Introduction

Retinoblastoma (RB) is the most common primary intraocular malignancy of childhood, affecting 1 in 15 000 children.1 The disease is caused by loss of function of both alleles of the RB tumour suppressor gene (RB1) in retinal progenitor cells. Children with the heritable form of this disease carry one mutated RB1 allele in their germ line, and loss of the remaining allele occurs somatically in developing retinal cells. These children typically develop bilateral, multifocal retinal tumours, and they are also predisposed to develop second primary tumours later in life, particularly bone and soft-tissue sarcomas.2 RB also occurs in a nonheritable form, in which biallelic inactivation of RB1 occurs somatically in a single retinoblast. In this form of the disease, retinal tumour development is invariably unifocal and unilateral.

RB usually develops in children under 6 years of age, and those who do not receive treatment die within 2–4 years after diagnosis. Although RB has the capacity to invade the brain through the optic nerve and to metastasize hematogenously, it has become one of the most curable pediatric solid tumours in developed countries. An overall survival rate between 90 and 95% can be achieved with combinations of chemotherapy, external beam radiotherapy, plaque radiotherapy, laser photocoagulation, cryotherapy, and enucleation.3 In the developing world, however, RB remains a frequently fatal disease.

Significant long-term morbidity and even secondary mortality have been associated with the application of traditional therapy. External beam radiotherapy induces midfacial hypoplasia, retinopathy, optic neuropathy, early cataract formation, and increased incidence of second tumours within the radiation field in children with heritable RB.2, 4, 5 For this reason, chemotherapy with local therapy has become the preferred treatment option for this disease in recent years.6 The current standard chemotherapeutic regimen for RB includes carboplatin, etoposide, and vincristine (CEV). These agents have been associated with adverse effects in some patients including febrile episodes, cytopenia, neutropenia, infections, gastrointestinal distress, and vincristine neurotoxicity.7, 8, 9, 10, 11 Etoposide and carboplatin are also mutagenic. Etoposide therapy is associated with risk for acute myelocytic leukaemia,12 and secondary leukaemias have so far been reported in two RB patients treated with etoposide.13, 14 Carboplatin may also act synergistically with etoposide in the induction of secondary leukaemias.15, 16 Because these agents are mutagenic, both of these drugs have potential to exacerbate second tumour risk in children with heritable disease.

Beta-lapachone is a natural o-naphthoquinone derived from the bark of the South American lapacho tree (Tabebuia avellanedae). This agent induces potent cytotoxic effects in a wide variety of malignant human cell types, including colon, lung, prostate, breast, pancreatic, ovarian, and bone cancers, as well as leukaemia, melanoma, and malignant glioma.17, 18, 19, 20, 21, 22 Beta-lapachone appears to exert a spectrum of anticancer effects, resulting in both apoptotic20, 23, 24 and necrotic17, 19 cell death. Beta-lapachone can directly inhibit DNA topoisomerases I and II,25, 26 and induce G1- and/or S-phase cell-cycle delay followed by apoptosis or necrosis.17, 18, 23 Beta-lapachone cytotoxicity may also be mediated by the activity of the NAD(P)H:quinone oxidoreductase enzyme NQO1.27 This agent has also been shown to selectively induce apoptosis in breast and prostate cancer cells but not in untransformed cells through a mechanism that appears to involve selective induction of the transcription factor E2F1, a regulator of both S-phase progression and apoptosis.24 To our knowledge, the effects of beta-lapachone in RB cell lines have not yet been investigated.

In this study, we tested the growth inhibitory effects of beta-lapachone in vitro in Y79, WERI-RB1, and RBM human retinoblastoma cell lines. Growth inhibitory effects of etoposide were also evaluated in Y79 cells and compared with those of beta-lapachone. The proapoptotic effects of beta-lapachone were evaluated by detection of caspase 3/7 activity, by enzyme-linked immunosorbent assay for nucleosome fragments, and by cellular morphological analysis.

Materials and methods

Agents

Beta-lapachone (Sigma Chemical Company, St Louis, MO, USA) was dissolved in a 20 mM dimethylsulphoxide (DMSO) stock solution and stored in aliquots at −20°C. Etoposide (intravenous injection formulation, Gensia Sicor Pharmaceuticals, Irvine, CA, USA) was purchased from the UCSF Inpatient Pharmacy and diluted in RPMI-1640 media immediately before use.

Cell lines

Y79, WERI-RB1 (UCSF Cell Culture Facility), and RBM (provided by MCM) human RB cell lines were used in these studies. Cell lines were maintained in RPMI-1640 media supplemented with 15% fetal bovine serum and 1% penicillin/streptomycin at 37°C in a humidified 5% CO2 atmosphere.

Growth inhibition assay

Y79, WERI-RB1, and RBM cells were seeded in ninety-six-well microtiter plates (10 000 cells/well) and treated with 0.1–10 μM beta-lapachone or with vehicle control (0.1% DMSO). Y79 cells were also treated with 0.2–20 μM etoposide. Vehicle control for etoposide assays consisted of anhydrous citric acid, polysorbate 80, polyethylene glycol, and dehydrated alcohol. Treated cells were maintained at 37°C for 96 h. At 96 h post-treatment, WST-1 Cell Proliferation Reagent (Roche, Indianapolis, IN, USA) was added to each seeded well and cells were incubated for an additional hour. Absorbance values were then measured in a VersaMax Microplate Reader (Molecular Devices, Sunnyvale, CA, USA). Absolute cell counts were calculated using a standard absorbance curve, and results were expressed as a percentage of vehicle-treated control values. All experiments were performed in triplicate.

Quantification of apoptosis by detection of caspase 3/7 activity

Ninety-six-well microtiter plates were seeded with Y79 cells as described above and treated with 1.9 μM beta-lapachone (the 50% cell growth inhibitory concentration (IC50) of this agent in Y79 cells, as determined by growth inhibition assay). At 8, 24, 48, 72, and 96 h post-treatment, Caspase-Glo 3/7 Reagent (Promega, Madison, WI, USA) was added to each seeded well and cells were incubated for an additional hour. Resulting luminescence values were measured in an Lmax luminometer (Molecular Devices), and results were expressed as fold-induction relative to vehicle-treated controls.

Quantification of apoptosis by enzyme-linked immunosorbent assay for nucleosome fragments

Ninety-six-well microtiter plates were seeded with Y79 cells as described above and treated with 1.9 μM beta-lapachone. At 24, 48, 72, and 96 h post-treatment, nucleosome enrichment in cytoplasmic cell fractions was quantified by Cell Death Detection ELISA Plus assay (Roche) as per manufacturer's instructions. Results were expressed as fold-increase relative to vehicle-treated control values.

Cellular morphology analysis

Y79 cells were treated in flasks at a density of 200 000 cells/ml with 1.9 μM beta-lapachone. At 24, 48, and 72 h post-treatment, cells were fixed in 3.7% paraformaldehyde, seeded into eight-well poly-L-lysine-coated chamber slides, and incubated for 30 min. Plated cells were then treated with 0.5% Triton X-100 for 5 min and stained with Hoechst 33258 (Sigma, St Louis, MO, USA) for 30 min at 37°C. Slides were then mounted with coverslips and viewed under a fluorescent microscope. For each time point, a representative chamber was selected, and all cells in that chamber were counted and categorized by morphological appearance according to previously described criteria28 (http://www.cyto.purdue.edu/cdroms/flow/vol4/index.htm). Cell viability was also evaluated at each time point by trypan blue exclusion assay with manual cell counting. All experiments were performed in triplicate.

Statistical analysis

Analysis of variance methods were used to determine the significance of mean dose-dependent growth inhibitory effects of beta-lapachone in Y79, Weri-RB1, and RBM cell cultures. Regression analyses were used to estimate the best fit for the linear component of the dose–response curve. The maximum R2-value identified the range of concentrations to include in the calculation. The IC50 was determined from the regression equation for each of the three cell lines.

Results

Beta-lapachone induces significant dose-dependent growth inhibition in RB cell lines

Beta-lapachone significantly inhibited proliferation of WERI-RB1 and RBM cells in a dose-dependent manner at 96 h post-treatment (P<0.0001 for each cell line, Figure 1). At the same time point, beta-lapachone also induced significant dose-dependent antiproliferative effects in Y79 cells (P<0.0001, Figure 1); however, these effects were observed at concentrations >1.4 μM, with enhanced proliferation of Y79 cells observed at lower concentrations, including 1.0 and 1.2 μM (Figure 1). The IC50 of beta-lapachone was 1.3 μM in WERI-RB1 cells, 0.9 μM in RBM cells, and 1.9 μM in Y79 cells. By comparison, we found that etoposide also induced dose-dependent antiproliferative effects at low micromolar concentrations in Y79 cells (Figure 2). The IC50 of etoposide was 1.2 μM, similar to the IC50 of beta-lapachone in Y79 (1.9 μM) and other RB cell types (see above).

Figure 1
figure 1

Growth inhibitory effects of beta-lapachone in RB cell lines at 96 h post-treatment. The IC50 of beta-lapachone was 1.3 μM in Weri-RB1 cells, 0.9 μM in RBM cells, and 1.9 μM in Y79 cells.

Full size image
Figure 2
figure 2

Growth inhibitory effects of etoposide in Y79 cells at 96 h post-treatment. The IC50 of etoposide was 1.2 μM, which was similar to the IC50 of beta-lapachone in this cell line.

Full size image

Beta-lapachone induces proapoptotic effects in Y79 RB cells

To determine whether the antiproliferative effects of beta-lapachone were mediated by induction of apoptosis, we assayed beta-lapachone-treated Y79 cells for the activities of caspases 3 and 7, which are markers for apoptosis but not necrosis. Y79 cells treated with 1.9 μM beta-lapachone showed a marked increase in caspase activity (Figure 3). At 72 h post-treatment, the mean caspase activity in beta-lapachone-treated cells was 3.5 times higher than in vehicle-treated controls (Figure 3).

Figure 3
figure 3

Caspase 3/7 activation in Y79 cells treated with 1.9 μM beta-lapachone. Beta-lapachone induced apoptosis-associated caspase 3/7 activity, which increased to 3.5 times vehicle-treated control values at 72 h post-treatment on average.

Full size image

We also quantified the degree of apoptosis-associated intranucleosomal DNA fragmentation in Y79 cells treated with 1.9 μM beta-lapachone. Consistent with the caspase analysis, we observed an enrichment of cytoplasmic nucleosome fragments beginning at 72 h post-treatment, which increased to 5.6 times vehicle-treated control values at 96-h post-treatment on average (Figure 4).

Figure 4
figure 4

Cytoplasmic nucleosome enrichment in Y79 cells treated with 1.9 μM beta-lapachone. Beta-lapachone induced apoptosis-associated intranucleosomal fragmentation, which increased to 5.6 times vehicle-treated control values at 96 h post-treatment on average.

Full size image

Morphological analysis of Y79 cells treated with 1.9 μM beta-lapachone confirmed an increased frequency of apoptotic cells relative to vehicle-treated controls (Figures 5 and 6). The frequency of cells with obvious apoptotic morphology (nuclear condensation and fragmentation, and cytoplasmic condensation) peaked at 48 h post-treatment, reaching 1.7 times control values on average. The mean per cent of apoptotic cells at this time point was 26.0±3.8% (mean±SD) in beta-lapachone-treated cells, compared with 9.7±0.3% in control cultures (Figure 5). Trypan blue analysis showed that reductions in cell viability peaked at 72 h post-treatment (54% of control, data not shown). Beginning at 72 h post-treatment, we also observed a large increase in the per cent of degenerate cells that took up Hoechst stain (29.7±2.5% in treated cultures vs 8.6±2.6% in control cultures, data not shown). These cells displayed advanced degeneration of the nucleus and stained faintly with Hoechst.

Figure 5
figure 5

Apoptotic cellular morphology detected by Hoechst staining in Y79 cells treated with 1.9 μM beta-lapachone. Beta-lapachone increased the mean per cent of apoptotic cell bodies in Y79 cultures to 1.7 times vehicle-treated control values at 48-h post-treatment. The mean per cent of apoptotic cells at this time point was 26.0±3.8 and 9.7±0.3% (mean±SD) in treated and control cultures, respectively.

Full size image
Figure 6
figure 6

Beta-lapachone-treated Y79 cells demonstrating apoptotic cellular morphology (arrows) by Hoechst staining. Cells were treated with 1.9 μM beta-lapachone for 48 h.

Full size image

Discussion

In the present study, beta-lapachone induced dose-dependent growth inhibition at low micromolar concentrations in all three RB cell lines tested, consistent with observations in other cancer cell lines.17, 18, 19, 20, 21, 22 We subsequently confirmed that growth inhibition was mediated, at least in part, by induction of apoptosis, as indicated by increased caspase 3/7 activity and intranucleosomal fragmentation. In addition, beta-lapachone-treated Y79 RB cells displayed cellular morphology characteristic of apoptosis, including chromatin condensation, nuclear fragmentation with bleb formation, and pyknotic bodies28 (http://www.cyto.purdue.edu/cdroms/flow/vol4/index.htm). At 72 h post-treatment, we noted the appearance of a significant number of degenerate cells that could not be characterized definitively, as reported previously.29 These degenerate cells were infrequently observed at the earlier time points (24 and 48 h post-treatment), suggesting late-stage degeneration characteristic of apoptosis in cultured cells.29

Beta-lapachone has recently been reported to induce apoptosis selectively in breast and colon cancer cells, but not in untransformed breast or colon epithelial cells.24 Low micromolar concentrations of beta-lapachone have also been observed to profoundly inhibit proliferation of myeloma cells with no apparent effect on normal peripheral blood mononuclear cells at the same concentrations, further indicating an unusual selectivity of beta-lapachone against cancer cells.30

The induction of apoptosis in breast and colon cancer cells by beta-lapachone has been reported to involve selective and rapid induction of the cell-cycle regulating factor, E2F1.24 E2F1 is one of the ‘activating’ members of the E2F family of transcription factors, which, along with E2F2 and E2F3, can induce a large set of genes required for S-phase progression.31, 32 E2F1 has been found to regulate both cellular proliferation and apoptosis, depending on the levels of E2F1 expressed. Higher levels of E2F1 appear to be required for apoptosis induction than for S-phase entry.31, 32, 33

Other studies in prostate and breast cancer cells have reported that beta-lapachone-induced apoptosis can occur through a mechanism involving NQOI-mediated cycling of this agent.27, 34, 35 NQOI is an enzyme that reduces quinone-containing agents such as beta-lapachone to more reactive forms. NQO1-deficient prostate cancer cells, and prostate and breast cancer cells treated with dicoumarol (an NQO1 inhibitor) are significantly more resistant to apoptosis than NQO1-expressing prostate cells after beta-lapachone exposure.27, 34 Furthermore, transfection of NQO1-deficient cells with an NQO1 expression plasmid markedly enhanced sensitivity to beta-lapachone.27, 34 However, blocking of NQO1 in various cancer cell types does not completely abrogate apoptosis, and although NQO1 plays a major role in beta-lapachone-induced apoptosis, alternate pathways also appear to be involved.27, 34, 35

Although our experiments were limited to evaluating the cytotoxic effects of beta-lapachone alone in RB cells, a number of studies have found significant synergistic effects of beta-lapachone with other chemotherapeutic agents.18, 35, 36, 37 Beta-lapachone induces G1/S checkpoint delay before inducing cell death, and remarkable in vitro and in vivo synergistic effects have been observed when this agent is administered in combination with taxol, which arrests cells in G2/M.18 Beta-lapachone has also been reported to enhance the cytotoxicity of alkylating agents in vitro, possibly owing to beta-lapachone-induced inhibition of DNA repair.37 Other studies have demonstrated that beta-lapachone can be effective as a single agent in cross-resistant cell lines.30, 38 The chemosensitizing effects of beta-lapachone could have application in the clinical management of RB, where chemotherapeutic resistance is frequently observed. Interestingly, beta-lapachone also acts synergistically with ionizing radiation in suppressing tumour growth in vivo, owing to radiation-induced upregulation of NQO1 activity in cancer cells.39

In our review of the literature, we found surprisingly few reports on the in vivo toxicity of beta-lapachone. In one study of a mouse model of human ovarian cancer, beta-lapachone administered intraperitoneally in an oil-based vehicle at doses up to 50 mg/kg (10 cycles, q.o.d.) induced significant antitumour effects with no reported systemic toxicity.18 Consistent yet more complete toxicity data have been reported in normal mice receiving intraperitoneal injections of beta-lapachone in beta-cyclodextrin inclusion complexes on a similar schedule.40 No toxicity was observed in mice treated with 50 mg/kg beta-lapachone in beta-cyclodextrin/water. However, 85 and 100% lethality was observed in mice treated with 60 and 70 mg/kg, respectively. Interestingly, preliminary investigations of mice that died in that study have been unable to detect major damage to vital organs.40

The potent cytotoxic effects of low doses of beta-lapachone in RB cells observed in the present study and the recently reported high degree of selectivity of beta-lapachone towards cancer cells in general suggest that beta-lapachone could have utility in the treatment of RB. To better evaluate efficacy and toxicity of this agent, we are currently conducting in vivo studies in a transgenic murine model of this disease.