Targeting the NPY/NPY1R signaling axis in mutant p53-dependent pancreatic cancer impairs metastasis

Abstract

Pancreatic cancer (PC) is a highly metastatic malignancy. More than 80% of patients with PC present with advanced-stage disease, preventing potentially curative surgery. The neuropeptide Y (NPY) system, best known for its role in controlling energy homeostasis, has also been shown to promote tumorigenesis in a range of cancer types, but its role in PC has yet to be explored. We show that expression of NPY and NPY1R are up-regulated in mouse PC models and human patients with PC. Moreover, using the genetically engineered, autochthonous KP^R172HC mouse model of PC, we demonstrate that pancreas-specific and whole-body knockout of Npy1r significantly decreases metastasis to the liver. We identify that treatment with the NPY1R antagonist BIBO3304 significantly reduces KP^R172HC migratory capacity on cell-derived matrices. Pharmacological NPY1R inhibition in an intrasplenic model of PC metastasis recapitulated the results of our genetic studies, with BIBO3304 significantly decreasing liver metastasis. Together, our results reveal that NPY/NPY1R signaling is a previously unidentified antimetastatic target in PC.

Fig. 1.. NPY expression is up-regulated in both mouse models of PC and human patients with PC.

(A) Schematic representation of the two genetically engineered PC mouse models, KP^floxC and KP^R172HC (16). (B) Q-RT-PCR results of the NPY signaling ligands (i) Npy, (ii) Pyy, and (iii) Ppy from whole tissue samples of normal pancreas and KP^R172HC tumors (n = 3). (C) Representative images and quantification of IHC analysis of NPY protein expression of normal pancreas and KP^floxC and KP^R172HC tumors (n = 3). Scale bars, 200 μm. (D) Representative image of NPY IHC of KP^R172HC liver metastases. Scale bar, 200 μm. (E) Representative images of normal pancreas and KP^floxC and KP^R172HC tumors stained with RNAscope for Npy (red) and Npy1r (blue). Scale bars, 25 μm. (F) Representative images of KP^floxC tumor and KP^R172HC tumor with matched liver metastases stained with RNAscope for Krt19 (red) and Npy (blue, top panel) or Npy1r (blue, bottom panel). Scale bars, 50 μm for the primary tumor and 25 μm for the liver metastases. Zoomed insets are 10 μm by 10 μm. (G and H) Gene expression data from the TCGA database assessed through OncoDB (60) for (G) NPY in human PDAC relative to normal pancreas and (H) NPY1R, NPY2R, NPY4R, and NPY5R expression from human PDAC. Normal pancreas (n = 200) and PDAC (n = 178). (I) (i) Representative images and (ii) quantification of IHC analysis of NPY protein expression of human PDAC tumors relative to patient matched normal pancreas (n = 4). Scale bars, 200 μm. Means ± SEM. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ****P < 0.0001 by an unpaired parametric t test or a one-way ANOVA with multiple comparisons.

Fig. 2.. Pancreas-specific Npy1r genetic ablation decreases metastasis to the liver in the KP^R172HC mouse model.

(A) Schematic showing the pancreas-specific Npy1r knockout KP^R172HC model (knockout tissue in blue). (B) Kaplan-Meier analysis of survival of KP^R172HC mice for the three different pancreas-specific Npy1r knockout genotypes, Npy1r WT, Npy1r^flox/+, and Npy1r^flox/flox (n ≥ 20 mice per genotype). (C) Quantification of visible macrometastases in the liver at the study end point. (D) (i) Representative images and (ii) H&E images of sections through the livers at the study end point of the pancreas-specific Npy1r knockout. Metastases are outlined in green. Data for Npy1r WT were the same for both pancreas-specific and whole-body Npy1r knockout survival studies and are also shown in Fig. 3D. Scale bars, 5 mm. Means ± SEM. ns, P ≥ 0.05; **P < 0.01; ***P < 0.001 by a one-way ANOVA or Kaplan-Meier survival analysis.

Fig. 3.. Whole-body Npy1r genetic ablation decreases metastasis to the liver in the KP^R172HC mouse model.

(A) Schematic showing the whole-body Npy1r knockout KP^R172HC model (knockout tissue in blue). (B) Kaplan-Meier analysis of survival of KP^R172HC mice for the three different whole-body Npy1r knockout genotypes, Npy1r WT, Npy1r^+/−, and Npy1r^−/− (n ≥ 19 mice per genotype). (C) Quantification of visible macrometastases in the liver at the study end point. (D) (i) Representative images and (ii) H&E images of sections through the livers at the study end point of the whole-body Npy1r knockout. Metastases are outlined in green. Data for Npy1r WT were the same for both pancreas-specific and whole-body Npy1r knockout survival studies and are also shown in Fig. 2D. Scale bars, 5 mm. Means ± SEM. ns, P ≥ 0.05; *P < 0.05; **P < 0.01 by a one-way ANOVA or Kaplan-Meier survival analysis.

Fig. 4.. Pharmacological NPY1R inhibition decreases KP^R172HC cell motility but not KP^floxC motility in vitro.

(A) Schematic showing the low metastatic KP^floxC model and highly metastatic KP^R172HC PC mouse models, from which cancer cells were derived and used in the following experiments. (B) Quantification of (i) Npy, (ii) Pyy, and (iii) Ppy expression using Q-RT-PCR of cancer cells isolated from the KP^floxC and KP^R172HC models. (C) Schematic showing CDM generation, decellularization, and cancer cell seeding in vitro. (D) Representative binary images of KP^R172HC cells migrating on CDMs upon BIBO3304 or vehicle treatment. (E) Quantification of KP^R172HC cell total distance traveled, average speed, average distance from the origin, and maximum distance from the origin treated with vehicle or 1 μM BIBO3304 over 8 hours (n = 3). (F) Representative x-y tracks of KP^R172HC cells over 8 hours upon treatment with vehicle or 1 μM BIBO3304 (n = 3, 10 cell tracks per polar plot). (G) Representative binary images of KP^floxC cells migrating on CDMs upon BIBO3304 or vehicle treatment. (H) Quantification of KP^floxC cell total distance traveled, average speed, average distance from the origin, and maximum distance from the origin treated with vehicle or 1 μM BIBO3304 over 8 hours (n = 3). (I) Representative x-y tracks of KP^floxC cells over 8 hours upon treatment with vehicle or 1 μM BIBO3304 (n = 3, 10 cell tracks per polar plot). Scale bars, 100 μm. Means ± SEM. ns, P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001 by an unpaired parametric t test.

Fig. 5.. Pharmacological NPY1R inhibition decreases metastatic burden within the liver in vivo.

(A) Schematic showing the treatment schedule for the intrasplenic xenograft experiment with BIBO3304 or vehicle control (saline). (B) Quantification of the number of visible liver metastases per mouse (n ≥ 13 mice per treatment). (C) Quantification of the KRT19⁺ surface area and the number of metastases >50,000 μm² per tissue depth upon treatment with vehicle or BIBO3304. (D) (i) Representative images (top panel) and (ii) IHC images (bottom panel) of KRT19-stained livers, outlining liver tissue in pink and metastases in green, upon treatment with vehicle or BIBO3304. Scale bars, 5 mm. Means ± SEM. *P < 0.05; **P < 0.01 by an unpaired parametric t test.

Fig. 6.. Early NPY1R inhibition decreases metastatic burden within the liver in vivo.

(A) Schematic showing the treatment schedule for the early intrasplenic xenograft experiment with BIBO3304 or vehicle control (saline). (B) Quantification of the number of visible liver metastases per mouse (n ≥ 8 mice per treatment). (C) Quantification of the KRT19⁺ surface area and the number of metastases >10,000 μm² per tissue depth upon treatment with vehicle or BIBO3304. (D) Representative IHC images of KRT19-stained livers, outlining liver tissue in pink and metastases in green, upon treatment with vehicle or BIBO3304. Scale bars, 5 mm. (E) Quantification and representative images of cleaved caspase-3 (CC3) IHC in liver metastases on day 5 following intrasplenic KP^R172HC cancer cell injection and treatment with vehicle or BIBO3304 (n ≥ 8 mice per treatment). (F) Quantification and representative images of Ki67 IHC in liver metastases on day 5 following intrasplenic KP^R172HC cancer cell injection and treatment with vehicle or BIBO3304 (n ≥ 8 mice per treatment). Scale bars, 150 μm. Means ± SEM. ns, P ≥ 0.05; *P < 0.05; ***P < 0.001 by an unpaired parametric t test.