Sulfide (Na2S) and Polysulfide (Na2S2) Interacting with Doxycycline Produce/Scavenge Superoxide and Hydroxyl Radicals and Induce/Inhibit DNA Cleavage

Doxycycline (DOXY) is an antibiotic routinely prescribed in human and veterinary medicine for antibacterial treatment, but it has also numerous side effects that include oxidative stress, inflammation, cancer or hypoxia-induced injury. Endogenously produced hydrogen sulfide (H2S) and polysulfides affect similar biological processes, in which reactive oxygen species (ROS) play a role. Herein, we have studied the interaction of DOXY with H2S (Na2S) or polysulfides (Na2S2, Na2S3 and Na2S4) to gain insights into the biological effects of intermediates/products that they generate. To achieve this, UV-VIS, EPR spectroscopy and plasmid DNA (pDNA) cleavage assay were employed. Na2S or Na2S2 in a mixture with DOXY, depending on ratio, concentration and time, displayed bell-shape kinetics in terms of producing/scavenging superoxide and hydroxyl radicals and decomposing hydrogen peroxide. In contrast, the effects of individual compounds (except for Na2S2) were hardly observable. In addition, DOXY, as well as oxytetracycline and tetracycline, interacting with Na2S or other studied polysulfides reduced the •cPTIO radical. Tetracyclines induced pDNA cleavage in the presence of Na2S. Interestingly, they inhibited pDNA cleavage induced by other polysulfides. In conclusion, sulfide and polysulfides interacting with tetracyclines produce/scavenge free radicals, indicating a consequence for free radical biology under conditions of ROS production and tetracyclines administration.


Introduction
Endogenously produced hydrogen sulfide (H 2 S) and polysulfides (H 2 S n ) affect many physiological and pathological processes, such as hypertension, atherosclerosis, heart failure, diabetes, inflammation, asthma, burn injuries, sepsis, angiogenesis, cancer, and neurodegenerative diseases. H 2 S and H 2 S n have the beneficial effects under conditions of oxidative stress by reacting with reactive To demonstrate how the time-dependent degradation of polysulfides influences their biological activities, we studied their ability to reduce the • cPTIO radical. The potency of Na 2 S 2 , Na 2 S 3 and Na 2 S 4 to reduce • cPTIO was strongly time-dependent. Incubation of 100 µM Na 2 S 2 , Na 2 S 3 or Na 2 S 4 for 15 s followed by addition of 100 µM • cPTIO caused its fast reduction in <1 min.
When polysulfides were incubated for 20, 40 or 70 min prior to • cPTIO addition, their reducing properties markedly decreased over the time (Figure 3 and Figure S1). In case of Na 2 S 2 , they were even negligible after 70 min of incubation ( Figure 3A). Extreme time-dependent instability of polysulfides in aqueous solutions should be taken into careful consideration when handling polysulfide solutions and working with exogenously added polysulfides, particularly in setting requiring a long incubation time.
Molecules 2019, 24, x FOR PEER REVIEW 3 of 20 to reduce • cPTIO was strongly time-dependent. Incubation of 100 µ M Na2S2, Na2S3 or Na2S4 for 15 s followed by addition of 100 µ M • cPTIO caused its fast reduction in < 1 min.
When polysulfides were incubated for 20, 40 or 70 min prior to • cPTIO addition, their reducing properties markedly decreased over the time (Figures 3 and S1). In case of Na2S2, they were even negligible after 70 min of incubation ( Figure 3A). Extreme time-dependent instability of polysulfides in aqueous solutions should be taken into careful consideration when handling polysulfide solutions and working with exogenously added polysulfides, particularly in setting requiring a long incubation time. . Spectra were recorded every 30 s for 40 min. The first spectrum was recorded 15 s after thawing of 10 mM Na2S2 stock. Arrow indicates decrease of ABS at 280 and 300 nm. The first spectrum is indicated by the solid red line, which is followed each 30 s by long dash red, medium dash red, short dash red, dotted red, full blue line, long dash blue, medium dash blue, etc. Insets: Kinetics of changes in ABS at 230 nm (black), 300 nm (red), 600 nm (blue) and 900 nm (green). Representative time-dependent UV-VIS spectra of 100 µM Na 2 S 2 in 100 mM sodium phosphate, 100 µM diethylenetriaminepentaacetic acid (DTPA), pH 7.4, at 37 • C. Spectra were recorded every 30 s for 40 min. The first spectrum was recorded 15 s after thawing of 10 mM Na 2 S 2 stock. Arrow indicates decrease of ABS at 280 and 300 nm. The first spectrum is indicated by the solid red line, which is followed each 30 s by long dash red, medium dash red, short dash red, dotted red, full blue line, long dash blue, medium dash blue, etc. Insets: Kinetics of changes in ABS at 230 nm (black), 300 nm (red), 600 nm (blue) and 900 nm (green).
Molecules 2019, 24, x FOR PEER REVIEW 3 of 20 to reduce • cPTIO was strongly time-dependent. Incubation of 100 µ M Na2S2, Na2S3 or Na2S4 for 15 s followed by addition of 100 µ M • cPTIO caused its fast reduction in < 1 min.
When polysulfides were incubated for 20, 40 or 70 min prior to • cPTIO addition, their reducing properties markedly decreased over the time (Figures 3 and S1). In case of Na2S2, they were even negligible after 70 min of incubation ( Figure 3A). Extreme time-dependent instability of polysulfides in aqueous solutions should be taken into careful consideration when handling polysulfide solutions and working with exogenously added polysulfides, particularly in setting requiring a long incubation time. . Spectra were recorded every 30 s for 40 min. The first spectrum was recorded 15 s after thawing of 10 mM Na2S2 stock. Arrow indicates decrease of ABS at 280 and 300 nm. The first spectrum is indicated by the solid red line, which is followed each 30 s by long dash red, medium dash red, short dash red, dotted red, full blue line, long dash blue, medium dash blue, etc. Insets: Kinetics of changes in ABS at 230 nm (black), 300 nm (red), 600 nm (blue) and 900 nm (green).

Formation of the O2 •− and • OH Radicals by Na2S or Na2S2 Interacting with DOXY
Since DOXY was reported to modulate in vivo biological conditions that involve ROS [24,31,32,37], we studied its interaction with radicals using EPR spectroscopy. In control spin trap experiments, 5-tert-butoxycarbonyl-5-methyl-1-pyrroline-N-oxide (BMPO) did not have EPR spectrum in the buffer solution during 11 min of observation ( Figure 4A1,A2). Minor intensity of EPR spectra (mostly • BMPO-OOH/OH, i.e., O2 •− and • OH were trapped by BMPO) were observed when Representative time-dependent UV-VIS spectra of 100 µM Na 2 S 3 (A) and 100 µM Na 2 S 4 (B) in 100 mM sodium phosphate, 100 µM DTPA, pH 7.4, at 37 • C. Spectra were recorded every 30 s for 40 min. The first spectrum was recorded 15 s after thawing of 10 mM Na 2 S 3 or 10 mM Na 2 S 4 stocks. Arrows indicate decrease of ABS at 280 nm and increase at 500 nm. For details on colors, see Figure 1. Insets: Kinetics of changes in ABS at 230 nm (black), 300 nm (red), 600 nm (green) and 900 nm (cyan). After 40 min incubation of Na 2 S 3 or Na 2 S 4 samples (A,B), 100 µM • cPTIO was added and the spectra were recorded for further 40 min. Kinetics of decrease of ABS at 900 nm after addition of 100 µM • cPTIO (time zero) to 40 min incubated Na 2 S 3 or Na 2 S 4 samples are shown as Insets-blue lines.

Formation of the O 2
•− and • OH Radicals by Na 2 S or Na 2 S 2 Interacting with DOXY Since DOXY was reported to modulate in vivo biological conditions that involve ROS [24,31,32,37], we studied its interaction with radicals using EPR spectroscopy. In control spin trap experiments, 5-tert-butoxycarbonyl-5-methyl-1-pyrroline-N-oxide (BMPO) did not have EPR spectrum in the buffer solution during 11 min of observation ( Figure 4A1,A2). Minor intensity of EPR spectra (mostly • BMPO-OOH/OH, i.e., O 2 •− and • OH were trapped by BMPO) were observed when Na 2 S (500 µM) was added to BMPO ( Figure 4B1,B2). This is in agreement with our previous study, in which 2 mM Na 2 S formed mostly the • BMPO-OOH/OH radicals [8]. EPR spectra of the • BMPO-adducts for DOXY (250 µM) were of negligible intensity only ( Figure 4C1,C2), indicating that virtually no radical is Na2S (500 µ M) was added to BMPO ( Figure 4B1, B2). This is in agreement with our previous study, in which 2 mM Na2S formed mostly the • BMPO-OOH/OH radicals [8]. EPR spectra of the • BMPOadducts for DOXY (250 µ M) were of negligible intensity only ( Figure 4C1,C2), indicating that virtually no radical is produced by DOXY. However, EPR spectra could clearly be observed and their intensity increased over the time in case of the Na2S/DOXY mixture. At constant concentration of DOXY (250 µ M), the spectral intensity of the • BMPO-adducts increased over the time with the increasing concentration of Na2S (100-500 µ M) ( Figure 4D1-F2), and the • BMPO-adducts spectra were stable for at least another 22 min ( Figure S2A1-B2). The results revealed formation of oxygen radicals during Na2S/DOXY interaction. To find out which radicals were trapped by BMPO, we simulated the accumulated spectra. The best fit was obtained when the hyperfine coupling constants for • BMPO-OH and • BMPO-OOH shown in Table 1 were used ( Figures 5A,B). The constants are similar to those reported by Zhao et al. [38]. The simulation confirmed that the O2 •− and • OH radicals, approximately at equimolar ratio, were formed during the Na2S/DOXY interaction ( Figure 5C). Since the first spectrum was recorded 110 ± The intensities of the time-dependent EPR spectra (A1-F1) and detailed spectra (A2-F2) are comparable, as they were measured under identical EPR settings.
To find out which radicals were trapped by BMPO, we simulated the accumulated spectra. The best fit was obtained when the hyperfine coupling constants for • BMPO-OH and • BMPO-OOH shown in Table 1 were used ( Figure 5A,B). The constants are similar to those reported by Zhao et al. [38]. The simulation confirmed that the O 2 •− and • OH radicals, approximately at equimolar ratio, were formed during the Na 2 S/DOXY interaction ( Figure 5C). Since the first spectrum was recorded 110 ± 15 s after sample preparation, we cannot exclude a possibility of trapping of other radicals with life-times shorter than 110 s.    Table 1 were used for simulation. Table 1. Hyperfine coupling constants of the BMPO spin adducts elucidated from the simulations of experimental spectra measured in the buffer solutions. • BMPO-OOH and • BMPO-OH were simulated based on two conformers.  Table 1 were used for simulation.

BMPO-adduct
The EPR spectrum of the • BMPO-adducts was not seen for 10, 25, 100 or 500 µM Na 2 S 2 ( Figure 6A1,A2 and Figure S2C1-E2). However, addition of DOXY to Na 2 S 2 caused a gradual increase of the intensity of EPR spectra of the • BMPO-adducts over the time ( Figure 6B1-D2). Concentration-dependent EPR spectra intensities were qualitatively different for the Na 2 S 2 /DOXY and Na 2 S/DOXY mixtures. At the constant DOXY concentration (250 µM), the intensities of the • BMPO-adducts were high at 10-25 µM Na 2 S 2 , decreased with increasing concentration of Na 2 S 2 , and diminished at 100 µM Na 2 S 2 ( Figure 6B1-E2). These results indicate that polysulfide Na 2 S 2 interacting with DOXY generates oxygen radicals in a bell-shaped manner; the radicals are produced at low Na 2 S 2 concentration (low Na 2 S 2 /DOXY molar ratio), but they are scavenged at higher Na 2 S 2 concentrations (higher Na 2 S 2 /DOXY molar ratio). Simulated spectra revealed that O 2 •− and • OH were formed by the Na 2 S 2 /DOXY mixture ( Figure 5E). The EPR spectrum of the • BMPO-adducts was not seen for 10, 25, 100 or 500 µ M Na2S2 (Figures 6A1, A2 and S2 C1-E2). However, addition of DOXY to Na2S2 caused a gradual increase of the intensity of EPR spectra of the • BMPO-adducts over the time ( Figures 6B1-D2). Concentrationdependent EPR spectra intensities were qualitatively different for the Na2S2/DOXY and Na2S/DOXY mixtures. At the constant DOXY concentration (250 µ M), the intensities of the • BMPO-adducts were high at 10-25 µ M Na2S2, decreased with increasing concentration of Na2S2, and diminished at 100 µ M Na2S2 ( Figures 6B1-E2). These results indicate that polysulfide Na2S2 interacting with DOXY generates oxygen radicals in a bell-shaped manner; the radicals are produced at low Na2S2 concentration (low Na2S2/DOXY molar ratio), but they are scavenged at higher Na2S2 concentrations (higher Na2S2/DOXY molar ratio). Simulated spectra revealed that O2 •− and • OH were formed by the Na2S2/DOXY mixture ( Figure 5E).

Formation of the O 2
•− and • OH Radicals by Na 2 S Interacting with DOXY in the Presence of Hydrogen Peroxide (H 2 O 2 ) EPR spectra of minor intensity were observed for 1 mM H 2 O 2 ( Figure 7A1,A2). Addition of 250 µM DOXY or 250-500 µM Na 2 S to 1 mM H 2 O 2 lead to no observable change in EPR spectra ( Figure 7B1-D2). However, the EPR intensity of the • BMPO-adducts notably increased, when the mixture of 250/250 or 500/250 µM/µM Na 2 S/DOXY was added to 1 mM H 2 O 2 ( Figure 7E1-F2).
Bell-shaped time-dependent spectra intensities indicate an initial radical production by Na 2 S/DOXY followed by scavenging of the • BMPO-adducts at the later times. The intensities of the time-dependent EPR spectra (A1-F1) and detailed spectra (A2-F2) are comparable, as they were measured under identical EPR settings.

Formation of the O 2
•− and • OH Radicals by Na 2 S 2 Interacting with DOXY in the Presence of H 2 O 2 Low intensity EPR spectra were observed for 1 mM H 2 O 2 after addition of 10 or 25 µM Na 2 S 2 ( Figure 8A1-B2). This intensity even decreased upon addition of higher Na 2 S 2 concentration (500 µM) ( Figure 8C). The results indicate that Na 2 S 2 decomposed H 2 O 2 forming • OH, which was trapped by BMPO. At high Na 2 S 2 concentration (500 µM), it likely scavenged the formed • OH and/or • BMPO-OH adducts. The intensity of the • BMPO-adducts noticeably increased, when the mixture of Na 2 S 2 /DOXY 10/250, 25/250, 50/250 and 100/250, but not 500/250 µM/µM, was added to 1 mM H 2 O 2 ( Figure 8D1-H2). This production of the radicals at low Na 2 S 2 /DOXY ratio and scavenging them at high(er) ratio indicates concentration-dependent bell-shaped production/inhibition of the radicals. Simulated spectra revealed that the O 2 •− and • OH radicals were produced by Na 2 S 2 /DOXY in the presence of H 2 O 2 ( Figure 5F).  (H1 and H2). The intensities of the time-dependent EPR spectra (A1-H1) and detailed spectra (A2-H2) are comparable, as they were measured under identical EPR settings.

Potency of Compounds to Reduce the • cPTIO Radical
In the previous section, we showed that Na 2 S and Na 2 S 2 interacting with DOXY possess different time-and concentration-dependent potency to produce and scavenge radicals. Since the Na 2 S/DOXY/H 2 O 2 and Na 2 S 2 /DOXY mixtures behave in the bell-shaped manner in these reactions, initially producing radicals followed by scavenging them, we next studied the potency of Na 2 S/DOXY and Na 2 S 2 /DOXY to reduce the • cPTIO radical as a model radical system. Decrease of • cPTIO ABS at 560 nm was used to measure reduction of the • cPTIO radical. We compared the effect of DOXY with other tetracycline derivatives, OXYT and TETR, and antibiotics, fusaric acid (FUSA) and norfloxacin (NORF), to reduce the • cPTIO radical alone and in combination with Na 2 S and polysulfides. For this purpose, we extended the set of studied polysulfides with Na 2 S 3 and Na 2 S 4 .
2.5.1. Tetracyclines, but neither FUSA nor NORF, Reduce the • cPTIO Radical in the Presence of Na 2 S Na 2 S (400 µM) reduced • cPTIO (100 µM) by <1% after 20 min ( Figure 9A). DOXY, OXYT and TETR, on their own had only minor effects in this context ( Figure 9B). However, in the presence of Na 2 S they reduced • cPTIO. The reducing potency of the mixture increased with increasing concentration of DOXY (50-200 µM; Figure 9A). When the potency of the mixtures of DOXY, OXYT or TETR (400 µM) with Na 2 S (200 µM) was compared, the following order was obtained: Na 2 S/TETR>Na 2 S/DOXY~Na 2 S/OXYT>>Na 2 S~0 ( Figure 9B). It was of interest to know if fluoroquinolone antibiotic NORF (400 µM) or fungal toxin FUSA (400 µM) can reduce • cPTIO. These compounds per se or in the presence of Na 2 S (400 µM) had minor effects, as they reduced • cPTIO <4% after 20 min ( Figure S3). spectra revealed that the O2 and OH radicals were produced by Na2S2/DOXY in the presence of H2O2 ( Figure 5F).

Potency of Compounds to Reduce the • cPTIO Radical
In the previous section, we showed that Na2S and Na2S2 interacting with DOXY possess different time-and concentration-dependent potency to produce and scavenge radicals. Since the Na2S/DOXY/H2O2 and Na2S2/DOXY mixtures behave in the bell-shaped manner in these reactions, initially producing radicals followed by scavenging them, we next studied the potency of Na2S/DOXY and Na2S2/DOXY to reduce the • cPTIO radical as a model radical system. Decrease of • cPTIO ABS at 560 nm was used to measure reduction of the • cPTIO radical. We compared the effect of DOXY with other tetracycline derivatives, OXYT and TETR, and antibiotics, fusaric acid (FUSA) and norfloxacin (NORF), to reduce the • cPTIO radical alone and in combination with Na2S and polysulfides. For this purpose, we extended the set of studied polysulfides with Na2S3 and Na2S4.

Ability of the Polysulfide/Tetracyclines Mixture to Reduce the • cPTIO Radical
Since polysulfides Na 2 S 2 , Na 2 S 3 and Na 2 S 4 at 100 µM concentration reduced • cPTIO in <1 min (Figure 3), we used lower 40 µM concentrations to study their effects in a mixture with tetracyclines. All tetracyclines potentiated ability of Na 2 S 2 and Na 2 S 3 to reduce • cPTIO ( Figure 10A,B). In case of Na 2 S 4 , the effects were less pronounced ( Figure 10C). It is noteworthy that the extent and rate of the polysulfides' ability to reduce • cPTIO depends on an amount of sulfurs atoms. Efficiency of Na 2 S, Na 2 S 2 , Na 2 S 3 and Na 2 S 4 (40 µM) to reduce • cPTIO (100 µM) was 0%, 35%, 63% and 87% respectively, and the rate of reduction might be different depending on sulfur atoms ( Figure 10D).
Since polysulfides Na2S2, Na2S3 and Na2S4 at 100 µ M concentration reduced cPTIO in < 1 min (Figure 3), we used lower 40 µ M concentrations to study their effects in a mixture with tetracyclines. All tetracyclines potentiated ability of Na2S2 and Na2S3 to reduce • cPTIO ( Figure 10A,B). In case of Na2S4, the effects were less pronounced ( Figure 10C). It is noteworthy that the extent and rate of the polysulfides' ability to reduce • cPTIO depends on an amount of sulfurs atoms. Efficiency of Na2S, Na2S2, Na2S3 and Na2S4 (40 µ M) to reduce • cPTIO (100 µ M) was 0%, 35%, 63% and 87% respectively, and the rate of reduction might be different depending on sulfur atoms ( Figure 10D).

Tetracyclines Cleave pDNA in the Presence of Na2S, but Inhibit pDNA Cleavage Induced by Polysulfides
To put into the biological frame our findings on free radical producing/scavenging interaction between tetracyclines and reactive sulfur species, which seem to be time-and concentrationdepended, we used well-characterized radical-induced pDNA cleavage assay. Tetracyclines (0.05-2.5 mM) alone have virtually no pDNA damaging effects. However, in the presence of Na2S (0.5 mM) the cleavage potencies robustly increased in the following order: DOXY > TETR ≥ OXYT >> FUSA ~ 0 ( Figure 11). Interestingly, the Na2S/DOXY mixture exhibited the pDNA damaging effects with the bell-shaped characteristics. NORF was not used in this study due to low solubility in the reaction buffer at the listed concentrations. All studied polysulfides slightly cleaved pDNA similarly to our previous findings on Na2S4 [8]. However, tetracyclines (DOXY, OXYT and TETR) in the presence of the polysulfides inhibited their pDNA cleavage potency in the concentration-dependent manner ( Figure 12).
All tetracyclines potentiated ability of Na2S2 and Na2S3 to reduce • cPTIO ( Figure 10A,B). In case of Na2S4, the effects were less pronounced ( Figure 10C). It is noteworthy that the extent and rate of the polysulfides' ability to reduce • cPTIO depends on an amount of sulfurs atoms. Efficiency of Na2S, Na2S2, Na2S3 and Na2S4 (40 µ M) to reduce • cPTIO (100 µ M) was 0%, 35%, 63% and 87% respectively, and the rate of reduction might be different depending on sulfur atoms ( Figure 10D).

Tetracyclines Cleave pDNA in the Presence of Na2S, but Inhibit pDNA Cleavage Induced by Polysulfides
To put into the biological frame our findings on free radical producing/scavenging interaction between tetracyclines and reactive sulfur species, which seem to be time-and concentrationdepended, we used well-characterized radical-induced pDNA cleavage assay. Tetracyclines (0.05-2.5 mM) alone have virtually no pDNA damaging effects. However, in the presence of Na2S (0.5 mM) the cleavage potencies robustly increased in the following order: DOXY > TETR ≥ OXYT >> FUSA ~ 0 ( Figure 11). Interestingly, the Na2S/DOXY mixture exhibited the pDNA damaging effects with the bell-shaped characteristics. NORF was not used in this study due to low solubility in the reaction buffer at the listed concentrations. All studied polysulfides slightly cleaved pDNA similarly to our previous findings on Na2S4 [8]. However, tetracyclines (DOXY, OXYT and TETR) in the presence of the polysulfides inhibited their pDNA cleavage potency in the concentration-dependent manner ( Figure 12).

Tetracyclines Cleave pDNA in the Presence of Na 2 S, but Inhibit pDNA Cleavage Induced by Polysulfides
To put into the biological frame our findings on free radical producing/scavenging interaction between tetracyclines and reactive sulfur species, which seem to be time-and concentration-depended, we used well-characterized radical-induced pDNA cleavage assay. Tetracyclines (0.05-2.5 mM) alone have virtually no pDNA damaging effects. However, in the presence of Na 2 S (0.5 mM) the cleavage potencies robustly increased in the following order: DOXY > TETR ≥ OXYT >> FUSA~0 ( Figure 11). Interestingly, the Na 2 S/DOXY mixture exhibited the pDNA damaging effects with the bell-shaped characteristics. NORF was not used in this study due to low solubility in the reaction buffer at the listed concentrations. All studied polysulfides slightly cleaved pDNA similarly to our previous findings on Na 2 S 4 [8]. However, tetracyclines (DOXY, OXYT and TETR) in the presence of the polysulfides inhibited their pDNA cleavage potency in the concentration-dependent manner ( Figure 12).

Na2S Did Not Modify Inhibitory Effect of DOXY on Growth of Escherichia Coli Cells
To examine whether exogenously added Na2S has an observable effect on living cells undergoing DOXY treatment, we measured the growth of E. coli cells in the presence of DOXY and several concentrations of Na2S. Growth of bacterial culture was measured as change in the optical density (OD) at 600 nm within six hours. As shown (Figure 13), the presence of exogenous source of Na2S had no significant effect on bacterial cells treated with 50 nM or 100 nM DOXY.

Na 2 S Did Not Modify Inhibitory Effect of DOXY on Growth of Escherichia Coli Cells
To examine whether exogenously added Na 2 S has an observable effect on living cells undergoing DOXY treatment, we measured the growth of E. coli cells in the presence of DOXY and several concentrations of Na 2 S. Growth of bacterial culture was measured as change in the optical density (OD) at 600 nm within six hours. As shown (Figure 13), the presence of exogenous source of Na 2 S had no significant effect on bacterial cells treated with 50 nM or 100 nM DOXY.

Na2S Did Not Modify Inhibitory Effect of DOXY on Growth of Escherichia Coli Cells
To examine whether exogenously added Na2S has an observable effect on living cells undergoing DOXY treatment, we measured the growth of E. coli cells in the presence of DOXY and several concentrations of Na2S. Growth of bacterial culture was measured as change in the optical density (OD) at 600 nm within six hours. As shown (Figure 13), the presence of exogenous source of Na2S had no significant effect on bacterial cells treated with 50 nM or 100 nM DOXY.  Figure 13. The effects of Na 2 S on bacterial growth in the presence of DOXY. Representative growth curves of parallel samples derived from three independent experiments show no significant effects of Na 2 S on E. coli cells undergoing DOXY treatment. Control (n = 2; dash green, green), 50 nM DOXY (n = 2; dash black and black), 10 µM Na 2 S (short dash blue), 25 µM Na 2 S (long dash blue), 50 µM Na 2 S (blue), 50/10 nM/µM DOXY/Na 2 S (short dash red), 50/25 nM/µM DOXY/Na 2 S (long dash red), 50/50 nM/µM DOXY/Na 2 S (red), 100/50 nM/µM DOXY/Na 2 S (dash pink) and 100 nM DOXY (pink). Growth of bacterial culture was measured as change in optical density at 600 nm (OD 600 ) within six hours.

Practical Use of Polysulfides versus Na 2 S
Our study confirmed and underlined that polysulfides (or their anion form) relative to Na 2 S are unstable in buffer solutions at 37 • C and their effects strongly depend on time of their storage or incubation (Figures 1-3). This should be taken into careful consideration particularly in a long(er) time experimental settings. The exact species formed by polysulfides in the aqueous solutions are still unclear [35]. Based on the differences in the time-dependent UV-VIS spectra (ABS at 600 and 900 nm; Figures 1 and 2), we suggest that different species are formed from Na 2 S 2 and Na 2 S 3 or Na 2 S 4 . ABS of HShas a peak at 232 nm. Since ABS at 232 nm decreased for all polysulfides over the time, we suggest that during the polysulfides incubation no, or negligible, amount of HSmolecules are formed (Figures 1 and 2, Insets). The decrease of ABS at 600-900 nm by • cPTIO added to 40 min of incubated Na 2 S 3 and Na 2 S 4 solutions indicates disturbing of the light scattering sulfur-containing species by • cPTIO. Since decrease of ABS at 900 nm (Figure 2, Inset-blue line) correlated with the decrease of the • cPTIO radical concentration ( Figure 3B, black line), we may assume that electron transfer from the light scattering sulfur-containing species to • cPTIO occurs.

Interaction of Na 2 S and Polysulfides with Tetracyclines
Here, we provide the first evidence that Na 2 S and polysulfides interact with DOXY and that this interaction produces/scavenges the O 2 •− and • OH radicals. Mechanistic details of these interactions remain unknown yet, but we propose that the Na 2 S/DOXY and Na 2 S 2 /DOXY product ( (Figures 7 and 8). In our previous study, Na 2 S 4 was more potent scavenger of radicals and producer of • OH by decomposition of H 2 O 2 compared to Na 2 S [8]. Our present results indicate that Na 2 S 2 and Na 2 S 4 possess similar properties in this manner. The bell-shaped production of O 2 •− and • OH and scavenging of the • BMPO-OOH/OH radicals by increasing ratio of Na 2 S 2 /DOXY ( Figure 6) can be explained by reduction potency of high concentration of Na 2 S 2 , as documented by reduction of the • cPTIO radical, which is potentiated by DOXY ( Figure 10A). At high Na 2 S 2 /DOXY ratio (100/250 µM/µM), • BMPO-OOH/OH spectra were not detected due to reduction of radicals during Na 2 S 2 /DOXY interaction and/or reduction of • BMPO-OOH/OH. In our previous study, we reported that Na 2 S 4 possesses higher potency to reduce • cPTIO than Na 2 S [8]. In the present study, we found high • cPTIO reducing properties, being comparable to that of Na 2 S 4 , for other two polysulfides, Na 2 S 2 and Na 2 S 3 (Figures 3 and 10). Based on comparison of the time-dependent ability of Na 2 S and polysulfides to reduce • cPTIO ( Figure 10D), we may assume that the effectiveness in reducing of the • cPTIO radical depends on sulfur(s) chemical configuration/arrangement in H 2 S n (n ≥ 1) species. Several studies showed that tetracyclines produce and inhibit radicals in different, mostly non-physiological conditions. Tetracyclines during photochemical oxidation, autoxidation or oxidation with a Fenton reagent in the aqueous solution produced O 2 •− , • OH, H 2 O 2 or singlet oxygen [39][40][41][42].
In addition, DOXY can induce ROS production [43]. Kladna et al. found that in dimethyl sulfoxide DOXY and OXYT generated O 2 •− at low concentrations, but scavenged it at high concentrations [42].
However, under our experimental conditions and by using EPR spectra of spin trap BMPO, we could detect formation of the O 2 •− and • OH radicals only in the mixture of DOXY/Na 2 S or DOXY/Na 2 S 2 .
Based on our data, we speculate that sulfide and polysulfide interacting with tetracyclines may modulate O 2 •− redox chemistry.
Here, we provide evidence that H 2 S and polysulfides interacting with tetracyclines reduce • cPTIO and modulate cleavage of pDNA. Such properties are not generally adopted by all antibiotics, since FUSA and NORF do not notably interact with Na 2 S, based on their • cPTIO reduction inefficiency. In addition, FUSA do not damage pDNA alone or in the presence of Na 2 S. Hence, it can be proposed that radicals produced during Na 2 S/polysulfides-tetracyclines interaction (Figures 4-8) are involved in pDNA cleavage ( Figure 11) and that high reduction properties of the polysulfides/tetracyclines mixtures in comparison to Na 2 S/tetracyclines mixtures are responsible for the inhibitory effects of tetracyclines in the presence of polysulfides ( Figure 12). The polysulfides/tetracyclines mixtures can efficiently scavenge radicals before they reach and damage pDNA. Recently, Gallo et al. reported that OXYT induces DNA damage, the fact representing a possible risk for human and animal health [44]. In contrast, no pDNA damage caused by OXYT on its own was observed under our experimental conditions. Only in the mixture with Na 2 S, OXYT was able to induce DNA injury ( Figure 11D).
It was proposed that bactericidal activities of tetracyclines may results from their capability of producing ROS, and involvement of ROS in bactericidal activities has become the subject of extensive debate [33]. Therefore, it was of high priority to know if radicals produced during Na 2 S/DOXY interaction influence the antibacterial effects of DOXY. Since addition of Na 2 S has no significant effect on bacterial cells growing in the presence of DOXY (Figure 13), it could be supposed that radicals produced during Na 2 S/DOXY interaction do not have any important influence in this type of in vivo experiment at the given Na 2 S/DOXY concentrations. However, a complexity of in vivo system may rather mask existence, extend and contribution of certain reactions, particularly if they are backed up, to the resulting phenotype and additional more sophisticated experimental setup is required to adequately address this issue. One of the options would be to use strains that can produce H 2 S endogenously so that its steady-state levels are ensured throughout whole experiment.
To our best knowledge, there is no kind of information on interaction of DOXY with H 2 S or polysulfides. Numerous qualitatively and quantitatively different time-and concentration-dependent data, which we provide, indicate that the H 2 S/polysulfides-tetracyclines interactions are highly complex, and specific additional "chemical" approach is therefore needed to delineate them.

Possible Biological Consequences of the Na 2 S and Polysulfides Interacting with Tetracyclines
Maximum human plasma concentrations of DOXY are usually ranging from 1.5 to 7.0 µg/mL or from 3 to 14.6 µM [45,46]. Concentrations of endogenously produced polysulfides in HeLa cells are~0-120 nM [47] and H 2 S concentration can be higher than 1 µM [48]. However, in the very local environment presence of 1 molecule in 1000 nm 3 gives raise of 1.65 mM H 2 S or polysulfide concentration. We assume that H 2 S and polysulfide concentrations might be time-dependently higher in situ, where they are enzymatically produced and/or released from intracellular H 2 S stores [1]. Another source of H 2 S could be H 2 S donors, which are extensively studied as H 2 S releasing drugs [3,28]. The mutual administration of H 2 S donors with tetracyclines is challenge for the future studies and can contribute to understanding of a biological relevancy of the H 2 S/polysulfide-tetracycline interactions.
Tetracyclines have several positive and negative biological effects in which free radicals might play a role. For example, DOXY protects human intestinal cells or renal function from hypoxia/reoxygenation injury, improves cardioprotection [24,32,37], protects against ROS-induced mitochondrial fragmentation and isoproterenol-induced heart failure [31], inhibits mitochondrial biogenesis and alters energy metabolism [49,50]. DOXY induces cell death and prevents the proliferation of several types of cell by inducing ROS production [43]. OXYT induces oxidative damage in liver and kidney [51,52], oxidative stress and immunosuppression in rainbow trout [53], and modulates inflammation, apoptosis and cancer [54][55][56][57]. Whether and how H 2 S/polysulfides-tetracycline interaction plays a role in these biological effects is a challenge for the future research.

EPR of the • BMPO-adducts
To study an involvement of radicals in Na 2 S/polysulfides/DOXY interaction, EPR study of spin trap BMPO was used and conducted in accordance with previously reported protocols [8]. To the solution (final concentrations) of BMPO (30 mM), DTPA (100 µM) in sodium phosphate buffer (50 mM, pH 7.4, 37 • C), aliquots of the compounds were added. The sample was mixed for 5 s and transferred to a standard cavity aqueous EPR flat cell. The first EPR spectrum was recorded 110 ± 15 s after mixing the sample. The sets of individual EPR spectra of the • BMPO spin-adducts were recorded as 15 sequential scans, each 42 s, within a total time of 11 min. EPR spectra of the • BMPO spin-adducts were measured on an EMX spectrometer (Bruker, Rheinstetten, Germany) X-band~9.

UV-VIS of • cPTIO
To 900-990 µL solution of 100 mM sodium phosphate, 100 µM DTPA buffer (pH 7.4, 37 • C), the final concentrations of the studied compounds were added (final volume 1 mL). UV-VIS spectra (900-190 nm) were recorded 40 (80) × 30 s using a Shimadzu 1800 spectrometer (Kyoto, Japan) at 37 • C (blank was H 2 O). For our study, the • cPTIO extinction coefficient at 560 nm of 930 M −1 cm -1 was used. Scavenging of the • cPTIO radical by the studied compounds was determined as a decrease of ABS at 560 nm (absorption maximum of • cPTIO) minus ABS at 730 nm after subtraction of baseline absorbance [8].

pDNA Cleavage Assay
The pBR322 plasmid (N3033L, New England BioLabs Inc., Frankfurt, Germany) was used in pDNA cleavage assay that was performed according to our previous report [8]. In this assay, all samples contained 0.2 µg pDNA in a sodium phosphate buffer (25 mM sodium phosphate, 50 µM DTPA, pH 7.4, 37 • C). After addition of compounds, the resulting mixtures were incubated for 30 min at 37 • C. Afterwards, the reaction mixtures were subjected to 0.6% agarose gel electrophoresis. The samples were electrophoresed in TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA, pH 8.0) at 5.5 V/cm for 2 h. Gels were stained with Gel Red™Nucleic Acid Gel Stain and photographed using the Odyssey Fc Imaging System (LI-COR Biotechnology, Bad Homburg, Germany). The integrated densities of two identified pBR322 forms (supercoiled and nicked circular form) in each lane were quantified using Image Studio analysis software (LI-COR Biotechnology, Bad Homburg, Germany) to estimate pDNA cleavage efficiency.

Conclusions
We present evidence that sulfide and polysulfides interact with tetracyclines and produce/scavenge free radicals. Some of the radical producing/scavenging properties display the bell-shaped behaviour that is dependent on time and/or concentration of the mixture components. Since H 2 S, and probably polysulfides, are endogenously produced in all organs, it may be suggested that some of the biological effects of tetracyclines are due to their interaction with H 2 S and/or polysulfides. Our results indicate that further studies of the biological effects of the H 2 S/polysulfide combination with tetracyclines may help to understand their possible mutual role in a "free radical signaling" and the combination may be useful in pathological states in which radicals play a negative role.
Supplementary Materials: The following are available online. Figure S1: Time-dependent reduction of the • cPTIO radical by Na 2 S 3 . Figure S2: EPR spectra of the • BMPO adducts for Na 2 S 2 , Na 2 S and DOXY and their mutual combinations. Figure S3: Time-dependent reduction of the • cPTIO radical by the studied compounds.