Is pharmacological, H2S-induced 'suspended animation' feasible in the ICU?

By definition, ‘suspended animation’ is a hypometabolic state characterized by the “the slowing of life processes by external means without termination” [1]. Various mammalian species are capable of nearly completely shutting down their vital functions in order to survive otherwise lethal environmental conditions, such as prolonged impairment of O2 supply and/or extreme temperatures. First described and studied in patients as “hibernation artificielle” induced by the so-called “cocktail lytique” during the Indochina war in the early 1950 s, for obvious reasons the concept of inducing such a hypometabolic condition has attracted special interest in intensive care and emergency medicine. Originally, organ-protection, in particular for the central nervous system (CNS), was demonstrated when suspended animation was induced by rapidly cooling experimental animals to core body temperatures of about 10–15 °C using ice-cold infusions and/or cardiopulmonary bypass (CPB). Given the potential undesired adverse effects of hypothermia per se, e. g., metabolic acidosis, coagulopathy, prolonged inflammation, and impaired host defense, any pharmacological measure allowing for a therapeutic on-demand induction of suspended animation would be of particular interest. Moreover, more recently, it was suggested that the reduced visceral organ function present in critically ill patients and/or after overwhelming hyperinflammation could be referred to as an adaptive mechanism to maintain ATP-homeostasis due to reduced energy expenditure rather than to irreversible organ failure [2]. A landmark paper by Blackstone et al. produced much excitement among researchers in the field of shock and critical illness: These authors demonstrated that mice inhaling hydrogen sulfide (H2S) reversibly decreased their energy expenditure, which was associated with a fall in core temperature [3]. In the meantime, numerous pre-clinical studies have been published on the possible organ-protective effects of H2S, the available data being equivocal depending on the model used and the type of shock investigated. In this context in particular, the impact of H2S effects on energy metabolism remains a matter of debate. Therefore, the present chapter reviews the available data on H2S-induced on-demand hypometabolism, and its relation (directly as well as a possible consecutive drop in body temperature) to organ-protective properties of H2S.


Introduction
By defi nition, 'suspended animation' is a hypometabolic state characterized by the "the slowing of life processes by external means without termination" [1]. Various mammalian species are capable of nearly completely shutting down their vital functions in order to survive otherwise lethal environmental conditions, such as prolonged impairment of O 2 supply and/or extreme temperatures. First described and studied in patients as "hibernation artifi cielle" induced by the so-called "cocktail lytique" during the Indochina war in the early 1950 s, for obvious reasons the concept of inducing such a hypometabolic condition has attracted special interest in intensive care and emergency medicine. Originally, organ-protection, in particular for the central nervous system (CNS), was demonstrated when suspended anima tion was induced by rapidly cooling experimental animals to core body temperatures of about 10-15 °C using ice-cold infusions and/or cardiopulmonary bypass (CPB). Given the potential undesired adverse eff ects of hypothermia per se, e. g., metabolic acidosis, coagulopathy, prolonged infl ammation, and impaired host defense, any pharmacological measure allowing for a therapeutic on-demand induction of suspended animation would be of particular interest. Moreover, more recently, it was suggested that the reduced visceral organ function present in critically ill patients and/or after overwhelming hyperinfl ammation could be referred to as an adaptive mechanism to maintain ATP-homeostasis due to reduced energy expenditure rather than to irreversible organ failure [2]. A landmark paper by Blackstone et al. produced much excitement among researchers in the fi eld of shock and critical illness: Th ese authors demonstrated that mice inhaling hydrogen sulfi de (H 2 S) reversibly decreased their energy expenditure, which was associated with a fall in core temperature [3]. In the meantime, numerous pre-clinical studies have been published on the possible organ-protective eff ects of H 2 S, the available data being equivocal depending on the model used and the type of shock investigated. In this context in particular, the impact of H 2 S eff ects on energy metabolism remains a matter of debate. Th erefore, the present chapter reviews the available data on H 2 Sinduced on-demand hypometabolism, and its relation (directly as well as via a possible consecutive drop in body temperature) to organ-protective properties of H 2 S.

Rodent models
In their above-mentioned murine study, Blackstone et al. demonstrated, in awake, spontaneously breathing animals, that exposure to incremental, sub-toxic gaseous H 2 S concentrations (20-80 ppm) dose-dependently decreased energy expenditure within a few minutes as assessed by calorimetric measurement of whole-body O 2 uptake and CO 2 production. Th is fall in metabolic activity was associated with bradypnea and consecutive hypothermia, with core temperature falling to levels close to ambient values [3]. After washout of H 2 S, all these metabolic and cardiopulmonary eff ects were completely reversible, and animals showed no apparent sequelae. Subsequently, Volpato et al. reported that the reduced metabolic activity went along with bradycardia and, consequently, reduced cardiac output, whereas blood pressure and stroke volume remained unaff ected [4]. Maintaining normothermia by external warming attenuated the metabolic depressor eff ect, but did not completely blunt the cardiovascular response [4]. Various other rodent models confi rmed these observations: Inhaling gaseous H 2 S [5]- [12] and infusing the soluble sulfi de salts, NaSH or Na 2 S [6], [13], [14], also induced a reversible reduction in energy expenditure with a subsequent fall in core temperature. Under stress conditions resulting from injurious mechanical ventilation [8], [13], ischemia/reperfusion [7], [9], [12], endotoxin challenge [11], or bacterial sepsis [14], this eff ect coincided with attenuation of lung [8], [12]- [14], liver [9], kidney [7] and heart [12] injury. Most importantly, survival was improved after otherwise lethal stress states, e. g., hemorrhagic shock [6] and exposure to hypoxic hypoxia (fraction of inspired O 2 [FiO 2 ] 5 %) [5]. In addition to anti-oxidant, anti-infl ammatory, and antiapoptotic properties, H 2 S was associated with better maintenance of mitochondrial integrity and function [7], [15], [16]: Treatment with either gaseous H 2 S treatment or injection of Na 2 S prevented mitochondrial swelling, loss of crypts [7], [15], and, at least under hypothermic conditions, outer mitochondrial membrane rupture as documented by the lack of responsiveness of the mitochondrial respiratory chain to stimulation with exogenous cytochrome c [16].
It should be noted that most of the above-mentioned murine data originate from experiments in awake, spontaneously breathing animals. Consequently, the role of anesthesia for a putative H 2 S-induced suspended animation remains unclear. Currently, scarce literature is available comparing the eff ects of anesthesia and H 2 S per se. In spontaneously breathing mice, Li et al. demonstrated that H 2 S (80 and 250 ppm) produced the same metabolic depression as 0.3 and 0.9 % of isofl urane, respectively, however, without any anesthesia-related muscle atonia. Strikingly, when combining these two interventions, H 2 S even antagonized the isofl urane-induced metabolic depression [17]. Finally, in mecha nically ventilated mice under continuous intravenous (i.v.) anesthesia, the metabolic depressor eff ect of H 2 S was completely blunted when normothermia was main tained [16].

Large animal species and humans
Any metabolic depressant property of H 2 S seems to be dependent on the animal size: In rats the H 2 S-induced decrease in O 2 uptake was several-fold lower than in mice [18]. In larger species (swine, sheep), various authors failed to confi rm any H 2 S-related reduction in metabolic activity at all, regardless of whether inhalation of gaseous H 2 S or injection of sulfi de salts were studied [19]- [22]. Moreover, in sheep, Derwall et al. [23] demonstrated that during administration of gaseous H 2 S via an extracorporeal, veno-arterial membrane oxygenator to avoid any airway mucosa damage related to the gas inhalation [24], [25], whole body O 2 uptake, CO 2 production, and cardiac output remained within the physiological range. At the highest doses administered (300 ppm), H 2 S did not aff ect calorimetric energy expenditure either, but caused pulmonary vasoconstriction associated with arterial hypotension and metabolic acidosis [23]. Finally, in human volunteers, inhalation of 10 ppm H 2 S during exercise decreased O 2 uptake, and this eff ect was referred to a toxic reduction in maximal aerobic capacity rather than to a regulatory eff ect on mitochondrial respiration, as evidenced by a tendency for muscle lactate to increase and citrate synthase activity to decrease [26]. Consequently, it was questioned whether any therapeutic potential of the H 2 S-induced "suspended animation"-like hypometabolism observed in mice and rats could be transferred to the clinical setting [27], [28]. On the other hand, when external measures to prevent hypothermia were withheld, Na 2 S-related organ-protection after kidney ischemia/reperfusion-injury [29] or hemorrhage and resuscitation [30] coincided with a progressive decrease in core temperature ( Figure 1). Moreover, in the latter experiments, immediate post-mortem liver tissue mitochondrial activity showed a tendency towards both reduced oxidative phosphorylation and maximal O 2 uptake in the uncoupled state, and, in particular, a signifi cantly decreased "leak respiration", i. e., the respiratory activity necessary to compensate for the proton leakage, slipping, and cation-exchange along the inner mitochondrial membrane ( Figure 2). In other words, H 2 S supplementation under these conditions provided protective reduction rather than toxic inhibition of cellular respiration.
How can these diverging fi ndings be reconciled? Under stress conditions, e. g., in response to hypoxia or circulatory shock, small rodents can reduce their energy expenditure as a result of decreased 'non-shivering thermogenesis' [31], due to modulation of the uncoupling protein-1, mostly in the brown adipose tissue [32]. In these species, non-shivering thermogenesis represents a large proportion of total O 2 uptake, which can be rapidly decreased without aff ecting ATP formation [31]. Th is response is independent of any pharmacological intervention, and represents a unique protective adaptation present in numerous mammals [31] and even in humans, e. g., in neonates and during cold acclimatization [32]. However, due to the high area/volume ratio and, consequently, the higher heat dissipation, it is inversely related to body size [31], i. e., to the ratio of O 2 consumption and body weight. Two phenomena support this latter notion: i) No matter the species, newborns present with more pronounced hypoxia-induced hypometa bolism than do adults [31]; ii) when the ratio of O 2 consumption and body weight per se is low (e. g., in adults of larger species), normoxic O 2 uptake (e. g., during exercise [31]) may be associated with hypoxiainduced hypometabolism. Hence, if possible at all, achieving a suspended animation-like status in larger animals and humans will be more diffi cult and require much more time because of the small surface area/mass ratio: In fact, in anesthetized and mechanically ventilated swine, after four hours of Na 2 S infusion whole body O 2 uptake and CO 2 production started to decrease, subsequently resulting in a moderate decrease in core temperature at ten hours of drug infusion [23] (Fig. 1).
No matter the current debate on the feasibility of pharmacological induction of whole body suspended animation in larger animals, inducing hypometabolism to hibernate isolated organs and, thereby, prolong their tolerance against tissue ischemia or hypoxia remains an attractive option, in particular for organ transplantation.
Numerous studies in rodents have demonstrated that H 2 S administration improved kidney, liver heart, and lung function and attenuated histological damage after orthotopic organ transplant. Th is benefi cial eff ect of H 2 S administration (NaSH 0.5 mmol/l over 10 minutes before and immediately after initiation of reperfusion) was confi rmed in isolated porcine kidneys ex vivo undergoing normothermic reperfusion with autologous blood after 25 minutes of warm ischemia and subsequently 18 hours of storage at 4 °C [33].  [29]: Dark blue squares, vehicle n = 10; blue circles, Na 2 S n = 9; all data are mean ± SD, § designates p < 0.05 between groups); (b) hemorrhage and resuscitation (data are adapted from [30]: black squares, vehicle n = 14; dark blue squares, Na 2 S started two hours before hemorrhage, n = 10; light blue squares, Na 2 S started simultaneously with hemorrhage, n = 11; blue triangles, Na 2 S started immediately after hemorrhage, n = 10; all data are mean ± SD, § designates p < 0.05 'simultaneous' treatment vs. vehicle). Note that in both experimental series at least four hours of drug infusion were necessary to achieve a signifi cant decrease in body temperature.  Hypothermia Equivocal data are available whether hypothermia, caused by a possible H 2 S-related fall in energy expenditure and/or due to external cooling measures, assumes importance for organ protection achieved during H 2 S administration. Inhaling H 2 S prior to myocardial ischemia at concentrations that had no metabolic depressant eff ect (10 ppm) attenuated organ damage, but to a lesser degree than concentrations that reduced energy expenditure (100 ppm) [12], suggesting that hypometabolism may indeed enhance the organ-protective properties of H 2 S. Of note, in that study as well as in others demonstrating H 2 S-related organ production coinciding with reduced metabolic activity, hypothermia was prevented [5], [7], [9], [14], [15] in order to elucidate the impact of a simultaneous drop in core temperature. Moreover, organ protection and improved survival were also shown to be in part [12], [13], [15], [34], [35] or even completely [8], [11], [36], [37] independent of any H 2 Sinduced metabolic depression at all. Finally, data obtained in large animal (swine or sheep) models of shock resulting from ischemia/reperfusion [29], [38]- [42], hemorrhage and resuscitation [30], or burn injury [36] also suggested that the benefi cial eff ects of infusing Na 2 S were at least in part independent of metabolic depression and/or a fall in core temperature. Hence, any moderate hypothermia observed simultaneously with H 2 S-induced organ-protection may also be due to attenuation of systemic infl ammation rather than to reduced energy expenditure per se. In other words, such fi ndings raise a 'chicken and egg' problem, which can be attributed to the so-called Q10 eff ect, i. e., the two to three fold reduction in all chemical reactions and thus metabolism associated with a 10 °C-reduction of body temperature [31]: As an example, during otherwise lethal porcine hemorrhage, therapeutic hypothermia was associated with reduced concentrations of pro-infl ammatory cytokines [43]. Th e potential of H 2 S acting as a metabolic depressant in larger species independent of any anti-infl ammatory and antioxidant property still remains unsettled: In the abovementioned swine study showing an H 2 S-induced drop in O 2 uptake and CO 2 production as well as a consecutive moderate fall in core temperature, animals underwent a short period of aortic occlusion, which did not cause any increase in the blood levels of pro-infl ammatory cytokines or markers of oxidative and nitrosative stress [23].
Irrespective of the question as to whether or not there is cause-eff ect relationship between H 2 S-related organ protection and coinciding hypometabolism and/or hypothermia, hypothermia does assume importance for H 2 S-induced eff ects on substrate utilization and mitochondrial function. It is well-established that H 2 S toxicity is due to inhibition of mitochondrial respiration resulting from blockade of the complex IV of the respiratory chain, i. e., cytochrome c oxidase [44]. When compared to normothermia, hypothermia (27 °C) increased the Na 2 S concentrations necessary to induce inhibition of mitochondrial respiratory activity (from < 1 μM to 2-4 μM), and nearly doubled the Na 2 S con centrations required for a 50 % reduction in mitochondrial respiratory activity [16], [45]. Hypothermia may also infl uence the eff ect of H 2 S on substrate utilization and, thereby, may even improve the yield of the mitochondrial respiration: In anesthetized and ventilated mice, during normothermia, inhaling 100 ppm H 2 S did not aff ect endogenous glucose production (as calculated from the rate of appearance of 1,2,3,4,5,6-13 C 6 -glucose during continuous i.v. isotope infusion), whole body CO 2 production, or direct, aerobic glucose oxidation rate (as derived from VCO 2 and the expiratory 13 CO 2 / 12 CO 2 ratio) (Fig. 3). However, under hypothermic (core temperature 27 °C) conditions, the rate of direct, aerobic glucose oxidation increased, suggesting a shift toward preferential carbohydrate utilization [16] (Fig. 3). Such a switch in fuel utilization is associated with an improved yield of oxidative phosphorylation: Th e ATP synthesis/O 2 consump tion ratio is higher for glycolysis than for βoxidation, because nicotinamide adenine dinucleotide (NADH) as an electron donor provides three coupling

phosphorylation (OxPhos); and maximal O 2 uptake in the uncoupled state (ETS) in immediate post-mortem liver biopsies of animals undergoing hemorrhage and resuscitation treated with vehicle and Na 2 S infusion started simultaneously with the initiation of blood withdrawal.
For protocol details, see [30]. All data are mean ± SD of O 2 uptake in pmol/s/mg tissue; dark blue columns: vehicle, n = 9; light blue columns: Na 2 S, n = 10. sites rather than just two from FADH 2 [46]. During cecal ligation and puncture-induced septic shock, the metabolic eff ects of inhaled H 2 S partially disappeared: Inhaled H 2 S aff ected neither the sepsis-induced metabolic acidosis [34] nor glucose utilization (Figure 3), nor the responsiveness to stimulation with exogenous cytochrome c oxidase. Nevertheless, H 2 S did normalize the sepsis-related increase in "leak respiration" -which was less pronounced during hypothermia -thus allowing for better maintenance of mitochondrial function (Figure 4). It is unclear whether the lack of eff ect of H 2 S on the mitochondrial respiratory chain was due to the septic challenge per se and/or to the ongoing treatment: During sepsis, all mice needed continuous i.v. norepinephrine to achieve target hemodynamics characterized by a normotensive and hyperdynamic circulation. In turn, norepinephrine incubation was associated with impairment of tissue mitochondrial respiration.

Timing and dose
No matter the importance of hypometabolism for the organ-protective properties of H 2 S administration per se, Data for sham-surgery are adapted from [16]. All data are mean ± SD, n = 8-11 per group, # designates p < 0.05 vs. normothermia, § designates p < 0.05 CLP vs. sham, $ designates p < 0.05 H 2 S vs. vehicle.  c the questions of timing and -due to the potential toxic inhibition of mitochondrial respiration -dosing of H 2 S remain unsettled. Clearly, there are plenty of data available showing that inhalation of H 2 S gas and/or the injection of NaSH or Na 2 S can prevent organ damage when administered prior to or at least simultaneously with the initiation of shock. However, the very few studies comparing a pre-and post-treatment design in mice showed marked reduction [7] or even complete disappearance [37] of the protective potency. In swine undergoing long-term hemorrhage and resuscitation, the results were even more curious [47]: Primed-continuous Na 2 S administration (initial bolus of 0.2 mg/kg, followed by 1 mg/kg/h over 12 hours of resuscitation) improved survival when compared to vehicle (survival: 71 %), regardless of whether the Na 2 S infusion was started two hours before (pre-treatment: survival 100 %) or simultaneously with (survival 91 %) the initiation of blood withdrawal, or at the start of re-transfusion of shed blood (post-treatment: survival 90 %) [30]. However, a signifi cant decrease in core temperature (Fig. 1b) and organ protection were only present in the group of animals treated simultaneously with the initiation of hemorrhage. Apparently, both the cumulative H 2 S dose as well as the rate of its generation assume importance for the eff ects on metabolism and organ protection, in particular under low fl ow conditions and/or circulatory shock: In swine undergoing cardiac arrest, primedcontinuous Na 2 S (0.3 mg/kg followed by 0.3 mg/kg/h over two hours) injected one minute after the start of cardiopulmonary resuscitation (CPR) reduced blood pressure and cardiac output during early resuscitation [21]. Increasing the Na 2 S dose (1.0 mg/kg followed by 1.0 mg/kg/h) was associated with impaired neurological recovery. Even injection of comparable total amounts may have markedly diff erent eff ects due to the diff erent rate of H 2 S generation: In vitro slow H 2 S release from the H 2 S donor GYY4137 exerted anti-infl ammatory and -apoptotic eff ects, whereas short-term, high peak free sulfi de levels resulting from incubation with NaSH induced the opposite response [48]. In vivo, this concept was confi rmed in swine undergoing myocardial ischemia/ reperfusion injury: A primed-continuous Na 2 S infusion was superior to bolus injection [39].

Conclusions
Th e concept of "buying time in suspended animation" [49] has been discussed in the literature for more than a century. Originally induced by rapid external body cooling, any pharmacological measure allowing for a therapeutic, on demand induction of 'suspended animation' is of particular interest because of the undesired side eff ects of hypothermia per se. Th erefore, the landmark paper demonstrating that inhaling H 2 S could induce a reversible, suspended animation-like hypometabolism [3], produced much excitement among researchers in the fi eld of shock and critical illness. Numerous pre-clinical studies are currently available on H 2 S-related organ protection, but the eff ects on energy metabolism remain a matter of debate. In this context, the well-established toxic blockade of cytochrome c oxidase by H 2 S may assume particular importance. Most studies so far suggest that the benefi cial eff ects of H 2 S are at least in part independent of an H 2 S-induced metabolic depression and, in particular, any decrease in core temperature. However, other data suggest that H 2 S-related hypometabolism may enhance the organ-protective properties. Th e mechanism behind H 2 S-induced hypometabolism is still not fully understood, and, moreover, the feasibility of H 2 S-induced suspended animation in larger animals has been questioned. Clearly, if possible at all, achieving a suspended animation-like status in larger animals and humans will be more diffi cult and require much more time because of the small surface area/mass ratio. Again the available data are equivocal, suggesting that at least hibernating isolated organs remains an option. Even in larger species, data on the eff ects of H 2 S on mitochondrial function and morphology suggest that its supplementation during circulatory shock provides  protective reduction rather than toxic inhibition of cellular respiration. Finally, according to the currently available literature, neither inhalation of gaseous H 2 S nor injection of the soluble sulfi de salts, NaSH or Na 2 S, is likely to become part of clinical practice because of damage to the airway mucosa and possibly toxic peak sulfi de concentrations, respectively, but slow H 2 Sreleasing molecules may enable these limitations to be overcome. Hence, there is "nothing rotten about hydrogen sulfi de's medical promise" [50], and H 2 S clearly remains a "hot molecule" [51] in the fi eld of research for a possible pharmacological induction of suspended animation-like hypometabolism.