Photodynamic Therapy in Retinoblastoma: Effects of Verteporfin on Retinoblastoma Cell Lines

Abstract

purpose. In contrast to the excellent survival rates of the malignant childhood tumor retinoblastoma (RB), morbidity is high in patients with this disease because of the enucleation or loss of retinal areas caused by current bulb-saving therapies. The authors aimed to preclinically assess the effects of photochemotherapy using second-generation photochemotherapeutics as a prerequisite to develop a promising therapeutic alternative. This therapy implies intravenous application of a photosensitizer activated locally by light of the appropriate wavelength. Activation leads to the formation of free radicals, vascular occlusion, and death of affected cells in the area of irradiation. The photosensitizer verteporfin is approved for the therapy of neovascularizations, such as age-related maculopathy.

methods. The uptake of verteporfin in RB cell lines was investigated. Established RB cell lines, an RB subline resistant to etoposide, and dissociated cells from a primary RB were incubated with verteporfin and irradiated with activating laser light. Proliferation was measured at different time points after application.

results. All five RB cell lines investigated incorporated verteporfin, and nanomolar concentrations were sufficient for effective killing. At lower doses, surviving cells started to proliferate again after several days, but verteporfin 50 ng/mL and 100 J/cm² were sufficient for irreversible killing. High verteporfin concentrations caused cell death with little to no irradiation. Etoposide-resistant cells and primary tumor cells had a comparable susceptibility to photodynamic therapy (PDT) as established parental cell lines.

conclusions. PDT using verteporfin efficiently kills chemotherapy-resistant and nonresistant retinoblastoma cell lines and primary tumor cells in vitro, and it warrants further preclinical evaluation as a therapeutic option for the treatment of retinoblastoma.

The common malignant childhood tumor retinoblastoma (RB) is treated by photocoagulation, cryotherapy, or irradiation. Other methods are used as additions or alternatives to combat special problems, such as additional chemotherapy in the case of hematogenous spreading or involvement of the central nervous system. ¹ In some cases, local treatment such as photocoagulation or cryotherapy follows primary chemoreduction. Most large tumors must be enucleated. ^{2 3}  

The greater than 95% survival rate contrasts the high morbidity caused by therapy. ⁴ All current therapeutic options lead to the loss of retinal areas if the eye bulb can be saved. Additionally, many patients with hereditary retinoblastoma develop nonocular secondary tumors after chemotherapy or irradiation. ^{5 6} New therapeutic insights must be identified to reduce patient morbidity. 

In photodynamic therapy (PDT), a primarily nontoxic photosensibilizator is administered intravenously. Subsequent activation of the dye by low-energy, nonthermal laser light leads to the formation of reactive oxygen species that cause oxidative stress and cell death within the immediate vicinity. ⁷ PDT is of interest for organs, including the brain, bronchial system, and intestine, that must retain microscopic tissue structure to maintain functionality. ^{8 9} By focal activation of the photosensitizer, it is possible to selectively treat the tumor. The photoactive substance is more enriched in the tumor than in the surrounding tissue because of the higher metabolic activity, more permeable vessels, and lower lymphatic drain of the tumor. The antitumor effect of PDT results from the combination of direct cytotoxicity and vascular obliteration. Success rates from 70% to 95% have been described based on tumor type. ¹⁰  

PDT was first attempted on retinoblastoma cell lines using hematoporphyrin and white light. ¹¹ A sensitizing period of at least 3.5 hours, 20 μM photosensitizer, and 6 μW/mm² irradiation were required to induce cell death in Y-79 cells. PDT using hematoporphyrin derivative (HPD) was used against human retinoblastoma xenografts in the anterior eye chamber of athymic nude mice and resulted in marked tumor cytotoxicity. ¹² Amelanotic melanoma was heterotransplanted to the iris in rabbits and treated with HPD. HPD accumulated only in vascularized structures of the eye, in the photoreceptor cell outer segments of the retina, and in the sclera in low concentrations. ¹³ High-dose PDT did not damage lens, cornea, aqueous, or vitreous but led to permanent, nonprogressive acute retinal damage. Using this first generation of photosensitizers, the benefit of PDT was negligible considering the severe side effects. 

The second-generation photosensitizer verteporfin (benzoporphyrin derivative monoacid ring A), a chlorin-type molecule, has been approved by the United States Food and Drug Administration for treating ocular tumors, neovascularization (as in age-related maculopathy [ARM]), and other macular diseases. ^{14 15} Worldwide, more than 200,000 patients have already been treated with verteporfin. It efficiently absorbs light of 680- to 695-nm wavelengths and is able to penetrate blood, melanine, and fibrotic tissue. ⁷ A nonthermal diode laser is the most suitable light source for verteporfin PDT, eliminating the possibility of thermal damage to eye structures. Verteporfin is encapsulated in liposomes, thus increasing stability and reducing renal and hepatic elimination. ¹⁶ Liposomal packaging makes verteporfin lipophile and able to bind LDL in the blood plasma. LDL-bound verteporfin enters cells through the LDL receptor by receptor-mediated endocytosis. ¹⁷ Fast-growing cells, such as tumor cells and neovascularizations, accumulate more verteporfin than normal cells by selective extravasation because of leaky vasculature, providing high target specificity and negligible adverse effects to normal tissues. ^{18 19 20} Verteporfin accumulates rapidly in vascular endothelial cells, ^{17 19 21} choroid, pigmented retina, and photoreceptor outer segments ²² but only negligibly in cornea, lens, or vitreous. The half-life of verteporfin is 5 to 6 hours and the substance is cleared from the body within 24 hours. ²³ Verteporfin has already been shown, through the use of electron microscopy and proliferation assays, to cause apoptosis in the established retinoblastoma cell line SO-RB₅₀. ²⁴  

Materials and Methods

Cell Lines

Retinoblastoma cell lines Y-79 and WERI-Rb1 were obtained from the German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany). RB247C3, RB355, and RB383 were kindly provided by B. Gallie (Hospital for Sick Children, Toronto, Canada). All cell lines were cultivated in RB medium (DMEM supplemented with 15% FBS, 4 mM l-glutamine, 50 μM β-mercaptoethanol, 10 μg insulin/mL, and penicillin/streptomycin) at 37°C in 10% CO₂ and 80% humidity. 

Chemotherapy-Resistant Cell Lines

Chemotherapy-resistant subclones were produced by consecutively increasing the concentration of a chemotherapeutic drug in the culture medium over time until the IC₅₀ of the subclone was at least 10-fold higher than that of the parental cell line. A subclone of the WERI-Rb1 cell line resistant to etoposide was used in PDT experiments. 

Primary Tumors

Primary tumor material was isolated from the eye bulb using a syringe immediately after enucleation and passed through a 40 μm nylon cell strainer. Cells were washed, resuspended in an appropriate volume of RB medium, and immediately used for PDT. 

Western Blotting

Retinoblastoma cells were washed three times with PBS and lysed in ice-cold NP-40 buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 5 mM EDTA, 1% NP-40) containing inhibitor (Complete Protease Inhibitor Cocktail; Roche, Basel, Switzerland). Lysates were cleared by centrifugation, separated (50 mg total protein/lane) on a polyacrylamide gel system (NuPAGE BisTRIS; Invitrogen, Carlsbad, CA) gel (4%–12% polyacrylamide gradient), and electroblotted onto nitrocellulose. Filters were blocked with 5% nonfat dry milk powder in PBS, then incubated with the anti-LDL receptor monoclonal antibody (Oncogene, San Diego, CA) diluted 1:500 for 16 hours at 4°C. After washing 3 × 10 minutes with PBS containing 0.02% Tween-20, blots were incubated for 30 minutes with anti-mouse immunoglobulin (Dako, Glostrup, Denmark) diluted 1:1000 and with streptavidin-biotin-horseradish peroxidase complex (Amersham, Piscataway, NJ) for 1 hour at room temperature. Immunocomplexes were detected with enhanced chemiluminescence and film (ECL Plus kit and ECL Hyperfilm; Amersham). 

Demonstration of Cellular Uptake

For microscopic demonstration of cellular uptake, 2 μg/mL verteporfin was added to the culture medium. The intensity of the green color within the cells was observed using bright-field microscopy. 

PDT and Cell Proliferation Assay

The previously described MTT assay for proliferation ²⁵ was modified to avoid problems with aggregation of retinoblastoma cell lines cultured in 96-well plates. For each treatment group, 10⁷ cells were seeded in a volume of 15 mL into 9-cm Petri dishes. Verteporfin (Visudyne; Novartis, Basel, Switzerland) was added to each treatment group in the appropriate concentration. Cells were incubated for 1 hour, washed twice with 15 mL RB medium, resuspended in 5 mL RB medium, and transferred to a 12-well plate. PDT was carried out (Visulas 690; Carl Zeiss, Oberkochen, Germany) with a spot size of 5 mm at 689 ± 3 nm. Higher light doses were achieved using longer irradiation times. After the indicated treatment of cells, 100 μL aliquots were transferred to a 24-well plate together with 900 μL RB medium and were incubated for 24, 48, or 72 hours. Cells were incubated for 1 hour after the addition of 200 μL MTT solution (6 mg/mL in PBS) and were solubilized by the addition of 1 mL stop solution (10% SDS, 5% acetic acid in dimethyl sulfoxide). Absorbance at 570 nm was measured using a universal plate reader (EM800; Bio-Tek Instruments, Bad Friedrichshall, Germany). Each experiment was performed at least in triplicate. 

Results

Internalization of Verteporfin by Retinoblastoma Cell Lines

We assessed whether verteporfin would be taken up by the retinoblastoma cells once it reached the tumor. Verteporfin circulates in the blood bound to low-density lipoprotein (LDL), and the LDL receptor must be present for its cellular uptake. The LDL receptor was expressed in every retinoblastoma cell line investigated using Western blotting of whole cell homogenates (Fig. 1) . Because of the intensely green color of verteporfin, cellular uptake could be directly visualized using bright-field microscopy. All retinoblastoma cell lines investigated were intensely stained a few minutes after high doses of verteporfin were added to the culture medium. Representative results are shown for the cell lines RB247C3 and WERI-Rb1 (laser scanning microscopy) in Figure 2 . 

Verteporfin PDT of Established Retinoblastoma Cell Lines

To preclinically test the efficiency of verteporfin for the treatment of retinoblastoma, we first defined the optimal combination of verteporfin concentration and laser light dosage sufficient to markedly decrease retinoblastoma cell proliferation. Established retinoblastoma cell lines were treated with 0 to 1000 ng verteporfin/mL medium and then were exposed to different doses of laser light (Fig. 3) . Laser light alone had no effect on cell proliferation. In combination with laser irradiation, even the lowest verteporfin concentration tested (2 ng/mL) resulted in a marked decrease in cell proliferation. On the other hand, high concentrations of verteporfin (>100 ng/mL) produced toxic effects in cells even without laser treatment. Because it was not possible to completely eliminate ambient lighting during the experiment, toxic effects at high concentrations were most probably the result of the activation of a portion of the verteporfin molecules by residual room lighting. Cells continued to proliferate in the presence of 10 ng/mL verteporfin, even after irradiation with the highest light dose of 100 J/cm². Verteporfin concentrations lower than 50 ng/mL in combination with a range of laser light energy densities were also tested for their effect on cell proliferation over a period of 72 hours (Fig. 4) . All cells were killed by 100 J/cm² light after treatment with 50 ng/mL verteporfin, whereas untreated control cell proliferation decreased only slightly. 

To determine the light dose sufficient to maximally reduce cell proliferation, cells were treated with 50 ng/mL verteporfin and were exposed to up to 250 J/cm² doses of laser light. Cell proliferation was measured after 72 hours. The proliferation of treated cells was reduced by approximately 20% even without laser irradiation in comparison with untreated cells, indicating that dimmed room lighting during handling had some effect on verteporfin-treated cells (Fig. 5) . Treatment with 50 J/cm² decreased proliferation to approximately 36%, and 100 J/cm² laser light or more killed 100% of the treated cells. No cell proliferation was detected 72 hours after treatment with 250 J/cm² laser light. 

We investigated the potential long-term effects of PDT to determine whether the few remaining cells eventually die or resume proliferation and whether verteporfin-treated nonirradiated cells experience irreparable damage. Regardless of the primary inhibition of proliferation, verteporfin-treated nonirradiated cells grew as nontreated control cells until the cell density reached critical levels at day 15 and proliferation decreased (Fig. 6) . Cells irradiated with 50 J/cm² resumed proliferation after approximately 10 days, whereas 100 J/cm² led to the death of all cells. 

To investigate the susceptibility of different established retinoblastoma cell lines to PDT, we also examined the proliferation of WERI-Rb1, Y-79, RB247C3, RB355, and RB383 72 hours after irradiation with 0, 50, and 100 J/cm² and treatment with 50 ng/mL verteporfin. In all cases, proliferation decreased markedly in a dose-dependent manner (Fig. 7) . 

Verteporfin PDT in Chemoresistance and Primary Tumor Cells

A major problem in the therapy of retinoblastoma is chemotherapy resistance. Multidrug resistance-associated proteins such as lung resistance protein and P-glycoprotein are known to be expressed in retinoblastoma, ^{26 27} and verteporfin could potentially be actively transported out of the cells. We assessed the effect of PDT using verteporfin on resistant retinoblastoma cells to test whether this treatment would be as potent as against nonresistant retinoblastomas. We created a subline of the etoposide-sensitive WERI-Rb1 cell line through long-term culture in the presence of increasing etoposide concentrations. This subline is resistant to 10 times the etoposide IC₅₀ concentration for the parental cell line. Both the sensitive parental cell line and the resistant subline were exposed to 0, 50, and 100 J/cm² after treatment with 50 ng/mL verteporfin. Both cell lines showed marked decreases of cell proliferation after PDT (Fig. 8) , and there was no significant difference in the response of etoposide-sensitive and -resistant WERI-Rb1 cells to PDT (P = 0.916). Thus, at least in this case, chemotherapy resistance did not influence the efficacy of PDT. 

Another factor potentially influencing the susceptibility to PDT in comparison with primary tumor cells is the long-term culture of established retinoblastoma cell lines. Therefore, tumor cells isolated immediately after enucleation were treated with 50 ng/mL verteporfin and irradiated with 0, 50, and 100 J/cm². Proliferation was markedly reduced in a dose-dependent manner after 72 hours compared with untreated control cells (Fig. 9) . 

Discussion

We show here that drug uptake and treatment sensitivity as essential prerequisites for successful treatment of retinoblastoma with PDT using verteporfin are fulfilled. As shown in the first reports demonstrating the feasibility of fluorescence angiography for retinoblastoma, ²⁸ retinoblastomas are sufficiently vascularized as a prerequisite for treatment with intravenous applicable drugs, which are dependent on a local effect. Thus, verteporfin may principally reach the tumor through the vessels and has the potential to cause effects similar to those of age-related macular degeneration (AMD) treatment. 

Because verteporfin, like fluorescein, may leak from tumor vessels, we suggest that PDT using verteporfin has an effect on tumor cells and tumor capillaries. Compared with retinoblastoma cells in culture, PDT may be even more effective in tumors that are essentially dependent on blood vessels for nutrition, comparable to an antiangiogenic effect of drugs in other tumors. 

The LDL receptor as a prerequisite for verteporfin internalization is expressed in all retinoblastoma cell lines investigated, as demonstrated by Western blotting (Fig. 1) . This receptor shows increased expression on some malignant and proliferating cells and is a good target for LDL in combination with therapeutic drugs. ^{29 30} Expression of the LDL receptor in the retina was demonstrated in retinal pigment epithelium ³¹ and another potential receptor for verteporfin, the LDL receptor-related protein 6 (LRP6), in photoreceptor cells. ³² LDL has previously been used as an endogenous carrier of the photosensitizer chlorin E6 in the retinoblastoma cell line Y-79, where receptor-mediated uptake was demonstrated by satiability and competitive inhibition using free LDL. Covalent binding to LDL increased the uptake of chlorin E6 by a factor of 4 to 5. ³³  

A verteporfin concentration of 50 ng/mL and a light dose of 100 J/cm² caused irreversible damage in all retinoblastoma cell lines tested here (Fig. 7) . Interestingly, in previous reports, much higher concentrations of verteporfin were used (up to 2500 ng/mL) in combination with much lower energy densities (1.2–3.6 J/cm²). ²⁴ Our results indicate that lower verteporfin concentrations and higher energy densities lead to the formation of a sufficient number of free radicals, causing irreversible cell damage with fewer side effects in the surrounding tissue. In the clinical setting proposed by the supplier, 6 mg/m² verteporfin is used to treat AMD. This should also be a suitable concentration for the treatment of retinoblastoma. 

Verteporfin has been used worldwide to treat over 200,000 patients with AMD since 2002. One of its benefits is the absence of severe side effects. The low risk for severe photosensitivity reactions makes it necessary for the patient to avoid sunlight for up to 48 hours. Our results with retinoblastoma cells treated at suboptimal verteporfin PDT doses suggest that therapy of retinoblastomas may not irreversibly kill all the tumor cells. For this reason, multiple treatments may be necessary to completely eradicate retinoblastoma. 

Chemotherapy resistance is a major problem in retinoblastoma therapy. ^{34 35} Expression of multidrug resistance-related proteins ^{26 36} could enable retinoblastoma cells to secrete verteporfin through ABC transporter activity or similar mechanisms after uptake by LDL receptors. The etoposide-resistant WERI-Rb1 subline was as susceptible to verteporfin PDT as the parental etoposide-sensitive cell line (Fig. 8) . These results also suggest the efficacy of verteporfin PDT against chemotherapy-resistant retinoblastoma cells. 

Genomic alterations occurring in cells that are maintained in long-term culture may result in phenotypic differences of cell lines in comparison with the primary tumor cells. We tested the efficiency of verteporfin PDT on a primary tumor directly after enucleation to estimate whether PDT treatment would be as efficient as in established retinoblastoma cell lines. We did not detect any difference in the efficacy of verteporfin PDT in primary retinoblastoma cells compared with established retinoblastoma cell lines (Fig. 9) . 

In a first clinical trial in 1986, five retinoblastoma patients were treated with the first-generation photosensitizer HPD using an argon laser photocoagulator. ³⁷ High doses of photosensitizer (2.5 mg/kg) and light (2 mm spot size, 200 mW, 10 minutes) were required because activated HPD was not effectively causing cell death. The excessive irradiation resulted in vitreous hemorrhaging and retinal damage in this study. Between 1987 and 1990, Winther et al. ^{38 39 40 41 42 43 44 45} published a series of PDT experiments using retinoblastoma-like tumor cells growing under cell culture conditions or in the vitreous of rat eyes. They used EXP-5 cells established from a tumor induced by human adenovirus serotype-12 in rat retina. ⁴⁶ The hematoporphyrin derivate photofrin II and red light were used in these experiments. PDT efficiently killed the retinoblastoma-like cells. However, it also resulted in severe vascular damage. ^{41 42 44} Taken together, the preclinical and clinical uses of first-generation photosensitizers were characterized by several disadvantages: they required long incubation times (1–5 days) before irradiation and high doses of drug (2.5 mg/kg body weight) and light; PDT treatments produced severe side effects, including vascular damage, leukocyte infiltration of the conjunctiva and cornea, edema, and damage to the photoreceptor cells in the retina ⁴⁴ ; and only one third of the treated tumors could be controlled in vivo. ⁴⁷  

Novel second-generation photosensitizers, such as verteporfin, are characterized by better systemic tolerance and more favorable activation wavelengths. Encapsulation of the modified benzoporphyrin ring system into liposomes and modifications for cellular uptake by way of the LDL receptor enable more selective targeting to tumor cells. All these advantages reduce the necessary drug dose, incubation, and irradiation times. Combined with advances in laser technology, verteporfin PDT should be a realistic alternative to treat retinoblastoma. 

²  These authors contributed equally to the work presented here and should therefore be regarded as equivalent authors.

 Supported by grants from the Deutsche Forschungsgemeinschaft and the Jackstaedt-Foundation (AE).

 Submitted for publication August 7, 2007; revised January 14 and March 7, 2008; accepted May 9, 2008.

 Disclosure: H. Stephan, None; R. Boeloeni, None; A. Eggert, None; N. Bornfeld, None; A. Schueler, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Expression of LDL receptor in retinoblastoma cell lines. Western blot with monoclonal anti-LDL receptor antibody.

Figure 2.

 

Uptake of verteporfin by retinoblastoma cell lines. Bright-field microscopic images show representative results in RB247C3 cells (a) before, (b) 1 minute after, and (c) 15 minutes after application of 2 μg verteporfin/mL medium. Green: verteporfin. Overlay of phase-contrast and pseudocolored laser-scanning microscopic images of WERI-Rb1 cells (d) 5 minutes after the addition of verteporfin.

Figure 3.

 

Effect of different verteporfin concentrations on proliferation of established retinoblastoma cells. Cells were incubated with 0 to 1000 ng verteporfin/mL for 1 hour, irradiated with 0, 50, or 100 J/cm². Proliferation was measured up to day 15. Treatment with more than 100 ng verteporfin/mL led to a marked decrease in cell proliferation even without laser treatment. The best effect of PDT was achieved with 10 to 50 ng/mL.

Figure 4.

 

Effect of verteporfin concentration on the proliferation of the established retinoblastoma cell line WERI-Rb1. Cells were incubated with the indicated verteporfin concentrations for 1 hour, washed, and irradiated with 0, 50, and 100 J/cm². Cell proliferation rate was measured 72 hours after PDT; controls were cells without verteporfin application. In contrast to nonirradiated cells, treatment with 50 or 100 J/cm² led to a marked decrease in cell proliferation at a concentration of 2 ng/mL verteporfin.

Figure 5.

 

Proliferation of WERI-Rb1 72 hours after treatment with 50 ng/mL verteporfin and laser light in light doses as given. Controls are cells without previous verteporfin application. Note the decrease of verteporfin-treated cells without laser irradiation compared with controls.

Figure 6.

 

Long-time effect of PDT on retinoblastoma cell line proliferation. WERI-Rb1 cells were treated with 50 ng/mL verteporfin for 1 hour and irradiated with 0, 50, or 100 J/cm². Proliferation was measured up to 23 days after PDT, using untreated cells as controls. The decrease of proliferation in nonirradiated cells after day 15 was the result of saturation in cell density. With 100 J/cm², there was no rescue in cell proliferation as there was with 50 J/cm² after 10 days.

Figure 7.

 

Effect of 50 ng/mL verteporfin 72 hours after PDT on the proliferation of established retinoblastoma cell lines, depending on light dose. Controls are cells without previous verteporfin incubation. All cell lines showed similar effects; in all cases, treatment with verteporfin without irradiation led to a slight decrease in cell proliferation compared with controls because of ambient light. With 50 or 100 J/cm² of irradiation, there was a significant decrease in cell proliferation.

Figure 8.

 

Proliferation of wild-type, etoposide-sensitive, and etoposide-resistant WERI-Rb1 cells 72 hours after treatment with 50 ng/mL verteporfin and PDT, depending on different power densities. Controls are resistant or sensitive cells without previous verteporfin incubation. There was no statistical significant difference between sensitive and resistant cells.

Figure 9.

 

Effect of 50 ng/mL verteporfin on freshly isolated primary retinoblastoma tumor cells after 72 hours and different light doses in PDT. Controls are tumor cells without previous verteporfin incubation. PDT-treated cells showed markedly decreased cell proliferation compared with control.

The authors thank Kathy Astrahantseff for critically reading the manuscript.