Activation of TAp73 and inhibition of TrxR by Verteporfin for improved cancer therapy in TP53 mutant pancreatic tumors

Aim: TAp73 is a tumor suppressor, which compensates for p53 loss and induces apoptosis in tumors in response to genotoxic stress or small-molecule treatments. Pancreatic ductal adenocarcinoma has a late onset of the disease, responds poorly to the existing therapies and has a very low survival rates. Result: Here, using drug-repurposing approach, we found that protoporphyrin IX (PpIX) and benzoporphyrin derivative (BPD) monoacid ring A activate TAp73 and induce apoptosis in pancreatic cancer cells. PpIX and BPD induce reactive oxygen species and inhibit thioredoxin reductase 1. Conclusion: Thus, PpIX and BPD target cancer cells’ vulnerabilities namely activate TAp73 tumor suppressor and inhibit oncogenic Trx1. Our findings may contribute to faster repurposing of PpIX and BPD to treat pancreatic tumors.

Two pathways leading to the induction of potent apoptosis by protoporphyrin IX and Verteporfin in mutant TP53 pancreatic cancer cells. One pathway is through inhibition of TAp73/MDM2 interactions and induction of TAp73-related apoptosis and the second pathway converges on inhibition of oncogenic TrxR and accumulation of reactive oxygen species. and now BPD induce apoptosis in cancer cells harboring TP 53 gene mutations, which is via reactivation of TAp73 tumor suppressor and inhibition of TrxR1.

Materials & methods
Cell culture Human pancreatic cancer cell lines, Paca3 (wt TP53), MiaPaca2 (mutant TP53R248W) and Panc1 (mutant TP53R273H) were kindly provided by Dr Rainer Heuchel from Karolinska Institutet, Sweden. Mouse pancreatic ductal cancer cell line Panc02 (wt TP53) was provided by Dr Maximilian Schnurr from University of Munich, Germany and made available by Dr Yihai Cao from Karolinska Institutet, Sweden.
Normal human pancreatic ductal epithelial cells (HPDE) immortalized by HPV transformation [18] were also provided by Dr Heuchel. Cells were maintained in DMEM medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin solution (Sigma-Aldrich, Munich, Germany). The cells were cultured at 37 • C in a humidified incubator with 5% CO 2 . The cells were routinely checked for mycoplasma during the whole duration of the project.

Chemicals
PpIX and BPD were purchased from Sigma-Aldrich dissolved in 100% DMSO (Sigma-Aldrich) to a final concentration of 2 mg/ml and stored in dark in amber tubes at room temperature (PpIX) or at -20 • C (BPD) until further use.
MTT reagent (Sigma-Aldrich) was reconstituted in PBS (prepared from 1× tablets Sigma-Aldrich) to 5 mg/ml, filter-sterilized and stored at +4 • C for a week or at -20 • C until further use.

Verteporfin uptake & cellular localization
Cells were incubated with 2.5 μg/ml of PpIX or BPD reconstituted in cell culture media for 3 h, washed three times with PBS and co-stained with ER-Tracker™ Green for live-cell imaging (Molecular Probes, OR, USA) according to manufacturer protocol. 4 ,6-diamidino-2-phenylindole (DAPI; Molecular Probes) was used for nuclear counterstaining. An organelle marker was excited using a 488 nm laser and BPD was excited with 420 nm (emission at 690 nm). Images were taken using an Olympus IX-71 microscope and DeltaVision SoftWoRx.

Cell viability
Cell viability was assessed by MTT assay as provided by the manufacturer (Sigma-Aldrich) and as described previously [15]. Briefly, 3000 cells were seeded in 96-well plate and allowed to adhere overnight. The cells were treated with the increasing concentrations of the compounds (PpIX or BPD) or 25 or 5 μM cisplatin or GEM for 72 h and the viability was measured using MTT reagent according to the manufacturer protocol.
future science group www.future-science.com Thioredoxin reductase activity assay For the enzymatic activity of purified TrxR1 we followed the protocol previously described [20,21]. Briefly, 10 nM of purified TrxR1 [22] was pre-incubated with BPD at the indicated concentrations (0-20 μM) in 50 μl phosphate EDTA (PE) buffer supplemented with 150 μM NADPH for 30 min at room temperature. Next, 2.5 mM DTNB and 150 μM NADPH were added to the samples just before measurement. DTNB reduction was performed for 6 min by detection of TNB anion formation detected as a change in the absorbance at 412 nm. For cellular TrxR activity, Panc1 and MiaPaCa2 cells were treated with 2.5 μg/ml PpIX or BPD for 6 h, after which they were harvested and resuspended in a Tris-EDTA (TE) buffer containing a protease inhibitor cocktail (complete, mini protease cocktail tablets, Roche, Solna, Sweden). Cells were sonicated and the total protein concentration of the supernatant was determined using a Bradford reagent kit (Bio-Rad Laboratories, Solna, Sweden). Cellular TrxR1 activity was determined using the previously described endpoint Trx-dependent insulin reduction assay [20]. Briefly, total cellular protein (20 μg) was incubated with 15 μM recombinant human wt Trx in the presence of 297 μM insulin, 1.3 mM NADPH, 85 mM Hepes buffer (pH 7.6) and 13 mM EDTA for 30 min at 37 • C, in a total volume of 50 μl. The reaction was stopped by the addition of 200 μl of 7.2 M guanidine-HCl in 0.2 M Tris-HCl (pH 8.0) containing 1 mM DTNB. The activity was then determined by measuring absorbance at 412 nm using a VersaMax microplate reader (Molecular Devices, CA, USA) with a background absorbance reference for each sample, containing all components except purified TrxR, incubated and treated in the same manner.

Cell death detection
Propidum iodide (PI) and FITC-Annexin V (from BD Biosciences, CA, USA) staining was performed according to the manufacturer's protocols and FACS analysis was carried out using the CELLQuest software (CELLQuest, NJ, USA) as described previously [23].

ROS measurement
Cells were treaded with compounds for 16 h, washed with PBS and then the generation of ROS assessed by flow cytometry. ROS were detected using 2 ,7 -dichlorofluorescein (H 2 DCF; Sigma-Aldrich) and method described previously [21].
Hydroethidine staining was performed according to the manufacturer protocol (Life Technologies, Stockholm, Sweden). Data analysis was performed with the CELLQuest software (CELLQuest, BD Biosciences, Franklin Lakes, NJ, USA).

Western blotting
Western blotting was performed according to the standard protocol. Briefly, 100 μg of total cell lysate was subjected to electrophoresis, after transfer the following antibodies were used to detect proteins: anti-TAp73

BPD localize into the cytoplasm of Panc1 cancer cells
Pancreatic ductal adenocarcinomas (PDACs) are cancers of poor clinical outcome due to the development of chemoresistance and disease relapse with mutant p53 being associated with the aggressiveness of the disease. Whether the re-activation of TAp73 can compensate for the p53 loss in cancers with the TP 53 gain-of-function mutations has not been unequivocally demonstrated yet. Therefore, we tested the outcome of TAp73 activation in PDAC cancer cell lines harboring hot-spot TP 53 R273H mutation.
First, we assessed the subcellular localization of BPD in Panc1 mutant TP53R273H-expressing cancer cells ( Figure 1A). Fluorescence microscopy indicated that BPD (red fluorescence emission) is uptaken to high degree by cancer cells 3 h after incubation with the drug. Next, we found that verteporfin localized in large in cytoplasm and partially in the endoplasmic reticulum. This is in agreement with the previously published data. PpIX & BPD inhibit proliferation of pancreatic cancer cells but not of nontransformed human ductal epithelial cells To assess the impact of PpIX and BPD on the proliferation of pancreatic cancer cells we performed MTT viability assay. The standard treatment for the advanced pancreatic cancer includes monotherapy with GEM. Therefore, we used GEM in the most of the experiments for the comparative purposes.
MTT showed that PpIX and BPD inhibit cancer cells' proliferation in a dose-dependent fashion ( Figure 1B-E). In agreement with the previous data, cancer cells expressing wt p53 protein (PaCa3) were sensitive to 72 h-treatment with PpIX [15] ( Figure 1B) and showed even higher growth inhibition rate after BPD treatment. Interestingly, mutant p53-harboring cells (Panc1) were more sensitive to PpIX than wtp53-expressing PaCa3 cells and slightly less sensitive to BPD ( Figure 1C).
The calculated IC 50 value for PpIX was 4.4 μM for PaCa3 (wtTP53) cells and 3.5 μM for Panc1 (mtTP53R273H) and 2.2 and 2.8 μM for BPD, respectively. future science group www.future-science.com Of note, murine (Panc02) and human (MiaPaca2) pancreatic cancer cells were less sensitive to PpIX and BPD at the concentrations tested ( Figure 1D and data not shown) and normal human epithelial ductal cells were insensitive ( Figure 1E). On the other hand, all cell lines applied in the study were highly susceptible to low dose (25 μM) cisplatin and 5 μM GEM.
Consistent with the previous data, wtp53 cancer cells were sensitive to PpIX [15]. Interestingly, mutant p53 Panc1 cancer cells were the most sensitive among the cell lines tested. Based on the above and previous studies, we selected 2.5 μg/ml concentration for further studies on the cell death induction upon reactivation of TAp73 in mutant p53 pancreatic cancer cells.

PpIX & BPD induce apoptosis in PDAC cells
To investigate if the growth inhibition observed in the MTT viability assays was a consequence of the induction of apoptosis, we treated PaCa3 and Panc1 cells for 48 h with 2.5 μg/ml PpIX or BPD and performed flow cytometry analysis using PI and Annexin V stainings. We found that Panc1 undergoes high late apoptosis after PpIX treatment, as demonstrated by above 40% increase of PI positive cells relative to the untreated control. At this time-point, Annexin V positivity, indicating early apoptosis, increased by 12 or 9% after PpIX or BPD treatments, respectively ( Figure 2C & D). In the PaCa3 cells harboring wtTP53, we observed 10 and 17% increase in PI positive cells after PpIX or BPD treatment, respectively. The increase in Annexin V-positive cells was 5% for PpIX and 11% for BPD. Thus, both compounds induce early and late apoptosis in the tested cancer cell lines. Cisplatin, used as a positive control, induced high early apoptosis in PaCa3 (wtTP53) cells and early and late apoptosis in Panc1 (mtTP53) cells.

PpIX & BPD induce TAp73 & its pro-apoptotic targets
It has been demonstrated that PpIX reactivates TAp73 tumor suppressor and induces TAp73-dependent apoptosis in TP 53-null cancer cells [16]. To assess if the growth inhibition observed in pancreatic cancer cells upon treatment with PpIX and BPD is a result of the ongoing apoptosis, we evaluated the expression of TAp73 tumor suppressor and its pro-apoptotic targets in PaCa3 and Panc1 cancer cells. Western blotting showed that in the dark, both PpIX and BPD induce the levels of endogenous TAp73, which correlates with the accumulation of its proapoptotic targets PUMA, Bax and Bid ( Figure 3A-C) in both cell lines. Time-dependent accumulation of the cleaved PARP was detected in Panc1 cells. Next, in Panc1 cells, PpIX and BPD upregulated the expression of bax, bid, puma and noxa genes ( Figure 3D). This indicates that apoptosis is due to the activation of TAp73 tumor suppressor ( Figure 3A & C). Interestingly, we detected upregulation of the antioxidant response proteins, HO-1 and Nrf2 ( Figure 3A-D). PpIX and BPD did not induce PARP cleavage or accumulation of TAp73 and its proapoptotic targets in nontransformed HPDE cells at the concentrations tested ( Figure 3E). We have observed downregulation of TAp73 after PpIX and BPD treatment; however, the mechanism remains to be elucidated.

PpIX & BPD induce ROS in PDAC
Since we detected accumulation of the antioxidant response proteins after treatment with PpIX and BPD in pancreatic cancer cells; we thus sought to investigate if PpIX and BPD elevate ROS. We assessed the levels of ROS using DCFDA and HE probes. Figures 4A-C Figure 1A). Importantly, the compounds did not induce ROS in HPDE cells (Supplementary Figure 1B).

PpIX & BPD are inhibitors of thioredoxin reductase
PpIX was first identified as an inhibitor of thioredoxin reductase in an in vitro screen for the inhibitors of the enzymes [24]. Since we observed significant induction of ROS in cancer cells treated with PpIX or BPD and inhibition of TrxR is known to cause accumulation of ROS, we reasoned that apart from TAp73-dependent apoptosis, PpIX and BPD inhibit TrxR which might explain the elevated ROS levels detected in the treated cancer cells.
As demonstrated in Figure 5A, recombinant TrxR was inhibited by both, PpIX and BPD. To assess if PpIX and BPD inhibit TrxR activity in cancer cells also, we treated Panc1 cells with the increasing concentrations of PpIX or BPD for 6 h and assayed the activity using the end point Trx-dependent insulin-reduction assay. PpIX and BPD treatment inhibited TrxR1 in Panc1 cancer cells ( Figure 5B) as well as in other mutant p53 harboring cell line, MiaPaCa2 (mtTP53R248W). Taken together, these data demonstrate that BPD, similarly to PpIX, is a potent inhibitor of TrxR in pancreatic cancer cells.

Discussion
Several tyrosine kinases inhibitors were reported to have off-target effects. For example, BRAF inhibitor, vemurafenib, used for treating metastatic melanoma [25] with overall good clinical outcome [26], was manifested to induce unusual photosensitivity in patients [27]. A recent study by Klaeger et al. [28] showed that several kinase inhibitors, including vemurafenib, in addition to their designed targets effectively inhibit ferrochelatase, the last enzyme in the heme biosynthesis. The authors speculated that the accumulation of PpIX, resulting from inhibition of ferrochelatase might lead to the photosensitivity of the patients upon treatment with kinase inhibitors.
PpIX is a metabolite of γ-aminolevulinic acid, heme precursor and a pro-drug applied in the PDT of cancer [14]. We have shown in a drug-repurposing approach that PpIX re-activates wt p53 and TAp73 tumor suppressors and induces p53-or TAp73-dependent apoptosis in cancer cells. The mechanism of activation of p53 and TAp73 is via inhibition of p53/MDM2 and TAp73/MDM2(X) interactions [15,16]. Next, PpIX was identified in the in vitro screen as a potent, competitive inhibitor of thioredoxin reductase (TrxR) [24]. Thus, here we speculated that PpIX, exogenously delivered to pancreatic cancer cells harboring TP53 mutations, will induce apoptosis via activation of future science group www.future-science.com  the p53 protein family member, TAp73. We also reasoned that PpIX and its clinically applied analog BPD may simultaneously inhibit TrxR in pancreatic cancer cells.
The detailed analysis revealed that PpIX and BPD, known under the commercial name of Verteporfin, inhibit proliferation of pancreatic cancer cells without affecting nontransformed cells. The mechanism of growth inhibition was by the induction of apoptosis. Next, both compounds triggered antioxidant response and induced accumulation of ROS only in cancer cells but not in normal HPDE cells. The above prompted us to investigate if the robust induction of ROS is a consequence of the inhibition of TrxR. According to our findings, BPD inhibited the enzymatic activity of the recombinant thioredoxin reductase. Both PpIX and BPD inhibited the enzyme in Panc1 and MiaPaCa2 cancer cells as demonstrated by the Trx-dependent endpoint insulin reduction assay.
Inhibitors of TrxR are currently under preclinical development and are promising candidates for improved cancer therapy [29]. Several small molecules activating wt or mutant p53 were also shown to inhibit TrxR [21,30,31]. The small molecule APR-246, a mutant p53-reactivating compound in Phase II clinical development, is a pro-drug, converted in cells to methylene quinuclidinone [32]. Methylene quinuclidinone is a Michael acceptor and was demonstrated to target cysteines in the p53 core domain [33] and selenocysteine (Sec) residues in the C-terminal motif of TrxR1 [30]. Our findings demonstrate that PpIX and BPD inhibit TrxR. The mechanism by which this occurs remains to be elucidated. Structural studies on the heme-binding protein, heme oxygenase 2 (HO-2), revealed that in the oxidized state of cysteine residues in apoprotein, heme binds 2.5-fold more tightly than in the reduced state [34]. This, together with the finding that the Sec-to-Cys mutant of TrxR1 was resistant to inhibition by PpIX [24], strongly implies that Sec residue in TrxR might be the binding site of PpIX and BPD (Verteporfin). We showed that PpIX and now also BPD activates TAp73 in pancreatic cancer cells, which results in the induction of apoptotic cell death. In parallel, the compounds induce ROS, which contributes to cell killing. Induction of ROS levels is with high probability due to the inhibition of TrxR by PpIX and BPD.

Future perspective
Our findings might have an important clinical implication and could support fast repurposing of porphyrins in treating pancreatic cancer patients harboring TP53 mutations. Dual targeting of TAp73 and TrxR by small molecules that we have identified might bring therapeutic advantage in overcoming the development of the treatment-resistant disease. In particular, we foresee that the parallel activation of TAp73 and inhibition of TrxR will significantly increase the stress burden in already stressed cancer cells when compared with separate approaches, which will contribute to the selective tumor killing without affecting normal cells.
• PpIX and BPD generate reactive oxygen species in pancreatic cancer cells.
• PpIX and BPD inhibit thioredoxin reductase in vitro and in cancer cells.

Acknowledgments
The authors would like to thank Dr Elias Arnér for helpful discussions. Special thanks are addressed to Dr Alicja Sznarkowska and Sofia Parassia for valuable feedback. The authors would like to thank all of their colleagues for sharing their reagents. No writing assistance was utilized in the production of this manuscript.