Wiring specificity of neural circuits, how precise needs it to be?
A prerequisite for the proper function of almost any neural circuit is the establishment of the precise and organized connections between neurons, referred to as the wiring specificity. Chapter 7 of Principles of Neurobiology summarizes many of the key steps in the development of wiring specificity in the nervous system. The chapter then delves into detailing the molecular mechanisms and general principles of establishing wiring specificity. It begins by using the olfactory circuits in flies and mice as examples and extends the discussion to more complex brains, such as those in humans. While numerous mechanisms have been discovered, however, from an evolutionary perspective, ‘neuronal wiring only needs to be specified as precisely as is useful’. Being overly precise not only adds little to the function of neural circuits, but also increases the burden on every neuron when it is searching for the correct synaptic partners. So, the question arises: How precise does wiring specificity need to be?
In Chapter 7, two types of situations are discussed. The first one is that while two types of neurons are specifically connected, the individual neurons within each type do not need to be defined further. This scenario is exemplified by motor neurons and muscles in the neuromuscular system as well as for the olfactory receptor neurons and projection neurons in the fly’s olfactory system. The second situation pertains to two types of neurons whose connections are somewhat stochastic, as observed in the connections between the fly’s olfactory projection neurons and the mushroom body neurons. However, a quantitative characterization of the precision at the synaptic level has been absent. Do 100% of the synapses from one neuron all form connections with the correct partner? If not all, what is the percentage? Due to technical limitations, the answer, which provides an upper limit for the specificity question, has been lacking until now.
Two recent companion papers systematically analyzed the connectome of two adult fly brains (comprising three hemispheres) acquired by electron microscopy (Schlegel et al. and Dorkenwald et al.). Here, I focus on the paper by Schlegel et al that annotated cell types by comparing across different brains. Measuring wiring precision at the synaptic level requires two key steps: defining cell types and counting ‘functionally meaningful’ synapses between these cell types. As stated in the paper, ‘a cell type is a falsifiable hypothesis about biological variability within and across animals.’ To assign cell types, Schlegel et al established a systematic and hierarchical set of annotations, considering both the anatomical organization of the brain and the developmental origin and coarse morphology of neurons (Figure 1). In total, they assigned ‘3,166 high-confidence consensus cell type labels for 42,687 neurons from three different hemispheres and two different brains.’ These cross-matched neurons form the basis for measuring wiring precision.
Now returning to the question: How many synapses of one neuron are connected to the correct partners? Practically, what we care about is that how many synapses are formed in a tightly controlled manner and appear reliably across animals, versus how many synapses are formed in a less controlled manner, displaying less conservation across individuals. A simple heuristic concluded by Schlegel et al. is that ‘connections stronger than 10 unitary synapses or providing >1% of the input to a target cell have a greater than 90% chance to be preserved’ between two brains. Hence, these connections are more likely to be tightly controlled during development and considered functionally relevant.
Much more could be learned about wiring specificity from these well-annotated connectomes in adults. For instance, Schlegel et al. found that the variance between the left and right hemispheres is consistently smaller than the variance across different animals. Does this difference primarily arise from the distinctions between the two animals, or does it suggest some coordination, between the left and right hemispheres, of wiring specificity during development that awaits further illustration in future work?
Figure 1. Hierarchical annotation schema for the FlyWire dataset. Adapted from Schlegel et al. and Dorkenwald et al.
Reference:
1. Schlegel, P., Yin, Y., Bates, A. S., Dorkenwald, S., Eichler, K., Brooks, P., ... & Jefferis, G. S. X. E. (2023). A consensus cell type atlas from multiple connectomes reveals principles of circuit stereotypy and variation. Biorxiv.
2. Dorkenwald, S., Matsliah, A., Sterling, A. R., Schlegel, P., Yu, S. C., McKellar, C. E., ... & FlyWire Consortium. (2023). Neuronal wiring diagram of an adult brain. bioRxiv.