Figure 10. Structural connectivity between antennal lobe glomeruli and third order olfactory neurons.

(A) Input tuning curves for a set of EM reconstructions comprising 17 LHLNs, and 29 LHONs (Schlegel, Bates et al. ms in prep) whose dendrites are restricted to the LH. The x axis is rank ordered by the number of synaptic inputs from uniglomerular PNs from each of 51 glomeruli. (B) Dot plots showing distribution of number of input glomeruli for LHLNs, and LHONs, and the number of dendritic claws for KCs. KC claw number data was based on light microscopy (LM) data presented in Caron et al. (2013) or EM reconstructions in Zheng et al. (2018). Some LHNs receive no excitatory uniglomerular PN input; their feed-forward input may come from multiglomerular or non-olfactory projections to the LH. Group means are shown by large black dots, error bars indicate a single standard deviation from the mean. See also Materials and methods. In sum, a reasonable cut-off for the observed structural connectivity in our EM-reconstructed wiring diagram reveals that the average LHN and each KC may sample the same number of glomeruli, via uniglomerular excitatory olfactory PNs. LHNs exhibit a larger standard deviation; some LHNs may act in a specific labelled line for unique and behaviorally significant odors, others may need to sample a larger number of ethologically connected odors more broadly.