Death Pathways Associated with Photodynamic Therapy

Abstract

This report describes studies involving ER vs. lysosomal targeting and is designed to assess the initiation of different death pathways as a function of sub-cellular targeting and PDT dose. Photodamage directed at mitochondria or lysosomes initiates apoptosis, a death pathway generally considered to be irreversible. Photodamage that involves the ER can lead to another death pathway termed paraptosis. This does not involve caspase activation, can eradicate cell types with impaired apoptosis; at high levels of irradiation, apoptosis and necrosis were observed. Autophagy has a cytoprotective function unless lysosomes are targeted; loss of lysosomal integrity can interfere with the autophagic recycling processes.

INTRODUCTION

Photodynamic therapy (PDT) is a procedure for the selective eradication of neoplastic cells. This involves a photosensitizing agent that preferentially localizes in such cell types, light at a wavelength that corresponds to an absorbance band of the photosensitizer and molecular oxygen. Because of tissue optics, wavelengths at 630 - 750 nm are needed for this purpose. At shorter wavelengths, photons will be scattered and absorbed and tissue penetration is limited. Successful PDT involves both direct tumor killing, shutdown of the tumor vasculature and immunologic effects [1,2]. Death pathways are examined here using the phenothiazine EtNBS to selectively photosensitize lysosomes [3] and BPD (benzoporphyrin derivative, Visudyne) to target ER > mitochondria [4].

The pathway to cell death is a function of sites of sub-cellular photodamage. The principal consequences of PDT are apoptosis, paraptosis, necrosis and autophagy [5,6], with the latter generally considered to be cytoprotective. Necrosis can also occur at higher PDT doses. The latter is usually associated with loss of tissue specificity since most ‘normal’ cell types will also accumulate levels of photosensitizing agents sufficient to cause photodamage at high light doses [2].

We have shown that paraptosis can be a pertinent death pathway even when apoptosis is impaired [7,8]. More recently, data was obtained indicating that PDT doses required for tumor eradication, i.e., in the LD₉₀ to LD₉₉ range, can shift the mechanism of paraptosis from the ‘canonical’ pathway, i.e., the route initiated by certain drugs or biochemical manipulations [9-11]. What can be termed the ‘non-canonical’ route to paraptosis is found at the higher, more clinically-relevant PDT doses required for tumor eradication. Unlike the drug-induced pathway, ‘non-canonical’ paraptosis does not require a brief interval of protein synthesis or action of kinases that are involved in the ‘canonical’ route. At very high PDT doses, sufficient to result in cross-linking of ER proteins, we have observed that paraptosis can be impaired with cell death then associated with apoptosis and/or necrosis [12].

MATERIALS AND METHODS

Chemicals and supplies.

BPD (benzoporphyrin derivative, Verteporfin) was purchased from VWR (Cat No 1711461), Radnor PA. The phenothiazine 5-ethylamino-9- diethyl-aminobenzo [a]phenothiazinium chloride (EtNBS) was provided by Dr. Conor Evans, Wellman Labs, Massachusetts General Hospital/Harvard Medical School. Its preparation has been described [13]. Fluorescent probes were purchased from Thermo Fisher Scientific, Waltham MA. Other reagents were obtained from MilliporeSigma, St. Louis MO.

Cell culture and clonogenic assays.

OVCAR-5 cells were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum..The procedure for estimating photokilling by clonogenic assay has been described [6-8]. Viability was assessed by colony counting using an Oxford Optronix CelCount device. All experiments were performed in triplicate.

PDT protocols.

Cells were grown on 30 mm cover slips in sterile plastic dishes and treated with 0.5 μM BPD or 0.25 μM EtNBS for 1 h. Stock solutions were prepared in DMSO at 1 mM concentrations. The medium was replaced and dishes irradiated with a 600-watt quartz-halogen source. The light beam was passed through a 10 cm layer of water to remove IR and the bandwidth confined by an interference filter (Oriel, Stratford CT) to 690 ± 10 nm (BPD) or 650 ± 10 nm (EtNBS). Light doses were adjusted to provide specified photokilling effects as established by clonogenic analysis on triplicate dishes. Images were acquired 24 h after irradiation.

Microscopy.

To assess patterns of nuclear alterations, cover slips were labeled with Hö33342. Localization of photosensitizers was assessed using ErTracker Green (ErTrG or Lysotracker Green (LTG) as previously described [3,4,6,7,12]. Phase-contrast and fluorescence Images were acquired with a Nikon E-600 microscope using a Rolera EM-CCD camera and MetaMorph software (Molecular Devices, Sunnyvale CA). At least 3 images were acquired for each sample, with typical representations shown.

RESULTS AND DISCUSSION

The pattern of sub-cellular localization for BPD involves ER > mitochondria [4]. In contrast, the corresponding pattern for EtNBS (Fig. 1) predominantly involves lysosomes as can be seen by the coincidence of the red and green pixels in this image.

Figure 1.

Sub-cellular localization of EtNBS in OVCAR 5 cells: Panel a = EtNBS fluorescence, b = lysotracker green fluorescence, c = overlay.

Figure 2 illustrates the morphology of cells after increasing PDT doses using BPD or EtNBS along with corresponding images of untreated cells. As the PDT dose (using BPD) was increased there is an initial appearance of paraptosis characterized by a highly-vacuolated cytoplasm. As the light dose is increased, this was replaced by both apoptosis (induced by fragmented cells and nuclei) and necrosis (cells wholly permeable to Hö33342). In contrast, lysosomal photodamage resulted only in the appearance of apoptosis, as indicated by fragmentation of both cells and nuclear chromatin. The percentage of apoptotic cells increased with the light dose, with necrosis observed at the higher PDT doses. A more detailed depiction of the morphology of paraptosis vs. apoptosis is seen in [14].

Figure 2.

Left: (panels a and b) indicating phase-contrast and Hö33342 labeling patterns of untreated OVCAR-5 cells. Right: corresponding patterns 24 hr after specified PDT doses using BPD (c-j) or EtNBS (k-r) as the photosensitizing agent.

Dose-response data involving clonogenic assays are summarized in Fig. 3. The shoulder on the dose-response curve for BPD is attributed to the cytoprotective effect of autophagy [6]. In contrast, targeting lysosomes resulted in a simple log dose-response curve. Since lysosomes are a part of the autophagic process, photodamage appears to compromise the ability of autophagy to offer cytoprotection. In this figure, an indication of the death modes is provided as discussed below.

Figure 3.

PDT log dose-response curves for OVCAR-5 cells using EtNBS (A, ▴) or BPD (B, ●) as the photosensitizing agent. Pertinent death pathways are indicated: apoptosis, necrosis and paraptosis, with the latter showing regions where canonical (‘c’) or non-canonical ‘non-c’ effects were observed.

We have described a ‘canonical’ response to ER photodamage when the PDT dose is in the LD_10-50 range [12,15]. This involves MAPK and JNK-related enzymes, requires a brief interval of protein synthesis and results in relocation of the protein termed HMBG1 from nucleus to cell membrane [9-11]. At levels of photodamage exceeding an LD₅₀ dose, we have shown paraptosis is ‘non-canonical’ [12,15]. If the level of photodamage proceeds further, cross-linking of ER proteins is observed and paraptosis is impaired [12]. In the latter report, we examined paraptosis initiated by cyclosporin A (CsA). This agent induced ‘canonical’ paraptosis that can be inhibited by the protein synthesis antagonist cycloheximide or by rapamycin, an agent that induces autophagy. This is also true for paraptosis initiated by PDT doses in the range of an LD₅₀ effect or less. As the PDT dose approaches an LD₉₀ value photodamage from BPD then begins to involve predominantly apoptosis and necrosis, likely reflecting the increasing level of ER protein cross-linking.

Lockshin has pointed out that a lethally-damaged cell will take any available path to cell death [16]. This point is illustrated in the BPD study: when paraptosis is impaired by ER protein cross-linking, other death pathways will be initiated. The role of death pathways in the efficacy of PDT is often neglected, with phase-contrast images of photodamaged cells seldom shown. Studies in animal models often report only survival data. The mode(s) of photokilling and details of sub-cellular targeting can be important when considering optimization of protocols. We have shown that impaired apoptosis can affect PDT responses and this can be circumvented by choosing photosensitizing agents that also target the ER [7,8].