Up-regulation of human prostaglandin reductase 1 improves the efficacy of hydroxymethylacylfulvene, an antitumor chemotherapeutic agent

Abstract

Prostaglandin reductase 1 (PTGR1) is a highly inducible enzyme with enone reductase activity. Previous studies demonstrated the role of rat PTGR1 in the activation of acylfulvene analogs, a class of antitumor natural product derivatives. Of these, hydroxymethylacylfulvene (HMAF) was in advanced clinical development for the treatment of advanced solid tumors, including prostate, ovarian, and pancreatic cancers. However, the efficiency of human PTGR1 in activating acylfulvenes and its potential to enhance therapeutic efficacy have remained uncharacterized. In this study, human PTGR1 was polymerase chain reaction-cloned and purified. Conversion of HMAF to its cellular metabolite by the purified enzyme proceeded at a 20-fold higher rate than with the rat variant of the enzyme. The Km was 4.9 μM, which was 40-fold lower than for the rat variant and similar to the therapeutic dose. Human cell lines, including colon cancer lines, were transfected with a vector containing rat PTGR1 or human PTGR1, and cell viability was examined after dosing with HMAF. New data obtained in this study suggest that transfection with human PTGR1, or its induction in colon and liver cancer cell lines with 1,2-dithiol-3-thione, enhances susceptibility to the cytotoxic influences of HMAF by 2- to 10-fold. Furthermore, similar or enhanced enzyme induction and HMAF toxicity results from preconditioning cancer cells with the bioactive food components curcumin and resveratrol. The functional impact of PTGR1 induction in human cells and chemical-based strategies for its activation can provide important knowledge for the design of clinical strategies involving reductively activated cytotoxic chemotherapeutics.

Scheme 1.

Bioactivation of acylfulvenes by PTGR1. The structures of illudin M (1), illudin S (2), HMAF (3), and AF (4) are shown. Reduction at the C8 position results in the unstable cyclohexadiene intermediate 5, which may react with water or other cellular nucleophiles (R³).

Fig. 1.

Structures of D3T (top left) and the dietary components curcumin [(1E,6E)-1,7-bis(4-hydroxy-3-methoxyphenyl)-1,6-heptadiene-3,5-dione[rsqb] (bottom) and resveratrol (3,4′,5-stilbenetriol) (top right), three examples of Nrf2 triggers used in this study.

Fig. 2.

hPTGR1 markedly sensitized HEK293 cells toward HMAF. HEK293 cells transfected with pCEP4, pCEP4-rPTGR1, or pCEP4-hPTGR1 were challenged with HMAF at the indicated concentrations. Cell viability was assessed 24 h later. Data were obtained from three independent experiments; error bars represent S.E.

Fig. 3.

Overexpression of hPTGR1 in colon cancer cell line HCT15 significantly enhances HMAF cytotoxicity. A typical survival graph is shown. Error bars represent S.D. IC₅₀ was calculated from three independent experiments by nonlinear regression analysis (P < 0.05).

Fig. 4.

a, identification of potential Nrf-2 binding sites (indicated by fuchsia marks) by promoter sequence analysis 1000 base pairs upstream from the transcription starting site of PTGR1. Both variant sequences of PTGR1 with experimentally verified 5′ transcripts from the database ElDorado 12-2010 were used (Genomatix Gene2promoter software). b, preconditioning with 20 μM D3T for 48 h induced hPTGR1 expression in colon cancer cell SW620 and liver cancer cell HepG2 and enhanced HMAF cytotoxicity. A typical survival graph is shown. Error bars represent S.D. IC₅₀ was calculated from four independent experiments by nonlinear regression analysis (P < 0.05).

Fig. 5.

a, up-regulation of PTGR1 by curcumin, resveratrol, or D3T. Total RNA (24 h) after treatment at indicated concentrations were analyzed via quantitative reverse transcription-PCR (qRT-PCR). Error bars represent S.E. from three independent experiments (*, P < 0.05). Veh, vehicle. b, effects of pretreatment with D3T (20 μM), curcumin (10 μM), or resveratrol (15 μM) on HMAF toxicity in SW620 and HepG2 cells. Error bars represent S.E. from three independent experiments (P < 0.05 with the exception of SW620 + 10 μM curcumin).