Subcellular Phototoxicity of Photofrin-II and Lutetium Texaphyrin in Cells In Vitro

. Three cell types including bovine pulmonary artery endothelium cells (CPAE), rat kangaroo kidney cells (PTK2), and human larynx epidermoid carcinoma cells (Hep-2) were used to study subcellular localisation and phototoxicity of Photofrin-II and lutetium texaphyrin (Lu Tex). Cells were examined for fluorescence after administration of the photosensitisers. Subcellular regions were exposed with a laser microbeam system that used an argon ion laser pumped dye laser generating a 630 nm for Photofrin-II and 730 nm for Lu Tex. Fluorescence detection suggests that the Photofrin-II is bound primarily to the mitochondria with some diffuse fluorescence in the rest of the cytoplasm. The fluorescence in Lu Tex treated cells appears to be localised to the lysosomes. The percentage of damaged cells following light exposure to the different subcellular regions after Photofrin-II or Lu Tex treatment demonstrates that the nuclear region was the most sensitive target followed by the perinuclear region and peripheral cytoplasm region.


INTRODUCTION
Since the initial studies on photodynamic therapy (PDT) in 1972 [1,2], an enormous amount of basic and clinical research has suggested that PDT has the potential to be a significant cancer treatment modality [3,4]. E$cacy of PDT is related to the selective accumulation of the photosensitiser within the tumours and the subsequent induction of cytotoxic singlet oxygen following exposure to light of the appropriate wavelength. To date, most clinical PDT has used 630 nm light with the first generation sensitiser, Photofrin-II.
One aspect of the tumouricidal e#ectiveness of PDT is related to the depth of light penetration within the tumour mass. The e#ective penetration at 630 nm is 1-3 mm, whereas at 700-850 nm at least 6 mm penetration of light is observed [5]. A growing number of secondgeneration photosensitisers with longer wavelength absorption peaks are being synthesised. The texaphyrins are tripyrrolic pentaaza expanded porphyrins, which exhibit strong, low energy optical absorption in the 730-770 nm range [6]. One of these compounds, lutetium texaphyrin (Lu Tex) is a pure, water-soluble photosensitiser with a broad absorption band centred at 732 nm [7].
One of the suggested mechanisms of PDT destruction of tumours is that the vascular endothelium is damaged initially and tumour cells are destroyed secondarily as a result of structural damage to capillaries and functional disturbance in the microcirculation [8][9][10].
Within the individual cell, the subcellular sites of photodynamic damage may include the plasma membrane, lysosomes, and the mitochondria [11,12]. Biochemical analysis indicates that membrane bound mitochondrial enzymes such as cytochrome c, oxidase and succinate dehydrogenase are inactivated by PDT [13][14][15]. Damage to membranes of the endoplasmic reticulum also has been observed ultrastructurally [16]. In all of these studies, the cell damage was assayed by either microscopy or biochemical analysis after light exposure to the entire cell or cell population. Consequently, it was di$cult to determine if the observed subcellular damage was due to a primary light plus drug e#ect in the damaged organelle or cell region, or due to secondary e#ects subsequent to absorption of light at a primary site.
In a recent study using a laser microbeam to expose subcellular sites, it was possible to demonstrate that the perinuclear cytoplasm was highly photosensitised. This observation correlated with photosensitiser fluorescence in that region. Surprisingly, it was also determined that the nucleus was a very sensitive target following laser microbeam irradiation [17]. In the present study we have expanded these early studies to examine the subcellular phototoxicity of Photofrin-II, the most frequently used photosensitiser, and Lu Tex, a second generation photosensitiser.
In the present study, we exposed specific subcellular regions of photosensitised cells to light using a laser microbeam microscope system [17]. By correlating subcellular photosensitiser fluorescence, the region of light exposure, and subsequent cell response, we are able to achieve a more complete understanding of PDT-induced cytotoxicity.

Photosensitisers
A stock solution of Photofrin-II (Quadra Logic Technologies, Inc., Vancouver, British Columbia, Canada) was prepared according to the manufacturer's directions using sterilised 5% dextrose to give a final concentration of a 2.5 mg/ml. Lu Tex (Pharmacyclic Corp., Sunnyvale, CA, USA) was prepared in 5% mannitol and filter sterilised.
Just prior to cell exposure to the drug, culture media containing 0.15 g/ml of Photofrin-II or 1 g/ml of Lu Tex were prepared in the dark. The experimental cells were treated with either Photofrin-II or Lu Tex in the dark. After 4 h the medium containing photosensitiser was replaced by fresh medium free of photosensitiser.

Cells and Cell Culture
Three cell types were used in these studies.
Bovine pulmonary artery endothelium cells (CPAE) were obtained from the American Type Culture Collection (ATCC). This is an endothelial cell line derived from the main stem pulmonary artery of a young, female cow (Box taurus). The cells were grown in Eagle's minimal essential medium with 20% fetal bovine serum at 37 C and 5-7% CO 2 atmosphere. The cells were maintained and subcultured for 2-3 weeks before being discarded and replaced with fresh cells from the same passage.
Rat kangaroo (Potorous tridactylis) kidney cells (PTK 2 ) were originally obtained from the ATCC and grown as a monolayer in modified Eagle's minimal essential medium with 10% fetal bovine serum. The cells were subcultured once a week into T-25 culture flask and maintained in an incubator at 37 C with 5-7% CO 2 [18]. These cells have been growing in our lab for over 25 years and their flat nature facilitates selective microbeam exposure to subcellular targets.
Hep-2 cells from human larynx epidermoid carcinoma were originally obtained from the ATCC and grown as a monolayer in Eagle's minimum essential medium with 2 mM -glutamine and Earle's BSS adjusted to contain 1.5 g/l sodium bicarbonate, 0.1 mM nonessential amino acids, and 1.0 mM sodium pyruvate. The medium contained 10% fetal bovine serum. As with the CPAE and PTK 2 cells, these cells are relatively flat facilitating laser microbeam irradiation of selective subcellular regions.
Twenty-four hours prior to the experiments, cells were injected into standard Rose culture chambers at a density of 1 10 3 cells/ml. Prior to laser microirradiation, the cells were exposed to fresh culture medium containing 0.15 g/ml Photofrin-II or 1 g/ml Lu Tex for 4 h, and then exposed to fresh medium free of sensitiser. Fresh photosensitiser-free medium was applied to the cells approximately ten minutes before microirradiation. Cells exposed to the laser but not the sensitiser served as controls. In addition, cells treated with the photosensitiser and exposed to microscope illumination, but not exposed to laser irradiation, served as controls.
The chambers were covered with aluminium foil immediately after injecting the photosensitiser in order to protect the cells from ambient light. Single cells were selected for laser microirradiation that were normal in size, well attached to cover glass, and exhibiting healthy morphology (minimal number of cytoplasmic vacuoles). A Zeiss diamond marking objective was used to circle each preselected cell by etching a small circle (at 1 mm diameter) around the desired cell. Cells that did not have any neighbouring cells within the marked circle were chosen for irradiation. The same criteria for cell selection was applied to experimental and control cells.

Fluorescence Microscopy
A Zeiss Axiovert inverted fluorescence microscope was used to visualise subcellular fluorescence of the two sensitisers. Cells were examined for fluorescence at 1 h intervals for 4 h after the initial administration of the photosensitiser. For Photofrin-II, each cell was excited at 365 nm. For Lu Tex, a wavelength of 470 nm was used. In both cases excitation filters were employed with a 75 W arc lamp as light source. An image was recorded of the fluorescence within the cell using a cooled charge-couple device (CCD) (Princeton Instruments TE/CCD-576E/UV) and stored in IP Lab format in a Macintosh IIfx computer. A phase-contrast picture of each cell was also taken for reference. Images were digitally processed improving contrast and reducing background light.

Microirradiation of Subcellular Components
A laser microbeam system was used that employed an argon ion laser (Coherent Innova 90) pumped dye laser (Coherent 599 dye) using {2-{2-{4-(dimethylamino)phenyl}ethanol}-6-methyl-4H-pyran-4-ylidene}-propanedinitrile (DCM) to generate a 630 nm beam for Photofrin-II and LDS 722 (pyridine 2) to generate a 730 nm beam for Lu Tex. Each of these wavelengths matched the absorption peak of the respective sensitisers. The beam was directed through a Zeiss Axiovert inverted fluorescence microscope and focused to a near di#raction limited spot using a Zeiss Neofluar 100 phase-contrast objective with a numerical aperture of 1.3. For each cell, the laser beam was focused on the centre of the nucleus, the perinuclear cytoplasm or the peripheral cytoplasm. The laser spot was approximately 0.5 m in diameter for the 630 nm microbeam irradiation and 0.6 m for the 730 nm microbeam. A Newport Corporation model 1815 optical power meter and power detector model 818-UV (Irvine, CA) were used to measure laser power at the entrance to the objective. To determine the actual power reaching the irradiated sample, the dual objective transmittance measuring technique of Misawa et al. [19] was used. This method eliminates internal reflection errors that are encountered in a direct objective-to-power measurement in air. Using this method it was determined that each cell was exposed to 9 mW of power at a power density of 4.6 10 6 W/cm 2 for Photofrin-II and 3.2 10 6 W/cm 2 for Lu Tex. The total irradiance (energy densities) per experiment were obtained by varying exposure time. The parameters used are listed in Table 1 and Table 2.
Immediately before laser irradiation, a phase-contrast image of the cell was made using the CCD camera. Each chamber was labelled and placed in the CO 2 incubator after the irradiation. Each cell was followed and recorded at 24 h and 48 h after irradiation.

Fluorescence Detection
The CCD camera system provided images of subcellular fluorescence in photosensitiser-  treated and untreated control cells. Figure 1(a) is a phase-contrast image of CPAE cells 4 h after exposure to medium containing 0.15 g/ ml of Photofrin-II. Figure 1(b) is an image of the same cell demonstrating a large amount of fluorescence in the perinuclear area that extents towards the periphery of the cell.    It is evident that in both Lu Tex-treated CPAE and PTK 2 cells, fluorescence was dispersed throughout the cytoplasm and localised primarily in the lysosomes. Control CPAE and PTK 2 cells show either no detectable or weak autofluorescence.
In summary, the pattern of fluorescence of Photofrin-II is di#erent from that of Lu Tex. The fluorescence detected in Photofrin-II treated cells suggests that the sensitiser is bound primarily to the mitochondria with some di#use fluorescence in the rest of the cytoplasm. However, the fluorescence in Lu Tex treated cells appears to be localised to the lysosomes. In all three cell types treated with either Photofrin-II or Lu Tex, no fluorescence was detected in the nucleus.

Subcellular Phototoxicity
A laser power of 9 mW (2.3 10 6 W/cm 2 ) at the microscope focal point has been used in the subcellular microirradiation experiments. The total energy densities (ED) were varied by changing the duration of laser exposure (Tables 1 and 2). In the Photofrin-II experiments, a total of 242 CPAE cells were     irradiated in the following subcellular regions: nucleus, n=73; perinuclear cytoplasm, n=99; peripheral cytoplasm, n=70. For PTK 2 cells, a total of 258 cells were irradiated: nucleus, n=84; perinuclear cytoplasm, n=124; peripheral cytoplasm, n=50. For Hep-2 cells, a total of 262 cells were irradiated: nucleus, n=109; perinuclear cytoplasm, n=113; peripheral cytoplasm, n=40.
In the Lu Tex experiments, a total of 211 CPAE cells were irradiated: nucleus, n=55; perinuclear cytoplasm, n=69; peripheral cytoplasm, n=87. A total of 229 PTK 2 cells were irradiated: nucleus, n=62; perinuclear cytoplasm, n=95; peripheral cytoplasm, n=72. The representative example of the locations of the laser microspot in each cell region, and the cell status 24 h after light exposure are presented at Figs 6-10.
The raw data of cell status at 24 and 48 h after microirradiation for Photofrin-II are shown in Tables 3-5 and in Tables 6 and 7 Figure 11 is a graph of the combined data depicting the percentage of damaged CPAE cells after light exposure to the di#erent subcellular regions after Photofrin-II treatment. All the cells in the three di#erent subcellular irradiation groups were killed by the 300 s exposure, i.e. ED=138 10 7 J/cm 2 . At lower light doses the region that demonstrated the greatest sensitivity was the nucleus, followed by the perinuclear cytoplasm, and peripheral cytoplasm. Figure 12 summarises the results for PTK 2 cells after Photofrin-II treatment. It was interesting that there was no significant di#erence between irradiation of the nuclear region and the perinuclear cytoplasm region though both of these regions were more sensitive than the peripheral cytoplasm. Figure 13 presents the results of Hep-2 cells. The higher sensitivity of the nucleus over the perinuclear cytoplasm was evident with an exposure of 20-30 s, i.e. ED=9.2 10 7 -13.8 10 7 J/cm 2 . No di#erences were observed at the higher doses (up to 60 s exposure) or lower doses (down to 5-10 s exposure).
The results with Lu Tex are presented in Figs 14 and 15 for CPAE and PTK 2 cells, respectively. The results indicate that in both cell types, the nuclear region was the most sensitive target followed by the perinuclear region and the peripheral cytoplasm region.
No adverse e#ects to cell survival were observed in CPAE, Hep-2 and PTK 2 control cells treated by microscope illumination alone. Also, no light-alone control exposures in these three cell types resulted in cell damage even with the maximum total energy doses (Tables  3-7). In addition, it was found that the Lu Tex-treated CPAE and PTK 2 cells were more sensitive to light exposure than Photofrin-II. For example, the ED for nucleus irradiation needed to cause 100% CPAE and PTK 2 cell damage using Lu Tex were 1.6 10 7 J/cm 2 (5 s exposure) and 3.2 10 7 J/cm 2 (10 s exposure), respectively. Whereas with Photofrin-II, 82.8 10 7 J/cm 2 (180 s) and 13.8 10 7 J/cm 2 (30 s) were required to induce cell damage.

DISCUSSION
Early cell studies using haematoporphyrin derivative (HPD) demonstrated a perinuclear pattern of fluorescence due primarily to mitochondria, with little or no fluorescence in the peripheral cytoplasm or the nucleus [11]. In a more recent study on the subcellular phototoxicity of 5-aminolaevulinic acid (ALA) [17], it was found that ALA induced endogenous protoporphyrin IX production in CPAE, PTK 2 and primary culture neonatal rat myocardial cells as exhibited by fluorescence primarily confined to the mitochondria-rich perinuclear cytoplasm. A study by Uberriegler et al. [20] indicated that ALA treatment led to a bright fluorescing perinuclear region in W138 cells. The present fluorescence study with Photofrin-II together with these earlier studies confirm that this photosensitiser results in a perinuclear pattern of fluorescence. It has been reported that after  5  24  19  79  15  63  Yes  3  8  7  89  6  75  Yes  1  10  10  100  10  100  Peripheral cytoplasm  No  300  10  10  100  10  100  Yes  300  10  0  0  0  0  Yes  180  10  7  70  7  70  Yes  60  10  10  100  9  90  Yes  30  10  10  100 10 100 a Healthy cell: cells underwent at least one additional mitosis and cells with no division but morphologically normal, alive and healthy.
haematoporphyrin-mediated PDT, fluorescence and electron microscopy showed immediate structural changes in mitochondria, with progressive swelling and destruction of these organelles [13,14].
Our experimental results indicate that the light dose needed to kill the cells is less for Lu Tex than for Photofrin-II. Cells treated with either of these compounds are far more sensitive to light than cells exposed to ALA. Photofrin-II and Lu Tex are incorporated directly into cells and rapidly bind to mitochondria, lysosomes and other cytoplasmic components. ALA is converted intracellularly into monomeric protoporphyrin IX which is the photoactive species. Di#erences in cell phototoxicity between the three compounds may also be due to di#erences in cellular concentration of the compounds as well as the quantum e$ciency of singlet oxygen generation.
The present study confirms previous work on the subcellular phototoxicity of ALA [17], in which it was possible to demonstrate a correlation between perinuclear sensitiser fluorescence and subcellular site of phototoxicity. However, of greater interest is the finding that the nucleus is a very sensitive site despite the fact that most of the literature and the fluorescence localisation data suggest that the nucleus is not a primary target site of PDT.     This study validates the hypothesis that subcellular fluorescence can be correlated with cellular target site, i.e. perinuclear cytoplasm versus peripheral cytoplasm. However, the results demonstrating selective nuclear sensitivity raises basic questions with respect to localisation of photosensitiser in the nucleus and the role of the nucleus as a primary subcellular target. It is possible that the photosensitiser may be present at a concentration below the detection of the CCD camera system. Under this condition, it would be possible that the genetic material in the nucleus may be sensitive to even small amounts of generated singlet oxygen.

CONCLUSIONS
The fluorescence detection in CPAE, PTK 2 , and He-2 cells after administration of photosensitisers demonstrated that the Photofrin-II is bound primarily to the mitochondria with some di#use fluorescence in the rest of the cytoplasm. However, the Lu Tex appears to be localised to the lysosomes. The percentages of damaged cells following light exposure to the di#erent subcellular regions after Photofrin-II or Lu Tex treatment indicated that the nuclear region was the most sensitive target followed by the perinuclear and peripheral cytoplasmic region.