Increased killing of SCCVII squamous cell carcinoma cells after the combination of Pc 4 photodynamic therapy and dasatinib is associated with enhanced caspase-3 activity and ceramide synthase 1 upregulation

Photodynamic therapy (PDT) is not always effective as an anticancer treatment, therefore, PDT is combined with other anticancer agents for improved efficacy. The combination of dasatinib and PDT with the silicone phthalocyanine photosensitizer Pc 4 was assessed for increased killing of SCCVII mouse squamous cell carcinoma cells, a preclinical model of head and neck squamous cell carcinoma, using apoptotic markers and colony formation as experimental end-points. Because each of these treatments regulates the metabolism of the sphingolipid ceramide, their effects on mRNA levels of ceramide synthase, a ceramide-producing enzyme, and the sphingolipid profile were determined. PDT + dasatinib induced an additive loss of clonogenicity. Unlike PDT alone or PDT + dasatinib, dasatinib induced zVAD-fmk-dependent cell killing. PDT or dasatinib-induced caspase-3 activation was potentiated after the combination. PDT alone induced mitochondrial depolarization, and the effect was inhibited after the combination. Annexin V+ and propidium iodide+ cells remained at control levels after treatments. In contrast to PDT alone, dasatinib induced upregulation of ceramide synthase 1 mRNA, and the effect was enhanced after the combination. Dasatinib induced a modest increase in C20:1-and C22-ceramide but had no effect on total ceramide levels. PDT increased the levels of 12 individual ceramides and total ceramides, and the addition of dasatinib did not affect these increases. PDT alone decreased substantially sphingosine levels and inhibited the activity of acid ceramidase, an enzyme that converts ceramide to sphingosine. The data suggest that PDT-induced increases in ceramide levels do not correlate with ceramide synthase mRNA levels but rather with inhibition of ceramidase. Cell killing was zVAD-fmk-sensitive after dasatinib but not after either PDT or the combination and enhanced cell killing after the combination correlated with potentiated caspase-3 activation and upregulation of ceramide synthase 1 mRNA but not the production of ceramide. The data imply potential significance of the combination for cancer treatment.


Introduction
Photodynamic therapy (PDT), a non-invasive cancer treatment modality, can effectively eradicate local malignancies. PDT utilizes a light-absorbing photosensitizer, visible light and oxygen to generate reactive oxygen species that destroy malignant cellular targets (1). However, because tumors recur, PDT needs to be optimized to improve its therapeutic benefit. Dasatinib, a multi-kinase inhibitor, is an anticancer agent that has been successfully used for treatment of chronic myeloid leukemia (2). As a single agent evaluated in clinical trials for treatment of solid tumors, including advanced head and neck squamous cell carcinoma (HNSCC), however, dasatinib has not been shown to be successful (3). In combination with other chemotherapeutic agents or radiation dasatinib is a more effective anticancer treatment in vitro, in vivo and in clinical trials (4-7).

Increased killing of SCCVII squamous cell carcinoma cells after the combination of Pc 4 photodynamic therapy and dasatinib is associated with enhanced caspase-3 activity and ceramide synthase 1 upregulation
Bioactive sphingolipids have been implicated in drugand radiation-resistance, therefore targeting sphingolipid metabolism can contribute to increased effectiveness of the current treatment strategies (8). As shown in Fig. 1, the sphingolipid ceramide is generated in the de novo biosynthesis pathway, which includes a ceramide synthase-dependent addition of a fatty acyl group to dihydrosphingosine to form dihydroceramide. Ceramide is formed from dihydroceramide by a desaturase-dependent insertion of a double bond in the sphingosine backbone. Six mammalian ceramide synthases have been identified with distinct specificity for fatty acyl CoAs and functions (9). For example, C18-and C16-ceramide, containing an 18-and 16-carbon fatty acid, are generated by ceramide synthase 1 and 6, respectively, and induce HNSCC suppression and proliferation, respectively (10). Ceramide is deacylated by ceramidase, giving rise to sphingosine, and sphingosine is acted upon by sphingosine kinase to give rise to sphingosine-1-phosphate (S1P), an antiapoptotic sphingolipid.
We demonstrated that the knockdown of ceramide synthase 1 or 6 is associated with reduction in ceramides and dihydroceramides resulting in apoptotic resistance to PDT with Pc 4 (11,12). Dasatinib induces apoptosis via upregulation of ceramide synthases, including increased expression of ceramide synthase 1 gene (13). The combination of dasatinib and PDT with Pc 4 was tested for potential anticancer efficacy in SCCVII mouse squamous cell carcinoma cells, a preclinical model of HNSCC (14), using apoptotic markers, colony formation and ceramide metabolism as experimental end-points.
Cell culture and treatments. SCCVII cells, initially derived from the spontaneous abdominal wall tumor of a C3H mouse (15), were grown in RPMI medium containing 10% fetal bovine serum, 100 U/ml penicillin and 100 µg/ml streptomycin (Life Technologies). Cells were maintained at 37˚C in a 5% CO 2 atmosphere and were treated in the growth medium. For PDT experiments, after overnight incubation with Pc 4 at 37˚C, cells were irradiated with red light (2 mW/cm 2 ; λ max ~670 nm) using a light-emitting diode array light source (EFOS, Mississauga, ON, Canada) at the fluence of 200 mJ/cm 2 at room temperature and then incubated at 37˚C for indicated periods of time. For PDT + dasatinib, dasatinib was added to the cells 22 h prior to irradiation, unless indicated otherwise. After treatments, cells were collected on ice and processed for various analyses. Electrospray ionization/double mass spectrometry (MS) analysis. After extraction, sphingolipids were separated by high performance liquid chromatography, introduced to electrospray ionization source and then analyzed by double MS using TSQ Quantum Access Max triple stage quadrupole mass spectrometer (Thermo-Fisher Scientific, Pittsburg, PA, USA) as described previously (16).
RNA extraction and quantitative real-time polymerase chain reaction (RT-PCR). Total RNA isolation was performed with RNeasy ® Mini kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. cDNA was synthesized from 1 µg of the total RNA using iScript™ cDNA Synthesis kit (Bio-Rad, Hercules, CA, USA). The concentration and quality of total RNA preparations were evaluated spectrophotometrically. RT-PCR was performed on a Bio-Rad CFX96 detection system using Bio-Rad SsoFast Probes Supermix™ and TaqMan ® Gene Expression Assays (Life Technologies) with the primers for ceramide synthases 1, 2, 4, 5, 6, the housekeeping gene products RPL37A and hypoxanthine-guanine phosphoribosyltransferase (HGPRT), and the fluorophore probe FAM-490 (6-carboxyflurescein; all obtained from Life Technologies). Initial steps of RT-PCR were 30 sec at 85˚C, followed by 40 cycles consisting of a 5 sec at 95˚C, followed by 10 sec at 60˚C. Determination of the relative normalized expression of corresponding ceramide synthase mRNAs against the expression of housekeeping gene-encoded proteins RPL37A and HGPRT was performed by ∆∆C T provided by CFX96 manager software 3.0 from Bio-Rad.
Acid ceramidase activity assay. Acid ceramidase activity was performed as described previously (17). Cells were lysed under acidic condition (pH 4.5). Equal amounts of N-[9,10-3 H] D-e-C16-ceramide were mixed with 0.2% Triton X-100 and 0.4% cholate and dried down under nitrogen. The lipid film was dissolved by mixing and sonication in deionized water. After additions of acidic assay buffer (0.2 M acetic acid, 0.2 M sodium acetate and 0.5% Triton X-100, pH 4.5) and cell lysate, the reaction was carried at 37˚C for 1 h and stopped by adding Dole's alkaline solution. [ 3 H]palmitic acid, a hydrolytic product of acid ceramidase, was extracted and processed to calculate the enzyme activity (17). Quantitation of radioactivity was performed using LS 6500 multipurpose scintillation counter (Beckman Coulter, Brea, CA, USA).
DEVDase (caspase-3) activity assay. As described previously (18), DEVDase activity was determined in the cytosol by an assay based on the enzyme's cleavage of a fluorogenic derivative of the tetrapeptide substrate N-acetyl-Asp-Glu-Val-Asp (DEVD; Enzo Life Sciences, Farmingdale, NY, USA). The peptide sequence is based on the cleavage site Asp 216 of the caspase-3 substrate poly(ADP-ribose) polymerase (PARP). The fluorescence of the cleaved DEVD substrate was measured using a spectrofluorometer (F-2500 Hitachi, New York, NY, USA; 380 nm excitation, 460 nm emission).
Apoptosis detection. As previously described (11,20,21), to detect apoptosis, the exposure of phosphatidylserine in the outer leaflet of the cell membrane and cell membrane integrity loss were measured using Annexin V and DNA-binding propidium iodide fluorescent dyes (BD Biosciences), respectively. Early apoptotic (Annexin V + /propidium iodide -) were distinguished from late apoptotic or necrotic cells (Annexin V + /propidium iodide + ). The kit was obtained from BD Biosciences and the flow cytometric protocol was followed, as described by the manufacturer.
Clonogenic assay. Long-term cell viability was assessed using clonogenic assay according to the modified protocol that we described previously (20). Plating density was 250 cells/plate. Plating efficiency was 34% (n=16).
Protein determination. Protein content was determined by a modified Bradford assay (Bio-Rad) or, for acid ceramidase assay, by a bicinchoninic acid protein assay kit (Thermo-Fisher Scientific).

Statistical analysis.
Data are shown as the mean ± SEM. Statistical analyses were performed by Student's t-test. Significance was defined as a two-tailed p<0.05.

Results
PDT + dasatinib enhances overall cell killing. Effect of zVADfmk on PDT ± dasatinib-induced cell killing. To test whether killing of SCCVII cells is increased by the combination of Pc 4-PDT with dasatinib, colony formation assay was used as the experimental end-point. The treatments were first used as single agents to determine whether they induce dose-dependent cell killing. Incubation of SCCVII cells with 100, 200 and 500 nM dasatinib led to 2, 28 and 53% cell killing, respectively ( Fig. 2A). Similarly, PDT with 100, 250 or 500 nM Pc 4, at the light fluence of 200 mJ/cm 2 , induced 16, 26 and 87% cell killing, respectively ( Fig. 2A). Treatment of cells with the combination of PDT and dasatinib, each used at LD <30, led to 72% cell killing, which was significantly greater than that of each treatment alone (Fig. 2B). Because both agents are apoptotic inducers (11,12,(22)(23)(24)(25)(26), the requirement of caspases in cell killing by each agent and the combination was assessed using the pan caspase inhibitor zVAD-fmk. As shown in Fig. 2B, zVAD-fmk substantially inhibited cell killing after dasatinib, but not after either PDT alone or the combination. Overall, the data demonstrate that each agent induces dose-dependent cell killing, the combination enhances cell killing, and that, unlike PDT alone or the combination, dasatinib induces zVAD-fmkdependent cell killing.
Dasatinib-induced caspase-3 activation is inhibited by zVADfmk. The combination potentiates PDT-or dasatinib-induced activation of caspase-3 in the absence of appearance of other apoptotic markers. Because zVAD-fmk inhibits effector caspases, including caspase-3 (27), we verified caspase-3 as a zVAD-fmk target in dasatinib-induced cell death using DEVDase assay. Caspase-3 activation began at 2 h and peaked at 24 h after dasatinib (not shown). As depicted in Fig. 3A, dasatinib induced activation of DEVDase was abolished by zVAD-fmk. The data suggest that zVAD-fmk-sensitive cell killing after dasatinib involves caspase-3.
To further assess induction of apoptosis after treatments, caspase-3 activation, mitochondrial depolarization and the appearance of Annexin V + and propidium iodide + cells were determined. PDT induced a dose-dependent activation of caspase-3 and the effect was potentiated after PDT + dasatinib (Fig. 3B). PDT alone induced mitochondrial depolarization and the effect was inhibited after the combination (Fig. 3C). Annexin V + and/or propidium iodide + cells remained at control levels after treatments (Fig. 3D). We validated that SCCVII cells display depolarized mitochondria and undergo apoptosis using camptothecin as a positive control (Fig. 3C and D). The results show that the combined treatment augments PDT or dasatinib-induced caspase-3 activation in the absence of appearance of other apoptotic markers.
Dasatinib-induced upregulation of mRNA ceramide synthase 1 is enhanced after the combination. Dasatinib upregulates expression levels of ceramide synthase genes 1, 2, 5 and 6 (13). We showed that knockdown of ceramide synthase 1 or 6 leads to apoptotic resistance to PDT (11,12). To test whether ceramide synthases are affected by treatments, mRNA levels of ceramide synthases 1, 2 4, 5 and 6 were measured using RT-PCR. As depicted in Fig. 4A, ceramide synthase 1 mRNA levels were upregulated after dasatinib and the effect was further increased after PDT + dasatinib. PDT alone did not significantly increase ceramide synthase 1 levels. None of the treatments had any effect on mRNA levels of ceramide synthase 2, 4, 5 and 6 ( Fig. 4A). The data show that dasatinib-induced upregulation of ceramide synthase 1 mRNA levels is enhanced after the combination.
Effects of treatments on the sphingolipid profile. To assess whether enhanced upregulation of ceramide synthase 1 mRNA after PDT + dasatinib is associated with increased ceramide production, ceramide levels were measured using MS. Dasatinib alone induced a modest increase in C20:1-and C22-ceramide but had no effect on total ceramide levels (Table I and Fig. 4B). PDT alone increased the levels of all 12 individual ceramides that were measured, as well as total ceramides. PDT-induced increases in ceramide levels were not, for the most part, further changed after the combination, with the exception of attenuated levels of C26-and C26:1-ceramide after PDT + dasatinib compared to individual treatments (Table I and Fig. 4B).
The effects of treatments on other sphingolipids were also measured by MS and are shown in Table I. Unlike dasatinib, and irrespective of the presence of dasatinib, PDT increased the levels of C16-dihydroceramide, an intermediate from the de novo sphingolipid biosynthesis pathway. Moreover, unlike dasatinib, and irrespective of the presence of dasatinib, PDT induced an 87% decrease in sphingosine levels. PDT also induced a 62% decrease in S1P levels, and the effect was not changed after the combination. Overall, the data show that in SCCVII cells PDT induced substantial changes in the sphingolipid profile that were not modulated by the addition of dasatinib.
Acid ceramidase activity is inhibited after PDT. The substantial decrease in sphingosine levels after PDT could result from  inhibition of ceramidase, a sphingosine-producing enzyme (Fig. 1). The inhibition of acid ceramidase has been implicated in radiosensitization of prostate cancer cells (28). We measured the activity of acid ceramidase in cell lysates from untreated and PDT-treated cells. As depicted in Fig. 4C, acid ceramidase activity was reduced by 52% post-PDT. Thus, PDT-induced decrease in sphingosine correlated with inhibition of acid ceramidase.

Discussion
Dasatinib-induced caspase-3 activation and cell killing was zVAD-fmk-dependent. PDT induced caspase-3 activation and the effect was potentiated after the combination. As shown previously (29), PDT-induced caspase-3 activation was abolished by zVAD-fmk. However, PDT-or PDT + dasatinib-induced cell killing was zVAD-fmk-insensitive. zVAD-fmk-insensitivity of PDT is consistent with the findings that caspase-3 is not required for the lethal effect of PDT (30). Incidentally, zVAD-fmk has been reported to upregulate caspase-9 activity in cell death after etoposide (27). We found no activation of caspase-9 after PDT in SCCVII cells (31). Therefore, it is unlikely that zVAD-fmk would have such an effect, especially since the inhibitor was used in our studies at a non-toxic, and comparably, lower concentration than in the had no effect on acid ceramidase activity. All the data are shown as the mean ± SEM, n=4-5. The significance (p<0.05) is shown as follows: * treated is different from control; + (PDT + dasatinib) is different from PDT or dasatinib alone. CERS, ceramide synthase; Con, untreated control; Das, dasatinib.
etoposide study, i.e., 25 vs. 50 µM, respectively. Overall, the data suggest that, unlike PDT or the combination, dasatinib requires caspase-3 activation for cell killing. Apparently, PDT rather than dasatinib determined zVAD-fmk sensitivity of cell killing after the combination.
Dasatinib-induced upregulation of ceramide synthase 1 mRNA correlated with increased production of C20:1-and C22-ceramide concomitant with activation of caspase-3. This is consistent with the finding that ceramide synthase 1 gene is upregulated during apoptosis after dasatinib (13). However, there was no correlation between combinationinduced enhanced ceramide synthase 1 mRNA upregulation and ceramide production, suggesting that the enzyme might modulate other cellular functions. Ceramide synthase 1 has been associated with sensitization to cisplatin via activation of p38 mitogen-activated protein kinase (MAPK) and concomitant translocation of the enzyme from the endoplasmic reticulum to the Golgi apparatus (32). Activation of p38 MAPK is critical for the antileukemic effects of dasatinib (33). Activation of p38 MAPK pathway has been associated with triggering apoptosis after Pc 4-PDT (34). The potential link between ceramide synthase 1 upregulation, p38 MAPK pathway and sensitization to PDT should be addressed in our future studies.
In HNSCC in vitro and in vivo models ceramide synthase 1-dependent C18-ceramide production has a proapoptotic role (35) and inhibits xenograft growth (36), respectively. We showed in HNSCC cells that knockdown of CERS1 induced apoptotic resistance to PDT and reduced the levels of total ceramide and several individual ceramides, including C18-ceramide (12). Our findings that the levels of the whole spectrum of ceramides are increased after PDT in the absence of upregulation of ceramide synthase mRNAs imply the involvement of other enzymes. Accordingly, in the present study we have demonstrated that PDT-induced inhibition of acid ceramidase correlates with decrease in sphingosine and increase in ceramide levels, concomitant with activation of caspase-3. Consistent with these findings, upregulation of acid ceramidase confers radioresistance in prostate cancer cells (28). As suggested in the same study, inhibition of acid ceramidase could be a potential target for treatment of cancers with overexpressed acid ceramidase.
The present study shows for the first time enhanced additive killing of SCCVII cells after the combination of PDT and dasatinib and paves the way for testing the combination in vivo. This combination has the potential to achieve what is hoped by combination therapy, i.e. maximizing the efficacy of each single anticancer agent while minimizing their systemic toxicity through the delivery of lower drug doses (37). This will have to be validated in vivo. Regardless, our novel findings imply the translational potential of the combination for cancer treatment.