Extrinsic Sensory Nerves

Extrinsic gut sensory neurons are separate from intrinsic sensory neurons within the enteric nervous system and are specifically designed to pick up nociceptive input, e.g. harmful chemical stimulation or high acidity. Their cell bodies are not within the gut wall, but within so-called dorsal root ganglia (DRG), nestled within the intervertebral foramina just outside the spinal cord, at thoracic, lumbar and sacral levels. Different DRG connect to different levels of the gut, with thoracolumbar DRG carrying information from the small intestine and proximal colon via afferents in splanchnic nerves, while lumbosacral DRG receive information from the distal colon and rectum via afferents within pelvic nerves. Within splanchnic and pelvic nerves, DRG afferents run alongside autonomic fibres, the cell bodies of which are located in celiac and mesenteric ganglia.

Classically, extrinsic nociceptive sensory neurons were primarily distinguished electrophysiologically, classified according to their firing properties. The gut is innervated by C- and Aδ fibres, which are unmyelinated and thinly myelinated fibres, respectively. Molecularly, there are a number of different C and Aδ fibre classes, with C fibres falling into two broad categories: those that use peptidergic neurotransmitters like CGRP and Substance C and those that do not. Within those categories, many other sub-types have since been defined with transcriptomic techniques.

In mouse, retrograde tracing experiments are starting to help define sensory neuron subtypes (Hockley et al. Gut, 2019, Wolfson et al. Cell, 2023). However, it is important to note that only colonic afferents have been captured in this way, i.e. after retrograde tracing, sacral, lumbar and thoracic dorsal root ganglia were harvested. Small intestinal nociceptive afferents have yet to be explicitly retrogradely traced for RNA sequencing, with most other extrinsic sensory neuron datasets in the gut focused on vagal afferents, and thus nodose ganglia (e.g. Zhao et al. Nature 2022). Vagal afferent activation does not produce conscious pain perception, but rather sensations like nausea or satiety.

In human, there are big atlases (e.g. Bhuiyan et al. bioRxiv, 2025 & Jung et al. Nat Comms, 2023) with easily accessible browsers (https://painseq.shinyapps.io/u19humandrgatlas/ & http://research-pub.gene.com/XSpeciesDRGAtlas/) that show what C fibre and putative Aδ-fibre subtypes express more generally – but these are based on post-mortem tissue from whole dorsal root ganglia – so will not just contain neurons which innervate the gut.

To study sensory neuron function in vitro, we are fortunate that adult sensory neurons grow happily in a dish for at least a week, even when harvested post-mortem from human donors. These can then be studied with electrophysiological techniques as well as with calcium imaging. Work with rodent primary sensory neurons is therefore very common, while a few select centres world-wide have ethics to use sensory neurons from human donors.

On the flip side, primary cultures contain a high proportion of non-neuronal cells from within the sensory ganglia (i.e. cell types that are absent from the microenvironment of the gut). Only in mouse do we know how to entirely remove these.

Another option that is therefore popular at the moment is the culture of induced pluripotent stem-cell derived sensory neurons. These cultures are human and can be very pure, but do not recapitulate all features of native sensory neurons. RNA sequencing datasets (e.g. Li et al. Pain 2025) are freely available for scientists to determine ahead of time whether this model contains what they need for their specific research question.

Text by Franziska Denk, King’s College London