Effects of HVP Status on Responsiveness to Ionizing Radiation vs. Photodynamic Therapy in Head and Neck Cancer Cell lines

Abstract

Efficacy of ionizing radiation (I/R) was compared with phototoxic effects of photodynamic therapy (PDT) in vitro using two cell lines derived from patients with head and neck squamous cell carcinoma (HNSCC). A cell line derived from a donor with a human papilloma virus (HPV) infection was more responsive to I/R but significantly less responsive to PDT than a cell line derived from an HPV-free patient. Cell death after I/R in the HPV(+) cell line was associated with increased DEVDase activity, a hallmark of apoptosis. The HPV(−) line was considerably less responsive to I/R, with DEVDase activity greatly reduced, suggesting an impaired apoptotic program. In contrast, the HPV(−) cells were readily killed by PDT when the ER was among the targets for photodamage. While DEVDase activity was enhanced, the death pathway appears to involve paraptosis until the degree of photodamage reached the LD₉₉ range. These data suggest that PDT-induced paraptosis can be a death pathway for cells with an impaired apoptotic program.

This image demonstrates the ability of photodynamic therapy (PDT) directed at ER/mitochondria to eradicate a head & neck cancer cell line (WSU12) from a patient with a human papilloma virus infection. These cells were relatively unresponsive to ionizing radiation, typical of HPV (+) tumors. In contrast, cells from an HPV (−) patient (UP154) were responsive to ionizing radiation but not to PDT.

INTRODUCTION

HPV infection is associated with enhanced responsiveness of head & neck cancers to ionizing radiation (I/R ) (1–3). In this study, we examined responsiveness to I/R vs. PDT using two HNSCC lines in cell culture representing the HPV(+) and HPV(−) phenotypes. The data indicate that the HPV(−) line has an impaired apoptotic program but can be eradicated by PDT-induced paraptosis. Prior studies had indicated that subcellular targets for photodamage can influence death modes (4–5). Apoptotic death results when mitochondria, lysosomes or the ER are targets for photodamage while death involving paraptosis can occur after ER photodamage (6–7). The latter process involves formation of an extensive series of cytoplasmic vacuoles, mainly derived from ER membranes, ultimately leading to loss of viability. Details on the pathway(s) to paraptosis are incomplete but there is evidence that activation of MAP kinases is involved (8). We have reported on the ability of ER photodamage to initiate paraptosis in neoplastic cells (7). The present study was undertaken after exploratory studies suggested an inverse relationship between responsiveness to I/R vs. PDT in HNSCC cell lines varying in HPV status. Since this issue is intended as a memorial issue dedicated to the accomplishments of Thomas Dougherty, it is perhaps noteworthy that his first experiments upon arrival at the Roswell Park Cancer Institute involved searching for agents that would sensitize tumor cells to ionizing radiation. In this report, evidence is presented that PDT can effectively eradicate cells with an impaired response to I/R. No doubt Tom would have appreciated this news.

MATERIALS AND METHODS

Chemicals and supplies.

BPD (benzoporphyrin derivative, Verteporfin) was obtained from VWR (Cat No 1711461), Radnor PA. NPe6 was provided by Prof. Kevin M. Smith, Louisiana State University. Fluorescent probes were provided by Thermo Fisher Scientific. Other reagents were obtained from Sigma-Aldrich or Calbiochem Corp. and were of the highest available purity.

Cell cultures.

The WSU12 cell line, negative for HPV, was established at Wayne State University. UP154, an HPV(+) cell line, was derived at the University of Pittsburgh. These were cultured in RPMI1640 and MEM, respectively, with media supplemented with 10% fetal bovine serum.

Ionizing radiation.

Cells were irradiated with 2–6 Gy using a gantry-mounted Best Theratronics Gammabeam 500 at dose rate of 1 Gy/min at room temperature under atmospheric oxygen conditions.

Photodynamic therapy.

Cultures were incubated at 37̊C in 0.5 μM BPD or 20 μM NPe6 for 1 h. The media was replaced and the cultures irradiated at 690 nm (BPD) or 660 nm (NPe6) at specified intervals as described previously (7). Light doses were calculated using a Scientech H410 Power & Energy meter. Over a range from 600 to 700 nm at a 10 nm bandwidth, this QH lamp provides a 2 mW/cm² light dose.

Microscopy.

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. A thermo-electrically cooled stage was used to keep the temperature at 15̊C to minimize further metabolic changes during observation.

Detection of singlet oxygen formation.

Cells were incubated for 60 min in media containing (where specified) BPD (0.5 μM) + 10 μM DADB, a fluorescent probe for singlet oxygen formation (9,10). The fluorescence signal from intracellular DADB is diminished when the probe encounters singlet molecular oxygen, providing a means for assessing ¹O₂ formation. DADB fluorescence was detected at wavelengths centered at 525 nm upon excitation at 360–400 nm. A 600 nm low-pass filter was inserted into the excitation beam to eliminate fluorescence from BPD. Images were analyzed using Metamorph software. A ‘thresholding’ program was used to assess pixel intensity in cytoplasmic loci, avoiding the bright punctate spots that represent DADB sequestered in lysosomes where it will be unaffected by BPD photodamage. Data are reported in terms of average pixel density ± SD as described in Ref. 10.

Clonogenic survival.

After treatment in 35 mm dishes, cultures were allowed to grow in a 5% CO₂ atmosphere at 37̊ for 7–10 days. Colonies were fixed in 70% ethanol, stained with 1% Crystal Violet (2 h, ambient temperature) and counted using an Oxford Optronix Ltd GelCount system as previously described (11). All such studies were performed in triplicate.

Other assays.

Cross-linking of an ER chaperone protein (binding immunoglobulin protein, BiP/GRP 78) was determined on western blots (12) 1 h after irradiation at which time cross-linking hed reached a steady-state. Protein extracts were prepared with modified RIPA buffer (50mM Tris-HCl, pH 7.4, 150mM NaCl, 0.25% deoxycholic acid, 1% NP-40, 1mM EDTA. 1mM PMSF, 1mM sodium orthovanadate, 1mM sodium fluoride, 1X protease inhibitor cocktail purchased from Roche Cat # 11873580001). Immunoblot analysis of BiP was performed using a rabbit monoclonal antibody against BiP (Cell Signaling Tech. Cat.# 3177). DEVDase results are reported in terms of μmol of product formed/mg protein/hour as described in Ref. 13.

RESULTS AND DISCUSSION

PDT vs. I/R response as a function of HPV status

A comparison of survival of WSU12 and UP154 cells after exposure to I/R vs. PDT is shown in Fig. 1. The photosensitizer was BPD, an agent known to target mitochondria and ER for photodamage (14). The HPV(+) UP154 cell line was substantially more responsive to the lethal effects of XRT than were HPV(−) WSU12 cells. A 6 Gy radiation dose eradicated >99% of the UP154 population while killing only ~80% of WSU12 cells. In contrast, a 135 mJ/ cm² PDT dose reduced clonogenicity of WSU12 by 90% while having only a slight phototoxic effect on UP154 cells. Measurements of DEVDase activity provide information on the conversion of pro-caspases 3 and 7 to their active forms. This is usually taken as an indication of an apoptotic process (15), but morphologic studies were not consistent with this assumption. This suggests that the caspase activity threshold for induction of apoptotic death may vary with the genetic background of the cells.

Fig. 1.

Clonogenic cell survival after ionizing irradiation (6 Gy) or PDT (BPD, 135 mJ/cm²) on WSU12 and UP154 cells. Data represent average ± SD of three determinations. See Fig. 3 for information on clonogenicity of untreated cell populations.

In related studies to be described in detail elsewhere, we observed that WSU cells have a high autophagic flux while UP154 cells have little or none. A dose-response curve assessing PDT efficacy (Fig. 2) indicates a substantial shoulder for WSU12 cells, consistent with the cytoprotective effects of autophagy on photokilling as previously reported (11). The HPV(+) UP154 cell line showed only a weak response to PDT with no significant shoulder apparent. Based on clonogenic data, we calculate the LD₅₀ light dose for WSU12 to be approx. 90 mJ/ cm²; LD₉₀, 130 mJ/ cm²; LD₉₉, 200 mJ/ cm².

Fig. 2.

Clonogenic survival of WSU12 and UP154 cells as a function of light dose with BPD as the photosensitizing agent. Data represent average ± SD in triplicate determinations.

When the NPe6 used as the photosensitizing agent, there was no detectable effect on survival (Fig. 3) using a 135 mJ/cm² light dose that is usually adequate for initiation of apoptosis (16). NPe6 targets only lysosomes for photodamage and induces cell death via a route that, in murine hepatoma cells, involves release of lysosomal proteases into the cytoplasm that ultimately triggers an apoptotic response (16).

Fig. 3.

Clonogenicity survival of WSU12 and UP154 cells photosensitized with NPe6 and irradiated (135 mJ/cm²). Data represent average ± SD for three determinations.

Assessing cytoplasmic BPD fluorescence revealed that differences in phototoxicity of between UP154 and WSU12 were not correlated with any impairment in BPD accumulation under experimental conditions employed: 0.5 μM BPD, 60 min incubations at 37̊C (Fig. 4). The fluorescent probe DADB was used to assess formation of intracellular ¹O₂ after irradiation as described in the figure legend (10). Formation of this ROS led to significant quenching of DADB fluorescence. In this study, we observed some partition of DADB to lysosomes where BPD does not initiate photodamage. These sites were excluded from the image analysis by a thresholding program. Differences in phototoxicity shown in Fig. 1 were not correlated with any differences in cytoplasmic ¹O₂ formation (Fig. 5). An analysis of cytoplasmic pixel brightness using the ‘region statistics’ option of the Metamorph acquisition program (10) is shown in Fig. 6. There was no significant difference between data obtained with WSU 12 vs. UP154 cells.

Fig. 4.

Assessing BPD uptake by fluorescence microscopy. WSU12 (panel a) and UP154 (panel b) cells were treated with 0.5 μM BPD for 60 min at 37̊C.

Fig. 5.

Assessing ROS formation in WSU12 (a, c) and UP154 (b, d) cells loaded with the DADB + BPD as described in the text, then irradiated (90 mJ/ cm²). DADB fluorescence (535 nm) was measured in control cells (a, b) and cells directly after irradiation (c, d) using 360–400 nm excitation.

Fig. 6.

DADB oxidation by singlet oxygen as indicated by pixel brightness. All acquisitions involved a 100 ms image acquisition time. Data represent mean pixel intensity ± SD in control cells and cells after BPD-induced PDT involving a 135 mJ/ cm² light dose (LD₉₀). Thresholding was used to limit measurement to areas corresponding to the cell cytoplasm, excluding lysosomes, as described in the text. The decrease in pixel intensity is proportional to the extent of photobleaching of the probe by singlet oxygen.

Photodamage, paraptosis and apoptosis

Apoptotic nuclei were not frequently observed 1 day after irradiation of WSU12 cells at an LD₉₀ PDT dose (Fig. 7), although there was a substantial increase in DEVDase activity (Fig. 1). There was, however, an extensive pattern of cytoplasmic vacuole formation, the hallmark of paraptosis (Fig. 7 a, b). If we increased the light dose to an LD₉₉ level (200 mJ/ cm²), apoptotic cells were apparent 24 h later (Fig. 7c, d). These results suggest the hierarchy of death pathways discussed below.

Fig. 7.

Morphology (a, c) and images of the nuclear fluorescence probe Ho33342 (b, d) 1 day after irradiation of WSU12 cells using BPD as the photosensitizing agent. Light dose = 135 (a, b) or 200 (c, d) mJ/ cm².

Paraptosis and MAP kinases

Paraptosis is associated with ER photodamage (7) and involves the action of MAP kinases (8). At an LD₅₀ PDT dose, we observed paraptotic vacuole formation in WSU12 cells that was partly impaired by antagonists of MAPK/JNKs or MAPK/ERKs (Fig. 8). At an LD₉₀ level of photokilling, however, no loss of vacuole formation was observed (Fig. 9). Results summarized in Fig. 10 indicate that MAPK antagonists can partially reverse photokilling at moderate (LD₅₀) PDT doses, but not in at LD₉₀ levels. These results suggest that there may be multiple pathways to paraptosis, some of which do not involve MAP kinase activation.

Fig. 8.

Morphology of WSU12 (a-c) and UP154 (d-f) cells. Untreated cells (a, d), 1 day after photosensitization with NPe6 (b, e) or BPD (c, f). Light dose = 135 mJ/ cm² at 660 nm (NPe6) or 690 nm (BPD).

Fig. 9.

Effects of MAPK antagonists (50 nM) on WSU12 cells 1 day after irradiation with LD₅₀ (a-c) or LD₉₀ (d-f) PDT doses as described in the text. No additions were made to samples shown in panels a and d; cells were treated with U0126 (panels b and e) or SP600125 (panels c and f). LD₅₀ = 90 mJ/ cm²; LD₉₀ = 135 mJ/ cm².

Fig. 10.

Clonogenicity of WSU12 cells after photosensitization with BPD and irradiation at LD₅₀ or LD₉₀ light doses. Effects of addition of U0126 (b, e) or SP600125 (c, f) are shown. Data represent average ± SD for three determinations. LD₅₀ = 90 mJ/ cm², LD₉₀ = 135 mJ/ cm².

ER protein cross-linking and its implications

To assess the ability of ER photodamage to initiate cross-linking of ER proteins, we examined this phenomenon in a typical protein termed BiP. Cell lysates were prepared 1 hr after irradiation. Immunoblot analysis of BIP under reducing conditions detected high molecular-weight bands at multiples of 78 kDa. The extent of cross-linking was proportional to the light dose (Fig. 11). These results, along with data shown in Figs. 2, 7, 9 and 10, show that the level of ER protein cross-linking at an LD₅₀ light dose is minimal (Fig. 10) and that paraptosis and photokilling can be partly reversed by MAPK antagonists (Fig. 9). Even after a 135 mJ/cm² light dose (an LD₉₀ effect, Fig. 2), the morphology of dying cells was still mainly that of paraptosis (Fig. 7, panels a and b). But when the PDT dose was increased from an LD₅₀ to an LD₉₀ level, paraptosis and photokilling could no longer be partly reversed by MAPK antagonists (Figs. 9 and 10). The level of ER protein cross-linking at these light doses was still minimal (Fig. 11). When the light dose was increased to 240 mJ/cm², a >LD₉₉ effect, the outcome was mainly apoptotic (Fig. 7, panels c and d) with BiP cross-linking readily detectable.

Fig. 11.

Western blots indicating the extent of cross-linking of the ER protein BiP in WSU12 cells 1 h after irradiation at specified light doses.

CONCLUSIONS

Data shown in Fig. 1 are consistent with reports that HPV infection can promote responsiveness of head & neck cancer cell lines to I/R (1–3). The lack of DEVDase activity in HPV(−) cells implies an impaired apoptotic response to radiation damage. The finding that radiation-resistant HPV(−) cells respond to photodamage suggests that PDT can be used to control/eradicate neoplastic cells with an impaired apoptotic program if appropriate intracellular targets are chosen.

We had previously identified paraptosis as a death pathway evoked after ER photodamage (7). In the present study, morphologic evidence for paraptosis was observed after photodamage to the HPV(−) WSU12 cell line using BPD but not an agent (NPe6) that targets only lysosomes (Fig. 8). Data shown here are consistent with the proposal that the HPV(−) lines have an impaired pathway to apoptosis after XRT (Fig. 1). A very high PDT dose was required for initiation of apoptosis (Fig. 6). At lower PDT doses, paraptosis predominates. Initiation of paraptosis after ER photodamage may explain the responsiveness of an HPV(−) cell line to PDT (Fig. 1). Studies on additional cell lines will be needed to assess the global relevance of this observation.

Examples of paraptosis have been associated with increased MAPK activity (8).

In this study, we show that paraptosis is impaired and viability partially preserved by MAPK antagonists at LD₅₀ but not at LD₉₀ levels. As has been reported before, ER protein cross-linking can occur after ER photodamage (17). In this study, we have observed that paraptotic death become no longer reversible by MAPK antagonists as the level of ER photodamage increases. These results indicate that there are both MAPK-dependent and -independent pathways to paraptosis after ER photodamage. In this study, only one ER protein was examined. Other proteins could show different levels of sensitivity to ER photodamage. Data shown in Fig. 9 indicate that apoptosis becomes a predominant mode of photokilling at high PDT doses where there is ER protein cross-linking. This may restrict the ability of ER proteins to undergo necessary configurational changes for vacuole formation. In this regard, Lockshin has observed that a dying cell will take any available pathway to death (18). When paraptosis becomes unavailable, apoptosis may be activated.

We propose that PDT may be an effective means for eradication of malignant cell lines with an impaired apoptotic program via death via paraptosis. While apoptosis was observed at high PDT doses, at an LD₉₀ level of photokilling, paraptosis was the major death pathway. Ongoing studies with two additional HPV (−) HNSCC cell lines UM19 and HPV(+) HNSCC cell line UP90 confirmed our findings that HPV(−) HNSCC is radioresistant but PDT-sensitive while HPV(+) cells are sensitive to I/R but resistant to PDT. An unexpected finding was the impaired PDT response in HPV(+) cell lines. There appears to be no simple explanation for this result at present, i.e., we observed no impaired photosensitizer uptake or decreased ROS formation. There have been few examples of PDT ‘resistance’ (19). Exploring the mechanism whereby UP154 cells survive high-dose PDT is a topic for further study. In this context, it is important to remember that the efficacy of PDT depends on factors other than direct photokilling, e.g., shutdown of the tumor vasculature (20). Clinical reports on efficacy of PDT in conjunction with surgery have not noted any loss of efficacy associated with HPV status (21).

Paraptosis is slowly gaining more attention as a route to cell death in cancer chemotherapy. This process can be a route to cell death in cells with an impaired apoptotic program and suggests that photosensitizing agents whose targets for photodamage include the ER may be especially effective for tumor eradication.