Mitochondria and carbon monoxide: cytoprotection and control of cell metabolism – a role for Ca²⁺?

Abstract

Carbon monoxide (CO) is an endogenously produced gasotransmitter with important biological functions: anti‐inflammation, anti‐apoptosis, vasomodulation and cell metabolism modulation. The most recognized cellular target for CO is the mitochondria. Physiological concentrations of CO generate mitochondrial reactive oxygen species (ROS), which are signalling molecules for CO‐induced pathways. Indeed, small amounts of ROS promote cytoprotection by a preconditioning effect. Furthermore, CO prevents cell death by limiting mitochondrial membrane permeabilization, which inhibits the release of pro‐apoptotic factors into the cytosol; both events are ROS dependent. CO also increases the ability of mitochondria to take up Ca²⁺. Mitochondrial metabolism is modulated by CO, namely by increasing TCA cycle rate, oxidative phosphorylation and mitochondrial biogenesis, which, in turn, increases ATP production. CO's modulation of metabolism might be important for cellular response to diseases, namely cancer and ischaemic diseases. Finally, another cytoprotective role of CO involves the control of Ca²⁺ channels. By limiting the activity of T‐type and L‐type Ca²⁺ channels, CO prevents excitotoxicity‐induced cell death and modulates cell proliferation. Several questions concerning Ca²⁺ signalling, mitochondria and CO can be asked, for instance whether CO modulation of cell metabolism would be dependent on the mitochondrial Ca²⁺ uptake capacity, since small amounts of Ca²⁺ can increase mitochondrial metabolism. Whether CO controls Ca²⁺ communication between mitochondria and endoplasmic reticulum is another open field of research. In summary, CO emerges as a key gasotransmitter in the control of several cellular functions of mitochondria: metabolism, cell death and Ca²⁺ signalling.

Abbreviations

ANT

adenine nucleotide translocator protein

CO

carbon monoxide

CORM

CO‐releasing molecule

HO

haem oxygenase

MCU

mitochondrial calcium uniporter

MMP

mitochondrial membrane permeabilization

ROS

reactive oxygen species

Biology of carbon monoxide

Carbon monoxide (CO) was first identified in human exhaled breath in 1949 by Sjöstrand and colleagues (Sjostrand, 1949). Only two decades later, endogenous production of CO was clarified through the identification of the enzyme haem oxygenase (HO), which is able to break down the haem group (Tenhunen et al. 1968). In mammals two isoforms of HO have been described: the inducible HO‐1 and the constitutively expressed HO‐2. Along with CO, biliverdin and iron are produced (Ryter et al. 2006). Furthermore, HO activity is closely related to the maintenance of tissue homeostasis and cytoprotection (Gozzelino et al. 2010). Only during the 1980s was it found that CO can be a vasomodulator molecule in isolated heart (McFaul & McGrath, 1987) and vascular smooth muscle cells (Lin & McGrath, 1988). At the start of the new millennium, endogenous and exogenous CO started to be widely explored as a vasomodulator and anti‐inflammatory and anti‐apoptotic agent (Brouard et al. 2000; Otterbein et al. 2000; Motterlini et al. 2002). Since then, a great deal of literature has been published on the biological activities of this gasotransmitter, as well as its potential clinical use with patent applications (for review see Ryter, 2006; Bannenberg & Vieira, 2009; Motterlini & Otterbein, 2010; Queiroga et al. 2014). Several strategies for administrating CO in a more physiologically relevant manner have been developed through advances in the chemistry of CO‐releasing molecules (CORMs), small organic or organometallic molecules able to deliver CO in a controlled way (Romão et al. 2012).

Dual role for CO in mitochondria: generation of small amount of ROS and anti‐oxidant effect

Generation of reactive oxygen species (ROS) is a key signalling event in CO‐induced cellular pathways (Bilban et al. 2008). CO promotes mitochondrial ROS production, which is important for promoting anti‐inflammatory responses (Zuckerbraun et al. 2007), for stimulating mitochondrial biogenesis (Suliman et al. 2007 b) and for protection against cell death (Vieira et al. 2008; Queiroga et al. 2010). Furthermore, the anti‐inflammatory role of CO is lost in ρ⁰ cells (cells without functional mitochondria), as well as in macrophages where mitochondrial complex III has been chemically inhibited (Zuckerbraun et al. 2007). Therefore, CO‐induced ROS generation and signalling leads to activation of cellular endogenous mechanisms of defence, which are linked to preconditioning and cytoprotection (Bilban et al. 2008). CO is claimed to act by preconditioning activation in several distinct models: cardioprotection (Stein et al. 2012), neuroprotection (Vieira et al. 2008; Queiroga et al. 2012) and anti‐inflammation (Sun et al. 2008). Taking this evidence together, many biologically beneficial functions of CO are signalled by small amounts of mitochondrial ROS.

On the other side of the coin, CO also appears as an anti‐oxidant molecule, and in particular it has been described as acting as a mitochondrial mild‐uncoupling factor. Mitochondrial mild uncoupling can avoid the excessive ROS production, in particular whenever mitochondria are under unfavourable conditions. Administration of low amounts of CORM‐3 increased mitochondrial respiration with an associated decrease in mitochondrial potential (Lo Iacono et al. 2011). This CO‐induced uncoupling effect was reversed by an inhibitor of complex II (malonate) and by inhibitors of uncoupling proteins or adenine nucleotide transporters. Moreover, CO limits hydrogen peroxide production, but only when mitochondrial respiration is initiated by complex II, by inhibiting the reverse electron transfer at the level of complex I (Lo Iacono et al. 2011). These experimental conditions are not physiological, but these studies have made an important contribution for disclosing mitochondrial CO mechanisms. In a more detailed study, the same authors have suggested that the uncoupling effect on mitochondrial respiration of CORM‐3 occurs by direct interaction with the mitochondrial phosphate carrier (PiC). This interaction increases the phosphate and proton concentrations in the matrix space, leading to mitochondrial swelling, reduction of mitochondrial membrane potential and also increased O₂ consumption. Another phosphate carrier, dicarboxylate carrier (DIC), may be involved, but with a regulator role and not as the main CO target (Long et al. 2014). Still, one cannot discard the possibility that the anti‐oxidant activity of CO is triggered by ROS generation at low levels, via activation of cellular anti‐oxidant machinery.

Carbon monoxide involved in mitochondrial membrane permeabilization and cell death control

Mitochondria play a central role in many cellular functions, from energy status to cell fate. Mitochondrial membrane permeabilization (MMP) is one of the earliest events in the case of disruption of intracellular homeostasis and a point of no return in control of cell death (Galluzzi et al. 2012). Indeed, several factors are released from the intermembrane space into the cytosol triggering different cell processes, such as apoptosis and regulated necrosis. High intracellular concentrations of Ca²⁺ and its overload in mitochondria is one of the most described inducers of MMP and consequently cell death (Kroemer et al. 2007). The first evidence showing the involvement of Ca²⁺ and MMP was published in the 1990s when permeability transition pore (PTP) was firstly characterized and Ca²⁺ was demonstrated to permeabilize the inner membrane by binding on its matrix side (Zoratti & Szabo, 1995). Then the direct implication of Ca²⁺‐induced PTP opening and cell death were largely described (Zamzami et al. 1996; Yang & Cortopassi, 1998; Marzo et al. 1998). Moreover, although it has been accepted for more than 50 years that mitochondria are important organelles in Ca²⁺ buffering and signalling, only in 2011 was the mitochondrial calcium uniporter (MCU) identified (Baughman et al. 2011; De Stefani et al. 2011). This finding opened new avenues in the research field of Ca²⁺, mitochondria and cell death.

Concerning CO and MMP, Piantadosi and co‐workers observed that rats continuously exposed to 50 p.p.m. of CO for 1, 3 and 7 days presented different responses to Ca²⁺ in isolated mitochondria from liver. At days 1 and 3, isolated mitochondria from CO‐treated rats were more sensitive to Ca²⁺, presenting mitochondrial swelling at lower levels of this cation when compared with isolated mitochondria from control rats. Nevertheless, after 7 days the Ca²⁺ concentrations needed for triggering mitochondrial swelling were the same for control and CO‐treated rats, indicating a compensatory effect of CO (Piantadosi et al. 2006). Thus, these data indicated that CO somehow modulates the way Ca²⁺ regulates function and membrane permeabilization of mitochondria. In non‐synaptic mitochondria isolated from rat cortex (Queiroga et al. 2010) and from liver (Queiroga et al. 2011), CO prevented Ca²⁺‐induced mitochondrial membrane permeabilization, namely by inhibiting mitochondrial swelling, mitochondrial depolarization and mitochondrial inner membrane permeabilization. Furthermore, CO prevented astrocytic cell death by inhibiting MMP. Low amounts of CO promoted mitochondrial ROS generation, with an increase in levels of oxidized glutathione, which induces glutathionylation of adenine nucleotide translocator protein (ANT) (Queiroga et al. 2010). The physiological function of ANT is the transport of ATP/ADP through the mitochondrial inner membrane, while in response to stress, ANT can form a lethal pore (Marzo et al. 1998; Vieira et al. 2000) in a thiol‐dependent manner (Costantini et al. 2000). Furthermore, in isolated mitochondria from mouse liver, CO increased mitochondrial capacity to take up Ca²⁺, which has been assessed by directly measuring the Ca²⁺ concentration in the extramitochondrial supernatant (S.R. Oliveira, C.S.F Queiroga and H.L.A. Vieira, unpublished data). In summary, CO appears to increase the mitochondrial capacity to take up Ca²⁺ and to limit Ca²⁺‐induced mitochondrial membrane permeabilization, thus inhibiting cell death (Fig. 1).

Figure 1. Carbon monoxide and Ca²⁺: CO decreases the activity of L‐ and T‐type Ca²⁺ channels and modulates intracellular Ca²⁺ levels .

Ca²⁺ can be taken up by mitochondria, and low amounts of mitochondrial Ca²⁺ increase pyruvate dehydrogenase (PDH) activity and ATP production. CO targets mitochondria leading to (i) enhancement of mitochondrial metabolism, reinforcement of oxidative phosphorylation and an increase in ATP production; and (ii) the prevention of Ca²⁺‐induced mitochondrial membrane permeabilization (MMP), and thus inhibition of release of pro‐apoptotic factors into the cytosol and, ultimately, cell death. It can be hypothesized that (i) CO affects Ca²⁺ signalling between endoplasmic reticulum and mitochondria, (ii) CO regulates the mitochondrial calcium uniporter (MCU), and (iii) CO reinforces mitochondrial metabolism by regulating mitochondrial Ca²⁺ concentrations.

CO modulates mitochondrial metabolism

In hypoxia and ischaemia–reperfusion related pathologies, such as cerebral stroke, cardiac arrest or myocardium infarct, cellular metabolic adaptation is crucial for cell survival, namely the balance between oxidative and glycolytic metabolism. In addition, tumour progression strongly depends on cellular metabolic changes, also known as the Warburg effect, where cancer cells preferentially use glucose through glycolysis for increasing their proliferative capacity (Hanahan & Weinberg, 2011). Mitochondria are key organelles in cellular metabolism for cell energy supply and it is known that they are also the main target of CO. Thus, modulation of cell metabolism by CO arises as a relevant subject for studying this gasotransmitter as a cytoprotective factor and vehicle to avoid cancer progression.

In a porcine model of cardiac ischaemia and reperfusion, during coronary occlusion, the ratio between lactate production and glucose consumption decreased in CO‐treated animals, indicating that CO maintains mitochondrial metabolism during the ischaemic phase of the myocardium (Ahlström et al. 2009). Likewise, hearts of CO‐treated pigs that were subjected to cardiopulmonary bypass with cardioplegic arrest showed significantly higher ATP and phosphocreatine levels at 1 h of reperfusion, indicating CO's improvement of mitochondrial metabolism (Lavitrano et al. 2004). By stimulating mitochondrial biogenesis and improving cell metabolism, CO limits progressive toxic cardiomyopathy induced by doxorubicin (Suliman et al. 2007 a). Likewise, in a model of mouse lethal sepsis, CO promoted the animal's survival by supporting mitochondrial energy metabolism via activation of mitochondrial biogenesis (Lancel et al. 2009). In hepatocytes, endogenous CO derived from HO‐1 activity increases ATP production (Tsui et al. 2005). This metabolic improvement is essential for CO to rescue mice from fulminant hepatitis; CO‐increased ATP levels were associated with a marked reduction of tumour necrosis factor‐α‐induced apoptosis in the liver of d‐galactosamine‐sensitized mice (Tsui et al. 2007). Exogenous CO treatment increased intracellular ATP levels in primary culture of mouse astrocytes (Almeida et al. 2012). In the presence of exogenous CO, the ratio between lactate production and glucose consumption decreased, oxygen consumption increased and the mitochondrial population was higher, indicating that this CO‐induced metabolic improvement is due to an increase in oxidative phosphorylation levels. Furthermore, CO also modulated specific enzymatic activity of cytochrome c oxidase in a biphasic response: in the first 5 min after CO administration there was a slight decrease on its activity, while in the following 30 min, 3 h and 24 h there was an enhancement (Queiroga et al. 2011; Almeida et al. 2012). Thus, one can speculate that a compensatory response must exist: by partially and reversibly inhibiting cytochrome c oxidase, CO promotes ROS signalling with a reinforcement of this enzyme activity and promotion of mitochondrial biogenesis. Indeed, in cardiomyocytes, CO's stimulation of mitochondrial biogenesis is dependent on mitochondrial ROS generation and signalling (Suliman et al. 2007 b).

Finally in a recent work, Wegiel and colleagues have demonstrated that CO limits human prostate cancer progression by manipulation of cancer cell metabolism, promoting an anti‐Warburg effect, which sensitizes cells to chemotherapy. Indeed, CO targets mitochondrial activity, which was assessed by higher oxygen consumption levels and free radical generation, leading to mitochondrial collapse, cell growth arrest and apoptosis induced by chemotherapy (Wegiel et al. 2013). Furthermore, in human breast cancer cells, HO‐1 induction or exogenous CO exposure decreases cell proliferation (Lee et al. 2014). The same anti‐proliferative action of CORM‐2 was observed in pancreatic cancer (Vítek et al. 2014). Furthermore, tagged nanoemulsions containing CORM‐2 for targeted B‐cell lymphoma treatment in in vitro and in vivo experiments proved to be efficient (Loureiro et al. 2015). Nevertheless, CO administration for cancer treatment is not trivial, since up‐regulation of HO‐1 was already observed in several human and murine cancer types and linked to their intrinsic resistance (Yin et al. 2014). To date, there are contradictory facts about the manipulation of the CO/HO‐1 levels in cancer therapy. HO‐1‐conferred chemoresistance may depend on the tumour state and type, and the localization of HO‐1 and its signalling pathway (Chau, 2015; Dey et al. 2015; Tan et al. 2015). For example, in the case of the anti‐Warburg effect described by Wegiel and coworkers in prostatic cancer, HO‐1 accumulates in the nucleus and presents little enzymatic activity, which corresponds to low CO production (Wegiel et al. 2013).

In summary, HO activity‐derived CO or exogenously administrated CO manipulates cell metabolism and supports oxidative phosphorylation and mitochondrial respiration, which improves ATP production and cellular energy status (Fig. 1).

Carbon monoxide and Ca²⁺ channels

Modulation of ion channels by CO has been extensively studied throughout the last 15 years (for further information there is a recent and interesting review by Peers et al. (2014)). CO is involved in cellular Ca²⁺ signalling and in the modulation of Ca²⁺ channels. Inhibition of L‐type Ca²⁺ channels by CO conferred cytoprotection on cardiomyocytes in a mitochondrial ROS‐dependent manner (Scragg et al. 2008). In HEK293 cells stably expressing the human cardiac α‐Ca²⁺ channel, CORM‐2 reversibly inhibited voltage‐independent Ca²⁺ channels; still, whenever the mitochondrial specific antioxidant Mito Q was used, there was a reversion of CO's effect (Dallas et al. 2009). Thus, one can speculate that the cardioprotective role of HO‐1 could be due to CO's inhibition of L‐type Ca²⁺ channels.

Furthermore, T‐type Ca²⁺ channels are implicated in many pathophysiological roles such as cell proliferation, contractility and excitability. Indeed, CO also prevented the activity of Cav3.2 T‐type Ca²⁺ channels when expressed in NG108‐15 cells or rat primary cultures of sensory neurons, in a thioredoxin‐dependent manner (Boycott et al. 2013). Likewise, in smooth muscle cells, HO‐1 regulates cell proliferation via CO‐mediated inhibition of T‐type Ca²⁺ channels (Duckles et al. 2014). There are also data correlating CO with the regulation of intracellular organelle Ca²⁺ stores. Indeed, treatment with CORM‐2 in pancreatic acinar cells resulted in a slight increase in intracellular Ca²⁺, partially mediated by the release of Ca²⁺ from the endoplasmic reticulum. In this case, the slight Ca²⁺ rise functioned as a cytoprotective signal promoting NO production via the Gq–phospholipase‐C–inositol 1,4,5‐trisphosphate receptor pathway (Moustafa & Habara, 2014). Likewise, in human platelets, CO prevented Ca²⁺ release from sarcoplasmic/endoplasmic reticulum, as well as decreasing Ca²⁺ entry to the intracellular space (Gende, 2004). In apparent contrast, chronic rat exposure to CO to mimic air pollution increases cardiac arrhythmia following cardiac stress due to altered Ca²⁺ homeostasis. CO increased mitochondrial ROS production up to toxic levels, leading to oxidative stress, which resulted in sarcoplasmic reticulum Ca²⁺ leak (André et al. 2011). This is a good example of how CO concentration and period of exposure matter in regard to the physiological relevance of this gasotransmitter. It would be of great interest to discover whether physiological exposure to CO also modulates Ca²⁺ homeostasis at the sarcoplasmic reticulum and how.

Finally, it is widely recognized that mitochondria and the endoplasmic reticulum are key organelles for buffering Ca²⁺ and are extremely sensitive to variations in this cation's concentration (Herzig et al. 2013). In addition, mitochondria are a privileged cellular target for CO. Nevertheless, few data are available in the literature demonstrating a direct effect of CO on mitochondrial Ca²⁺ signalling. Further studies correlating mitochondrial Ca²⁺ signalling and CO's biological functions are urgently needed.

Future directions

Metabolic adaption to a suboptimal environment is crucial for cell survival and homeostasis. In order to produce a rapid response to external stimuli, fast signalling is needed. Ions and ion channels are rapidly recruited, in particular Ca²⁺. Ca²⁺ is a universal messenger, linking the extracellular space with organelles and the nucleus. An increase in intracellular Ca²⁺ levels is associated with a multitude of pathophysiological conditions, and understanding these molecular and cellular mechanisms may provide novel therapeutic options.

Since the 1970s it has been known that mitochondria are not a simple sink for Ca²⁺, and that this cation can modulate their functions. In the case of respiration, Ca²⁺ can be taken up by mitochondria and accelerate their respiration complexes, including pyruvate dehydrogenase, in turn leading to higher levels of ATP production (Denton, 2009; Griffiths & Rutter, 2009). In addition, it is established that CO reinforces oxidative phosphorylation and mitochondrial metabolism and that CO prevents Ca²⁺‐induced mitochondrial membrane permeabilization and cell death. Therefore, one can speculate that CO's modulation of mitochondrial Ca²⁺ uptake capacity can also be important for supporting mitochondrial ATP production by increasing the TCA cycle rate and mitochondrial respiration in a rapid and efficient manner. Likewise, CO could potentially control the MCU, for instance via ROS signalling. Still, Ca²⁺ communication between mitochondria and the endoplasmic reticulum is far from being well understood. Further work is necessary to address these issues.

It is worth mentioning that O₂ concentration under experimental conditions is a key factor in CO biology since CO and O₂ chemically compete for the same targets. Indeed, most of the in vitro approaches used are under hyperoxia and do not represent physiological oxygen concentrations in tissues. Thus, the in vitro effects of CO must be corroborated with data generated in vivo, or in vitro under physiological O₂ levels, in particular those concerning cell metabolism and cancer progression.

Finally, two pleiotropic and fast signalling molecules – Ca²⁺ and ROS – appear to be closely related to CO biology, at least in the early response phase. Novel findings involving Ca²⁺ and ROS signalling in CO's pathways are essential for the development of new therapeutic strategies based on this gasotransmitter.

Additional information

Competing interests

There is no conflict of interests.

Author contributions

All authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

This work was supported by the Portuguese Fundação para a Ciência e Tecnologia (FCT) grant FCT‐ANR/NEU‐NMC/0022/2012, H.L.A.V.’s FCT support IF/00185/2012, C.S.F.Q.’s SFRH/BPD/88783/2012 fellowship and S.R.O.’s SFRH/BD/51969/2012 fellowship both funded by FCT.

Biographies

Sara R. Oliveira graduated in 2011 from the Instituto Superior Técnico, Universidade de Lisboa, with a master's degree in Biological Engineering, with a master's thesis on yeast apoptosis. Currently she is a third year PhD student at the Center of Neuroscience and Cell Biology (CNC), Universidade de Coimbra and CEDOC, Universidade Nova de Lisboa with a shared supervision from Carlos B. Duarte and Helena L. A. Vieira.

Cláudia S. F. Queiroga did her PhD thesis on the molecular mechanisms underlying cerebral ischaemia with supervision from Paula M. Alves and Helena L. A. Vieira in iBET and ITQB of the Universidade Nova de Lisboa, and currently has a postdoctoral fellowship with Helena L. A. Vieira in the Laboratory of Cell Death and Disease, CEDOC, Universidade Nova de Lisboa, Portugal.

Helena L. A. Vieira did her PhD with Guido Kroemer at the Centre National pour la Recherche Scientifique (CNRS), France, in the fields of Cell Biology and Oncobiology; then she moved back to Portugal with two postdoctoral fellowships with Paula M. Alves in iBET and ITQB of the Universidade Nova de Lisboa in the biotechnology field followed by the neurobiolgy field. In 2011, she started her own group at CEDOC. Their common interests are neurobiology, cell death control, mitochondria and carbon monoxide biology.