Verapamil enhances the antitumoral efficacy of oncolytic adenoviruses

Abstract

The therapeutic potential of oncolytic adenoviruses is limited by the rate of adenovirus release. Based on the observation that several viruses induce cell death and progeny release by disrupting intracellular calcium homeostasis, we hypothesized that the alteration in intracellular calcium concentration induced by verapamil could improve the rate of virus release and spread, eventually enhancing the antitumoral activity of oncolytic adenoviruses. Our results indicate that verapamil substantially enhanced the release of adenovirus from a variety of cell types resulting in an improved cell-to-cell spread and cytotoxicity. Furthermore, the combination of the systemic administration of an oncolytic adenovirus (ICOVIR-5) with verapamil in vivo greatly improved its antitumoral activity in two different tumor xenograft models without affecting the selectivity of this virus. Overall, our findings indicate that verapamil provides a new, safe, and versatile way to improve the antitumoral potency of oncolytic adenoviruses in the clinical setting.

Figure 1

Verapamil enhances the release, spread, and cytotoxicity of Ad5 in vitro. (a) Viral production and release kinetics of Ad5 combined with 40 µmol/l verapamil in A549 cells. Viral content of the total (CE) and extracellular (SN) fractions were analyzed at the indicated time points. Mean values (n = 3) ± SD are plotted.*Significant (P = 8.1 × 10⁻⁵ and P = 0.03 at 48 and 64 hours postinfection, respectively) compared to SN of Ad5. (b) Viral release of Ad5 in combination with verapamil in CAF1, SkMel-28, and NP-9 cells. The time point at which the biggest difference in virus release was observed is shown. *Significant (P = 0.015 for CAF1 and P = 0.04 for NP-9) compared to the release of Ad5. (c) Plaque size of Ad5 in the presence of verapamil (30 µmol/l final concentration) in A549 cells at day 7 postinfection. (d) Comparative cytotoxicity of Ad5 ± verapamil in SkMel-28, A549, and NP-9 tumor cell lines. IC₅₀ values (TU/cell of Ad5 required to cause a reduction of 50% in cell viability) for each condition are shown. TU, transducing units.

Figure 2

Effect of verapamil on viral protein expression and dependence on ADP. (a) Verapamil treatment does not modify adenovirus early and late protein expression pattern. A549 cells were infected with Ad5 with or without verapamil in the extracellular medium, and expression of E1A, E3/19K, L4, and fiber proteins was analyzed at 16, 24, 40, and 48 hours postinfection. (b) Viral release kinetics of Ad5 and AdADP⁻ in A549 cells treated with verapamil. Supernatant viral content was quantified at the indicated time points. Mean values (n = 3) ± SD are plotted. *Significant (P = 0.005 and P = 0.001 at 40 and 48 hours p.i., respectively) compared to the release of Ad5 and ^#significant (P = 0.03, P = 0.003, and P = 0.02 at 48, 64, and 72 hours p.i., respectively) compared to the release of AdADP⁻. (c) Comparative plaque size of rec700 and dl732 in the presence and absence of verapamil in A549 cells at day 6 p.i. ADP, adenovirus death protein; p.i., postinfection.

Figure 3

Dependence of the enhanced release phenotype on the calcium blocking activity of verapamil. (a) Ad5 release at 40 hours postinfection in the presence of calcium channel blockers. Mean values (n = 3) ± SD are plotted.*Significant compared to the release of Ad5 (P = 0.025, P = 0.023, and P = 0.003 for verapamil, amlodipine, and diltiazem, respectively). (b) Plaque size of Ad5 in the presence of calcium channel blockers: verapamil (V), amlodipine (A), and diltiazem (D) at 30 µmol/l. Pictures of representative plaques at 7 days postinfection (p.i.) are shown. (c) Virus production and release kinetics in the presence or absence of calcium in the extracellular medium. Mean values (n = 3) ± SD are plotted.

Figure 4

Potential death mode triggered by verapamil and comparison of the effect of verapamil with that of other drugs that synergize with adenovirus. (a) Apoptosis activation status in Ad5-infected A549 cells in the presence or absence of verapamil. PARP cleavage was detected by western blot at 40 hours postinfection. Cells incubated with Staurosporine (SSP, 0.2 µmol/l) are used as positive control of apoptosis induction. (b) Anti-LC3 western blot in A549 cells in the presence of calcium channel blockers. A549 cells were infected with Ad5 and incubated with normal medium or medium containing rapamycin (positive control of autophagy induction), verapamil (V), amlodipine (A), or diltiazem (D). (c) Virus release in the presence of rapamycin. A549 cells were incubated with increasing concentrations of rapamycin or verapamil, and LC3 expression was assessed by western blot 40 hours p.i. At the same conditions, the extracellular and total virus produced were quantified at 40 hours p.i. Mean values (n = 3) ± SD are plotted. *Significant compared to the release of Ad5 (P = 0.04). (d) Plaque size of Ad5 in the presence of 30 µmol/l verapamil (V) or 5 µmol/l cisplatin (CDDP), 200 pg/ml docetaxel (Doce), 10 nmol/l RAD001, or 10 µmol/l temozolomide (TMZ) (concentration that gave 10% growth inhibition). Cells were stained 7 days postinfection and pictures of representative plaques are shown. p.i., postinfection.

Figure 5

The selectivity of ICOVIR-5 is maintained in the presence of verapamil in an immunocompetent model in vivo. (a) Percent of body weight variation after systemic administration of PBS or ICOVIR-5 alone or combined with daily i.p. injection of 20 mg/kg of verapamil. (b) Liver E1A expression analyzed by immunohistochemistry of representative frozen liver sections of mice treated with PBS or 5 × 10¹⁰ vp of AdwtRGD, or ICOVIR-5 + daily verapamil (V) at day 5 postinfection (AdwtRGD groups were killed at day 3 due to toxicity). (c) Mean values of AST (aspartate animotransferase) and ALT (alanine aminotransferase) in serum and (d) lymphocyte and platelet concentrations in peripheral blood at day 5 postinjection (day 3 postinjection for the groups injected with AdwtRGD at 5 × 10¹⁰ vp) of the doses indicated. Mean values of five animals per group ± SD are plotted. *Significant (P < 0.05) compared to PBS and ^#significant compared to AdwtRGD. PBS, phosphate-buffered saline.

Figure 6

Verapamil enhances the antitumoral activity of ICOVIR-5 in vivo. (a) Nude mice with A549 tumor xenografts were treated with PBS, PBS combined with daily 20 mg/kg verapamil i.p. injection, or a single dose of 5 × 10¹⁰ vp of ICOVIR-5 alone or combined with verapamil. Percent of tumor growth ± SEM is plotted. *Significant (P = 0.0019) compared to PBS; ^§Significant (P = 0.016) compared to PBS + V; and ^#Significant (P = 0.019) compared to ICOVIR-5. (b) Antiadenovirus immunostaining of frozen A549 tumor sections treated with PBS, PBS + verapamil, ICOVIR-5, or ICOVIR-5 + daily verapamil at day 13 after virus administration. (c) Nude mice with SkMel-28 tumor xenografts were treated with PBS, PBS combined with daily 20 mg/kg verapamil i.p. injection, or a single dose of 5 × 10¹⁰ vp of ICOVIR-5 alone or combined with verapamil. Percent of tumor growth ± SEM is plotted. *Significant (P = 0.0009) compared to PBS; ^§Significant (P = 0.0015) compared to PBS +V; and ^#Significant (P = 0.017) compared to ICOVIR-5. (d) Comparative size of SkMel-28 tumor xenografts treated with PBS, PBS + verapamil, ICOVIR-5, or ICOVIR-5 + verapamil at day 52 after virus injection. PBS, phosphate-buffered saline.