Perineural invasion is a common feature of pancreatic ductal adenocarcinoma (PDAC). Here, we investigated the effect of perineural invasion on the microenvironment and how this affects PDAC progression. Transcriptome expression profiles of PDAC tissues with different perineural invasion status were compared, and the intratumoral T-cell density and levels of neurotransmitters in these tissues were assessed. Perineural invasion was associated with impaired immune responses characterized by decreased CD8+ T and Th1 cells, and increased Th2 cells. Acetylcholine levels were elevated in severe perineural invasion. Acetylcholine impaired the ability of PDAC cells to recruit CD8+ T cells via HDAC1-mediated suppression of CCL5. Moreover, acetylcholine directly inhibited IFNγ production by CD8+ T cells in a dose-dependent manner and favored Th2 over Th1 differentiation. Furthermore, hyperactivation of cholinergic signaling enhanced tumor growth by suppressing the intratumoral T-cell response in an orthotopic PDAC model. Conversely, blocking perineural invasion with bilateral subdiaphragmatic vagotomy in tumor-bearing mice was associated with an increase in CD8+ T cells, an elevated Th1/Th2 ratio, and improved survival. In conclusion, perineural invasion–triggered cholinergic signaling favors tumor growth by promoting an immune-suppressive microenvironment characterized by impaired CD8+ T-cell infiltration and a reduced Th1/Th2 ratio.

Significance:

These findings provide a promising therapeutic strategy to modulate the immunosuppressive microenvironment of pancreatic ductal adenocarcinoma with severe perineural invasion.

Pancreatic ductal adenocarcinoma (PDAC) remains one of the most malignant tumors, and it is predicted to become the second leading cause of cancer-related death by 2030 (1). Nearly 80% of the patients with PDAC are diagnosed at a locally advanced or distant stage, which eliminates the chance of radical surgery. The progression of systemic treatment has contributed a limited increase in survival, but the 5-year survival rate has remained at 8% (2). While the immunotherapies have demonstrated promising outlooks in many solid tumors (3, 4), the application of such therapies to PDAC has been frustrating (3, 5, 6). Therefore, new insights into the biology and genetics of PDAC are urgently needed to identify therapeutic targets.

Perineural invasion is a characteristic feature of PDAC that contributes to local recurrence and a poor prognosis for patients. Although perineural invasion has been reported in a variety of cancers (7), the incidence of perineural invasion remains the highest in PDAC, ranging from 80% to 100% (8). Invasion of the surrounding nerves by cancer cells not only provides a route for metastasis but also promotes tumorigenesis and development via the cross-talk between neuron cells and tumor cells (9–14).

To achieve this goal, we analyzed the gene expression alterations associated with perineural invasion status in tumor tissue and we found that perineural invasion was associated with impaired immune responses. Growing evidences suggest that the peripheral nervous system profoundly alters the immune response in both inflammatory diseases and cancers (15). Effected autonomic neuron signaling, especially cholinergic signaling, could regulate lymphocyte trafficking (16), cytokine production (17), and immune cell differentiation (18) via neurotransmitters. Among the neurotransmitters of peripheral nerve system, acetylcholine is the best-characterized anti-inflammatory modulator (17). Here, we demonstrated that increased level of acetylcholine could impair PDAC cells to recruit CD8+ T cell and directly inhibit IFNγ production of CD8+ T cells and favor Th2 over Th1 differentiation. Taken together, we for the first time revealed that perineural invasion contribute to the immune-suppressive microenvironment in PDAC.

Patients and tissues

Pancreatic tissue samples were obtained from patients diagnosed with PDAC following tumor resection at Renji Hospital (Shanghai, P.R. China). The clinicopathologic characteristics of Renji cohort 1 (n = 50) were deposited in the Gene Expression Omnibus (GEO, GSE102238) and the characteristics of Renji cohort 2 (n = 311) are available in Supplementary Table S1. Serum samples from healthy subjects were also obtained from Renji Hospital (Shanghai, P.R. China). Our study was approved by the Research Ethics Committee of Renji Hospital, School of Medicine, Shanghai Jiao Tong University (Shanghai, P.R. China) with the approved protocol number RA-2019-116. Written informed consent was provided by all the patients before enrollment. Follow-up information is shown in the Supplementary Materials and Methods.

Definition and severity score of perineural invasion

Perineural invasion is defined as cancer cells present along nerves (cells circumscribed at least 33% of the nerve) and/or within the epineural, perineurial, and endoneurial space of the neuronal sheath. The severity of perineural invasion was determined on the basis of the degree and frequency of neural cancer cell invasion as previously reported (19). The degree of perineural invasion was scored as absent/0, perineural/1, or intraneural/2 and the frequency of perineural invasion was scored as absent/0, low/1, frequent/2, and excessive/3. The severity score was the product of the degree score and the frequency score of perineural invasion, and severe perineural invasion was defined as a severity score of 4–6. Tissues with severe perineural invasion were classified into “with perineural invasion group” and others were classified into “without perineural invasion group.”

Animal model studies

All animal experiments were undertaken in accordance with the NIH Guide for the Care and Use of Laboratory Animals. All manipulations were performed under an approved protocol number 20141204 assigned by the Research Ethics Committee of East China Normal University. Animals that had a weight loss exceeding 20% of the first recorded body weight were removed from this study.

Pancreatic cancer orthotopic model and treatments

C57BL/6 and BALB/cA-nu/nu male mice that were 6 to 8 weeks old were used in this study. To establish an orthotopic pancreatic cancer model, 2 × 106 luciferase-expressing FC1199 or Panc02 cells suspended in 100 μL of DMEM media were mixed with 5% Matrigel and were injected into the body of the pancreas. In the drug-treated group, 20 mice were randomly divided into four groups. One group received carbacholine (1 mg/kg body weight), another group received atropine (1 mg/kg body weight) and carbacholine (1 mg/kg body weight), third group received mecamylamine hydrochloride (1 mg/kg body weight) and carbacholine (1 mg/kg body weight), and control group was treated with DMSO, all the reagents were dissolved in DMSO and administrated intraperitoneally. The drugs were intraperitoneally injected every 2 days and the tumor volume was monitored by the IVIS Spectrum (Caliper Life Sciences) every 5 days. In the surgical treatment group, 20 mice were randomly divided into two groups. One group received bilateral subdiaphragmatic vagotomy, pyloroplasty and simultaneous tumor cell injection. The other group received pyloroplasty and simultaneous tumor cell injection as sham operation. After surgery, the tumor volume was monitored by the bioluminescence IVIS imaging every 7 days until mice death or the 28 day after surgery. The emission images were quantified by living image software, version 4.5.3.

Pancreatic cancer transgenic model and vagotomy operation

LSL-Kras+/G12D; LSL-Trp53+/R172H; Pdx1-Cre (KPC) mice were characterized by spontaneously developed PDAC (20). LSL-Kras+/G12D; Pdx1-Cre (KC) mice and LSL-Kras+/G12D; LSL-Trp53+/R172H; Pdx1-Cre (PC) mice were crossed to generate KPC mice, which were then confirmed by nucleic acid electrophoresis. Pyloroplasty with or without bilateral subdiaphragmatic vagotomy was performed on KPC mice at 3 months of age. Mice were sacrificed 4 weeks after surgery, and the pancreas of each was removed and fixed for IHC staining.

Cell culture and reagents

Human PDAC cell lines Mia Paca-2 and BxPC-3 and murine PDAC cell lines FC1199 and Panc02 were all from Shanghai Cancer Institute, Renji Hospital, School of Medicine, Shanghai Jiao Tong University (Shanghai, P.R. China) in 2018. All cell lines used in this study were authenticated by short tandem repeat analysis. The length of time between cell line thawing and use in experiment did not exceed 4 weeks (two or more passages). The details are provided in the Supplementary Materials and Methods.

High-performance liquid chromatography

Human PDAC frozen tissue was weighted and homogenized on ice. The homogenized tissue (50 μL) was vortexed with 4% perchloric acid solution (100 μL) then centrifugated for 10 minutes (4°C, 9,600 × g). The supernatants were collected for high-performance liquid chromatography (HPLC) analysis. Acetylcholine and norepinephrine were measured using Shimadzu Nexera UHPLC (Shimadzu Europa GmbH) coupled to Triple Quad 5500 Mass Spectrometer (AB Sciex). ACE C18-PFP column (Hichrom Ltd.) was used for fractionation. Ions in the mass spectrometer system were acquired in the positive mode using a Turbo V ESI interface and either MRM.

Accession number

The gene microarray data have been deposited in the Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/) under the accession code GEO: GSE102238.

Statistical analysis

Statistical analysis in this study was performed by SPSS 16.0 (IBM). Survival analysis was calculated by the Kaplan–Meier method and analyzed by the log-rank test. The correlation between CD8A+ T-cell density and clinical parameters was examined by Fisher exact test. Univariate and multivariate Cox regression analyses were performed using the Cox proportional hazards model. Student t tests or one-way ANOVA were used for comparisons between groups. P < 0.05 was defined as statistically significant. Values are expressed as the means ± SEM or means ± SD.

Perineural invasion was associated with an immune-suppressive microenvironment in PDAC

To investigate the influence of perineural invasion on PDAC and its microenvironment, we employed a gene microarray to analyze the gene expression alterations associated with perineural invasion status in tumor tissue with a complete microenvironment. Patients with PDAC from Renji cohort 1 (n = 50) were divided into two groups based on the severity and frequency of neural invasion as reported in a previous study (19); 56% of the patients (n = 28) were diagnosed with severe perineural invasion (with perineural invasion group), and the others (n = 22) were in the “without perineural invasion group.” Then, 1,172 differentially expressed genes were identified by gene expression profiling (P < 0.05; Supplementary Table S2). Gene set enrichment analysis (GSEA) was performed to identify the biological process associated with perineural invasion. Consistent with the pathology, gene sets related to nervous system development and axonogenesis were positively correlated with severe perineural invasion status. Interestingly, gene sets negatively correlated with perineural invasion status were mostly associated with immune system activity, including immune cell chemotaxis and response to immune processes such as response to bacterium and IFNγ (Fig. 1A and B; Supplementary Fig. S1).

Figure 1.

Perineural invasion is negatively correlated with gene sets associated with immune response. A, GSEA based on the gene expression profiles of PDAC tissues with severe perineural invasion compared with that without severe perineural invasion. Gene ontology (GO) terms of biological process database (c5.bp.v4.0) that positively or negatively associated with perineural invasion status are listed according to the normalized enrichment score (NES). B, Representative GSEA plots positively or negatively associated with perineural invasion. Nom. P, normalized P value. C, The correlation between the expression of signature genes of cytotoxic T cell, Th2 cell, and the status of perineural invasion. The blue dots represent the genes negatively associated with perineural invasion status, while the red dots represent the positive correlation and the shade of color represents the fold change of gene differently expressed in two groups. The size of the dots represents the significance in gene differential expression. The data are analyzed from GEO database GSE102238 and are demonstrated by Cytoscape.

Figure 1.

Perineural invasion is negatively correlated with gene sets associated with immune response. A, GSEA based on the gene expression profiles of PDAC tissues with severe perineural invasion compared with that without severe perineural invasion. Gene ontology (GO) terms of biological process database (c5.bp.v4.0) that positively or negatively associated with perineural invasion status are listed according to the normalized enrichment score (NES). B, Representative GSEA plots positively or negatively associated with perineural invasion. Nom. P, normalized P value. C, The correlation between the expression of signature genes of cytotoxic T cell, Th2 cell, and the status of perineural invasion. The blue dots represent the genes negatively associated with perineural invasion status, while the red dots represent the positive correlation and the shade of color represents the fold change of gene differently expressed in two groups. The size of the dots represents the significance in gene differential expression. The data are analyzed from GEO database GSE102238 and are demonstrated by Cytoscape.

Close modal

Because intratumoral immune cells are heterogeneous, we adopted the immune landscape analysis to determine which type of the immune cells might be the most affected at the transcriptome level within Renji cohort 1 (21). We found that the signature genes of cytotoxic cells were remarkably negatively correlated with perineural invasion, while and the Th2 signature genes were positively correlated with perineural invasion (Fig. 1C). We further confirmed the relationship between perineural invasion status and the density of intratumoral cells [CD8+ T cells, Th1 cells, Th2 cells, and regulatory T (Treg) cells], by IHC staining of samples from Renji cohort 2. The results showed that the density of CD8+ T cells and Th1 cells were dramatically decreased, and the density of Th2 cells concomitantly increased in PDAC tissue with severe perineural invasion (Fig. 2A). There was no significant relationship between perineural invasion status and the density of Treg cells (Fig. 2B). Furthermore, we also performed IHC staining of CD68 and CD163 in tissue microarray to evaluate the relationship between TAM and perineural invasion in PDAC. And there is no significant correlation between the perineural invasion status and the density of CD68+ cells and CD163+ cells (Supplementary Table S3). In-line with previous reports (22), low infiltration of CD8+ T cells is associated with severe T-cell classification and shorter overall survival (Supplementary Table S1; Supplementary Fig. S2).

Figure 2.

The density of intratumoral T cells varied dependent on the perineural invasion status. A, Representative image of IHC staining for nerve (PGP9.5), CD8+ T cell (CD8A), Th1 cell (T-bet), and Th2 cell (GATA3) in serial sections of PDAC tissue diagnosed with or without severe perineural invasion (PNI). B, The proportion of high or low density of CD8+ T cell, Th1 cell, Th2 cell, and Treg cell in PDAC TMA (n = 311). Fisher exact test was performed in the analysis of statistical significance.

Figure 2.

The density of intratumoral T cells varied dependent on the perineural invasion status. A, Representative image of IHC staining for nerve (PGP9.5), CD8+ T cell (CD8A), Th1 cell (T-bet), and Th2 cell (GATA3) in serial sections of PDAC tissue diagnosed with or without severe perineural invasion (PNI). B, The proportion of high or low density of CD8+ T cell, Th1 cell, Th2 cell, and Treg cell in PDAC TMA (n = 311). Fisher exact test was performed in the analysis of statistical significance.

Close modal

Increased acetylcholine impaired tumor cell recruitment of CD8+ T cells by reducing tumor-derived CCL5 production, which was dependent on HDAC1-mediated histone deacetylation

It has been reported that the peripheral nervous system can change the immune response via neurotransmitters in both inflammatory diseases and cancers (15). We wondered whether perineural invasion altered neurotransmitter secretion and resulted in immune system dysfunction. To test this hypothesis, we performed HPLC to detect the concentrations of norepinephrine and acetylcholine, the neurotransmitters for the sympathetic and parasympathetic nervous systems. As shown in Fig. 3A, the concentration of acetylcholine was remarkably increased in tissue with severe perineural invasion (328.2 ± 71.18 pg/mL vs. 94.15 ± 24.92 pg/mL; P = 0.0043). There was a trend of increased norepinephrine in the perineural invasion group, but the variation was not statistically significant (621 ± 136.3 pg/mL vs. 398.7 ± 85.2 pg/mL; P = 0.1774).

Figure 3.

Increased acetylcholine impairs the ability of tumor cells to recruit CD8+ T cells by reducing CCL5 secretion via HDAC1-mediated histone deacetylation. A, The concentration of acetylcholine (Ach) and norepinephrine (NE) in the PDAC tissue diagnosed with (n = 15) or without (n = 15) severe perineural invasion (PNI) by HPLC. B, The chemotaxis assay using the DMEM medium and supernatant of pancreatic cancer cell line (MiaPaca-2) pretreated with acetylcholine (10 μmol/L) or in combination with atropine (1 μmol/L) or mecamylamine (MEC, 1 μmol/L). C, The chemotaxis assay using the DMEM medium and supernatant of isolated PSCs pretreated with acetylcholine (10 μmol/L) or in combination with atropine (1 μmol/L) or mecamylamine (1 μmol/L). D, The correlation between the expression of CD8+ T-cell signature genes CD8A and the expression of chemokines based on the data in Renji cohort 1 and TCGA PAAD database. E, The in situ hybridization shows colocalization of CCL5 and CK19 but not αSMA in human PDAC tissue at mRNA level. F, Immunofluorescence staining for αSMA/CCL5 and CK19/CCL5 shows colocalization of CCL5 in the pancreatic duct cells but not stroma cell. G, The expression of CCL5 in human PDAC cell lines treated with acetylcholine (10 μmol/L) and/or the antagonist of AChRs (atropine, 1 μmol/L; mecamylamine, 1 μmol/L) detected by qRT-PCR (left) and ELISA (right). Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test). H, The expression of genes at mRNA level in human PDAC cell lines (Mia Paca-2) treated with nicotine (10 μmol/L) or DMSO detected by qRT-PCR. Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test). I, The expression of genes at mRNA level in human PDAC cell lines treated with nicotine (10 μmol/L) or nicotine in combination with EPZ-6438 (1 μmol/L) or romidepsin (1 μmol/L) detected by qRT-PCR. Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test). J, Expression of HDAC1, H4Ac, H3Ac, H4, and H3 at protein level in human PDAC cell lines treated with nicotine (10 μmol/L) or nicotine with romidepsin (1 μmol/L). K, ChIP assays using the antibody target H3ac and H4ac performed on human PDAC cell lines. Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, no significance).

Figure 3.

Increased acetylcholine impairs the ability of tumor cells to recruit CD8+ T cells by reducing CCL5 secretion via HDAC1-mediated histone deacetylation. A, The concentration of acetylcholine (Ach) and norepinephrine (NE) in the PDAC tissue diagnosed with (n = 15) or without (n = 15) severe perineural invasion (PNI) by HPLC. B, The chemotaxis assay using the DMEM medium and supernatant of pancreatic cancer cell line (MiaPaca-2) pretreated with acetylcholine (10 μmol/L) or in combination with atropine (1 μmol/L) or mecamylamine (MEC, 1 μmol/L). C, The chemotaxis assay using the DMEM medium and supernatant of isolated PSCs pretreated with acetylcholine (10 μmol/L) or in combination with atropine (1 μmol/L) or mecamylamine (1 μmol/L). D, The correlation between the expression of CD8+ T-cell signature genes CD8A and the expression of chemokines based on the data in Renji cohort 1 and TCGA PAAD database. E, The in situ hybridization shows colocalization of CCL5 and CK19 but not αSMA in human PDAC tissue at mRNA level. F, Immunofluorescence staining for αSMA/CCL5 and CK19/CCL5 shows colocalization of CCL5 in the pancreatic duct cells but not stroma cell. G, The expression of CCL5 in human PDAC cell lines treated with acetylcholine (10 μmol/L) and/or the antagonist of AChRs (atropine, 1 μmol/L; mecamylamine, 1 μmol/L) detected by qRT-PCR (left) and ELISA (right). Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test). H, The expression of genes at mRNA level in human PDAC cell lines (Mia Paca-2) treated with nicotine (10 μmol/L) or DMSO detected by qRT-PCR. Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test). I, The expression of genes at mRNA level in human PDAC cell lines treated with nicotine (10 μmol/L) or nicotine in combination with EPZ-6438 (1 μmol/L) or romidepsin (1 μmol/L) detected by qRT-PCR. Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test). J, Expression of HDAC1, H4Ac, H3Ac, H4, and H3 at protein level in human PDAC cell lines treated with nicotine (10 μmol/L) or nicotine with romidepsin (1 μmol/L). K, ChIP assays using the antibody target H3ac and H4ac performed on human PDAC cell lines. Bar graphs presented as mean ± SD (n = 3; unpaired two-tailed Student t test; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, no significance).

Close modal

Because both pancreatic tumor cells (23) and pancreatic stellate cells (PSC; ref. 24) can recruit immune cells, the supernatant of pancreatic cancer cells and primary human PSCs were used to test the effects of acetylcholine on the recruitment of CD8+ T cells. The gating strategy is illustrated in Supplementary Fig. S3A. CD8+ T cells showed a strong chemotactic response toward cancer cell that could be inhibited by the pretreatment with acetylcholine. We also found that mecamylamine, an antagonist of nicotinic acetylcholine receptor (nAChR), could abolish the suppressive effect of acetylcholine. While atropine, the antagonist of muscarinic acetylcholine receptor (mAChR), did not show similar effects (Fig. 3B; Supplementary Fig. S3A and S3B). The chemotactic response of CD8+ T cells to PSCs was relatively weaker, and pretreatment of PSCs with acetylcholine had no impact on recruitment (Fig. 3C). Furthermore, the direct effect of acetylcholine on the chemotaxis of CD8+ T cells was excluded (Supplementary Fig. S3C).

The spatial distribution of intratumoral CD8+ T cells is controlled by the following chemokines: CCL2, CCL3, CCL4, CCL5, CXCL9, and CXCL10 (25, 26). We performed correlation analysis between CD8A and those chemokines using both Renji cohort 1 and The Cancer Genome Atlas (TCGA) PAAD database, as shown in Fig. 3D, CCL5 is the most relevant chemokine in terms of relation to CD8A (R2 = 0.7943 in the Renji cohort and R2 = 0.7961 in the TCGA database). In addition, we verified that rhCCL5 directly promoted CD8+ T-cell migration in vitro (Supplementary Fig. S3D). Because CCL5 could be produced by various cells in the tumor microenvironment (TME), RNAscope in situ hybridization analysis was performed with CCL5, KRT19, and ACTA2 probes. The results revealed that CCL5 mRNA accumulated in the KRT19-positive tumor cells, but not in ACTA2-positive stroma cells (Fig. 3E), which was consistent with the results of immunofluorescence staining results (Fig. 3F).

Furthermore, acetylcholine obviously reduced CCL5 expression in cancer cells at both the mRNA and protein levels (Fig. 3G). To provide functional evidence, a CCL5 neutralizing antibody was used in chemotaxis assays. Treatment with the CCL5 neutralizing antibody as well as with nicotine abolished CD8+ T-cell migration. Nicotine could no longer suppress CD8+ T-cell migration in the presence of CCL5-neutralizing antibody (Supplementary Fig. S3E). Thus, acetylcholine impaired tumor cells recruitment of CD8+ T cells by reducing CCL5 expression.

It has been reported that allogeneic fetuses and placentas avoid rejection by epigenetic silencing chemokine genes during pregnancy (27). We wondered whether epigenetic regulation contributed to the immune escape in PDAC. We found that nicotine specifically increased the expression of histone deacetylase 1 (HDAC1), which is one of the key epigenetic modulating enzymes (Fig. 3H). Then, we confirmed that the HDAC inhibitor romidepsin, but not the EZH2 inhibitor EPZ-6438, could rescue the downregulation of CCL5 induced by nAChR activation (Fig. 3I). Moreover, nicotine decreased the level of acetylated histone H3 (H3Ac) and H4 (H4Ac), while romidepsin rescued the nicotine-induced downregulation of H3Ac and H4Ac in cancer cells (Fig. 3J). Furthermore, we confirmed that the H3Ac and H4Ac occupancy of the promoter region of CCL5 using chromatin immunoprecipitation (ChIP) assays (Fig. 3K). The above results demonstrate that acetylcholine and nicotine increased HDAC1 expression via nAChR. Increased HDAC1 led to the decreased levels of H3Ac and H4Ac, which suppressed CCL5 transcription.

Acetylcholine inhibited IFNγ production in CD8+ T cells and promoted Th1 to Th2 switching via nAChRs in a dose-dependent manner

Because it has been reported that nicotine tips the Th1/Th2 balance toward a Th2-dominant phenotype in many inflammatory diseases (28–30), we assessed the direct role of acetylcholine in activated T cells by treating them with carbacholine, a stable analogue of acetylcholine. Consistent with previous reports, we found that carbacholine markedly impaired the IFNγ production by CD8+ T cells in a dose-dependent manner (Fig. 4A). In addition, carbacholine treatment significantly attenuated the proportion of IFNγ+ CD4+ T cells (Th1) and increased the proportion of IL4+ CD4+ T cells (Th2) proportion in a dose-dependent manner (Fig. 4B). We also found that only mecamylamine could rescue carbacholine-induced IFNγ reduction of CD8+ T cells and Th1 to Th2 switching, indicating that this effect was mediated by nAChRs (Fig. 4C and D).

Figure 4.

Acetylcholine decreased IFNγ production of CD8+ T cell and promoted Th1 to Th2 switch through nAChRs. A and B, Frequency of IFNγ+ CD8+ T cell, IFNγ+ CD4+ T (Th1) cell, and IL4+ CD4+ T (Th2) cell in the presence of carbachol (1–10 μmol/L). C and D, Frequency of IFNγ+ CD8+ T cell, IFNγ+ CD4+ T (Th1) cell, and IL4+ CD4+ T (Th2) cell in the presence of carbachol (10 μmol/L) or carbachol in combination with atropine (1 μmol/L) or mecamylamine (MEC, 1 μmol/L). Bar graphs are presented as mean ± SD (n = 3; unpaired two-tailed Student t test; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, no significance).

Figure 4.

Acetylcholine decreased IFNγ production of CD8+ T cell and promoted Th1 to Th2 switch through nAChRs. A and B, Frequency of IFNγ+ CD8+ T cell, IFNγ+ CD4+ T (Th1) cell, and IL4+ CD4+ T (Th2) cell in the presence of carbachol (1–10 μmol/L). C and D, Frequency of IFNγ+ CD8+ T cell, IFNγ+ CD4+ T (Th1) cell, and IL4+ CD4+ T (Th2) cell in the presence of carbachol (10 μmol/L) or carbachol in combination with atropine (1 μmol/L) or mecamylamine (MEC, 1 μmol/L). Bar graphs are presented as mean ± SD (n = 3; unpaired two-tailed Student t test; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, no significance).

Close modal

Cholinergic signaling promoted an immune-suppressive microenvironment to favor tumor growth in vivo

To verify the immune-suppressive role of cholinergic signaling in vivo, an orthotopic pancreatic cancer model was used. As shown in Fig. 5A, luciferase-transfected FC1199 cells were injected into the pancreas of C57BL/6 mice at 7 weeks of age; then, 1 week later, the mice were randomized into four groups, which received different treatments. On week 10, the bioluminescence IVIS imaging revealed that carbacholine treatment significantly increased tumor volume, while the tumor volumes in the carbacholine combination with mecamylamine group were almost the same as those of the control group (Fig. 5B and C). Then, the mice were sacrificed, and tumor-infiltrating immune cells were isolated for flow cytometry analysis. The gating strategies for CD4+ and CD8+ T cells are shown in Fig. 5D. We found that carbacholine significantly reduced the density of intratumoral CD4+ and CD8+ T cells, which could be abolished by mecamylamine (Fig. 5E). We also noticed that carbacholine reduced the proportion of intratumor IFNγ+ CD8+ T cells and Th1 cells, it increased the proportion of Th2 cells (Fig. 5F and G). Moreover, we also determined the proportions of Th17 cell, neutrophils, and monocytes/macrophages in the tumor tissue. We found that carbacholine also increased the proportion of Th17 cells via nAChRs but did not change the proportion of neutrophils and monocytes/macrophages (Supplementary Fig. S4A). In addition, we concurrently detected the proportion of IFNγ+ CD8+ T cells, Th1 cells, Th2 cells, Th17 cells, neutrophils, and monocytes/macrophages in the spleen of those mice. The results suggested that the immune-suppressive role of cholinergic signaling was limited to tumor tissue and did not influence the systemic immune response (Supplementary Fig. S4B).

Figure 5.

Cholinergic signaling reprogramed immune microenvironment to favor tumor growth via nicotinic receptor in vivo. A, Luciferase-expressing FC1199 or Panc02 cells (2 × 106) were injected into the body of pancreas of C57BL/6 or BALB/cA-nu/nu mice (n = 20) at age of 7 weeks. At age of 9 weeks, 20 mice were randomly divided into four groups and, respectively, received intraperitoneally injected triweekly as follows, carbacholine (1 mg/kg body weight), atropine (1 mg/kg body weight), and carbacholine (1 mg/kg body weight), mecamylamine hydrochloride (1 mg/kg body weight) and carbacholine (1 mg/kg body weight), and control group was treated with DMSO. Mice were euthanized at age of 10 weeks. B, Representative image of IVIS spectrum of mice at age of 9 weeks. C, The quantification of emission images for mice at age of 9 weeks. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). D, The gate strategies of counting beads, CD4+ T cells, and CD8+ T cells. E, The relative quantity of CD4+ T cells and CD8+ T cells in the tumor tissue normalized to tissue weight. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). F and G, The proportion of IFNγ+ CD8+ T cell, IFNγ+ CD4+ T (Th1) cell, and IL4+ CD4+ T (Th2) cell in different groups. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). H, Representative image of IVIS spectrum of C57BL/6 mice at age of 8 and 10 weeks and the quantification of emission images. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). I, The density of CD8A+ T cells, T-bet+ T cells (Th1), and GATA3+ T cells (Th2) was determined by IHC in the tumor lesion. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). J, Representative image of IVIS spectrum of BALB/cA-nu/nu mice at age of 8 and 10 weeks and the quantification of emission images. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA; *, P < 0.05; **, P < 0.01; ns, no significance).

Figure 5.

Cholinergic signaling reprogramed immune microenvironment to favor tumor growth via nicotinic receptor in vivo. A, Luciferase-expressing FC1199 or Panc02 cells (2 × 106) were injected into the body of pancreas of C57BL/6 or BALB/cA-nu/nu mice (n = 20) at age of 7 weeks. At age of 9 weeks, 20 mice were randomly divided into four groups and, respectively, received intraperitoneally injected triweekly as follows, carbacholine (1 mg/kg body weight), atropine (1 mg/kg body weight), and carbacholine (1 mg/kg body weight), mecamylamine hydrochloride (1 mg/kg body weight) and carbacholine (1 mg/kg body weight), and control group was treated with DMSO. Mice were euthanized at age of 10 weeks. B, Representative image of IVIS spectrum of mice at age of 9 weeks. C, The quantification of emission images for mice at age of 9 weeks. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). D, The gate strategies of counting beads, CD4+ T cells, and CD8+ T cells. E, The relative quantity of CD4+ T cells and CD8+ T cells in the tumor tissue normalized to tissue weight. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). F and G, The proportion of IFNγ+ CD8+ T cell, IFNγ+ CD4+ T (Th1) cell, and IL4+ CD4+ T (Th2) cell in different groups. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). H, Representative image of IVIS spectrum of C57BL/6 mice at age of 8 and 10 weeks and the quantification of emission images. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). I, The density of CD8A+ T cells, T-bet+ T cells (Th1), and GATA3+ T cells (Th2) was determined by IHC in the tumor lesion. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA). J, Representative image of IVIS spectrum of BALB/cA-nu/nu mice at age of 8 and 10 weeks and the quantification of emission images. Bar graphs presented as mean ± SEM (n = 5; ordinary one-way ANOVA; *, P < 0.05; **, P < 0.01; ns, no significance).

Close modal

To further confirm the assumption that cholinergic signaling promotes the growth of tumors through reliance on the reprogrammed immune microenvironment, we employed Panc02 cells to establish an orthotopic pancreatic cancer model in mice with an intact immune system (C57BL/6) as well as in thymus absent nude mice (BALB/cA-nu/nu). The experimental scheme is shown in Fig. 5A. We found similar tumor-promoting effects of carbacholine in C57BL/6 mice (Fig. 5H). As shown in Supplementary Fig. S4C, IHC staining was performed to determine the immune cell–infiltrating pattern in tumor tissue. Consistent with previous results, cholinergic signaling decreased the density of CD8+ T cells and Th1 cells, while it increased the density of Th2 cells (Fig. 5I). Of note, carbacholine failed to promote tumor growth in the nude mice (Fig. 5J), indicating that the tumor-promoting effects might depend on the suppression of the immune system, especially T cells.

Moreover, pyloroplasty and bilateral subdiaphragmatic vagotomy was performed in KPC mice to verify the immune-suppressive role of the parasympathetic nerve (Supplementary Fig. S5). As demonstrated in Fig. 6A, surgery was performed at 3 months of age, and the effects of the vagotomy were examined 1 month after surgery by analyzing the density of immune cells that infiltrated the tumor. Representative IHC staining of CCL5, CD8A, T-bet, and GATA3 is shown in Fig. 6B. IHC analysis revealed that vagotomy significantly increased the density of CD8+ T cells (P = 0.0115) and decreased the density of GATA3+ T cells (Th2; P = 0.0010). There was a trend that vagotomy could increase the density of T-bet+ T cells (Th1); however, because of the limited sample size, the result did not reach statistical significance (P = 0.0568; Fig. 6C). To investigate the effects of vagotomy on tumor growth and survival, we performed simultaneous tumor cell injection and pyloroplasty with or without bilateral subdiaphragmatic vagotomy (Fig. 6D). We found that vagotomy significantly inhibited tumor growth and improved survival in an orthotopic pancreatic cancer model (Fig. 6E and F).

Figure 6.

Vagotomy changed immune microenvironment and inhibited tumor growth in vivo. A, Pyloroplasty and bilateral subdiaphragmatic vagotomy were performed on KPC mice (n = 4) at age of 3 months. KPC mice in the sham operation group (n = 4) received pyloroplasty at age of 3 months as negative control. Mice were euthanized at age of 4 months and pancreas was fixed in paraffin for IHC staining. B, Representative image of IHC staining for CCL5, CD8A, T-bet, and GATA3 in pancreas from KPC mice. C, The density of CD8A+ T cells, T-bet+ T cells (Th1), and GATA3+ T cells (Th2) was determined by IHC in the tumor lesion. Bar graphs presented as mean ± SD (n = 4; unpaired two-tailed Student t test). D, Pyloroplasty and bilateral subdiaphragmatic vagotomy were performed simultaneously with tumor cell injection on C57BL/6 mice (n = 10) at age of 7 weeks. Mice in the sham operation group (n = 10) received pyloroplasty and tumor cell injection at age of 7 weeks as negative control. The tumor volume was monitored by the bioluminescence IVIS imaging every 7 days until mice death or the 28 day after surgery. E, Representative image of IVIS spectrum of C57BL/6 mice at age of 9 weeks and the quantification of emission images. Bar graphs presented as mean ± SEM (Sham operation group, n = 6; Vagotomy group, n = 9; ordinary one-way ANOVA). F, Kaplan–Meier survival curve showing that the survival of Sham operation group (n = 10) and Vagotomy group (n = 10). **, P < 0.01.

Figure 6.

Vagotomy changed immune microenvironment and inhibited tumor growth in vivo. A, Pyloroplasty and bilateral subdiaphragmatic vagotomy were performed on KPC mice (n = 4) at age of 3 months. KPC mice in the sham operation group (n = 4) received pyloroplasty at age of 3 months as negative control. Mice were euthanized at age of 4 months and pancreas was fixed in paraffin for IHC staining. B, Representative image of IHC staining for CCL5, CD8A, T-bet, and GATA3 in pancreas from KPC mice. C, The density of CD8A+ T cells, T-bet+ T cells (Th1), and GATA3+ T cells (Th2) was determined by IHC in the tumor lesion. Bar graphs presented as mean ± SD (n = 4; unpaired two-tailed Student t test). D, Pyloroplasty and bilateral subdiaphragmatic vagotomy were performed simultaneously with tumor cell injection on C57BL/6 mice (n = 10) at age of 7 weeks. Mice in the sham operation group (n = 10) received pyloroplasty and tumor cell injection at age of 7 weeks as negative control. The tumor volume was monitored by the bioluminescence IVIS imaging every 7 days until mice death or the 28 day after surgery. E, Representative image of IVIS spectrum of C57BL/6 mice at age of 9 weeks and the quantification of emission images. Bar graphs presented as mean ± SEM (Sham operation group, n = 6; Vagotomy group, n = 9; ordinary one-way ANOVA). F, Kaplan–Meier survival curve showing that the survival of Sham operation group (n = 10) and Vagotomy group (n = 10). **, P < 0.01.

Close modal

PDAC is considered as a “cold tumor” that typically lacks of intratumoral effector lymphocytes and therefore shows less response to immunotherapy (3, 5, 6). However, converting a “cold tumor” into a “hot tumor” makes checkpoint therapy possible in pancreatic cancer (31, 32), paving a promising path toward immunotherapy for PDAC. Our work demonstrated that cholinergic signaling might partly contribute to the limited intratumoral effector T-cell infiltration by HDAC1-mediated histone deacetylation. This finding suggested that HDAC1 might be the potential target for immunotherapy. Actually, growing number of phase I/II clinical trials evaluating epigenetic therapies in PDAC, including HDAC inhibitors (HDACi), are in progress. Vorinostat, romidepsin, and panobinostat are all FDA-approved HDACis in the field of cutaneous T-cell lymphoma and multiple myeloma (33). Notably, precursor studies have shown that hyperactivity of HDAC proteins promotes proliferation and epithelial–mesenchymal transition in PDAC (34–36), and a combination of the gemcitabine and HDACi demonstrated a synergistic effect in clinical trials (37, 38). On the basis of these results, we proposed that synchronous targeting of cancer cells as well as the immune microenvironment by HDACi might be a promising therapeutic option that deserves further investigation. Furthermore, a large number of studies suggest that cholinergic anti-inflammatory pathway is a robust regulator of cytokine-mediated inflammatory disease. The α7 nAChR (α7nAChR), expressed on macrophages and other immune cells, was identified as a main mediator of cholinergic anti-inflammatory pathway (39, 40). It is reported that antagonist of α7nAChR enhances the severity of murine pancreatitis, while agonist of α7nAChR (GTS-21) alleviates pancreatitis (41). However, the role of α7nAChR in the content of immune surveillance remains unknown and deserves further study.

We also noticed that acetylcholine could directly decrease the production of IFNγ via nAChRs in CD8+ T cells, which was consistent with the results of a previous study (42). Acetylcholine released from postganglionic fibers of parasympathetic nerves and even immune cells themselves acts as an anti-inflammatory modulator (17), which has been reported in sepsis (43), arthritis (44), inflammatory bowel syndrome (45), and other models of inflammatory and autoimmune disorders (15). The anti-inflammatory function relies on nAChRs. In the α7 AChR-knockout model, an excessive cytokine response was observed (17, 46), and vagus nerve stimulation failed to suppress cytokine production (46). In addition, cholinergic neural signals could also inhibit leukocyte trafficking (16) and influence naïve T-cell differentiation (47). Tracey summarized this acetylcholine-mediated immune modulatory function as a “cholinergic anti-inflammatory pathway,” which is recognized as an important efferent arm of the inflammatory reflex (48). Our data are consistent with the theory of the cholinergic anti-inflammatory pathway and further expand it into the field of cancers. We propose that the cholinergic anti-inflammatory pathway might be a physiologic brake for inflammation; however, cancers take advantage of this “brake” to achieve the immune escape.

The pancreas is an organ with both endocrine and exocrine functions. Its abundant innervation makes the rapid and precise regulation of the pancreas possible, but it might also contribute to the high incidence of perineural invasion. Recent studies highlight the role of the peripheral nerve in tumor initiation and progression. In prostate cancer and gastric cancer, denervation reduced tumor incidence and progression (49, 50). In addition, sensory neuron ablation prevented perineural invasion, delayed pancreatic intraepithelial neoplasia formation and ultimately prolonged survival in transgenic mice with PDAC (51). Of note, prostate cancer, gastric cancer, and PDAC are cancers with relatively high incidence of perineural invasion. These previous works focused on the nerve cell and cancer cell interaction; although Zhao and colleagues have noticed that inflammation-related pathways were activated in the stomach of INS-GAS mice after vagotomy (50), the direct effects of vagotomy on the immune-microenvironment remains unknown. In our study, we first revealed that vagotomy enhanced the infiltration of CD8+ T cells and increased the Th1/Th2 ratio in KPC mice, while hyperactivation of cholinergic signaling contributed to tumor growth via nAChRs in a T-cell–dependent manner. Recently, Renz and colleagues reported that activation of mAChR, especially M1 mAChR, suppressed tumorigenesis through MAPK and PI3K/AKT signaling in PDAC (52). Their conclusions seem to be the opposite of ours; however, the roles of mAChRs and nAChRs are complicated and sometimes controversial. In immune cells, knocking out mAChRs leads to the impaired inflammatory functions, including impaired T-cell differentiation (53) and attenuated cytokine production (54, 55), which are contrary to results obtained following the knockout of nAChRs. Furthermore, Dr. L.-W. Wang's works focused on the role of mAChRs in tumor cells during the tumor initiation process, and we focused on the role of nAChRs in the immune-suppressive microenvironment. The exact functions of acetylcholine need to be further addressed in the specific context of AChRs, TME components, and disease stages.

In this study, we showed that acetylcholine level was elevated in the case of severe perineural invasion and activation or blockade of cholinergic signaling in vivo could regulate the immune response, respectively. However, the limitation is that there is no mature model for perineural invasion in the field of PDAC. Instead, owing to the pathophysiologic and phenotypic similarity to human pancreatic cancer, KPC mice models are widely used in the study of perineural invasion. It is demonstrated that KPC mice exhibited neural plasticity alterations including increased neural density, enhanced pain-associated behavior, and elevated expression of neurotrophic factor expression in dorsal root ganglia neurons compared with wild-type mice, which is in accordance with the neuropathic alterations of PDAC (56, 57). Of note, it is reported that nearly 30% of KPC mice present perineural invasion, which is the histopathologic hallmark of human PDAC (10). Furthermore, KPC mice demonstrated a human-like collagen deposition adjacent to pancreatic intraepithelial neoplasia and PDAC lesions (58). Therefore, in this study, we used genetically engineered KPC mice, which possess similar dense stroma and neural plasticity alterations, to verify the role of cholinergic signaling on tumor TME by vagotomy or sham operation in vivo, and in orthotopic transplantable tumor models, tumor cells were directly injected into the pancreas and grew vigorously in dozens of days. Zheng and colleagues used this model to analyze neural innervation in PDAC, it is suggested that the nerves in the tumor environment remained intact in the first 10 days after tumor injection and then the expansive tumor growth crushed the nerves soon (10). Owing to the rapidly progressive disease course, orthotopic tumor model is insufficient to develop human-like neural plasticity and collagen deposition. Because of these defects of this model, we used exogenous agonists and antagonists of AChRs to study the role of cholinergic signaling on immunosuppressive microenvironment.

In summary, we demonstrated that perineural invasion–associated cholinergic signaling hyperactivation impaired CD8+ T-cell accumulation by epigenetic downregulation of CCL5 in cancer cells, direct inhibition of IFNγ production in CD8+ T cells, and promotion of a Th1 to Th2 switch (Fig. 7). Our findings reveal a vital role for perineural invasion in the immunosuppressive microenvironment, thus providing a potential target for immunotherapy.

Figure 7.

Model depicting perineural invasion promoting an immune-suppressive TME in PDAC. Increased acetylcholine from vagus nerve in pancreatic cancer impaired CD8+ T-cells' accumulation by epigenetic downregulation of CCL5 in cancer cells, direct inactivation of CD8+ T cells, and promotion of Th1 to Th2 switch.

Figure 7.

Model depicting perineural invasion promoting an immune-suppressive TME in PDAC. Increased acetylcholine from vagus nerve in pancreatic cancer impaired CD8+ T-cells' accumulation by epigenetic downregulation of CCL5 in cancer cells, direct inactivation of CD8+ T cells, and promotion of Th1 to Th2 switch.

Close modal

No potential conflicts of interest were disclosed.

Conception and design: Z.-G. Zhang, Y.-W. Sun

Development of methodology: M.-W. Yang, L.-Y. Tao, J.-Y. Yang, Y.-W. Sun

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M.-W. Yang, L.-Y. Tao, J.-Y. Yang, Y.-M. Huo, D.-J. Liu, J. Li, R. He, C. Lin, W. Liu, Q. Li, S.-H. Jiang, L.-P. Hu, G.-A. Tian, J. Shi, L.-W. Wang

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M.-W. Yang, L.-Y. Tao, J.-Y. Yang, D.-J. Liu, X.-L. Fu, R. He, C. Lin, G.G. Xiao, J. Xue

Writing, review, and/or revision of the manuscript: M.-W. Yang, L.-Y. Tao, J.-Y. Yang, Q. Li, S.-H. Jiang, L.-P. Hu, G.G. Xiao, J. Xue, Z.-G. Zhang, Y.-W. Sun

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y.-M. Huo, J. Li, W. Liu, J.-F. Zhang, R. Hua, Q. Li, S.-H. Jiang, G.-A. Tian, X.-X. Zhang, N. Niu, P. Lu, J. Shi, L.-W. Wang, Z.-G. Zhang, Y.-W. Sun

Study supervision: J.-F. Zhang, R. Hua, N. Niu, G.G. Xiao, J. Xue, Z.-G. Zhang, Y.-W. Sun

Other (carried out in vivo experiments): Y.-S. Jiang

This work was supported by grants from National Natural Science Foundation of China (grant number 81802317 to W. Yang; 81874175 to Y.-W. Sun; 81702739 to X.-L. Fu; 81702844 to Y.-M. Huo; 81702726 to W. Liu; 81602414 to J.-Y. Yang; and 81672358 to Z.-G. Zhang), grant from the Shanghai Science and Technology Committee (grant number 17411952100 to Y.-W. Sun), the Shanghai Municipal Education Commission—Gaofeng Clinical Medicine Grant Support (20181708 to Z.-G. Zhang), Program of Shanghai Academic/Technology Research Leader (19XD1403400 to Z.-G. Zhang), Science and Technology Commission of Shanghai Municipality (18410721000 to Z.-G. Zhang), and Shanghai Municipal Health Bureau (2018BR32 to Z.-G. Zhang).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Rahib
L
,
Smith
BD
,
Aizenberg
R
,
Rosenzweig
AB
,
Fleshman
JM
,
Matrisian
LM
. 
Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the United States
.
Cancer Res
2014
;
74
:
2913
21
.
2.
Siegel
RL
,
Miller
KD
,
Jemal
A
. 
Cancer statistics, 2017
.
CA Cancer J Clin
2017
;
67
:
7
30
.
3.
Brahmer
JR
,
Drake
CG
,
Wollner
I
,
Powderly
JD
,
Picus
J
,
Sharfman
WH
, et al
Phase I study of single-agent anti-programmed death-1 (MDX-1106) in refractory solid tumors: safety, clinical activity, pharmacodynamics, and immunologic correlates
.
J Clin Oncol
2010
;
28
:
3167
75
.
4.
Hodi
FS
,
Butler
M
,
Oble
DA
,
Seiden
MV
,
Haluska
FG
,
Kruse
A
, et al
Immunologic and clinical effects of antibody blockade of cytotoxic T lymphocyte-associated antigen 4 in previously vaccinated cancer patients
.
Proc Natl Acad Sci U S A
2008
;
105
:
3005
10
.
5.
Brahmer
JR
,
Tykodi
SS
,
Chow
LQ
,
Hwu
WJ
,
Topalian
SL
,
Hwu
P
, et al
Safety and activity of anti-PD-L1 antibody in patients with advanced cancer
.
N Engl J Med
2012
;
366
:
2455
65
.
6.
Royal
RE
,
Levy
C
,
Turner
K
,
Mathur
A
,
Hughes
M
,
Kammula
US
, et al
Phase 2 trial of single agent Ipilimumab (anti-CTLA-4) for locally advanced or metastatic pancreatic adenocarcinoma
.
J Immunother
2010
;
33
:
828
33
.
7.
Liebig
C
,
Ayala
G
,
Wilks
JA
,
Berger
DH
,
Albo
D
. 
Perineural invasion in cancer: a review of the literature
.
Cancer
2009
;
115
:
3379
91
.
8.
Bapat
AA
,
Hostetter
G
,
Von Hoff
DD
,
Han
H
. 
Perineural invasion and associated pain in pancreatic cancer
.
Nat Rev Cancer
2011
;
11
:
695
707
.
9.
Amit
M
,
Na'ara
S
,
Gil
Z
. 
Mechanisms of cancer dissemination along nerves
.
Nat Rev Cancer
2016
;
16
:
399
408
.
10.
Jurcak
NR
,
Rucki
AA
,
Muth
S
,
Thompson
E
,
Sharma
R
,
Ding
D
, et al
Axon guidance molecules promote perineural invasion and metastasis of orthotopic pancreatic tumors in mice
.
Gastroenterology
2019
;
157
:
838
50
.
11.
Hayakawa
Y
,
Sakitani
K
,
Konishi
M
,
Asfaha
S
,
Niikura
R
,
Tomita
H
, et al
Nerve growth factor promotes gastric tumorigenesis through aberrant cholinergic signaling
.
Cancer Cell
2017
;
31
:
21
34
.
12.
Ayala
GE
,
Dai
H
,
Powell
M
,
Li
R
,
Ding
Y
,
Wheeler
TM
, et al
Cancer-related axonogenesis and neurogenesis in prostate cancer
.
Clin Cancer Res
2008
;
14
:
7593
603
.
13.
Marchesi
F
,
Piemonti
L
,
Fedele
G
,
Destro
A
,
Roncalli
M
,
Albarello
L
, et al
The chemokine receptor CX3CR1 is involved in the neural tropism and malignant behavior of pancreatic ductal adenocarcinoma
.
Cancer Res
2008
;
68
:
9060
9
.
14.
Gil
Z
,
Cavel
O
,
Kelly
K
,
Brader
P
,
Rein
A
,
Gao
SP
, et al
Paracrine regulation of pancreatic cancer cell invasion by peripheral nerves
.
J Natl Cancer Inst
2010
;
102
:
107
18
.
15.
Pavlov
VA
,
Chavan
SS
,
Tracey
KJ
. 
Molecular and functional neuroscience in immunity
.
Annu Rev Immunol
2018
;
36
:
783
812
.
16.
Huston
JM
,
Rosas-Ballina
M
,
Xue
X
,
Dowling
O
,
Ochani
K
,
Ochani
M
, et al
Cholinergic neural signals to the spleen down-regulate leukocyte trafficking via CD11b
.
J Immunol
2009
;
183
:
552
9
.
17.
Borovikova
LV
,
Ivanova
S
,
Zhang
M
,
Yang
H
,
Botchkina
GI
,
Watkins
LR
, et al
Vagus nerve stimulation attenuates the systemic inflammatory response to endotoxin
.
Nature
2000
;
405
:
458
62
.
18.
Wong
CH
,
Jenne
CN
,
Lee
WY
,
Léger
C
,
Kubes
P
. 
Functional innervation of hepatic iNKT cells is immunosuppressive following stroke
.
Science
2011
;
334
:
101
5
.
19.
Ceyhan
GO
,
Bergmann
F
,
Kadihasanoglu
M
,
Altintas
B
,
Demir
IE
,
Hinz
U
, et al
Pancreatic neuropathy and neuropathic pain–a comprehensive pathomorphological study of 546 cases
.
Gastroenterology
2009
;
136
:
177
86
.
20.
Jiang
SH
,
Li
J
,
Dong
FY
,
Yang
JY
,
Liu
DJ
,
Yang
XM
, et al
Increased serotonin signaling contributes to the Warburg effect in pancreatic tumor cells under metabolic stress and promotes growth of pancreatic tumors in mice
.
Gastroenterology
2017
;
153
:
277
91
.
21.
Bindea
G
,
Mlecnik
B
,
Tosolini
M
,
Kirilovsky
A
,
Waldner
M
,
Obenauf
AC
, et al
Spatiotemporal dynamics of intratumoral immune cells reveal the immune landscape in human cancer
.
Immunity
2013
;
39
:
782
95
.
22.
Fukunaga
A
,
Miyamoto
M
,
Cho
Y
,
Murakami
S
,
Kawarada
Y
,
Oshikiri
T
, et al
CD8+ tumor-infiltrating lymphocytes together with CD4+ tumor-infiltrating lymphocytes and dendritic cells improve the prognosis of patients with pancreatic adenocarcinoma
.
Pancreas
2004
;
28
:
e26
31
.
23.
Guirnalda
P
,
Wood
L
,
Goenka
R
,
Crespo
J
,
Paterson
Y
. 
Interferon gamma-induced intratumoral expression of CXCL9 alters the local distribution of T cells following immunotherapy with Listeria monocytogenes
.
Oncoimmunology
2013
;
2
:
e25752
.
24.
Ene-Obong
A
,
Clear
AJ
,
Watt
J
,
Wang
J
,
Fatah
R
,
Riches
JC
, et al
Activated pancreatic stellate cells sequester CD8+ T cells to reduce their infiltration of the juxtatumoral compartment of pancreatic ductal adenocarcinoma
.
Gastroenterology
2013
;
145
:
1121
32
.
25.
Bronte
V
,
Bria
E
. 
Interfering with CCL5/CCR5 at the tumor-stroma interface
.
Cancer Cell
2016
;
29
:
437
39
.
26.
Harlin
H
,
Meng
Y
,
Peterson
AC
,
Zha
Y
,
Tretiakova
M
,
Slingluff
C
, et al
Chemokine expression in melanoma metastases associated with CD8+ T-cell recruitment
.
Cancer Res
2009
;
69
:
3077
85
.
27.
Nancy
P
,
Tagliani
E
,
Tay
CS
,
Asp
P
,
Levy
DE
,
Erlebacher
A
. 
Chemokine gene silencing in decidual stromal cells limits T cell access to the maternal-fetal interface
.
Science
2012
;
336
:
1317
21
.
28.
Chen
T
,
Yan
YE
,
Liu
S
,
Liu
HX
,
Yan
HY
,
Hou
LF
, et al
Increased fetal thymocytes apoptosis contributes to prenatal nicotine exposure-induced Th1/Th2 imbalance in male offspring mice
.
Sci Rep
2016
;
6
:
39013
.
29.
Nizri
E
,
Irony-Tur-Sinai
M
,
Lory
O
,
Orr-Urtreger
A
,
Lavi
E
,
Brenner
T
. 
Activation of the cholinergic anti-inflammatory system by nicotine attenuates neuroinflammation via suppression of Th1 and Th17 responses
.
J Immunol
2009
;
183
:
6681
8
.
30.
Wu
S
,
Zhou
Y
,
Liu
S
,
Zhang
H
,
Luo
H
,
Zuo
X
, et al
Regulatory effect of nicotine on the differentiation of Th1, Th2 and Th17 lymphocyte subsets in patients with rheumatoid arthritis
.
Eur J Pharmacol
2018
;
831
:
38
45
.
31.
Lutz
ER
,
Wu
AA
,
Bigelow
E
,
Sharma
R
,
Mo
G
,
Soares
K
, et al
Immunotherapy converts nonimmunogenic pancreatic tumors into immunogenic foci of immune regulation
.
Cancer Immunol Res
2014
;
2
:
616
31
.
32.
Kim
VM
,
Blair
AB
,
Lauer
P
,
Foley
K
,
Che
X
,
Soares
K
, et al
Anti-pancreatic tumor efficacy of a Listeria-based, Annexin A2-targeting immunotherapy in combination with anti-PD-1 antibodies
.
J Immunother Cancer
2019
;
7
:
132
.
33.
Brien
GL
,
Valerio
DG
,
Armstrong
SA
. 
Exploiting the epigenome to control cancer-promoting gene-expression programs
.
Cancer Cell
2016
;
29
:
464
76
.
34.
von Burstin
J
,
Eser
S
,
Paul
MC
,
Seidler
B
,
Brandl
M
,
Messer
M
, et al
E-cadherin regulates metastasis of pancreatic cancer in vivo and is suppressed by a SNAIL/HDAC1/HDAC2 repressor complex
.
Gastroenterology
2009
;
137
:
361
71
.
35.
Edderkaoui
M
,
Chheda
C
,
Soufi
B
,
Zayou
F
,
Hu
RW
,
Ramanujan
VK
, et al
An inhibitor of GSK3B and HDACs kills pancreatic cancer cells and slows pancreatic tumor growth and metastasis in mice
.
Gastroenterology
2018
;
155
:
1985
98
.
36.
Edderkaoui
M
,
Xu
S
,
Chheda
C
,
Morvaridi
S
,
Hu
RW
,
Grippo
PJ
, et al
HDAC3 mediates smoking-induced pancreatic cancer
.
Oncotarget
2016
;
7
:
7747
60
.
37.
Tan
J
,
Cang
S
,
Ma
Y
,
Petrillo
RL
,
Liu
D
. 
Novel histone deacetylase inhibitors in clinical trials as anti-cancer agents
.
J Hematol Oncol
2010
;
3
:
5
.
38.
Hessmann
E
,
Johnsen
SA
,
Siveke
JT
,
Ellenrieder
V
. 
Epigenetic treatment of pancreatic cancer: is there a therapeutic perspective on the horizon?
Gut
2017
;
66
:
168
79
.
39.
Wang
H
,
Yu
M
,
Ochani
M
,
Amella
CA
,
Tanovic
M
,
Susarla
S
, et al
Nicotinic acetylcholine receptor α7 subunit is an essential regulator of inflammation
.
Nature
2003
;
421
:
384
.
40.
Olofsson
PS
,
Katz
DA
,
Rosas-Ballina
M
,
Levine
YA
,
Ochani
M
,
Valdés-Ferrer
SI
, et al
α7 nicotinic acetylcholine receptor (α7nAChR) expression in bone marrow-derived non-T cells is required for the inflammatory reflex
.
Mol Med
2012
;
18
:
539
43
.
41.
van Westerloo
DJ
,
Giebelen
IA
,
Florquin
S
,
Bruno
MJ
,
Larosa
GJ
,
Ulloa
L
, et al
The vagus nerve and nicotinic receptors modulate experimental pancreatitis severity in mice
.
Gastroenterology
2006
;
130
:
1822
30
.
42.
Fujii
YX
,
Fujigaya
H
,
Moriwaki
Y
,
Misawa
H
,
Kasahara
T
,
Grando
SA
, et al
Enhanced serum antigen-specific IgG 1 and proinflammatory cytokine production in nicotinic acetylcholine receptor α7 subunit gene knockout mice
. 
2007
;
189
:
69
74
.
43.
Wang
H
,
Liao
H
,
Ochani
M
,
Justiniani
M
,
Lin
X
,
Yang
L
, et al
Cholinergic agonists inhibit HMGB1 release and improve survival in experimental sepsis
.
Nat Med
. 
2004
;
10
:
1216
21
.
44.
Levine
YA
,
Koopman
FA
,
Faltys
M
,
Caravaca
A
,
Bendele
A
,
Zitnik
R
, et al
Neurostimulation of the cholinergic anti-inflammatory pathway ameliorates disease in rat collagen-induced arthritis
.
PLoS One
2014
;
9
:
e104530
.
45.
Meregnani
J
,
Clarençon
D
,
Vivier
M
,
Peinnequin
A
,
Mouret
C
,
Sinniger
V
, et al
Anti-inflammatory effect of vagus nerve stimulation in a rat model of inflammatory bowel disease
.
Auton Neurosci
2011
;
160
:
82
9
.
46.
Wang
H
,
Yu
M
,
Ochani
M
,
Amella
CA
,
Tanovic
M
,
Susarla
S
, et al
Nicotinic acetylcholine receptor α7 subunit is an essential regulator of inflammation
.
Nature
2003
;
421
:
384
8
.
47.
Qian
J
,
Galitovskiy
V
,
Chernyavsky
A
,
Marchenko
S
,
Grando
SA
. 
Plasticity of the murine spleen T-cell cholinergic receptors and their role in in vitro differentiation of naive CD4 T cells toward the Th1, Th2 and Th17 lineages
.
Genes Immun
2011
;
12
:
222
30
.
48.
Tracey
KJ
. 
The inflammatory reflex
.
Nature
2002
;
420
:
853
9
.
49.
Magnon
C
,
Hall
SJ
,
Lin
J
,
Xue
X
,
Gerber
L
,
Freedland
SJ
, et al
Autonomic nerve development contributes to prostate cancer progression
.
Science
2013
;
341
:
1236361
.
50.
Zhao
CM
,
Hayakawa
Y
,
Kodama
Y
,
Muthupalani
S
,
Westphalen
CB
,
Andersen
GT
, et al
Denervation suppresses gastric tumorigenesis
.
Sci Transl Med
2014
;
6
:
250ra115
.
51.
Saloman
JL
,
Albers
KM
,
Li
D
,
Hartman
DJ
,
Crawford
HC
,
Muha
EA
, et al
Ablation of sensory neurons in a genetic model of pancreatic ductal adenocarcinoma slows initiation and progression of cancer
.
Proc Natl Acad Sci U S A
2016
;
113
:
3078
83
.
52.
Renz
BW
,
Tanaka
T
,
Sunagawa
M
,
Takahashi
R
,
Jiang
Z
,
Macchini
M
, et al
Cholinergic signaling via muscarinic receptors directly and indirectly suppresses pancreatic tumorigenesis and cancer stemness
.
Cancer Discov
2018
;
8
:
1458
73
.
53.
Zimring
JC
,
Kapp
LM
,
Yamada
M
,
Wess
J
,
Kapp
JA
. 
Regulation of CD8+ cytolytic T lymphocyte differentiation by a cholinergic pathway
.
J Neuroimmunol
2005
;
164
:
66
75
.
54.
Fujii
YX
,
Tashiro
A
,
Arimoto
K
,
Fujigaya
H
,
Moriwaki
Y
,
Misawa
H
, et al
Diminished antigen-specific IgG 1 and interleukin-6 production and acetylcholinesterase expression in combined M 1 and M 5 muscarinic acetylcholine receptor knockout mice
.
J Neuroimmunol
2007
;
188
:
80
5
.
55.
Xu
Z-P
,
Yang
K
,
Xu
G-N
,
Zhu
L
,
Hou
LN
,
Zhang
WH
, et al
Role of M3 mAChR in in vivo and in vitro models of LPS-induced inflammatory response
.
Int Immunopharmacol
2012
;
14
:
320
7
.
56.
Winston
JH
,
He
Z-J
,
Shenoy
M
,
Xiao
SY
,
Pasricha
PJ
. 
Molecular and behavioral changes in nociception in a novel rat model of chronic pancreatitis for the study of pain
.
Pain
2005
;
117
:
214
22
.
57.
Stopczynski
RE
,
Normolle
DP
,
Hartman
DJ
,
Ying
H
,
DeBerry
JJ
,
Bielefeldt
K
, et al
Neuroplastic changes occur early in the development of pancreatic ductal adenocarcinoma
.
Cancer Res
2014
;
74
:
1718
27
.
58.
Bailey
J
,
Hendley
A
,
Lafaro
K
,
Pruski
MA
,
Jones
NC
,
Alsina
J
, et al
p53 mutations cooperate with oncogenic Kras to promote adenocarcinoma from pancreatic ductal cells
.
Oncogene
2016
;
35
:
4282
8
.