Photosensitization, Uptake, and Retention of Phenoxazine Nile Blue Derivatives in Human Bladder Carcinoma Cells1

The overall goal of our research is to develop effective new photosensitizers for tumor-selective photodynamic therapy. Phenoxazine dyes, including several Nile blue analogues, are known to localize selectively in animal tumors. Structural modifications yielded several series of analogues with substantially higher '02 yields and different photochem ical and physicochemical properties. This study examined the photosensitization potency, cellular uptake, and retention of these derivatives in human bladder carcinoma cells (MGH-U1) in culture. Nile blue deriva tives containing halogens and/or sulfur substitutes were selected to exhibit different '(); yields, pKa values, and hydrophobicities. The effectiveness of these derivatives in mediating photokilling of tumor cells in vitro corresponded well with the 'O2 yields of these compounds, indicating that structural modifications which resulted in increased '(), yields enhanced potency in mediating photocytotoxicity in vitro. Using derivatives (sat-NBS and sat-NBS-6I) with the highest 'i >. quantum yield (0.35 and 0.821), over 90% cell kill was achieved at a sensitizer concentration of 5 x Kl" M, about 3 orders of magnitude more effective than hematoporphyrin derivative, the only sensitizer currently available clinically. This result suggests that some of the oxazine deriv atives could potentially be effective photosensitizers. The correspondence between 'Oj yield and photosensitizing potency, together with results showing enhanced photocytotoxicity in the presence of I>.-Oand reduced photocytotoxicity under hypoxic conditions, strongly suggests that the generation of '()? is a major mechanism mediating the photocytotoxic effect. The uptake of Nile blue derivatives by cells in culture exhibited a pattern of rapid initial uptake followed by a gradual increase in cellular dye contents. The uptake does not correlate directly with the individual pKa values or hydrophobicities of the derivatives, indicating that the structural modifications that increased '( ).. yields did not significantly alter the uptake and retention of Nile blue derivatives. The highly concentrative uptake by and slow efflux from dye-loaded cells were consistent with an active mechanism for the cellular accumulation of these dyes. On the other hand, the retention of the compounds was directly proportional to dye concentration in the medium over a 1000fold range of concentrations, and the uptake could proceed at tempera tures below 2°C;these observations excluded endocytosis or a carriermediated mechanism for the uptake. The uptake was also unaffected by the presence of serum in the medium. Based on these results, we hypoth esize that Nile blue derivatives transport across the cell membrane possibly as deprotonated forms and, upon entering the cell, either parti tion into lipophilic areas of the cell membranes and/or become seques tered in certain intracellular organelles. INTRODUCTION PDT1 is an ¡nvestigational cancer treatment procedure. The therapy is effective against a wide range of solid tumors, includReceived 8/10/90; accepted 11/30/90. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' This work was supported by grants from the National Cancer Institute (CA 32259), the Beinecke Foundation, and the Thomas Anthony Pappas Charitable Foundation and by the Rowland Institute of Science. 1To whom requests for reprints should be addressed. 5The abbreviations used are: PDT, photodynamic therapy; PBS. phosphatebuffered saline at pH 7.4; 'O¡,singlet oxygen. The designations of the Nile blue derivatives are listed in Table 1. ing carcinomas of the skin, esophagus, lung, bladder, and breast (1-8). The theoretical basis of the treatment relies on a dual selectivity to achieve a high therapeutic ratio, the selectivity of the photosensitizer to localize in higher concentrations in malignant versus surrounding nonmalignant tissues and the selectivity of the treatment light to activate specifically the exogenous sensitizer and destroy the tumor. Currently, many investigators are searching for new photosensitizers to expand the utility of this treatment. Approaches for this search include dyes that are chemically well defined, with high photoreactivity at excitation wavelengths above 600 nm and with high tumor selectivity (8). The overall goal of our research is to develop photosensitizers with high tumor selectivity. Our investigations have centered on a group of compounds related to the phenoxazine dye Nile blue. This investigation stems from results of early studies by Lewis et al. (9-11), who, in an extensive search for dyes that selectively stain tumors and retard tumor growth, found a number of phenoxazines, including several Nile blue analogues, which cause intense, diffuse staining of viable tumor cells but not of the tumor stroma or of surrounding tissues. These dyes exhibit relatively low systemic toxicity, and some of them temporarily retard tumor growth. These findings were further substantiated by studies of Riley (12) and of Bates and Kershman (13). The former tested the effect of 55 different dyes on transplantable mouse carcinomas and sarcomas and found that Nile blue stained tumors and retarded their growth. The latter investigators observed selective intense staining of Nile blue 2B on a number of transplanted and carcinogen-induced murine brain tumors. These early studies strongly suggested that phen oxazines may constitute a special class of dyes that are selec tively localized in tumors. More recently, it was shown that the tumor-localizing phen oxazines, such as Nile blue A and Nile blue 2B, have very low 'O2 yields because of inefficient intersystem crossing of excited states (14). Structural modifications of Nile blue A, including halogenation, sulfur substitution, and saturation of the angular D-ring, resulted in a number of derivatives with increased 'O2 yields (14, 15). These derivatives have a strong absorbance within the therapeutic window of light (600-1200 nm) and relatively low systemic toxicity and thus may be effective pho tosensitizers for PDT. Preliminary experimental results showed that several of these derivatives are effective in mediating the selective photokilling of tumor cells in culture (14). The com bination of high photoactivity, strong absorption of red wave lengths, and possible high tumor selectivity makes these com pounds promising candidates as photosensitizers for the selec tive PDT of tumors. Furthermore, these derivatives possess different physicochemical properties, such as pKa and hydrophobicity, and are capable of undergoing protonation-deprotonation conversion depending on the pH of the environment (15, 16). These compounds, therefore, may be useful in the investi gation of the possible effects of these factors on cellular dye 1109 on June 12, 2017. © 1991 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from NILE BLUE DERIVATIVES AS POTENTIAL PHOTOSENS1TIZERS uptake and might provide insight into the mechanism for selec tive localization of compounds in tumors. The main purposes of the current study were (a) to evaluate the photosensitization potencies of various Nile blue derivatives in mediating cell killing in vitro and to identify potentially effective photosensitizers for PDT and (b) to examine the cellular uptake and retention of derivatives with different chem ical properties to shed light on the mechanism of selective localization of Nile blue in tumors. The results reveal that the effectiveness of these derivatives in mediating photokilling of tumor cells corresponded with their 'O2 yields, thus suggesting that those with high 'Û2yields can be effective photosensitizers and that 'O2 is probably a major mechanism of the photocytotoxic effect induced by these compounds. The uptake of Nile blue derivatives was rapid, was highly concentrative, was pro portional to dye concentration in the medium, and could occur at temperatures below 2°C.Furthermore, uptake was not af fected by the presence of serum, nor did it correlate with the pKa or hydrophobicity of the dyes. These results indicate that the uptake does not occur by endocytosis or through a carriermediated mechanism. Details of the process are yet to be defined. MATERIALS AND METHODS Nile Blue Derivatives. Previous reports (14-16) showed that structural modifications of Nile blue A resulted in derivatives with enhanced 'O2 yields and different physicochemical properties. Deriva tives used in this investigation were available from previous studies (15, 16), and their '(). quantum yields, absorption maxima, pKa values, and partition coefficients in octanol and PBS are listed in Table 1. Tumor Cells. The experimental system used for this study is a human bladder tumor cell line, MGH-U1, which is a subculture from a wellestablished bladder cancer cell line T-24 (17). MGH-U1 cells grow routinely in McCoy's 5A medium supplemented with 5% fetal calf serum. Under this culture condition, the cells have a doubling time of 14 h and a colony formation efficiency of about 50%. Determination of Photocytotoxicity. Because of the wide differences in photoactivity among various Nile blue derivatives to be tested, we varied the dye concentration rather than the light dose in the evaluation of the sensitizing effect of these dyes in mediating photocytotoxicity in vitro. MGH-U1 cells (2 x IO6 in a 100-mm dish) were incubated with 10 ml of each of the derivatives, ranging from 10~8 to 10~5M, for 30 min at 37°C.Extracellular dye was removed, and light treatment followed immediately. The treatment light source was a Polaroid pro jector with filters allowing light of 590-700 nm to pass through. The power density of this light source at a distance of 10 cm was 8-10 mW/ cm2. The light dose used for the treatment was 4.8 J/cm2 over a treatment period of 8 to 10 min. The number of viable cells, as determined by the colony-forming assay, was used to assess the photodynamic effect. Controls for the study were untreated cells and those treated with light or sensitizer only. Photosensitization in D2O and in Hypoxia. To determine whether D2O can enhance the photodynamic effect of Nile blue derivatives, cells were permitted to take up the photosensitizer as above (NBA-6I, 5 x 10~7M; NBS-6I, 2 x 10"'M) and were incubated with D2O/PBS for 30 min at 37'C to allow exchange of D2O with the cellular water. Light treatment was performed in the presence of D2O at light doses ranging from 1.2 to 7.2 J/cm2, and a colony-forming assay was used to assess the result of the treatment. A hypoxic condition was created as described (18), by placing the cell culture plate in an air-tight chamber and perfusing the chamber with humidified nitrogen. The oxygen tension in the chamber was reduced to 21 ±2(SD) ppm after about 30 min of gassing. Oxygen tension was monitored from the effluent gas flow for the duration of the hypoxia using an oxygen tension meter (Thermox 1; Thermo Laboratory Instrument, Inc.). Light treatment proceeded while the cells in the chamber were maintained at this reduced oxygen tension by continuous gassing during photoirradiation. The colony assay was used to assess the results of the treatment. Controls for both D2O and hypoxia experiments were cells treated with photosensitizer and light without D2O and those treated under atmospheric oxygen tension. Dye Extraction and Quantitation. For quantitation of dye in cells, culture plates were washed twice with Dulbecco's PBS to remove excess dyes. Cells were removed from plates with 0.1% EDTA, dis solved in concentrated HC1, and extracted with chloroform:methanol (1:1) acidified with 0.5% glacial acetic acid. Concentrations of the dye in the extracts were measured by fluorescence spectroscopy in acidic chloroform:methanol at the maximal excitation and emission wave lengths for each derivative: NBA, 620 nm, 660 nm; NBA-6I, 610 nm, 670 nm; NBS, 650 nm, 690 nm; NBS-6I, 660 nm, 690 nm; sat-NBS, 620 nm, 660 nm; sat-NBS-6I, 650 nm, 670 nm. In this study, the overall dye recovery from the cellular experiments was usually between 40 and 50%. Up to about 50% of the dyes were bound to the plastic surface and could not be recovered. Uptake Studies. The patterns of uptake of the Nile blue deriva tives by MGH-U1 cells were determined by plating 2 x IO6cells per 60-mm dish and allowing them to attach overnight. Dyes (2 ml) at 2.5 x 10~6M in phenol red-free and serum-free McCoy's 5A medium were added to the cells, and cellular dye concentrations were determined at different time intervals for up to 80 min while the cells were incubated at 37°C. To examine the effect of dye concentration on the uptake of Nile blue derivatives, two experiments were performed. In the first, the uptake of NBA-6I over time by cells was determined at 4 different dye concentrations in the medium: 0.156, 0.625, 2.5, and 10 nM. In the second, the uptake of NBA-6I at 30 min was examined at various dye concentrations ranging from 10~7to 10~4M. To determine whether cellular uptake of Nile blue derivatives can proceed at temperatures below 2°C,cells in culture, washing buffer, and dye-containing medium were precooled on ice to reach a tempera ture between 0 and 2°Cbefore the uptake was allowed to proceed on ice for various intervals. Efflux Studies. The patterns of efflux of Nile blue derivatives from the cells were determined by first plating the cells at 2 x 10' per 60mm dish and permitting them to attach overnight. The cells were then incubated with serum-free McCoy's 5A medium containing 2.5 x 10~6 Mdye for 30 min at 37°Cto allow uptake of the dye. The dye-containing medium was removed, and the cell culture plates were washed twice and overlayered with the serum-free, dye-free medium. Cellular dye concentrations were determined at various time intervals for up to 6 h. To reduce reuptake of the effluxed dye, fresh dye-free medium was replaced every hour throughout the efflux period.