Many infectious diseases alter intestinal functions. Symptoms of such diseases often include diarrhea and constipation that are associated with changes in the enteric nervous system (ENS), which regulates the propulsion of content within the gastrointestinal tract. In this issue of the JCI, Janova et al. (1) explored mechanisms by which West Nile virus (WNV) produced dysmotility in mice as a model for infectious insults to the gut. The authors built on their earlier work implicating T cells in the death of enteric neurons leading to dysmotility (2). Janova and authors showed that adult neurogenesis followed the acute response to infection (1). These findings suggest that neurogenesis may be the key to recovery from all neurotoxic change to the ENS, whether triggered by infection, cancer chemotherapy (e.g. oxaliplatin) (3), inflammation (4), or other insults (5). Typically, enteric neurotoxicity leads to persistent dysfunction, so initiation of neurogenic recovery is vitally important for human health and quality of life.

Janova et al. (1) used intersecting approaches involving transgenic mice and blocking antibodies. WNV infected enteric neurons and glia, leading to cell death induced by helper (CD4+) and cytotoxic (CD8+) T cells. By post-infection day seven, whole intestinal transit time, neuron numbers, and the density of neural and glial networks were affected. The authors systematically explored the roles of both T cell classes, showing that other immune cells (neutrophils, B cells, nonresident macrophages/monocytes) had little or no role in the response to WNV (1). This finding contrasts with other viruses, such as herpes simplex virus, which induces various immune cells (6, 7). Notably, resident macrophages reduced damage (1) and have been proposed in other contexts to provide neuroprotection (8). Given that nonresident macrophages accumulated within affected ganglia, the mediators of pathogen-induced neurotoxicity probably did not reflect the changes in immune cell numbers. Notably, the mechanism by which the T cells produced injury predominantly involved cytotoxic and death receptor pathways, via perforin and Fas ligand, respectively (1).

The loss of neuron cell bodies due to WNV appears to be relatively nonspecific, as both myenteric and submucosal plexus neurons are affected, with major neuronal subtypes, including calretinin- and nNOS-positive neurons, being equally vulnerable. Loss of diverse neuronal groups probably accounts for dysmotility, as cell bodies are where input from other neurons is integrated, so their death blocks physiological action potential signaling, even though the decentralized axons persist for several days. Thus, dead neurons would be unable to participate in network activity, or contribute output to smooth muscles, or activate other effector tissues. Since some calretinin neurons are excitatory motor neurons and most nNOS neurons are inhibitory motor neurons (5), neural control of the smooth muscle becomes compromised and is likely to produce localized strictures with proximal accumulation of content where nNOS innervation is particularly affected (1). Such strictures substantially delay transit.

It should be noted, however, that loss of submucosal neurons (1) is expected to result in dysfunction due to disrupted neurogenic water and electrolyte secretion, rather than dysmotility. Furthermore, general loss of enteric neurons would alter neurally driven immune activity, mucus secretion, and epithelial cell production in the crypts, among many other functions (5).

Loss of cell bodies has other, less straightforward effects by depriving surviving neurons of some synaptic targets and depriving others of inputs. This synaptic loss may have retrograde effects on surviving neurons and lead to rewiring within the remaining circuit. However, there are many different functional subtypes of neurons mixed together within any myenteric ganglion (5), so axons of surviving neurons will have a variety of synaptic targets in different intermingled enteric circuits from which to choose. Thus, new connections may aid or impede the recovery of function depending the appropriateness of these synapses. Whether such opportunistic synaptogenesis occurs within a damaged ENS is unknown, but is a key question for all enteric neuropathies (9).