Differential Expression of Polyamine Pathways in Human Pancreatic Tumor Progression and Effects of Polyamine Blockade on Tumor Microenvironment

Abstract

Pancreatic cancer is the fourth leading cause of cancer death. Existing therapies only moderately improve pancreatic ductal adenocarcinoma (PDAC) patient prognosis. The present study investigates the importance of the polyamine metabolism in the pancreatic tumor microenvironment. Relative mRNA expression analysis identified differential expression of polyamine biosynthesis, homeostasis, and transport mediators in both pancreatic epithelial and stromal cells from low-grade pancreatic intraepithelial neoplasia (PanIN-1) or primary PDAC patient samples. We found dysregulated mRNA levels that encode for proteins associated with the polyamine pathway of PDAC tumors compared to early lesions. Next, bioinformatic databases were used to assess expression of select genes involved in polyamine metabolism and their impact on patient survival. Higher expression of pro-polyamine genes was associated with poor patient prognosis, supporting the use of a polyamine blockade therapy (PBT) strategy for inhibiting pancreatic tumor progression. Moreover, PBT treatment of syngeneic mice injected intra-pancreatic with PAN 02 tumor cells resulted in increased survival and decreased tumor weights of PDAC-bearing mice. Histological assessment of PBT-treated tumors revealed macrophage presence and significantly increased expression of CD86, a T cell co-stimulatory marker. Collectively, therapies which target polyamine metabolism can be used to disrupt tumor progression, modulate tumor microenvironment, and extend overall survival.

Keywords: CD86; DFMO; immune suppression; macrophage; pancreatic ductal adenocarcinoma; polyamine blockade therapy; polyamine metabolism; polyamine transport inhibitor; survival; tumor microenvironment.

Figure 1

Relative mRNA expression of select polyamine biosynthesis, homeostasis, and transport genes altered in PDAC. Expression of each indicated gene in the epithelial (Epi) and stroma compartments isolated from human PanIN-1 or PDAC samples by laser capture microdissection are represented as log2 scale of transcripts per million (TPM). Approximately 1000 cells per sample were captured and analyzed per materials and methods descriptions. p value: * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001.

Figure 2

Polyamine dysregulation model in PDAC. Graphical summary of key players in polyamine regulation within a cell, highlighting upregulated (red) and downregulated (blue) gene expression products in PDAC epithelium versus stroma. Abbreviations: Arg: Arginine; ARG2: Arginase 2; ATP13A3: an ATPase involved in polyamine transport; AZIN1: Antizyme Inhibitor 1; CAV1: Caveolin 1; GPC1: Glypican 1; Orn: Ornithine; OAZ1: Antizyme 1; ODC1: Ornithine Decarboxylase 1; Put: Putrescine; SLC3A2: Solute transporter 3A2 (subunit of the diamine exporter DAX); SMS: Spermine Synthase; Spd: Spermidine; Spm: Spermine; SRM: Spermidine Synthase; y + LAT: cationic amino acid transporter, subunit of the diamine exporter DAX; OAT: Ornithine Aminotransferase; PAO: Polyamine Oxidase; SMOX: Spermine Oxidase; MYC: Myc Proto-Oncogene Protein; SAT1: Spermidine/Spermine N1-Acetyltransferase 1; AMD1: Adenosylmethionine Decarboxylase 1; MTA: 5′-methylthioadenosine; dcSAM: decarboxylated S-adenosylmethionine; SLC12A8: solute carrier family 12 member 8; N1Ac-Spd: N1-Acetylspermidine; PA: polyamine/polyamine metabolites.

Figure 3

Select polyamine gene expression in human PDAC tumors were associated with patient survival and prognosis. MYC, SMS, AZIN1, OAZ1, and ATP13A3 expression levels were correlated to patient outcomes in the TCGA dataset. (A) Survival of patients in the upper, median, and lower tertiles of gene expression was plotted in Kaplan–Meier curves. (B) Hazard ratio associated with high (upper tertile) gene expression versus either low (lower tertile) or medium (median tertile) gene expression was quantified as a ratio and depicted in a heatmap. (C) Schematic of polyamine blockade therapy strategy showing the nodes of DFMO and PTI intervention, which inhibit polyamine synthesis and import, respectively.

Figure 4

Survival was increased and tumor weights were decreased in PBT treated PAN 02 tumor-bearing mice. (A) Kaplan–Meier curves and associated median survival of mice treated as indicted (n > 9 mice per group). DFMO treatment alone showed a significant improvement in survival compared to control (Log-Rank test, p = 0.0029), although PBT showed the greatest improvement in survival (Log-Rank test, p = 0.001) among the treatment groups. (B) Fixed termination timepoint study showing tumor weights from treated mice (n > 8 mice per group). p value: * < 0.05, ** < 0.01, *** < 0.001.

Figure 5

Pathological assessment shows differences in tumor phenotype and presence of infiltrating immune cells in the mice treated with DFMO and/or PTI (Trimer44NMe). (A) Representative Hemotoxylin and Eosin (H&E) stained sections assessed for tumor phenotype and microenvironment with respect to 0.25% (w/v) DFMO and/or 4 mg/Kg PTI (Trimer44NMe) treatments. (B) Quantification of expression of F4/80 and CD86 across 1% (w/v) DFMO and/or 1.8 mg/kg PTI (Trimer44NMe) treatment groups. (C) Representative immunohistochemistry images of F4/80 and CD86 stained 1% (w/v) DFMO and/or 1.8 mg/kg PTI (Trimer44NMe)-treated pancreatic tumor sections imaged at 20× and 40× magnification. The 20× images contain inlays representing area captured at 40× magnification. Scale bars correspond to 100 μm. p value: ** < 0.01, *** < 0.001, **** < 0.0001.