Extended Data Fig. 8: Divergent regulation of l/vlPAG by two distinct vLGN-derived circuits: vLGN-l/vlPAG and vLGN-LHA-l/vlPAG pathways engage non-overlapping cellular populations. Related to Fig. 8.
a, Left, rabies virus tracing the l/vlPAG receiving input from the LHA. Right, injection site of l/vlPAG illustrating the star cells (yellow). Scale bar, 200 μm (upper), 50 μm (bottom). b, Representative images of the areas projecting to the l/vlPAG that receive input from the LHA. Scale bar, 200 μm. Right, quantifications of neurons projecting to the l/vlPAG in different brain regions (n = 4 mice). c, Specific injection of l/vlPAG postsynaptic neurons with DIO-hM3Dq or DIO-eGFP. Scale bar, 200 μm. d-f, Changes in food intake during the daytime (d) and the nighttime (e, eGFP vs. hM3Dq P = 0.6079) in chow-fed mice; n = 8 mice. Weekly time course of body weight (f, 2 weeks P = 0.6414, 3 weeks P < 0.0001); n = 8 mice. g, h, 1-h (g, P < 0.0001) and 6-h (h, P = 0.0217) food intake test in chow-fed mice; n = 8 mice. i-l, Depressive- and anxiety-like behaviors in different experimental groups; n = 7 mice. OFT (i; Center time, P = 0.5367, Total time, P = 0.2673), EPM (j, P = 0.4570), forced swim test (FST) (k, P = 0.6917), and tail suspension test (TST) (l, P = 0.9249). m, Mice were injected with l/vlPAG postsynaptic neurons using DO-hM3Dq or DO-eGFP, which do not target neurons receiving direct inputs from the LHA. Scale bar, 200 μm. n, o, OFT (n; Center time, P < 0.0001, Total time, P = 0.0001), EPM (o, P < 0.0001) in different experimental groups; n = 7 mice. Two-way ANOVA with Sidak’s multiple-comparisons test was used in d, e, and f. Two-tailed unpaired t-test was used in g-o. *P < 0.05; **P < 0.001; ***P < 0.0001; ns, no significant difference. All error bars indicate the mean ± s.e.m. Detailed statistical information is available in Supplementary Table 1.
