Fig 2: Representative immunostaining for CGRP and protein gene product (PGP) in the proximal (A–C) and distal (D–F) esophageal mucosa.

Fig 3: TNX expression in mouse and human neural structuresRepresentative immunohistochemical images taken from whole mounts and sections of WT mouse (A) and human distal colon (B). TNX-IR was commonly expressed in calretinin-positive neurones in the mouse myenteric plexus (MP) (A-merge) and human submucous plexus (SP) (B-Merge), while TNX- and NOS-positive cells in the MP showed few co-labelled cell bodies in both mouse and human colon (A and B-Merge). TNX-IR and CGRP expression is mutually exclusive but in close apposition whereby CGRP-positive fibres surround TNX-positive myenteric neurones in mouse and human colon (A and B-Merge). TNX specificity is confirmed by the absence of TNX in TNX-KO mouse colon while calretinin-positive neurones are present in the submucous plexus (C). TNX was not found in the mucosa while calretinin fibres are present (C). D, average number of neuronal cell bodies per field of view (fov) that co-labelled TNX plus calretinin/CGRP/NOS and ChAT in SP and MP in mouse and human colon respectively. TNX co-expressed significantly more with calretinin (55.5%) and ChAT (51.6%) than NOS in mouse MP neurones, with none in SP and only one neuron in MP. In human colon TNX-IR co-expressed significantly more with ChAT in the MP (100%) and SP (52.58) than NOS (D). CGRP did not co-label with any TNX-IR cell bodies in the SP and only one in MP in mouse and human (D). Images on the first three panels (A) were obtained on a confocal microscope and were taken from a compressed optical slice in a z stack. All other images were obtained from sections. Data shown as percentages of total neurons positive for TNX and respective markers. Statistical analysis was performed using an unpaired Student's t-test for each region, P < 0.05. Scale bar in all panels = 50 µm.

Fig 4: Neuronal sprouting in TNX‐KO mucosaRepresentative immunohistochemical images taken from sections of WT mouse colonic mucosa (A, C) and myenteric plexus (B, D). CGRP‐IR is shown in green (A) where there was a threefold proliferation of CGRP‐IR nerve endings in the colonic mucosa of TNX‐KO mice (P < 0.0001) (E). This was also observed with PGP9.5 in red (C), although this increase was smaller for PGP‐IR endings (P = 0.0224) (E). No changes were seen in either marker in the myenteric plexus (P = 0.3950, P = 0.1368) (F). Data are shown as individual values with standard deviation. Statistical analysis was performed using an unpaired Student's t test (P < 0.05). Scale bar in all panels = 20 μm.

Fig 5: TNX is expressed in neural structures in the stomach Representative immunohistochemical images taken from whole mount mouse fundus labelled with TNX and calretinin in IGLE surrounding the MP but not in cell bodies (A). TNX-immunoreactivity was commonly found in calretinin-immunoreactive fibres in the smooth muscle layer of the stomach (B, merge). The cell bodies of vagal afferents in the nodose ganglia also showed positive TNX and calretinin co-labelling in sections (C, merge). TNX positive endings found in the mouse nuclus tractus solitarius (NTS) (D). Unlike mouse, human fundus sections showed TNX- and calretinin-immunoreactive cell bodies in the myenteric plexus (MP) (E, merge) that were distinct from CGRP fibres (F, merge). TNX also co-labelled calretinin-positive endings in the circular smooth muscle (G, merge). All images from mouse and human stomach are whole mount confocal z-images except for nodose ganglia (NG) and circular muscle (CM) images, which were taken with an epifluorescence microscope. DMN, dorsal motor nucleus. Scale bars in (A–D) = 25 µm, (E–G) = 30.