|Home | About | Journals | Submit | Contact Us | Français|
Low concentrations of carbon monoxide (CO) protect hepatocytes against apoptosis and confers cytoprotection in several models of liver. Mitochondria are key organelles in cell death control via their membrane permeabilization and the release of pro-apoptotic factors.
Herein, we show that CO prevents mitochondrial membrane permeabilization (MMP) in liver isolated mitochondria. Direct and indirect approaches were used to evaluate MMP inhibition by CO: mitochondrial swelling, mitochondrial depolarization and inner membrane permeabilization. Additionally, CO increases mitochondrial reactive oxygen species (ROS) generation, and their scavenging, by ß-carotene addition, decreases CO protection, which reveals the key role of ROS. Interestingly, cytochrome c oxidase transiently responds to low concentrations of CO by decreasing its activity in the first 5 min, later on there is an increase of cytochrome c oxidase activity, which were detected up to 30 min.
CO directly prevents mitochondrial membrane permeabilization, which might be implicated in the hepatic apoptosis inhibition by this gaseoustransmitter.
Carbon monoxide (CO) is usually considered a harmful and toxic molecule due to its high affinity to heme proteins. However, recent evidences show that low doses of CO can be cytoprotective, presenting several biological properties, namely, anti-apoptosis, anti-proliferation, anti-inflammation and vasodilatation . Furthermore, CO is an endogenous product of heme degradation by heme-oxygenase (HO), generating free iron and biliverdin as by-products. In fact, HO system is essential for tissue response to diverse pathological contexts, aiming at restoring and/or maintaining cellular homeostasis .
In hepatocytes and/or liver models, CO appears to act as an anti-apoptotic molecule. By stimulating ATP production, CO activates p38 MAPK signalling, preventing apoptosis in human hepatocytes . CO rescues mice from fulminant hepatitis, presenting a marked reduction of TNF-alpha-induced apoptosis  or via NO generation . In primary cultures of rat hepatocytes, CO limits cytotoxicity induced by glucose deprivation through suppression of ERK MAPK activation . In an endotoxic shock model, CO protects hepatocytes from apoptosis by augmenting iNOS expression . It is also described that superoxide anion-induced apoptosis is inhibited by CO via limiting JNK activity . CO treatment protects hepatocytes from cell death by inducing NF-kB activation, which is dependent on ROS generation, since inhibition of ROS generation (via anti-oxidant addition or by using respiratory deficient cells) reverses CO-induced cytoprotection . Among all publications showing CO as anti-apoptotic molecule in hepatic model, only Kim and colleagues  have mentioned the involvement of mitochondria. CO protects hepatocytes from TNF-alpha/Actinomycin D-induced apoptosis by activating NF-kB, which is associated with a reduction in cytochrome c release from mitochondria . However, no data demonstrate the direct role of CO into isolated liver mitochondria.
Mitochondria play a key role in the intrinsic pathways of apoptosis. Many pro-apoptotic factors are confined in the inter-membrane space, and upon mitochondrial membrane permeabilization (MMP) these factors are released into the cytosol and cell death becomes an irreversible process . MMP marks a point of no return in the apoptotic intrinsic pathways by activating both caspase-dependent and caspase-independent mechanisms. The rupture of mitochondrial membrane also leads to the functional impairment of mitochondria, bioenergetic and redox crisis with ATP depletion and strong oxidative stress . Therefore, mitochondria become a crucial target to modulate cell death in several models.
Based on the following facts: (i) CO is an anti-apoptotic molecule in several hepatic models, hepatocytes and/or liver and (ii) mitochondria are central executers of cell death process, via the mitochondrial membrane permeabilization (MMP); we explored the direct effect of CO into isolated liver mitochondria (MMP modulation) and the involvement of ROS in this process. MMP was assessed by mitochondrial depolarization, inner membrane permeabilization and mitochondrial swelling.
In order to evaluate the toxicity of carbon monoxide (CO) on isolated liver mitochondria, swelling and depolarization assays were performed with different doses of CO (10 to 500 μM). Up to 100 μM and for 30 minutes, CO is not able to trigger mitochondrial swelling. However CO triggers swelling at concentrations of 250 and 500 μM (Figure (Figure1A),1A), indicating that at higher concentrations this gas induces mitochondrial damage. Because mitochondrial depolarization is an earlier event compared to swelling, lower concentrations of CO have induced lost of ΔΨm, only at 10 μM there is no depolarization induction (Figure (Figure1B).1B). Pre-treatment with 10, 50 or 100 μM of CO for 15 minutes partially prevents mitochondrial swelling induced by calcium (Figure (Figure1C).1C). Since 10 μM presents the highest protection and does not induce mitochondrial depolarization, this concentration is the optimal one to be used in all other assays.
CO partially inhibits mitochondrial swelling (Figure (Figure2A)2A) when liver isolated mitochondria were exposed to 10 μM of CO during 15 minutes at room temperature prior to addition of calcium to induce MMP (Figure (Figure2A2A and and2C,2C, left panel) and of atractyloside, a ligand of ANT that prevents ADP/ATP translocation and induces its pore forming function (Figure (Figure2A2A and and2C,2C, right panel) [13-15]. Swelling quantification analysis was performed at different time points for calcium or atractyloside because of their different mode of action and kinetics (Figure (Figure2C,2C, left and right panel). Loss of ΔΨm, or mitochondrial depolarization, induced by atractyloside or calcium was also prevented by prior addition of CO at 10 μM (Figure (Figure2B)2B) and quantification analysis was done (Figure (Figure2D)2D) for calcium (left panel) and for atractyloside (right panel). Changes in the inner membrane permeability (the opening of a large channel for molecules up to ~800 Da) were assessed by an enzymatic assay based on the accessibility of citrate synthase, which is a soluble matrix enzyme . The atractyloside induction of inner membrane permeabilization is partially prevented by CO (Figure (Figure3).3). Due to the nature of this enzymatic assay, the quantification of the effect has to be done by calculating the slopes of the different curves over time, which were normalised relatively to Ca2+ 5 μM (Table (Table1).1). CO pre-treatment decreases the slope compared to Ca2+ 5 μM (85% vs 100%) and Atra 300 μM (81% vs 145%). This reduction demonstrates CO delay in the induction of inner membrane permeabilization. Taken together, CO acts directly on mitochondria in order to limit their membrane permeabilization, shown by its prevention of swelling, depolarization and inner membrane permeabilization.
Mitochondrial generated reactive oxygen species (ROS) are described as imperative signalling molecules for CO biological functions [17-20]. Thus, the role of ROS in CO-induced protection in isolated liver mitochondria was also evaluated. First, it was verified whether CO increases ROS generation in isolated liver mitochondria in a dose-response manner and this generation is limited by pre-treatment with ß-carotene (Figure (Figure4A).4A). To disclose ROS role in this system, their level augmentation was prevented by treating mitochondria with ß-carotene prior to CO addition and MMP induction. Indeed, CO inhibition of mitochondrial swelling (Figure (Figure4B),4B), depolarization (Figure (Figure4C)4C) and inner membrane permeabilization (Figure (Figure4D4D and Table Table2)2) were, at least, partially prevented by the anti-oxidant ß-carotene. Thus, at mitochondrial level, ROS appear to be important signalling molecules for CO mode of action during inhibition of mitochondrial membrane permeabilization.
Major intracellular sources of ROS are the oxidative phosphorilation complexes. Moreover, accordingly to the literature, CO is described to prevent cytochrome c oxidase (COX) activity . Thus, CO direct effect on the COX activity was assessed in isolated liver mitochondria. CO at 10 μM transiently prevents COX activity up to 10 minutes, while 30 minutes later the effect is inverted; CO slightly accelerates COX activity (Figure (Figure55).
Carbon monoxide has been described to be involved in protection of hepatocytes against cell death. Kim and colleagues  have demonstrated that CO decreases Bcl-2 family proteins translocation into mitochondria, limiting cytochrome c release into the cytosol . Despite the crucial role of mitochondria in cell death control and the potent anti-apoptotic property of CO in hepatocytes, the direct effect of CO in isolated liver mitochondrial membrane permeabilization has never been reported before. Herein it is shown that low doses of this gaseous molecule prevent mitochondrial membrane permeabilization. Recently, we have shown that CO limits mitochondrial membrane permeabilization in non-synaptic mitochondria isolated from rat brain cortex , which was accompanied by inhibition of cytochrome c release from mitochondria and by glutathionylation of adenine nucleotide translocase (ANT). In this cerebral model, CO protects astrocytes against cell death and ROS generation appears to be important for this pathway . Furthermore, in other systems, it is generally recognized that several CO biological functions are dependent on mitochondrial ROS generation and signalling [17-21]. In the present work, ROS also emerge as significant molecules involved in the signal transduction at the mitochondrial sub-cellular level. Inhibition of mitochondrial ROS generation by an anti-oxidant addition (ß-carotene) reverses CO prevention of liver mitochondrial membrane permeabilization (Figure (Figure4),4), which confirms their key role. Still one might hypothesize that ROS promote post-translational modifications on mitochondrial proteins, as described for ANT glutathionylation in non-synaptic mitochondria .
The most accepted hypothesis for CO-induced mitochondrial ROS production is via partial inhibition of cytochrome c oxidase, accumulating electrons at complex III level. The generated anion superoxide is rapidly converted into hydrogen peroxide . According to our results, and using low concentrations of CO (10 μM), COX inhibition occurs only up to 10 minutes after CO treatment (Figure (Figure5).5). One might speculate that this transient inhibition assures sufficient ROS generation to signal protective pathways, although not enough to induce damage. On the other hand, after 30 minutes COX activity is enhanced by CO treatment (Figure (Figure5).5). Interestingly it is in accordance to our previous data showing an increase on ATP/ADP translocase activity of ANT  or a mitochondrial hyperpolarization by low concentrations of CO . In summary, low doses of CO appear to accelerate mitochondrial oxidative phosphorylation and oxygen consumption. Another hypothesis to be considered is whether transient inhibition of COX activity also decreases calcium uptake protecting mitochondria against MMP.
Further studies are needed to elucidate the mechanisms implicated in ROS signalling, in particular how CO modifies and/or accelerates mitochondrial oxidative phosphorylation and oxygen consumption.
Thus, for the first time, it was demonstrated that CO inhibits MMP in isolated liver mitochondria, by preventing mitochondrial swelling, mitochondrial depolarization and the opening of a non-specific pore through inner membrane. Additionally, small amounts of ROS generation are essential for signalling MMP inhibition by CO. In conclusion, it can be hypothesized that part of the CO's anti-apoptotic property in hepatocytes and/or liver is due to its capacity to limit mitochondrial membrane permeabilization, preventing the release of pro-apoptotic factors into the cytosol.
Mitochondria were isolated from mouse liver (C57, female, 6-12 week old, Instituto Gulbenkian de Ciência, Portugal) by differential centrifugation and purified on Percoll gradient, according to . Mitochondrial protein was quantified using BCA assay (Pierce, Illinois). All mitochondrial assays were performed under atmospheric air, without oxygen level control.
Fresh stock solutions of CO gas were prepared daily and carefully sealed immediately after. PBS (Phosphate Buffered Saline) was saturated by bubbling 100% of CO gas during 30 minutes to produce 10-3 M stock solution. The concentration of CO in solution was determined spectrophotometrically, as previously described . CO compressed gas at 100% was purchased from Linde, Germany.
ROS generation was monitored by the conversion of 2',7'-dichlorofluorescein diacetate (H2DCFDA, Invitrogen, UK) to fluorescent 2', 7'-dichlorofluorescein (DCF). 25 μg of mitochondrial protein was incubated with 5 μM of H2DCFDA and 10, 50, 100 or 250 μM of CO or 500 μM of hydrogen peroxide, in swelling buffer. Fluorescence (λexc: 485 nm, λem: 530 nm) was measured using Biotek Synergy 2 Spectrofluorimeter during 30 minutes at 37°C. ROS generation was calculated as an increase over baseline levels, determined for untreated cells and considering 100% of ROS generation with 500 μM of hydrogen peroxide. In some cases, β-carotene (1 μM) was added to isolated mitochondria 10 minutes prior CO treatment.
25 μg of mitochondrial protein was diluted in swelling buffer for swelling (decrease in optical density at 540 nm) or depolarization rhodamine 123 (1 μM) fluorescence dequenching assay containing or not 10 μM of CO for 15 min of incubation at room temperature, as described in . In some cases, β-carotene (1 μM) was added to isolated mitochondria 10 minutes prior CO treatment.
Mitochondrial swelling was assessed by the decrease in optical density at 540 nm measured for 30 minutes at 37°C, using Biotek Synergy 2 Spectrofluorimeter. 100% of swelling is calculated based on the optical density decrease between non-treated and 12.5 μM Ca2+ or 300 μM atractyloside treated mitochondria.
For depolarization assessment by Rhodamine 123 dequenching, 6.25 μM of Ca2+ or 300 μM of atractyloside were added. The fluorescent measurements (λexc: 485 nm, λem: 535 nm, Biotek Synergy 2 Spectrofluorimeter) were followed at 37°C and are expressed in percentage relative to the positive control 5 μM of Ca2+ or 300 μM atractyloside (100%) at the indicated time point, as described in .
Citrate synthase activity assay was used to assess the inner membrane permeability according to . Upon inner mitochondrial membrane permeabilisation acetyl-CoA is able to enter into mitochondrial matrix, reacting with citrate synthase. 5, 5'-dithio-bis 2-nitrobenzoic acid (DTNB) and deacetyled acetyl-CoA reaction gives 5-thio-2-nitrobenzoate (TNB) which can be followed by absorbance at 412 nm. Briefly, 25 μg of protein from isolated mitochondria was incubated with CO (10 μM) in swelling buffer containing 100 μM of DTNB, 300 μM of acetylCoA and 1 mM of oxaloacetate. Inner membrane permeabilisation was induced by atractyloside at 300 μM or Ca2+ at 5 μM. Whenever the case, β-carotene (1 μM) was was added 10 minutes prior CO treatment. The absorbance at 412 nm was acquired for 20 minutes, using Biotek Synergy 2 Spectrofluorimeter. For slope calculation, absorbance values for each different condition were normalized by the absorbance value corresponding to the control non-treated mitochondria.
Cytochrome c oxidase (COX) activity was determined using a kit from Sigma CYTOCOX1. It is a colorimetric assay based on the oxidation of ferrocytochrome c to ferricytochrome c by COX. The reaction can be followed by a decrease in the absorbance at 550 nm, at 25°C. Briefly, mitochondria were treated with 10 μM of CO at 37°C for 5, 10, 15 and 30 minutes. The absorbance at 550 nm was acquired using Spectrophotometer DU-530, under the following conditions: 230 μg of mitochondrial protein was incubated with 0.45 mM Tris-HCl containing 12 mM of sucrose, 9 mM Tris-HCl containing 100 mM KCl and 0.01 mM of ferrocytochrome c, during 1 minute, with 10 seconds of interval. For each sample, COX activity was expressed in mU/mL.
Mitochondrial data is presented as a representative result of at least three independent batchs or assays. All values are mean ± SD, n ≥ 3. Error bars, corresponding to standard deviation, are represented in the figures. Statistical comparisons were performed using ANOVA: single factor, with p < 0.05, n ≥ 3. p < 0.05 means that samples are significantly different at a confidence level of 95%.
CO: carbon monoxide; HO: heme-oxygenase; ROS: reactive oxygen species; MMP: mitochondrial membrane permeabilization; Atra: atractyloside; ΔΨm: mitochondrial membrane potential; H2DCFDA: 2',7'-dichlorofluorescein diacetate; β-c: β-carotene; RT: room temperature; COX: cytochrome c oxidase
CSFQ has made the acquisition and interpretation of data and drafted the manuscript; ASA performed experiments important for the re-estructure of the manuscript; PMA has been involved in revising the manuscript; CB has made contributions to conception of the manuscript and to revise it critically and HLAV conceived of the study, participated in its design, coordination and acquisition of data, and helped to draft the manuscript. All authors read the paper and approved the final manuscript.
This work was supported by the Portuguese Fundação para a Ciência e Tecnologia grants PTDC/SAU-NEU/64327/2006 and PTDC/SAU-NEU/098747/2008. H. Vieira and C. Queiroga have grants from SFRH/BPD/27125/2006 and SFRH/BD/43387/2008. The authors express their gratitude to João Seixas from Alfama, Portugal, for measurements of CO in solution.