Pharmacologic Ascorbate Primes Pancreatic Cancer Cells for Death by Rewiring Cellular Energetics and Inducing DNA Damage

Abstract

The clinical potential of pharmacologic ascorbate (P-AscH^-; intravenous delivery achieving mmol/L concentrations in blood) as an adjuvant in cancer therapy is being reevaluated. At mmol/L concentrations, P-AscH^- is thought to exhibit anticancer activity via generation of a flux of H₂O₂ in tumors, which leads to oxidative distress. Here, we use cell culture models of pancreatic cancer to examine the effects of P-AscH^- on DNA damage, and downstream consequences, including changes in bioenergetics. We have found that the high flux of H₂O₂ produced by P-AscH^- induces DNA damage. In response to this DNA damage, we observed that PARP1 is hyperactivated. Using our unique absolute quantitation, we found that P-AscH^- mediated the overactivation of PARP1, which results in consumption of NAD⁺, and subsequently depletion of ATP leading to mitotic cell death. We have also found that Chk1 plays a major role in the maintenance of genomic integrity following treatment with P-AscH^-. Hyperactivation of PARP1 and DNA repair are ATP-consuming processes. Using a Seahorse XF96 analyzer, we demonstrated that the severe decrease in ATP after challenging with P-AscH^- is because of increased demand, not changes in the rate of production. Genetic deletion and pharmacologic inhibition of PARP1 preserved both NAD⁺ and ATP; however, the toxicity of P-AscH^- remained. These data indicate that disruption of bioenergetics is a secondary factor in the toxicity of P-AscH^-; damage to DNA appears to be the primary factor. IMPLICATIONS: Efforts to leverage P-AscH^- in cancer therapy should first focus on DNA damage.

Figure 1.. Through formation of H₂O₂, P-AscH⁻ activates Chk1 in response to DNA damage.

(A) Western blot analyses show the marker of DNA damage (phosphorylation of histone H2AX, i.e. γH2AX) and DNA replication stress (phosphorylation at S345 of Chk1; P-Chk1) at 60 min following treatment with P-AscH⁻. Prexasertib was used to determine the contribution of Chk1 in DNA damage response after exposure to P-AscH⁻, while extracellular catalase (200 U mL⁻¹) was used to investigate the involvement of H₂O₂. MIA PaCa-2 or PANC-1 cells were pre-treated with prexasertib for 18 h then incubated with P-AscH⁻ + prexasertib for 1 h. The amounts of prexasertib were 50 amol cell⁻¹ (MIA PaCa-2) and 150 amol cell⁻¹ (PANC-1), while the amounts of ascorbate were 7 pmol cell⁻¹ (MIA PaCa-2) and 15 pmol cell⁻¹ (PANC-1). Results are representative of three independent experiments. (B, C) Following treatment of P-AscH⁻ + prexasertib, cell viabilities of MIA PaCa-2 and PANC-1 were evaluated by clonogenic survival assay. The protocol for treatment is similar to panel A (n = 3, mean ± SEM; * p < 0.05, **** p < 0.0001 vs. untreated control; † < 0.05, ††† p < 0.001, †††† p < 0.0001 vs. prexasertib + P-AscH⁻).

Figure 2.. P-AscH⁻ depletes cellular ATP and NAD⁺ via formation of H₂O₂.

(A-D) MIA PaCa-2, PanC-1, 339 or H6c7 cells were incubated with varying amounts of P-AscH⁻ (0 – 40 pmol cell⁻¹) for 1 h; levels of intracellular ATP and NAD⁺ were measured immediately after (n = 3; mean ± SEM). (E, F) To demonstrate the role of H₂O₂, MIA PaCa-2 cells were treated with P-AscH⁻ (14 pmol cell⁻¹) ± bovine catalase (200 unit mL⁻¹) for 1 h, then levels of intracellular ATP and NAD⁺ were determined (n = 3; mean ± SEM). (G, H) The amount of ascorbate required (ED₅₀) to reduce the level of cellular ATP or NAD⁺ by 50% was determined for MIA PaCa-2, PANC-1, 339 and H6c7 cells on the basis of the dose-response curves shown in panels A – D. The ED₅₀’s for ascorbate are plotted vs. the rate constants for removal of extracellular H₂O₂ (k_cell).

Figure 3.. Repair of DNA occurs prior to restoration of cellular energetic state.

(A) nDNA lesions introduced by P-AscH⁻ are fewer and repaired more efficiently than lesions in mtDNA. The ordinate is the change in frequency compared to untreated control,“0”, i.e. lesion frequency per fragment = -ln (A_D/A_Ct), where A_D = amplification of treatment, A_Ct = amplification of control. (B-D) Intracellular ATP as well as NAD⁺ and NADH are depleted by P-AscH⁻. Each is restored but more slowly than the repair of nDNA. MIA PaCa-2 cells were treated with P-AscH⁻ (14 pmol cell⁻¹) for 1 h and then were allowed to recover in fresh media. Responses were determined at indicated times (0 – 24 h following treatment). The dashed line in panels B – D represents the untreated control. Lesions of DNA, n = 4; ATP, NAD⁺, and NADH, n = 3; mean ± SEM.

Figure 4.. Treatment with P-AscH⁻ does not affect the rate of energy production in MIA PaCa-2 or PANC-1 cells.

(A, B) The rate of oxygen consumption (OCR), and (C, D) rate of extracellular acidification (ECAR) were determined immediately after treatment. MIA PaCa-2 or PANC-1 cells were treated with a lethal amount of ascorbate (50 pmol cell⁻¹; 20 mM) for 1 h. OCR and ECAR were determined using a Seahorse XF96 Analyzer (n = 4, each biological replicate consists of ten independent wells, mean ± SEM; * p < 0.05 vs. untreated control).

Figure 5.. P-AscH⁻ activates PARP1, resulting in depletion of NAD⁺ and ATP.

(A) MIA PaCa-2 or PANC-1 cells were incubated with 10 or 30 pmol cell⁻¹ of ascorbate, respectively, for 30 min; then activation of PARP1 was determined by generation of PAR (green) using immunofluorescence microscopy. Nuclei were counter-stained with TO-PRO-3 (blue). Extracellular catalase (200 U mL⁻¹) was used to investigate the contribution of H₂O₂ in activation of PARP1 (n = 3; magnification = x63; Scale bar = 20 μm). (B) MIA PaCa-2 cells were injected into the flank of mice to form tumors. Mice were treated with normal saline (control) or ascorbate (4 g kg⁻¹) twice daily for 5.5 days. After final treatment on day 5.5, mice were sacrificed and tumors harvested. The formation of PAR in tumor tissues was determined by western blot analysis. Each lane represents tissue from individual tumors from different mice. (C) MIA PaCa-2 cells were pre-treated with olaparib (50 fmol cell⁻¹) for 18 h then exposed to P-AscH⁻ (7 pmol cell⁻¹) + olaparib (50 fmol cell⁻¹) for 30 min. Formation of PAR were determined immediately after treatment with western blot analysis. Results are representative of three independent experiments. (D, E) MIA PaCa-2 cells were incubated with olaparib (50 fmol cell⁻¹) for 18 h following by treatment of P-AscH⁻ (7 pmol cell⁻¹) + olaparib (50 fmol cell⁻¹) for 1 h. Intracellular ATP and NAD⁺ were determined immediately after treatment (n = 3; mean ± SEM; *** p < 0.001, **** p < 0.0001 vs. untreated control). (F) To verify the involvement of PARP1 in the depletion of ATP and NAD⁺, MIA PaCa-2 PARP1 KO cells were generated with CRISPR/Cas9 technology. The western blot analysis demonstrated a successful knockout of PARP1. (G, H) CRISPR Control (CRISPR Ctrl) or PARP1 knockout MIA PaCa-2 cells were exposed to P-AscH⁻ (10 pmol cell⁻¹) for 1 h. Intracellular ATP and NAD⁺ were observed immediately after treatment (n = 3; mean ± SEM; * p < 0.05, **** p < 0.0001 vs. untreated control; NS = no significant).

Figure 6.. Disruption to bioenergetics is not the principal factor for anti-cancer activity of P-AscH⁻.

(A, B) Olaparib did not inhibit the cytotoxicity of P-AscH⁻ in MIA PaCa-2 or PANC-1 cells. Cells were pre-treated with olaparib for 18 h then treated with P-AscH⁻ + olaparib for 1 h. Olaparib was at 50 fmol cell⁻¹ (MIA PaCa-2) and 150 fmol cell⁻¹ (PANC-1), while the amounts of ascorbate were 7 pmol cell⁻¹ (MIA PaCa-2) and 15 pmol cell⁻¹ (PANC-1). Cell viability was determined by clonogenic survival assay (n = 3; mean ± SEM; **** p < 0.0001 vs. untreated control; †††† p < 0.0001 vs. olaparib + P-AscH⁻). (C) The activation of Chk1 and phosphorylation of histone H2AX were detected in both MIA PaCa-2 and PANC-1 cells at 1-h following treatment. The protocol for treatment was as in panel A and B. (D) CRISPR Ctrl or PARP1 knockout MIA PaCa-2 cells were treated with different concentrations of P-AscH⁻ (0 – 10 pmol cell⁻¹) for 1 h. Cell viability was evaluated using a clonogenic survival assay (n = 3; mean ± SEM).

Figure 7.. Proposed model for anti-cancer activity of ascorbate.

Pharmacological ascorbate (P-AscH⁻) functions as a pro-drug for the delivery of a significant flux of H₂O₂ to tumors. In the presence of redox active metals, P-AscH⁻ generates high-fluxes of H₂O₂ in the extracellular space of tumors. This H₂O₂ is readily transported across the plasma membrane via peroxiporins; it attacks DNA through the site-specific Fenton reaction. This DNA damage activates PARP1, leading to depletion of NAD⁺ and ATP. In order to maintain genomic integrity, the DNA damage from this high-flux of H₂O₂ leads to activation of Chk1. Our data show that the DNA damage caused by the H₂O₂ produced by P-AscH⁻ appears to be the principal biochemical change responsible for the anti-cancer activity of P-AscH⁻. The disruption of bioenergetics is secondary. The combination of P-AscH⁻ and agents that target DNA-repair (e.g. olaparib, prexasertib) could be a promising approach for the treatment of pancreatic cancer.