Photodynamic antimicrobial chemotherapy activity studies of 3-pyridyl substituted tetraarylchlorins and tetraarylbacteriochlorins

Abstract

The photodynamic antimicrobial chemotherapy activity (PACT) properties of 3-pyridyl substituted tetraarylchlorin (Chl1) and tetraarylbacteriochlorin (BChl1) have been assessed against Gram-(+) S. aureus and Gram-(−) E. coli strains. The pyridyl nitrogens were quaternized with methyl iodide to form water-soluble tetracationic species (Chl1-Q and BChl1-Q) that are more suitable for use against Gram-(−) bacteria. Metal insertion reactions with a main-group heavy-atom ion, Sn(IV), were attempted to assess whether the singlet oxygen quantum yield and the PACT activity properties could be enhanced in this manner. This proved successful with the chlorin ligand (Sn^IVChl1), but purification and characterization was problematic in the context of the bacteriochlorin. Upon photoirradiation with Thorlabs LEDs, Chl1 and Chl1-Q exhibited high Log₁₀ reduction values of 8.98 against S. aureus, while Sn^IVChl1 exhibited Log₁₀ reduction of 8.96 and 8.98 against both S. aureus and E. coli. This is noteworthy since neutral photosensitizers are normally ineffective against Gram-(−) strains. The cationic BChl1-Q species exhibits favorable photocytotoxicity towards S. aureus with a Log₁₀ reduction value of 6.61.

Introduction

Photodynamic antimicrobial chemotherapy (PACT) has emerged as an alternative approach for treating bacteria that have developed resistance to antibiotics.^1–3 A photosensitizer (PS) dye is photoexcited with incident light of an appropriate wavelength to produce cytotoxic reactive oxygen species (ROS), which induce cell death. After photoexcitation and intersystem crossing to the triplet manifold, the T₁ state of a photosensitizer dye reacts with oxygen molecules to form ¹O₂, via the Type II energy transfer reaction, or to form other ROS via the Type I electron transfer mechanism.^4,5 In this study, tetra-3-pyridylchlorin (Chl1) and tetra-3-pyridylbacteriochlorin (BChl1) were chosen as the main target dyes (Figure 1), since the parent tetra-3-pyridylporphyrin ligand absorbs relatively weakly in the 500−650 nm region. In recent years, we have reported the PACT activity properties of a series of different tetraarylchlorins,^6–11 since the lower symmetry of the chlorin (Chl) ligand results in an intensification and red shift of the Q band into the red region of the visible, making the dyes more suitable for use as PS dyes. Bacteriochlorins, with two oppositely aligned reduced pyrrole bonds, absorb light in the near-infrared region (NIR) at ca. 750 nm well within the therapeutic window. ¹² This enhances tissue penetration when excited by an appropriate light source,^13,14 making the dyes potentially suitable for treating infections of deep-seated tissue wounds. Naturally formed bacteriochlorins (BChls) have recently been reported to exhibit promising properties in this regard. ¹⁵ NIR photons can provide direct access to biological targets because they are the most invasive and least harmful to human tissues. The phototherapeutic window lies between 620 and 850 nm, where light can penetrate tissue to depths of 1−3 cm.^14,16 Bacteriochlorins have ideal properties for use as PS dyes, including ease of synthesis, solubility in a variety of solvents, high singlet oxygen quantum yields, long triplet state lifetimes and most importantly, nontoxic and rapid clearance out of the body after therapy, unlike most Food & Drug Administration (FDA) approved PSs.^17–19

Figure 1.

Structures of Chl1, Chl1-Q, Sn^IVChl and BChl1.

A key advantage of introducing pyridyl rings is that the nitrogen atom can be quaternized to form a water-soluble tetracationic species (Chl1-Q and BChl1-Q) that are more suitable for PACT applications against Gram-(−) bacteria, since they can readily penetrate their lipopolysaccharide outer membrane. ²⁰ Metal insertion reactions were also carried out with the Sn(IV) ion to form trans-axially ligated complexes. This reaction proceeded successfully with the chlorin ligand to form Sn^IVChl1, but not with the bacteriochlorin. The in vitro PACT activities of these dyes and complexes were compared against S. aureus and E. coli, typical Gram-(+) and -(−) bacterial strains, respectively.

Experimental section

Materials

All chemicals used in this study were purchased from commercial sources and were used as received unless otherwise stated. Pyridine, p-toluenesulfonyl hydrazide (p-TsNHNH₂), 3-pyridinecarboxaldehyde, 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ), tin(II) chloride dihydrate, pyrrole, propanoic acid, 9,10-dimethylanthracene (DMA), and 5,10,15,20-tetraphenylporphyrinato zinc(II) (ZnTPP) were purchased from Sigma Aldrich. Reagent grade chloroform, dichloromethane (DCM), dimethylformamide (DMF), methanol (MeOH), spectroscopic grade dimethyl sulfoxide (DMSO), diethyl ether, iodomethane, deuterated dimethyl sulfoxide (DMSO-d₆), hexane, and deuterated chloroform (CDCl₃) were obtained from Merck. Aluminum oxide neutral was purchased for column chromatography from Merck.

Phosphate-buffered saline (PBS) solution pH 7.4 was prepared using appropriate amounts of NaCl, KCl, Na₂HPO₄, and KH₂PO₄ in ultra-pure water obtained from an Elga Purelab Chorus 2 (RO/DI) system. Agar bacteriological BBL Mueller Hinton broth (HG000C24.500) and nutrient agar (HG0000C1.500) were purchased from Merck. S. aureus and E. coli were purchased from Davies Diagnostics and Microbiologics, respectively. Ultra-pure water was always obtained from an Elga Purelab Chorus 2 (RO/DI) system in all contexts.

Synthesis

Chl1 was synthesized according to literature procedures (Figure 2). ²¹ A mixture of propanoic acid (150 ml), 3-pyridinecarboxaldehyde (500 mg, 4.6 mmol), and pyrrole (313 mg, 4.6 mmol) in a round-bottom flask equipped with a condenser was refluxed for 4 h. The reaction mixture was cooled in ice and neutralized with NaOH to yield a crude precipitate, which was filtered, washed with ultrapure water (3 × 100 mL), and dried in vacuo. The crude product was purified on an aluminum oxide neutral column using chloroform as the eluent, yielding the porphyrin as a purple solid. The free base porphyrin was used in the next step without further purification or analysis. In a 100 mL round-bottom flask, porphyrin (150 mg, 1 mmol), K₂CO₃ (335 mg, 10.0 mmol), and p-TsNHNH₂ (180 mg, 4 mmol) were dissolved in dry pyridine (50 mL) and refluxed for 6 h. The same amount of p-TsNHNH₂ was added every 2 h. The reaction was stopped when a peak at 742 nm began to appear, corresponding to a doubly reduced bacteriochlorin product. After completion of the reaction, the reaction mixture was cooled and extracted with chloroform. Trace amounts of DDQ were added to the chloroform layer until the band at 742 nm disappeared. The crude product was added on a neutral aluminum oxide column with chloroform as the eluent to obtain the pure chlorin.

Figure 2.

Synthesis of Chl1, Chl1-Q, Sn^IVChl, BChl1 and BChl1-Q from the parent porphyrin ligand (Por1). The reduced peripheral pyrrole bonds of the chlorin and bacteriochlorin ligands are highlighted in red. Failed syntheses to form Sn^IVChl and Sn^IVBChl are also described.

Chl1 (Yield: 82%). ¹H NMR (400 MHz, DMSO-d₆) δ 9.73 (d, J = 10.0 Hz, 4H), 9.47–9.34 (m, 4H), 9.21 (d, J = 2.2 Hz, 2H), 9.09–8.95 (m, 4H), 8.65–8.41 (m, 8H), 4.30 (s, 4H), −1.58 (s, 2H). MS (MALDI-TOF): m/z for [M]⁺ = 620.18 (calc. 620.58); UV-Vis (DMSO) λ, nm (log ɛ): 653 (3.51) nm.

Quaternized chlorin Chl1-Q was synthesized as follows (Figure 2): In a 25 mL round-bottom flask containing iodomethane and DMF (1:2), chlorin (20 mg, 0.003 mmol) was added and stirred under warm temperature (50°C) for 24 h. The mixture was poured into 50 ml diethyl ether, centrifuged and dried in vacuo.

Chl1-Q (Yield: 84%). ¹H NMR (80 MHz, DMSO-d₆) δ 9.78 (s, 2H), 9.35 (d, J = 6.2 Hz, 4H), 9.04 (s, 6H), 8.55–8.25 (m, 4H), 7.72 (s, 6H), 4.46 (s, 4H), 2.66 (s, 12H), −3.31 (s, 2H). MS (MALDI-TOF): m/z for [M]⁺ = 680.86 (calc. 680.34). UV-Vis (DMSO) λ, nm (log ɛ): 658 (3.33) nm.

Sn^IVChl was prepared through a metal insertion reaction of Chl1 (Figure 2). Chl1 (80 mg, 0.1 mmol) and SnCl₂ (80 mg, 0.3 mmol) were dissolved in 1:1 CHCl₃/MeOH and refluxed for 4 h. The mixture was cooled, evaporated to dryness, and dissolved in CHCl₃. The CHCl₃ layer was washed with 50 mL ultra-pure water and dried. The resulting solid was recrystallized from CHCl₃/hexane to provide Sn^IVChl as a green solid.

Sn^IVChl (Yield: 79%). ¹H NMR (80 MHz, CDCl₃) δ 9.04 (d, J = 5.2 Hz, 8H), 8.86 (s, 6H), 8.17 (d, J = 5.4 Hz, 8H), 4.59 (s, 4H). MS (MALDI-TOF): m/z for [M−Cl + H]⁺ 772.23 (calc. 806.29). UV-Vis (DMSO) λ, nm (log ɛ): 628 (4.85) nm.

Attempts were made to quaternize this complex to form a quaternized species of Sn^IVChl1 (Sn^IVChl-Q) in a similar manner to Chl1-Q (Figure 2), but these failed to produce a product that could be characterized satisfactorily by UV-visible absorption spectroscopy, ¹H NMR spectroscopy and MALDI-TOF MS despite repeated attempts.

BChl1 was synthesized similarly to the corresponding chlorins, as shown in Figure 2. In a 100 mL round-bottom flask, porphyrin (235 mg, 1 mmol), K₂CO₃ (524 mg, 10 mmol), and p-TsNHNH₂ (280 mg, 4 mmol) were dissolved in dry pyridine (50 mL) and refluxed for 12 h. The same amount of p-TsNHNH₂ was added every 3 h. The reaction was stopped when the chlorin peak at 652 nm disappeared. After reaction completion, the reaction mixture was cooled and washed with water and chloroform. The crude product was purified by silica gel chromatography using ethyl acetate/petroleum ether (10:1) to yield the target bacteriochlorin compound.

BChl1 (Yield: 70%) ¹H NMR (80 MHz, CDCl₃): δ 9.32 (s, 4H), 8.88 (d, J = 6.4 Hz, 4H), 8.58 (s, 8H), 8.05–7.82 (m, 4H), 4.00 (s, 8H), −3.83 (s, 2H). MS (MALDI-TOF): m/z for [M]⁺ 622.32 (calc. 622.26).

BChl1 was quaternized similarly to chlorin Chl1 (Figure 2). BChl1 (30 mg, 0.038 mmol), iodomethane, and DMF (1:2) were added to a 25 mL round-bottom flask and stirred at 50°C for 24 h. The mixture was poured into 30 ml diethyl ether, centrifuged and dried in vacuo to give the desired BChl1-Q product.

BChl1-Q (Yield: 72%). ¹H NMR (600 MHz, DMSO) δ 9.22 (d, J = 6.6 Hz, 8H), 8.43 (d, J = 6.7 Hz, 8H), 7.03 (d, J = 2.2 Hz, 4H), 4.47 (s, 8H), 2.89 (s, 12H), −1.53 (s, 2H). MS (MALDI-TOF): m/z for [M + DMF]⁴⁺ 188.23 (calc. 188.86).

The crude BChl1 (0.1 mmol) product and SnCl₂ (0.3 mmol) were dissolved in 50 mL DMF or MeOH/CHCl₃ and refluxed for 4 h, in an attempt to form its Sn(IV) complex (Sn^IVBChl1) in a similar manner to what is described above for Sn^IVChl1 (Figure 2). The mixture was cooled, evaporated to dryness, and dissolved in CHCl₃. The CHCl₃ layer was washed with 50 mL Millipore water and dried. The dried product was purified through alumina-neutral chromatography. UV-visible absorption spectra for the crude products were measured. The lower-energy Q band between 720−800 nm of the BChl starting material had disappeared. The BChl appeared to have oxidized to form a porphyrin, but this was not investigated in depth.

Instrumentation

Ground state UV-visible absorption spectra were collected on a Shimadzu UV-2550 spectrophotometer or an EvolutionTM 350-UV-vis spectrophotometer from Thermo Fischer ScientificTM. Mass spectra were obtained using a Bruker Auto-FLEX III Smartbeam MALDI-TOF mass spectrometer using α-cyano-4-hydroxycinnamic acid as a matrix in either negative or positive ion mode with an m/z range of 400−1500 amu. A Varian Eclipse spectrofluorimeter was used for fluorescence quantum yield studies. Excitation was carried out in the B-band region for both chlorins and bacteriochlorins. Bruker AMX 400 MHz and 80 MHz benchtop NMR spectrometers were used to obtain ¹H NMR data. The spectra were obtained at ambient temperature using deuterated solvents.

Singlet oxygen quantum yield studies were carried out using an Ekspla NT 342B-20-AW laser with an Nd:YAG that pumps a 420−2300 nm optical parametric oscillator (OPO) (355 nm, 2.0 mJ/7 ns, 20 Hz) to provide monochromatic light at a crossover wavelength for the standard and sample solutions to ensure that the optical density is the same for both solutions. A Thermo Scientific Evolution 350 spectrometer was used to monitor the degradation of the singlet oxygen quencher (DMA) at regular time intervals, enabling slope values for −ln(A/A₀) versus time to be determined. The triplet-state lifetimes spectra were recorded in DMSO at 470 nm using an Edinburgh Instruments LP980 spectrometer and an Ekspla NT-342B laser equipped with an OPO to provide an excitation wavelength of 430 nm (2.0 mJ excitation energy, 7 ns pulse duration, and a 20 Hz repetition rate). The solutions were degassed with nitrogen for 20 min prior to the measurements, and absorbance was maintained at ca. 1.5 for the B band. Exponential curve fitting of the decay curve using OriginPro 8 software provided the triplet lifetimes.

PACT activity studies

Gram-(+) S. aureus (ATCC^® 25923^TM) and Gram-(–) E. coli (ATCC^® 25922^TM) strains were used for the PACT studies. Both bacteria were grown on agar plates according to the manufacturer's guidelines to obtain the respective bacterial colonies. Single colonies were inoculated into freshly prepared Luria nutrient broth; the resulting cultures were swirled and placed in a shaking incubator (37°C, 200 rpm) for 18 h and 6 days for S. aureus and E. coli, respectively. An aliquot of the culture was transferred into freshly prepared broth (4 mL) and further incubated at 37°C. The optical density of the bacterial culture was taken regularly to ensure mid-logarithmic growth (OD 600 nm ∼ 0.6−0.7), and the broth culture was removed by centrifugation at 3000 rpm for 15 min. The independent bacterial pellets of S. aureus and E. coli were washed three times with PBS to remove residual nutrient broth. After resuspension in 4 mL PBS, they were diluted to 1:1000 (v/v) with PBS to afford the working stock solutions. A Ledetect 96 microplate reader was used to measure the optical density of bacteria in the 96-well plate used during the PACT activity studies.

The PACT studies were carried out using the previously reported viable count method. ²² The compounds studied are Chl1, Chl1-Q, Sn^IVChl, BChl1 and BChl-Q. Only Chl1-Q and BChl1-Q are soluble in water. Hence, the antimicrobial studies were performed with 1% DMSO in PBS. With regards to the experimental procedure, appropriate aliquots of stock solutions of the Chl and BChl dyes were mixed with 60 mL of the individual bacterial suspensions of S. aureus and E. coli in PBS to provide the concentrations of PS dye requiredfor concentration and time dependence studies. The mixtures were placed in a shaking incubator set at 37°C for 45 min. Half (1 mL) of the mixture was irradiated at 625 or 660 nm for chlorins and 730 nm for BChls in a 24-well plate at predefined time intervals (t_n, where n = 0, 15, 30, 45, and 60 min) with the appropriate Thorlabs LED, while the other 1 mL of the mixture was placed in 24 well plates and kept in the dark for the same time-intervals. The bioillumination chamber of a Modulight^® Medical Laser (ML) 7710–680 system, fitted with an appropriate Thorlabs LED, was used during PACT studies. Thorlabs M625L3 (1.4 mW.cm⁻²), M660L3 (1.7 mW.cm⁻²) and M730L3 (1.0 mW.cm⁻²). LEDs with maximum outputs at 625, 660 and 730 nm, respectively, were selected by matching the LED output profiles with the lowest-energy Q bands of the PS dyes. On this basis, the M625L3 LED was used for Sn^IVChl1, while the M660L3 LED was used for Chl1 and Chl1-Q, and the M730L3 LED was used for BChl1 and BChl1-Q. After treatment, 100 µL of the samples exposed to either light or dark conditions were inoculated onto agar plates in aseptic conditions. The plates were then placed in an incubator at 37°C for 18 h before colony count determination. The colony-forming units (CFU/mL) values for the bacteria were counted using a Scan^® 500 automatic colony counter.

Photostability experiments

Photostability studies were performed in quartz cuvettes (1 × 1 cm). The solutions were irradiated for 30 min at regular time intervals (6−30 min) using a Thorlabs M730L3 (1.0 mW.cm⁻²) LED for BChl1 and BChl1-Q, a Thorlabs M625L3 LED (1.4 mW.cm⁻²) for Sn^IVChl1, and a Thorlabs M660L3 (1.7 mW.cm⁻²) LED for Chl1 and Chl1-Q in the same arrangement used for the PACT activity studies. The solutions were prepared in the dark by dissolving 1 mg (6 mL) of the sample in 1% DMSO/PBS and transferred into a quartz cuvette. The photobleaching of the dyes at the B band maxima was monitored as a function of time.

Results and discussion

Synthesis

Chl1, Chl1-Q, Sn^IVChl, BChl1 and BChl1-Q were prepared and purified according to previously reported literature procedures.^{6,7,23 1}H NMR spectroscopy and MALDI-TOF MS measurements were conducted to confirm their successful synthesis and ensure their purity.

UV-visible absorption spectra

The reduction of a peripheral pyrrole bond of the porphyrin ligand to form a chlorin results in significant changes in the UV-visible absorption spectra, since the lower symmetry lifts the degeneracy of the e_g* LUMO, resulting in a mixing of the allowed and forbidden properties of the B and Q bands. ⁹ For this reason, free base Chls can be readily identified spectroscopically by their intense Q band between 650−670 nm in the red region of the visible.^6–11 They also have an intense B band in the blue region of the spectrum (ca. 420 nm). This holds for Chl1, as shown in Figure 3 and Table 1. Quaternization to form Chl1-Q results in a red shift of the Q band by 5 nm. Introducing a heavy metal ion into the core of Chl1 results in a blue shift of the intense lower energy Q band of Sn^IVChl1 to 628 nm.

Figure 3.

(a) normalized UV-visible absorption spectra of Chl1, Chl1-Q and Sn^IVChl1. (b) Fluorescence emission spectra for Chl1, Chl1-Q, and Sn^IVChl1. (c) Normalized UV-visible absorption spectra of BChl1 and BChl1-Q. Red asterisks highlight bands associated with the presence of Chl1-Q.

Table 1.

Absorption and emission data for Chl1, Chl1-Q, Sn^IVChl1, BChl1 and BChl1-Q in DMSO.

	λabs (nm)	λem (nm)	SS (cm−1)
Chl1	417, 513, 601, 653	657	90
Chl1-Q	423, 512, 602, 658	664	140
SnIVChl1	427, 599, 628	636	200
BChl1	355, 378, 521, 747	–	–
BChl1-Q	355, 381, 517, 762	–	–

Bacteriochlorins,^12,23–25 which have two oppositely arranged reduced peripheral pyrrole bonds, have an even more intense Q band maximum in the red region of the spectrum in the 730−770 nm region. The Q band for BChl1 lies at 747 nm. BChl1 exhibits a symmetry-split B band in the blue region of the spectrum, between 340 and 390 nm. For quaternized BChl1-Q, the Q band maximum is more red-shifted and lies at 762 nm (Figure 3). The longest-wavelength absorption band observed for these BChls enhance the suitability of the PS for PACT of deep-seated tissue wounds, since near-infrared light can provide sufficient tissue penetration in this spectral region. It proved difficult to obtain spectra of BChl1 and BChl1-Q in DMSO solution that did not contain weak bands relating to a minor chlorin impurity (Figure 3). Similar issues have been faced in previous studies on BChls.^23,25 PACT experiments were carried out in this context to determine, in proof-of-principle terms, whether further studies are merited in this context.

Fluorescence spectroscopy

The fluorescence excitation and emission spectra of the chlorins are shown in Figure 3 and Table 1. No emission was observed for the BChls. The emission spectra of the Chls exhibit a single band maximum when excited at the Soret band. The fluorescence emission maximum for Chl1 lies at 657 nm, while that of Chl1-Q lies at 664 nm. The fluorescence emission band maximum of Sn^IVChl1 lies at 636 nm. Due to the heavy atom effect induced by the central metal ion Sn(IV) introduced into the central cavity of Sn^IVChl1, a significant fluorescence quenching is observed. This is attributed to the heavy atom effect imparted by the Sn(IV) ion.^26–29 The chlorins have relatively small Stokes shift (SS) values of 90−200 cm⁻¹, due to their relatively rigid structures. The fluorescence quantum yields of Chl1, Chl1-Q, and Sn^IVChl1 are presented in Table 2. All fluorescence quantum yield measurements were conducted in DMSO using the comparative method. ZnTPP (Φ_F = 0.039 ³⁰ ) was used as the standard. As expected, the Φ_F value of Sn^IVChl1 is significantly lower than those of Chl1 and Sn^IVChl1. This is due to the central metal ion that enhances the rate of ISC.

Table 2.

Photophysicochemical properties of Chl1, Chl1-Q, Sn^IVChl1, BChl1 and BChl1-Q in DMSO, and photobleaching (Photo) percentages after 30 min photoirradiation in 1% DMSO in PBS.

	ΦF	ΦΔ	τT (μs)	Photo (%)
Chl1	0.10	0.24	192	3
Chl1-Q	0.13	0.48	226	5
SnIVChl1	< 0.01	0.60	47	2
BChl1	–	0.61	85	4
BChl1-Q	–	0.83	14	5

Singlet oxygen quantum yields

The effectiveness of a PS drug for use in PACT depends on its ability to generate sufficient singlet oxygen. Hence, oxygen photosensitization studies were conducted, since this cytotoxic agent triggers biological pathways that inactivate bacteria.^31–33 Singlet oxygen quantum yield studies were conducted in DMSO using DMA as a singlet oxygen scavenger for both Chls and BChls. A comparative method against a known standard was used. ZnTPP (Φ_Δ = 0.53 ³⁴ ) was used as the standard. The degradation of the singlet oxygen scavenger band was observed at 10 s intervals for 60 s, as shown in Figure 4, using Chl1-Q as a typical example. After the −ln(A/A₀) data were plotted against time, the slope for each compound was calculated by linear regression. Table 2 lists the Φ_Δ values.

Figure 4.

Degradation of DMA in the presence of (a) Chl1-Q and (b) BChl1-Q at regular time intervals. The insets provide plots of −ln(A/A₀) values for DMA vs photoirradiation time in the presence of Chl1, Chl1-Q, Sn^IVChl1, BChl1 and BChl1-Q.

As expected, Sn(IV) complex Sn^IVChl1 exhibited higher singlet oxygen quantum yields than free base chlorins Chl1 and Chl1-Q. This is due to the heavy atom effect, which enhances the rate of ISC. It is well-known that only monomeric forms of photosensitizers tend to exhibit high singlet oxygen quantum yields, so the trans-axial ligation of Sn^IVChl1 may also be a factor. BChl1-Q has a higher Φ_Δ value than BChl1 (Table 2). Among the chlorins studied, the observed trend is Sn^IVChl1 > Chl1-Q > Chl1, while BChl1 and BChl1-Q provide the order BChl1-Q > BChl1. The absence of spectral changes in the 650–700 region nm for chlorins and 700–720 nm for BChls during the singlet-oxygen quantum yield measurements provided further evidence that these compounds exhibit favorable photostability.

Photostability

A photosensitizer dye with favorable properties should also exhibit high photostability, with little or no photobleaching, since this reduces photosensitizer concentration and can generate problematic photodegradation products. When singlet oxygen is formed, photosensitizer dyes are known to degrade upon photoexcitation, so it is crucial to assess the degree of photodegradation that occurs under the conditions used for the PACT activity experiments. The photostability properties were studied using a Thorlabs M730L3 (1.0 mW.cm⁻²) LED for BChl1 and BChl1-Q, a Thorlabs M625L3 LED (1.4 mW.cm⁻²) for Sn^IVChl1, and a Thorlabs M660L3 (1.7 mW.cm⁻²) LED for Chl1 and Chl1-Q.

Photoirradiation was conducted for 30 min in PBS under the same conditions used for the PACT activity studies. The dyes were solubilized with 1% DMSO, since Chl1, Sn^IVChl1, and BChl1 do not fully dissolve in PBS alone. Although Chls and BChls are known to be relatively unstable, ³⁵ only 3% photobleaching of Chl1 was observed during exposure to a 660 nm LED. Chl1-Q exhibited 5% photobleaching (Figure 5 and Table 2), while Sn^IVChl1 underwent 2% photobleaching during exposure to the 625 nm LED. As shown in Table 2, BChl1 and BChl1-Q also exhibited high photostability with photobleaching of 4 and 5%, respectively, during exposure to the 730 nm LED.

Figure 5.

Percentage photostability of Chl1, Chl1-Q, Sn^IVChl1, BChl1 and BChl1-Q in 1% DMSO in PBS under the photoirradiation conditions used for the PACT studies.

Triplet lifetimes

Flash photolysis experiments were carried out for Chl1, Chl1-Q, Sn^IVChl1, BChl1 and BChl1-Q in nitrogen-purged DMSO. Figure 6 shows the triplet absorption decay curve of BChl1-Q, as a typical example. Population of the triplet state of a PS dye after photoexcitation is usually required to facilitate the formation of ROS via Type I and/or Type II mechanisms. Chl1 and Chl1-Q have relatively long-lived triplet lifetimes of 192 and 226 µs, respectively. In contrast, Sn^IVChl1 has a shorter triplet lifetime value of 47 µs. In the context of BChls, the parent free base BChl1 had the longer triplet lifetime value of 85 μs, while the value for BChl1-Q is 14 µs. τ_T values on the μs timescale in Table 2 indicate that the Chl and BChl PS dyes studied have relatively long-lived triplet states, further enhancing their potential utility as PS dyes in PACT.

Figure 6.

Triplet state absorption decay curve of BChl1-Q in N₂ purged DMSO.

PACT activity studies

Optimization studies of the photosensitizer concentration were conducted against S. aureus and E. coli over 60 min irradiation times (Figures 7 and 8). A control experiment was also performed to demonstrate that the presence of 1% DMSO does not affect the bacterial cells. The concentration studies were performed for Chl1, BChl1 and the quaternized Chl1-Q and BChl1-Q species. A concentration of 8.0 µM was selected for the Chls for PACT activity studies against both S. aureus and E. coli, whereas a concentration of 16 µM was used for the BChls. PACT activities for both S. aureus and E. coli were evaluated for 60 min at 15 min intervals. Plate count analyses for the Chls and BChls against both S. aureus and E. coli are shown in Figure 7. As anticipated, the chlorin PS dyes exhibited greatly enhanced toxicity towards S. aureus when irradiated with LED light, resulting in 0% bacterial survival after 60 and 45 min for Chl1 and Chl1-Q (Figure 8 and Table 3), respectively. It is noteworthy, however, that significant dark toxicity effects were also observed for Sn^IVChl1. In contrast, Sn^IVChl1 exhibited lower dark toxicity against E. coli, with about 15% bacterial survival, since it has an extra cell membrane that is not readily penetrated by non-cationic PS dyes. ²⁰

Figure 7.

Images of the S. aureus and E. coli colonies for BChl1 (B1) and BChl1-Q (B3) after an irradiation time of 60 min using a Thorlabs M730L3 LED. Images of the control cells are also provided on the left.

Figure 8.

PACT activity studies for the Chl and BChl PS dyes against S. aureus and E. coli in both the absence (dark) and presence (light) of photoirradiation with Thorlabs M625L3 (1.4 mW.cm⁻²), M660L3 (1.7 mW.cm⁻²) and M730L3 (1.0 mW.cm⁻²) LEDs for 60 min in 15 min intervals.

Table 3.

Log₁₀ reduction values in the dark and during the PACT activity studies with Chl1, Chl1-Q, Sn^IVChl1, BChl1 and BChl1-Q against S. aureus and E. coli, respectively.

	Dark		Light
	S. aureus	E. coli	S. aureus	E. coli
Chl1	0.13	0.05	8.96@60 a	0.16
Chl1-Q	3.50	0.15	8.96@45	0.17
SnIVChl1	8.96@30	0.68	8.96@30	8.98@30
BChl1	1.14	0.40	2.04	0.47
BChl1-Q	0.92	0.42	1.75	6.61@45

^a

Time in min at which complete inactivation was achieved is provided after the @ symbol.

The Log₁₀ reduction values can be used to quantify the results (Figure 8 and Table 3). According to Food and Drug Administration (FDA) standards, a PS dye should have a Log₁₀ reduction > 3 to be considered an antibacterial agent.^36–38 Chl1 and Chl1-Q exhibited Log₁₀ reduction > 3 against both S. aureus. These studies demonstrate that these Chls are more effective towards S. aureus than E. coli. The best photoinactivation results were observed for Chl1, Chl1-Q and Sn^IVChl1, which exhibited Log₁₀ reductions of 8.96 against S. aureus, while Sn^IVChl1 also had a Log₁₀ reduction of 8.98 against E. coli. Although E. coli is usually a more difficult bacterial strain to treat due to its outer cell wall, the introduction of the central metal ion greatly enhanced the PACT activity compared to Chl1 and Chl1-Q. It is unusual for neutral PS dyes to exhibit such high Log₁₀ reduction values against Gram-(−) bacteria. Similar PACT activity data were reported recently by Mack and co-workers for other Sn(IV) tetraarylchlorin complexes with methylthiophenyl and thien-2-yl rings.^7,8 BChl1 was determined to have Log₁₀ reduction values < 0.47 against E. coli and < 2.04 against S. aureus, while BChl1-Q performed well against E. coli with a Log₁₀ reduction of 6.61. This further demonstrates the effectiveness of positively-charged PS dyes in this context. ³⁹

Conclusion

Tetra-3-pyridyl-substituted chlorin dyes Chl1, Chl1-Q, and Sn^IVChl1, and bacteriochlorins BChl1 and BChl1-Q were successfully synthesized and characterized. The structures of the dyes were confirmed by mass spectrometry, and UV-visible and ¹H-NMR spectroscopy. Chl1, Chl1-Q, and Sn^IVChl1 have Φ_Δ values of 0.24, 0.48, and 0.60, respectively, while those of BChl1 and BChl1-Q are 0.61 and 0.83. Since Chls and BChls are known to be relatively unstable, the photostability of the dyes studied was investigated under the same conditions used for the PACT activity studies. All the Chl and BChl dyes are highly photostable, with < 6% photobleaching after 30 min of photoirradiation under the same conditions used for the PACT activity studies. Chl1 and Chl1-Q have longer triplet-state lifetimes than Sn^IVChl1. Triplet-state lifetimes in the µs range enhance their suitability for use in singlet oxygen PS dye applications. PACT studies were performed for both the Chl and BChl dyes against Gram-(+) S. aureus and Gram-(−) E. coli strains. Upon photoirradiation with Thorlabs LEDs, Chl1 and Chl1-Q had Log₁₀ reduction values of 8.96 against S. aureus. Sn^IVChl1 exhibited a high Log₁₀ reduction of 8.96 in the dark, as well as during the PACT activity studies, with a Log₁₀ reduction of 8.98 against E. coli. This is noteworthy since neutral PS dyes are normally ineffective against Gram-(−) strains. Similar data have been reported previously for other Sn(IV) tetraarylchlorins.^7,8 The cationic BChl1-Q species exhibits favorable photocytotoxicity towards E. coli with a Log₁₀ reduction of 6.61. These results suggest that tetrapyridyl-substituted chlorins and bacteriochlorins merit further in-depth study in the context of PACT.