(A and B) SBFSEM images of retinal terminals in P42 control (A, labeled in green) and Lrrtm1−/− (B, labeled in red) dLGN. RG synapses are depicted in insets a1-a4 (Ctl) and b1-b4 (Lrrtm1−/−). In insets, each retinal terminal is depicted in a unique color. Similar colors in different insets do not represent axonal branches of the same RGC. 3D reconstruction of retinal terminals and relay cell dendrite are depicted on the right. The black arrows denote the position of dendrites stemming from relay cells somas. (C) Percentage of retinal terminals contributing to forming RG synapses with 1, 2, 3 or ≥4 distinct retinal terminals in P42 Lrrtm1−/− and control dLGN. Data represent Mean ±SEM; ***p<0.0001, by ANOVA. (D–F) Quantification of terminal size (D), actual number of active zones per terminal (E) and the active zones: terminal area ratio for simple and complex RG synapses in dLGN of Lrrtm1−/− and control mice (NT = nerve terminal). Data represent Mean ±SEM; ***p<0.0001, by ANOVA. (G and H) Retinal terminals were multicolor-labeled by injecting 1–2 µl brainbow AAVs into the vitreous humor of Lrrtm1−/−::Calb2-Cre and control mice. (I) Color analysis of clustered retinal terminals in wild type and Lrrtm1−/− mutants revealed a lower level of complex RG synapses in mutants. Scale bar, 10 µm (A and the 3D image), 20 µm (G), 1 µm (insets).