Combined Targeting of NAD Biosynthesis and the NAD-dependent Transcription Factor C-terminal Binding Protein as a Promising Novel Therapy for Pancreatic Cancer

Abstract

Cancer therapies targeting metabolic derangements unique to cancer cells are emerging as a key strategy to address refractory solid tumors such as pancreatic ductal adenocarcinomas (PDAC) that exhibit resistance to extreme nutrient deprivation in the tumor microenvironment. Nicotinamide adenine dinucleotide (NAD) participates in multiple metabolic pathways and nicotinamide phosphoribosyl transferase (NAMPT) is one of the key intracellular enzymes that facilitate the synthesis of NAD. C-terminal binding proteins 1 and 2 (CtBP) are paralogous NAD-dependent oncogenic transcription factors and dehydrogenases that nucleate an epigenetic complex regulating a cohort of genes responsible for cancer proliferation and metastasis. As adequate intracellular NAD is required for CtBP to oligomerize and execute its oncogenic transcriptional coregulatory activities, we hypothesized that NAD depletion would synergize with CtBP inhibition, improving cell inhibitory efficacy. Indeed, depletion of cellular NAD via the NAMPT inhibitor GMX1778 enhanced growth inhibition induced by either RNAi-mediated CtBP1/2 knockdown or the CtBP dehydrogenase inhibitor 4-chlorophenyl-2-hydroxyimino propanoic acid as much as 10-fold in PDAC cells, while untransformed pancreatic ductal cells were unaffected. The growth inhibitory effects of the NAMPT/CtBP inhibitor combination correlated pharmacodynamically with on-target disruption of CtBP1/2 dimerization, CtBP2 interaction with the CoREST epigenetic regulator, and transcriptional activation of the oncogenic target gene TIAM1. Moreover, this same therapeutic combination strongly attenuated growth of PDAC cell line xenografts in immunodeficient mice, with no observable toxicity. Collectively, our data demonstrate that targeting CtBP in combination with NAD depletion represents a promising therapeutic strategy for PDAC.

Significance: Effective precision therapies are lacking in PDAC. We demonstrate that simultaneous inhibition of NAD metabolism and the oncoprotein CtBP is potently effective at blocking growth of both PDAC cells in culture and human PDAC-derived tumors in mice and should be explored further as a potential therapy for patients with PDAC.

FIGURE 1

NAMPT is highly expressed and regulates NAD in PDAC cells. A, Schematic illustrating the intersection of NAD biosynthesis and CtBP protein–protein interactions and related oncogenic transcriptional function. B, Immunoblot analysis of CtBP1/2 and NAMPT in hTERT-HPNE (HPNE) and indicated PDAC cell lines. Vinculin served as loading control. Representative immunoblots shown; N = 3 independent experiments. C, Quantification of total NAD [NAD(H) + NAD⁺] in Panc-1, PaTu8988T, and HPNE cells after treatment with Vehicle (Veh) or increasing concentrations of GMX1778 for 24 hours, using commercial assay kit (Sigma-Aldrich). Values represent the percent change in total NAD concentration relative to vehicle (Veh) treatment for each cell line. P values were determined by Student t test. D, GMX1778 IC₅₀ values for half-maximal reduction of total cellular NAD concentration in the indicated cell lines. ^#indicates the IC₅₀ could not be calculated and beyond the upper range of the titration. Error bars indicate the SEM. N = 3 independent experiments. *, P < 0.05; **, P < 0.01 for comparison with vehicle-only treatment.

FIGURE 2

Genetic depletion or pharmacologic inhibition of CtBP sensitizes PDAC cells to GMX1778 growth inhibition. A, GMX1778 EC₅₀ values for PaTu8988T or SUIT2 cells expressing shGFP, shCtBP1, or shCtBP2 as measured by 72 hours MTT assay. B, qPCR analysis of TIAM1 (CtBP target gene) mRNA expression after treatment of PaTu8988T cells expressing shGFP or shCtBP1 with Vehicle (0) or indicated concentrations of GMX1778 for 24 hours. The calculation was performed using the ΔΔCt approach. P values were determined by Student t test for comparison with vehicle-treated shGFP cells. C, Representative colony formation assay of PaTu8988T cells treated with Vehicle (top row) or 2 nmol/L GMX1778 (bottom row) for 24 hours, followed by treatment with Vehicle (plate 5) or indicated concentrations of 4-Cl-HIPP (plates 6–8) for 7 days, followed by crystal violet staining. D, Colony formation assay in C was quantified by measuring A₅₉₂ of solubilized crystal violet. P values were determined by Student t test for comparison with vehicle-only treated cells. E, Calculation of GI₅₀ values for 4-Cl-HIPP treatment of PaTu8988T cells with or without 2 nmol/L GMX1778 derived from colony formation assay in C. ^#indicates the EC₅₀ or GI₅₀ could not be calculated and beyond the upper range of the titration of GMX1778 or 4-Cl-HIPP, respectively. Error bars indicate ± 1 SD. N = 3 independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.005.

FIGURE 3

Combined NAMPT/CtBP inhibition attenuates physiologic CtBP dimerization and interaction with transcriptional coregulators. A, Panc-1 cells were treated with Vehicle (lanes 2–4) or 10 nmol/L GMX1778 (lanes 5–7) for 24 hours, followed by addition of Vehicle (0) or indicated concentrations of 4-Cl-HIPP for 48 hours, followed by incubating live cells in 0.25 µmol/L DSG and CtBP2 immunoblotting of cross-linked cell lysates. Results of a representative experiment from three independent experiments are shown. B, Densitometric quantitation of immunoblot in A. P values were determined by Student t test. C, Panc-1 cells were treated with Vehicle (Veh) or 10 nmol/L GMX1778 for 24 hours followed by addition of Vehicle or 250 µmol/L of 4-Cl-HIPP for 48 hours. Lysates of treated cells were then IPd with anti-CtBP2 antibody or control IgG (last lane) and IPs immunoblotted with CtBP1 or CoREST antibodies. Immunoblots of the input lysates are shown in the bottom. Vinculin was used as a loading control. Results of a representative experiment from three independent experiments are shown. D, Densitometric quantitation of immunoblot in C. P values were calculated using one-way ANOVA using Tukey post hoc correction. Error bars indicate ±1 SD. *, P < 0.05; **, P < 0.01; ***, P < 0.005 for comparison with vehicle-only treatment.

FIGURE 4

Combined NAMPT/CtBP inhibition potently represses CtBP transcriptional regulation and inhibits cellular motility. A, qPCR analysis of TIAM1 (CtBP target gene) mRNA expression after indicated treatments for 24 hours in PaTu8988T cells. The calculation was performed using the ΔΔCt approach. P values were determined by Student t test. Error bars indicate ±1 SD. *, P < 0.05; **, P < 0.01 for comparison with vehicle-only treatment. B, Patu8988T cells were seeded into a glass bottom cell culture dish, treated with Vehicle or 2 nmol/L GMX1778 for 24 hours followed by addition of Vehicle or 250 µmol/L of 4-Cl-HIPP for 48 hours and then stained with crystal violet to detect cells that had migrated from the central groove toward the periphery of the dish. Results of a representative experiment from three independent experiments are shown.

FIGURE 5

Combined NAMPT/CtBP inhibition abrogates growth of PDAC xenografts in vivo. A, Four cohorts of 5 NSG mice each were subcutaneously xenografted with 1 × 10⁶ Panc-1 cells and after 2 weeks of tumor growth, mice were treated three times per week for 4 weeks (see inset) by intraperitoneal injection of: (i) Vehicle (100 mmol/L NaHCO₃); (ii) 4-Cl-HIPP (100 mg/kg); (iii) GMX1778 (30 mg/kg); or (iv) the combination of 4-Cl-HIPP and GMX1778. B, Tumor volume (mm³) was measured weekly using calipers and P values were calculated using multiple unpaired Student t test. Error bars indicate ±1 SD. ****, P < 0.0001 for comparison with vehicle-only treatment. C, Mean tumor weights determined at necropsy at the end of experiment on day 28 and P values were calculated using multiple unpaired Student t test. Error bars indicate ±1 SD. *, P < 0.05 for comparison with vehicle-only treatment.