Distribution and neurochemical identification of pancreatic afferents in the mouse

Abstract

Dysfunction of primary afferents innervating the pancreas has been shown to contribute to the development of painful symptoms during acute and chronic pancreatitis. To investigate the distribution and neurochemical phenotype of pancreatic afferents, Alexa Fluor-conjugated cholera toxin B (CTB) was injected into the pancreatic head (CTB-488) and tail (CTB-555) of adult male mice to label neurons retrogradely in both the dorsal root ganglia (DRG) and nodose ganglia (NG). The NG and DRG (T5-T13) were processed for fluorescent immunohistochemistry and visualized by using confocal microscopy. Spinal pancreatic afferents were observed from T5 to T13, with the greatest contribution coming from T9-T12. The pancreatic afferents were equally distributed between right and left spinal ganglia; however, the innervation from the left NG was significantly greater than from the right. For both spinal and vagal afferents there was significantly greater innervation of the pancreatic head relative to the tail. The total number of retrogradely labeled afferents in the nodose was very similar to the total number of DRG afferents. The neurochemical phenotype of DRG neurons was dominated by transient receptor potential vanilloid 1 (TRPV1)-positive neurons (75%), GDNF family receptor alpha-3 (GFRalpha3)-positive neurons (67%), and calcitonin gene-related peptide (CGRP)-positive neurons(65%) neurons. In the NG, TRPV1-, GFRalpha3-, and CGRP-positive neurons constituted only 35%, 1%, and 15% of labeled afferents, respectively. The disparity in peptide and receptor expression between pancreatic afferents in the NG and DRG suggests that even though they contribute a similar number of primary afferents to the pancreas, these two populations may differ in regard to their nociceptive properties and growth factor dependency.

Fig. 1

Pancreatic neurons in the mouse were retrogradely labeled from the head and/or tail portions of the pancreas by using Alexa Fluor-conjugated cholera toxin B (CTB). CTB labeling produced dense cytoplasmic staining among mostly small to medium-sized neurons. Neurons projecting exclusively to the pancreatic head contained only CTB-488 (green somata, A,D,G), those projecting exclusively to the pancreatic tail contained only CTB-555 (red somata, B,E,H) and those projecting to both areas of the pancreas contained both CTB-488 and 555 (yellow somata, C,F,I). J: Reconstructed montage of a mouse pancreas injected with CTB-488 in the pancreatic head (H) and CTB-555 in the tail (T). The stomach (S), pylorus (P), and duodenum (D) are also present. Scale bar = 20 μm in I (applies to A–I); 100 μm in J.

Fig. 2

Approximately the same number (A) and percentage (B)of CTB-positive neurons was observed in the nodose (NG) and dorsal root ganglia (DRG). The greatest number of spinal pancreatic neurons was observed from T9 to T12. The majority of all pancreatic afferents projected only to the pancreatic head (black bars), with similar, significantly smaller proportions projecting to the pancreatic tail (white bars) or to both the tail and head (gray bars). *, P < 0.05; **, P < 0.01; ***, P < 0.0001 vs. head only.

Fig. 3

A: The left nodose ganglia (NG) contained a significantly higher percentage of all pancreatic neurons than did the right NG; however, no differences were observed between the left and right dorsal root ganglia (DRG), regardless of spinal level. B: The same pattern held true for neurons innervating only the pancreatic head. C: The opposite was observed for neurons retrogradely labeled from the pancreatic tail, with a significantly higher percentage being observed in the right NG. There was also a significantly higher percentage of pancreatic tail neurons in the left T12 DRG than the right. D: No differences in dually projecting neurons were observed between left and right ganglia for either NG or DRG.

Fig. 4

Immunohistochemistry using antisera to TRPV1, CGRP, and GFRα3, along with IB₄ binding, was performed on nodose ganglia to identify neurochemically cholera toxin B (CTB)-positive afferents innervating the pancreas. A small percentage of pancreatic vagal neurons expressed TRPV1 (arrows; A–H) or bound IB₄ (open arrows; A–D); however, very few (<2%, not shown) showed reactivity with both markers. CGRP was also expressed by a very small percentage of pancreatic vagal neurons (arrows, I–L); however, GFRα3 was uniformly not expressed among pancreatic vagal neurons (E–L). Scale bar = 20 μm in L (applies to A–L).

Fig. 5

Immunohistochemistry using antisera to TRPV1, CGRP, and GFRα3, along with IB₄ binding, was performed on dorsal root ganglia to identify neurochemically cholera toxin B (CTB)-positive afferents innervating the pancreas. The vast majority of pancreatic spinal neurons expressed TRPV1 (arrows and arrowheads; A–H); however, very few bound IB₄ (not shown, C), and none were found to both express TRPV1 and bind IB₄ (D). CGRP was also expressed by the majority of pancreatic spinal neurons (arrows and arrowheads, I–L), as was GFRα3 (arrowheads, E–L). GFRα3 was expressed by nearly all TRPV1-positive (arrowheads, H) and CGRP-positive neurons (arrowheads, L). Scale bar = 20 μm in L (applies to A–L).

Fig. 6

Specificity of staining was tested for each antiserum. The TRPV1 antiserum labels primarily small and medium-sized DRG neurons in wild-type mice (A) but did not react with tissue sections taken from TRPV1 knockout mouse DRG (B). GFRα3 antiserum also labels many small and medium-sized DRG neurons in wild-type mice (C); however, staining was absent in DRG sections taken from GFRα3 knockout mice (D). CGRP staining is observed in many DRG neurons in wild-type mice (E), and preabsorption with 10 μM CGRP peptide abolished all antibody staining (F). Scale bar = 20 μm in B (applies to A,B), D (applies to C,D), and F (applies to E,F).

Fig. 7

The neurochemical phenotype of cholera toxin B (CTB)-positive neurons innervating the pancreatic head was evaluated in nodose (NG) and dorsal root ganglia (DRG). A,B: TRPV1 was expressed by a significantly larger percentage of DRG neurons than NG. A small percentage of NG neurons bound IB₄, and less than 2% expressed both TRPV1 and bound IB₄ (A). Only 2.5% of pancreatic DRG neurons bound IB₄, and none of these expressed TRPV1 (A). C: CGRP had a similar significantly higher expression among pancreatic DRG neurons, compared with NG. GFRα3 was also expressed by the majority of pancreatic DRG neurons and was practically absent among NG neurons (<1%, B,C). Most pancreatic DRG neurons that expressed TRPV1 or CGRP also expressed GFRα3; however, this neurochemical phenotype was not observed among pancreatic NG neurons (B,C). ND, not detected. *, P < 0.05; **, P < 0.01; ***, P < 0.0001 vs. nodose.