The nervous system of C. elegans comprises a total complement of only 302 neurons, yet contains the foundational circuit principles of vertebrate nervous systems, which, in humans,  contain about 86 billion neurons (Goriely, 2024). In the mid-1960s, Sydney Brenner pioneered the use of C. elegans as a genetic model system to dissect nervous system development and function (Brenner, 1974). Since then, extensive research has generated comprehensive datasets detailing both the structural connectivity of the nervous system (White et al., 1986; Cook et al., 2019) and single-cell gene expression patterns across development (Taylor et al., 2021). These data are curated in publicly accessible databases such as WormAtlas (Altun et al., 2009) and CeNGEN ( Caenorhabditis elegans Neuronal Gene Expression Network) (Taylor et al., 2021).

Gap junctions are specialized ion channels that facilitate electrical coupling between cells (Sohl et al., 2005). They are made up of innexin subunits (connexins in vertebrates), of which there are 25 in C. elegans (Starich et al., 1996; Simonsen et al., 2014). In this study, I analyse the expression of innexins in identified motor neurons and command interneurons. Specifically, I examine the spectrum of expressed innexins at coupled electrical synapses, where striking molecular asymmetries suggest these junctions function as rectifying diodes (Starich et al., 2009; Liu et al., 2020).

The expression maps of the 25 C. elegans innexins reveal two general patterns: generalized expression, such as unc-7 and unc-9 which are co-expressed across 69.9% of all neurons, and restricted expression, where certain innexins exhibit relatively high expression in small neuronal subsets, such as inx-21 and inx-22 which are uniquely expressed in SMD neurons (Taylor et al., 2021).

A generic gap junction is broadly considered a symmetrical structure that permits bidirectional passage of ions or small molecules between coupled cells (Sohl et al., 2005). However, functionally asymmetric electrical synapses have been described that preferentially permit current flow in one direction (Phelan et al., 2008; Starich et al., 2009; Liu et al., 2020). Such gap junctions are structurally asymmetric, most straightforwardly achieved by incorporating differing innexin subunits into the paired hemichannels of adjacent cells (Palacios-Prado et al., 2014).

Figure 1A details examples of molecular asymmetry in neural circuits associated with major motor neurons innervating body wall muscles (VA, VB) and head muscles (SMB, SMD).  These classes of neuron all have Extended Longitudinal Neurites (ELNs).  One prominent form of asymmetry involves the stoichiometric ratio of the widely expressed innexins unc-7 and unc-9 . While the mean organism-wide unc-7 / unc-9 ratio across co-expressing neurons is 1.60, major command interneurons AVA and AVB exhibit ratios of 6.15 and 5.65, respectively—nearly four times the neuronal baseline. Even more prominent innexin asymmetries are demonstrated by the restricted expression of inx-21 and inx-22 in SMD and eat-5 in SMB and SAA.

Strikingly, these asymmetrical gap junctions selectively involve circuitry tied to motor neurons driving body (VA, VB) and head (SMB, SMD) locomotion ( Fig. 1B, C). All these motor neuron classes share the hallmark anatomical feature of ELNs: morphologically undifferentiated longitudinal processes situated distal to their neuromuscular output zones ( Fig. 1D, E). It has long been posited that these distal extensions act as proprioceptors, transducing local body bending into downstream muscle activation during undulatory locomotion (White et al., 1986; Wen et al., 2012).

Rectifying gap junctions function effectively as blocking diodes that restrict ionic current flow to a single direction (Shui et al., 2020). For motor neurons possessing dual motor-proprioceptive modalities (VA, VB, SMB, SMD), diode rectification would be vital to prevent locally generated, stretch-activated electrical signals from back-propagating antidromically into critical central interneurons, such as the major network hub RIB ( Fig. 1C ). Supporting this hypothesis, major motor neurons lacking extended proprioceptive process extensions, such as RMD, exhibit no corresponding innexin asymmetries in their gap junctions.

Combining connectomic mapping with cell-specific transcriptomic profiling can provide key functional insights into neural circuitry. Such findings provide testable hypotheses to guide future electrophysiological experiments and computational modelling aimed at deciphering how a nervous system generates the behaviours that control an organism.

Transcriptomic and Connectomic Data Analysis

Single-cell RNA sequencing expression levels (Transcripts Per Million, TPM) for all 25 Caenorhabditis elegans innexins across 133 identified neuronal cell types were extracted from the adult CeNGEN dataset (Taylor et al., 2021). Co-expression frequencies and mean expression ratios were calculated using custom Python scripts running in Google Colab (Python 3.10; pandas v2.0, numpy v1.24). Neuronal co-expression of unc-7 and unc-9 was defined as non-zero TPM values (>0.0 TPM) in both genes across the 133 annotated neuron classes. Anatomical connectivity, chemical and electrical synapse counts, and motor neuron process morphologies were cross-referenced against the EM connectome datasets (White et al., 1986; Cook et al., 2019) and WormAtlas (Altun et al., 2009).