Perineural mast cells are specifically enriched in pancreatic neuritis and neuropathic pain in pancreatic cancer and chronic pancreatitis

Abstract

Background: Pancreatic neuritis is a histopathological hallmark of pancreatic neuropathy and correlates to abdominal neuropathic pain sensation in pancreatic adenocarcinoma (PCa) and chronic pancreatitis (CP). However, inflammatory cell subtypes that compose pancreatic neuritis and their correlation to the neuropathic pain syndrome in PCa and CP are yet unknown.

Methods: Inflammatory cells within pancreatic neuritis lesions of patients with PCa (n = 20) and CP (n = 20) were immunolabeled and colorimetrically quantified with the pan-leukocyte marker CD45, with CD68 (macrophages), CD8 (cytotoxic T-lymphocytes), CD4 (T-helper cells), CD20 (B-lymphocytes), NCL-PC (plasma cells), neutrophil elastase, PRG2 (eosinophils), anti-mast cell (MC) tryptase and correlated to pain sensation. Perineural mast cell subtypes were analyzed by double immunolabeling with MC chymase. Expression and neural immunoreactivity of protease-activated receptor type 1 (PAR-1) and type 2 (PAR-2) were analyzed in PCa and CP and correlated to pain status of the patients.

Results: In PCa and CP, nerves were predominantly infiltrated by cytotoxic T-lymphocytes (PCa: 35% of all perineural inflammatory cells, CP: 33%), macrophages (PCa: 39%, CP: 33%) and MC (PCa: 21%, CP: 27%). In both entities, neuropathic pain sensation was associated with a specific increase of perineural MC (PCa without pain: 14% vs. PCa with pain: 31%; CP without pain: 19% vs. CP with pain: 34%), not affecting the frequency of other inflammatory cell subtypes. The vast majority of these MC contained MC chymase. PAR-1 and PAR-2 expression did not correlate to the pain sensation of PCa and CP patients.

Conclusion: Pancreatic neuritis in PC and CP is composed of cytotoxic T-lymphocytes, macrophages and MC. The specific enrichment of MC around intrapancreatic nerves in neuropathic pain due to PCa and CP suggests the presence of MC-induced visceral hypersensitivity in the pancreas. Therefore, pancreatic and enteric neuropathies seem to share a similar type of neuro-immune interaction in the generation of visceral pain.

Figure 1. Inflammatory cell subsets composing pancreatic neuritis in pancreatic adenocarcinoma (PCa).

(A) Pancreatic neuritis was identified by means of a hematoxylin-counterstained section which was immunolabeled with the pan-neuronal marker PGP9.5 or by means of a hematoyxlin-eosin (H&E)-stained section. Inflammatory cells composing pancreatic neuritis were identified by means of immunolabeling with the pan-leukocyte marker CD45, with CD68 (macrophages), CD8 (cytotoxic T-lymphocytes), CD4 (T-helper cells), CD20 (B-lymphocytes), NCL-PC (plasma cells), neutrophil elastase, PRG2 (eosinophils), anti-mast cell tryptase (MC-T) and correlated to pain sensation. (B–C) In PCa, CD68+ macrophages, CD8⁺ cytotoxic T lymphocytes and mast cells (MC) dominate the neural inflammatory cell population. “N” stands for the identified nerve. Between three to five nerves with pancreatic neuritis were analyzed from each patient. All images at 200x magnification. Scale bars indicate 50 µm.

Figure 2. Inflammatory cell subsets composing pancreatic neuritis in chronic pancreatitis (CP).

In chronic pancreatitis, pancreatic neuritis showed a similar composition to pancreatic cancer, comprising CD68+ macrophages, CD8⁺ cytotoxic T lymphocytes and mast cells (MC). “N” stands for the identified nerve. Between three to five nerves with pancreatic neuritis were analyzed from each patient. All images at 200x magnification. Scale bars indicate 50 µm.

Figure 3. Impact of neuropathic pain upon the composition of pancreatic neuritis in pancreatic adenocarcinoma (PCa).

Pain status of PCa patients did not affect the relative distribution of the majority of perineural inflammatory cell subsets in PCa, but it was only mast cells which were specifically enriched around intrapancreatic nerves of PCa patients with pain when compared to patients with no pain. “N” stands for the identified nerve. Between three to five nerves with pancreatic neuritis were analyzed from each patient. All images at 200x magnification. Scale bars indicate 50 µm.

Figure 4. Impact of neuropathic pain upon the composition of pancreatic neuritis in chronic pancreatitis (CP).

Similar to PCa, also in CP, patients with pain demonstrated a greater relative amount of mast cells (MC) around intrapancreatic nerves than patients without pain. Furthermore, this increase in the amount of perineural MC was paralleled by decreased amounts of cytotoxic CD8⁺ T lymphocytes around intrapancreatic nerves. “N” stands for the identified nerve. Between three to five nerves with pancreatic neuritis were analyzed from each patient. All images at 200x magnification. Scale bars indicate 50 µm.

Figure 5. Neural invasion in pancreatic cancer (PCa) and perineural mast cell infiltration.

Intrapancreatic nerves in PCa tissues were evaluated for the presence of neural invasion (NI) and classified into the NI degrees “0 / no invasion”, “1 / peri-neural invasion” and “2 / endo-neural invasion). In our cohort, the relative perineural mast cell amounts did not differ between the varying degrees of NI in human PCa. Results are presented as median (minimum; maximum).

Figure 6. Protease-activated-receptor (PAR) type 1 (PAR-1) and type 2 (PAR-2) in human pancreatic cancer (PCa) and chronic pancreatitis (CP).

(A) Expression of PAR-1 and PAR-2 was compared between normal human pancreas (NP), CP and PCa tissues via qRT-PCR and did not differ between these three entities. Expression was normalized first to the housekeeping gene cyclophilin B and then to NP. (B) In PCa, the tissue levels of PAR-1 and PAR-2 did not differ between patients with pain versus patients without pain. (C) Similarly, also in CP, there was no difference in the tissue mRNA levels of PAR-1 and PAR-2 in patients with no pain versus with pain. (D) Intrapancreatic nerves in PCa were analyzed for the immunoreactivity for PAR-1 and PAR-2 and correlated to the pain status of patients. Here, pain sensation was not associated with differences in the immunoreactivity of intrapancreatic nerves for PAR-1 or PAR-2. (E) In analogy with PCa, also in CP, patients with pain exhibited similar immunoreactivities in nerves for PAR-1 and PAR-2 as patients without pain.

Figure 7. Analysis of mast cell (MC) phenotype in NP, PCa and CP.

(A) Human pancreatic tissue samples from NP, PCa and CP were double-immunolabeled for perineural MC-tryptase (red) and MC-chymase (green). In all three entities, the vast majority of perineural MC demonstrated double immunoreactivity (yellow in overlay) for MC-tryptase and -chymase. In CP, there were significantly greater relative amounts of double-immunoreactive MC among patients with pain than among those without pain. The white scale bars indicate 100 µm. (B) In accordance with increased perineural MC-tryptase in painful PCa and CP, the perineural MC chymase immunoreactivity was significantly greater among PCa and CP patients with pain than in patients without pain. Results are expressed as median (Minimum; Maximum).