Metabolic and cardiac signaling effects of inhaled hydrogen sulfide and low oxygen in male rats

Abstract

Low concentrations of inhaled hydrogen sulfide (H₂S) induce hypometabolism in mice. Biological effects of H₂S in in vitro systems are augmented by lowering O₂ tension. Based on this, we hypothesized that reduced O₂ tension would increase H₂S-mediated hypometabolism in vivo. To test this, male Sprague-Dawley rats were exposed to 80 ppm H₂S at 21% O₂ or 10.5% O₂ for 6 h followed by 1 h recovery at room air. Rats exposed to H₂S in 10.5% O₂ had significantly decreased body temperature and respiration compared with preexposure levels. Heart rate was decreased by H₂S administered under both O₂ levels and did not return to preexposure levels after 1 h recovery. Inhaled H₂S caused epithelial exfoliation in the lungs and increased plasma creatine kinase-MB activity. The effect of inhaled H₂S on prosurvival signaling was also measured in heart and liver. H₂S in 21% O₂ increased Akt-P^Ser473 and GSK-3β-P^Ser9 in the heart whereas phosphorylation was decreased by H₂S in 10.5% O₂, indicating O₂ dependence in regulating cardiac signaling pathways. Inhaled H₂S and low O₂ had no effect on liver Akt. In summary, we found that lower O₂ was needed for H₂S-dependent hypometabolism in rats compared with previous findings in mice. This highlights the possibility of species differences in physiological responses to H₂S. Inhaled H₂S exposure also caused tissue injury to the lung and heart, which raises concerns about the therapeutic safety of inhaled H₂S. In conclusion, these findings demonstrate the importance of O₂ in influencing physiological and signaling effects of H₂S in mammalian systems.

hydrogen sulfide (h₂s), although known as an industrial and environmental toxicant, is now recognized as an important physiological signaling molecule in mammalian systems (29, 34). Like nitric oxide (NO) and carbon monoxide (CO), H₂S is produced endogenously in mammalian cells by specific enzymes. H₂S is generated as a product of the two trans-sulfuration pathway enzymes cystathionine β-synthase and cystathioinine γ-lyase, as well as 3-mercaptopyruvate sulfurtransferase and cysteine aminotransferase (28, 35). Importantly, H₂S may have potent cytoprotective effects against oxidative injury as several studies show that sulfide-generating chemicals like sodium hydrosulfide (NaHS) and sodium sulfide (Na₂S) prevent cardiac damage from ischemia-reperfusion injury (7, 26, 30, 37). Lefer and colleagues (8, 16) showed that Na₂S pretreatment reduced infarct size and prevented deficits in cardiac output following coronary artery occlusion in mice. While the molecular mechanisms responsible for H₂S-mediated cardioprotection are not known, H₂S activation of the prosurvival Akt signaling pathway is implicated as a key metabolic target in heart (52).

New studies also demonstrate that H₂S has global effects on cellular metabolism and bioenergetics. Previous studies by Roth and colleagues (3) showed that when mice were exposed to 80 ppm H₂S, they experienced a marked depression in metabolism. Breathing rate and core body temperature decreased during H₂S exposure and then returned to baseline upon removal of H₂S with no apparent ill effects. While the precise mechanism for these responses is not known, hypometabolism may be attributed, in part, to the ability of H₂S to reversibly inhibit mitochondrial respiration (27). Additionally, H₂S can inhibit O₂ transport by binding to heme to form ferric sulfide species (46). The formation of sulfhemoglobin decreases the affinity of hemoglobin for O₂ and also inhibits O₂ transport (10). As a consequence, cellular respiration, mitochondrial electron flow, and ATP formation are inhibited as H₂S and HS⁻ ligate the heme a₃ of cytochrome c oxidase in either the ferrous or ferric states (38). Typically, these inhibitory effects on cellular respiration become fatal only at H₂S concentrations above several hundred parts per million; however, at lower concentrations (<80 ppm), eye, nose, and throat irritation has been noted (11). While the potential of H₂S to induce a hypometabolic state may have broad therapeutic implications for trauma medicine and organ transplantation, acute perturbations of lung and other organ systems following inhaled H₂S have not been evaluated (3).

Previous work from this laboratory has shown that O₂ tension affects the vasoactive actions of H₂S; e.g., under normoxic conditions H₂S causes vasodilation, while under hyperoxic conditions H₂S causes vasoconstriction (31). As mitochondrial respiratory rates can be affected by changes in O₂ concentrations (17, 50) it is likely that O₂ tension is a key factor determining the degree of H₂S-mediated hypometabolism in mammals. Taking these points into consideration, it was hypothesized that reduced O₂ tension would increase H₂S-mediated hypometabolism in vivo. To test this, male Sprague-Dawley rats were exposed to 80 ppm H₂S in 21% or 10.5% O₂ to determine the effect of O₂ tension on the ability of H₂S to induce a temporary and reversible hypometabolic state. Further, because H₂S is a toxic gas at high concentrations, the effect of H₂S on lung tissue and heart and liver enzymes was assessed.

MATERIALS AND METHODS

Exposure protocol for H₂S-induced hypometabolism.

Male Sprague-Dawley rats (∼300 g) were exposed to the following inhalation gas mixtures for 6 h: 1) 21% O₂ alone; 2) 80 ppm H₂S + 21% O₂; 3) 10.5% O₂ alone; and 4) 80 ppm H₂S + 10.5% O₂. After exposure, rats were allowed to recover for 1 h in room air before tissue and blood were collected for experimental measures. Detailed diagrams illustrating the exposure setup and experimental design are provided in Fig. 1, A and B, respectively. Digital mass flow controllers (Sierra Instruments, Monterey, CA) were used to blend gas mixtures to the target O₂ and H₂S concentrations from three gas tanks containing 5% H₂S in N₂ gas, N₂ gas alone, and air, respectively. Nitrogen was used as the diluent gas for treatments. Gas flow rate was 3 l/min through a custom-built 6-liter glass exposure chamber (2) and the inflow rate of H₂S gas did not alter gas phase O₂ concentration. Nitrogen and air source gases were humidified before countercurrent injection of H₂S followed by chamber inflow. An O₂ analyzer (SA-3 AEl Technologies, Pittsburgh, PA) and an H₂S analyzer (RM17, Interscan, Chatsworth, CA) calibrated with a 50.8 ppm H₂S gas standard (Scott Specialty Gases, Plumsteadville, PA) were used to monitor O₂ and H₂S concentrations, respectively, of both input and output chamber gas. The concentration of H₂S (80 ppm) was verified in the chamber by using the H₂S gas analyzer. Exposures were performed at the same time each day (9:00 AM to 3:00 PM) with lights on and were conducted in a chemical safety hood with the ambient temperature approximately 24–25°C. Food was withheld during exposures. All animal protocols were approved by the UAB Institutional Animal Care and Use Committee, and studies were performed in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH Publication No. 86–23).

Fig. 1.

Diagram illustrating the H₂S exposure system and experimental protocol. A: all gas flow rates were maintained by mass flow controllers. Carrier gases were humidified before converging with the H₂S gas line and the H₂S concentration was confirmed by using a H₂S gas analyzer. Rats were exposed to one of the gas mixtures in an air-tight 6-liter chamber maintained in a chemical safety fume hood. B: physiological parameters were measured 5 min before beginning the exposure experiments and then once per hour for 6 h during the exposure. Note that measurements were made while the animal continued to breathe the appropriate gas mixture through a nose cone. After 6 h, the animal was removed from the exposure chamber and allowed to recover in a separate cage while breathing room air. After 1 h of recovery, a final measurement was recorded for all physiological parameters. Statistical analyses were performed on measurements take before, during (at 5-h time point), and after exposures.

Physiological measurements.

A MouseOx pulse oximeter (STARR Life Sciences, Oakmont, PA) was used to noninvasively measure heart rate, ventilation rate, and hemoglobin saturation 5 min before the exposure, once per hour during the 6-h exposure, and 1 h after the exposure (Fig. 1B). The oximeter measures heart rate by detecting changes in infrared light caused by changes in blood volume that accompany each cardiac pulse due to cyclical pulse distention. Breathing causes a similar pulse distention, albeit at a lower frequency and magnitude. The oximeter also measures arterial blood oxygenation by detecting the fractional saturation of hemoglobin. To measure these physiological parameters during exposures, rats were removed from the chamber and a face cone was placed over the rat's nose to continue gas exposure. The oximeter was placed on the right hind foot and the rats were acclimated to the presence of the face cone and oximeter for 5 min. This time period allowed for readings to stabilize before data were recorded. In the event that physical movement of the rat caused an unstable signal in the ventilation rate measure, breathing rate was recorded by visual counting for a 1-min period. An infrared thermometer was placed in the ear to measure inner ear temperature. One hour following exposure, rats were anesthetized with ketamine/xylazine (75 mg/25 mg ip) and tissues were prepared for the following measurements.

Lung histopathology.

Lungs were cannulated via midline incision and insertion into the trachea, resected en bloc, and fixed with formalin using 25 cmH₂O hydrostatic pressure (1 h) via tracheal instillation for histopathology analyses (20). Following fixation, a cross section of lung tissue (cut perpendicular to the main airway path) was removed and used for hematoxylin and eosin (H and E) staining. Lung sections were assessed for airway damage by a lung pathologist, blinded to the experimental design, using the following pathological scoring system: “0” = normal columnar airway epithelia consisting of a majority of ciliated cells with a smaller number of nonciliated cells present; “1” = up to 25% of airway epithelium with evidence of epithelial cell injury (cell swelling/vacuolization or very thin epithelia); “2” = 25–50% of airway epithelium with evidence of cell injury; “3” = 50–75% of airway epithelium with evidence of cell injury; and “4” = >75% of airway epithelium with evidence of cell injury. The alveolar regions of the lung were evaluated for increased numbers of cell nuclei in alveolar region and the presence of extravascular erythrocytes.

Plasma chemistries.

Following exposures rats were allowed to recover for 1 h in room air before blood was collected for serum preparation. Potential H₂S-mediated damage to heart and liver was assessed by measuring plasma creatine kinase-MB (CK-MB) and alanine aminotransferase (ALT) activities, respectively, with commercial kits (Pointe Scientific, Canton, MI). Serum CK-MB activity increases after myocardial infarction; thus it can be used as a laboratory test for heart injury (48). These enzyme-coupled spectrophotometric assays monitor NAD⁺ reduction (CK-MB) and NADH oxidation (ALT) at 340 nm. The resulting rate of change in absorbance is directly proportional to enzyme activity.

Immunoblotting for Akt and GSK-3β.

Immunoblotting was performed by loading equal amounts (20 μg) of heart and liver homogenate protein onto 10% SDS-PAGE gels followed by transferring proteins to nitrocellulose membranes by standard techniques. Total Akt, phosphorylated Akt (Akt-P^Ser473), total GSK-3β, and phosphorylated GSK-3β (GSK-3β-P^Ser9) were measured using a 1:2,000 dilution of their respective primary antibodies (Cell Signaling Technology, Danvers, MA). After incubations with appropriate horseradish peroxidase-conjugated secondary antibodies, protein bands were visualized with chemiluminescence detection and were quantified using Quantity One software (Bio-Rad, Hercules, CA).

Statistical analysis.

Data are expressed as means ± SE for n ≥ 3 animals per measure. The level of statistical significance was set at P < 0.05. Statistical differences between groups were determined using two-factor analysis of variance (ANOVA) and one-factor repeated-measures (RM) ANOVA, where appropriate. The Holm-Sidak method for all pairwise multiple comparisons was used to identify which groups were statistically different within ANOVAs (SigmaStat 11, Systat Software Chicago, IL). Results from two-factor ANOVA for each physiological measurement are included within Figs. 2–5 (panel B), whereas results from a one-factor RM ANOVA for physiological measurements are compiled in Table 1. Post hoc statistics performed for two-factor ANOVA (physiological measurements) are included as letter notations within the bar graphs (Figs. 2–5; panel A) to show which experimental groups are different from each other within each of the three designated experimental measurement periods; i.e., before (5 min before), during (5 h time point), and after (1 h after) exposures. For simplicity, one time point (5 h) was used to determine differences among groups “during” exposures by 2-factor ANOVA (Figs. 2–5). It was at this time point when maximal effects (if any) were noted for H₂S and/or 10.5% O₂; thus this time point was used for analyses. Statistical significance for the histopathological scoring of the lung was determined by using ANOVA on ranks.

Fig. 2.

Effect of H₂S and/or 10.5% O₂ on body temperature. A: body temperature was measured before, during, and after exposures with an infrared thermometer placed in the ear. The “before” measure was taken ∼5 min before starting the exposure protocol, the “during” measure was taken 5 h into the exposure, and the “after” measure was taken 1 h after the exposure protocol had ended. B: results presented are P values obtained from 2-factor ANOVA with H₂S effect (absent vs. present), O₂ effect (21% vs. 10.5% O₂ tension), and interaction (H₂S × O₂). Results from post hoc statistical analyses: ^aP < 0.001 compared with 10.5% O₂ within during (5 h); ^bP < 0.001 compared with H₂S + 21% O₂ within during (5 h); and ^cP = 0.011 compared with 21% O₂ within after. Data represent means ± SE for n = 4 animals per treatment group.

Fig. 3.

Effect of H₂S and/or 10.5% O₂ on heart rate. A: heart rate was measured before, during, and after exposures with a pulse oximeter. The “before” measure was taken ∼5 min before starting the exposure protocol, the “during” measure was taken 5 h into the exposure, and the “after” measure was taken 1 h after the exposure protocol had ended. B: results presented are P values obtained from 2-factor ANOVA with H₂S effect (absent vs. present), O₂ effect (21% vs. 10.5% O₂ tension), and interaction (H₂S × O₂). Results from post hoc statistical analyses: ^aP = 0.008 compared with 21% O₂ within during (5 h); ^bP < 0.001 compared with 10.5% O₂ within during (5 h); and ^cP = 0.003 compared with 21% O₂ within after (1 h post exposure). Data represent means ± SE for n = 4 animals per treatment group. NS, not significant.

Fig. 4.

Effect of H₂S and/or 10.5% O₂ on ventilation rate. A: ventilation rate was measured before, during, and after exposures with a pulse oximeter. The “before” measure was taken ∼5 min before starting the exposure protocol, the “during” measure was taken 5 h into the exposure, and the “after” measure was taken 1 h after the exposure protocol had ended. B: results presented are P values obtained from 2-factor ANOVA with H₂S effect (absent vs. present), O₂ effect (21% vs. 10.5% O₂ tension), and interaction (H₂S × O₂). Results from post hoc statistical analyses: ^aP < 0.001 compared with 10.5% O₂ within during (5 h) and ^bP < 0.001 compared with H₂S + 21% O₂ within during (5 h). Data represent means ± SE for n = 4 animals per treatment group.

Fig. 5.

Effect of H₂S and/or 10.5% O₂ on hemoglobin saturation. A: hemoglobin saturation was measured before, during, and after exposures with a pulse oximeter. The “before” measure was taken ∼5 min before starting the exposure protocol, the “during” measure was taken 5 h into the exposure, and the “after” measure was taken 1 h after the exposure protocol had ended. B: results presented are P values obtained from 2-factor ANOVA with H₂S effect (absent vs. present), O₂ effect (21% vs. 10.5% O₂ tension), and interaction (H₂S × O₂). Results from post hoc statistical analyses: ^aP = 0.025 compared with 21% O₂ within during (5 h); ^bP = 0.025 compared with 10.5% O₂ within during (5 h); and ^cP < 0.001 compared with H₂S + 21% O₂ within during (5 h). Data represent means ± SE for n = 4 animals per treatment group.

Table 1.

P values from one-factor repeated-measures ANOVA on physiological measurements taken before, during, and after exposure to H₂S and/or 10.5% O₂

Groups	Body Temperature	Heart Rate	Ventilation Rate	Hemoglobin Saturation
21% O2	NS	NS	NS	NS
H2S + 21% O2	NS	0.010	NS	0.03
10.5% O2	<0.001	NS	0.041	0.002
H2S + 10.5% O2	<0.001	0.013	0.009	0.008

Results presented are P values obtained from 1-factor repeated-measures ANOVA. Body temperature, heart rate, ventilation rate, and hemoglobin saturation data are from Figs. 2–5, respectively. NS, not significant (P > 0.05).

RESULTS

Body temperature.

Unlike previous results obtained in mice (3), 80 ppm H₂S inhaled in 21% O₂ did not significantly decrease body temperature in rats (Fig. 2A and Table 1). However, H₂S combined with 10.5% O₂ caused a statistically significant decrease in body temperature at the 5-h time point, which returned to preexposure levels after 1 h of room air recovery (Fig. 2A and Table 1). Two-factor ANOVA demonstrated significant H₂S and O₂ effects on body temperature during exposure (Fig. 2B). There was also a significant interaction between H₂S and O₂ during exposure (Fig. 2B), indicating that the impact on body temperature was dependent on the level of both factors.

Heart rate.

Heart rate was significantly decreased by exposure to 80 ppm H₂S under both O₂ tensions (Fig. 3A and Table 1). Note that heart rate did not return to preexposure levels in both H₂S treatment groups (Fig. 3B), suggesting a possible toxic effect of H₂S on the heart. Additionally, the heart rate was unaffected by low O₂ alone (Fig. 3A and Table 1).

Ventilation rate.

Both H₂S in 10.5% O₂ and 10.5% O₂ alone significantly decreased the ventilation rate in rats (Fig. 4A and Table 1). Two-factor ANOVA demonstrated significant O₂ and H₂S effects (Fig. 4B). Like body temperature, the decrease in breathing rate was greater in the H₂S in 10.5% O₂ group than that in the 10.5% O₂ alone group (Fig. 4A). The drop in ventilation rate in the H₂S in 10.5% O₂ group returned to preexposure levels when rats were reexposed to room air (Fig. 4A).

Hemoglobin saturation.

H₂S in 10.5% O₂ and 10.5% O₂ alone significantly decreased hemoglobin saturation (Fig. 5A). Once the exposure protocol ended, hemoglobin saturation rapidly returned to preexposure levels (>90% hemoglobin saturation) in the H₂S in 10.5% O₂ and 10.5% O₂ alone groups (Fig. 5A). Two-factor ANOVA demonstrated a significant O₂ effect and interaction between H₂S and O₂, indicating the effect on hemoglobin saturation was dependent on the level of both factors (Fig. 5B).

Lung histopathology.

Normal airway epithelial composition of rats exposed to 21% O₂ is shown in Fig. 6A. The columnar airway epithelium consisted of a majority of ciliated cells with a smaller number of nonciliated cells present. Airways from rats exposed to 10.5% O₂ appeared similar to airways from rats exposed to 21% O₂ (Fig. 6B). Exposure to H₂S at both 21% and 10.5% O₂ caused damage to both the airways of the lung (Fig. 6, C and D). Epithelial cell swelling and exfoliation were observed in the airways following exposure to H₂S. The H₂S in 10.5% O₂ group exhibited less damage in the airways compared with the H₂S in 21% O₂ group (Fig. 6F).

Fig. 6.

Effect of H₂S and/or 10.5% O₂ on lung histopathology. Representative light micrographs of airways from rats exposed to 21% O₂ (A), 10.5% O₂ (B), H₂S + 21% O₂ (C), and H₂S + 10.5% O₂ (D). Lungs were collected 1 h after the cessation of exposures, inflation-fixed with formalin, embedded in paraffin, and stained with hematoxylin and eosin to detect pathological changes. Magnification bar = 30 μm. Histopathology scores for both large (E) and small (F) airways are shown in dot plots with mean provided as lines. Results from statistical analyses: ^aP < 0.005 compared with 21% O₂; ^bP = 0.006 compared with 10.5% O₂; and ^cP = 0.011 compared with H₂S + 21% O₂. Data represent results from n = 3–6 animals per treatment group.

Heart and liver injury.

Using plasma CK-MB as a marker of heart injury, we observed a significant increase in plasma CK-MB in the H₂S in 10.5% O₂ group compared with the other groups (Fig. 7A). Two-factor ANOVA showed significant H₂S and O₂ effects (Fig. 7C). Analysis also showed a significant interaction between H₂S and O₂, indicating that the effect on CK-MB was dependent on the level of both factors (Fig. 7C). Post hoc analyses demonstrated that the plasma levels of CK-MB were significantly higher in the H₂S in 10.5% O₂ group compared with 10.5% O₂ alone and H₂S in 21% O₂ groups (P < 0.001). In contrast, plasma ALT activity was elevated to the same level in both 10.5% O₂ groups with a significant O₂ effect (Fig. 7, B and C). Plasma ALT levels were significantly increased in the H₂S in 10.5% O₂ group compared with the H₂S in 21% O₂ group (P = 0.005). It should be noted that the H₂S- and low O₂-mediated elevations in plasma ALT compared with 21% O₂ control are small and likely represent no damage to liver tissue. These results suggest that the combination of H₂S with lower O₂ increases the potential toxicity of H₂S in the heart.

Fig. 7.

Effect of H₂S and/or 10.5% O₂ on heart and liver enzymes. Plasma levels of creatine kinase-MB (CK-MB; A) and alanine aminotransferase (ALT; B) were measured after exposures to assess effects on heart and liver injury, respectively. C: results presented are P values obtained from 2-factor ANOVA with H₂S effect (absent vs. present), O₂ effect (21% vs. 10.5% O₂ tension), and interaction (H₂S × O₂). Results from post hoc statistical analyses: ^aP < 0.005 compared with H₂S + 21% O₂ and ^bP < 0.001 compared with 10.5% O₂. Data represent means ± SE for n = 6 animals per treatment group.

Akt and GSK-3β signaling.

While phosphorylation status of Akt and GSK-3-β were affected by H₂S and/or 10.5% O₂ in the heart, activation of this pathway was unaffected in liver (data not shown). Phosphorylation of Akt was higher in the H₂S in 21% O₂ group than in the 21% O₂ control group (Fig. 8, A and C). Two-factor ANOVA showed a significant effect of O₂ on Akt phosphorylation (Table 2) with lower phosphorylation observed in the two 10.5% O₂ groups than in the two 21% O₂ groups (Fig. 8, A and C). A significant interaction was also observed for phosphorylated and total Akt protein levels (Table 2). This significant interaction was due to statistically significant differences between the H₂S in 21% O₂ and H₂S in 10.5% O₂ groups for both phosphorylated Akt and total Akt. Note that the statistical results obtained for phosphorylated Akt were mirrored when results were normalized to total Akt protein (Fig. 8G and Table 2, compare p-Akt to p-Akt/Akt results). Like Akt, the phosphorylation state of GSK-3β was primarily altered by O₂ tension (Fig. 8, B and D). Two-factor ANOVA showed a significant O₂ effect on GSK-3β phosphorylation (Table 2) with lower phosphorylation observed in the 10.5% O₂ groups than in the 21% O₂ group (Fig. 8, B and D). While H₂S alone numerically increased GSK-3β phosphorylation by 43% compared with the 21% O₂ control group this was not statistically significant (Table 2). The level of total GSK-3β protein was similar in all treatment groups (Fig. 8F). The statistical results obtained for phosphorylated GSK-3β were mirrored when results were normalized to total GSK-3β protein (Fig. 8F and Table 2, compare p-GSK-3β to p-GSK-3β/GSK-3β results). These results suggest that H₂S-dependent cytoprotection in the heart, mediated by Akt signaling, can be negatively influenced by oxygen.

Fig. 8.

Effect of H₂S and/or 10.5% O₂ on Akt and GSK-3β phosphorylation in heart. Representative immunoblots of phosphorylated (top) and total (bottom) Akt (A) and GSK-3β (B) in heart samples from rats exposed to 21% O₂, 80 ppm H₂S + 21% O₂, 10.5% O₂, and 80 ppm H₂S + 10.5% O₂. C, E, and G: densitometry analyses of Akt-P^Ser473, total Akt, and Akt-P normalized to total Akt protein, respectively. D, F, and H: densitometry analyses of GSK-3β-P^Ser9, total GSK-3β, and GSK-3β-P normalized to total GSK-3β protein, respectively. Results from post hoc statistical analyses: ^aP < 0.01 compared with 21% O₂; ^bP < 0.001 compared with 10.5% O₂; and ^cP < 0.05 compared with H₂S + 21% O₂. Data represent means ± SE for n = 4 animals per treatment group. Note that the results from the 2-factor ANOVA for these data are provided in Table 2.

Table 2.

P values from two-factor ANOVA on Akt and GSK-3β in heart following exposure to hydrogen sulfide (H₂S) and/or 10.5% O₂

	p-Akt	Akt	p-Akt/Akt	p-GSK-3β	GSK-3β	p-GSK-3β/GSK-3β
H2S effect	NS	NS	NS	NS	NS	NS
O2 effect	<0.0001	NS	0.0003	<0.001	NS	<0.001
H2S × O2	<0.0001	0.042	0.0076	NS	NS	NS

Results presented in this table are P values obtained from 2-factor ANOVA with H₂S effect (absent vs. present), O₂ effect (21% vs. 10.5% O₂ tension), and interaction (H₂S × O₂). p-Akt data are from Fig. 8C, total Akt from Fig. 8E, ratio of p-Akt/total Akt from Fig. 8G, p-GSK-3 β from Fig. 8D, total GSK-3β from Fig. 8F, and ratio of p-GSK-3β/total GSK-3β from Fig. 8H. NS, not significant (P > 0.05).

DISCUSSION

The beneficial effects of hypometabolism are currently being investigated as new medical treatments for trauma and disease. Hibernating mammals dramatically decrease metabolic rate and body temperature to survive long periods when food supply is low and demands to maintain body temperature in a cold environment are great (9, 40). Importantly, upon arousal hibernators recover without signs of tissue injury (41). Therefore, the ability to pharmacologically induce a safe hibernation-like state could provide an important new clinical approach to mitigate organ damage and improve survival in trauma patients. Several candidate hypometabolism-inducing agents including 2-deoxyglucose, 5-adenosine monophosphate (5′-AMP), 3-iodothyronamine, and H₂S have been studied in a variety of experimental animal models (see reviews in 5, 33). While results from these studies are encouraging, knowledge about the molecular mechanisms and signaling pathways involved in pharmacologically induced hypometabolism is limited. Therefore, the aim of this current study was to increase understanding regarding the hypometabolic effects induced by H₂S.

Recent studies show that mammals that do not hibernate (e.g., laboratory mice) can enter a reversible hypometabolic state when exposed to air containing low levels of H₂S (80 ppm). This “suspended animation” or hypometabolic state was first reported by Blackstone et al. (3), who reported that mice exposed to 80 ppm H₂S in room air (21% O₂) for 6 h had significantly decreased heart rates with body temperatures decreased to ambient temperature levels. Importantly, when H₂S exposure ceased and mice were returned to room air, these physiological measures returned to pre-H₂S exposure levels and mice behaved normally. Similarly, Zapol and colleagues (47) investigated the impact of ambient temperature on H₂S-induced hypometabolism. In these studies, they observed that mice breathing 80 ppm H₂S at a higher ambient temperature (35°C) to prevent hypothermia also had decreased heart and respiratory rates. This finding indicates that the cardiac and metabolic effects of H₂S are likely independent from changes in body temperature.

Previously, our laboratories reported that the biological actions of H₂S can be modulated by O₂ concentrations. Using isolated rat aorta preparations, H₂S concentrations that mediate rapid vasoconstriction at high O₂ levels (200 μM) cause vasodilation at lower O₂ levels (40 μM) (31). Herein, we extended this concept to determine whether the published hypometabolic responses elicited by H₂S in vivo (3, 47) may be influenced by changes in O₂ tension. We observed that the ability of inhaled H₂S to lower metabolic function in the rat is enhanced when administered at a lower O₂ tension (10.5%) than room air (21% O₂). Body temperature and ventilation rate were not decreased in rats breathing 80 ppm H₂S in 21% O₂ compared with previous studies in mice (3, 47). Only when rats were exposed to H₂S in 10.5% O₂ did we observe a significant decrease in body temperature compared with preexposure levels (Fig. 2). Similar results were observed on respiration rate (Fig. 4).

The ability of lower O₂ to induce H₂S-mediated hypometabolism in the rat may be related, in part, to the fact that O₂ decreases H₂S bioavailability. Previous studies have shown that H₂S spontaneously reacts with O₂ to form sulfide oxidation products (43, 44), which are less bioactive than H₂S and rapidly cleared from the body. Preliminary studies from our laboratory support this concept as H₂S disappearance occurred at a faster rate in normoxic than in hypoxic tissues (unpublished data). The impact of O₂ on H₂S activity may be important when considering potential molecular mechanisms responsible for H₂S-dependent hypometabolism. For example, several groups have proposed that mitochondria are a prime target of H₂S with reversible inhibition of cytochrome c oxidase mediating H₂S hypometabolism (12). Important to the present study is the observation that while H₂S binding at the heme a₃ site in cytochrome c oxidase is noncompetitive with O₂, mitochondria exposed to low O₂ tensions are more sensitive to H₂S inhibition (19, 25). Interestingly, Lee and colleagues (14) have reported that 5′-AMP injection in mice induces a reversible hypometabolic state in mice triggered by reduced O₂ transport by red blood cells. They show that red blood cell uptake of 5′-AMP increases intracellular levels of 2,3-bisphosphoglycerate and dexoyhemoglobin resulting in decreased oxygen transport, body temperature, heart rate, and oxygen consumption (V̇o₂). Similarly, we observed decreased hemoglobin saturation in the H₂S in 10.5% O₂ group (Fig. 5). Taken together, this work supports the concept that hypometabolism would be enhanced in rats breathing H₂S in reduced O₂ tensions.

The ability of H₂S to induce hypometabolism in rats at normoxia is mild, if not absent, compared with mice. Why mice are more sensitive to H₂S than rats is not known; however, others have shown insensitivity to H₂S-mediated hypometabolism in other mammalian species, especially those that are larger than mice (15, 19). Consequently, there could be a dilution effect linked to body mass. For example, a mouse has a higher ventilation rate to body mass such that a mouse inhales more H₂S for mass distribution relative to a rat. However, increasing the concentration of H₂S to 160 ppm did not enhance hypometabolism in rats (unpublished data). It is important to note that body temperature may not be the most suitable measure of metabolic rate in mammals, as body temperature largely remains constant even under active conditions with an elevated metabolism (18). For example, there are thermoregulatory mechanisms; e.g., increased skin blood flow, which prevent an elevation in body temperature during periods of increased metabolic rate (18). However, a decrease in metabolism will cause a decrease in body temperature because heat, like CO₂, is a product of mitochondrial substrate oxidation (39). Thus reduced metabolism will decrease body temperature under conditions of constant ambient temperature. In this study with H₂S, the drop in body temperature likely occurred because the energy (i.e., heat) generated under a suppressed metabolic rate failed to compensate for the heat loss into the environment. Importantly, H₂S-dependent thermogenic deficits will likely be more pronounced in smaller animals (mice) with a greater surface area-to-body mass ratio and kept in a cooler environment (3) compared with larger mammals (rats) kept under a warmer environment as in the present study. Therefore, a decrease in metabolism in the rat may not produce as a dramatic a decrease in body temperature as that observed in mice exposed to H₂S in 21% O₂ (3). Taking these considerations into account, body temperature may not be the most sensitive measure of metabolic rate for these types of studies. Other, more direct measures of metabolism, such as V̇o₂ and carbon dioxide production (V̇co₂) may reveal that H₂S in 21% O₂ did induce a significant hypometabolic state in the rat.

Altered susceptibility to H₂S may also be due to species-specific differences in the expression of H₂S detoxifying systems. To date, H₂S removal is believed to take place in mitochondria through a series of reactions using H₂S as a respiratory substrate while oxidizing it to thiosulfate (22, 32). This H₂S disposal mechanism has been called the sulfide quinone reductase system or SQR (22). Rhodanese (thiosulfate sulfurtranferase) is implicated to be a key component in the H₂S detoxification system in mitochondria. Rhodanese has high activity in mitochondria and, in addition to its function in cyanide detoxification, rhodanese catalyzes the transfer of sulfane sulfur between proteins and other thiol-containing compounds. Interestingly, the cyanide detoxification capacity of rhodanese is decreased under low O₂ tensions (25). Therefore, it is predicted that H₂S disposal by the SQR may be decreased under low O₂ effectively increasing H₂S concentrations locally in mitochondria and enhancing the hypometabolic effect of H₂S.

Relevant to this study is the fact that rhodanese activity in humans is lower than activity measured rats, (23, 49) with rats having rhodanese activity comparable to that measured in lug worms living in sulfide-rich mud (22). As a result, rats may be much less sensitive to H₂S, thus necessitating a lower O₂ tension to obtain hypometabolism. Thus, for a given species, the ratio of administered H₂S to O₂ tension may be an important factor determining the degree of H₂S responsiveness rather than simply the concentration of administered H₂S alone. Similarly, the hypometabolic effect of H₂S may not solely be due to rhodanese but also related to the ratio of cytochrome c oxidase to rhodanese. When considered together as enzymes competing for the same substrate (i.e., H₂S), the cytochrome c oxidase:rhodanese ratio may also contribute to species-specific responses to H₂S hypometabolism and organ-specific responses, as well. This concept is consistent with the possibility that high and low susceptibilities of mice and rats to H₂S, respectively, may be related to a higher cytochrome c oxidase:rhodanese ratio in mice and a lower ratio in rats. Future studies are needed to determine this ratio in mice and rats tissues and how this relates to sensitivity to the hypometabolic effect of H₂S.

While H₂S had little effect on body temperature and respiration when administered in 21% O₂, inhaled H₂S caused a significant decrease in heart rate at both O₂ tensions. Heart rate remained depressed and did not return to pre-H₂S exposure levels after the 1 h recovery period at room air. Moreover, we found a significant increase in the plasma activity of CK-MB in the H₂S in 10.5% O₂ group, indicating heart damage. In contrast, liver enzymes were little affected by H₂S. It is possible that increased sensitivity of the heart to H₂S is due to a higher dose of H₂S exposure and lower levels of H₂S detoxification enzymes (e.g., rhodanese) in the heart compared with liver (23). Also, because the heart has a higher bioenergetic demand than liver, it is expected that the ratio of cytochrome c oxidase:rhodanese is higher in the heart than in the liver, which may contribute, in part, to increase sensitivity and potential cardiotoxicity of H₂S under low O₂ conditions.

We also observed lung damage in rats exposed to H₂S in 21% and 10.5% O₂. This has not been reported in published hypometabolism studies using inhaled H₂S. Epithelial swelling and exfoliation were the predominant pathological events observed in the lungs following H₂S exposure. Because rats were exposed to H₂S through inhalation, local concentrations of H₂S in the lung epithelial lining fluid and epithelial cells during the exposure would be expected to be higher than in distal tissues. This could lead to cell death and exfoliation possibly due to the effect of H₂S to reduce disulfide bonds, a critical structural component of many cellular proteins. These H₂S-mediated posttranslational modifications of protein thiols may also be involved in heart injury. Respiratory gas exchange was not compromised in rats, as hemoglobin saturation returned to preexposure values after H₂S exposure. Thus it is likely that H₂S exposure did not compromise basal lung functions in this model, but the possibility that H₂S could sensitize the lung to inflammatory injury or secondary insults is a potential cautionary note for inhaled H₂S therapy.

While H₂S cytoprotection has been demonstrated in numerous studies, it is important to restate that H₂S and other sulfide compounds have a narrow therapeutic window (4). Indeed, studies by Wang and colleagues have shown a toxic proapoptotic effect of H₂S in several models (51). Results presented herein also support a potential damaging effect of H₂S on the lung and heart, even while inducing the proposed protective hypometabolic state. In contrast to most investigations that use single bolus administration of sulfide chemicals (NaHS or Na₂S), this study exposed rats to H₂S gas for 6 h. Thus the cumulative dose may, in fact, be higher than that experienced in other studies assessing the effect of H₂S on similar metabolic and signaling pathways. Importantly, while we were able to reproduce findings from other studies showing activation of the prosurvival Akt signaling pathway in heart from the H₂S in 21% O₂ group, we found decreased phosphorylation in hearts of rats exposed to H₂S in 10.5% O_2. This signaling block in the H₂S in 10.5% O₂ group may have contributed to the heart injury observed; however, as Akt is involved in multiple cellular processes, decreased Akt phosphorylation should not be equated to cytotoxicity. Similarly, GSK-3β function is multifaceted and regulated by multiple kinases, not just Akt (42). In support of these new findings, Akt phosphorylation is decreased in heart, muscle, and liver of hibernating ground squirrels without a change in total Akt content (1, 6). These findings strongly suggest that processes mediated by Akt are suppressed during hypometabolic states for energy conservation. Moreover, because low O₂ and H₂S exposure are known to independently inhibit mitochondrial respiration (13), it is also not surprising that Akt and GSK-3β phosphorylation were diminished under these conditions where there will be less ATP available for phosphorylation of cellular signaling molecules. While it is possible that other metabolic or nucleotide imbalances may be responsible for H₂S-dependent regulation of Akt, there is little evidence in the literature delineating what pathways upstream of Akt are impacted by H₂S. Manna and Jain (36) showed that Na₂S and its precursor l-cysteine increases cellular levels of phosphatidylinositol 3,4,5-triphosphate (PIP3) in adipocytes, which activates Akt. Whether PIP3-mediated signaling in heart is altered by inhaled H₂S is not known. Similarly, little is known regarding what other signaling pathways are impacted by H₂S-mediated hypometabolism in vivo. Inhibition of mitochondrial function by H₂S is predicted to increase the AMP:ATP ratio, which could lead to activation of AMP-activated protein kinase (AMPK), a key sensor of cellular energy homeostasis. Once activated allosterically by AMP and via phosphorylation by upstream kinases, AMPK functions to turn-off ATP-consuming pathways and turn-on ATP producing pathways (45). While activation of brain AMPK was observed by Na₂S in a model of ischemia-reperfusion (37), reports of AMPK activation during mammalian hibernation are conflicting (21, 24). Therefore, future studies need to be directed at investigating the effects of H₂S-dependent hypometabolism on key energy sensing signaling pathways with special consideration given to understanding the impact of O₂ tension. In summary, the biological outcomes of H₂S, mediated by phosphorylation of these key signaling molecules and others are likely to be influenced by O₂ and effects on mitochondrial metabolism. Thus careful consideration of the local tissue and/or cellular environment will likely dictate, in part, whether H₂S exposure mediates a cytoprotective or cytotoxic effect in vivo.

Concluding remarks.

A growing body of evidence supports H₂S and other sulfide-containing chemicals as being potential therapeutics by their ability to regulate prosurvival signaling pathways and induce a hypometabolic-like state. The results presented in this study demonstrate that the biological actions and outcomes of H₂S can be influenced by key environmental factors; specifically O₂. We observed that unlike previous studies in mice, rats were largely insensitive to H₂S-induced hypometabolism unless O₂ tension was decreased. This highlights an important species difference, perhaps mediated by differences in activities of key enzymes involved in H₂S detoxification or those which use H₂S as a substrate. This concept has ramifications for H₂S use and efficacy in humans as biological outcomes obtained in mice or rats may not be easily translated to humans. These studies also demonstrate damaging effects of H₂S on lung and heart and call into question inhaled H₂S as an appropriate route of administration. Moreover, the effect of H₂S-mediated on key signaling molecules like Akt is also influenced by O₂ with differential activation seen with H₂S in 21% and 10.5% O₂. Taken together, the results from this study should serve as a cautionary note to investigators regarding potential toxicity of H₂S. These data also highlight the need for increased investigations focused on determining the contribution of other metabolic or environmental factors to H₂S-induced hypometabolism. Continued investigations will likely lead to improved understanding of this unique signaling molecule and its newly discovered physiological actions.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants T32 HL-007918 (A. Stein) and R01-HL-092857 (J. E. Doeller/S. M. Bailey).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: A.S., E.M.P., R.P.P., D.W.K., J.E.D., and S.M.B. conception and design of research; A.S., Z.M., and J.P.M. performed experiments; A.S., M.V.F., J.E.D., and S.M.B. analyzed data; A.S., M.V.F., E.M.P., D.W.K., J.E.D., and S.M.B. interpreted results of experiments; M.V.F. provided analyses of lung histopathology; A.S., Z.M., and S.M.B. prepared figures; A.S. and S.M.B. drafted manuscript; A.S., J.P.M., M.V.F., E.M.P., R.P.P., J.E.D., and S.M.B. edited and revised manuscript; A.S., J.E.D., and S.M.B. approved final version of manuscript.