Mechanisms underlying the gut–brain communication: How enterochromaffin (EC) cells activate vagal afferent nerve endings in the small intestine

How the gastrointestinal tract communicates with the brain, via sensory nerves, is of significant interest for our understanding of human health and disease. Enterochromaffin (EC) cells in the gut mucosa release a variety of neurochemicals, including the largest quantity of 5‐hydroxytryptamine (5‐HT) in the body. How 5‐HT and other substances released from EC cells activate sensory nerve endings in the gut wall remains a major unresolved mystery. We used in vivo anterograde tracing from nodose ganglia to determine the spatial relationship between 5‐HT synthesizing and peptide‐YY (PYY)‐synthesizing EC cells and their proximity to vagal afferent nerve endings that project to the mucosa of mouse small intestine. The shortest mean distances between single 5‐HT‐ and PYY‐synthesizing EC cells and the nearest vagal afferent nerve endings in the mucosa were 33.1 ± 14.4 µm (n = 56; N = 6) and 70.3 ± 32.3 µm (n = 16; N = 6). No morphological evidence was found to suggest that 5‐HT‐ or PYY‐containing EC cells form close morphological associations with vagal afferents endings, or varicose axons of passage. The large distances between EC cells and vagal afferent endings are many hundreds of times greater than those known to underlie synaptic transmission in the nervous system (typically 10–15 nm). Taken together, the findings lead to the inescapable conclusion that communication between 5‐HT‐containing EC cells and vagal afferent nerve endings in the mucosa of the mouse small intestinal occurs in a paracrine fashion, via diffusion.


INTRODUCTION
The gut-brain axis has been of great interest to many aspects of medical science for a number of decades (Aziz & Thompson, 1998;Camilleri, 1990;Mayer & Raybould, 1990;Mearin et al., 1991).In recent years, extraordinary interest has developed regarding the mechanisms by which gut-brain communication via sensory nerves can control health and disease (Cryan et al., 2019;Krieger et al., 2022;Margolis et al., 2021).It is now appreciated that sensory nerve pathways between the gut and brain have significant roles in numerous physiological processes that underlie homeostasis, including hunger and satiety (Borgmann et al., 2021), responses to fluid and food intake (Bai et al., 2019;Shechter & Schwartz, 2018), gastric acid secretion (Mediavilla, 2020;Raybould et al., 1990), and responses to stress and mood (Guerrero-Hreins et al., 2021).In fact, a variety of neurodegenerative diseases have also been linked to disruptions of the gut-brain axis (Yu & Li, 2022).Precisely how sensory nerves within the gut-brain axis contribute to health and disease is a major unresolved issue.Understanding how the major sensory nerve pathways in the gut-brain axis are activated and controlled within the body holds great promise for understanding a number of aspects of homeostasis, including potential neurodegenerative diseases.
Vagal afferents represent one of the two major sensory nerves in the body that provide an extensive primary afferent innervation to visceral organs (Berthoud et al., 1992(Berthoud et al., , 1997;;Mazzone & Undem, 2016).
The functional role of vagal afferents in the upper gut is generally considered to involve monitoring food and water intake and control of satiety.However, there is some evidence that vagal afferents in the upper gut also participate in nociceptive signaling (Janig et al., 2000;Yu et al., 2005).Precisely how vagal afferent nerve endings are activated by ingested luminal contents, via mechanical distension or chemical stimuli, or the microbiome remains uncertain.
Enteroendocrine cells (EEC) are specialized cells found within the GI tract (Gribble & Reimann, 2019;Latorre et al., 2016;Song et al., 2023), stomach and pancreas that release a variety of neurotransmitters and neurohormones that play a major role in body homeostasis (Bai et al., 2022;Martin, Lumsden, et al., 2017;Martin, Young, et al., 2017;Song et al., 2023;Spencer & Keating, 2022).Recent studies have suggested that direct synaptic communication exists between EEC and vagal afferent nerve endings in the GI mucosa (Kaelberer et al., 2018), where it has been proposed that glutamate released from EEC as a neurotransmitter, causing fast synaptic transmission onto the terminals of vagal afferent endings (Kaelberer et al., 2018).Similarly, it has been proposed that serotonin is released from enterochromaffin (EC) cells (a subset of EEC) which activates spinal afferent nerve endings, via fast synaptic transmission (Bellono et al., 2017).However, neither study provides morphological evidence for synapses between EEC and vagal or spinal afferent nerve endings in situ.Recently, it was found that anterograde labeling from dorsal root ganglia labeled spinal afferent nerve endings in the colonic mucosa but these nerve endings did not form close morphological associations with EC cells that would indicate synapses occur (Dodds et al., 2022).If EC cells make synaptic connections with vagal afferent terminals in the mucosa as proposed, we rationalized that anterograde labeling from nodose ganglia should label vagal afferent nerve endings in the mucosa that also form synapses with EC cells.
Here, we identified single vagal afferent axons and their nerve terminal endings in the mucosa of mouse small intestine.Then, we determined their spatial relationship with serotonergic EC cells-the most abundant type of EEC in the gut (Koo et al., 2021).The findings show no preferential close physical contacts between 5-hydroxytryptamine (5-HT)-containing EC cells and vagal afferent nerve endings in the mucosa, nor their associated varicosities along vagal afferent axons of passage.
Rather, these data suggest that the release of substances from 5-HTcontaining EC cells occurs via diffusion to neighboring vagal afferent nerve endings.

Ethical approval
All experimentation in this study was approved by the Animal Welfare

Anterograde tracer technique from nodose ganglia
Male and female C57BL/6 mice were anesthetized by isoflurane inhalation (2%-3% in 1 L/min O 2 ), and an incision 20 mm in rostro-caudal axis was made along the dorsal surface.An incision was made into the right side to expose nodose ganglion, where dextran biotin (20%; #D1956, F I G U R E 1 Schematic showing dextran biotin injection unilaterally into anesthetized mouse, to selectively label vagal afferent axons and endings in the small intestine. Molecular Probes) was injected using fine glass micropipettes (inner diameter: ∼5 µm; #TW150-4, World Precision Instruments) (Figure 1).This injection procedure was modified from previous studies injecting nodose ganglia in mice (Raab et al., 2003).
For consistency and comparison between animals, right nodose ganglia were injected with ∼200-500 nL dextran biotin via a custom-made nitrogen-driven spritz system (Biomedical Engineering; Flinders University of South Australia; Bedford Park, SA, Australia) that applied 10-15 psi nitrogen pulses to the micropipette for 1 s duration at 0.3 Hz (Figure 1).Following the injection of the ganglion, the skin was sutured closed.Animals were given a period of 10 days to recover, at which point they were euthanized by exsanguination whilst under deep inhaled isoflurane anesthesia.

Immunohistochemistry
Full-length small intestine excised from euthanized mice was cut longitudinally along the mesenteric border, then pinned mucosal side

Image acquisition and analysis
After fixation, preparations of small intestine were initially viewed with an Olympus IX71 epifluorescence microscope using appropri- Moreover, randomly chosen varicosities were also measured to nearest EC cell-varicosities were chosen at random on the vagal afferent then EC cells overlayed.

Experimental design and statistical analysis
Data are expressed as mean ± standard error of the mean.N values in the text refer to the number of animals used, and n refers to the number of observations of a given measurement.All statistics were generated using Prism (GraphPad Software, Inc.), and statistical significance was accepted for p values <.05.Distances between all vagal afferent nerve endings, or varicosities, and EC cells were compared for original and randomized images using two-way Mann-Whitney tests.

RESULTS
Of 20 mice with unilateral (right sided) dextran biotin injections to nodose ganglia, 6 mice had prominent labeling of single axons throughout whole mount preparations of small intestine: 5 of these animals showed nerve axons and endings in the smooth muscle and mucosa, while 1 animal (Figure 2a-e) had a single nerve axon that gave off multiple endings in the mucosa without detectable collaterals in the muscle.The latter gave off six discrete mucosal nerve endings in six different single villi, with no detectable axons of passage in myenteric ganglia (Figure 2b).In general, vagal afferent axons were relatively sparse across the long length of small intestine.Labeled axons were brightly filled that entered the gut wall singly but not multiply along the gut, allowing the branching patterns of single vagal afferents and their endings to be observed in high resolution in the intestine.For example, Figure 2a shows a single vagal afferent axon entering the small intestine via the mesentery that ran outside myenteric ganglia for ∼700 µm circumferentially before branching to innervate 8 individual mucosal villi (Figure 2a).A 3D reconstruction shows a side on view of the ramifications of single vagal afferent axons through the mucosa (Figure 3a-f).
This single vagal afferent axon was smooth with no varicosities for the first ∼400 µm on entering the intestine, becoming varicose more distally toward the terminal endings.In total, n = 56 vagal afferent nerve endings in the small intestinal mucosa were identified across all animals (N = 6).Dextran injections applied to the nodose ganglion outer surface without penetrating, giving opportunity to be taken up by neighboring nerves, did not give rise to detectable labeling (N = 3).This suggests labeling was specific for nodose neurons.

Characteristics of vagal afferent nerve endings in the mucosa
Three morphological types of vagal afferent endings were identified that we described as simple, branched-varicose, and spiral types.Sim- ple type endings consisted of single axons fine terminal endings that lack complex specializations, with relatively few varicosities.Spiral endings consisted of fine axons that wrapped around into a circular (spiral) formation, where they also developed increased numbers of swellings prior to the terminals that were distinct from single varicosities.The mean diameter of the swellings of spiral endings was 3.94 ± 0.24 µm (n = 19; N = 6).When measurements were taken of the diameter of final vagal afferent axon terminal in the mucosa, this was 1.2 ± 0.12 µm (n = 25; N = 5).Spiral-like vagal afferent endings developed extensive dense varicosities again of increased size along their terminal endings (Figures 4a-h5a-c, and 6a-d).Branched varicose endings consisted of axons that had more than a single axon terminal, often branching a number of times with extensive varicosities on their endings (Figure 7a-e).None of the vagal afferent endings or axons were CGRP immunoreactive.Figure 3e shows CGRP immunoreactive axons (of unknown origin) extending to the villus tip, as well as neighboring vagal afferent axons (magenta) lacking CGRP.The CGRP+ axons are likely of spinal origin, with cell bodies in dorsal root ganglia (Spencer et al., 2014).

TA B L E 1
The characteristics of diameters of vagal afferent axons and varicosities in the mucosa.

Depth of EC cells in the mucosa and relationship to vagal afferents in mucosa
In intact full thickness small intestine preparations, confocal imaging through the full depth of mucosa revealed that EC cells were distributed throughout the depth of the mucosa, ranging 50-90 µm from the apical tip of villi (mean: 67 ± 3.9 µm; n = 10; N = 5).Although single vagal afferent axon terminals extended much of the distance to the end of single villi, single EC cells were never identified that extended all the way to the villus tip (Figure 2c).The distances between EC cells and vagal afferent endings in the mucosa and varicosities along vagal afferent axons of passage are plotted in Figure 8.

DISCUSSION
Understanding the gut-brain axis requires a clear knowledge of the mechanisms by which sensory stimuli applied to the gut wall are transduced into nerve action potentials by extrinsic afferents, including those of spinal and vagal origin (Brookes et al., 2013;Furness et al., 2014;Kim et al., 2022).Here, we identified single vagal afferent nerve axons and their terminal endings that project to the mucosa of mouse small intestine and then quantified their 3D spatial relationship with EC cells that store and release large quantities of 5-HT.Anterograde labeling from nodose ganglia allowed us to selectively label vagal afferent neurons that project to the gut in fresh-fixed, full thickness wholemount preparations, without the need for antibodies or cell culturing.Our major finding suggests that there is a lack of close physical contact between 5-HT-and PYY-containing EECs and vagal afferent nerve endings in the mucosa.
Previous studies have suggested that EEC communicate via synapses with extrinsic sensory nerves (Bellono et al., 2017;Bohorquez et al., 2015).Such conclusions are supported by in vitro culturing of isolated neurons with EEC in a dish (Bohorquez et al., 2015) and by antagonists used in functional studies that suggested EEC form synapses with vagal afferents (Bohorquez et al., 2015).However, no electron microscopy or morphological data confirmed synapses existed in situ.To address this issue, we opted for anterograde labeling to identify mucosal-projecting vagal afferent endings in situ.
Some studies show cross sections of intestine to imply a close spatial relationship between EC cells and nerve axons.From our experience, sections cannot reveal any information in the Y-axis (depth); hence, it is never clear how far EC cells may lie from any nerve axons.A major advantage of the current study was the use of 3D confocal Imaris software to precisely quantify the X, Y, and Y orientations of single EC cells with respect to nearest vagal afferent nerve endings.It was immediately clear that EC cells were not closely aligned in any axis with single EC cells.On some occasions, we found vagal axons of passage that passed close to single EC cells (e.g., Figure 5c).In these planes, it was difficult to discern if the axons touched the EC cells on route to their final ending.Hence, in these cases, we rotated the images (c.f. Figure 6a-d), and it was found that axons passed over EC cells with a separation distance of about 1-3 µm, without making contact with EC cells.In no case, did we find any evidence to support a "neuropod" hypothesis, where a vagal afferent ending terminated with the basolateral border of an EC.These findings suggest that serious caution should be exercised when extrapolating data from culture dishes to intact whole intestine in situ.

Morphological differences between mucosal endings in spinal afferent and vagal afferent
Within the mucosa, we identified three different morphological types of nerve ending, consisting of simple, branched varicose, and spiraltype endings (Figure 9).Simple endings consisted of bare axons with F I G U R E 9 Cartoon representing the three morphological types of vagal afferent ending identified in the mucosa of the mouse small intestine.These endings consisted of simple, branched varicose, and spiral-type endings.
few varicosities and an absence of multiple bifurcations.The branched varicose endings consisted of single axons that bifurcate multiple times, with axon terminals that did not align parallel to neighboring axons and had extensive varicosities (Figure 7).Spiral-type endings consisted of single axons that branched toward their endings into a spiral-like formation, with multiple swellings or nodules prior to the terminals, as shown in Figures 4-6.Indeed, Powley et al. (2011) identified three types of vagal afferent nerve endings in the upper small intestine and antrum from a cohort of rats and mice.The authors described these endings as villus afferents, crypt afferents, and antral gland afferents (Powley et al., 2011).
The characteristics of vagal afferent endings we identified are similar to those reported in the Powley study (Berthoud et al., 1995;Powley et al., 2011).
A notable feature of vagal afferent axons in the small intestine is that unlike spinal afferent axons that ramified through many rows of myenteric ganglia, vagal afferent axons followed a course that ran largely outside of myenteric ganglia (Figure 2b).The vagal mucosal afferent endings identified in small intestine here were vastly different in their morphological features compared to mucosal endings of spinal afferents we identified in the mouse colon (Spencer et al., 2014).In the colon, mucosal endings were consistently lacking extensive varicosities, wrapped around the crypts, and ended in simple endings without any complex morphological specializations (Dodds et al., 2022).Where the sensory transduction sites are concentrated in these endings is not clear.In contrast, the mucosal endings of vagal afferents identified here in mouse small intestine contained multiple large varicosities within the last ∼100 µm of their axon terminal and consistently axon diameters also increased from around 1.5 to 3.5 µm where the axon formed a spiral-like ending (e.g., Figures 4a-h and 5a-c).
Hence, we focused on these cells.We did not study CCK (I cell).It is possible that one may speculate that synapses occur between EEC and afferent nerves when the endocrine hormone/transmitter is a peptide, not an amine.This seems highly improbable because EC cells that synthesize and release the 5-HT also synthesize and release peptides, like substance P (Grun et al., 2015;Haber et al., 2017).
Distances between extrinsic nerves and EEC subtypes are instead likely to differ by gut region (i.e., upper vs. lower gut), EEC or nerve type, species, and perhaps under stimulated or pathophysiological con-ditions (e.g., inflammation).For instance, in addition to the studies mentioned above, it has been shown that many nerves expressing vasoactive intestinal peptide in rat stomach closely approach (within 2 µm) ghrelin-containing EEC (Hunne et al., 2019).Early electron microscopy studies have also demonstrated distances of 150-250 nm between one EC cell and a catecholaminergic (sympathetic) motor nerve terminal in guinea pig small intestine (Lundberg et al., 1978), although there was no indication of the number of times this result was reproduced.
The minimum distance we found between a vagal afferent nerve ending and the closest 5-HT-containing EC cell in mouse colon was about 1 µm.Synapses in the nervous system of vertebrates and invertebrates occur with distances between pre-and postsynaptic membranes at conventional synaptic junctions, which is 15-25 nm (Kandel et al., 2000).This means that the closest distances we found between a spinal afferent nerve ending or varicosity and 5-HT-containing EC cell were in the order of 25 times further than accepted distances for direct synaptic transmission to take place.
Antibodies to CGRP introduce uncertainty as to the origin of the cell bodies of any axons or endings because they could be either intrinsic or extrinsic to the gut wall.Furthermore, with such dense labeling, it is impossible to follow the course of a single axon as it weaves its way down and throughout the mucosa.We circumvented these concerns in the present study using our in vivo anterograde tracer technique that allowed us to be certain we were labeling only vagal afferents, and from which single axons could be identified.It is known that EC cells have a high rate of renewal as they travel from the crypt stem cell niche upward to the villi (Gehart et al., 2019).We have previously shown that ∼50% of colonic EC cells renew every 2 weeks (Wei et al., 2021).With this high turnover rate, if EC cells did form synapses with any nerve endings, it would mean that vagal afferent endings would need to be constantly making new synapses and searching for newly formed EC cells.

Differences between EC cell-nerve communication to other parts of the nervous system
Ribbon synapses are well known to form between rods and cones as well as bipolar cells in the retina (Moser et al., 2020).Moreover, inner hair cells of the cochlea form multiple monosynaptic connections with bipolar spiral ganglion cells; see Moser et al. (2020), for review.It may have been thought that EC cells communicate with the nervous system in the gut wall in a similar manner as ribbon synapses in the eye or inner hair cell in the cochlea.None of our findings suggest any specialized synapses form between EC cells and any part of vagal afferents, at least in the small intestine of mice.Indeed, we recently showed a similar result for spinal afferent endings in the mucosa of mouse colon (Dodds et al., 2022).Similar findings were reported in Koo et al. (2021), where it was stated, We did not find specific relationships between nerve fibers and the processes of colonic 5-HT cells.Importantly, these authors noted that basal processes were not present in 5-HT-containing EC cells of the small intestine.Moreover, Berthoud and colleagues studied CCK-containing EECs and stated: Most labeled vagal afferent axons, which distributed strictly within the crypt and villous lamina propria were at distances of tens to hundreds of microns to the nearest CCK-IR cell.(Berthoud & Patterson, 1996).They concluded that their..findings strongly support the idea that CCK released from entero-endocrine cells acts on vagal sensory in a paracrine fashion (Berthoud & Patterson, 1996).
In a recent study on the mouse colon, different populations of EC cells were identified, termed..open (O) cells, with apical processes that reached the lumen, and closed (C) cells, not contacting the lumen.. (Kuramoto et al., 2021).We found that in the small intestine, the vast majority of 5-HT-containing EC cells lay ∼50 µm from the lumen.
Although we did not study the EC cell morphology in detail, the EC cells labeled in this study were analogous to the C2 and C3 type EC cells described in the colon by Kuramoto et al. (2021) that consisted of either a single process or no process, where both cell types were closed to the lumen.Basal processes in EC cells were not observed in our experiments, consistent with findings of Koo et al. (2021) in the small intestine.
Based on the low number of preparations that showed labeled vagal axons after nodose injections, one could conclude that the vagal innervation of the intestine is relatively weak.This is a reasonable assumption.It should also be noted that we only injected one nodose ganglion (right side), and very small volumes of dextran tracer were used (few hundred nanoliters).This likely contributes to why less than half of injected animals showed labeled vagal afferents.Indeed, it is possible that some axon branch points occur outside the observation areas, and that vagal afferents from the left nodose may comprise more fibers that send collaterals to different anatomical compartments.In previous studies, when we injected similar volumes of dextran biotin into dorsal root ganglia, extensive labelings of large numbers of spinal afferent axons were identified in the large intestine (Spencer et al., 2014).

CONCLUSIONS
Taken together, these findings suggest that releases of neurochemicals/neurohormones from 5-HT-containing and PYY-containing EECs are likely to activate vagal afferent endings in a paracrine fashion that does not involve fast synaptic transmission.
Committee (AWC) of Flinders University of South Australia (approval #3999), and all protocols carried out in accordance with the National Health and Medical Research Council (NHMRC) Australian code for the care and use of animal for scientific purposes (8th edition, 2013) and recommendations from the NHMRC Guidelines to promote the well-being of animals used for scientific purposes (2008).
ate discriminating filters and imaged at 4×-40× magnification with a CoolSNAP camera (Roper Scientific) via AnalySIS Image 5.0 computer software (Olympus-SIS).Preparations were imaged for quantitative analysis using a Zeiss LSM 880 confocal microscope (Carl Zeiss) with 20× oil immersion lenses (numerical aperture, nA) at 0.8.Z-stacks were scanned at 0.8 µm steps through the full thickness of preparations.Distances between mucosal vagal afferent nerve endings and EC cells were measured with Imaris x64 version 8.4.1 software (Bitplane AG).Full volume rendered images were rotated in 3D to find the shortest distances between identified vagal afferent nerve endings and varicosities and the closest surfaces of EC cells.The length between these points was calculated manually using Measurement Points; a function automatically calibrated by Imaris based on the magnification encoded within the raw image file.Distances between EC cells and the closest varicosities of mucosal vagal afferent axons were recorded.
Confocal micrograph taken from a wholemount preparation of small intestine after anterograde labeling from nodose ganglia.A single vagal afferent axon is shown with multiple spiral endings in the mucosa (see yellow asterisk).(b) Anterograde labeled axon (magenta) overlaid with calcitonin gene-related peptide (CGRP) immunoreactivity (green).The vagal axon did not follow the path of myenteric ganglia.The single axon gave off nerve endings that occupied ∼1 mm 2 .Panels (a and b) are confocal Z projection images taken through a depth of 139 µm (0.8 µm image stacks).Part (c) shows an expanded image taken from the region represented by box c in panel.Parts (b, d, and e) show the pathway of single vagal afferent axon that had entered the intestine and bifurcates in a number of branches that follow interganglionic spaces, that is, not directly traversing through myenteric ganglia.The red arrow in panel (a and b) shows the location of the single axon entering the intestine, via the mesentery.

F
I G U R E 3 (a) Confocal micrograph of an anterogradely labeled mucosal vagal afferents shown in a 3D side on perspective.These axon terminals project into single villi and arise from a single vagal afferent nerve cell bodies in nodose ganglion.The endings shown in (a) are the same endings shown in Figure 2A.Part (b) shows 5-hydroxytryptamine (5-HT) immunoreactivity and enterochromaffin (EC) cells labeled.It can be seen that EC cells are dispersed over a depth of ∼60 µm.Part (c) shows anterograde labeled vagal afferent endings in mucosa and EC cells (cyan).It is clear that the projections of mucosal endings does not communicate directly with EC cells (see arrow).(d) Calcitonin gene-related peptide (CGRP) immunoreactivity.(e) Superimposed CGRP and anterograde labeled endings.It is shown that the vagal afferent axons are not CGRP immunoreactive.Part (f) shows superimposed images of CGRP, 5-HT (EC cells), anterograde labeled endings.This confocal image was taken through a Z projection depth of 90 µm (using 0.8 µm stacks).

F
Vagal afferent mucosal endings.The image in Figure 2A is shown on expanded scale here in panel (a).The spiral-like vagal afferent mucosal endings are apparent.Part (b) shows 5-hydroxytryptamine (5-HT) (cyan) staining for enterochromaffin (EC) cells and their relationship to spiral-like endings.Boxes a-d are shown on expanded scale below.Box a in panel (b) is shown expanded in panel (c).Two spiral endings are shown in panel (c) neither of which communicate directly to EC cells.Part (d) shows the expanded region represented by the box b in panel (b).The yellow arrow shows a single varicosity along an axon of passage close to an EC cell.Part (e) shows panel (d) rotated.The yellow arrow in (e) shows a space between the varicosity shown in panel (d) and the EC cell.Hence, in panel (d), the space at the arrow is about 3-5 µm between EC cell and vagal axon.Part (f) shows the expanded region represented by the box c in panel (b).Part (g) shows the region represented by the box d in panel (b).The spiral-like ending projects away from the nearest EC cell.An axon of passage has an apparent close varicosity contact (see yellow arrow).This image in (g) is rotated in (h) to show that the axon (yellow arrow) is not touching the EC cell (see arrow).
potential areas with the closest contact between any varicosity along any vagal axons and the nearest EC cell.With the spatial limitations of confocal microscopy, we found 8 out of the 1005 EC cells where F I G U R E 5 a) Vagal afferent ending shown on expanded scale from part of Figure 2a.The bifurcation of the axon is shown where discrete spiral-like endings are observed.Part (b) shows 5-hydroxytryptamine (5-HT) stain (cyan) and the enterochromaffin (EC) cells superimposed on the vagal afferent endings.Part (c) shows and expanded region from the box shown in panel (b).The enlarged varicosities are apparent on the spiral-like ending.The yellow arrow indicates a possible contact between an EC cell and vagal afferent axon of passage.F I G U R E 6 Rotated confocal micrograph showing the spatial relationship between 5-hydroxytryptamine (5-HT)-containing enterochromaffin (EC) cells and vagal afferent endings.The image shown in the dotted box in Figure 5B is shown on expanded scale in panel (a), with rotation.An axon terminal in the vagal spiral ending appears to contact an EC cell (see yellow arrow).(b) Rotation of the image in panel (a) shows that the axon indicated by the arrow in panel (a) is not touching the EC cell (see arrow in b).Part (c) shows the same spiral ending in panel (a) rotated again showing apparent contact with an EC cell.However, this is again rotated in panel (d), where the yellow arrow shows a space of about 2 µm between the vagal afferent axon and EC cell.

F
I G U R E 7 (a) Anterogradely labeled vagal afferent axon and endings (magenta) in the mucosa and enterochromaffin (EC) cells (cyan).(a) In this animal, the spiral-like vagal afferent endings have extensive enlarged varicosities at the terminal endings.Part (b) shows an expanded image of the region highlighted by the box a in panel (a).The vagal afferent axon extends past the EC without making contact and develops enlarged varicosities in this spiral-like ending.Part (c) shows and expanded image of the region represented in the box b in panel (a).It can be seen that the vagal endings do not make targeted synapses with EC cells.Part (d) shows an expanded segment from the region represented in box a from panel (c).The enlarged varicosities are shown in the spiral-like endings.Part (e) shows a rotated imaged of panel (b), confirming that the vagal endings do not make contact with nearest EC cells.it appeared a single EC cell lay very close to a single varicosity on a vagal afferent axon of passage (Figures

F
I G U R E 8 (a) Measurements of shortest distance between vagal afferent terminals to enterochromaffin (EC) cells-containing 5-hydroxytryptamine (5-HT) and peptide-YY (PYY) and shortest distance between varicosities to EC cells that contain 5-HT or PYY.Part (b) shows the proportion of examples with the shortest distance between randomly chosen vagal afferent varicosity to nearest 5-HT-containing EC cell.Part (c) shows the proportion of examples with the shortest distance from randomly chosen vagal afferent nerve endings to nearest 5-HT-containing EC cells.