Unraveling the Photodynamic Activity of Cationic Benzoporphyrin-Based Photosensitizers against Bladder Cancer Cells

In this study, we report the preparation of new mono-charged benzoporphyrin complexes by reaction of the appropriate neutral benzoporphyrin with (2,2′-bipyridine)dichloroplatinum(II) and of the analogs’ derivatives synthesized through alkylation of the neutral scaffold with iodomethane. All derivatives were incorporated into polyvinylpyrrolidone (PVP) micelles. The ability of the resultant formulations to generate reactive oxygen species was evaluated, mainly the singlet oxygen formation. Then, the capability of the PVP formulations to act as photosensitizers against bladder cancer cells was assessed. Some of the studied formulations were the most active photosensitizers causing a decrease in HT-1376 cells’ viability. This creates an avenue to further studies related to bladder cancer cells.


Introduction
Cancer is a broad term that defines a group of diseases that can develop almost anywhere in the body, induced by the uncontrolled overgrowth of abnormal cells resulting in DNA mutations and consequently to the destruction of normal tissues. Malignant disorders are the second leading cause of death worldwide [1,2].
Currently, the most used treatment approaches are based on surgical resection of the tumoral mass, radio-, immuno-, and chemotherapy. However, these treatments display several disadvantages, such as severe radiation damage, limited applicability, lack of specificity, and severe side effects [2].
Concerning the chemotherapeutic approach, platinum-based drugs are the most used drugs against solid tumors such as bladder, testicular, ovarian, lung, neck, or head [3][4][5][6]. The mechanism of action of platinum-based drugs is based on their capability to bind DNA strands, which avoids the DNA strand from unzipping, by blocking the replication process, having as a consequence the malignant cell death [7][8][9][10][11].
In 1978, cisplatin was the first FDA-approved platinum-based drug to be used as an anticancer agent. After this milestone, other platinum-based drugs, namely carboplatin and oxaliplatin, were developed and also approved throughout the world [3,12]. Some others have regulatory approval only in some countries (e.g., nedaplatin, miriplatin, loboplatin, or heptaplatin) or are currently under clinical trials [9,13]. Despite the wide use of platinumbased drugs, the treated patients experienced severe side effects related to their poor The synthesis of the positively monocharged benzoporphyrin derivatives 2 and 3 required the previous preparation of the scaffolds 1a,b following procedures already reported by us. Briefly, 2-formyl-5,10,15,20-tetraphenylporphyrin reacts with the adequate 3-or 4-acetylpyridine in the presence of NH 4 OAc and catalytic amounts of La(OTf) 3 in refluxing toluene for 4 h, under N 2 atmosphere [87,88].
Then, the precursors 1a,b reacted with (2,2 -bipyridine)dichloroplatinum(II) in refluxing CHCl 3 /MeOH (2:1) mixture for 24 h. After purification by column chromatography, the expected 2a or 2b derivatives were obtained in 68 and 87% yield, respectively. In the synthesis of the benzoporphyrin-platinum complex 2a, it was recovered 18% of the starting benzoporphyrin 1a. The lower reactivity of derivative 1a and, consequently, the low yield obtained for 2a are probably related with a hindrance effect due to the bulkiness of the (2,2 -bipyridine)chloroplatinum(II) moiety and the proximity of the nitrogen atom of the pyridyl unit to the benzoporphyrin core. An extension of the reaction time from 24 to 48 h led to a slight improvement in the yield of benzoporphyrin-platinum complex 2a to 72%.
It is worth noting that all the attempts to prepare the analog benzoporphyrin-platinum complex bearing a 2-substituted pyridyl moiety failed, probably due to an even higher steric hindrance effect induced by the bulkiness of the (2,2 -bipyridine)chloroplatinum(II) moiety and the proximity of the pyridyl nitrogen atom with the benzoporphyrin core.
Compounds 3a and 3b were prepared in 97 and 98% yield, respectively, from the corresponding neutral derivative 1a,b by alkylation reaction with iodomethane in DMF at 40 • C for 24 h (Scheme 1). This is a typical and well-established protocol to prepare porphyrinoids bearing pyridinium moieties and, once again, it revealed to be effective for preparing the benzoporphyrin derivatives 3a,b. of this type of cancer can benefit from the easy light delivery via insertion of a light source into urethra [86].

Synthesis
The synthesis of the positively monocharged benzoporphyrin derivatives 2 and 3 required the previous preparation of the scaffolds 1a,b following procedures already reported by us. Briefly, 2-formyl-5,10,15,20-tetraphenylporphyrin reacts with the adequate 3-or 4-acetylpyridine in the presence of NH4OAc and catalytic amounts of La(OTf)3 in refluxing toluene for 4 h, under N2 atmosphere [87,88].
Then, the precursors 1a,b reacted with (2,2′-bipyridine)dichloroplatinum(II) in refluxing CHCl3/MeOH (2:1) mixture for 24 h. After purification by column chromatography, the expected 2a or 2b derivatives were obtained in 68 and 87% yield, respectively. In the synthesis of the benzoporphyrin-platinum complex 2a, it was recovered 18% of the starting benzoporphyrin 1a. The lower reactivity of derivative 1a and, consequently, the low yield obtained for 2a are probably related with a hindrance effect due to the bulkiness of the (2,2′-bipyridine)chloroplatinum(II) moiety and the proximity of the nitrogen atom of the pyridyl unit to the benzoporphyrin core. An extension of the reaction time from 24 to 48 h led to a slight improvement in the yield of benzoporphyrin-platinum complex 2a to 72%.
It is worth noting that all the attempts to prepare the analog benzoporphyrin-platinum complex bearing a 2-substituted pyridyl moiety failed, probably due to an even higher steric hindrance effect induced by the bulkiness of the (2,2′-bipyridine)chloroplatinum(II) moiety and the proximity of the pyridyl nitrogen atom with the benzoporphyrin core.
Compounds 3a and 3b were prepared in 97 and 98% yield, respectively, from the corresponding neutral derivative 1a,b by alkylation reaction with iodomethane in DMF at 40 °C for 24 h (Scheme 1). This is a typical and well-established protocol to prepare porphyrinoids bearing pyridinium moieties and, once again, it revealed to be effective for preparing the benzoporphyrin derivatives 3a,b. The structures of compounds 2a,b and 3a,b were confirmed by NMR spectroscopy and mass spectrometry data (see Figures S1-S22. The mass spectra of the mono-charged benzoporphyrin derivatives 2a,b and 3a,b exhibit the m/z peak corresponding to the [M + 2H] + or [M + 2H] +• molecular ion. However, it is important to point out that, for all the compounds synthesized, the corresponding [M + 2H] + and [M + 2H] +• species are formed in the gas phase due to the reduction in one of the β-pyrrolic positions. Similar results were already observed by us in a previous publication [89]. The 1 H NMR analysis of the derivatives 2a,b and 3a,b supports also the proposed structures with the resonances of six β-pyrrolic protons appearing in the aromatic region, between δ 8.98 and δ 8.60 ppm. In all the 1 H NMR spectra, the distinguishing singlet at ca. δ −2.7 ppm generated by the resonance of the N-H protons from the free-base benzoporphyrin core is also observed.
The 1 H NMR spectra of compound 3a,b show a singlet at around δ 4.67 ppm generated by the resonances of the methyl groups' protons confirming the formation of the pyridinium moiety. All the remaining signals generated by the resonances of the protons from the benzoporphyrin moieties, as well as from the phenyl ring at the meso-positions appear in the aromatic region, being the most deshielded signals generated by the protons near the nitrogen atom at the pyridinium unity.
In the 1 H NMR of compounds 2a,b, the most deshielded signals (ca. 9.6 ppm) are due to the resonances of the protons from the 6 positions of the 2,2 -bipyridine unit, while the protons from the pyridyl units are shielded by the presence of the platinum core when compared with the ones from the neutral precursors. The signals generated by the remaining protons from the benzoporphyrin core are not significantly affected by the 2,2bipyridine)chloroplatinum(II) unit and display similar chemical shifts to the ones observed for the corresponding 3a,b derivatives.
The absorption, steady-state fluorescence emission, and excitation spectra of the new derivatives were recorded in N,N-dimethylformamide (DMF) solution at 298 K. The UV-Vis spectra of compounds 2a,b and 3a,b were not significantly affected by the modification performed into the benzoporphyrinic ring, presenting the typical features of free-base porphyrin derivatives due to π-π* transitions [90][91][92]. Both series 2a,b and 3a,b exhibit a strong Soret band at ca. 427 nm assigned to allowed S 0 → S 2 transitions and two Q bands at approximately 520 and 595 nm due to S 0 → S 1 transitions. Additionally, absorption bands centered from 282 to 324 nm were observed for derivatives 2a,b due to the ligand to metal charge transfer (LMCT) transition from the bipyridine moiety to the platinum ion. The fluorescence emission spectra of compounds 2 and 3 obtained after excitation at approximately 595 nm present two bands centered at ca. 660 and 720 nm (    The resemblance between the absorption and excitation spectra rules out the presence of emissive impurities. The large Stokes shift (ca. 66 nm) displayed by both prepared series of benzoporphyrin derivatives are indicative of a change in the electronic nature of the excited state compared with that of the ground state. The fluorescence quantum yields (Φ F ) determined by the internal reference method with respect to a solution of 5,10,15,20tetraphenylporphyrin (TPP) in DMF as standard (Φ F = 0.11) [93,94] are shown in Table 1. The Φ F values range from 0.05 to 0.07, and no noticeable differences were induced by the presence of the cyclometalated (2,2 -bipyridine)platinum(II) moieties or the methyl groups.

Incorporation into PVP Micelles
Benzoporphyrin derivatives 2 and 3 were used to prepare polyvinylpyrrolidone (PVP) formulations aiming to avoid aggregation phenomena in aqueous medium due to their low hydrophilic character (miLog P: 8.45-9.91) [95]. This is a low-cost approach that requires the dissolution of both PS and N-vinylpyrrolidone (VPD) in CHCl 3 solution, stirring the resulting mixture for 2 h at room temperature, and then solvent removal under a nitrogen flow. The resulting residue, after being maintained for 48 h at 40 • C, was dissolved in water and submitted to dialysis affording the expected PVP-PS formulations PVP-2a,b and PVP-3a,b. It is worth noting that the PVP-PS formulations prepared retained the photophysical features previously discussed for PS 2 and 3 without noticed changes due to their incorporation into PVP micelles (Table 1, Figures 1C,D and S23C,D).
VPD was selected as the monomer to prepare PVP formulations due to their already reported features, namely, pharmacokinetic and pharmacological properties, non-toxicity, and water-solubility of the obtained micelles [96]. This strategy allows to improve the hydrophilicity of biologically active drugs and is being efficiently used to solubilize neutral porphyrin-base PS in water with positive effects in photodynamic processes [97][98][99][100]. Moreover, this carrier demonstrated to be non-toxic for both normal and cancer cells after PDT treatment [100]. However, to the best of our knowledge, this approach has not been used with benzoporphyrin-type derivatives.

Photostability and Singlet Oxygen Generation
Photostability is a relevant parameter to evaluate the PS potential to be used in photodynamic processes such as PDT. The photostability assays for PVP-PS formulations PVP-2a,b and PVP-3a,b were performed in PBS by monitoring the Soret band decay (λ max = 425 nm) after irradiation with white light at an irradiance of 20 mW·cm −2 for different irradiation periods. After 30 min of irradiation, formulations PVP-2b and PVP-3a,b showed a Soret band absorption decay ranging from 11 to 16%, while, for PVP-2a, the decrease was 28% (Table S1). As such, it is possible to conclude that the two synthetic strategies used to modify the benzoporphyrin core and the incorporation of the obtained benzoporphyrin derivatives into PVP micelles allowed to afford PVP formulations with adequate photostability.
Besides photostability, another relevant feature for a PS to be used in PDT is its capability to generate ROS, namely, singlet oxygen ( 1 O 2 ) [101]. The generation of 1 O 2 by the PVP-PS formulations was qualitatively determined by monitoring at 415 nm, the photooxidation of the 1 O 2 quencher 1,3-diphenylisobenzofuran (DPiBF) to the colorless o-dibenzoylbenzene, after the Diels-Alder-like reaction [102][103][104]. The irradiations of each PVP-PS formulation in DMF and in the presence of dioxygen were performed at a fluence of 11 mW·cm −2 and the results obtained from the DPiBF time-dependent photodecomposition are summarized in Figure 2. It is worth noting that, for these assays, a PVP-TPP formulation was prepared to be used as reference, since 5,10,15,20-tetraphenylporphyrin (TPP) is pointed out as a good singlet oxygen generator [105].
The PVP-2a,b formulations showed to be better 1 O2 generators than PVP-3a,b formulations, despite both series displaying worse capability than the one presented by the PVP-TPP formulation. The capability of PVP-3a,b formulations is 10% of the one exhibited by the reference and is not significantly affected by the position of the charge in the pyridinium unit. The ability of the PVP-2a and PVP-2b formulations to produce 1 O2 is 60 and 45% lower when compared with the reference, respectively, but even so, it is 4-to 5-fold higher than the ones presented by the PVP-3a,b formulations. Yet, for the PVP-2a,b for- It is worth noting that, for these assays, a PVP-TPP formulation was prepared to be used as reference, since 5,10,15,20-tetraphenylporphyrin (TPP) is pointed out as a good singlet oxygen generator [105].
The PVP-2a,b formulations showed to be better 1 O 2 generators than PVP-3a,b formulations, despite both series displaying worse capability than the one presented by the PVP-TPP formulation. The capability of PVP-3a,b formulations is 10% of the one exhib-ited by the reference and is not significantly affected by the position of the charge in the pyridinium unit. The ability of the PVP-2a and PVP-2b formulations to produce 1 O 2 is 60 and 45% lower when compared with the reference, respectively, but even so, it is 4-to 5-fold higher than the ones presented by the PVP-3a,b formulations. Yet, for the PVP-2a,b formulations, the position of the charge and the (2,2 -bipyridine)chloroplatinum(II) unit influences the PS production of 1 O 2 , with 2b being the one with the better yield of 1 From the analysis of Figure 2, it is obvious that the absorbance of DPiBF, when irradiated in the absence of a PS, remains almost unchanged, as well as in the presence of just PVP. These results revealed the potential of the PVP-PS formulations prepared to be used in PDT and prompted us to evaluate their efficiency as PSs against bladder cancer cells.     Table 2.

Cell Viability after PDT Treatment with PVP-2a,b and PVP-3a,b Formulations
The photodynamic effect of the PVP-2a,b and PVP-3a,b formulations was evaluated in bladder cancer cell line HT-1376 at 2.5, 5.0, 10.0, and 12.5 µM. The cell line was incubated in the dark for 4 h with the PVP formulations and then irradiated with white light for 40 min with an irradiance of 20 mW.cm −2 . The cell viability was accessed by the MTT colorimetric assay after 24 h of PDT protocol. The results obtained are presented in Figure 4, and the IC50PDT values (for a fluence rate of 20 mW·cm −2 ) of all formulations are plotted in Table 2.   The results showed that all PVP formulations caused a decrease in the HT-1376 cell viability, being possible to observe that the phototoxicity increased with the PS concentration. PVP-3a,b formulations showed to be the most active PSs causing a decrease in HT-1376 cell viability higher than 80% for the maximum concentration. This superior efficiency is also proved by their lower IC50 PDT values (5.58 and 5.51 µM for PVP-3a and PVP-3b, respectively). Although also highly efficient, the phototoxicity values of PVP-2a,b formulations were lower, causing a reduction in the HT-1376 cell viability of around 70% for the highest concentration. This fact could be explained by the lower PVP-2a,b accumulation inside HT-1376, when compared with the internalization of PVP-3a,b into the bladder cancer cell line.
The same protocol without the irradiation procedure was performed to evaluate the cytotoxic effect of all formulations. As expected, no cytotoxicity was observed in  Figure S24).
It is well known that the efficiency of PDT depends on the intrinsic efficacy of the PS, and there are many PS properties that need to be taken into account, such as 1 O 2 generation, aggregation, and photodegradation behavior [106]. Although being the least efficient in the 1 O 2 generation, it is noteworthy that formulations PVP-3a,b were the ones that demonstrated higher photostability and higher internalization into the cancer cell line. The conjugation of these important properties could explain the higher efficiency of these formulations in the decrease in the HT-1376 cell viability after the PDT procedure.
Moreover, it is also important to note that, in this particular case, the insertion of a 2,2'-bipyridine-platinum moiety into the benzoporphyrin macrocycle did not enhance the PDT effect in the cancer cell line.

General Remarks
1 H and 13 C NMR spectra were recorded on a Bruker Avance 300 spectrometer at 300.13 MHz and a Bruker Avance 500 spectrometer at 500.12 and 125.77 MHz, respectively. CDCl 3 was used as solvent and tetramethylsilane (TMS) as internal reference. Chemical shifts are expressed in δ (ppm) and the coupling constants (J) are expressed in Hertz. HRMS were recorded on a VG AutoSpec M mass spectrometer using MeOH as solvent and 3-nitrobenzyl alcohol (NBA) as matrix. The UV-Vis spectra were recorded on an UV-2501 PC Shimadzu spectrophotometer using DMF as solvent. Fluorescence emission spectra were recorded on a Horiba Jobin-Yvon Fluoromax 3 spectrofluorometer and fluorescence quantum yields of compounds 2a,b and 3a,b and PVP-PS formulations PVP-2a,b and PVP-3a,b were measured by using a solution of TPP in DMF as a standard (Φ F = 0.11). Flash chromatography was carried out using silica gel (230-400 mesh), and preparative thin-layer chromatography was carried out on 20 × 20 cm glass plates coated with silica gel (1 mm thick). The reactions were routinely monitored by thin-layer chromatography (TLC) on silica gel precoated F254 Merck plates.

Synthesis of the Benzoporphyrin Precursors 1
Precursors 1a and 1b were synthesized according to the previous procedures described. The structures of both porphyrin-based PS were confirmed by 1 H-NMR spectroscopy and mass spectrometry, and the data are in accordance with the data reported [87].

Synthesis of Porphyrin-Platinum(II) Complexes
(2,2 -Bipyridine)dichloroplatinum(II) (13.7 mg, 32.4 µmol) was added to a solution of the adequate benzoporphyrin 1a or 1b (20 mg, 27 µmol) in a CHCl 3 /MeOH mixture (2:1, 1.5 mL) in a sealed tube. The reaction mixture was stirred at 100 • C for 24 h. Then, 0.2 M aqueous saturated solution of KPF 6 was added to the reaction mixture, and the precipitate obtained, corresponding to the PF 6 − salt, was filtered, dissolved in CH 2 Cl 2 , and washed with distilled water, and the organic layer was collected. The solvent was evaporated under reduced pressure, and the crude purified by column chromatography using CH 2 Cl 2 /MeOH (98:2) as the eluent. The benzoporphyrin-platinum(II) complexes 2a and 2b were obtained pure after crystallization from CH 2 Cl 2 /hexane.

Synthesis of Cationic Benzoporphyrins 3a,b
The appropriate neutral benzoporphyrin 1a,b (20 mg, 27 µmol) was dissolved in DMF (1.0 mL) and to the solution was added an excess of iodomethane (0.1 µL, 1.6 mmol). The resulting mixture was stirred at 40 • C for 18 h, and, after this period, diethyl ether was added. The precipitate obtained was filtered through a cotton pad and washed with diethyl ether. Then, the solid was dissolved with a CH 2 Cl 2 /MeOH (95:5) mixture, the solvent evaporated, and the expected compounds 3a and 3b were obtained in almost quantitative yield, after hexane/CH 2 Cl 2 crystallization.

General Procedure to Prepare PVP-PS Micelles
Chloroform solutions of N-vinylpyrrolidone (100 mg) and compounds 2a,b or 3a,b (10% w/w) were mixed in a Becker and the solution stirred for 2 h at room temperature for a full homogenization. Then, the solvent was evaporated under nitrogen flow and the reddish-brown solid obtained was dried in an oven at 40 • C for 48 h. The resulting residues were dissolved in 2 mL of water and submitted to dialysis in distilled water at pH 7. After this approach, PVP-2a,b and PVP-3a,b formulations were obtained.

Singlet Oxygen Generation
To evaluate the ability of PVP formulations (PVP-2a,b, and PVP-3a,b) to generate singlet oxygen ( 1 O 2 ), in a 1 × 1 cm cuvette, we prepared 3 mL solutions, each one containing a PS (0.5 µM) and DPiBF (50 µM) in DMF. The solutions were irradiated with a red-light LED board (630 ± 20 nm) at an irradiance of 11 mW·cm −2 for 15 min at room temperature under gentle magnetic stirring. Control assays using a DPiBF solution at 50 µM and the PVP and PVP-TPP formulations (0.5 µM) and just a DPiBF (50 µM) were also performed. HT-1376 cells were seeded (9.4 × 10 4 cells.cm −2 ) in 96-well cell culture plates and maintained in culture medium under an air atmosphere containing 5% of CO 2 overnight. The cells were washed twice with PBS and incubated with 2.5, 5.0, 10.0, and 12.5 µM of PVP-2,ab and PVP-3a,b formulations for 4 h in the dark. The cells were then washed twice with PBS and covered with 100 µL of fresh medium. The cells were irradiated for 40 min with white light delivered by an illumination system (LC-122 LumaCare, London, UK) equipped with a halogen/quartz 250 W lamp coupled to the selected interchangeable optic fiber probe (400-800 nm) at a fluence rate of 20 mW·cm −2 . After irradiation, the cells were incubated in a humidified incubator with 5% of CO 2 atmosphere and 95% of air. After 24 h of the PDT protocol, cell phototoxicity was determined by measuring the ability of cancer cells to reduce 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide (MTT, Sigma), to a colored formazan using a microplate reader (Synergy HT, Biotek, Winooski, VT, USA). The data were expressed in percentage of control (i.e., optical density of formazan from cells not exposed to PVP formulations).
The dark toxicity of PVP-2,ab and PVP-3a,b formulations was evaluated under the same protocol, though without the irradiation procedure.

Statistical Analysis
The results are presented as mean of at least 3 independent assays with 3 replicates per assay. The statistical analysis was performed with GraphPad Prism (GraphPad Software, San Diego, CA, USA). Statistical significance among the conditions was assessed using the nonparametric Mann-Whitney test.

Conclusions
In summary, two different approaches to prepare mono-charged benzoporphyrinbased Ps were efficiently developed. The reaction of the neutral precursors with (2,2bipyridine)dichloroplatinum(II) allows preparing the corresponding benzoporphyrinplatinum(II) modified at the isoindole-type unit in good-to-excellent yields, while the alkylation with iodomethane gives the cationic benzoporphyrins in almost quantitative yields. All the mono-cationic benzoporphyrin derivatives prepared were successfully incorporated in PVP micelles, allowing to improve their solubility in aqueous medium.
Both compounds and PVP-PSs formulations display photophysical features typical of free-base porphyrin derivatives, which are not noticeably affected by the different moieties inserted into the benzoporphyrin core. The PVP-PSs formulations prepared are stable when irradiated with white light and all are able to generate singlet oxygen. However, the PVP formulations prepared with the benzoporphyrin-platinum(II) derivatives exhibit better performance in the 1 O 2 generation.
Under the context of PDT evaluation, PVP-3a,b formulations demonstrated higher photostability, higher internalization into the cancer cell line, and, consequently, were the most active PSs causing a decrease in HT-1376 cell viability higher than the corresponding formulations with benzoporphyrin-platinum(II) derivatives. Moreover, the synthetic approach to prepare the mono-cationic derivatives 3a,b exhibits a much better cost-effectiveness relationship when compared with the route to prepare the corresponding derivatives 2a,b, due to affording almost quantitative yields for both compounds, as well as the high cost associated with (2,2 -bipyridine)chloroplatinum(II) or its synthesis.
Additionally, none of the formulations tested exhibit dark toxicity for HT-1376 cell line, suggesting that the phototoxic effect is due to the reactive oxygen species production under irradiation. These promising results encourage further in vitro and in vivo test studies of benzoporphyrin derivatives as prototypes of future PDT agents.  Table S1: photostability data of PVP formulations.