F (z-scored) 5 s before and after approaches, escapes, freeze bouts, and stretch-attend postures in syn-GCaMP6 (L) and CCK-GCaMP6 (M) populations (syn, n = 9; CCK, n = 13 for freeze, n = 12 for other behaviors). (L, M) Mean dF/F (z-scored) in the safer zone (one-third of assay near safer wall) and threat zone (two-thirds of assay distal from safer wall) in syn-GCaMP6 and CCK-GCaMP6 mice. Pan-neuronal l/vlPAG activity was increased in the threat zone compared to the safer zone (syn, n = 9; paired t-test, **p=0.0024), whereas CCK-specific activity was decreased in the threat zone compared to the safer zone (CCK, n = 12; paired t-test, *p=0.0223). Errorbars: mean ± SEM.

Figure 8—figure supplement 1. GCaMP6f expression and fiber placement in the lateral/ventrolateral periaqueductal gray (l/vlPAG) in coronal brain sections.

GCaMP6f expression is depicted in shading and fiber placement is represented by dots for (A) l/vlPAG-synapsin-expressing (l/vlPAG-syn) mice and (B) l/vlPAG-cholecystokinin-expressing (l/vlPAG-CCK) mice. All fibers were placed at a 15° angle. All viral injections and fiber placements were unilateral and counterbalanced across left and right PAG. Center of fiber was determined at widest point in histological slices.

Figure 8—figure supplement 2. No correlation between speed and df/F.

(A, B) Spearman’s correlation of speed and fiber photometry df/F of (A) lateral/ventrolateral periaqueductal gray-synapsin-expressing (l/vlPAG-syn) or (B) l/vlPAG-cholecystokinin-expressing (l/vlPAG-CCK) population during exposure to toy rat (syn, n = 5; CCK, n = 6; one-sample t-test). Errorbars: mean ± SEM.

Figure 8—figure supplement 3. Lateral/ventrolateral periaqueductal gray-synapsin-expressing (l/vlPAG-syn) and cholecystokinin (CCK) activity during exposure to a control toy rat.

(A) Example heatmaps showing z-scored dF/F in syn-GCaMP6 (left) or CCK-GCaMP6 (right) during exposure to a control toy rat. The toy rat was confined to the right of the map, as indicated by the red bar. (B, C) Mean dF/F (z-scored) during approaches toward the toy rat (top) or escapes from the toy rat (bottom) within 10 spatial bins of varying distance from the safer wall of syn-GCaMP6 (B, n = 5) or CCK-GCaMP6 (C, n = 6) mice (syn-approach, n = 7361 samples; syn-escape, n = 1884 samples; CCK-approach, n = 3632 samples; CCK-escape, n = 1128 samples). (D, E) Analysis of data shown in (B) and (C), respectively. Comparisons of mean dF/F during approaches and escapes from toy rat from samples in the left or right side of the enclosure. The toy rat was located in the right side of the enclosure. Average activity in bins on right side is greater than average activity in left side during escapes in syn-GCaMP6f mice (left side, n = 5 bins, right side, n = 5 bins; unpaired t-test, *p=0.0486). (F, G) Mean dF/F (z-scored) 5 s before and after approaches, escapes, freeze bouts, and stretch-attend postures in syn-GCaMP6 (F, n = 5) and CCK-GCaMP6 (G, n = 6) populations. (H, I) Mean dF/F (z-scored) in the empty zone and zone containing the toy rat in syn-GCaMP6 (H) and CCK-GCaMP6 (I) mice (syn, n = 5; CCK, n = 6; paired t-test, ns = not significant). Errorbars: mean ± SEM.