Novel photosensitizers trigger rapid death of malignant human cells and rodent tumor transplants via lipid photodamage and membrane permeabilization

Abstract

Background: Apoptotic cascades may frequently be impaired in tumor cells; therefore, the approaches to circumvent these obstacles emerge as important therapeutic modalities.

Methodology/principal findings: Our novel derivatives of chlorin e(6), that is, its amide (compound 2) and boronated amide (compound 5) evoked no dark toxicity and demonstrated a significantly higher photosensitizing efficacy than chlorin e(6) against transplanted aggressive tumors such as B16 melanoma and M-1 sarcoma. Compound 5 showed superior therapeutic potency. Illumination with red light of mammalian tumor cells loaded with 0.1 µM of 5 caused rapid (within the initial minutes) necrosis as determined by propidium iodide staining. The laser confocal microscopy-assisted analysis of cell death revealed the following order of events: prior to illumination, 5 accumulated in Golgi cysternae, endoplasmic reticulum and in some (but not all) lysosomes. In response to light, the reactive oxygen species burst was concomitant with the drop of mitochondrial transmembrane electric potential, the dramatic changes of mitochondrial shape and the loss of integrity of mitochondria and lysosomes. Within 3-4 min post illumination, the plasma membrane became permeable for propidium iodide. Compounds 2 and 5 were one order of magnitude more potent than chlorin e(6) in photodamage of artificial liposomes monitored in a dye release assay. The latter effect depended on the content of non-saturated lipids; in liposomes consisting of saturated lipids no photodamage was detectable. The increased therapeutic efficacy of 5 compared with 2 was attributed to a striking difference in the ability of these photosensitizers to permeate through hydrophobic membrane interior as evidenced by measurements of voltage jump-induced relaxation of transmembrane current on planar lipid bilayers.

Conclusions/significance: The multimembrane photodestruction and cell necrosis induced by photoactivation of 2 and 5 are directly associated with membrane permeabilization caused by lipid photodamage.

Figure 1. Potency of compounds 2 and 5 as antitumor photosensitizers.

Mice bearing s.c. transplant of B16 melanoma (A) or rats bearing s.c. transplant of M-1 sarcoma (B) were injected i.p. with PBS or 5 mg/kg of 2 or 5. Ninety minutes after injection the tumors were illuminated with red light (see MATERIALS AND METHODS). The tumor volume was measured at indicated time intervals. One representative experiment out of a total of three is shown. *p<0.01 between the ‘compound 5’and the ‘compound 2’ groups.

Figure 2. Intracellular distribution of compound 5.

The C6 rat glioma cells were stained with the Golgi marker BODIPY FL C5–ceramide (A–C), the ER marker brefeldin A, BODIPY FL (D–F) the lysosomal dye LysoTracker®Green DND-99 (G–I), or the nuclei marker SYBR Green (J–L). Then 0.3 µM of 5 was added for 15 min. After removal of the drug cells were analyzed by confocal scanning microscopy. A, D, G and J: autofluorescence of 5. B, Golgi marker; E, ER marker; H, lysosomal marker, K, nuclei marker, C, F, I and L: overlays of A and B, D and E, G and H and J and K, respectively. Shown are the optical sections at the midst of the nucleus. The experiments were performed at 37°C, 5%CO2. In A–I, insets in the upper right corners show Golgi, ER and lysosomes at a higher magnification. Bar, 10 µm.

Figure 3. Time course of death-associated events after LI of cells loaded with compound 5.

The C6 rat glioma cells were incubated with 1 µM of 5 for 15 min. at 37°C, 5%CO₂. A–D: fluorescence of 5. E–H: ROS generation detected with DCFH-DA (4 µM, 5 min.). I–K: changes of mitochondrial morphology and a decrease of ΔΨ_m detectable with MitoTracker Red CM-H₂XRos (0.5 µM, 5 min.). L, staining with PI (15 µM). Fluorescence of MitoTracker Red CM-H₂XRos and PI was detected in the same channel. M: overlay of A, E and I. N: overlay of B, F and J. O: overlay of C, G and K. P: overlay of D, H and L. Shown are optical sections. Bar, 20 µM.

Figure 4. Fragmentation of mitochondria after PDT with compound 5.

Top panel: phase contrast, bottom panel: fluorescence of MitoTracker Red CM-H₂XRos. Time after LI is shown in upper left corners. Arrow, intact mitochondria; arrowheads, fragmented mitochondria. Bar, 5 µm.

Figure 5. Sodium azide and trolox inhibit the increase of intracellular ROS level after PDT with compound 5.

The C6 glioma cells were loaded for 15 min with DCFH-DA and 5 in the absence (A–D) or presence of 10 mM NaN₃ (E–H) or 10 µM trolox (I–L). A, C, E, G, I and K, phase contrast; B, D, F, H, J and L, epifluorescence. Images were taken before (upper panel) or after (bottom panel) LI. Note green staining in D and its disappearance in H and L. Bar, 10 µm.

Figure 6. Oxidative photodamage of liposomes in the presence of compound 5.

The SRB-containing EggPC liposomes were suspended in PBS and either mock-treated or exposed to red light in the presence of 350 nM of 5 alone or 5 and 10 mM NaN₃. A line bar and an arrow indicate the time of LI and the addition of 0.1% Triton X-100, respectively. Liposome photodamage was registered as an increase of SRB fluorescence in the extra-liposomal milieu. Shown is one representative experiment out of 3 independent replicates.

Figure 7. Time- and concentration dependence of liposome permeabilization.

The SRB-loaded EggPC liposomes were suspended in PBS and illuminated in the presence of 35 nM of 5 (curve 1) or chlorin e₆ (curve 2) (top panel; one representative experiment out of 4 with similar results). Bottom panel: dependence of SRB efflux (% at 100 s after LI) on indicated concentrations of each compound (bottom panel; mean ± S.D. of 3 experiments). Differences between the values of SRB efflux for the respective concentrations of compounds are statistically significant (p<0.01).

Figure 8. Dependence of liposome photodamage on lipid composition.

SRB-liposomes made of EggPC (curve 1) or DPhPC (curve 2) were resuspended in PBS and exposed to red light in the presence of 350 nM of 5. Note SRB leakage from EggPC- but not from DPhPC-containing liposomes. Shown is one representative experiment out of three independent replicates.

Figure 9. Differential permeability of a lipid membrane for compounds 2 and 5.

Shown are time courses of electrical current after application of a voltage jump of V = 100 mV (at t = 0, ‘on’ response) in the presence of 10 nM of 5 or 2 and after switching off the voltage to zero (at t = 0.36 sec, ‘off’ response). Planar BLMs were formed from DPhPC (see MATERIALS AND METHODS). The experiments were performed in the buffer containing 10 mM Tris-HCl, pH 7.4 and 100 mM KCl. Shown is one representative experiment out of 3 with essentially the same results.