Nitroreductase-activatable photosensitizers for selective antimicrobial photodynamic therapy

Abstract

Antimicrobial photodynamic therapy (aPDT) utilizes light, oxygen and a photosensitizer (PS) to enact cell death via the production of reactive oxygen species (ROS). This mechanism of cell death, via oxidative stress, has allowed aPDT to be effective against antibiotic-resistant bacterial strains, with the development of resistance being minimal as no specific pathway is targeted. While promising, as ambient light can activate PSs, damage to mammalian tissues can occur, leading to drug-induced photosensitivity. To mitigate this, we developed a nitroreductase-activatable PS containing a quenching group that inhibits fluorescence and ROS. Upon reaction with nitroreductase, the quenching group can be liberated, restoring fluorescence and ROS production. As nitroreductase is not present in healthy mammalian tissues but expressed in many bacteria, photosensitivity of mammalian cells can be reduced. Herein, the synthesis and photophysical characterization of the nitroreductase-activatable PS, DB2, is described. DB2 was quenched compared to the free PS, DB1, and activation both in vitro by purified nitroreductase and in the gram-positive bacterial strain, Bacillus subtilis, was confirmed by fluorescence recovery. Cell viability studies in B. subtilis showed low dark toxicity and an IC₅₀ of 0.16 μM under 10-minute irradiation (530 nm, 42 mW cm⁻²). Minimal toxicity was observed under the same conditions in mammalian cell cultures demonstrating the potential of DB2 to mitigate photosensitivity and provide a promising approach for aPDT.

---

With the constant use of antibiotics to treat infections, antimicrobial resistance (AMR) continues to rise, and it is estimated that over 10 million deaths per year will be caused by AMR by 2050.¹ To combat this trend, alternative treatments such as antimicrobial photodynamic therapy (aPDT) are being developed. Antimicrobial PDT is a subset of PDT, a clinically approved cancer treatment requiring light, oxygen and a chromophore photosensitizer (PS).^2–4 Upon irradiation with light, a PS can absorb the incoming photon entering an excited singlet state. From here, the PS can undergo intersystem crossing into an excited triplet state where they interact with nearby biomolecules and oxygen to generate cytotoxic reactive oxygen species (ROS), such as singlet oxygen (¹O₂), leading to cell death via oxidative stress.^2–4 Additionally, some PSs can relax from the singlet excited state to the ground state via fluorescence, enabling detection. For aPDT, bacteria, fungi, viruses and protozoa are targeted instead of cancerous tissues and can provide certain advantages over traditional antibiotics as the generation of ROS invokes cell death by a different mechanism (i.e. oxidative stress).^5–7 Therefore, aPDT has been demonstrated to be effective against antibiotic-resistant bacterial strains and, as no specific enzyme or pathway is targeted, the development of resistance in bacteria is also slower when compared to the usage of antibiotics.^5–7 This has led to the development of many PSs for aPDT summarized in numerous review articles.^8–10 Although a promising strategy, aPDT can have adverse side effects as PSs can accumulate in healthy mammalian tissues. This may cause unwanted cell damage and death leading to photosensitivity in patients.^8–10

Cationic PSs have demonstrated higher efficacy towards bacteria,^10–12 and PS-conjugates to antibiotics,^13–15 nanoparticles,^16–18 and peptides^19,20 are being developed to promote permeability into bacteria. However, in both cases, the PS is always in an “ON” state, whereby ambient light can activate the PS, leading to photosensitivity. To mitigate this, compounds containing quenched PSs, known as activatable PSs (aPSs), are being pursued. By introducing an enzyme-activatable trigger group that can also quench the PS, aPSs are effectively in an “OFF” state producing low levels of fluorescence and ROS production when irradiated.^21,22 Upon reaction with the target enzyme, the trigger group can be cleaved liberating a free PS in its “ON” state and recovering its fluorescence and ROS production.^21,22 As a result, only cells (i.e. bacteria) that express the target enzyme can activate the aPS and generate sufficient ROS upon irradiation to enact cell death. Non-targeted cells (i.e. mammalian cells) will contain the inactive aPS producing less ROS and, therefore, minimize photosensitivity. As fluorescence is also quenched for aPSs, this method can also be used to diagnose the presence of the targeted cells, when activated, and to verify specificity.^21,22

To develop an aPS selectively for aPDT, nitroreductase (NTR), an enzyme capable of reducing nitro groups into amines using nicotinamide adenine dinucleotide (NADH) as a cofactor, was chosen for two reasons. First, NTR is expressed in many bacterial strains such as Escherichia coli, Staphylococcus aureus, and Bacillus subtilis, while not being expressed in healthy mammalian cells.²³ This has led to NTR prodrugs^24,25 and NTR-activated fluorophores^26,27 being used for the treatment and diagnostics of bacteria. Secondly, NTR can interact with the 4-nitrobenzyl trigger group, a known fluorescence and ROS quencher. By tethering 4-nitrobenzyl onto a PS, we reasoned that the NTR-aPSs would have low fluorescence and ROS production until reaction with NTR (Fig. 1), thereby providing higher selectivity when compared to using a PS alone. It should be noted that NTR can also be expressed in the hypoxic, low-oxygen environments commonly found within tumours, which has led to many NTR-aPSs being synthesized and tested for PDT applications.^28,29 However, to our knowledge, studies using aPSs for aPDT are limited and only one known exploiting NTR has been reported, which used a hemicyanine-based PS that had a low singlet oxygen quantum yield and targeted intracellular bacteria.²² Herein, the synthesis and photophysical characterization of DB2, a fluorescein-based NTR-aPS are described. Fluorescein-based PSs can have high singlet oxygen and fluorescence quantum yields. By appending the 4-nitrobenzyl group onto 4′,5′-dibromofluorescein (DBFCN), a known PS, the fluorescence and ROS production is quenched. After incubation with purified NTR and NADH, cleavage of the trigger group was observed by the release of the fluorescent product. Bacterial cell studies were then conducted in the gram-positive Bacillus subtilis model strain where DB2 demonstrated activation and managed to kill cells with an absolute IC₅₀ value of 0.16 μM with 10 minutes of irradiation (530 nm, 41.8 mW cm⁻²). With the broad number of bacterial strains that express NTR, and the rising health concerns caused by AMR bacteria, the development of novel NTR-aPSs such as DB2 can further build upon the foundation for aPDT.

Fig. 1 Mechanism of action for NTR-aPSs. Upon reaction with NTR, the nitro group is reduced to an amine, leading to cleavage and the release of the free PS.

Results and discussion

The synthesis of DB2 proceeded in two steps starting with commercially available DBFCN (Scheme 1). To direct the 4-nitrobenzyl group towards the hydroxyl on the xanthene core, the first step involved protecting the carboxylic acid with a methyl group forming a methyl ester, DB1. Nucleophilic substitution of DB1 to nitrobenzyl bromide yielded DB2. All ¹H and ¹³C NMR and MS data can be seen in the supplementary information.

Scheme 1 Synthesis of DB2. The carboxylic acid on DBFCN is converted to a methyl ester in the first step to direct nucleophilic substitution of the 4-nitrobenzyl group towards the hydroxyl group in the subsequent step.

To determine DB2's potential as an NTR-aPS, the photophysical properties were determined and compared to DB1, the expected product from the NTR reaction (Fig. 1). All photophysical characterizations were conducted using DMSO stocks of the respective compounds diluted in phosphate buffered saline (PBS) to a final concentration containing 2% DMSO unless otherwise specified. The absorbance of DB1 showed a slight redshift of 5 nm (λ_abs = 508 nm) compared to DBFCN. With the addition of the 4-nitrobenzyl group, the absorbance of DB2 showed peak broadening with a significant 38 nm blueshift (λ_abs = 470 nm) (Fig. 2a). To monitor fluorescence quenching, the absorbances of the compounds were matched at 0.1 AU at 503 nm and fluorescence spectra were obtained (Fig. 2b), showcasing that DB2 has lower fluorescence than DB1. To further confirm the quenching capabilities of the 4-nitrobenzl group, the fluorescence quantum yields (Φ_f) were then obtained in PBS. DB2 was ∼11-fold quenched compared to DB1 (0.024 ± 0.002 and 0.26 ± 0.01 for DB2 and DB1, respectively, using Rhodamine B (Φ_f = 0.5 in ethanol)³⁰ as a standard, confirming that the 4-nitrobenzyl group acts as a quencher. The singlet oxygen quantum yields (Φ_Δ) were then measured using 9,10-anthracenediyl-bis(methylene)dimalonic acid (ABDA), a selective ¹O₂ sensor that decreases in absorbance at 401 nm upon reaction with ¹O₂. Consistent with fluorescence quenching, DB2 was ∼2-fold quenched compared to DB1 (Φ_Δ = 0.38 ± 0.04 and 0.20 ± 0.02 for DB1 and DB2 respectively) (Fig. 2c) using Eosin Y (EY) as a control (Φ_Δ = 0.50 in H₂O)³¹ under irradiation at 490 nm (28.0 mW cm⁻²).

Fig. 2 Photophysical characterizations of DB2 in PBS (2% DMSO). (a) Normalized absorbance spectra of compounds. (b) Fluorescence spectra of compounds. (c) ¹O₂ generation of compounds monitored by the absorbance of ABDA at 401 nm under 490 nm irradiation (28.0 mW cm⁻²). EY was chosen as the standard and ABDA under irradiation serves as a control. Experiments were conducted in triplicate.

With the quenching of fluorescence and ROS generation observed in DB2 compared to DB1, the extinction coefficients (ε) were determined in DMSO by quantitative NMR (Fig. S1†) at their respective λ_abs in DMSO (Fig. S2†) and used to determine the concentrations for subsequent experiments. The difference in ε between the two compounds can further enhance the quenching observed as DB2 has a significantly lower ε (1,700 M⁻¹ cm⁻¹) compared to DB1 (123 000 M⁻¹ cm⁻¹) when compared at 531 nm in DMSO. All photophysical properties are summarized in Table 1.

Table 1 Summary of the photophysical data for DB1 and DB2

	DB1	DB2
a Extinction coefficients were determined at the respective maximum λabs in DMSO.
λ abs (nm, PBS)	507	470
λ em (nm, PBS)	534	537
λ abs (nm, DMSO)	531	464
ε (M^−1 cm^−1, DMSO)^a	123 000	39 000
Φ f (PBS)	0.26 ± 0.01	0.024 ± 0.002
Φ Δ (PBS)	0.38 ± 0.04	0.20 ± 0.02

---

To confirm that NTR can recognize DB2 and release DB1, an in vitro fluorescence time-course assay was conducted. Upon incubating purified NTR, NADH, and DB2, the fluorescence increased over the 10-minute incubation period in contrast to a sample containing only NADH and DB2 (Fig. 3a). The fluorescence spectra before and after the incubation were also obtained (Fig. 3b) and the λ_em matched that of DB1. To further confirm the release of DB1, the λ_abs of the product after incubation also matched DB1 with a similar peak shape showcasing its release (Fig. 3c). The increase in absorbance observed after activation also highlights the additional quenching caused by the change in ε between the two compounds. No changes in fluorescence and absorbance were observed upon incubating DB2 in the absence of NTR (Fig. S3†).

Fig. 3 Nitroreductase reaction with DB2 in PBS (2% DMSO). (a) Fluorescence time-course of DB2 with NTR and NADH. (b) Fluorescence spectra of DB2 before and after 10-minute incubation with NTR and NADH. (c) Absorbance spectra of DB2 before and after 10-minute incubation with NTR and NADH.

With this promising in vitro data, DB2 was incubated separately with two NTR-containing model bacteria, Bacillus subtilis (Gram-positive strain) and Escherichia coli (Gram-negative strain), to monitor the release of DB1via observing the fluorescence increase over time. Utilizing an initial optical density (OD₆₀₀) of 0.05 AU at 600 nm for both bacterial strains, 1 μM of DB2 was incubated independently with both strains and compared to controls with bacteria only and DB2 only for 24 hours. As seen in Fig. 4a, B. subtilis showed a fluorescence increase suggesting activation and release of the free PS, DB1. However, no substantial difference in fluorescence was observed with E. coli (Fig. S4†), which we attributed to the permeability barrier formed by the dual cell membrane of E. coli preventing entry of DB2.³² Therefore, B. subtilis was chosen for further experiments to determine if DB2 can enact bacterial cell death.

Fig. 4 Cell experiments with DB2. (a) Activation of DB2 in B. subtilis monitored by fluorescence. Experiments were conducted in triplicate. (b) Cell viability curve of DB2 with B. subtilis under dark or light conditions (10 minutes, 530 nm, 41.8 mW cm⁻²). Experiments were conducted in triplicate. (c) Cell viability of DB1 and DB2 in MRC-9 cells under dark or light conditions (10 minutes, 530 nm, 41.8 mW cm⁻²). Experiments were conducted in triplicate.

Utilizing an initial OD₆₀₀ of 0.05, a variety of concentrations of DB2 were incubated with B. subtilis at room temperature for 4 hours and left in the dark or irradiated at 530 nm for 10 minutes (41.8 mW cm⁻²). The OD₆₀₀ of the dark control (B. subtilis only) was then measured to calculate the dilution factor required to reach a final OD₆₀₀ of 2.5 × 10⁻⁴ for plating. All samples were diluted with LB broth, plated into 6-well plates containing agarose, and spread evenly with glass beads. The plates were incubated overnight at 37 °C and then the colonies were counted to assess cell viability (Fig. S5†). DB2 gave an absolute IC₅₀ value of 0.16 μM under the irradiation conditions with dark toxicity being observed in concentrations >3.2 μM demonstrating its dependence on light and its potential for antimicrobial PDT (Fig. 4b).

Furthermore, DB2 was tested in MRC-9 fibroblast cells, a normal lung cell line, to determine if toxicity would be observed in mammalian cells under normoxic (21% oxygen) conditions. The MRC-9 cells were incubated for four hours with DB2 (0 or 0.32 μM) and treated under dark or the same irradiation conditions (530 nm, 41.8 mW cm⁻²). A thiazolyl blue tetrazolium bromide (MTT) assay was used to assess cell viability, and no toxicity was observed in the mammalian cells under dark conditions. A mild decrease in cell viability (77 ± 8% survival) was observed in cells containing DB2 and under irradiation (Fig. 4c) but when compared to the cell viability of B. subtilis under the same concentration and irradiation conditions (3.5 ± 0.8% survival), a difference of 21-fold can be seen. This is comparable to other antimicrobial techniques and showcases the selectivity DB2 has towards targeting bacteria compared to mammalian cells.³³ We also tested DB1 under the same conditions and observed no light or dark toxicity. Since DB1 contains not NTR trigger group, we hypothesize its lack of light-dependent toxicity may be due to poor cell permeability. Similar results with DB2 were also obtained in the mammalian, HeLa cervical cancer cell line (Fig. S6†),

To demonstrate that NTR expression aids DB2 in causing bacterial cell death, the compound EY2 was synthesized using similar synthetic procedures as DB2 except EY served as the base PS (Fig. 5a, Scheme S1†). EY was chosen due to the similarities it has with DBFCN except, instead of having two bromines, EY contains four. Therefore, it was hypothesized that the extra steric bulk around the ether bond would hinder the recognition of the 4-nitrobenzyl group by NTR making EY2 unable to be activated. In vitro experiments were conducted and EY2 exhibited similar photophysical properties to DB2 (Fig. S7–S9, Table S1†). However, when incubating EY2 with purified NTR, no fluorescence recovery was observed (Fig. S10†) demonstrating the inability of NTR to recognize EY2. Furthermore, when EY2 was incubated with B. subtilis, no substantial fluorescence was observed until 8 hours (Fig. 5b). Finally, the cell viability of B. subtilis was tested under the same conditions with EY2 and less cell death was observed compared to DB2 showcasing the dependence of NTR and, generally, that activation of the aPS can lead to improvements in IC₅₀ values.

Fig. 5 Cell experiments with EY2 control. (a) Structure of EY2. (b) Activation of EY2 in B. subtilis monitored by fluorescence and compared with DB2. Experiments were conducted in triplicate. (c) Cell viability of EY2 in B. subtilis under dark or light conditions (10 minutes, 530 nm, 41.8 mW cm⁻²) compared with DB2. Experiments were conducted in triplicate.

Conclusion

By utilizing DBFCN as the base PS and appending the 4-nitrobenzyl trigger group, the NTR-aPS DB2 was designed and synthesized. The fluorescence and ROS production of DB2 is effectively in an “OFF” state, with lower Φ_f (0.024 ± 0.002) and Φ_Δ (0.20 ± 0.02), relative to DB1 (Φ_f = 0.26 ± 0.01, Φ_Δ = 0.38 ± 0.04), which is the product of the NTR reaction. When incubating with purified NTR enzyme, DB2 demonstrated a release of DB1, monitored by fluorescence recovery, showcasing activation and turning the probe “ON”. Although DB2 did not show activation in gram-negative E. coli, likely due to an inability to penetrate the cell envelope, fluorescence recovery was observed in gram-positive B. subtilis indicating internalization and activation of the compound. In B. subtilis, DB2 demonstrated an absolute IC₅₀ value of 0.16 μM with dark toxicity being observed at >0.32 μM. In MRC-9 cells, DB2 once again demonstrated low dark toxicity and had a cell viability of 77 ± 8%, showcasing a 21-fold difference compared to B. subtilis treated under the same conditions. Finally, EY2 was synthesized as a control to mimic DB2 that could not be recognized by NTR. Under the same conditions, EY2 was unable to be activated and demonstrated lower efficacy in killing B. subtilis, showcasing the potential of using aPSs.

Future works include testing DB2 against clinically relevant gram-positive bacterial strains such as Staphylococcus aureus^34,35 and in bacterial biofilms to mimic disease models more accurately.^36,37 The synthesis of DB2 with different NTR trigger groups such as 5-nitrothiophene and 5-nitrofuran can also be pursued.³⁸ By varying the trigger group, the fluorescence and ROS quenching, permeability into cells, and the time scale for activation with NTR can be altered, which may lead to more effective NTR-aPSs. Overall, the development of DB2 further validates the use of aPSs for aPDT to mitigate photosensitivity and provide a solution to the rising trend in AMR.

Methods

Synthetic materials and instrumentation

All chemicals and solvents were purchased from commercial suppliers and used without further purification. ¹H NMR spectra were obtained using a Bruker 400 MHz NMR spectrometer while ¹³C NMR spectra were obtained using a 500 MHz Agilent DD2 NMR spectrometer. High-resolution mass spectroscopy (HRMS) was obtained using the Agilent 6538 UHD. Cell lines, MRC-9 and HeLa were purchased from ATCC.

Synthesis of DB1

4′,5′-Dibromofluorescein (123 mg, 0.25 mmol) was dissolved in a 2 : 1 mixture of CHCl₃ and MeOH (15 mL). 1-Ethyl-3-(3-dimethyllaminopropyl)carbodiimide hydrochloride (48 mg, 0.25 mmol) and hydroxy-benzotriazole (34 mg, 0.25 mmol) were stirred at room temperature for 30 minutes. 4-(Dimethylamino)pyridine (122 mg, 1 mmol) was added and refluxed overnight at 65 °C. The reaction was then evaporated under vacuum and purified by silica chromatography (100% DCM to DCM/MeOH 90 : 10) to yield an orange solid (107 mg, 85% yield). ¹H NMR (400 MHz, DMSO) δ 8.23 (d, J = 7.8 Hz, 1H), 7.91–7.75 (m, 2H), 7.51 (d, J = 7.5 Hz, 1H), 6.90–6.75 (m, 4H), 3.59 (s, 3H). ¹³C NMR (126 MHz, DMSO) δ 165.11, 151.32, 133.59, 133.24, 130.74, 130.70, 130.26, 129.33, 128.96, 115.49, 52.40. HRMS (ESI-MS): calculated for C₂₁H₁₂Br₂O₅ ([M − H]⁻) = 500.8973 Da, found 500.8971 Da.

Synthesis of DB2

DB1 (100 mg, 0.2 mmol), 4-nitrobenzyl bromide (172.8 mg, 0.8 mmol), and tetrabutylammonium bromide (129 mg, 0.4 mmol) were dissolved in DCM (5 mL). K₂CO₃ (55.3 mg, 0.4 mmol) was dissolved in water (5 mL) and added to the above solution. The reaction was then stirred vigorously for 18 hours at room temperature. The reaction was then extracted with DCM and brine. The organic layer was dried with sodium sulphate, filtered, and evaporated under vacuum before purification by silica chromatography (100% DCM to DCM/MeOH 98 : 2) to yield an orange solid (28 mg, 22% yield). ¹H NMR (400 MHz, DMSO) δ 8.31 (d, J = 8.5 Hz, 2H), 8.25 (d, J = 7.8 Hz, 1H), 7.94–7.75 (m, 4H), 7.52 (d, J = 7.5 Hz, 1H), 7.25 (d, J = 9.1 Hz, 1H), 6.99 (d, J = 9.0 Hz, 1H), 6.87 (d, J = 9.8 Hz, 1H), 6.60 (d, J = 9.7 Hz, 1H), 5.55 (s, 2H), 3.60 (s, 3H). ¹³C NMR (126 MHz, DMSO) δ 177.51, 165.11, 158.81, 154.78, 150.61, 149.55, 147.28, 143.52, 133.39, 133.35, 130.84, 130.73, 130.41, 130.05, 129.32, 128.29, 123.80, 117.73, 116.22, 111.28, 102.20, 99.37, 69.86, 52.44. HRMS (ESI-MS): calculated for C₂₈H₁₇Br₂NO₇ ([M + H]⁺) = 637.9450 Da, found 637.9435 Da.

Synthesis of EY1

Synthesized using procedures for DB1, the compound was an orange-red solid (92% yield). ¹H NMR (400 MHz, DMSO) δ 8.21 (d, J = 7.8 Hz, 1H), 7.85 (dt, J = 32.8, 7.6 Hz, 2H), 7.51 (d, J = 7.5 Hz, 1H), 7.00 (s, 2H), 3.61 (s, 3H). ¹³C NMR (126 MHz, DMSO) δ 165.16, 152.48, 151.02, 133.24, 133.08, 130.79, 130.68, 130.45, 129.58, 129.52, 128.40, 123.60, 112.49, 99.83, 52.47. HRMS (ESI-MS): calculated for C₂₁H₁₀Br₄O₅ ([M − H]⁻) = 656.7183 Da, found 656.7172 Da.

Synthesis of EY2

Synthesized using procedures for DB1, the compound was a red solid (24% yield). ¹H NMR (400 MHz, DMSO) δ 8.31 (m, 3H), 7.98–7.83 (m, 4H), 7.53 (d, J = 8.2 Hz, 1H), 7.33 (s, 1H), 7.20 (s, 1H), 5.33 (s, 2H), 3.65 (s, 3H). ¹³C NMR (126 MHz, DMSO) δ 165.03, 156.33, 150.18, 148.83, 147.45, 143.10, 133.65, 132.23, 131.17, 130.94, 130.68, 130.09, 129.31, 128.99, 125.78, 123.69, 120.20, 119.35, 114.09, 107.79, 100.94, 73.74, 52.61. HRMS (ESI-MS): calculated for C₂₈H₁₅Br₄NO₇ ([M + H]⁺) = 793.7660 Da, found 793.7641 Da.

Photophysical characterization instrumentation

All absorbance spectra were obtained on a Shimadzu UV-1800 UV Spectrophotometer and blanked with the respective solvent before measurements. Normalized spectra were obtained by normalizing to the λ_max^abs of the respective compound. All fluorescence emission spectra and fluorescence time courses were obtained on a Shimadzu RF-6000 Spectrofluorophotometer. Quartz cuvettes with a 1.0 cm path length were used and purchased from Starna Scientific Ltd. Two LEDs, a 490 nm (M490L4) and a 530 nm (M530L3), were used for irradiation and were purchased from Thorlabs. The irradiation powers were determined using a Newport Optical Power meter Model 1916-R.

Fluorescence quantum yield determination

Fluorescence quantum yields of DB1 and DB2 in PBS were measured using rhodamine B (Φ_f = 0.50 in ethanol)³⁰ as the reference standard and determined using the following equation:
where: S = sample, R = reference, n = refractive index, I = integrated fluorescence intensity, A = absorbance, Φ = fluorescence quantum yield.

Subscripts s and r represent sample and reference standard, respectively.

Rhodamine B was measured in ethanol (n = 1.36) while samples were measured in PBS (n = 1.33). The absorbance of rhodamine B was taken at the maximal wavelength (λ_max) indicated in Table 1 to be 0.1 AU. Absorbance values of DB1 and DB2 were all ≤ 0.1 AU to avoid inner filter effects. Emission spectra of each sample with excitation at maximal wavelength (3 nm bandwidth) was taken with emission measured from λ_max + 5 nm to 800 nm (1 nm emission bandwidth) at 1 nm s⁻¹ scan speed. Each emission spectra was integrated by Riemann sums at 1 nm intervals from λ_max + 5 nm to 800 nm.

¹O₂ Quantum yield determination (ABDA)

The absorbance of all compounds and the respective control, EY (Φ_Δ = 0.50 in H₂O),³¹ were matched at 490 nm to an absorbance of ∼0.1 AU in PBS containing 2% DMSO. A 10 mM stock of ABDA was prepared in DMSO and diluted into the sample to an absorbance of ∼0.3 AU at 401 nm. Absorbance at 401 nm was monitored over the irradiation period (490 nm, 28.0 mW cm⁻²) to obtain the slopes of degradation of ABDA and calculate the singlet oxygen quantum yield using the following equation:
where: S = sample, R = reference, A = absorbance, m = slope of ABDA degradation.

The equation does not consider differences in refractive indexes or irradiation powers as the experiments were all conducted in MeOH with the same LED. All measurements were conducted in triplicate.

Determination of extinction coefficients by qNMR

Compounds were dissolved in 392 μL DMSO-d₆ and 8 μL of a 500 μM stock of 1,4-dioxane in DMSO-d₆ was added. This provided a final concentration of 10 μM of 1,4-dioxane as the standard. The sample was then analyzed using a Bruker 400 MHz spectrometer and the ratio of the integrals of the chemical peaks was used to determine the concentration of the compound. UV-vis spectra were then obtained in anhydrous DMSO with a series of dilutions to generate a slope of the absorbance vs. concentration to determine the ε using Beer–Lambert Law (A = εlc, where l is 1 cm). All measurements were done in triplicate.

NTR assay

Stocks of all compounds were kept in DMSO and all absorbance and fluorescence spectra were obtained at a final concentration of PBS (2% DMSO). Nitroreductase enzyme expressed in E. coli was purchased from Sigma-Aldrich and a final concentration of 200 μM NADH and 500 nM NTR was used for all experiments. Fluorescence time courses were obtained upon exciting at the λ_abs and monitoring the λ_em for DB1 and EY1 respectively.

Bacterial strains and preparation

Colonies of E. coli (N99 pXG10sf-GFP) and B. subtilis (wild-type) were thawed from frozen concentrated stocks. Bacteria were inoculated from an agar plate into 5 mL of LB broth (ampicillin and tetracycline antibiotics added to N99 E. coli) and grown at 37 °C overnight with shaking at 250 rpm. The next day, bacteria were regrown by diluting the overnight culture 100× into fresh LB broth (5 mL total volume) (antibiotics added for N99 E. coli). Bacteria were grown at 37 °C for an additional 2 hr with shaking at 250 rpm until they reached an exponential phase (OD₆₀₀ = 0.10–0.50). Standardized suspensions of bacteria were prepared for all experiments to the same OD₆₀₀.

Fluorescent assay with bacteria

After the bacteria were cultured overnight and regrown, an initial OD₆₀₀ = 0.05 AU was used for fluorescent assays. Compound stocks in DMSO were added to the bacterial solution to a final concentration of 1 μM and contained a final concentration of 2% DMSO. Fluorescence was monitored by exciting the λ_abs and monitoring the λ_em of either DB1 or EY1 every 2 hours for the first 8 hours and one final timepoint at 24 hours. Bacteria were kept in the dark and at room temperature to prevent irradiation from ambient light and to minimize the growth of the bacteria. All measurements were conducted in triplicate.

Cell viability B. subtilis

After culturing and regrowing B. subtilis, an initial OD₆₀₀ = 0.05 AU was used and incubated with a range of concentrations of DB2 or EY2 (0–3.2 μM) at room temperature for 4 hours and either left in the dark or irradiated with a 530 nm for 10 minutes (41.8 mW cm⁻²). The OD₆₀₀ of the dark control (B. subtilis only) was measured to calculate the dilution factor required to reach a final OD₆₀₀ of 2.5 × 10⁻⁴ for plating. All samples were diluted with LB broth and 5 μL of the diluted sample was spread using glass beads onto 6-well cell culture plates (35 mm diameter, cell star, cat. no. 657160) prepared previously by melting agarose (3 mL) and leaving to solidify at room temperature. Plates were left to incubate at 37 °C overnight before counting the colonies formed. Experiments were conducted in triplicate.

Mammalian (HeLa and MRC-9) cell culture conditions

HeLa cells were cultured and maintained in a 75 cm² culture flask (Nunc™ EasYFlask™ 75 cm² Nunclon™ Delta Surface) at 37 °C and 5% CO₂ atmosphere in a Thermo Scientific Forma Steri-Cycle CO₂ Incubator. The cells were grown in Dulbecco's Modified Eagle Medium (DMEM) with sodium pyruvate, 4.5 g L⁻¹ glucose and L-glutamine (Wisent Inc.) supplemented with 10% fetal bovine serum and 1% antibiotics/antimycotics (complete growth medium).

MRC-9 cells were grown in minimum essential medium Eagle's (EMEM) with L-alanine, L-proline, L-serine, 1.5 g L⁻¹ sodium bicarbonate and sodium pyruvate, Earle's salts, L-glutamine, and glycine (Wisent Inc.) supplemented with 10% fetal bovine serum and 1% antibiotics/antimycotics (complete growth medium).

Cell viability with DB1 or DB2 (MTT assay)

MRC-9 and HeLa cells were seeded at a density of 10 000 cells per well into 96-well plates (Thermo Scientific Nunclon™ Delta Surface) and incubated with 200 μL of EMEM or DMEM, respectively, overnight. The following day, the media was removed, and the cells were washed with D-PBS and replaced with 200 μL opti-minimal essential medium (Opti-MEM™, Thermo Scientific) containing 0 or 0.32 μM DB2 and DB1 to a final concentration of 2% DMSO. Plates were incubated at 37 °C for four hours, washed with D-PBS and the media was replaced with 200 μL of corresponding media. Plates were either kept in the dark or irradiated with a 530 nm LED for 10 minutes before incubating at 37 °C overnight. The following day, 20 μL of a 5 mg mL⁻¹ solution of thiazolyl blue tetrazolium bromide (MTT) in D-PBS was added to each well and incubated for 3 hours. The media was removed, and 150 μL DMSO was added to dissolve the formazan products. Plates were then read at 565 nm using a BioTek Synergy™ HTX multi-mode microplate reader. Absorbance values were then used to calculate the cell viability, and all viability experiments were conducted in triplicate.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

M. T. T. acknowledges support from the Ontario Graduate Scholarship and Queen Elizabeth II – Graduate Scholarship in Science and Technology. I. R. L.-M. acknowledges support from a Doctoral Research Award from Cancer Research Society (Grant 1330571) and a Queen Elizabeth II/Dina Gordon Malkin Graduate Scholarship in Science and Technology.

Notes and references

1. J. O'Neill, Tackling drug-resistant infections globally: Final report and recommendations. HM Government and Wellcome Trust, London, 2016, https://amr-review.org/sites/default/files/160518_Final%20paper_with%20cover.pdf (accessed 2024-06-01).
2. T. C. Pham, V. N. Nguyen, Y. Choi, S. Lee and J. Yoon, Recent Strategies to Develop Innovative Photosensitizers for Enhanced Photodynamic Therapy, Chem. Rev., 2021, 121(21), 13454–13619,  DOI:10.1021/acs.chemrev.1c00381.
3. Y. Y. Huang, S. K. Sharma, T. Dai, H. Chung, A. Yaroslavsky, M. Garcia-Diaz, J. Chang, L. Y. Chiang and M. R. Hamblin, Can Nanotechnology Potentiate Photodynamic Therapy?, Nanotechnol. Rev., 2012, 1(2), 111–146,  DOI:10.1515/NTREV-2011-0005.
4. J. H. Correia, J. A. Rodrigues, S. Pimenta, T. Dong and Z. Yang, Photodynamic Therapy Review: Principles, Photosensitizers, Applications, and Future Directions, Pharmaceutics, 2021, 13(9), 1–16,  DOI:10.3390/pharmaceutics13091332.
5. H. Mahmoudi, A. Bahador, M. Pourhajibagher and M. Y. Alikhani, Antimicrobial Photodynamic Therapy: An Effective Alternative Approach to Control Bacterial Infections, J. Lasers Med. Sci., 2018, 9(3), 154–160,  DOI:10.15171/JLMS.2018.29.
6. Y. Liu, R. Qin, S. A. J. Zaat, E. Breukink and M. Heger, Antibacterial Photodynamic Therapy: Overview of a Promising Approach to Fight Antibiotic-Resistant Bacterial Infections, J. Clin. Transl. Res., 2015, 1(3), 140–167,  DOI:10.18053/jctres.201503.002.
7. S. Rajesh, E. Koshi, K. Philip and A. Mohan, Antimicrobial Photodynamic Therapy: An Overview, J. Indian Soc. Periodontol., 2011, 15(4), 323–327,  DOI:10.4103/0972-124X.92563.
8. R. B. Cecatto, L. S. Magalhaes, M. F. S. D. Rodrigues, C. Pavani, A. Lino-dos-Santos-Franco, M. T. Gomes and D. F. T. Silva, Methylene Blue Mediated Antimicrobial Photodynamic Therapy in Clinical Human Studies: The State of the Art, Photodiagn. Photodyn. Ther., 2020, 31, 101828,  DOI:10.1016/j.pdpdt.2020.101828.
9. Y. Jao, S. J. Ding and C. C. Chen, Antimicrobial Photodynamic Therapy for the Treatment of Oral Infections: A Systematic Review, J. Dent. Sci., 2023, 18(4), 1453–1466,  DOI:10.1016/j.jds.2023.07.002.
10. J. Ghorbani, D. Rahban, S. Aghamiri, A. Teymouri and A. Bahaador, Photosensitizers in Antibacterial Photodynamic Therapy: An Overview, Laser Ther., 2018, 27(4), 293–302,  DOI:10.5978/islsm.27_18-RA-01.
11. M. Klausen, M. Ucuncu and M. Bradley, Design of Photosensitizing Agents for Targeted Antimicrobial Photodynamic Therapy, Molecules, 2020, 25, 5239,  DOI:10.3390/molecules25225239.
12. E. M. Digby, T. Ma, W. R. Zipfel, J. N. Milstein and A. A. Beharry, Highly Potent Photoinactivation of Bacteria Using a Water-Soluble, Cell-Permeable, DNA-Binding Photosensitizer, ACS Infect. Dis., 2021, 7(11), 3052–3061,  DOI:10.1021/acsinfecdis.1c00313.
13. I. Nieves, C. Hally, C. Viappiani, M. Agut and S. Nonell, A Porphycene-Gentamicin Conjugate for Enhanced Photodynamic Inactivation of Bacteria, Bioorg. Chem., 2020, 97, 103661,  DOI:10.1016/J.BIOORG.2020.103661.
14. G. Feng, Y. Yuan, H. Fang, R. Zhang, B. Xing, G. Zhang, D. Zhang and B. Liu, A Light-up Probe with Aggregation-Induced Emission Characteristics (AIE) for Selective Imaging, Naked-Eye Detection and Photodynamic Killing of Gram-Positive Bacteria, Chem. Commun., 2015, 51, 12490,  10.1039/c5cc03807c.
15. L. Huang, M. Wang, Y. Y. Huang, A. El-Hussein, L. M. Wolf, L. Y. Chiang and M. R. Hamblin, Progressive Cationic Functionalization of Chlorin Derivatives for Antimicrobial Photodynamic Inactivation and Related Vancomycin Conjugates, Photochem. Photobiol. Sci., 2018, 17, 638,  10.1039/c7pp00389g.
16. R. C. C. de Pietra, R. C. Cruz, C. N. Melo, L. B. Rodrigues, P. C. Santos, G. P. M. Bretz, B. M. Soares, G. R. de Sousa, M. V. L. Ferreira, P. S. Cisalpino, P. P. Magalhães, L. de M. Farias and M. Pinotti, Evaluation of Polymeric PLGA Nanoparticles Conjugated to Curcumin for Use in APDT, Braz. J. Pharm. Sci., 2017, 53(2), e16043,  DOI:10.1590/S2175-97902017000216043.
17. V. Klepac-Ceraj, N. Patel, X. Song, C. Holewa, C. Patel, R. Kent, M. M. Amiji and N. S. Soukos, Photodynamic Effects of Methylene Blue-Loaded Polymeric Nanoparticles on Dental Plaque Bacteria, Lasers Surg. Med., 2011, 43(7), 600–606,  DOI:10.1002/LSM.21069.
18. G. S. De Paula, M. C. Oliveira, L. S. Sales, M. Boriollo, L. K. A. Rodrigues, M. Nobre-dos-Santos and C. Steiner-Oliveira, Antimicrobial Photodynamic Therapy Mediated by Methylene Blue Complex to β-Cyclodextrin Reduces Early Colonizing Microorganisms from the Oral Biofilm, Photodiagn. Photodyn. Ther., 2021, 34, 102283,  DOI:10.1016/j.pdpdt.2021.102283.
19. G. A. Johnson, E. A. Ellis, H. Kim, N. Muthukrishnan, T. Snavely and J. P. Pellois, Photoinduced Membrane Damage of E. Coli and S. Aureus by the Photosensitizer-Antimicrobial Peptide Conjugate Eosin-(KLAKLAK)2, PLoS One, 2014, 9(3), e91220,  DOI:10.1371/JOURNAL.PONE.0091220.
20. K. Yang, B. Gitter, R. Rüger, G. D. Wieland, M. Chen, X. Liu, V. Albrecht and A. Fahr, Antimicrobial Peptide-Modified Liposomes for Bacteria Targeted Delivery of Temoporfin in Photodynamic Antimicrobial Chemotherapy, Photochem. Photobiol. Sci., 2011, 10, 1593–1601,  10.1039/C1PP05100H.
21. Z. Li, Q. Feng and J. Shen, A β-Lactamase-Activatable Photosensitizer for the Treatment of Resistant Bacterial Infections, Chin. Chem. Lett., 2024, 35(11), 109602,  DOI:10.1016/J.CCLET.2024.109602.
22. W. Zhou, X. Da, Y. Jian, Y. Peng, X. Liu, Y. Xu, Y. Wu, X. Wang and Q. Zhou, Nitroreductase-Responsive Photosensitizers for Selective Imaging and Photo-Inactivation of Intracellular Bacteria, Chem. – Eur. J., 2024, 30(16), e202303766,  DOI:10.1002/CHEM.202303766.
23. M. D. Roldán, E. Pérez-Reinado, F. Castillo and C. Moreno-Vivián, Reduction of Polynitroaromatic Compounds: The Bacterial Nitroreductases, FEMS Microbiol. Rev., 2008, 32(3), 474–500,  DOI:10.1111/J.1574-6976.2008.00107.x.
24. H. A. J. Hibbard and M. M. Reynolds, Synthesis of Novel Nitroreductase Enzyme-Activated Nitric Oxide Prodrugs to Site-Specifically Kill Bacteria, Bioorg. Chem., 2019, 93, 103318,  DOI:10.1016/J.BIOORG.2019.103318.
25. K. A. Pardeshi, T. A. Kumar, G. Ravikumar, M. Shukla, G. Kaul, S. Chopra and H. Chakrapani, Targeted Antibacterial Activity Guided by Bacteria-Specific Nitroreductase Catalytic Activation to Produce Ciprofloxacin, Bioconjugate Chem., 2019, 30(3), 751–759,  DOI:10.1021/ACS.BIOCONJCHEM.8B00887.
26. S. Jiao, S. Yang, X. Meng and C. Wang, One Step Synthesis of Red-Emitting Fluorescence Turn-on Probe for Nitroreductase and Its Application to Bacterial Detection and Oral Cancer Cell Imaging, Spectrochim. Acta, Part A, 2020, 241, 118637,  DOI:10.1016/J.SAA.2020.118637.
27. J. W. Yoon, S. Kim, Y. Yoon and M. H. Lee, A Resorufin-Based Fluorescent Turn-on Probe Responsive to Nitroreductase Activity and Its Application to Bacterial Detection, Dyes Pigm., 2019, 171, 107779,  DOI:10.1016/J.DYEPIG.2019.107779.
28. F. Xu, H. Li, Q. Yao, H. Ge, J. Fan, W. Sun, J. Wang and X. Peng, Hypoxia-Activated NIR Photosensitizer Anchoring in the Mitochondria for Photodynamic Therapy, Chem. Sci., 2019, 10(45), 10586–10594,  10.1039/C9SC03355F.
29. Z. Liu, F. Song, W. Shi, G. Gurzadyan, H. Yin, B. Song, R. Liang and X. Peng, Nitroreductase-Activatable Theranostic Molecules with High PDT Efficiency under Mild Hypoxia Based on a TADF Fluorescein Derivative, ACS Appl. Mater. Interfaces, 2019, 11(17), 15426–15435,  DOI:10.1021/ACSAMI.9B04488.
30. T. Karstens and K. Kobs, Rhodamine B and Rhodamine 101 as Reference Substances for Fluorescence Quantum Yield Measurements, J. Phys. Chem., 1980, 84, 1871–1872,  DOI:10.1021/j100451a030.
31. R. W. Redmond and J. N. Gamlin, A Compilation of Singlet Oxygen Yields from Biologically Relevant Molecules, Photochem. Photobiol., 1999, 70(4), 391–475,  DOI:10.1111/J.1751-1097.1999.TB08240.X.
32. C. Maher and K. A. Hassan, The Gram-Negative Permeability Barrier: Tipping the Balance of the in and the out, MBio, 2023, 14(6), e01205,  DOI:10.1128/MBIO.01205-23.
33. E. H. H. Wong, M. M. Khin, V. Ravikumar, Z. Si, S. A. Rice and M. B. Chan-Park, Modulating Antimicrobial Activity and Mammalian Cell Biocompatibility with Glucosamine-Functionalized Star Polymers, Biomacromolecules, 2016, 17, 1170–1178,  DOI:10.1021/ACS.BIOMAC.5B01766.
34. G. Y. C. Cheung, J. S. Bae and M. Otto, Pathogenicity and Virulence of Staphylococcus Aureus, Virulence, 2021, 12(1), 547–569,  DOI:10.1080/21505594.2021.1878688.
35. S. Y. C. Tong, J. S. Davis, E. Eichenberger, T. L. Holland and V. G. Fowler, Staphylococcus Aureus Infections: Epidemiology, Pathophysiology, Clinical Manifestations, and Management, Clin. Microbiol. Rev., 2015, 28(3), 603–661,  DOI:10.1128/CMR.00134-14.
36. A. Zhao, J. Sun and Y. Liu, Understanding Bacterial Biofilms: From Definition to Treatment Strategies, Front. Cell. Infect. Microbiol., 2023, 13, 1137947,  DOI:10.3389/FCIMB.2023.1137947/BIBTEX.
37. H. K. N. Vyas, B. Xia and A. Mai-Prochnow, Clinically Relevant in Vitro Biofilm Models: A Need to Mimic and Recapitulate the Host Environment, Biofilm, 2022, 4, 100069,  DOI:10.1016/J.BIOFLM.2022.100069.
38. T. D. Gruber, C. Krishnamurthy, J. B. Grimm, M. R. Tadross, L. M. Wysocki, Z. J. Gartner and L. D. Lavis, Cell-Specific Chemical Delivery Using a Selective Nitroreductase-Nitroaryl Pair, ACS Chem. Biol., 2018, 13(10), 2888–2896,  DOI:10.1021/ACSCHEMBIO.8B00524.

---

Footnote

† Electronic supplementary information (ESI) available: Photophysical characterizations of all compounds, including extinction coefficient graphs, in vitro assays, E. coli experiments with DB2, images of colony formation for B. subtilis treatments, synthetic scheme for EY2, ¹H and ¹³C NMR spectra, and MS data. See DOI: https://doi.org/10.1039/d4md00890a

---