The solid samples of 0.08 ICG (0.33%; pure <98%; Frontier Scientific™) and 0.08 mg/mL^-1MtB (0.33%; pure <98%, Valerus™), positive control, were mixed in distillated water (di-Milli-Q), aluminum foil covered and stored in dark at 22°C after 5 min ultra-sonication to avoid aggregation. The chosen concentrations of 0.08 mg/mL^-1 for ICG and MtB are commonly utilized for clinical PDT activation in the oral cavity.

The laser source (PDT) was a diode array laser from D-touch™, Syneron Lasers (Israel) emitting at 630–810 nm. The nominal energy was 0.1–0.5 W. The 0.08 mg/mL^-1ICG exposures was performed under wavelength of 810 nm, average power: 500 mW, beam diameter: 3.0 cm, and power density: 134 J/cm². The 0.08 mg/mL^-1MtB exposure was performed under wavelength of 630 nm laser light (SIX Laser TSC™, Atlantis Lasers, Bulgaria), average power: at 100 mW, beam diameter: 3.0 cm, and power density: 15 J/cm². The used laser tip was 400-micron fiber. Each solution sample was irradiated 60 seconds/dark at peak-to-peak power fluctuation (<0.2%) wavelengths.

The EPR analyses (Bruker, X-band-EMXmicro) were employed to detect ROS and RNS generated radicals during the PS agent treatment at 23°C.

A spin-trapping agent, 2,2’-azobis-2-methyl-propanaimidamide dichloride (AAPH, >97%) dissolved in phosphate buffered saline (PBS) (pH=7.4), at a 10 mM (100 µL) concentration was used directly to generate alkyl radicals in mixed with 0.08 mg/mL^-1ICG, 0.08 mg/mL^-1MtB; 0.08 mg/mL^-1ICG + PDT and 0.08 mg/mL^-1MtB + PDT combinations, by stirring at 23°C, in either aerobic conditions. Then, 60 µL, 0.1 mM N-tert-butyl-a-phenylnitrone (PBN) was added to the mixtures. After incubating at 40°C for 5 minutes in a water bath, the sample was examined triplicate at different time intervals, at 1, 3, 30, and 60 min, by center field 3513G, microwave power 2.05 mW, modulation amplitude 10 G, five scans per sample.^[13]

4-hydroxy-TEMPO (TEMPOL, >97%) dissolved in phosphate buffered saline (PBS) (pH=7.4), at a concentration of 0.2 mM was used directly to generate •O₂ radicals in mixed with 0.08 mg/mL^-1ICG, 0.08 mg/mL^-1MtB and in combinations 0.08 mg/mL^-1ICG+PDT, 0.08 mg/mL^-1MtB+PDT, by stirring at 23°C, in either aerobic conditions. Then, 50 µL, 0.2 mM TEMPOL was added to the mixtures. After incubating at 40°C for 5 minutes in a water bath, the sample was examined triplicate at different time intervals, at 1, 3, and 5 min, by center field 3513G, microwave power 2.05 mW, modulation amplitude 10 G.^[14] The PS effect in vitro was evaluated by the equation:

Scavenged TEMPOL/ O₂⁻ = [I/Io]×100%,

where: Io – a double integrated plot of the TEMPOL/·O₂⁻ adduct registered in the control; I – the double integrated plot of the TEMPOL/·O₂⁻ spin adduct registered in the tested sample.

A spin-trapping agent, 50 µL, 0.08 mM 1,3-diphenylisobenzofuran (DPBF), dissolved in 2 mL ethanol was used directly to generate singlet oxygen (¹O₂) in mixtures with 0.08 mg/mL^-1ICG, 0.08 mg/mL^-1MtB and in combinations 0.08 mg/mL^-1ICG+PDT, 0.08 mg/mL^-1MtB+PDT, by stirring at 23°C, in either aerobic conditions. The decrease in spectra intensity at the 630-810 nm wavelength corresponds to the singlet oxygen ¹O₂ interaction with DPBF.^[15]

The remaining statistical analyses were performed using Statistica v. 7.0, (StaSoft, Inc., USA) and the results are given as mean ± standard error (SE). The EPR spectral processing was performed using Win-EPR and Simfonia software as averages of three replicates. Statistical analysis was performed using a one-way ANOVA and the Student t-test to determine differences, and p<0.05 value was considered statistically significant.

Laser non-invasive application produces ¹O₂, •O₂⁻, H₂O₂, and •OH radicals, leading to the destruction of pathogenic microbes and cancer cells.^{[1, 19]} Girotti^[19] commented on the fact, subjected to photooxidative stress, unsaturated membrane lipids are directly attacked by ¹O₂ or by reduced species in cells, i.e. the lipids are directly responsible for the PS amphiphilicity and its localization in the membrane bilayer. Subsequent reactions of oxidation potentially alter the structure and efficiency of the bound proteins, nucleic acids, and other active molecules. Non-invasive laser therapy activates membrane oxidation processes and re-activates various oxidative pathways, photooxidative stress also. This possibility highlights the need for PS to neutralize photooxidative stress.^{[1, 13, 19]}

The alkyl radicals scavenging capacity after the PDT application was verified by EPR spectroscopy in the presence of the AAPH-PBN spin-probe (Fig. 2).

The samples containing MtB showed significant signal minimization in both probes, before (21.5%) and after PDT activation (27.1%); i.e. MtB minimally reduces AAPH-PBN-induced alkyl radicals and photooxidative stress, before PDT, and fails to suppress oxidative changes after laser activation. Before laser activation, the signal observed for ICG probe was stable (g=2.0054±0.0001), within 30% to 42.7%, after 30-60 minute incubation period (Fig. 2). As shown in Fig. 2, the laser activation of ICG probe proceeded with no changes in the recorded g-factor=2.0054±0.0001, showing significant AAPH-PBN scavenging and AAPH-PBN reduction, and a stable alkyl adduct signal (the spectra are not shown). Samples containing ICG after 810 nm activation did not show a typical sextet signal, proving that ICG competitively scavenges and neutralizes AAPH-generated alkyl radicals. Compared to the initial AAPH-PBN as control (100%) and the laser induction (69.4%), the maximum detected AAPH-PBN-spin adducts were in the range of 68.4% for ICG, after 810 nm. Therefore, the photodynamic activation of ICG showed high reproducibility, especially after incubation lasting 30 minutes, followed by a sharp reduction of laser-activated oxidative changes, in vitro. Consistent with our observations, Alander et al.^[21] commented that the ICG decomposition is due to ¹O₂ and the ¹O₂ is immediately bound to the decomposition ICG-itself products; i.e. ICG with respect to clinical application can be used without photo-toxicity worry and ¹O₂ over-production. In this regard, C-centered alkyl radicals combined with oxygen can easily transform into alkyl radicals, and AAPH-PBN or the alkyl spin-adduct, proves the C-centered alkyl radical formation in the laser activation^{[22, 23]}, and the rapid alkyl radical reduction from ICG. Notably, ICG modulates the accumulated alkyl radicals and photooxidative stress, associated with the laser induction, and the maximum being directly dependent on time, at 30 minutes.

Exogenous ROS production, in particular the oxygen-centered radicals •O₂⁻ and •OH, damage biomolecules and induces inflammatory diseases through various oxidative mechanisms. PDT involves the ROS generation in the target tissue through a combination of O₂, light, and PS agents. The photosensitizers absorb laser activation (650 nm-850 nm) and transfers electrons or electronic energy through two reaction mechanisms to produce ROS.^[24] Types I and II reaction mechanisms involve charge transfer from the photosensitizer to the oxygen molecule, generating primarily ROS, as superoxide (•O₂⁻) or hydroxyl (•OH) radicals formation, or singlet oxygen (¹O₂) activation. The local photooxidative stress, produced by laser, produces ROS (in particular •O₂⁻ radicals, i.e. activated spontaneous type I mechanism) in the illuminated tissue, while sparing normal cells.^[24–26] The photosensitizers such as hydro-ICG are charged and membrane-impermeable molecules, and their physical characteristics are suitable for measuring in vitro extracellular ROS production.^{[24, 25]}

Firstly, we investigated the ability of ICG and MtB to quantify the •O₂⁻ and •OH radical in vitro (i.e. the probability of crossing over the type I) under PDT/ laser accumulation, and its sensitivity was compared against spin-trapping agents (Fig. 3).

The •O₂⁻ production formed on 810 nm laser accumulation at 25°C of both photosensitizers was evaluated by using a highly selective method that involves the reduction to stable nitroxide radical, TEMPOL, which is easily detectable by EPR.^[14] As expected, ICG solution containing TEMPOL resulted in the formation of an almost symmetrical, nearly equal-intensity three-lined spectrum, with relative peak-to-peak ratio of 1:2:1, evidencing the free radical formation of the TEMPOL – •O₂⁻, with 16.3 G hyperfine splitting constants (spectra are not present). ICG solution under 810 nm activation increased intensity of the triplet signal to 67.23%, at 1 minute incubation. The maximum detected TEMPOL spin-adducts as inhibited •O₂⁻ production in the range of 70.92 %, show high reproducibility, at 3 minutes, and the results are comparable to laser non-activated ICG(Figs 2a, 2b). However, there were no significant differences between the MtB solution in both before and under laser illumination, on the 1-5 min incubation. The MtB solution showed weak signal and spectra minimization, confirmed the lower detecting accuracy to •O₂⁻. In addition, it indicated that the laser induction possessed the higher detecting accuracy of ICG solution and a wider range of •O₂⁻ concentration (p<0.05), comparing to MtB solution. Martins et al.^[27] noted that molecular O₂ in aerobic conditions of samples preparation strongly affected on the formation of TEMPOL spin-adducts and additional •O₂⁻ production. Therefore, we could conclude that at a short laser activation time all the produced •O₂⁻ is completely consumed by ICG and residual photooxidative stress is suppressed.^[28] In addition, it is reasonable to believe that in complex biosystems, laser-activated ICG will reduce the overproduction of oxygen-centered radicals and possibly other pro-oxidants that may disable the therapeutic penetration and enhanced antibacterial activity of ICG+aPDT^[28], but also generally reduce oxidative stress pathways.

The DPBF photooxidation, involves different reactive oxygen formation, as bi-radicals, oxy-radicals and peroxy-radicals, and DPBF with PS interactions should be considered^[15] through free radical deposition-induced mechanisms. Fig. 4 shows the ICG and MtB sensitivity towards singlet oxygen (¹O₂) production assessed by using DPBF spin-trap, before and under laser activation (Figs 4а, 4b), in vitro. The ¹O₂ concentration, scavenged by ICG was almost proportional to 810 nm PDT activation. The stable intensity of ICG spectra (spectra are not shown) under 810 nm laser activation, and the significant decreasing ¹O₂ concentration (87.04 % at 3 minutes) show that ICG possessed a high sensitivity and stable reduction towards ¹O₂. MtB alone, and MtB + laser activation showed significantly weak spectra signal and spectra minimization, in both, before and under laser illumination, on the 1-5 min (Figs 4a, 4b). The maximum ¹O₂ scavenging activity at MtB alone, and after photo-activation, were at region, 21.02% and 29.05% at 3 minutes, respectively. The results show that 0.08 mg/mL^-1ICG concentration combined with 810 nm laser irradiation has a wider detection range of ¹O₂ than MtB + laser activation, even at significantly low free-radical concentrations. Therefore, the photoactivated ICG could function as singlet oxygen ¹O₂ probes (p<0.05), stops the formation of different oxy-, per-oxy, hydro-peroxy radicals; hydroxy polymer radical^{[15, 27, 29, 30]} and directly converts DPBF spin-adducts (Figs 4a, 4b). The photoactivated ICG work more accurately to stop the photo-oxidative degradation initiated by ROS compared to no photosensitizing MtB. In contrast, it is possible low light illumination to formatted ¹O₂ as dominant process, easily to convert into •O₂^−[30] that reduces the formation of both, DPBF spin-adducts and PBN spin-adducts. Probably, our experiment supports the theory of oxygen-centered radical and ¹O₂ destruction, in vitro. Consistent with our fundings, Montazerabadi et al.^[2] confirmed that when treating MCF-7 human breast cancer cells line in the dark, no effect of ICG was confirmed, while increased cell survival occurred upon ICG photoactivation. As discussed previously^{[27, 30, 31]}, the reactive oxygen species reduction supports the therapeutic use in the context of PDT to control tissue growth or promoting tumor tissue cell death.

Therefore, ICG as an electron-excited PS at 810 nm, localized in a lipid medium, directly binds and interacts with O₂ molecules, produces •O₂– radicals and ¹O₂ radicals, but at additional cellular mechanism neutralizes the residual production of alkoxyl and peroxyl radicals. In addition, ICG binds rapidly to plasma proteins (lipoproteins) without altering protein structures, which explains protein non-toxicity and immediate •O₂– radicals and ¹O₂ uptake.^{[2, 21, 31, 32]} On the other hand, it has been commented that ICG changes its molecular structure – forms aggregates, whose absorption properties vary depending on light intensities, temperature and dissolution time in a solvent.‌^{[31, 32]} We assume that the photoactivated ICG achieved in vitro is under ideal conditions for the reaction system. We proposed that non-fading of ICG staining after 810 nm illumination, and maximum ROS scavenging effect occurs without structural changes in the PS molecule, without the aggregates formation and at optimal temperature.