Exclusion criterion for inappropriate synapses that are formed between TA proprioceptors and Gs motor neurons, density maps for TA and Gs motor neuron pools, and no change in appropriate synapses in C3 ^−/− and CD47 ^−/− mice. (A) Criterion for exclusion of synapses from our analysis. Image depicting an excluded synapse labelled in DsRed (red) and colocalized with VGluT1 (green) because it was not opposed to a dendrite of a TA motor neuron (shown in green; see graph where there is no overlap between green and those in red and blue) in a P5 PV::FlpO;Ai65 mouse at P5. Scale bar: 10 μm. (B) Z-stack confocal images (total: 3.5 μm) from a Gs motor neuron pool in PV::FlpO;Ai65 and PV::FlpO;Ai65;C3^−/− mouse injected with rAAV6-GFP virus in Gs muscle and CAV2^CRE in the TA muscle. Scale bar: 20 μm (C) Percentage of motor neurons (MNs) with inappropriate synapses from TA proprioceptors to Gs motor neurons. Significance: *P < 0.05; unpaired t test. (D) Identification of inappropriate synapses in PV::FlpO;Ai65;C1q^−/− mice, labelled by DsRed (red) and VGluT1 (blue) on TA motor neurons (GFP; green). Yellow arrows show dendritic and somatic inappropriate synapses in higher magnification images. Scale bar: 20 μm. (E) Percentage of motor neurons with inappropriate synapses in WT and C1q^−/− mice (WT, N = 3 mice; C1q^−/−, N = 4). **P < 0.001; unpaired t test. (F) Density maps for TA (red) and Gs (blue) motor neuron pools in WT, C3^−/−, and CD47^−/− mice. Maps represent the ventro-dorsal and mediolatearal distribution of MNs between L4 and L5 (450 μm total). The value 0 in y-axis represents the central canal. (G and H) Quantification of “hot area” distribution of each type of motor neuron, TA (G) and Gs (H). ns: no significance, one-way ANOVA with multiple comparisons using Bonferroni’s test. D, dorsal; V, ventral; M, medial; L, lateral. (I) Schematic illustration of experimental design. The TA muscle was co-injected with AAV6-GFP (blue) and CAV2-Cre (red) viruses using PV::FlpO^+/+;Ai65^+/+ mice at birth (P0) and the spinal cord was examined at P5. (J) Single optical plane confocal images from L4 motor neurons at P5 in CD47^−/− (top) and C3^−/− (bottom) mice. Immunostaining against GFP (blue), homonymous proprioceptive synapses (amplified by an antibody against DsRed; immunoreactive signal is shown in red) and VGluT1⁺ synapses (white). GFP labels TA motor neurons only. DsRed labels the fibers and synapses of proprioceptive neurons originating in the TA muscle. Fluorochrome-separated images are shown for clarity for DsRed only and VGluT1⁺ only synapses, as well as merged images (motor neurons are GFP⁺, proprioceptive fibers/synapses are DsRed⁺, VGluT1 for synapses only) (scale bar: 10 μm). Insets represent higher magnification images of individual synapses; white arrows indicate colocalization between DsRed and VgluT1⁺ signals (inset scale bars: 4 μm). (K) Percentage of homonymous proprioceptive synapses (VGluT1⁺ and DsRed⁺) per TA motor neuron soma, relative to the total number of VGluT1⁺ synapses per motor neuron soma. At least five TA motor neurons were examined per mouse. WT, N = 3 mice; C3^−/−, N = 3; CD47^−/−, N = 3. ns: no significance; one-way ANOVA, post hoc Bonferroni’s test.