Fig. 6 |. Activation of PB→aNB→S1 circuits during NREM sleep drives the development of chronic pain.

a, Experimental design for in vivo Ca²⁺ imaging of PB neurons projecting to aNB. b, Representative images of PB→aNB projection neurons expressing GCaMP6s (left; the boxed region shown in the bottom panel), Ca²⁺ traces (middle), and Ca²⁺ activity in NREM before and after SNI (right; n = 234 cells from five mice; ****P < 0.0001). c, Experimental design for optogenetic inhibition of PB→aNB projections and Ca²⁺ imaging of aNB axons in S1. Opto-inhibition was performed during ZT2–10, in wake or NREM, for 5 days following SNI. d, Ca²⁺ activity of aNB→S1 axonal boutons during wake and NREM 14 days after SNI (n = 123, 105, 103, 110 boutons from three mice per group; wake, P < 0.0001, = 0.0004, 0.0005, 0.0073; NREM, ****P < 0.0001, = 0.014). e, Nociceptive thresholds (mean ± SEM) under various conditions (n = 5 mice per group). Daily inhibition of PB→aNB projections in NREM, but not wake, reduces punctate (NREM vs. SNI, P < 0.0001, = 0.0001, P < 0.0001), dynamic (P = 0.0001, < 0.0001, < 0.0001), cold (P = 0.0003, < 0.0001, = 0.0022), and hot (P = 0.0004, 0.0002, 0.0012) allodynia 14, 21, and 28 days after SNI. f, Representative two-photon images of DRG neurons expressing GCaMP6s (left), Ca²⁺ traces in NREM (middle), and Ca²⁺ activity before and after SNI (n = 140 cells from four mice; ****P < 0.0001). In b, d, f, box plot bounds and center, quartiles and median; whiskers, min and max. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; NS, not significant; by two-sided Friedman’s test followed by Dunn’s test (b, f), Kolmogorov-Smirnov test (d), two-way ANOVA followed by Bonferroni’s test (e). Detailed statistics in Supplementary Table 1.