2.1.

Preparation of Photosensitizers

According to our early published protocols,^14,15 the unsymmetrical $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ was prepared via the activation of the carboxylic acid group on ZnPc-COOH by 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide and N-hydroxysuccinimide, and a coupling to pentalysine. The synthesized $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ was characterized by ${}^1\mathrm{H}$ NMR [AV-400, Bruker, 400 MHz, in dimethyl sulphoxide (DMSO)], ${}^{13}\mathrm{C}$ NMR (75 MHz, in DMSO), FTIR (Magna-IR 750, Nicolett), and high-resolution mass spectra–electrospray ionization (DECAX-30000 LCQ Deca XP mass spectrometer). The UV–vis absorption spectrum of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ in dimethylfomamide was recorded from 300 to 800 nm using quartz cuvettes with 1-cm path length on a Lambda-35 UV/vis spectrometer (PerkinElmer, Massachusetts) in DMSO. Analytical high-performance liquid chromatography (HPLC) was carried out on a C-18 reversed-phase HPLC system (Dalian Elite Analytical Instruments Co., Ltd., Dalian, China; Column: SinoChrom ODS-BP $250\times 4.6\mathrm{mm}$, $5\mu \mathrm{m}$), using a linear gradient from 50% to 100% (v/v) of methanol:acetonitrile ($1:1$) at a flow rate of $1\mathrm{ml}/\min$. The UV–vis absorption spectrum of photosensitizer $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ in DMSO (Fig. 1) was typical of ZnPc with the strongest absorption at 678 nm (extinction coefficient of $\mathrm{118,380}\mathrm{L}·{\mathrm{mol}}^{-1}·{\mathrm{cm}}^{-1}$). Furthermore, singlet oxygen (${}^1{\mathrm{O}}_2$) is believed to be the major cytotoxic agent involved in photodynamic therapy. The quantum yield of singlet oxygen generation of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ was measured in reference to ZnPc, which has a quantum yield of 0.67 in DMSO.¹⁶

Fig. 1

UV–vis absorption spectrum of photosensitizer $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ in DMSO.

Zinc phthalocyanine tetrasulfonate ($\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$) with an absorption peak at 680 nm was kindly provided by Professor Naisheng Chen of Fuzhou University, China.

2.2.

Microorganism and Culture Condition

As the standard bacteria used in antibacterial activity tests according to the National Standard of the People’s Republic of China in detection and control of pathogens (GB4789.28-2013), P. acnes (ATCC 6919) and S. aureus subsp. aureus (ATCC 6538) were purchased from Beijing Zhongyuan Ltd. S. aureus were cultured in nutrient agars, and P. acnes were cultured in thioglycollate medium in anaerobic bags. Thioglycollate medium is a semisolid nutrient medium containing a low concentration of agar to prevent convection of oxygen from the surface. The thioglycollate medium was boiled before use in order to drive off the oxygen and inoculated without excessive shaking of the medium after cooling.

A luminescent strain of S. aureus Xen29 (NCTC8532), a derivative of the biofilm forming S. aureus 12600 that possesses a stable copy of the modified Photorhabdus luminescens luxABCDE operon at a single integration site on the bacterial chromosome, was purchased from Caliper Life Sciences, Inc., and grown at 37°C using Luria–Bertani broth containing kanamycin ($200\mu \mathrm{g}/\mathrm{ml}$ to select for resistance encoded by the plasmid) to an absorbance of 0.5 at 600 nm corresponding to $1.44\times {10}^8\text{organisms}/\mathrm{ml}$.

2.3.

Cellular Uptake of Photosensitizer by Bacteria

Aliquots of microorganism suspension [${10}^6$ colony forming unit (CFU)/ml] were incubated in 96-multiwell plates (Falcon) with photosensitizer $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ at different concentrations (${10}^{-7}$, ${10}^{-6.5}$, and ${10}^{-6}\mathrm{M}$) for 1 h at 37°C. The exponentially growing cells were then washed with sterile phosphate buffer saline (PBS) before lysis with NaOH [0.1 N, 1.0 ml with 1% sodium dodecyl sulfate (SDS)] to give a homogeneous solution. The fluorescence of the cell extract was measured on a microplate reader (Synergy 4, BioTek Instruments). The concentration of cellular protein was determined using a bicinchoninic acid protein assay kit (Pierce, Thermo Fisher Scientific). Standard curves were made with cell lysates treated as above with known added amounts of bovine serum albumin. Results are expressed as nmol of phthalocyanine per mg cell protein.

2.4.

Antimicrobial Studies in Cultured Bacteria

Bacteria suspensions ($\sim {10}^6\mathrm{CFU}/\mathrm{ml}$) were incubated in the dark at room temperature for 1 h with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ (at concentrations of ${10}^{-7}$, ${10}^{-6.5}$, ${10}^{-6}$, ${10}^{-5.5}$, and ${10}^{-5}\mathrm{M}$) followed by the light exposure using a light-emitting diode (LED) light source (SunDynamic, Inc., Qingdao) of 680 nm and with a power of 100 mW for 1 or 2 min (i.e., light dosages at 3 or $6\mathrm{J}/{\mathrm{cm}}^2$). After illumination of the appropriate wells, cell viability was determined by incubating agar plates overnight after applying 0.1 ml from one of the tubes in a bacterial dilution series. The percentage of cell survival was determined by comparing the colony counts from treated plates to the colony counts from control plates that were incubated without the conjugate and were kept in dark for periods equal to irradiation times.

2.5.

Antimicrobial Studies in Experimental Animals

Sprague–Dawley rats (purchased from Shanghai SLAC Laboratory Animal Co. Ltd., Shanghai, China) were maintained and handled in accordance with the recommendations of an Institutional Animal Care and Use Committee (IACUC). The animals were allowed free access to water throughout the course of the experiments. The experimental procedures on rats were in accordance with PerkinElmer IACUC guidelines and approved veterinarian requirements for animal care and use. Rats ($n=16$) weighing $150\pm 2\mathrm{g}$ were anesthetized with an intraperitoneal injection of pentobarbital ($30\mathrm{mg}/\mathrm{kg}$, i.p.) before they were shaved on the back and treated with depilatory cream. Two excisional wounds were then made along the dorsal surface, using surgical scissors, by carefully opening the epidermal layer of the skin with one side connected. Each wound measured $8\times 8\mathrm{mm}$ with at least 5 mm of unbroken skin between wounds. The bottom of the wound was panniculus carnosus, with no visible bleeding. Localized infection on the wounds was induced according to our previously published protocol¹³ with slight changes by inoculating a suspension ($25\mu l$) of midlog phase bioluminescent Xen29 S. aureus ($\sim {10}^8\text{cells}/\mathrm{ml}$ in PBS) into each wound of the rats. The bacteria was allowed to attach to the tissue for 30 min, then $25\mu \mathrm{l}$ of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ (1 mM in PBS) or $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ (1 mM in PBS) was added into the wounds. After a further 30 min were given to allow the conjugates to bind to and penetrate the bacteria under subdued room lighting, eight rats were illuminated using a 680-nm light source (SunDynamic, Inc., Qingdao) with an irradiance of 100 mW for $5\min /\text{wound}$, giving a total light fluence of $15\mathrm{J}/{\mathrm{cm}}^2$ for each wound. The other eight rats were kept in their cages in the dark during the whole experimental period.

To measure the bacteria load after PACT treatment, we cut out the epidermis layer of the skin from rats under anesthesia. After being shredded into small pieces, the tissues were mixed with PBS, and the bioluminescence signals were measured on a Synergy™ 4 multimode microplate reader (BioTek Instruments, Winooski, Vermont). The viability of bacteria after PACT treatment was measured by comparing the luminescence signal to that of a control group treated with saline. All rats were kept in cages after the treatment. Their wound areas were measured in two dimensions each day for 12 days. To precisely observe the early wound healing after different treatments, the laser speckle contrast imaging (LSCI) system (MoorFLPI-2, Moor Instruments Ltd., United Kingdom) was used to monitor cutaneous blood flow in the wound area after PACT treatment ($\sim 2$ weeks later). The spatiotemporal characteristics of blood flow under the wound area were addressed.

3.1.

Preparation of Photosensitizer ZnPc-(Lys)₅

The structure of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ was confirmed by NMR spectroscopy and mass spectrometry as described previously.^13–15 Furthermore, the quantum yield for singlet oxygen photogeneration for $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ in DMSO was determined to be $0.64\pm 0.02$, which is similar to that for ZnPc (0.67),¹⁶ showing that the coupling of a positively charged pentalysine group does not significantly alter the generation of singlet oxygen.

3.2.

Cellular Uptake of Photosensitizer by Bacteria

The uptake profile of a photosensitizer is important in the clinical setting in order to adjust the dose for effective treatment.¹⁷ It is believed that bacteria P. acnes and S. aureus are the main pathogenic factors involved in some skin infections, like cellulitis, abscesses, and postsurgical infections. In the current study, we measured the uptake of photosensitizer $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ in the above-mentioned bacteria. Meanwhile, the negatively charged ZnPc compound, $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$, was used as a control. The uptake of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ was shown in a dose-dependent manner with higher concentration of ZnPc leading to higher cellular uptake on both bacteria we measured (Fig. 2). Compared with the anionic photosensitizer $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$, $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ exhibited much higher cellular uptake amounts in both bacteria (S. aureus and P. acnes). This enhanced uptake amount of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ is likely due to the binding of the positively charged $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ to the outer membrane of bacteria that bears a strong negative charge.¹⁸ Furthermore, we noticed that the uptake amount of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ (at $1\mu \mathrm{M}$) in S. aureus ($1.97\mathrm{nmol}/\mathrm{mg}$ protein) was about 2.7 times more than that in P. acnes ($0.73\mathrm{nmol}/\mathrm{mg}$ protein).

Fig. 2

Dose-dependent cellular uptakes of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ by bacteria (a) S. aureus and (b) P. acnes. Bacteria were incubated for 1 h with various concentrations of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ and were washed and dissolved in 0.1-N NaOH—1% SDS for quantization of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ using fluorescence signals.

3.3.

Antimicrobial Studies in Cultured Bacteria

With increasing numbers of small, naturally occurring antibacterial peptides being discovered, the mechanism of action of these peptides is being intensively investigated.^19–21 A common factor in all the structures is the polycationic charge due to lysine, histidine, and arginine residues in the amino acid sequence, and the polycationic charge is probably responsible for their initial binding to bacteria. In this study, we conjugated a pentalysine moiety to ZnPc to selectively target bacteria for photodestruction and tested the antibacterial activities of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ under two different light dosages (3 and $6\mathrm{J}/{\mathrm{cm}}^2$) against S. aureus.

Figure 3 shows the antibacterial activities of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ [Fig. 3(a)] and $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ [Fig. 3(b)] against S. aureus at different light dosages (3 and $6\mathrm{J}/{\mathrm{cm}}^2$, respectively). It is obvious that $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ inhibited the bacterial growth in light dosage-dependent and photosensitizer dose-dependent manners [Fig. 3(a)]. In contrast, $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ did not show significant toxicity to S. aureus under the same conditions with the same light and photosensitizer dosages [Fig. 3(b)]. The pronounced photodynamic effect of the cationic $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ on S. aureus may be due to the electrostatic attraction between the photosensitizer and the negatively charged membrane of the bacterium.

Fig. 3

(a) Dose-dependent antimicrobial effects of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ and (b) $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ on S. aureus under two different light doses (3 and $6\mathrm{J}/{\mathrm{cm}}^2$, respectively).

Moreover, $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ did not show significant toxicity to the bacterium in the absence of light illumination (data not shown), suggesting a broad safety margin of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ with $10\text{-}\mu \mathrm{M}$ concentration followed by a light dosage at $6\mathrm{J}/{\mathrm{cm}}^2$.

We further investigated the antimicrobial activities of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ against P. acnes, the primary pathogen responsible for cystic acne, under a light dose at $6\mathrm{J}/{\mathrm{cm}}^2$ using a colony-counting method. It showed that irradiation of P. acnes cells treated with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ (at a concentration of $1\mu \mathrm{M}$) caused a 5 to 6 log reduction of the CFU (Fig. 4). In comparison, the survival fraction of P. acne was about only 4 log reduction of the CFU, which was much smaller than that of S. aureus. This result was in line with the different amount of cellular uptakes we observed (Fig. 2).

Fig. 4

Effects of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ ($1\mu \mathrm{M}$) on organisms related to bacterial skin disorder under a light dose at $6\mathrm{J}/{\mathrm{cm}}^2$. Survival fractions are expressed as ratios of CFU from bacteria treated with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ or $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ and light over CFU of bacteria treated with neither.

3.4.

Antimicrobial Studies in Experimental Animals

The method using genetically engineered bacteria that emit luminescence to monitor bacterial numbers and viability in real time in living animals so far has been demonstrated in several models.^22,23 In the current studies, we evaluated the in vivo efficacy of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ on skin disorder-related bacteria using a localized infection animal model by inoculating the luminescent strain of S. aureus (Xen29) into excisional wounds on the back of rats. We estimated that the luminescence signal was linearly proportional to live bacterial CFU in the range of ${10}^3$ to ${10}^8$ CFU. In our experiments, about 2 million CFU of the luminescent S. aureus from a midlog culture in $25\mu \mathrm{l}$ were applied onto each wound ($8\times 8\mathrm{mm}$) made on the back of anesthetized rats. The antibacterial efficacy of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ on skin disorder-related bacteria was evaluated by measuring the bacterial load right after the PACT treatment and monitoring the wound size for 12 days after the PACT treatment in order to detect the wound healing rate. We observed that $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ dramatically reduced the bacterial load in the wounds upon light irradiation (680 nm, $15\mathrm{J}/{\mathrm{cm}}^2$), whereas $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ without light irradiation or with the anionic photosensitizer $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ (in the presence or absence of light irradiation) did not show an antibacterial effect upon the excisional wounds on the back of rats (Fig. 5). These results show that the cationic $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ has a strong antimicrobial effect in animal models and is more effective than anionic photosensitizer $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$.

Fig. 5

Reduction of viable luminescent S. aureus strain (Xen29) in infected incision wounds on rats upon incubation with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ ($1\mu \mathrm{M}$) and light exposure (680 nm, $15\mathrm{J}/{\mathrm{cm}}^2$). Data points are means of values from wounds on rats, and bars are standard error of the mean (SEM).

In the presence of light, $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ dramatically promoted wound healing in rats from the second day after the treatment compared with $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ (Fig. 6). This accelerating effect on wound healing is much stronger in the presence of light (680 nm, $15\mathrm{J}/{\mathrm{cm}}^2$) compared to the case without the irradiation. Such wound-healing promotional activities were previously reported with two different photosensitizers (5-aminolevulinic acid at $5\mathrm{mg}/\mathrm{kg}$ and hematoporphyrin derivative at $5\mathrm{mg}/\mathrm{kg}$) in an open excision wound model with rats.²⁴ Interestingly, two other reports^25,26 showed that PACT treatment did not have any effect on wound healing. The reason for these conflicting results is not yet clear; it could probably be explained by the use of different photosensitizers and requires further study.

Fig. 6

Effects of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ ($1\mu \mathrm{M}$) on wound recovery of S. aureus infected excisional wounds with rats in the presence or absence of light irradiation (680 nm, $15\mathrm{J}/{\mathrm{cm}}^2$). Each wound was infected with $2\times {10}^6$ CFU S. aureus 30 min before the photodynamic treatment with photosensitizers. Wounds were measured in two dimensions every day after the infection, and the areas were calculated. Data points are means of values from the corresponding wound on rats and bars are SEM.

We noticed that the wound recovery rate was significantly enhanced from the second day to the fifth day after PACT with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ (Fig. 6) and slowed down afterward. This observation is possibly due to the metabolism of $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ and suggests that PACT with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ applied repeatedly for a certain interval of time ($\sim \text{five days}$) would promote the recovery of infected incision wounds on rats.

To further evaluate the wound healing process in bacterial skin infections with PACT treatment, we imaged the blood flow under the wound area upon all wounds healing (2 weeks later) using an LSCI system, which has been used widely to image cutaneous blood flow.^27,28 We observed that PACT with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ in the presence of light irradiation (680 nm, $15\mathrm{J}/{\mathrm{cm}}^2$) significantly reduced the blood flow under the wound area ($\sim 1.4$-fold blood flow as normal skin, Fig. 7(B-c), whereas PACT with $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$ showed $\sim 2.8$-fold blood flow as normal skin in the presence or absence of light irradiation [Fig. 7(B-a)]. These results demonstrated that PACT with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ significantly relieved the symptoms of inflammation during the wound healing process [Fig. 7(C)]. Furthermore, we observed that PACT with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ in the absence of light irradiation partially reduced the blood flow under the wound area [$\sim 2.2$-fold blood flow under normal skin, Fig. 7(B-b)], indicating that ambient light illumination alone was helpful in promoting the wound healing process under PACT with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$.

Fig. 7

In vivo speckle imaging of S. aureus infected excisional wounds with rats treated with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ or $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$. (A) Schematic diagram of the laser speckle imaging system. (B) Three representative speckle images of the wound-skin rats treated with ZnPc-S4 (a) or $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ with (c) or without (b) irradiation (680 nm, $15\mathrm{J}/{\mathrm{cm}}^2$). (C) Analysis of the blood flow ratio under wound-skin of rats 2 weeks after PACT with $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$ or $\mathrm{ZnPc}\text{-}{\mathrm{S}}_4$.

The LSCI system used in the current study allows for real-time monitoring of the skin microvascular structural and functional information. It shows great potential for monitoring blood flow, but the spatial resolution suffers from the scattering of tissue.^28,29 A combination method of LSCI and skin optical clearing is now attracting increasing attention.^28,30,31 The physiological effect of the optical-clearing agents on skin, such as inflammation, seems limited but requires further evaluation.³²

Typically, PACT treatment uses laser light sources. For example, a recent report showed the effectiveness of laser therapy with a wavelength of 850 nm in the prevention of caries and gingivitis in adolescents.³³ We used LED light sources with a power of 100 mW in this study and found it strong enough to produce significant antimicrobial photodynamic effects, in combination with our cationic photosensitizer $\mathrm{ZnPc}\text{-}{(\mathrm{Lys})}_5$. The use of battery-powered, inexpensive, and portable LED light sources may make possible a low-cost and efficient deployment of systems to be used for wound decontamination.