Dorkenwald, S., Turner, N. L., Macrina, T., Lee, K., Lu, R., Wu, J., … & Seung, H. S. (2022). Binary and analog variation of synapses between cortical pyramidal neurons. Elife, 11, e76120. https://elifesciences.org/articles/76120
Serial Section Transmission Electron Microscopy and Connectomic Analysis
Mouse Visual Cortex
The authors traced and proofread 1960 synaptic connections among 334 pyramidal neurons in a serial section TEM volume covering layer 2/3 of mouse V1. They quantified the size, and thus functional strength (Holler 2021), of each of these synapses. Plotting a histogram of these synapse sizes showed, instead of the expected log-normal distribution, a highly skewed distribution that was better fit by a mixture of two log-normal distributions. They then analyzed a subset containing 320 of these synapses from 160 matched pairs (two synapses that share the same pre- and post-synaptic cells). Such matched synapses have identical learning history and thus could be expected to be correlated in size. Plotting the histogram of the geometric mean of each matched pair showed an even clearer bimodal distribution which the authors showed could be modeled by assuming each pair had the same binary value (small or large) with an uncorrelated analog noise on top. This is evidence for a Hebb-style learning rule with only two synaptic strength states (small vs large). An excellent example of EM connectomics providing novel insight into one of neuroscience’s most fundamental questions.
“Here, we show that synapse size, when restricted to synapses between L2/3 pyramidal cells, is well modeled by the sum of a binary variable and an analog variable drawn from a log-normal distribution. Two synapses sharing the same presynaptic and postsynaptic cells are known to be correlated in size. We show that the binary variables of the two synapses are highly correlated, while the analog variables are not.”
