Unless otherwise specified, all materials were used as received from commercial suppliers. The reagents were chemically or analytically pure. ¹H and ¹³C NMR spectra were tested on a BRUCKER AVANCE 500 NMR spectrometer. Tetramethylsilane (TMS) is the internal standard. High resolution mass spectrometers (HRMS) were tested on Agilent 6,550 Q TOF and Thermo Fisher-QE. Compound 1 (5,10,15,20-tetrakis(4-nitrophenyl)porphyrin) and compound 2 (5,10,15,20-tetrakis(4-aminophenyl)porphyrin) were prepared as described in the literature (Semeikin et al., 1982).

The general procedure for the synthesis of compounds 4a–d is as follows: A mixture of acid Boc-Orn(Boc)-OH (1.00 equiv), HATU (2.00 equiv), and N, N-Diisopropylethylamine (4.00 equiv) in N,N-Dimethylformamide (5 ml) was stirred at room temperature for 30 min. Then, 5,10,15,20-Tetrakis(4-aminophenyl)-21H,23H-porphine (0.125 equiv-0.40 equiv) was added to the acid mixture and the solution was stirred overnight at room temperature. The reaction was diluted with water (10 ml) and extracted with dichloromethane (3 × 10 ml). The combined organic layers were washed with brine, dried with anhydrous sodium sulfate, and concentrated. Purification with column chromatography on silica gel (200–300 mesh) gave the compounds 3a–d (yield 47–80%). Trifluoroacetic acid (TFA/CH₂Cl₂, 1/1, v/v) was added to the solution of compounds 3a–d (1 equiv) in dried CH₂Cl₂. The reaction was stirred for 30 min at room temperature and concentrated. Into the residue was added ethyl ether (10 ml) and a green solid was obtained by filtration. The solid was suspended in water (5 ml) and neutralized with ammonium hydroxide. The purple solid (4a–d) was obtained by filtration and washed with water (3 × 5 ml) and CH₂Cl₂ (3 × 5 ml; yield 54–83% for the two steps). The NMR and HRMS spectra of compounds 4a–d are available in the Supplementary material.

The photosensitizers were dissolved in dimethyl sulfoxide or phosphate buffered saline (PBS) to get the stock solution (500 mM). Light of 650 nm was outputted via an optic fiber from a semiconductor laser (WSLS-650/532-500 m-200 M-H2, Wavespectrum, China). The transmitted energy was calibrated with an optical power meter (LM1; Carl Zeiss). The irradiance of 3.33 mW/cm² and doses of 0, 2, 4, 6, 8, 10, and 12 J/cm² were used in the in vitro experiments while 100 mW/cm² and 50 J/cm² were used in vivo.

The MRPM used in the study was isolated from diabetic foot wounds of clinical patients at the First Teaching Hospital of Tianjin University of Traditional Chinese Medicine in 2016. The bacterial strain was inoculated in Luria Bertani (LB) liquid medium (10 ml) and cultured in a shaker at 37°C until the logarithmic growth phase. Then the bacterial solution was inoculated on LB solid medium and incubated overnight at 37°C. A single colony was selected and inoculated in a 5 ml liquid medium and cultured overnight (37°C, 220 rpm). After centrifugation, the concentration of bacteria was adjusted to 1 × 10⁸ CFU/ml and stored at 4°C.

We determined the antimicrobial susceptibility profile of the MRPM strain by following the methods in Clinical and Laboratory Standards Institute documents M07 and M100 (CLSI M07-A10 and M100-ED32). The broth macrodilution method was used to obtain the minimal inhibitory concentration (MIC) of antibiotics. The antimicrobial agents were diluted to different concentrations with Mueller-Hinton broth culture medium in aseptic test tubes, and then the bacteria were added to the test tube with a final concentration of 10⁵ CFU/ml. After inoculation, the test tube was placed in an incubator at 37°C for 24 h. The minimal concentration of the antimicrobial agents making bacterial suspension visibly change from turbid to clear was the MIC.

The standard curves were prepared as follows: The photosensitizers were dissolved in 0.1 M-1% SDS. The final concentrations were 31.25, 15.62, 7.81, 3.91, 1.95, 0.98, 0.49, and 0.24 μM. Then, 100 μl of the solution was added to a 96-well culture plate and four replicate wells were prepared for each concentration. The fluorescence intensity of each well was measured at an excitation wavelength of 418 nm and an emission wavelength of 658 nm. The standard curve was plotted according to the fluorescence intensity and concentrations.

Next, 1 ml of bacterial suspension in the logarithmic growth phase was centrifuged (9,000 rpm, 2 min) and then the supernatant was decanted, washed once with PBS, and resuspended in PBS to the corresponding optical density (OD₆₀₀ = 0.6–0.8). Then, 10 μM of each compound, which was dissolved in PBS, was added and incubated for 0 min, 10 min, 20 min, 30 min, 40 min, 80 min, 160 min, and 320 min in the dark. The suspensions were centrifuged (9,000 rpm, 2 min) and washed with PBS. The bacteria were treated with 1% SDS for 24 h. The fluorescence intensity was measured. The amount of photosensitizer uptake by bacteria was calculated according to the standard curve. The change of the uptake amount along varying concentrations (0–31.25 μM) was measured using the same method, only the incubation time was 40 min. Three sets of independent experiments were performed.

Measures of 100 μl of MRPM suspension (OD₆₀₀ = 0.1) and 100 μl of the photosensitizers (the final concentration = 15.62 μM) were loaded into a 96-well plate. Plates were kept in the dark for 40 min and then illuminated with a dose of light (0, 2, 4, 6, 8, 10, and 12 J/cm²) and a wavelength of 650 nm. After illumination, 100 μl of the aliquots were taken from each well and pipetted into a tube containing 900 μl of sterilized distilled water to make a 10⁻¹ dilution. In addition, 10⁻², 10⁻³, 10⁻⁴, 10⁻⁵, and 10⁻⁶ dilutions were prepared using the same method. Each dilution (100 μl) was loaded onto LB agar plates for culture. The colony-forming units (CFU) were statistically analyzed and the bacterial survival fractions were calculated according to the ratios of CFU of the administration groups and control group. Three sets of independent experiments were performed.

In a series of test tubes, MRPM suspension (500 μl) and photosensitizers (500 μl) were added. The final concentrations of photosensitizers and MRPM were 1.96, 3.91, 7.81, 15.62, 31.25, 62.50, 125, 250, and 500 μM and 10⁶ CFU/ml. The mixture was incubated in a shaker (180 rpm) for 40 min in the dark and then irradiated by laser (650 nm, 8 J/cm²). After being cultured in the dark for 24 h, the degree of turbidity of the bacterial solution was observed. Then, 100 μl samples from the tubes without turbidity were added to LB agar plates and incubated at 37°C for 24 h. The minimum bactericidal concentration (MBC) was determined when no more than five colonies were observed on the plate. Three sets of independent experiments were performed.

The bacteria (OD₆₀₀ = 0.1) and photosensitizers (0, 1.95, 3.91, 7.81, 15.62, and 31.25 μM) were incubated in the dark at 37°C for 40 min. The suspensions were then illuminated (650 nm, 8 J/cm²). The remaining procedure is the same as that detailed in the section “The effects of irradiation energy density on PACT.” Three sets of independent experiments were performed.

The wound healing effect was evaluated using BALB/C mice. The animal experimental protocols were all approved by the Animal Ethical and Welfare Committee of Tianjin Medical University Cancer Institute and Hospital. The weight of BALB/C mice from Beijing HFK Bioscience Co., Ltd. was 20–22 g. Mice in each group were anesthetized via intraperitoneal injection of 10% chloral hydrate (300 mg/kg). After local skin preparation on the back, the skin was incised in its entirety to the deep fascia and a 1.0 cm diameter circular incision was created. Then the suspension (50 μl) of MRPM (2 × 10⁹ CFU/ml) in PBS was inoculated onto the surface of the wound and then smeared with an inoculating loop. Next, gauze was used to bandage the wound. After 48 h, when the wound infection model was successful, it was used in the next experiment.

Four groups including 24 mice with wound infections were involved in the experiments in vivo: (A) no treatment control; (B) light alone (50 J/cm²); (C) 4d-PACT; and (D) 4d alone (without illumination). The mice in groups C and D were treated with 50 μl of 4d (100 μM in 0.9% NaCl solution) which was injected into the wound. After 40 min in the dark, the mice wounds of groups B and C were irradiated with the light (650 nm, 50 J/cm²) while the other areas of the mice were shielded from light. The next day, the mice were irradiated again at the same dose of light without photosensitizer administration. The above process was defined as one PACT treatment. The PACT treatment was performed three times.

The day the mice were first treated with photosensitizers was considered day 1. On days 1, 2, 3, 4, 5, 6, 7, 8, 10, and 12, the wound area was calculated by measuring the length and width. The rate of wound healing of the mice was obtained by the following formula. Wound healing ratio = unhealed wound area/initial wound area. The viability of mice after treatment was observed and recorded. Meanwhile, the number of bacterial colonies in the wound at different time points was counted according to the following methods. A sterile cotton swab was dipped into normal saline and evenly smeared the whole wound from one side to the other side, and immediately put into an aseptic tube. During the detection, the ratio of normal saline was diluted to 10⁻⁶, and then 100 μl of the bacterial solution was loaded onto LB agar plates for culture. The colony morphology was observed and counted.

Data were analyzed by using SPSS 19.0 and were expressed as mean ± standard error of mean (SEM) or mean ± standard deviation (SD). Significant differences between sample means were determined by least significant difference t-tests or χ² test.

New cationic photosensitizers against MRPM were designed and synthesized by the condensation of 5,10,15,20-tetrakis(4-aminophenyl)porphyrin (2) with Boc-protected ornithine as depicted in Scheme 1. The 5,10,15,20-tetrakis(4-aminophenyl)porphyrin (2) was prepared according to the method in the literature (Semeikin et al., 1982), and 3a–d were synthesized by condensation of 5,10,15,20-tetrakis(4-aminophenyl)porphyrin with Boc-Orn(Boc)-OH catalyzed by HATU/DIPEA at room temperature. By adjusting the ratio of reactants, ornithine-porphyrin conjugates containing one to four ornithine side chains were successfully synthesized. Then, by using trifluoroacetic acid (TFA), the t-butyloxy carbonyl group was removed to get photosensitizers 4a–d.

Energy dose is a main parameter essential for the therapeutic effect of photosensitizers. We evaluated the effect of laser energy density on photoinactivation abilities. MRPM suspensions containing compound 4c or 4d and the control group were irradiated with different energy densities. Laser irradiation alone had no significant inactivation efficiency against MRPM in the control group. The killing effect of photosensitizers increased with the increase of laser energy density, and it tended to be stable when the energy reached 8 J/cm² (Figure 1). Moreover, based on our previous study on porphyrin derivatives (Meng et al., 2015; Xu et al., 2016; Rong et al., 2021; Yin et al., 2022), the best wavelength for treatment was 650 nm, which has strong penetration in tissue. Therefore, the condition of 650 nm, 8 J/cm² was chosen for the subsequent experiments.

The amount of photosensitizer entering the bacteria has an important influence on the PACT efficacy. Therefore, the uptake experiment was performed using the fluorescence spectrophotometry method (Figure 2). Figure 2A shows that the uptake amount of photosensitizers by MRPM increased with the increasing concentration. Photosensitizer 4d, linking four ornithine groups in porphyrin, showed the highest affinity for MRPM. It took about 40 min when the uptake amount of photosensitizers by MRPM reached the maximum and then entered the plateau period (Figure 2B). Therefore, in the following evaluation of antibacterial activity, 40 min was selected as the time point of incubation, that is, bacteria and photosensitizers were incubated for 40 min in the dark and then irradiated.

The MIC and MBC of ornithine-porphyrin photosensitizers were determined to evaluate the photodynamic antibacterial activity against MRPM preliminary. The MRPM used in this experiment was isolated from the clinic. The antimicrobial susceptibility profile of the MRPM strain showed that it was resistant to a variety of antibiotics (Table 1). However, the MIC and MBC experiments demonstrated that photosensitizers 4c and 4d exhibited an inhibitory effect at concentrations of 15.62 μM and 7.81 μM under laser irradiation conditions (Table 2) when their dark toxicities were relatively weak (MIC ≥500 μM).

To further investigate the photodynamic bactericidal activities against MRPM of the ornithine-porphyrin photosensitizers, the dose-dependent photoinactivation effects were evaluated using the screened conditions. Generally, the MRPM suspension was incubated with the photosensitizers in the dark at 37°C for 40 min. The final concentrations of the photosensitizers included 0, 1.95, 3.91, 7.81, 15.62, and 31.25 μM and the irradiation energy was 8 J/cm². Figure 3 shows the photoinactivation effects and dark toxicity against MRPM differed among the photosensitizers. The photodynamic inactivation effect against bacteria by photosensitizers was significantly dose-dependent, which increased with the increase of concentration (Figure 3A). When the dark toxicity of all compounds was very low, this indicated the photosensitizers had no bactericidal effect without irradiation (Figure 3B).

The cationic photosensitizers killed gram-negative bacteria more efficiently than neutral and negatively charged analogs by damaging the membrane, which is due to the ionic interaction between the cationic groups on photosensitizers and the negatively charged lipopolysaccharides or peptidoglycans on the bacterial cell wall. Therefore, the membrane integrity of ornithine-porphyrin conjugates was investigated by fluorescence microscopic imaging and measuring the release of intracellular components. Since compounds 4a and 4b both have weak activities, one of the compounds, 4b, was chosen as a representative to be investigated.

Figure 4 shows the fluorescence microscopy imaging of MRPM including the control (A), 4b-PACT (B), 4c-PACT (C), and 4d-PACT (D) groups. Acridine orange (AO) can penetrate the intact cell membrane and intercalate into DNA to make it emit bright green fluorescence. Ethidium bromide (EB) can only penetrate the damaged cell membrane, embed into DNA, and emit orange-red fluorescence. The fluorescence image of the control group exhibited dispersed green fluorescent particles, indicating the bacteria were alive and the membrane was not damaged. The 4b-PACT group had green and little red fluorescent particles, indicating the cell membrane of a small number of bacteria was ruptured. Most of the bacteria of 4c-PACT and 4d-PACT were dead with a red color due to the damage to the membrane. The results showed that the photosensitizers could destroy cell membranes to kill bacteria and the PACT effect of 4c and 4d was stronger than that of 4b, which may be related to their physical and chemical properties and affinity to bacteria.

In addition, the destruction of bacteria membrane integrity could result in intracellular components including small ions and large molecules being released from bacteria. Large molecules such as DNA and RNA have strong UV absorption at the wavelength of 260 nm. Hence, the change of absorbance at the wavelength of 260 nm can reveal the integrity of the bacterial membrane (Sahu et al., 2009). After treatment with PS-PACT, the bacterial content exudation of the 4c group and 4d group were increased by 116.65, and 356.67%, respectively, while other groups including the PBS treatment group, light alone (8 J/cm²), and photosensitizer alone group changed from 2.01 to 9.31% (Table 3). These results demonstrated that 4c-PACT and 4d-PACT could cause disruption of bacterial cell membranes and changes in the permeability of MRPM.

The antibacterial effect of 4d-PACT in vivo was evaluated using a MRPM-infected wound mouse model. After the establishment of the model, the mice were randomly divided into four groups: no treatment control; light alone (50 J/cm²); 4d-PACT; and 4d alone (without illumination). The mice in the 4d-PACT group were treated with photosensitizers and irradiated. On the next day, the mice were irradiated again at the same dose of light without photosensitizer administration. The day the mice were first treated was considered day 1 (Figure 5A). The results showed that 4d-PACT was effective in the treatment of wounds infected by MRPM (Figure 5B). Compared with the control group with no treatment, the mice of the 4d-PACT treatment group showed rapid wound healing, especially between days 3 and 8 (p < 0.05). In the light alone (50 J/cm²) and 4d alone groups, rapid wound healing happened from day 5 to 8 and day 7 to 10, respectively, which was later than that of the 4d-PACT treatment group (Figure 6). The mean wound area ratios on day 8 of the control group, 4d-PACT treatment group, light alone group, and dark toxicity group were 0.68 ± 0.08, 0.16 ± 0.02, 0.23 ± 0.07, and 0.42 ± 0.05, respectively. The wound healed almost completely on day 12, although wound healing slowed down after day 8. These results indicated photosensitizer 4d had better photodynamic efficiency on MRPM-infected wound healing.

The bactericidal effect to MRPM of 4d-PACT was evaluated by counting the viable bacteria in wound tissue on days 1, 2, 3, 4, 5, 6, 7, and 8 (Figure 7). Compared with the other three groups, the 4d-PACT treatment group showed a significant decrease in the number of bacteria. However, there was little decrease in bacteria in the dark toxicity group and 50 J/cm² laser irradiation group, which indicated that 4d-PACT accelerated wound healing by inhibiting the proliferation of MRPM.

Two mice died on day 2 and one died on day 3 in the control group, while a total of two mice died in the light alone (50 J/cm²) group, and a total of two mice also died in the 4d alone (dark toxicity) group. The mice in the 4d-PACT group all survived during the experiment, which indicated that 4d-PACT was more efficient in the treatment of MRPM infection (Figure 8). The change in the body weight of the mice is shown in Figure 9. On the whole, the mice had varying degrees of weight gain over 12 days, while there was a small decrease in body weight in the first 2 days after the establishment of the wound bacterial infection model. Compared with the control group, the mice in the 4d-PACT treatment group lost less weight in the first 2 days and gained more rapidly in the later stage. This showed that the body recovery of mice treated with 4d-PACT was faster than that of the other three groups, which was consistent with the results of wound healing rate and wound colony count.