Effects of Manganese Porphyrins on Cellular Sulfur Metabolism

Manganese porphyrins (MnPs), MnTE-2-PyP5+, MnTnHex-2-PyP5+ and MnTnBuOE-2-PyP5+, are superoxide dismutase (SOD) mimetics and form a redox cycle between O2 and reductants, including ascorbic acid, ultimately producing hydrogen peroxide (H2O2). We previously found that MnPs oxidize hydrogen sulfide (H2S) to polysulfides (PS; H2Sn, n = 2–6) in buffer. Here, we examine the effects of MnPs for 24 h on H2S metabolism and PS production in HEK293, A549, HT29 and bone marrow derived stem cells (BMDSC) using H2S (AzMC, MeRho-AZ) and PS (SSP4) fluorophores. All MnPs decreased intracellular H2S production and increased intracellular PS. H2S metabolism and PS production were unaffected by cellular O2 (5% versus 21% O2), H2O2 or ascorbic acid. We observed with confocal microscopy that mitochondria are a major site of H2S production in HEK293 cells and that MnPs decrease mitochondrial H2S production and increase PS in what appeared to be nucleoli and cytosolic fibrillary elements. This supports a role for MnPs in the metabolism of H2S to PS, the latter serving as both short- and long-term antioxidants, and suggests that some of the biological effects of MnPs may be attributable to sulfur metabolism.

While early research on these compounds focused on their SOD-mimetic and antioxidant attributes, recent evidence suggests that substantial biological activity is derived from their paradoxical oxidant activity through an increase in hydrogen peroxide (H 2 O 2 ) in cells. This activates cytoprotective antioxidant defenses in healthy cells, presumably mediated to a large extent through the peroxidation of Keap-1 which frees it from Nrf2 allowing the latter to translocate to the nucleus and activate the antioxidant defenses. Peroxide also creates lethal oxidative stress in malignant cells [1,4]. However, unlike their ability to dismute superoxide, in this instance MnPs are believed to produce H 2 O 2 by MnP cycling between one-electron reduction by a cellular reductant such as ascorbic acid (AA) and one-electron oxidation by oxygen. Ascorbic acid reduces Mn(III)P to Mn(II)P and oxygen reoxidizes it to Mn(III); this also reduces oxygen to superoxide. The superoxide produced from this reaction then dismutes, either spontaneously or catalyzed by SOD or MnPs, to produce O 2 and H 2 O 2 . H 2 O 2 then oxidizes regulatory protein cysteines thereby conferring the respective cytoprotective effects in healthy cells and cytotoxic reactions in malignant ones [1].
We have previously shown that both SOD1 and SOD2 and the above MnPs oxidize hydrogen sulfide (H 2 S) to form polysulfides in buffer [5,6]. In the present study we examine the effects of MnPs on endogenous sulfur metabolism in a variety of cell lines. We show that MnPs decrease the rate of endogenous H 2 S production and increase intracellular polysulfides in cells. These results suggest that MnPs oxidize intracellular H 2 S to form polysulfides, which is consistent with our previous observations of these reactions in buffer. We also observed that these reactions were independent of either ascorbic acid or hydrogen peroxide, suggesting that peroxide may not be involved in polysulfide production during H 2 S oxidation. Polysulfides formed by these reactions are potent antioxidants and can also initiate cytoprotective effects through persulfidation of Keap-1 and liberation of Nrf2. Collectively, these experiments support the hypothesis that MnPs affect intracellular sulfide metabolism and that the products from these reactions can directly contribute to the antioxidant effects of the MnPs.

Effects of MnPs on Sulfur Metabolism in HEK293 Cells
The effects of MnTE, MnTnHex and MnTnBuOE on intracellular H 2 S (AzMC fluorescence) and polysulfides (SSP4 fluorescence) were examined in HEK 293 cells in standard incubator conditions (21% O 2 , 5% CO 2 , 74% N 2 ) that are hyperoxic for these cells and under more physiological oxygen (physioxia) conditions (5% O 2 , 5% CO 2 , 90% N 2 ) in gas permeable 96 well plates. Physioxia alone consistently increased H 2 S but had insignificant effects on polysulfides in otherwise untreated cells ( Figure 1). All three MnPs at 3 µM decreased intracellular H 2 S and more than doubled polysulfides in both hyperoxic and physioxic environments and MnP-stimulated polysulfide production was greater in physioxic cells than it was in hyperoxic cells ( Figure 1).
Similar results were observed with 0.3 μM MnPs (Supplemental Figure S1) and with 1 μM MnTE (not shown). Polysulfides were still significantly increased, albeit less so, at 0.03 μM MnP although the effects on H2S were less pronounced (Supplemental Figure S2). Figure 2 shows the time-and concentration-dependent effects of MnTE on H2S (AzMC) and polysulfides (SSP4) in HEK 293 cells expressed as the percentage change from before to after MnTE treatment. The magnitude of all responses increased with MnTE concentration. However, the effects of MnTE on both H2S and polysulfides appeared to be independent of the oxygen status of the cells and relatively independent of the duration of MnTE exposure.   . Intracellular H 2 S were increased in physioxia compared to hyperoxia (p < 0.001), whereas polysulfides were unaffected. All MnPs decreased intracellular H 2 S and increased intracellular polysulfides in both O 2 environments (p < 0.001). Mean +SE, n = 16 wells per treatment, error bars may be covered by symbols.
Similar results were observed with 0.3 µM MnPs (Supplemental Figure S1) and with 1 µM MnTE (not shown). Polysulfides were still significantly increased, albeit less so, at 0.03 µM MnP although the effects on H 2 S were less pronounced (Supplemental Figure S2). Figure 2 shows the time-and concentration-dependent effects of MnTE on H 2 S (AzMC) and polysulfides (SSP4) in HEK 293 cells expressed as the percentage change from before to after MnTE treatment. The magnitude of all responses increased with MnTE concentration. However, the effects of MnTE on both H 2 S and polysulfides appeared to be independent of the oxygen status of the cells and relatively independent of the duration of MnTE exposure. Similar results were observed with 0.3 μM MnPs (Supplemental Figure S1) and with 1 μM MnTE (not shown). Polysulfides were still significantly increased, albeit less so, at 0.03 μM MnP although the effects on H2S were less pronounced (Supplemental Figure S2). Figure 2 shows the time-and concentration-dependent effects of MnTE on H2S (AzMC) and polysulfides (SSP4) in HEK 293 cells expressed as the percentage change from before to after MnTE treatment. The magnitude of all responses increased with MnTE concentration. However, the effects of MnTE on both H2S and polysulfides appeared to be independent of the oxygen status of the cells and relatively independent of the duration of MnTE exposure.     All three MnPs also increased cellular polysulfides and this appeared to occur in nucleoli and what appeared to be fibrillary-like structures, although the specific intracellular locations where these occurred have not yet been identified. These results confirm that MnPs affect endogenous sulfur concentrations and suggest that the MnPs consume H2S in the production of polysulfides as they do in cell-free conditions [6].

Effects of MnPs on Sulfur Metabolism in Mesenchymal Stem, Bovine Pulmonary Artery Smooth Muscle, A549 and HT29 Cells
The effects of oxygen tension and 1 μM MnTE, MnBuOE and MnHex on human bone marrow derived stem cells are shown in Figure 4. H2S production was significantly (p < 0.05) greater in 5% (physioxic) O2 than in 21% (hyperoxic) O2 at all but 24 h, whereas the effects of O2 on polysulfides were less pronounced. All three MnPs decreased cellular H2S production and increased polysulfides. All three MnPs also increased cellular polysulfides and this appeared to occur in nucleoli and what appeared to be fibrillary-like structures, although the specific intracellular locations where these occurred have not yet been identified. These results confirm that MnPs affect endogenous sulfur concentrations and suggest that the MnPs consume H 2 S in the production of polysulfides as they do in cell-free conditions [6].

Effects of MnPs on Sulfur Metabolism in Mesenchymal Stem, Bovine Pulmonary Artery Smooth Muscle, A549 and HT29 Cells
The effects of oxygen tension and 1 µM MnTE, MnBuOE and MnHex on human bone marrow derived stem cells are shown in Figure 4. H 2 S production was significantly (p < 0.05) greater in 5% (physioxic) O 2 than in 21% (hyperoxic) O 2 at all but 24 h, whereas the effects of O 2 on polysulfides were less pronounced. All three MnPs decreased cellular H 2 S production and increased polysulfides.
H 2 S was also increased by the lower physioxic O 2 tension in bovine pulmonary artery smooth muscle cells, adenocarcinomic human alveolar basal epithelial (A549) cells, and human colon cancer (HT29) cells, with minimal effects on cellular polysulfides ( Figure 5). Consistent with the previous findings, 1 µm MnTE decreased H 2 S and increased polysulfides in all three cells and this appeared to be unaffected by O 2 tension ( Figure 5). Collectively, these results show that MnPs decrease intracellular H 2 S and increase intracellular polysulfides irrespective of cell type and oxygen status. H2S was also increased by the lower physioxic O2 tension in bovine pulmonary artery smooth muscle cells, adenocarcinomic human alveolar basal epithelial (A549) cells, and human colon cancer (HT29) cells, with minimal effects on cellular polysulfides ( Figure 5). Consistent with the previous findings, 1 μm MnTE decreased H2S and increased polysulfides in all three cells and this appeared to be unaffected by O2 tension ( Figure 5). Collectively, these results show that MnPs decrease intracellular H2S and increase intracellular polysulfides irrespective of cell type and oxygen status.   H2S was also increased by the lower physioxic O2 tension in bovine pulmonary artery smooth muscle cells, adenocarcinomic human alveolar basal epithelial (A549) cells, and human colon cancer (HT29) cells, with minimal effects on cellular polysulfides ( Figure 5). Consistent with the previous findings, 1 μm MnTE decreased H2S and increased polysulfides in all three cells and this appeared to be unaffected by O2 tension ( Figure 5). Collectively, these results show that MnPs decrease intracellular H2S and increase intracellular polysulfides irrespective of cell type and oxygen status.

Effects of Ascorbic Acid on MnTE Oxidation of H 2 S in HEK and A549 Cells
The biological actions of MnPs are thought to be due to redox cycling where oxidized Mn(III)P is reduced by a reductant such as ascorbic acid (AA) and then re-oxidized by oxygen which generates superoxide [4,[8][9][10]. The superoxide then spontaneously (or catalyzed by SOD or MnPs if present) dismutes to form peroxide and oxygen with the ultimate biological effects exerted by the peroxide. To determine if MnPs could redox cycle using O 2 and AA to produce ROS that then would oxidize H 2 S, we treated HEK 293 and A549 cells with 1 µM MnTE and 1 mM AA and monitored H 2 S and polysulfide metabolism in 21% O 2 . As shown in Figure 6A,B, MnTE decreased H 2 S and increased polysulfides in both cell lines consistent with previous observations (cf. Figures 1 and 5). AA alone did not affect H 2 S in HEK293 cells, but reduced H 2 S in A549 cells and increased polysulfides in both cell lines. AA plus MnTE decreased polysulfides compared to MnTE alone in HEK 293 cells but did not affect A549 cells. These results suggest that MnTE does not redox cycle with AA and O 2 to form ROS that subsequently oxidizes H 2 S in either HEK293 or A549 cells.

Effects of Ascorbic Acid on MnTE Oxidation of H2S in HEK and A549 Cells
The biological actions of MnPs are thought to be due to redox cycling where oxidized Mn(III)P is reduced by a reductant such as ascorbic acid (AA) and then re-oxidized by oxygen which generates superoxide [4,[8][9][10]. The superoxide then spontaneously (or catalyzed by SOD or MnPs if present) dismutes to form peroxide and oxygen with the ultimate biological effects exerted by the peroxide. To determine if MnPs could redox cycle using O2 and AA to produce ROS that then would oxidize H2S, we treated HEK 293 and A549 cells with 1 μM MnTE and 1 mM AA and monitored H2S and polysulfide metabolism in 21% O2. As shown in Figure 6A,B, MnTE decreased H2S and increased polysulfides in both cell lines consistent with previous observations (cf. Figures 1 and 5). AA alone did not affect H2S in HEK293 cells, but reduced H2S in A549 cells and increased polysulfides in both cell lines. AA plus MnTE decreased polysulfides compared to MnTE alone in HEK 293 cells but did not affect A549 cells. These results suggest that MnTE does not redox cycle with AA and O2 to form ROS that subsequently oxidizes H2S in either HEK293 or A549 cells.

MnP Metabolism of H2S in Buffer is Independent of Ascorbic Acid and H2O2
To determine if MnP generation of polysulfides from H2S requires AA, we evaluated the effects of 1 mM AA on the reaction between 1 μM MnPs and H2S in buffer in 21% O2. As shown in Figure  7A, polysulfide production was not increased by incubation of any MnP with AA in combination with either 100 or 300 μM H2S. We then incubated H2S with H2O2 in the presence or absence of MnTE ( Figure 7B-D). H2O2 alone from 10 μM to 1 mM concentration-dependently increased polysulfide production in normoxia and hypoxia, albeit at a slow rate. MnTE increased polysulfide production, however, and this was concentration-dependently inhibited by 100 μM and 1 mM H2O2. These results suggest that H2O2 does not augment MnTE catalysis of H2S, and it appears to be inhibitory at higher concentrations [6].

MnP Metabolism of H 2 S in Buffer is Independent of Ascorbic Acid and H 2 O 2
To determine if MnP generation of polysulfides from H 2 S requires AA, we evaluated the effects of 1 mM AA on the reaction between 1 µM MnPs and H 2 S in buffer in 21% O 2 . As shown in Figure 7A, polysulfide production was not increased by incubation of any MnP with AA in combination with either 100 or 300 µM H 2 S. We then incubated H 2 S with H 2 O 2 in the presence or absence of MnTE ( Figure 7B-D). H 2 O 2 alone from 10 µM to 1 mM concentration-dependently increased polysulfide production in normoxia and hypoxia, albeit at a slow rate. MnTE increased polysulfide production, however, and this was concentration-dependently inhibited by 100 µM and 1 mM H 2 O 2 . These results suggest that H 2 O 2 does not augment MnTE catalysis of H 2 S, and it appears to be inhibitory at higher concentrations [6].

Discussion
In the present studies, we show that MnPs produced polysulfides in a variety of cell lines and that this appeared to result from MnP oxidation of endogenously produced H2S. This hypothesis is also supported by our observation that MnPs decrease cellular H2S. In addition, the effects of MnPs on cellular H2S and polysulfides were concentration-dependent and could be observed at concentrations as low as 0.03 μM. This illustrates the efficacy of MnPs in cellular sulfur metabolism. Our studies also suggest that MnP-catalyzed H2S oxidation is efficient at relatively low levels of O2 and H2S.

Cellular Oxygen and Sulfur Availability
Reducing ambient oxygen from 21% to 5% increased intracellular H2S (AzMC fluorescence), consistent with previous studies showing that oxygen-dependent metabolism of H2S is a mechanism for short-and long-term cellular oxygen sensing [7,11]. H2S is constitutively produced from cysteine catabolism via the transsulfuration pathway and metabolized in the mitochondrion, initially by the mitochondrial enzyme sulfur:quinone oxidoreductase [12]. As H2S is oxidized by SQR, electrons are transferred to complex III via ubiquinone and shuttled down the electron transport chain (ETC) to ultimately reduce oxygen at complex IV. As oxygen levels fall, electron flux down the ETC decreases and as H2S oxidation can no longer keep pace with production, H2S increases. This mechanism inversely couples oxygen availability to H2S concentration.
A fall in oxygen can also increase H2S via other mechanisms. There is a large store of polysulfides in cells [13][14][15][16][17]. A fall in oxygen will increase the reducing environment in cells and reductants such as ascorbate and thioredoxin can then release H2S from polysulfides and thiosulfate [18][19][20][21]. As we

Discussion
In the present studies, we show that MnPs produced polysulfides in a variety of cell lines and that this appeared to result from MnP oxidation of endogenously produced H 2 S. This hypothesis is also supported by our observation that MnPs decrease cellular H 2 S. In addition, the effects of MnPs on cellular H 2 S and polysulfides were concentration-dependent and could be observed at concentrations as low as 0.03 µM. This illustrates the efficacy of MnPs in cellular sulfur metabolism. Our studies also suggest that MnP-catalyzed H 2 S oxidation is efficient at relatively low levels of O 2 and H 2 S.

Cellular Oxygen and Sulfur Availability
Reducing ambient oxygen from 21% to 5% increased intracellular H 2 S (AzMC fluorescence), consistent with previous studies showing that oxygen-dependent metabolism of H 2 S is a mechanism for short-and long-term cellular oxygen sensing [7,11]. H 2 S is constitutively produced from cysteine catabolism via the transsulfuration pathway and metabolized in the mitochondrion, initially by the mitochondrial enzyme sulfur:quinone oxidoreductase [12]. As H 2 S is oxidized by SQR, electrons are transferred to complex III via ubiquinone and shuttled down the electron transport chain (ETC) to ultimately reduce oxygen at complex IV. As oxygen levels fall, electron flux down the ETC decreases and as H 2 S oxidation can no longer keep pace with production, H 2 S increases. This mechanism inversely couples oxygen availability to H 2 S concentration.
A fall in oxygen can also increase H 2 S via other mechanisms. There is a large store of polysulfides in cells [13][14][15][16][17]. A fall in oxygen will increase the reducing environment in cells and reductants such as ascorbate and thioredoxin can then release H 2 S from polysulfides and thiosulfate [18][19][20][21]. As we did see a slight decrease in polysulfides when cells were in 5% oxygen (Figure 1), it is possible some H 2 S was derived from polysulfides. H 2 S also induces HIF-1α [22] and HIF-1α in turn increases expression of H 2 S-producing enzyme cystathione γ-lyase suggesting the potential for complex interplay between these hypoxia response systems that should be pursued in the future.
Our studies also suggest that oxygen availability is not a limiting factor on H 2 S oxidation. We have previously shown that MnP oxidation of H 2 S in buffer consumes O 2 , even at very low O 2 tensions [6]. We also estimated that HEK293 cells in 21% O 2 produce approximately 2-4 × 10 −16 mol of H 2 S per hour per cell and this would require only 0.2-0.3% of the total O 2 consumed by these cells [23]. Therefore, it is likely that there is sufficient O 2 in cells, even at 5% O 2 , to support the additional oxidation of H 2 S by MnPs.

MnP Utilization of Intracellular Sulfur
It is evident that an increase in MnP concentration is associated with a decrease in cellular H 2 S production and an increase in polysulfide production ( Figure 2). We propose that the decrease in H 2 S is the result of MnP-driven oxidation of H 2 S to polysulfides. It is less clear what the source of H 2 S is and where it comes from.
It is generally accepted that the concentration of H 2 S in cells or plasma is less than 1 µM as the human nose can detect as little as 0.5-1 µM H 2 S in solution [24] and there is no appreciable noxious odor of rotten eggs emanating from healthy animal cells or blood. However, the potential significance of H 2 S as a significant source of polysulfides stems not from the absolute concentration of H 2 S but rather the rate of H 2 S production relative to its metabolism and the extensive intracellular sulfur pool from which H 2 S can be derived.
HEK293 cells have an approximate cell diameter of nearly 14 µm [25] and a cell volume of 1.5 × 10 −9 mL. If each cell produces 3 × 10 −16 mol of H 2 S per cell [23], intracellular H 2 S could theoretically increase to 200 µM in one hour if H 2 S was not metabolized. Mitochondrial oxidation of H 2 S to sulfate normally keeps pace with H 2 S production but this can quickly change as evidenced by the rapid rise in H 2 S associated with hypoxia [11] and by the longer term responses observed in the present and other studies [7]. While it is tempting to propose that MnP-oxidation diverts H 2 S generated from cysteine catabolism away from mitochondrial oxidation, this does not appear to be the case. Figure 2 shows that while MnP-catalyzed polysulfide formation was observed in cells in both 21% and 5% O 2 , when polysulfide production was normalized to the percentage change in absolute fluorescence, the effect of MnPs on both AzMc and SSP4 fluorescence appeared to be independent of oxygen status. This suggests that H 2 S is not diverted from normal mitochondrial metabolism, nor dependent on hypoxia-increased cellular reductants or coupled to HIF-1α.
Recent evidence suggests that hydropersulfides (RSSH) and polysulfides (RSS n R', where R and/or R' are an alkyl or H and n > 1) are far more prevalent in cells than previously thought and that they exist in a dynamic equilibrium with H 2 S and other small inorganic thiols as well as cysteine (CysS n H), glutathione (GS n H) and protein (PS n H) per-and polysulfides [15,[26][27][28][29][30]. Thus, there appears to be a large source/sink for H 2 S in cells that may be utilized by MnPs. The identity of this source and the mechanism by which H 2 S is mobilized from it, remains to be identified.

Antioxidant Function of MnP-Generated Polysulfides
A well known therapeutic attribute of MnPs is their ability to protect healthy cells and increase susceptibility of malignant cells to radiation therapy. This process has been attributed in part to increased H 2 O 2 which, in a process catalyzed by MnP, oxidizes signaling proteins such as NFκB [4,10,31]. Our results suggest that persulfides produced by MnP oxidation of H 2 S could have a similar effect with the added bonus of direct oxidant scavenging by persulfides.
Polysulfides generated from MnP oxidation of H 2 S could contribute to both short-and long-term antioxidant effects. Short-term benefits are derived from the direct antioxidant effects of polysulfides. Counterintuitive as it may seem, oxidization of H 2 S, which is the most reduced form of sulfur (oxidation state −2) to per-and polysulfides (oxidation states −1 or 0) produces far better antioxidants than H 2 S [15,27,32]. We have also demonstrated this antioxidant potency with the garlic-derived polysulfides [33]. Addition of garlic oil (GO) or diallyl trisulfide (DATS), the active polysulfide in GO, increased antioxidant activity in HEK 293 cells and this was correlated with an increase in intracellular polysulfides, not an increase in H 2 S. The present studies show that much of the MnP-induced increase in cellular polysulfides occurs during the initial 4 h and this could provide a short-term benefit through the direct antioxidant activity of these molecules.

Mechanism of MnP-Catalyzed H 2 S Oxidation in Cells
Based on our studies in cell-free conditions, we recently proposed that polysulfides were produced by MnP-catalyzed redox cycling between H 2 S as the reductant and O 2 as the oxidant [6]. In these reactions two molecules of Mn(III) are reduced by two hydrosulfide anions which produces two hydrosulfide radicals (Equation (1)). The two hydrosulfide radicals react with each other producing hydrogen persulfide (Equation (2)) and the reduced Mn(II) is re-oxidized by O 2 forming Mn(III) and superoxide (O 2 •− ; Equation (3)) and the latter is then spontaneously, or catalyzed by MnP, dismuted to hydrogen peroxide (H 2 O 2 ) and water (Equation (4)).
However, we did not observe a consistent increase in MnP-catalyzed polysulfide production in cells in the presence of excess ascorbic acid or in buffer with either excess ascorbic acid or hydrogen peroxide, suggesting that neither ascorbate nor peroxide contribute to MnP-catalyzed polysulfide production in cells. This suggests that not only are the reactions given by Equations (1)-(4) the predominate reactions of MnP oxidation of H 2 S to polysulfides, but they also suggest that the cytoprotective effects of MnPs can be achieved through sulfur metabolism.

Efficacy of MnTE, MnTnHex and MnTnBuOE
The three MnPs appear to be nearly equally efficacious in producing polysulfides in cells as measured by SSP4 fluorescence, although the confocal studies suggested that MnTE was slightly less potent ( Figure 3). Two factors can contribute to MnP activity in cells, direct catalytic efficiency and bioavailability. We showed in our previous study in cell-free conditions that MnTE was the most efficacious followed by MnTnBuOE and MnTnHeX [6]. The difference between MnP activity in buffer and cells appears to be due to bioavailability as MnTE accumulates in 4T1 cells to a lower extent than the more lipophilic MnTnHex [9].
3.6. Potential Limitations of the Study 3.6.1. Specificity of Fluorophores As with any study employing fluorophores, specificity is always a concern as it is virtually impossible to account for all potential interfering substances. AzMC and SSP4 specificity have been examined relative to other sulfur compounds as well as reactive oxygen and nitrogen species and both have been shown to be sufficiently specific for their respective analysis [23,43,44]. Clearly, there needs to be continual awareness of potential interfering substances with all fluorescent probes, however, we are reasonably confident that our probes are acceptable reporters of their respective RSS as employed in this study.

Effects of Fluorophores on Sulfur Metabolism
AzMC, MeRho and SSP4 are irreversible fluorophores that were present throughout the entire experiment. In this way, they provide a historical record of sulfur metabolism over the entire experiment and the results are less sensitive to transient variations in the concentration of sulfur species and to the relatively slow response characteristics (tens of minutes) of the fluorophores. However, it is also recognized that when used this way they can also consume H 2 S and polysulfides and thereby affect sulfur metabolism. We do not know to what extent this could affect our results. However, our findings are consistent with our previous observations of MnP oxidation of H 2 S and polysulfide production in cell-free conditions [6], suggesting to us that our general conclusion that MnPs also produce polysulfides from H 2 S in cells is valid.

Summary
The biological activity of SOD-mimetic MnPs is often attributed to their involvement in ROS metabolism which may then affect cysteine thiols in glutathione and proteins. It is now evident that these compounds also oxidize H 2 S and form polysulfides, mainly H 2 S 2 . As H 2 S-driven catalysis appears to be able to function under relatively low oxygen levels, this would allow polysulfide formation to proceed in the relatively hypoxic environment of the mitochondrion, which is where MnPs preferentially accumulate and where it appears that much H 2 S is produced. Direct cytoprotective effects of MnP oxidation of H 2 S can be derived from the direct antioxidative properties of polysulfides as well as prolonged activation of Nrf2. The ability of MnPs to oxidize H 2 S in low O 2 environments may take on even more therapeutic significance in hypoxic cells, such as solid tumors. Phosphate buffer (PBS; in mM): 137 NaCl, 2.7, KCl, 8 Na 2 HPO 4 , 2 NaH 2 PO 4 . pH was adjusted with 10 mM HCl or NaOH to 7.4. Please note that we use H 2 S to denote the total sulfide added (sum of H 2 S + HS − ) usually derived from Na 2 S. Note: S 2− , often thought of as part of the H 2 S + HS − equilibrium, does not exist under these conditions [45].

Cells
Human embryonic kidney (HEK 293) cells, bovine pulmonary artery smooth muscle (BPASMC) cells, human bone marrow derived stem cells, adenocarcinomic human alveolar basal epithelial (A549) cells, and human colon cancer (HT29) cells were cultured and maintained at 37 • C in a 21% O 2 /5% CO 2 humidified incubator supplemented with DMEM (low glucose) containing 10% FBS and 1% Pen/Strep. After trypsinization, cells were transferred from T-25 tissue culture flasks to gas-permeable 96-well plates (Coy Laboratory Products, Inc. grass Lake, MI, USA) and grown to >75% confluence. AzMC (7-azido-4-methylcoumarin) and SSP4 were used to monitor H 2 S and polysulfide production, respectively. The maximum MnP concentration used in these experiments was 3 µM, as previous studies have shown that above this concentration, the MnPs optically interfere with AzMC and SSP4 fluorescence [6].
Cells cultured under the above conditions are 'hyperoxic' relative to their 'physioxic' environment in vivo [46]. Due to the fact that exposure to more physioxic conditions also affects sulfur metabolism [7], a number of additional experiments were performed at more appropriate oxygen levels (5% O 2 /5% CO 2 /balance N 2 ) in a model 856-HYPO hypoxia chamber (Plas Labs, Inc. Lansing, MI, USA) at 37 • C. They were covered with a standard 96-well plate cover before they were removed from the hypoxia chamber for plate reader measurements.

Confocal Microscopy
HEK 293 cells were cultured as above and transferred to 8 chamber well plates mounted on #1.5 German borosilicate coverglass (Fisher, Grand Island, NY, USA) and grown to~75% confluence. They were then treated with either the H 2 S-specific fluorophore MeRho-Az (10 µM; [47]) or the polysulfide-specific fluorophore, SSP4 (20 µM; [48]) and then either no MnP or 1 µM of MnTE, MnTnBuOE or MnTnHex for 24 h in 21% O 2 /5% CO 2 and examined with a Zeiss LSM 710 confocal microscope and 63X objective (1.4 NA) magnification. Imaging conditions were kept the same for all samples. Fluorescence intensity from a minimum of 15 images of whole cells of both MeRho-Az and SSP4 treated cells was quantified using Image J without any prior image adjustments; representative micrographs were uniformly brightened for clarity.

Fluorescence Measurements in Buffer
Compounds of interest were aliquoted into black 96 well plates in a darkened room and fluorescence was measured on the SpectraMax plate reader. Fluorescence was typically measured every 10 min over 90 min, although most reactions were completed within the first 10-20 min. Excitation/emission wavelengths for 3 ,6 -di(O-thiosalicyl)fluorescein (SSP4), and 7-azido-4-methylcoumarin (AzMC) were 482/515 and 365/450 nm, respectively, per manufacture's recommendations.

Calculations
Results are expressed as mean +SE. Statistical analysis was determined by one-way ANOVA with Holm-Sidak for multiple comparisons. Significance was assumed at p ≤ 0.05.