Gut–brain communication: Do sensory fibres in the vagus nerve innervate the lower gastrointestinal tract?

Over the past few years, there has been considerable interest in the role of the vagus nerve in health and disease, particularly regarding communication between the gastrointestinal (GI) tract and the brain; the so-called ‘gut–brain axis’. Accumulating evidence suggests that health and disease might be modified significantly by activity within vagal sensory fibres along the gut–brain axis (Doenyas et al., 2025; Ma et al., 2025). In this dynamic era of gut–brain research, we must have a clear understanding of which regions of the GI tract are innervated by the vagal and/or spinal afferent neurons (Spencer et al., 2025). The latest research from different research groups reveals non-trivial inconsistencies regarding the sensory innervation provided by the vagus nerve to the lower GI tract (distal colon and rectum). This region of the GI tract, defined as within 30 mm of the anal sphincter in mice, is commonly studied at a functional level in gut–brain research and visceral pain models. Hence, to advance the field of gut–brain research, it is essential that we have a clear understanding of which sensory pathways underlie gut–brain communication. The vagus nerve (cranial nerve X) is a mixed sensory and motor nerve that provides the primary parasympathetic motor supply to most internal organs. However, most axons (∼90%) in the vagus nerve are sensory, with cell bodies located in the nodose ganglia. These ganglia provide long axons that detect various sensory cues from internal organs, which then relay these signals to the brainstem. In the GI tract, vagal sensory fibres have been studied extensively in the upper gut (stomach and small intestine) (Berthoud et al., 2025; Wang et al., 2020), with comparatively fewer studies on the large intestine because the vagal sensory fibres provide a limited innervation to the distal gut (distal or descending colon and rectum) in vertebrate animals (Wang & Powley, 2000). This is supported by results obtained by anterograde tracing from nodose ganglia revealing that vagal afferent innervation decreases progressively along the length of the GI tract, with a sparse or absent innervation towards the terminal large intestine (distal colon and rectum; Figure 1; Spencer et al., 2024; Wang & Powley, 2000). In contrast to what has been taught for many years in medical schools, recent studies over the past 5 years from several different laboratories have suggested that the vagus nerve provides a robust sensory innervation to the terminal part of the GI tract, specifically the distal colon of mice (Kaelberer et al., 2018; Meerschaert et al., 2020; Osman et al., 2023; Wang & Tache, 2024; Wang et al., 2024). It has even been suggested that ≤25% of neurons in the nodose ganglia innervate the mouse colon (Osman et al., 2023), with ∼10% of nodose neurons innervating the distal colon (Meerschaert et al., 2020; Osman et al., 2023; Wang & Tache, 2024; Wang et al., 2024). The idea that vagal sensory nerves can innervate the proximal and mid colon is not new (Borgmann et al., 2021; Spencer et al., 2024; Wang & Powley, 2000). However, the idea that the terminal large intestine receives a significant population of vagal afferents differs from what has been understood previously. If correct, this would have important new ramifications for our understanding of the spinal and vagal afferent pathways that underlie sensory transmission along the gut–brain axis. The common technical approach adopted in recent studies is the use of retrograde tracing from the colonic wall to label vagal afferent nerve cell bodies. Retrograde tracing involves injection of a neuronal tracer into the colonic wall, where many different types of cells (including vagal sensory neurons) absorb the tracer, which is then transported back to the cell bodies in nodose ganglia after a period of days. Recent studies have used injections of cholera toxin B (CTB) into the distal colon of mice (Meerschaert et al., 2020; Osman et al., 2023; Wang & Tache, 2024; Wang et al., 2024). The studies differ slightly in the volume of tracer injected (usually between 3 and 5 µl (see Meerschaert et al., 2020; Osman et al., 2023; Wang & Tache, 2024) via two to five independent injection sites, at a distance of 15–30 mm from the anal sphincter. All these recent studies obtained similar findings; i.e. extensive labelling of nerve cell bodies in nodose ganglia (Meerschaert et al., 2020; Osman et al., 2023; Wang & Tache, 2024; Wang et al., 2024). These findings were interpreted as evidence that the vagus nerve provides a prominent sensory innervation to the distal colon of mice. It should be noted that in rodents, such as mice, the pelvic bone is used as an anatomical landmark to distinguish the distal colon (located rostrally) from the rectum (located caudally) (Wang et al., 2024), where the rectum is innervated by the rectal nerves (Spencer et al., 2025). A recent study by Hibberd et al. (2025) reproduced similar findings to other groups, in that injection of CTB-488 into the mouse distal colon labelled many sensory neurons in nodose ganglia (Meerschaert et al., 2020; Osman et al., 2023; Wang & Tache, 2024; Wang et al., 2024). Hibberd and colleagues reasoned that if their retrograde tracing results were an accurate reflection of the vagal afferent innervation to the distal colon, then anterograde tracing from nodose ganglia should also reveal vagal axons and their terminal endings in this region. However, anterograde tracing studies were not able to identify any vagal afferent axons in the distal colon or rectum (Figure 1; Hibberd et al., 2025). In fact, few vagal afferents innervated the proximal colon and progressively less in the mid colon, with no fibres caudal to the mid colon (Hibberd et al., 2025). This inconsistency between the results of retrograde and anterograde tracing prompted Hibberd and colleagues to inspect the injection site of CTB in the distal colon closely. What they found was alarming. They identified extensive CTB spread within the colon wall, as far as the proximal colon (Figure 2, 3; Hibberd et al., 2025) despite applying smaller volumes of tracer (1 µl) compared with other studies, via a single CTB injection from a fine glass micropipette (∼10 µm tip diameter). It is noteworthy that in some studies wide-bore (31 gauge) insulin syringes have been used to inject ≤10 µl of CTB into the colonic wall of mice, via multiple puncture sites (Meerschaert et al., 2020). These needles have an inner diameter of 133 µm (outer diameter, 261 µm), which is equivalent to the entire thickness of the mouse colon circular muscle. Other studies have used similar wide-bore 33 gauge Hamilton syringes in the distal colon of mice (Osman et al., 2023) and rats (Herrity et al., 2014). The assumption with retrograde tracing studies is that the injected tracer does not leak out of the gut wall from the injection site and that the tracer does not spread extensively within the gut wall, remaining spatially restricted at the injection site. However the study by Hibberd et al. (2025) suggests that retrograde tracing, even using single injection sites with tiny volumes of CTB (≤1 µl), leads to extensive non-specific spread throughout the colon (Figure 2, 3), providing no confidence that results obtained using retrograde tracing in their study should be considered credible. Hibberd et al. (2025) also noted that after anterograde tracing from nodose ganglia, no labelled vagal axons were identified in the bladder or uterus of mice, yet injections of CTB into the bladder and uterus extensively labelled sensory neurons in nodose ganglia, owing to spread of tracer within the abdominal cavity. This raises further concern regarding the specificity of using CTB as a retrograde tracer in visceral organs. The finding that anterograde tracing from nodose ganglia was unable to identify any vagal afferent axons in the distal colon or rectum of mice was supported by retrograde labelling using small volumes of another retrograde tracer, DiI, injected into the distal colon (Hibberd et al., 2025). Hence, it can be speculated that CTB-488 is particularly well absorbed in the viscera among retrograde tracers, raising significant doubts regarding the location of uptake. Overall, Hibberd et al. (2025) suggest that the vagus nerve provides either no sensory innervation to the distal region of GI tract of mice or a sparse sensory innervation of this region, consistent with the original suggestions raised by the work of others, decades earlier (Wang & Powley, 2000). Indeed, it has been taught for many years that the human colon receives a vagal afferent innervation as far as the right two-thirds of the transverse colon, with no vagal innervation of the descending colon or rectum (Browning et al., 2017; Grey, 2008). Transgenic reporter mice with selective labelling of vagal afferents (using the Phox2B reporter line) have helped to verify the extent of vagal afferents along the length of the GI tract. Borgmann et al. (2021) noted that ‘innervation density of PHOX2B vagal afferents decreased beyond the ileum’ and that ‘significantly fewer endings in muscular layers and sparse innervation of crypts’ were detected. In visceral pain models, distension of the distal colon and rectum is commonly used to elicit visceromotor responses. A recent functional study showed that colonic distension in conscious mice could activate neurons in the nucleus of the tractus solitarius (Wang et al., 2024). This was interpreted as evidence that vagal afferent endings in the distal colon were directly activated by colonic distension (Wang et al., 2024). However, colonic distension in conscious mice induces multiple behavioural changes, including mobility, writhing, facial grimace and vocalization, and all these stimuli can readily activate other cranial nerves (e.g. V, trigeminal nerve), VII (facial) and IX (glossopharyngeal), all of which provide direct inputs into the nucleus of the tractus solitarius, not only the vagus nerve. Thus, activating the nucleus of the tractus solitarius via colonic distension is not direct evidence that vagal afferents innervate the distal colon. Fully implantable vagal nerve stimulators have been approved by the US Food and Drug Administration as controlled therapies for a variety of human conditions, including epilepsy and treatment-resistant depression (O'Reardon et al., 2006). There has also been extensive testing of vagal nerve stimulators for a variety of psychiatric and inflammatory disorders (Breit et al., 2018; George et al., 2008), including treatment of rheumatoid arthritis (Koopman et al., 2016). This raises the importance of understanding which organs receive an innervation from the vagus nerve. The purpose of this article is not to disagree with the results of recent investigators who have used retrograde tracing from the distal GI tract. Hibberd et al. (2025) demonstrate that vagal afferent neurons can be retrogradely labelled from the distal colon of mice. The fundamental difference lies in the interpretation of the results. Hibberd et al. (2025) suggest that labelling nodose neurons from the distal colon and rectum using retrograde tracing provides unconvincing evidence that this region receives a prominent vagal afferent innervation. We would like to suggest that the field take extreme caution when interpreting the results obtained by retrograde tracing from visceral organs and that the distal colon and rectum receive an extensive spinal afferent innervation, with either no vagal afferents or an extremely sparse vagal afferent innervation to this region. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. None declared. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed. None.