Photodynamic therapy.

Photodynamic therapy involves administration of a tumor-localizing photosensitizing agent, which may require metabolic synthesis (i.e., a prodrug), followed by activation of the agent by light of a specific wavelength. This therapy results in a sequence of photochemical and photobiologic processes that cause irreversible photodamage to tumor tissues. Results from preclinical and clinical studies conducted worldwide over a 25-year period have established photodynamic therapy as a useful treatment approach for some cancers. Since 1993, regulatory approval for photodynamic therapy involving use of a partially purified, commercially available hematoporphyrin derivative compound (Photofrin) in patients with early and advanced stage cancer of the lung, digestive tract, and genitourinary tract has been obtained in Canada, The Netherlands, France, Germany, Japan, and the United States. We have attempted to conduct and present a comprehensive review of this rapidly expanding field. Mechanisms of subcellular and tumor localization of photosensitizing agents, as well as of molecular, cellular, and tumor responses associated with photodynamic therapy, are discussed. Technical issues regarding light dosimetry are also considered.

Systemic toxicity in mice induced by localized porphyrin photodynamic therapy.

An unexpected high level of acute lethality has been documented following Photofrin II-mediated photodynamic therapy (PDT) treatments which were localized to the hind leg of normal and tumor-bearing mice. Doses of PDT which induced lethality (10 mg/kg Photofrin II, 200-500 J/cm2) were in the range of doses required to obtain murine tumor cures. The percentage of lethality was proportional to the total light dose but was inversely proportional to the dose rate of delivered light. Comparable levels of acute toxicity were observed in four pigmented mouse strains (C57BL/6J, C3H/HeJ, DBA/1, and DBA/2) and in two albino mouse strains (BALB/c and Swiss Webster). Decreased sensitivity to PDT-induced lethality was observed in two pigmented mouse strains (B10D2/OSN and B10D2/NSN). The administration of warfarin, aspirin, indomethacin, or antihistamine had significant protective effects in terms of decreasing PDT-induced lethality. However, injection of cobra venom factor (to deplete C3 and C5 of the complement system) did not alter the lethality mediated by PDT. Histological profiles obtained 24 h following PDT demonstrated vascular congestion in the liver, kidney, lung, and spleen. Significant decreases in removable blood volume, core temperature, and spleen weight were also observed within 24 h of localized PDT treatment. These results indicate that PDT-induced lethality is consistent with a traumatic shock syndrome and suggest that endogenous vasoactive mediators of shock such as prostaglandins, thromboxanes, and histamine are associated with the lethality induced by localized PDT in mice.

Tissue distribution and photosensitizing properties of mono-L-aspartyl chlorin e6 in a mouse tumor model.

Mono-L-aspartyl chlorin e6 (NPe6) is a photosensitizer that possesses properties such as chemical purity and a major absorption band at 664 nm which are potentially exploitable for photodynamic therapy (PDT). The current investigation examined pharmacological and photosensitizing parameters of NPe6 in tumor and normal tissues in mice. [14C]NPe6 was used to obtain quantitative tissue distributions of the photosensitizer as a function of: (a) time following administration; (b) drug dose; (c) mode of drug administration; and (d) tumor size. The in vivo photosensitizing efficiency of NPe6 was compared directly to Photofrin II in experiments which evaluated tumor responses and induction of normal skin damage. Initial PDT experiments demonstrated that NPe6 was ineffective at inducing tumor cures when a 24-h time interval (between drug administration and light treatment) was used. However, PDT-induced tumor cures were obtained when NPe6 was administered 4-6 h prior to light exposure, and these NPe6-PDT treatment parameters were as effective as standard Photofrin II-mediated PDT. Interestingly, the level of PDT-induced normal skin damage was significantly greater for Photofrin II than for NPe6 at comparable drug and light doses. An analysis of pharmacological data and PDT time interval requirements suggests that plasma concentrations of NPe6 may be a more important predictive factor than tumor tissue levels of the photosensitizer for the production of PDT-mediated tumor cures. The results of this investigation indicate that NPe6 is an effective tumor photosensitizer with in vivo clearance properties that eliminate the side effect of prolonged normal skin photosensitization.

Isolation and initial characterization of mouse tumor cells resistant to porphyrin-mediated photodynamic therapy.

Photodynamic therapy (PDT)-resistant variants of the RIF-1 mouse tumor cell line have been isolated following a protocol of repeated porphyrin incubation and light treatments. Two porphyrin incubation procedures, employing either an extended (16 h) or a short (1 h) incubation, were used in order to obtain cell strains exposed to conditions with differing intracellular photosensitizer localization. Two clones from each PDT porphyrin incubation protocol were selected for in vitro and in vivo analyses based on degree of resistance and plating efficiency. Resistant variants had increased protein content and were larger than the parental RIF-1 cells. In vitro growth rates were similar for all cell strains. Both 16-h PDT-resistant variants exhibited modest resistance to ionizing radiation and one of the 16-h PDT-resistant variants demonstrated increased sensitivity to hyperthermia. The PDT-resistant variants did not exhibit a multidrug resistance phenotype nor did they have altered porphyrin uptake properties. The parental and resistant RIF cells had comparable basal levels of antioxidant enzymes, reduced glutathione and stress proteins, but the number of cells required to produce in vivo tumor growth in 50% of inoculated animals was increased for all PDT-resistant variants. The resistant cells exhibit a stable phenotype and should be useful in studies designed to define PDT mechanisms of action.

Determination of [3H]- and [14C]hematoporphyrin derivative distribution in malignant and normal tissue.

The synthesis and tissue-localizing ability of [14C]- and [3H]hematoporphyrin derivative (HPD) in mice have been described. Tissue levels and distributions were the same for both radioactive compounds, indicating that in vivo tritium exchange did not occur with [3H]HPD. The amount of [14C]HPD or [3H]HPD which localized in the transplanted tumor tissue of mice at various times following i.p. injection (10 mg/kg) was higher than in skin or muscle tissue but was less than in liver, kidney, or spleen tissue. These results tend to disprove the generalization that HPD accumulates in malignant tissue to a higher degree than in all normal tissue. It is also reported that gross visualization of porphyrin fluorescence cannot be correlated with actual tissue concentrations of the dye.

Differential cell photosensitivity following porphyrin photodynamic therapy.

Experiments were performed to determine if differences in porphyrin photosensitivity could be observed for cells with varying efficiency in DNA damage repair, as well as for cells which make up components of the vasculature. Photofrin II is undergoing current clinical evaluation for photodynamic therapy of solid tumors, and therefore the retention, dark toxicity, and photosensitizing effects of this drug on human DNA repair-deficient fibroblasts (ataxia telangiectasia and xeroderma pigmentosum) were compared to normal human fibroblasts. In addition, bovine cells of endothelial, smooth muscle, and fibroblast origin were compared for porphyrin retention, toxicity, and photosensitivity. All human fibroblasts exhibited porphyrin-induced dark toxicity, but there were no significant differences in photosensitization or porphyrin retention for any of these cell lines. However, bovine endothelial cells were considerably more photosensitive than smooth muscle or fibroblast cells treated under identical conditions. All bovine cells accumulated similar levels of porphyrin, and therefore the increased sensitivity of the endothelial cells was not due to differences in porphyrin retention. These results provide additional evidence that nuclear damage and/or repair is not a dominant factor in the cytotoxicity induced by porphyrin photosensitization. In addition, these results indicate that endothelial cell photosensitivity may play a role in the vascular damage observed following photodynamic therapy.

Porphyrin photosensitivity in cell lines expressing a heat-resistant phenotype.

In vitro sensitivity to porphyrin-mediated photodynamic therapy (PDT) has been examined in cell lines resistant to hyperthermia. Parental (HA-1) and heat-resistant (3012) Chinese hamster fibroblasts as well as parental (RIF-1) and temperature-resistant (TR-4, TR-5, and TR-10) mouse radiation-induced fibrosarcoma cells were evaluated for thermal and PDT sensitivity. Quantitative survival curves were generated, and porphyrin uptake properties were obtained for all cell lines. Significant resistance to hyperthermia (45 degrees C for varying exposure periods) was documented for the 3012 and temperature-resistant RIF cell strains when compared with the parent lines. Normal and heat-resistant clones, however, exhibited comparable levels of porphyrin uptake and photosensitivity. Our results indicate that cross-resistance between hyperthermia and PDT is not observed and that members of the Mr 70,000 heat shock protein family (which are elevated in the thermal-resistant cells and which may be associated with the heat-resistant phenotype) do not play a significant role in modulating PDT sensitivity. Mechanisms of in vitro cytotoxicity appear to be different for PDT and hyperthermia even though possible subcellular targets (such as the plasma membrane) and types of damage (protein denaturation) may be similar for the two modalities.

Metabolic properties and photosensitizing responsiveness of mono-L-aspartyl chlorin e6 in a mouse tumor model.

A mouse mammary tumor model was used to evaluate metabolic properties of the photosensitizer mono-L-aspartyl chlorin e6 (NPe6) and to determine the optimal time interval between drug administration and light treatment for effective photodynamic therapy (PDT). Photosensitizer metabolism was evaluated by comparing tissue distribution patterns of NPe6 having 14C atoms positioned on either the tetrapyrrole ring or on the aspartyl residue. High performance liquid chromatographic analysis of photosensitizer extracted from tumor tissue was also obtained as a function of time after drug administration. NPe6 distribution in tissue samples and pharmacological calculations of area under the curve were similar for both forms of [14]NPe6. Likewise, metabolic contaminants of NPe6 were not detected by high performance liquid chromatographic analysis following extraction of the photosensitizer from tumor tissue. Maximal in vivo PDT effectiveness was achieved when light treatments were started within 2 h of drug injection. PDT effectiveness was decreased by 50% when light treatments were initiated 6 h after drug injection and was abolished with a 12-h interval between NPe6 injection and light exposure. Responsiveness to NPe6-mediated PDT was correlated with photosensitizer levels in the plasma but not in tumor tissue. These results show that NPe6 was not metabolized following in vivo administration and that the responsiveness of NPe6 mediated PDT was associated with vascular clearance of the photosensitizer.

Photodynamic therapy mediated induction of early response genes.

Photodynamic therapy (PDT) generates reactive oxygen species which initiate the cytotoxic events of this tumor treatment. We demonstrate that PDT mediated oxidative stress induced a transient increase in the early response genes c-fos, c-jun, c-myc, and egr-1 in murine radiation-induced fibrosarcoma cells. Incubation of exponentially growing cells with porphyrin based photosensitizers in the dark also induced an increase in mRNA levels of early response genes. However, the xanthine photosensitizer, rose bengal, produced increased c-fos mRNA levels only following light treatment. Nuclear runoff experiments confirmed that the induction of c-fos mRNA is controlled in part at the level of transcription. Likewise, a chloramphenicol acetyltransferase reporter construct containing the major c-fos transcriptional response elements was inducible by porphyrin and PDT. Signal transduction pathways associated with PDT mediated c-fos activation were examined by treating cells with protein kinase inhibitors. Staurosporine and 1-(5-isoquinolinesulfonyl)-2-methylpiperazine inhibited PDT mediated c-fos activation while N-(2-guanidinoethyl)-5-isoquinoline-sulfonamide had no effect. In addition, quinacrine, which can inhibit phospholipase activity, blocked PDT induced c-fos mRNA expression. These results suggest that photosensitizer mediated oxidative stress acts through protein kinase-mediated signal transduction pathway(s) to activate early response genes.

Energetics and efficiency of photoinactivation of murine tumor cells containing hematoporphyrin.

Hematoporphryrin derivative at an intracellular concentration in TA-3 mouse mammary carcinoma cells of 0.6 or 0.9 mM required input of 3.0 X 3.6 X 10(9) quanta/cell of red light (620 nm) to achieve a 90% kill. At an intracellular concentration of 1.2 mM, this light requirement drops to 1.5 X 10(9)quanta/cell. The energy for this photodynamic process is about 100 times higher than that required for ionizing radiation to achieve the same level of kill for these cells. The quantum yield for singlet oxygen formation (the cytotoxic agent in most photodynamic processes) from hematoporphyrin derivative is 0.75 +/- 0.07 in ethanol but only 0.16 +/- 0.07 within TA-3 cells, indicating possible intracellular complexing and quenching.

Identification of singlet oxygen as the cytotoxic agent in photoinactivation of a murine tumor.

Singlet oxygen, a metastable state of normal triplet oxygen, has been identified as the cytotoxic agent that is probably responsible for in vitro inactivation of TA-3 mouse mammary carcinoma cells following incorporation of hematoporphyrin and exposure to red light. This photodynamic inactivation can be completely inhibited by intracellular 1,3-diphenylisobenzofuran. This very efficient singlet oxygen trap is not toxic to the cells nor does it absorb the light responsible for hematoporphyrin activation. We have found that the singlet oxygen-trapping product, o-dibenzoylbenzene, is formed nearly quantitatively intracellularly when both the furan and hematoporphyrin are present during illumination but not when only the furan is present during illumination. The protective effect against photodynamic inactivation of the TA-3 cells afforded by 1,3-diphenylisobenzofuran coupled with the nearly quantitative formation of the singlet oxygen-trapping product indicates that singlet oxygen is the probable agent responsible for toxicity in this system.

Glucose regulated protein induction and cellular resistance to oxidative stress mediated by porphyrin photosensitization.

Photodynamic therapy (PDT) utilizes a tumor localizing porphyrin photosensitizer in the clinical treatment of cancer. At a mechanistic level, porphyrin photosensitization generates reactive oxygen species which initiate oxidative damage to a wide spectrum of biomolecules. Cellular stress proteins are also increased following oxidative stress treatments. In the current study, we examined porphyrin photosensitizing parameters associated with induction of the glucose regulated family of stress proteins. Elevated levels of mRNA encoding glucose regulated proteins (GRPs) as well as increases in GRP protein synthesis were observed for mouse radiation induced fibrosarcoma cells exposed to an extended (16-h) porphyrin incubation prior to light exposure. However, a short (1-h) porphyrin incubation prior to light treatment (designed to produce comparable phototoxicity as PDT using the 16-h porphyrin incubation protocol) was associated with only minimal increases in GRP mRNA levels or GRP protein synthesis. The relationship between GRP levels and PDT sensitivity was examined in radiation induced fibrosarcoma cells pretreated with the calcium ionophore A-23187 in order to overexpress GRPs prior to photosensitization. Resistance to PDT was observed in cells overexpressing GRPs only under photosensitizing conditions associated with the extended porphyrin incubation protocol, and this response was not due to changes in cellular porphyrin uptake. In separate experiments, a transient elevation of GRP mRNA levels was observed in transplanted mouse mammary carcinomas following in vivo PDT treatments. Our results indicate that specific targets of oxidative damage (modulated by porphyrin incubation conditions) instead of generalized cellular exposure to reactive oxygen species are correlated with PDT mediated GRP induction. In this regard, GRP induction may be a useful in vivo biochemical marker of PDT mediated injury. These results also support the hypothesis that GRPs may play a role in modulating sensitivity to cellular stresses including certain types of oxidative injury.

Photodynamic therapy sensitivity is not altered in human tumor cells after abrogation of p53 function.

Photodynamic therapy (PDT) is an effective local cancer treatment that induces cytotoxicity through the intracellular generation of reactive oxygen species. The current study investigated whether abrogation of wild-type p53 expression modified the sensitivity of tumor cells to PDT-mediated oxidative stress. In these experiments, human colon (LS513) and breast (MCF-7) carcinoma cells exhibiting a wild-type p53 phenotype were directly compared to LS513 and MCF-7 cells with abrogated p53 function induced by stable integration of the human papillomavirus type 16 E6 viral oncoprotein. The effectiveness of this viral oncoprotein to target p53 for degradation was confirmed using a p53 transactivation reporter gene assay. Western analysis also confirmed attenuated expression of p53 in E6-transfected cells. Photosensitivity of PDT-treated cells was measured by a clonogenic assay and found to be equivalent for parental and p53-abrogated cells. PDT-mediated oxidative stress resulted in a rapid shift of pRb from a hyperphosphorylated form to a predominantly underphosphorylated form in parental cells that was not preceded by increases in p53 or p21 expression. Hypophosphorylated pRb was also observed in PDT-treated LS513/E6 and MCF-7/E6 cells, further indicating that p53 was not involved in this process. Delayed expression of p53 and p21 proteins was seen in parental cells 24-48 h after photosensitization. Cell cycle analysis showed that the abrogation of p53 had minimal effects on an observed PDT-induced G1 block. Rapid induction of apoptosis was documented in PDT-treated LS513 cells, whereas LS513/E6 treated cells exhibited reduced apoptosis in response to PDT. The MCF-7 cell lines exhibited a minimal apoptotic response to PDT. These results indicate that p53 expression does not directly modulate tumor cell sensitivity to PDT in either apoptosis-responsive (LS513) or nonresponsive (MCF-7) cells.

Comparison of mutagenicity and induction of sister chromatid exchange in Chinese hamster cells exposed to hematoporphyrin derivative photoradiation, ionizing radiation, or ultraviolet radiation.

Cell culture studies have been performed to compare the mutagenic potential and the induction of sister chromatid exchanges for hematoporphyrin derivative photoradiation, ionizing radiation, and UV radiation. The mutation frequency in Chinese hamster ovary cells at the hypoxanthine-guanine phosphoribosyltransferase locus was measured using resistance to 6-thioguanine. Phenotypic expression time prior to mutation selection was also examined. Treatment with either X-rays or UV was effective in producing mutants resistant to 6-thioguanine, but treatment with hematoporphyrin derivative photoradiation (at comparable toxicity levels) did not induce any mutagenic activity above background levels. The hematoporphyrin derivative incubation and photosensitization conditions used in this study did induce sister chromatid exchanges at frequencies comparable to those induced by X-rays but at lower frequencies than for UV treatments.

In vitro and in vivo light dose rate effects related to hematoporphyrin derivative photodynamic therapy.

In vitro and in vivo experiments were performed to evaluate the parameter of light dose rate as it relates to the efficiency of hematoporphyrin derivative (HPD)-induced photosensitization. Exponentially growing Chinese hamster ovary cells were incubated with HPD (25 micrograms/ml) and were then exposed to red light (630 nm) delivered at different dose rates. A total of five dose rates (0.5, 5.0, 15, 23, and 60 milliwatts/sq cm) were examined following a 1-hr HPD incubation, two dose rates (1 and 20 milliwatts/sq cm) were examined after a 12-hr HPD incubation, and three dose rates (0.4, 4, and 40 milliwatts/sq cm) were examined following a 16-hr incubation and a 30-min serum wash protocol. The effect of light dose rate was determined from cell survival curves obtained by standard clonogenic colony formation assays. Similar levels of cellular toxicity were obtained when cells from each HPD incubation group were treated with equal doses of red light delivered at different dose rates. For in vivo experiments, albino mice were given injections of HPD (7.5 mg/kg) and 24 h later the right hind leg of each mouse was treated with localized red light (630 nm). A total dose of 270 J/sq cm was delivered to the right hind leg at dose rates of 5, 25, or 125 milliwatts/sq cm. The resulting acute skin damage induced by HPD photosensitization was scored over a 30-day period, and skin response curves for the three dose rates were obtained. Comparable levels of damage were induced in each of the three experimental groups. The results obtained from both in vitro and in vivo studies indicate that the photosensitizing efficiency of HPD photodynamic therapy is not affected by nonthermal variations in clinically relevant dose rates of delivered light.

Expression of potentially lethal damage in Chinese hamster cells exposed to hematoporphyrin derivative photodynamic therapy.

Experiments were performed to determine whether the expression and/or repair of potentially lethal damage could be observed in mammalian cells exposed to hemataporphyrin derivative (HPD) photodynamic therapy (PDT). Photodynamic therapy was combined with posttreatment protocols known to inhibit the repair of potentially lethal damage in cells treated with X-rays, ultraviolet radiation, or alkylating agents. Potentiation of lethal damage from photodynamic therapy was induced by hypothermia (4 degrees C) following short (1 h) or extended (16 h) HPD incubation conditions. Caffeine potentiated the lethal effects of PDT only when cells were incubated with HPD for extended time periods. However, 3-aminobenzamide had no effect on the cytotoxic actions of PDT following either short or extended HPD incubations. Recovery from potentially lethal damage expressed by posttreatment hypothermia was complete within 1 h, while recovery from potentially lethal damage expressed by posttreatment caffeine required time periods of up to 24 h. The lack of effect of 3-aminobenzamide on expression of potentially lethal damage following photodynamic therapy may be related to direct inhibition of adenosine diphosphoribose transferase by photodynamic therapy. These results indicate that the expression and repair of potentially lethal damage can be observed in cells treated with PDT and will vary as a function of porphyrin incubation conditions.

Decreased expression and function of alpha-2 macroglobulin receptor/low density lipoprotein receptor-related protein in photodynamic therapy-resistant mouse tumor cells.

Parental and photodynamic therapy (PDT)-resistant mouse, radiation-induced fibrosarcoma cell lines were evaluated using mRNA differential display in an attempt to identify unique transcripts. We detected one transcript that was consistently present in the parental cells but absent in PDT-resistant cells. The transcript was cloned, sequenced, and identified as alpha-2 macroglobulin receptor/low density lipoprotein receptor-related protein (alpha-2 MR/LRP). Northern and Western immunoblot analysis confirmed that receptor expression was present in the parental cell line but barely detectable in PDT-resistant cells. Functionality of the receptor was evaluated by exposing cells to Pseudomonas exotoxin A. alpha-2 MR/LRP is responsible for Pseudomonas exotoxin A internalization, and only the parental cells exhibited toxin-mediated cytotoxicity. The binding and endocytosis of activated alpha-2 macroglobulin and lipoproteins by alpha-2 MR/LRP are consistent with modulating uptake and localization of photosensitizers. Our results demonstrate that PDT-resistant murine tumor cells exhibit minimal alpha-2 MR/LRP activity and suggest that this receptor plays a role in PDT sensitivity by modulating photosensitizer uptake and/or subcellular localization.

Hematoporphyrin derivative photoradiation therapy for the treatment of intraocular tumors: examination of acute normal ocular tissue toxicity.

Preclinical studies designed to define potential side effects resulting from the use of hematoporphyrin derivative (HPD) photoradiation therapy (PRT) as a modality for treating intraocular tumors have been performed. Pigmented rabbits were used to evaluate acute normal ocular tissue toxicity following single HPD PRT treatments in which the light was directed through the pupil and onto a 1-sq cm area of the retina. The treatment procedure consisted of the i.v. administration of HPD (1 to 10 mg/kg) followed 48 hr later by a 15-min exposure of localized red light [635 +/- 5 nm; 40 to 400 milliwatts/sq cm] generated by a free running rhodamine B dye laser pumped by a 5-watt argon laser. Toxicity to normal retinal tissue was documented using fundus photography, fluorescein angiography, and histological examination. The results of this study demonstrated that ocular damage in the form of retinal edema, detachment, and necrosis could be induced by clinically relevant doses of HPD PRT. The area of retinal damage was limited to the treatment field in all but the highest doses of HPD PRT. The histological results were in agreement with the visual observations in that abrupt and demarcated transition areas between injured and normal-appearing retina were observed. Care will have to be used in the delivery of light to the treatment field if HPD PRT is to be utilized for treatment of intraocular tumors.

Photodynamic therapy of intraocular tumors: examination of hematoporphyrin derivative distribution and long-term damage in rabbit ocular tissue.

Studies were performed to determine the distribution of hematoporphyrin derivative (HPD) in ocular structures and to characterize long-term damage associated with ocular HPD photodynamic therapy. Pigmented rabbits with an amelanotic melanoma heterotransplanted to the iris were used to obtain quantitative tissue levels of HPD as well as to document HPD localization by fluorescence microscopy. HPD was administered i.v., and tissue concentrations of HPD were determined by spectrofluorometry following porphyrin extraction. Vascular structures such as the tumor, iris, and choroid-retina as well as the aqueous fluid from eyes containing tumors demonstrated rapid HPD localization. The sclera had minimal HPD uptake, and the drug was not detected in avascular structures such as the lens or cornea. HPD was cleared from all ocular structures except the tumor and choroid-retina by 24 h following injection. Fluorescence microscopy data indicate that HPD remained in the avascular photoreceptor cell outer segments of the retina. Long-term damage was documented in rabbits which received HPD photodynamic therapy to a 1-sq cm area of retina via transpupillary light delivery. Acute damage to the exposed area of retina (in the form of a chorioretinal scar) could be induced. This damage was permanent but not progressive. Lens opacities were not observed, and the cornea, aqueous, and vitreous remained clear on all test eyes. The results from these studies suggest that HPD photodynamic therapy may provide a selective and safe approach to the treatment of ocular tumors.

Photodynamic therapy-mediated oxidative stress as a molecular switch for the temporal expression of genes ligated to the human heat shock promoter.

Oxidative stress associated with photodynamic therapy (PDT) is a transcriptional inducer of genes encoding stress proteins, including those belonging to the heat shock protein (hsp) family. The efficiency of PDT to function as a molecular switch by initiating expression of heterologous genes ligated to the human hsp promoter was examined in the present study. Selective and temporal reporter gene expression was documented after PDT in mouse radiation-induced fibrosarcoma cells stably transfected with recombinant vectors containing an hsp promoter ligated to either the lac-z or CAT reporter genes and in transfected radiation-induced fibrosarcoma tumors grown in C3H mice. Hyperthermia treatments were included as a positive control for all experiments. Expression vectors containing either human p53 or tumor necrosis factor (TNF)-alpha cDNA under the control of an hsp promoter were also constructed and evaluated. A p53 null and TNF-alpha-resistant human ovarian carcinoma (SKOV-3) cell line was stably transfected with either the p53 or TNF-alpha constructs. Inducible expression and function of p53 as well as inducible expression, secretion, and biological activity of TNF-alpha were documented after PDT or hyperthermia in transfected SKOV cells. These results demonstrate that PDT-mediated oxidative stress can function as a molecular switch for the selective and temporal expression of heterologous genes in tumor cells containing expression vectors under the control of an hsp promoter.