Retinal connectomics: Towards complete, accurate networks

Abstract

Connectomics is a strategy for mapping complex neural networks based on high-speed automated electron optical imaging, computational assembly of neural data volumes, web-based navigational tools to explore 1012–1015 byte (terabyte to petabyte) image volumes, and annotation and markup tools to convert images into rich networks with cellular metadata. These collections of network data and associated metadata, analyzed using tools from graph theory and classification theory, can be merged with classical systems theory, giving a more completely parameterized view of how biologic information processing systems are implemented in retina and brain. Networks have two separable features: topology and connection attributes. The first findings from connectomics strongly validate the idea that the topologies complete retinal networks are far more complex than the simple schematics that emerged from classical anatomy. In particular, connectomics has permitted an aggressive refactoring of the retinal inner plexiform layer, demonstrating that network function cannot be simply inferred from stratification; exposing the complex geometric rules for inserting different cells into a shared network; revealing unexpected bidirectional signaling pathways between mammalian rod and cone systems; documenting selective feedforward systems, novel candidate signaling architectures, new coupling motifs, and the highly complex architecture of the mammalian AII amacrine cell. This is but the beginning, as the underlying principles of connectomics are readily transferrable to non-neural cell complexes and provide new contexts for assessing intercellular communication.

High-Resolution Synaptic Connectomics

“…The proper motivation for connectomics is graph theory: the analysis of network connection patterns and how signals travel through networks [72]. Formal schemas for exploring network topologies and their component motifs (i.e., stereotyped components) can scale to massive levels.…”

“…For biological neural networks, there is some debate about the nature of ground truth. We have made the argument that ground truth datasets for testing imaging and algorithms ought to come from neural TeM samples with Nyquist-compliant resolutions [72]: images with resolutions superior to the structures being resolved. For synapses, gap junctions, adherens junctions, endocytosis sites, and discovery of new cell contact features, that means 2 nm pixel sizes or smaller [4].…”

“…Neurons that integrate signals from other cells are graph vertices, and the individual connections they form via synapses, gap junctions, or other spatially discrete signaling modalities are graph edges [72]. While classical approaches treat networks as electrical circuits, they are not loop circuits at all, but rather flow systems.…”

“…in this view, connectomics is not a viable effort. Morgan and lichtman [81] give a strong defense of connectomics on this point, but it can be further argued that the high diversity of modeling solutions obtained by Prinz et al [84] means that network topology and synaptic weights must both be demonstrably constrained to develop a functioning system of repeating units known to underlie vertebrate brain and retinal organization [72]. Ultimately, due to the massive diversity of neuronal classes, classical TeM sampling has been unable to completely specify any outflow module in the retina, much less than any other complex brain region.…”

Glycine receptor alpha subunit (GlyRa) specific inhibition contributes to ganglion cell signaling in mouse retina.

“…Once the types of RGCs are defined, the recombinant G protein-deleted rabies virus could be used to retrogradely trace the presynaptic partners of the infected RGCs (Farrow et al, 2013;Wall et al, 2010). Also, the retinal connectome constructed using electron microscopy data could be explored for candidates of the proposed circuits (Anderson et al, 2011;Marc et al, 2013).…”