Synthesis and Evaluation of Novel meso-Tetraphenyltetrabenzoporphyrins for Photodynamic Therapy

Abstract

To discover effective photosensitizers for photodynamic therapy (PDT), a series of new meso-tetraphenyltetrabenzoporphyrin (m-Ph₄TBP) derivatives were designed, prepared, and characterized. All m-Ph₄TBPs own two characteristic absorption bands in the range of 450–500 and 600–700 nm and have the ability to generate singlet oxygen upon photoexcitation. Most of the m-Ph₄TBPs demonstrated high photoactivity, among which compounds I₄, I₆, I₁₂, and I₁₃ induced apoptosis and also exhibited excellent photodynamic activities in vivo. Nonetheless, the liver organs of the I₄ and I₆–PDT groups showed clear calcifications, whereas the liver tissues of the other PDT groups showed no calcification. It was indicated that compared to phenolic m-Ph₄TBPs, glycol m-Ph₄TBPs exhibited superior biological safety in mice. According to comprehensive evaluations, m-Ph₄TBP I₁₂ displayed excellent photodynamic antitumor efficacy and biological safety and can be regarded as a promising antitumor drug candidate.

Photodynamic therapy (PDT) is a minimally invasive therapeutic method which has been approved for the treatment of several cancer types such as lung cancer, breast cancer, prostate cancer, and bladder cancer.¹⁻³ PDT involves three basic elements: photosensitizer (PS), light, and oxygen. The combination of these three elements results in the generation of reactive oxygen species (ROS) like hydroxyl radical (·OH) and singlet oxygen (¹O₂) which can interact with biological molecules to inhibit the growth of tumor cells and/or destroy tumor tissues through apoptosis or necrosis.⁴⁻⁷

Porphyrin and its derivatives play a crucial role among the marketed photosensitizers; nevertheless, they are not without drawbacks. As the first porphyrin-derived PS drug used in clinics with pronounced photodynamic antitumor activities, Photofrin II⁸⁻¹¹ consists of several oligomeric hematoporphyrins which can be retained in the skin of patients after administration and cause skin phototoxicity for more than 4 weeks. Another porphyrin-derived drug known as temoporfin, which has absorption at 652 nm, is unstable at room temperature and must be stored at −20 °C.¹²⁻¹⁴ 5-Aminolevulinic acid hydrochloride, which could be converted into protoporphyrin (IX) in vivo, and shows instability in aqueous solutions upon use as a pro-drug.¹⁵

Porphyrin, a conjugated tetrapyrrole ring, has a major Soret (S) band at 410 nm and four weak Q bands (500–650 nm).¹⁶ The penetration depth of 410 nm light through tissues is <0.5 mm, which is only suitable for superficial lesions,¹⁷ whereas the transmission depth of 600–800 nm light is 3–6 mm, which is suitable to treat larger sized and deeper seated tumors. Synthesis of porphyrin-based PSs with strong absorption in the range of 600–800 nm can be achieved by three methods, namely, the reduction of porphyrin into chlorin,¹⁸ modification of the substitution pattern,¹⁹ and expansion of the porphyrin ring,²⁰ which so far is the most effective strategy. Currently, there is a lot of literature about porphin (porphyrin) and its derivatives, while only a few papers have been reported about benzoporphin (benzoporphyrin) and its derivatives, including meso-tetraphenyl-substituted tetrabenzoporphyrins.

Barret et al.²¹ synthesized meso-tetraphenyl-tetrabenzoporphyrin (m-Ph₄TBP) with four benzene rings separately fused to four pyrroles of the porphyrin ring, but the resulted compounds were hard to purify because of their sparing solubility. Ruppel et al.²² synthesized a series of phenyl-substituted tetrabenzoporphyrins with strong absorption and high singlet oxygen quantum yields, but they did not investigate regarding photodynamic activity in vitro and in vivo, and most of their compounds are hydrophobic, which are unsuitable to prepare the subsequent solutions for administrations in vivo such as intravenous injection. To improve the water solubility of m-Ph₄TBP, Menard et al.²³ linked sugar units with m-Ph₄TBP, but the resulting compounds were difficult to purify, and no in vivo study was reported.

In our previous works, it was found that conjugating polar or nonpolar groups at the periphery of the porphin ring can ameliorate or enhance the druggability of PSs.^24,25 Therefore, a series of hydroxy-functionalized m-Ph₄TBPs were designed and synthesized, namely, compounds I_1–6 with hydroxy groups at the R⁵ or R⁶-position of the meso-phenyl moieties; I_7–12 with monoethylene glycol groups at the R⁵or R⁶-position of the meso-phenyl moieties; and I_13–15 with C3- to C5-alkyl chain glycol groups at the R⁶-position of meso-phenyl moieties. In addition, their photodynamic antitumor activities were evaluated in vitro and in vivo.

Compounds 3_A-O, a series of reaction intermediates of m-Ph₄TBPs, were obtained by incorporating phenol or C2- to C5-alkyl chain glycol groups into meso-phenyl moieties. This allowed researchers to investigate the effects of various meso-phenyl moieties on the photoactivities of m-Ph₄TBP. Through a substitution reaction between hydroxybenzaldehydes and acetic anhydride in an aqueous NaOH solution (0.5 M) at 0 °C, phenolic aldehydes protected with acetate groups (2_A-F) were prepared as shown in Scheme S1. By using a nucleophilic substitution reaction between hydroxybenzaldehydes and 2-bromoalkyl acetate in the presence of excess K₂CO₃ in N,N-dimethylformamide, aromatic aldehyde derivatives (2_G-L, N–O) were obtained. Compound 2_M was prepared by reacting acetated phenolic aldehydes with 3-bromo-propanol in N,N-dimethylformamide at 80 °C in the presence of Cs₂CO₃.²⁶

As an important intermediate, tetrahydroisoindole 1 was obtained from cyclohexene through four-step reactions with benzenesulfinic acid, N-iodosuccinimide, ethyl isocyanoacetate, and potassium hydroxide as reagents in sequence, according to literature.²⁷ All m-Ph₄TBP derivatives were prepared as shown in Scheme 1. Phenolic aldehydes protected with acetate groups were reacted with intermediate 1 under a catalytic amount of BF₃·Et₂O in DCM for 19 h at room temperature and then dehydrogenated by DDQ to give the corresponding tetracyclohexanoporphyrins (3_A-O) with a yield of 30–60%. 3_A-O was complexed with copper acetate in DMF at 80 °C for 0.5 h to give crude Cu-tetracyclohexanoporphyrins as a red solid. The obtained crude copper complexes were reacted with excess DDQ in toluene at 80 °C for 0.5 h to afford green Cu-m-Ph₄TBP complexes (4_A-O). 4_A-O was mixed with H₂SO₄–CF₃COOH (1:6, v/v) at 0 °C for 40 min to afford compounds I_1–6 as a green solid in a one-step reaction, while compounds I_7–15 were obtained in good yields after demetalation using CF₃COOH-H₂SO₄ and deprotection by potassium hydroxide solution in THF at 60 °C for 6 h.

Scheme 1.

Reagents and reaction conditions: (a) BF₃·OEt₂, DCM, rt, 19 h; then DDQ, 1 h; (b) Cu(OAc)₂, DMF, 80 °C, 0.5 h; (c) DDQ, toluene, 80 °C, 0.5 h; (d) H₂SO₄–CF₃COOH (1:6, v/v), 0 °C, 40 min; (e) 3M KOH-H₂O, THF, 60 °C, 6 h.

To study the influence of the molecular orbital energy gaps on the absorption wavelength of the glycol m-Ph₄TBP, DFT calculations were performed using B3LYP/6-31G (d). As shown in Figure 1, m-Ph₄TBPs I₈, I₉, and I₁₃ all had energy gaps (ΔE) between their lowest unoccupied molecular orbital (LUMO) and highest occupied molecular orbital (HOMO) that were all at the range of 2.36–2.37 eV smaller than those of 5, 10, 15, and 20-tetraphenylporphyrin (TPP: ΔE = 2.69 eV), suggesting that the porphyrin ring’s expansion could cause a sizable red-shift in the absorption wavelength. Furthermore, there were minor influences on the energy gap from peripheral groups that were distant from the π-conjugate system of m-Ph₄TBP.

Figure 1.

Energy diagram and frontier molecular orbitals HOMO and LUMO of the respective I₈, I₉, and I₁₃.

The UV–visible absorption of I_1–15 was tested at 25 °C with a concentration of 2 μM in DMSO (Figure 2). Each of the novel m-Ph₄TBPs displayed three absorption peaks at approximately 590, 645, and 700 nm, in addition to a strong peak at about 470 nm. For the following experiments, a 650 nm laser light was employed since the absorption wavelength of the Q bands of m-Ph₄TBPs was located at the phototherapeutic window, and a relatively greater molar extinction coefficient was observed at ∼648 nm. Surprisingly, all hydroxy-functionalized m-Ph₄TBPs displayed minor shifted absorptions of 0–7 nm at the maximum absorption peak in Q bands in Table 1, while they exhibited big differences on the molar extinction coefficients. All values of the molar extinction coefficients displayed that varying substituents of meso-phenyl moieties with hydroxy, methoxy, halogen, nitro, glycol, and C3- to C5-alky chain glycol groups can change the absorption intensities of m-Ph₄TBPs.

Figure 2.

(A, B) Absorption spectra of respective compounds at 2 μM in DMSO.

Table 1. Photophysical Properties of I_1-15 in Q Bands.

Compound	λabs [nm](ε [M–1 cm–1])	λabs [nm](ε [M–1 cm–1])
I1	467(52500)	598(2000)	646(7000)	698(1500)
I2	473(115000)	596(6500)	646(11500)	701(3000)
I3	472(182500)	590(10500)	640(21500)	698(5500)
I4	471(136000)	592(7500)	642(14500)	703(3000)
I5	473(165500)	598(17500)	649(23500)	703(5500)
I6	473(178000)	592(9500)	639(18500)	701(3500)
I7	466(126500)	595(5500)	646(17000)	699(4000)
I8	472(204500)	592(10500)	644(19500)	700(4000)
I9	472(225000)	595(10000)	646(22000)	701(4500)
I10	471(131000)	596(12500)	644(20000)	698(8000)
I11	473(271000)	590(17500)	642(32000)	702(9500)
I12	471(133500)	596(8500)	643(17500)	702(9500)
I13	472(216500)	593(13000)	642(23000)	700(6000)
I14	472(260000)	592(16000)	643(27000)	700(8000)
I15	472(313500)	593(16500)	643(30500)	700(7500)

ROS plays a significant role in damaging cancer cells during PDT treatment due to its strong oxidative activity. Photodecomposition of 1,3-diphenylisobenzofuran (DPBF) without PS was negligible, which was reported by the publication.²² The ROS generation of m-Ph₄TBPs I_1–15 was performed at room temperature using DPBF as a capture reagent. When a 650 nm laser light (5 mW/cm²) was used to irradiate the mixture of m-Ph₄TBPs and DPBF solution, a progressive decrease in the absorption intensity of DPBF at 417 nm was presented in Figure S1, and no decline in the absorption intensities of m-Ph₄TBPs at ∼470 nm was observed. All new m-Ph₄TBPs were able to produce ROS in Figure 3, and the singlet oxygen generation rate and quantum yield of I_1–15 were summarized in Table 2 with ZnPC as a standard (Φ_Δ = 0.56).²⁸

Figure 3.

(A–C) Linear plots fitted for the reactive oxygen species generation rate of I_1–15.

Table 2. Reactive Species Generation Rate and Quantum Yield of I_1–15.

Compound	K [s–1]	ΦΔ	Compound	K [s–1]	ΦΔ
I1	0.0010	0.05	I10	0.0106	0.32
I2	0.0058	0.20	I11	0.0084	0.37
I3	0.0049	0.13	I12	0.0075	0.18
I4	0.0083	0.25	I13	0.0127	0.22
I5	0.0023	0.07	I14	0.0079	0.18
I6	0.0057	0.23	I15	0.0102	0.21
I7	0.0031	0.07
I8	0.0078	0.19
I9	0.0049	0.17

Before cytotoxicity assays, the cellular uptake experiment was performed to determine the incubation time for the cells after the treatment of m-Ph₄TBP. Cellular uptake of respective m-Ph₄TBPs by Eca-109 cells was displayed in Figure S2A. The concentrations of most m-Ph₄TBPs increased dramatically within the first 4 h and reached a plateau within the following 8 h. We therefore considered that the incubation time of the m-Ph₄TBPs for MTT assays was 12 h.

The photoactive effects of compounds I_1–15 were examined using Eca-109 cells in vitro by an MTT assay (Figure 4). At a dosage of 4 μM, none of the compounds showed dark toxicity in the cells. All of the novel m-Ph₄TBPs were able to inhibit Eca-109 cells from growing after 8 J/cm² 650 nm laser irradiation (Figure 4A,B and Table 3). Among them, m-Ph₄TBPs I₄, I₆, I₉, I₁₂, and I₁₃ showed significant phototoxicity against Eca-109 cells. Compound I₁₂ displayed significant photodynamic activity by inhibiting Eca-109 cell proliferation with cell viability of ∼15% at a concentration of 1 μM. Compared to I₉, the ortho-fluorine-substituted glycol m-Ph₄TBP (I₁₂) exhibited stronger phototoxicity than I₉, which demonstrated that introducing ortho-fluorine at the meso-phenyl moiety is advantageous for enhancing the PDT efficiency. As the IC₅₀ values of I₄, I₆, I₉, I₁₂, and I₁₃ were higher than that of other compounds, those five m-Ph₄TBPs were selected for the following analysis.

Figure 4.

(A) Cell viability of I_1–15 upon irradiation with 650 nm laser light at the a dose of 8 J/cm² under a concentration ranging from 0 to 4 μM; (B) The dark-toxicity of representative m-Ph₄TBPs I₄, I₆, I₉, I₁₂, and I₁₃ at the concentrations ranging from 0 to 10 μM; (C) Flow cytometric assay of m-Ph₄TBPs I₄, I₆, I₉, I₁₂, and I₁₃ at 4 μM exposed to 8 J/cm² of 650 nm laser light. Q1: Annexin V (−) PI (+), necrotic cells; Q2: Annexin V (+) PI (+), late apoptotic cells; Q3: Annexin V (+) PI (−), early apoptotic cells; Q4: Annexin V (−) PI (−), lived cells; (D) The statistical results for the percentage of late apoptotic cells; (E) The statistical results for the percentage of necrotic cells.

Table 3. IC₅₀ Value of Compounds I_1-15 against Eca-109 Cells.

	IC50 (μM)				IC50 (μM)
Compound	0 J/cm2	8 J/cm2	PIa	Compound	0 J/cm2	8 J/cm2	PIa
I1	7.25	1.13	6.41	I10	9.63	1.48	>6.5
I2	9.55	>4	N/A	I11	>10	2.04	>4.90
I3	7.68	1.33	5.77	I12	8.08	0.39	20.72
I4	4.73	0.95	4.97	I13	6.89	0.48	14.35
I5	6.32	3.79	1.66	I14	9.17	1.32	6.94
I6	7.31	0.76	9.61	I15	>10	1.81	>5.52
I7	7.07	1.21	5.84	Verteporfin	3.27	0.41	7.97
I8	8.73	1.07	8.15	HMME	28.39	12.92	2.92
I9	8.43	0.81	10.41

^aPhototoxic index (PI): the ratio of dark to light toxicity.²⁹

To explore the mechanism of cell death induced by PDT, flow cytometry with Annexin VFITC/PI staining was performed. As shown in Figure 4C, m-Ph₄TBPs I₄, I₆, I₉, I₁₂, and I₁₃ could induce late apoptosis of cancer cells, and the late apoptosis rates were 15.8%, 30.4%, 59.8%, 82.4%, and 73.4%, respectively (Figure 4D). Furthermore, Eca-109 cells treated with I₄–PDT displayed a high necrosis percentage (42.1% in Figure 4E). The delightful result was found in m-Ph₄TBP I₁₂, which showed a good photodynamic effect in vitro with a high apoptosis rate of 83.22% and a lower necrosis rate of 13.1%.

For the detection of intracellular ROS, the fluorescent dye DCFH-DA was selected as a marker of ROS generation since this marker could be hydrolyzed into DCFH without fluorescence after entering into the cell and then oxidized by intracellular ROS to form DCF with high fluorescent signals. Eca-109 cells were coincubated with m-Ph₄TBPs (I₄, I₆, I₉, I₁₂, and I₁₃) at 4 μM for 24 h, and then added the marker, and coincubated for another 20 min, and exposed to a 650 nm laser at a dose of 8 J/cm². After being exposed to the 650 nm laser, all Eca-109 cells containing I₄, I₆, I₉, I₁₂, and I₁₃ showed green fluorescence (Figure 5), indicating that m-Ph₄TBPs could produce ROS in cells upon light irradiation.

Figure 5.

Intracellular ROS of I₄, I₆, I₉, I₁₂, and I₁₃ were detected at the dose of 8 J/cm² using 650 nm laser irradiation on Eca-109 cells recorded by fluorescence microscopy. Scale bar is 70 μm.

Using a Boyden chamber assay, the effect of compounds I₄, I₆, I₉, I₁₂, and I₁₃ on the invasiveness of Eca-109 cells was investigated.³⁰ As shown in Figure 6, treatment with m-Ph₄TBPs I₄, I₆, I₉, I₁₂, and I₁₃ at 4 μM for 24 h displayed significant anti-invasion ability by inhibiting Eca-109 cell invasion with invasion rates of ∼35%, ∼27%, ∼45%, ∼16%, and ∼17%, respectively. Compounds I₆ and I₁₂ with methoxy groups have strong anti-invasion abilities. In addition, varying the positions of substituents like I₉ and I₁₃ could change the anti-invasion ability of m-Ph₄TBPs.

Figure 6.

(A) Cell invasion assays were performed in Eca-109 cells after the treatments with m-Ph₄TBPs I₄, I₆, I₉, I₁₂, and I₁₃ for 24 h. Scale bar was 70 μm; (B) Cell invasion.

Compounds I₄, I₆, I₁₂, and I₁₃ were tested to evaluate their photodynamic efficiency in vivo. Balb/C nude mice carrying Eca-109 cells were divided into seven groups. The corresponding compound solution was injected into mice at a dose of 2 mg/kg, while the control and light groups obtained comparable amounts of saline. After injection, a 650 nm laser light irradiation with a dose of 120 J/cm² was carried out. The tumors in the control and light groups were larger compared to those in the PDT groups in Figure 7A,B. Notably, the in vitro assay results are consistent with those of tumor growth inhibition generated by PDT-treated. Remarkably, I₁₂–PDT displayed significant photodynamic effects among all of the PDT-treated groups.

Figure 7.

(A) Tumor images at 14 days post PDT; (B) Tumor weight at 14 days post PDT; *P < 0.05, **P < 0.01, ***P < 0.001; (C) H&E staining images of tumors in different groups at 14 days post PDT; Scale bar is 100 μm; (D) Apoptotic analysis of TUNEL staining images of tumor slices from mice after 14 d post-PDT; Scale bar was 50 μm.

Significant tumor tissue damages were observed in the I₄–PDT, I₆–PDT, I₁₂–PDT, I₁₃–PDT, and Verteporfin-PDT groups (Figure 7C), but the control and light groups showed no apparent destruction. Furthermore, TUNEL was used to investigate the in vivo anticancer mechanism of the m-Ph₄TBPs. The I₁₂–PDT group displayed a remarkably larger percentage of TUNEL-positive apoptotic cells, which were presented with green fluorescence, than the other PDT-treated groups in Figure 7D.

After 14 d post-PDT, in order to investigate the biosafety of m-Ph₄TBPs, major organ tissues were taken for histopathological examination (H&E) (Figure 8). According to Figure 8, some calcifications were formed in liver tissues 14 days post-PDT of m-Ph₄TBPs I₄ and I₆ while no calcifications were seen in the liver organs of other PDT groups. This suggests that phenolic m-Ph₄TBPs are harmful to the livers of mice. The H&E tissue staining results showed that the mice in the PDT-treated groups presented no alterations in any of the other major organs, including the heart, lung, kidney, or spleen.

Figure 8.

H&E staining images of other major organs (heart, lung, kidney, spleen, or liver) after 14 d post-PDT; Scale bar was 100 μm.

To sum up, series of m-Ph₄TBPs were designed, prepared, and tested. They all showed one absorption peak at about 470 nm and three absorption peaks at approximately 590, 645, and 700 nm. In addition, these novel compounds could be able to generate ROS and possess photodynamic activity toward Eca-109 cancer cells in vitro. Among these m-Ph₄TBPs, I₄, I₆, I₁₂, and I₁₃ showed more efficient phototoxicity in vitro and in vivo. Nonetheless, obvious calcifications were formed in liver tissues in I₄–PDT and I₆–PDT groups, whereas no calcification was seen in the liver organs of other PDT groups. This implies that compared to phenolic m-Ph₄TBPs, glycol m-Ph₄TBPs exhibited superior biosafety. Through comprehensive evaluations, I₁₂ has the best photodynamic antitumor efficacy and good biosafety in vivo, which can be regarded as a potential antitumor drug candidate for PDT.

Glossary

Abbreviations

DMSO

Dimethyl sulfoxide

DMF

Dimethylformamide

DCM

Dichloromethane

DDQ

2,3-Dichloro-5,6-dicyano-1,4-benzoquinone

DPBF

1,3-Diphenylisobenzofuran

MTT

3-(4,5-Dimethyl-2-thiazolyl)-2,5-diphenyl-2Hterazolium bromide

H&E

Hematoxylin and eosin

PDT

Photodynamic therapy

PSs

Photosensitizers

ROS

Reactive oxygen species

THF

Tetrahydrofuran

TFA

Trifluoroacetic acid

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.4c00202.

• Details of materials and methods was used in this study, including chemical synthesis and assay methods; ¹H and ¹³CNMR spectroscopic data for m-Ph₄TBPs I₁–I₁₅ along with procedures (PDF)

Author Contributions

The manuscript was written through contributions of all authors. All authors have read and agreed to the final version of the manuscript.

The authors declare no competing financial interest.