| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Surgery,
* University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA; and
Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts, USA
1Correspondence: University of Pittsburgh School of Medicine, NW653 MUH, 3459 Fifth Ave., Pittsburgh, PA 15213, USA. E-mail: zuckerbraunbs{at}upmc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
0 cells. CO inhibits cytochrome c oxidase, while maintaining cellular ATP levels and increasing mitochondrial membrane potential. The addition of antioxidants or inhibition of complex III of the electron transport chain by antimycin A attenuates the inhibitory effects of CO on lipopolysaccharide (LPS)-induced TNF-
and blocked CO-induced p38 MAPK phosphorylation, which we previously have shown to be important in the anti-inflammatory effects of CO.—Zuckerbraun, B. S., Chin, B. Y., Bilban, M., de Costa d’Avila, J., Rao, J., Billiar, T. R., Otterbein, L. E. Carbon monoxide signals via inhibition of cytochrome c oxidase and generation of mitochondrial reactive oxygen species.
Key Words: macrophage • p38 MAPK • tumor necrosis factor-alpha
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. The delivery of exogenous CO has been shown to be cytoprotective both in vivo and in vitro in models of vascular injury, ischemia/reperfusion, transplantation, endotoxemia, and hemorrhagic shock (3
4
5
6
7
8
9
10
11
12)
. Akin to NO, CO signaling has been illustrated primarily in vascular cells to function via activation of the enzyme-soluble guanylyl cyclase and the generation of cGMP (13)
. However, the level of activation of soluble guanylyl cyclase by CO is modest compared to that of NO. Cyclic GMP was shown to not be involved in the ability of CO to exert its potent anti-inflammatory effects in macrophages (14)
. Additional heme containing targets of CO action have also been identified, including cytochrome P450 enzymes, NO synthases, and cytochrome c oxidase, among others. Taille et al. have shown that CO from pharmacologic CO donors inhibits the production of ROS from NAD(P)H oxidase in platelet-derived growth factor (PDGF)-stimulated airway smooth muscle cells (15)
. Cytochrome c oxidase (complex IV) serves as the mitochondrial enzyme responsible for the reduction of oxygen into water as the final step of the electron transport chain (ETC) (16)
. CO has long been known to inhibit cytochrome c oxidase by competing with oxygen binding (17
18
19
20)
. The majority of evidence supporting this comes from studies using very high concentrations of CO. Inhibition of cytochrome c oxidase by decreased oxygen levels and/or via NO binding can increase generation of mitochondrial-derived ROS and thereby alter the cellular redox state (21
22
23
24)
. This generation of ROS has been linked to the activation of various adaptive signaling pathways, including the mitogen-activated protein kinase family members, JNK and p38, as well as NF-
B (24
25
26
27)
. The mechanisms of protection conferred by CO are only partially elucidated. We therefore hypothesized that exogenous CO imparts cytoprotection via inhibition of cytochrome c oxidase, favoring the generation of ROS, which, in turn, activate adaptive signaling pathways. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
0 cells) were generated by incubating wild-type cells in ethidium bromide for 2–3 wk in medium supplemented with pyruvate and uridine (28)
. The 0 cells were then selected by exposure of the mitochondrial inhibitors rotenone (1 µg/ml) and antimycin A (1 µg/ml), which are lethal to wild-type cells. Cells were considered 0 cells when there was no detectable cytochrome b by polymerase chain reaction (PCR). All CO exposures were performed at 250 ppm.
Cytochrome c oxidase microassay
Cytochrome c oxidase activity was determined in permeabilized whole cells using a microtiter assay as described by Chrzanowska-Lightowlers et al. (29)
. Briefly, RAW 264.7 cells were cultured in a 96-well tissue culture plate. Cells were exposed to standard incubation or CO (250 ppm) for 1 h. All traces of growth media were removed by aspiration. The cells were permeabilized by the addition of 50 µl/well of 0.01% saponin in water for 10 min. One-hundred microliters of substrate media were added to each well to give a final concentration of 4 mM 3,3'-diaminobenzidine (DAB), 100 µM reduced cytochrome c, 2 µg/ml catalase in 0.1 M Na phosphate, pH 7.0. Immediately after this addition, absorbances were measured over 15 min using a 450-nm filter. Control incubations were performed, which included cells plus 0.3 mM sodium azide to show specificity and wells with saponin and substrate but no cells to monitor background oxidation of DAB.
TNF- measurement
Levels of the cytokines TNF- were determined by ELISA (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions.
ATP measurement
ATP levels were determined using a CellTiter-Glo luciferase/luciferin reaction assay (Promega, Madison, WI, USA) as per the manufacturer’s protocol.
Phospho-p38 ELISA
Phosphorylation of p38 MAPK was determined using cellular activation of signaling ELISA kit as per the manufacturer’s instructions (Super Array Bioscience Corporation, Frederick, MD, USA).
ROS measurement
Intracellular ROS generation was assessed using 2',7'-dichlorofluorescin diacetate (DCF-DA; 5 µM). Microscopy or FACS was performed. Cells on coverslips were perfused under controlled O2 and CO conditions in a flow-through chamber at 37°C on an inverted fluorescent microscope. Images were acquired with an Olympus camera (excitation 488 nm, emission 535 nm). Intensities are reported as arbitrary fluorescent units compared to non-CO treated cells over time, after subtracting background. Data were taken as the average fluorescence for 5 groups of 5 cells.
Fluorescent-activated cell sorter analysis of DCF fluorescence
RAW 264.7 and THP-1 macrophages were incubated with 10 µM DCFH-DA (Invitrogen, Carlsbad, CA), for 30 min at 37°C prior to harvest time point. The cells were then washed, exposed to air or CO for 5–60 min, resuspended in FACS buffer (PBS+1%FBS), and analyzed on a fluorescent activated cell sorter (FACS), (BD Biosciences, San Diego, CA). Forward and sidescatter gates were set to include cells but to exclude debris and remaining unbound particles. Excitation was set at 488 nm, and emission was recorded on an FL1 detector (525±10 nm). Electronic compensation was used to eliminate spreading into adjacent fluorescence channels. Each experimental set was performed 5 times, and the data analyzed with CellQuestTM software (BD Biosciences, San Diego, CA, USA).
Determination of cellular glutathione content
Total glutathione was determined using a kit based on the oxidation of glutathione (GSH) by 5,5'-dithiobis(2-nitrobenzoic acid) as per the manufacturer’s instructions (U.S. Biological, Swampscott, MA, USA). Cells treated with 1 mM n-acetylcysteine or 1 mM DL-buthionine-[S,R]sulfoximine (BSO) were used as a positive and negative controls, respectively.
Determination of mitochondrial membrane potential
Mitochondrial membrane potential was determined using the vital mitochondrial dye JC-1, (5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolocarbocyanine iodide; Molecular Probes, Eugene, OR). In the cytosol and in the mitochondria at low membrane potential, the monomeric form of JC-1 fluoresces green (emission at 525 nm), whereas within the mitochondrial matrix at high membrane potentials, JC-1 forms aggregates that fluoresce red (emission at 590 nm). Samples were incubated with JC-1 at a final concentration of 1 µM at 37°C for the last 30 min of the experiment. Fluorescent microscopy was performed. Red and green fluorescence was quantified using MetaMorphTM (Molecular Dynamics, Sunnyvale, CA, USA). Results are expressed as the ratio of red:green fluorescence.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A, B). Additionally, we show a dose-response of CO (50–500 ppm) in increasing DCF fluorescence (Fig. 1C
). Cells that were pretreated with the nitric oxide synthase (NOS) inhibitor, L-NMMA exhibited the same increase in DCF fluorescence, suggesting that at least in RAW macrophages, this was not mediated by NO generation (data not shown). In addition to DCF fluorescence, the influence of CO on the generation of ROS was also determined by measuring changes in the cellular antioxidant glutathione. CO decreased levels of intracellular glutathione after 1 h of exposure, supporting the concept that CO was increasing ROS production (Fig. 1D
). However, levels of glutathione were restored to baseline by a 4-h time point after the initiation of CO, despite continued exposure. This may represent a cellular response with de novo glutathione synthesis in response to CO as suggested by Sawle et al. (30)
.
|
Anti-inflammatory effects of CO are dependent on ROS
It is well known that CO decreases production of inflammatory cytokines in RAW cells in response to LPS. The contribution of CO-induced ROS on these anti-inflammatory effects was assessed. It is important to note that LPS (10 ng/ml) treatment alone did not induce the generation of intracellular ROS (Fig. 2
A). Phorbol myristate acetate (PMA) was used as a positive control. Consistent with our previous findings, CO decreased LPS (10 ng/ml)-induced TNF- production by RAW 264.7 cells. Pretreatment with PEG-superoxide dismutase (SOD) (10 U/ml) and PEG-catalase (10 U/ml) reversed the attenuation of TNF- by CO, suggesting that ROS generated in response to CO were functioning in part as a second messenger of CO signaling (Fig. 2B
).
|
ROS function as second messengers of CO signaling
The anti-inflammatory effects of CO have been attributed to the specific amplification of LPS-induced p38 MAPK signaling pathway; however, the mechanism by which CO augments LPS-induced p38 MAPK phosphorylation is not known. We investigated the role of CO-induced ROS on p38 phosphorylation in the setting of LPS stimulation. To test this, RAW cells pretreated with PEG-SOD and PEG-catalase were exposed to CO in the presence and absence of LPS, and phosphorylated p38 was assessed by ELISA. LPS treatment increased p38 phosphorylation 2.37 ± 0.55-fold compared to untreated controls. As expected, CO augmented LPS-induced p38 phosphorylation to 5.7 ± 0.87-fold compared to untreated cells (P<0.05 compared to LPS alone). Pretreatment with SOD and catalase abrogated the CO-induced augmentation, resulting in only a 3.6 ± 0.64-fold increase in p38 phosphorylation (P<0.05 compared to LPS+CO-treated cells). SOD and catalase pretreatment did not significantly influence the effect of LPS stimulation on p38 phosphorylation (Fig. 2C
).
CO-induced ROS are generated by mitochondria
There are numerous sites of ROS production within cells including NAD(P)H oxidase, various cytochrome P450 enzymes, xanthine oxidase, and the mitochondrial electron transport chain. Several studies have demonstrated that NO influences cellular respiration by inhibiting cytochrome c oxidase, leading to the generation of superoxide and hydrogen peroxide from mitochondrial complexes I and III. The influence of CO on cytochrome c oxidase in isolated mitochondrial preparations has also been well described (18
, 31)
. Additionally, CO has been shown to potentiate fluorescence of a probe specific for mitochondrial-produced ROS in PDGF-stimulated airway smooth muscle cells (15)
. We hypothesized that CO at this low concentration (250 ppm) was targeting the heme-containing oxidases of the mitochondria resulting in a transient burst in ROS generation as electrons were favorably directed toward reacting with O2 to form superoxide. We first wanted to identify the cellular compartment and more specifically the mitochondria as the source of ROS. To test this, we generated respiration-deficient macrophages or rho zero cells (0). The absence of mitochondrial DNA was confirmed in these cells by PCR analysis of cytochrome b (Fig. 3
A). Exposure of 0 cells to CO resulted in a loss in ROS as measured by FACS analyses of DCF vs. wild-type respiration-intact cells (Fig. 3B
). These data allowed us to conclude that the mitochondria were the source of ROS. We next wanted to know where in the mitochondria, CO-induced ROS were being generated. To test this hypothesis, RAW cells were pretreated with the pharmacologic inhibitors antimycin A, apocynin, diphenyliodonium (DPI), and rotenone, which are known inhibitors of components of the electron transport chain or NAD(P)H oxidase; however, these agents are somewhat nonspecific in their inhibition. The influence of CO on DCF fluorescence was then assessed by FACS. The mitochondrial complex III inhibitor antimycin A significantly decreased CO-induced DCF fluorescence, while the other inhibitors did not. Interestingly, inhibition of a potential source of ROS, NAD(P)H oxidase with either apocynin or DPI did not inhibit CO-induced ROS generation, confirming that the mitochondria is the source of ROS (Fig. 3C
).
|
Cytochrome c oxidase is inhibited by exogenous CO
The influence of CO on cytochrome c oxidase activity, which is downstream in the chain from complex III was assayed in whole cell preparations. RAW 264.7 cells were exposed to CO (250 ppm) for 1 h, after which cytochrome c oxidase activity was determined. Compared to untreated controls, CO inhibited cytochrome c oxidase activity by 50% (Fig. 4
). The consequences of inhibition of cytochrome c oxidase by CO would be a reduction of the electron transport chain and the generation of superoxide anions by complexes III and I. This correlates with the findings that inhibition of complex III by antimycin A or studies in 0 cells diminishes ROS generation by CO.
|
CO leads to a hyperpolarization of mitochondrial membrane potential
Inhibition of respiration by CO could lead to a collapse in mitochondrial membrane potential, decreased production of mitochondrial ATP synthesis, and depletion of ATP levels ultimately leading to cell death. We initially sought to determine the influence of CO on cellular ATP levels. These data demonstrate that exogenous CO does not lead to a significant depletion of ATP (Fig. 5
A) nor does it affect cell viability (data not shown). This would suggest that in the setting of inhibition of oxidative phosphorylation by CO, that CO also paradoxically leads to increased ATP generation by glycolysis and/or decreased utilization of ATP. Beltran et al. illustrated that as long as the production of glycolytic ATP was possible, that inhibition of cytochrome oxidase by NO led to a hyperpolarization of the mitochondrial membrane (32)
. The influence of CO on membrane potential was also assayed. RAW cells were stained with the mitochondrial vital dye JC-1, subjected to fluorescent microscopy, and a red:green ratio determined. CO exposure for one hour consistently caused a hyperpolarization of membrane potential (Fig. 5B
). CO increased the red:green ratio to 121 ± 5% that of controls, which indicates mitochondrial membrane hyperpolarization. The dependence on the F1F0-ATPase for the hyperpolarization was assayed by adding oligomycin (6 µM). In the presence of oligomycin, the red:green ratio was only 104 ± 5% that of controls, which is consistent with what has been illustrated for NO (32)
.
|
Antimycin A reverses the anti-inflammatory effects of CO
We next sought to determine whether inhibition of CO-induced mitochondrial ROS would attenuate the anti-inflammatory effects of CO. We used LPS-induced TNF- as the prototypical proinflammatory molecule. RAW cells were treated with antimycin A alone or with LPS-treated cells had no significant influence on TNF- production. However, antimycin significantly attenuated the inhibitory effect of CO on LPS-induced TNF- production (Fig. 6
A). These data support the concept that the generation of mitochondrial ROS, specifically at complex III by CO is responsible, at least in part, for the anti-inflammatory effects of CO.
|
The influence of antimycin A on CO-induced p38 MAPK phosphorylation in LPS treated cells was also investigated as activation of the p38 pathway by CO is required for blockade of LPS-induced TNF- (14)
. Antimycin A reversed the increase in phosphorylation of p38 that is seen with CO, supporting the notion that the inhibition of cytochrome c oxidase (complex IV) by CO and the production of ROS by proximal complex III is responsible for the increased phosphorylation of p38 MAPK (Fig. 6B
) and more importantly provide, in part, an explanation for the anti-inflammatory effects of CO.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
0 cells, which lack the ability to carry out mitochondrial respiration. CO targets cytochrome c oxidase, while permitting cellular ATP generation and increases mitochondrial membrane potential. The addition of SOD and catalase to scavenge superoxide and hydrogen peroxide or inhibition of CO-induced ROS generation by antimycin A effectively attenuated the anti-inflammatory effects of CO and diminished CO-induced adaptive signaling which revolves around activation of p38 MAP kinase. These studies demonstrate that CO induces ROS production by inhibition of cytochrome c oxidase and that these ROS subsequently act as second messengers for CO signaling. The effect of CO on augmented p38 activation occurs only in the presence of the LPS stimulation with no effects on p38 in the absence of LPS. Ongoing investigations of CO-induced gene profiling by gene chip analyses in the absence of a stimulus shows that CO increases expression of a number of candidate genes (data not shown), many driven by ROS that may in part explain many of the downstream effects that occur after stimulation such as p38 activation and support what we conclude as a preconditioning of the cell by CO, in essence altering the phenotype toward one of protection. What remains unclear is what the factors are that determine select activation of p38 vs. the other MAPK, such as ERK or JNK or other signaling pathways, which occurs with CO in other cells in the presence of different stimuli (6
, 12
, 33
, 34)
.
The effects of CO demonstrated in these data are similar to those of NO on cytochrome c oxidase, ROS production, and mitochondrial membrane potential (23
, 24
, 32
, 35
36
37)
. Here, we demonstrate that the generation of ROS by CO was independent of NO, in that CO had the same effect in the presence of a NOS inhibitor. A growing body of work has illustrated that the effects of NO on cytochrome c oxidase results in the reduction of the electron transport chain leading to the generation of superoxide anions. Superoxide anions are rapidly converted to hydrogen peroxide by SOD, and hydrogen peroxide can activate numerous mechanisms that lead to cell protection or death. Studies have confirmed that with prolonged NO exposure, inhibition of respiration becomes persistent, and this occurs mainly at the level of the upstream complexes where superoxide is generated. Moncada et al. hypothesize that the early stage of inhibition of cytochrome c oxidase by NO leads to the generation of low levels of hydrogen peroxide and subsequent biochemical and genetic signaling mechanisms that lead to a process of cellular defense (24)
. However, unlike NO, CO does not react with the generated ROS. CO signaling, even with prolonged exposures, may resemble the early effects of NO on cellular respiration and elicitation of subsequent cytoprotective effects.
The investigations into the cellular targets for CO will continue, and no doubt lead to different findings in different settings as ROS in different cell types will elicit different activation of signaling cascades and gene expression patterns, but the data presented here suggest that one of the most basic mechanisms of action occurs at the level of the mitochondria involving the most fundamental of biochemical pathways. Taille et al. have shown that pharmacological CO donor can induce mitochondrial ROS and that scavenging ROS with N-acetylcysteine partially reversed the inhibitory effects of CO donor on PDGF-stimulated airway smooth muscle cell proliferation (15)
. Modulation of cytochrome c oxidase by CO and the subsequent effects on cellular respiration may explain many of the cytoprotective effects of CO that have been shown in multiple in vivo and in vitro models. Studies suggest that exogenous CO can promote a ‘quiescent’ state in many different cell types, following many different stimuli. For example, we have previously shown that CO can inhibit proliferation of vascular smooth muscle cells (6)
, TNF--induced apoptosis of hepatocytes (38)
, and LPS-induced activation of RAW 264.7 cells (14)
. We hypothesize that the CO decreases the response to inflammatory stimuli by preconditioning the cell. It should be noted that Srisook et al. demonstrated that pharmacological CO donor decreased DCF fluorescence of LPS (1 µg/ml)-stimulated RAW 264.7 cells (39)
. In contrast to this study, our data did not show evidence of increased intracellular ROS production in response to LPS (10 ng/ml) in this macrophage cell line. This may be explained in part by the lower dose utilized in our study. Additionally, LPS-stimulation may increase NAD(P)H oxidase activity leading to increased extracellular ROS. There may be a differential influence of CO on the intracellular and extracellular ROS levels of stimulated macrophages, similar to the results seen by Taille et al. (15)
.
The concept of CO creating "metabolic hypoxia" has been suggested in previous work (40)
. Nystul et al. has previously demonstrated that CO can rescue C. elegans embryos from hypoxia-induced damage by inducing ‘suspended animation’. These embryos tolerate anoxia and mild hypoxia but do not tolerate intermediate levels of hypoxia. CO exposure decreases hypoxic-induced injury in these embryos exposed to otherwise toxic levels of hypoxia. Furthermore, both NO and CO can diminish hypoxia-induced signaling such as HIF-1 activation (41)
. We have previously shown that CO can prevent hemorrhagic shock-induced hepatic hypoxia, which may be related to decreased utilization of oxygen by cytochrome c oxidase (9)
. These data raise several interesting questions regarding the interaction of O2, NO, and CO. In addition to the need to consider the affinities of oxygen, NO, and CO with cytochrome c oxidase, the oxygen equilibrium constants of NO synthases, heme oxygenases, and cytochrome c oxidase must also be taken into account.
These data demonstrate that ATP levels are maintained in the setting of inhibition of aerobic respiration. Additionally, there is a hyperpolarization of the mitochondrial membrane potential. The orientation of the F1F0-ATPase must reverse and hydrolyze cytoplasmic ATP to effectively extrude mitochondrial protons and maintain the membrane potential (32
, 35)
. Similar to that which has been shown for NO, the influence of CO on membrane potential is lost in the presence of oligomycin, an inhibitor of the ATPase. To provide ATP, anaerobic respiration would be maintained or increased.
The influence of CO on inhibition of cytochrome c oxidase and the generation of superoxide by complexes I and III of the electron transport chain are highly localized to the mitochondria. Thus superoxide is generated in close proximity to SOD, which can convert superoxide to hydrogen peroxide leading to protective signaling. Furthermore, the production of ROS by the electron transport chain is likely to be at lower levels of ROS production compared to that of NAD(P)H oxidase.
In conclusion, CO limits LPS-induced inflammation in RAW 264.7 cells. This is dependent, at least in part, on inhibition of cytochrome c oxidase and subsequent mitochondrial ROS production. The influence of CO, both from exogenous delivery and the endogenous production by HO, on aerobic and anaerobic respiration require further examination. Investigations of metabolic and bioenergetic effects of HO/CO will likely yield insight into this intriguing system.
Received for publication August 11, 2006. Accepted for publication November 28, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Ricci, V. Pastukh, J. Leonard, J. Turrens, G. Wilson, D. Schaffer, and S. W. Schaffer Mitochondrial DNA damage triggers mitochondrial-superoxide generation and apoptosis Am J Physiol Cell Physiol, February 1, 2008; 294(2): C413 - C422. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
P<0.05 vs. CO+lipopolysaccharide), indicating that the anti-inflammatory effect of CO is at least, in part, secondary to ROS signaling. C) SOD/catalase diminished the CO-induced p38 phosphorylation in LPS treated (10 ng/ml) cells assayed by p38 MAPK ELISA (*P<0.05 vs. LPS alone; #P<0.05 vs. CO+lipopolysaccharide).
50% that of controls.
