Membrane-Bound Redox Enzyme Cytochrome bd-I Promotes Carbon Monoxide-Resistant Escherichia coli Growth and Respiration

The terminal oxidases of bacterial aerobic respiratory chains are redox-active electrogenic enzymes that catalyze the four-electron reduction of O2 to 2H2O taking out electrons from quinol or cytochrome c. Living bacteria often deal with carbon monoxide (CO) which can act as both a signaling molecule and a poison. Bacterial terminal oxidases contain hemes; therefore, they are potential targets for CO. However, our knowledge of this issue is limited and contradictory. Here, we investigated the effect of CO on the cell growth and aerobic respiration of three different Escherichia coli mutants, each expressing only one terminal quinol oxidase: cytochrome bd-I, cytochrome bd-II, or cytochrome bo3. We found that following the addition of CO to bd-I-only cells, a minimal effect on growth was observed, whereas the growth of both bd-II-only and bo3-only strains was severely impaired. Consistently, the degree of resistance of aerobic respiration of bd-I-only cells to CO is high, as opposed to high CO sensitivity displayed by bd-II-only and bo3-only cells consuming O2. Such a difference between the oxidases in sensitivity to CO was also observed with isolated membranes of the mutants. Accordingly, O2 consumption of wild-type cells showed relatively low CO sensitivity under conditions favoring the expression of a bd-type oxidase.


Introduction
Carbon monoxide (CO) is a well-known gaseous molecule that has long been recognized to mediate important physiological processes when produced in low amounts .In eukaryotes, CO is formed endogenously as a byproduct upon the degradation of heme to biliverdin and iron catalyzed by heme oxygenase [23].In bacteria, this gaseous molecule is generated by homologs of eukaryotic heme oxygenases and via alternative CO-producing mechanisms [24].Interestingly, CO is considered as a probable signaling molecule between the host and the gut microbiome [24].Some bacteria can also use CO as a source of energy and carbon [25].High concentrations of CO are toxic, and some pathogenic bacteria were reported to be susceptible either to CO produced by the host heme oxygenases or to transition-metal-based CO-releasing molecules (CORMs) [26,27].CORMs developed to deliver physiologically relevant levels of CO experimentally or therapeutically [28,29] showed an additive effect when combined with other antibiotics in certain microbes [26,27].However, care should be taken as one of the most widely used CORMs, the water-soluble Ru-containing CORM-3, was reported to exert cytotoxic effects due to a thiol-reactive Ru(II) ion and releases little CO [30].Thus, the development of novel effective CO-releasing drugs is an urgent problem as the therapeutic use of CO has emerged as an antimicrobial strategy in medicine.Bacterial proteins, which contain a pentacoordinate high-spin heme in the 2 of 20 reduced state, should bind CO as a strong exogenous ligand, and this, in turn, should affect their function.The terminal oxidases of bacterial respiratory chains have such a heme in their active sites and therefore are among these proteins.These redox-active enzymes belong to class EC 7 translocases.Terminal oxidases catalyze four-electron reduction of O 2 to 2H 2 O at the expense of oxidation of the respiratory substrate, either quinol or ferrocytochrome c.This redox reaction is coupled to the generation of proton-motive force which serves as the driving force for ATP synthesis and other useful work.There are two different groups of these enzymes found in bacteria: heme-copper oxidases, including aa 3 -type cytochrome c oxidase and bo 3 -type quinol oxidase, and copper-lacking bd-type quinol oxidases, also called cytochromes bd .Heme-copper oxidases are true proton pumps, whereas cytochromes bd generate proton-motive force solely due to the vector transfer of protons along the intraprotein proton-conducting pathway, without a mechanism of proton pumping [47,[52][53][54].Escherichia coli is a ubiquitous member of the gut microbiome of humans and warm-blooded animals.Like many other aerobic bacteria, it contains the branched and flexible O 2 -dependent respiratory chain.In the chain, type I and type II NADH dehydrogenases transfer electrons from NADH to ubiquinone-8 and/or menaquinone-8.Ubiquinone-8 can also accept electrons from succinate via succinate dehydrogenase.Then, electrons from ubiquinol-8 and/or menaquinol-8 are transferred by three terminal oxidases, cytochromes bo 3 , bd-I, and bd-II, to the final electron acceptor, O 2 , to produce 2H 2 O [55][56][57][58][59].
In E. coli, the bo 3 , bd-I, and bd-II oxidases are encoded by the cyoABCDE, cydABX, and appCBX operons, respectively.With the use of a rigorous chemostat methodology and powerful modeling tools, it was shown that the cyoABCDE operon was maximally induced under fully aerobic conditions [60].The cydABX operon was maximally expressed at 56% aerobiosis [61].The appCBX operon was maximally expressed at 0% aerobiosis.Spectroscopic assays of oxidase levels confirmed these conclusions [60,62].Thus, at high aeration, cytochrome bo 3 is preferentially expressed, whereas a shift from aerobic to microaerobic conditions activates the expression of the bd-type cytochromes [55].The atomic structures of the three cytochromes were reported [43,44,[63][64][65][66][67].Cytochrome bo 3 , composed of four subunits, carries the ubiquinol-binding site and three metal redox groups, a low-spin heme b, a high-spin heme o 3 , and Cu B .Heme b accepts electrons from ubiquinol; heme o 3 and Cu B compose a binuclear site for O 2 reduction [55].Cytochromes bd-I and bd-II consist of four and three subunits, respectively.Both bd enzymes contain the ubiquinol/menaquinolbinding site (called the Q-loop) and three hemes, a low-spin b 558 , a high-spin b 595 , and a high-spin d, but no copper ion.Heme b 558 is the immediate electron acceptor for ubiquinol and/or menaquinol.Heme d serves as the O 2 -reducing site.The functional role of heme b 595 is not clear yet, but at least apparently it transfers electrons from heme b 558 to heme d [55].
It is worth mentioning that cytochrome c and chlorophyll a were reported to be suitable redox mediators for the development of enzymatic biofuel cell systems [68].In these systems, electron transfer from glucose oxidase to the electrode was facilitated through cytochrome c and chlorophyll a adsorbed on the electrode [68].In this regard, the applicability of cytochrome d and cytochrome o 3 for charge transfer from/to bacteria and in biofuel cells could also be evaluated.
A protective role of cytochrome bd-I against nitric oxide [69][70][71], peroxynitrite [72], and ammonia [73] was reported.Cytochrome bd-I likely provides NO resistance in E. coli for two reasons.The first reason is the extraordinarily high rate of NO dissociation from heme d 2+ : the k off values were reported to be 0.133 s −1 [69] and 0.163 s −1 [70].This is about 30 times higher than that for NO dissociation from heme a 3 2+ in the mitochondrial cytochrome c oxidase [74] and can explain why following inhibition by NO the oxygen reductase activity of the bd-I enzyme is restored much faster than that of the mitochondrial enzyme.The second plausible reason is the ability of a bd oxidase to rapidly convert NO into NO 2 -in turnover, although that was reported only for cytochrome bd from Azotobacter vinelandii so far [75].The fact that the bd-I oxidase is not inhibited by peroxynitrite can be due to its ability to catalytically scavenge peroxynitrite [72].The reaction likely occurs on the heme d active site, and at least four possible reaction mechanisms have been suggested [76].Cytochrome bd-I is not only resistant to but also activated by ammonia under alkaline conditions [73].In this case, ammonia is suggested to react with a few catalytic intermediates of the enzyme.In particular, NH 3 can promote the formation of the peroxy state (P) from the oxidized state (O).NH 3 also possibly reacts with the one-electron-reduced state (O 1 ) to produce the ferryl state (F).In these reactions, NH 3 presumably serves as a two-electron donor being oxidized to NH 2 OH [73].
Both bd-type oxidases also contribute to E. coli resistance to cyanide [77], hydrogen peroxide [78][79][80], and sulfide [77,81].On the contrary, the activity of the bo 3 oxidase was shown to be highly sensitive to inhibition by cyanide, sulfide, nitric oxide, and ammonia [70,73,77,81].Thus, a bd-type terminal oxidase endows E. coli and possibly other bacteria with resistance to the above-mentioned stressors and, being absent in eukaryotic cells, can serve as a good therapeutic target [82].
Increased expression of cytochrome bd is in fact a likely mechanism for survival used by the pathogenic microorganisms in the presence of reactive oxygen and nitrogen species generated by the host immune system to fight infection [59].In view of the fact that cytochrome bd-I exhibits tolerance to such reactive species, it is relevant to examine its potential resistance also to CO, which has been reported to be important in host-pathogen relationships [83][84][85][86] and is a heme ligand like NO.It has to be noted that data on the effect of CO on the function of bacterial terminal oxidases are limited and contradictory.Of the three E. coli oxidases, cytochrome bo 3 was shown to be the least sensitive to inhibition by CO if the enzymes were purified and detergent-solubilized [87].In contrast, according to a recent short report [88], cytochrome bd-I is more resistant to inhibition by CO than cytochrome bd-II and cytochrome bo 3 if CO is added to E. coli cell suspensions, at [O 2 ] = 150 µM.Bayly et al. also studied the physiological response of Mycobacterium smegmatis to CO [83].The respiratory chain of mycobacteria is known to contain two different terminal oxidases: cytochrome bcc-aa 3 supercomplex and cytochrome bd [89].Bayly et al. reported that in M. smegmatis cell cultures, the activity of cytochrome bd is resistant to CO while cytochrome bcc-aa 3 supercomplex is strongly inhibited by CO [83].Furthermore, M. smegmatis lacking the bd oxidase shows a significant growth defect in the presence of this gas [83].
In this work, we studied the effect of CO on aerobic respiration sustained by bo 3 , bd-I, or bd-II oxidases in cell cultures of E. coli respiratory mutants by varying the O 2 concentration at which CO was added, and we examined the ability of these cytochromes to sustain bacterial growth in the presence of CO.We also tested the CO inhibition of O 2 consumption of both isolated membranes of the mutants and wild-type cells grown under conditions favoring the expression of either cytochrome bo 3 or a bd-type oxidase.

Effect of CO on E. coli Aerobic Respiration
We examined the effect of CO on the aerobic respiration of E. coli cells of the three different mutant strains.The aerobic respiratory chain of each mutant contains only one terminal quinol oxidase: cytochrome bd-I, cytochrome bd-II, or cytochrome bo 3 .In each mutant, aerobic cellular respiration was supplied by endogenous electron-donor respiratory substrates; therefore, the use of an exogenous reducing system was not required.Figure 1A shows that 96.3 µM CO added at [O 2 ] = 100 µM inhibits O 2 consumption by bd-I-only E. coli cells to a small extent of 11.6 ± 1.1%.Under these experimental conditions, the CO inhibitory effect on the aerobic respiration of bd-II-only and bo 3 -only E. coli cells is much greater, 43.3 ± 7.6% and 44.3 ± 1.5%, respectively (Figure 1B,C).We compared the inhibition of respiration of the three mutants by increasing [CO] added at three different O2 concentrations, 50, 100, and 200 μM (Figures 2-4).In the case of bd-I-only cells, at [O2] = 50, 100, and 200 μM, the maximum inhibition percentage at a maximum concentration of added CO, 196.3 μM, appeared to be 39.6 ± 4.5%, 18.0 ± 7.8%, and 9.7 ± 4.9%, respectively (Figure 2).The respective values were 59.3 ± 11.5%, 50.0 ± 3.6%, and 47.0 ± 6.0% for bd-II-only cells (Figure 3A-C) and 85.6 ± 3.7%, 65.3 ± 17.2%, and 39.7 ± 11.5% for bo3-only cells (Figure 4A-C).Thus, at all O2 concentrations studied, in E. coli cell suspensions, cytochrome bd-I turned out to be much more resistant to inhibition by CO than cytochrome bd-II or cytochrome bo3.One can see that in each mutant, the degree of inhibition decreases with increasing [O2].This suggests competitive enzyme inhibition; i.e., in all three terminal oxidases, CO competes with the substrate, O2, for binding to the enzyme's active site under turnover conditions.We compared the inhibition of respiration of the three mutants by increasing [CO] added at three different O 2 concentrations, 50, 100, and 200 µM (Figures 2-4).In the case of bd-I-only cells, at [O 2 ] = 50, 100, and 200 µM, the maximum inhibition percentage at a maximum concentration of added CO, 196.3 µM, appeared to be 39.6 ± 4.5%, 18.0 ± 7.8%, and 9.7 ± 4.9%, respectively (Figure 2).The respective values were 59.3 ± 11.5%, 50.0 ± 3.6%, and 47.0 ± 6.0% for bd-II-only cells (Figure 3A-C) and 85.6 ± 3.7%, 65.3 ± 17.2%, and 39.7 ± 11.5% for bo 3 -only cells (Figure 4A-C).Thus, at all O 2 concentrations studied, in E. coli cell suspensions, cytochrome bd-I turned out to be much more resistant to inhibition by CO than cytochrome bd-II or cytochrome bo 3 .One can see that in each mutant, the degree of inhibition decreases with increasing [O 2 ].This suggests competitive enzyme inhibition; i.e., in all three terminal oxidases, CO competes with the substrate, O 2 , for binding to the enzyme's active site under turnover conditions.As the inhibition of respiration of bd-II-only and bo3-only E. coli mutants by CO was significant, we were able to obtain the apparent half-maximal inhibitory concentration values, IC50, for CO added at different O2 concentrations.At [O2] = 50, 100, and 200 μM, the respective IC50 values were 88.6 ± 9.3, 170.4 ± 15.0, and 230.2 ± 12.0 μM CO for bd-IIonly cells (Figure 3A-C) and 66.5 ± 10.0, 130.6 ± 14.0, and 330.1 ± 19.6 μM CO for bo3-only cells (Figure 4A-C).In view of the insignificant inhibition, it was not possible to obtain IC50 for bd-I-only cells.
With the IC50 values at different O2 concentrations, we were in a position to estimate the inhibition constants (Ki) for CO in the case of bd-II-only and bo3-only E. coli cells.To achieve this goal, the IC50 values acquired at different [O2] values were plotted as a function of [O2]/Km(O2).The data were fitted to the appropriate equation assuming a competitive mode of inhibition [91].For analysis purposes, the previously reported Km(O2) values, 2 μM (for cytochrome bd-II [90]) and 6 μM (for cytochrome bo3 [70]), were used.As a result .The data were fitted to the appropriate equation assuming a competitive mode of inhibition [91].For analysis purposes, the previously reported K m (O 2 ) values, 2 µM (for cytochrome bd-II [90]) and 6 µM (for cytochrome bo 3 [70]), were used.As a result of this analysis, we obtained the following K i values for CO: 2.5 ± 0.2 µM for bd-II-only cells (Figure 3D) and 8.4 ± 0.7 µM for bo 3 -only cells (Figure 4D).
We also investigated the effect of CO on the aerobic respiration of wild-type E. coli cells.Transcriptomic studies on wild-type E. coli cultured at different O 2 availabilities, coupled with biochemical determination of respiratory oxidase expression [60,61], showed that both the transcript abundance of cyoA and cydA and the expression of the corresponding bo 3 and bd cytochromes are observed under fully aerobic and microaerobic conditions, respectively.Consistently, a change in oxidase expression from cytochrome bo 3 to the bd-type cytochromes was reported to occur with cell growth, following a progressive reduction in O 2 availability in the medium [77].In agreement with Forte et al. [77], when studying cells in an early growth phase of the culture (OD 600 < 0.8), most of the respiration (50-70%) is sensitive to 50 µM sulfide, pointing to a prevalent expression of cytochrome bo 3 .Conversely, in a late growth phase of the culture (OD 600 > 2.5), when O 2 is limiting, sulfide has little effect on respiration, indicating a prevalent expression of the bd-type cytochromes.Accordingly, O 2 consumption of wild-type cells that were harvested at high OD 600 (the prevalent expression of a bd-type oxidase) and low OD 600 (the prevalent expression of cytochrome bo 3 ) showed low and high sensitivity to CO, respectively (Figure 5).
showed that both the transcript abundance of cyoA and cydA and the expre corresponding bo3 and bd cytochromes are observed under fully aerobic and m conditions, respectively.Consistently, a change in oxidase expression from bo3 to the bd-type cytochromes was reported to occur with cell growth, follo gressive reduction in O2 availability in the medium [77].In agreement with [77], when studying cells in an early growth phase of the culture (OD600 < 0.8) respiration (50-70%) is sensitive to 50 μM sulfide, pointing to a prevalent e cytochrome bo3.Conversely, in a late growth phase of the culture (OD600 > 2.5 limiting, sulfide has little effect on respiration, indicating a prevalent expressi type cytochromes.Accordingly, O2 consumption of wild-type cells that were high OD600 (the prevalent expression of a bd-type oxidase) and low OD600 (t expression of cytochrome bo3) showed low and high sensitivity to CO, respect 5).In addition, we tested the CO inhibition of O2 consumption of membra from mutants.As with cell cultures, CO effectively inhibits the O2 reductas both bd-II-and bo3-containing membranes, whereas the activity of bd-I-conta branes is relatively resistant to inhibition by CO (Figure 6).In addition, we tested the CO inhibition of O 2 consumption of membranes isolated from mutants.As with cell cultures, CO effectively inhibits the O 2 reductase activity of both bd-IIand bo 3 -containing membranes, whereas the activity of bd-I-containing membranes is relatively resistant to inhibition by CO (Figure 6).
The addition of N 2 to respiring cells and membranes did not significantly alter the O 2 consumption rate, indicating that the observed inhibitory effect on wild-type cells at low OD values as well as on bd-II and bo 3 mutant strains is due to CO (Supplementary Figures S1 and S2).

Effect of CO on E. coli Cell Growth
The observation that the degree of resistance of the O 2 consumption process by bd-Ionly E. coli cells to CO is quite high, as opposed to the high sensitivity for the gas displayed by bd-II-only and bo 3 -only cells, prompted us to examine whether cytochrome bd-I, beyond enabling aerobic respiration, promotes E. coli cell growth in the presence of CO.In order to determine the effect of CO on E. coli cell growth, we studied the growth of the three different respiratory mutant strains in the presence of either ~20% CO or ~20% N 2 as a control.Following the addition of ~20% CO to the E. coli strain expressing cytochrome bd-I as the only terminal oxidase at 60 min after the start of growth in air, a minimal effect on cell growth was observed (Figure 7A).On the contrary, the growth of both bd-II-only and bo 3 -only strains was severely impaired over the same time window after the addition of ~20% CO, compared to the control with ~20% N 2 (Figure 7B,C).Thus, these data suggest that, unlike cytochromes bd-II and bo 3 , the bd-I oxidase promotes E. coli growth in the presence of CO.
x FOR PEER REVIEW 8 of 21 The addition of N2 to respiring cells and membranes did not significantly alter the O2 consumption rate, indicating that the observed inhibitory effect on wild-type cells at low OD values as well as on bd-II and bo3 mutant strains is due to CO (Supplementary Figures S1 and S2).

Effect of CO on E. coli Cell Growth
The observation that the degree of resistance of the O2 consumption process by bd-I-

Discussion
We examined the effect of CO on the O2-dependent respiration of the E. coli respiratory mutants containing one of the three terminal oxidases.The experiments were performed at different concentrations of the enzyme substrate, O2.In all mutants tested, the

Discussion
We examined the effect of CO on the O 2 -dependent respiration of the E. coli respiratory mutants containing one of the three terminal oxidases.The experiments were performed at different concentrations of the enzyme substrate, O 2 .In all mutants tested, the inhibitory action of CO decreased with increasing [O 2 ].The fact that the degree of inhibition decreases as the substrate concentration increases clearly suggests that CO acts as a competitive inhibitor for the enzyme-catalyzed O 2 reduction reaction under steady-state conditions.CO apparently competes with O 2 for binding to a high-spin pentacoordinate ferrous heme in the enzyme's active site.Previous spectroscopic studies showed that the redox-active group that, in the reduced state, is able to bind both O 2 and CO is heme d in the bd-type oxidases and heme o 3 in the bo 3 oxidase [87,92,93].For this reason, we think that in the present enzyme-inhibition experiments with E. coli cells, heme d and heme o 3 are targets for CO in cytochromes bd-I/bd-II and cytochrome bo 3 , respectively.
We found that the bd-I oxidase is much less sensitive to CO than both the bd-II and bo 3 oxidases, as confirmed by measuring the effect of CO on O 2 consumption by both whole cells and isolated membranes of the respiratory mutants.Accordingly, the O 2 consumption of wild-type cells displayed low CO sensitivity under conditions favoring the expression of a bd-type oxidase and high CO sensitivity when cytochrome bo 3 was predominantly expressed.Consistently, cell growth proved to be almost unaffected by CO in an E. coli mutant strain expressing solely cytochrome bd-I, but was drastically inhibited after the addition of the gas to mutant strains expressing either cytochrome bo 3 or cytochrome bd-II as the only terminal oxidase.These results are in disagreement with an earlier study according to which O 2 consumption by the isolated cytochromes bd-I and bd-II is more sensitive to inhibition by CO than that by the isolated bo 3 oxidase [87].The measurement conditions, such as buffer, pH, and temperature, in [87] and in this work were the same, but the difference was the environment for the enzyme (detergent in [87] versus natural lipid bilayer in this study).Therefore, we assume that this inconsistency can be explained by differences in the protein environment which affect enzyme sensitivity towards CO.In this work, the oxidases were in vivo conditions, i.e., integrated into native bacterial membranes, whereas in [87], the enzymes were isolated and incorporated into detergent micelles.Indeed, more and more data are accumulating on how the lipid membrane environment can significantly affect the structure, function, and dynamics of various membrane proteins [94,95].For instance, regarding terminal oxidases, it was shown that the membrane environment modulates the ligand-binding characteristics of the E. coli cytochrome bd-I [96].According to another study, solubilization of the membrane-bound bovine cytochrome c oxidase leads to an increase in the binding affinity of the enzyme for cyanide by 100-1000 times [97].In addition, it was reported that in E. coli membranes, the bd-I oxidase is apparently in a supercomplex with other membrane-bound respiratory enzymes [56].Such protein-protein interactions could also modulate the sensitivity of cytochrome bd-I to CO.
We hypothesize that when the E. coli cytochromes are integrated into the native lipid bilayer, the binding affinities of the enzymes for CO and O 2 are such that CO binding to heme d in the bd-I oxidase is outcompeted by O 2 while the binding of the inhibitor to heme d in the bd-II oxidase or to heme o 3 in the bo 3 oxidase is not.The resistance of cytochrome bd-I to CO is in agreement with the extremely fast dissociation of CO from the oxidase (k off = 6.0 ± 0.2 s −1 for the fully reduced form of the CO-bound enzyme [69]).Such a high enzyme-ligand dissociation rate constant would indeed lead to the prompt restoration of respiration.The variation between the terminal oxidases in the affinity for the ligands may arise from differences in the structural organization of the O 2 -binding sites and their specific environment.Consistently, among the factors that can regulate the binding affinity of heme proteins for gaseous ligands are the chemical structure and geometry of the proximal axial ligand of the heme and its distal amino acid residues [98].Distal amino acid residues may either stabilize the bound diatomic gaseous molecule via weak interactions, including electrostatic effects, hydrogen bonds, van der Waals interactions, and the hydrophobic effect, or, contrariwise, destabilize ligand binding by providing steric constraints [98,99].
It is of interest to mention global E. coli responses to CO. Transcription factor measurements and modeling showed that gene expression is significantly perturbed by CO, with major consequences for energy metabolism, iron homeostasis, and amino acid metabolism [86].Genes encoding energy-transducing proteins are highly affected by CO via the global regulators, ArcA and FNR.CO inhibition of respiration results in over-reduction of the quinone pool; accumulation of the fermentation product, pyruvate; and enhanced expression of iron acquisition genes.Regarding the respiratory terminal oxidases, the transcriptomic analysis of wild-type E. coli exposed to CO gas revealed that under aerobic conditions there is a 5-to 10-fold decrease in the expression of the cyoABCDE operon, encoding cytochrome bo 3 .In contrast, there is a 4-fold increase in the expression of the cydABX operon, encoding cytochrome bd-I.Changes in the expression of the appCBX operon, encoding cytochrome bd-II oxidase, are slight [86].The CO inhibition data reported here are consistent with the transcriptional response to CO in wild-type cells reported by Wareham et al. [86].Indeed, it seems reasonable that following CO gas addition, the CO-sensitive bo 3 oxidase is downregulated and the CO-insensitive bd-I oxidase is upregulated.The bd-II oxidase normally is not expressed under aerobic conditions, and there is no need for its upregulation after CO treatment since the enzyme is sensitive to CO.
The difference in susceptibility to CO inhibition between cytochromes bd-I and bo 3 is indeed not unexpected since they belong to different superfamilies of terminal oxidases and, as reported before, also differ in resistance to other stressors, such as cyanide, sulfide, nitric oxide, and ammonia [70,73,77,81].However, the fact that cytochrome bd-I is much more resistant to CO than cytochrome bd-II in respiring E. coli cells is surprising.In order to explain this phenomenon, the following structural differences between the two bd enzymes have to be noted.First, the two proteins differ in the number of subunits.The bd-I oxidase is composed of four subunits, namely CydA, CydB, CydX, and CydY, whereas the bd-II oxidase contains one fewer subunit, being composed of AppC, AppB, and AppX.Notably, the extra subunit in cytochrome bd-I, CydY, shields the high-spin pentacoordinate heme b 595 from the lipid bilayer interface.This shielding prevents external ligands from accessing this potential ligand-binding site [64,65].Consistently, heme b 595 resistance to external ligand binding in cytochrome bd-I was previously identified in an MCD study [92].Since cytochrome bd-II has no CydY, direct access of gaseous molecules from the membrane lipid environment is possible [66,67].Therefore, in the bd-II oxidase, CO could possibly bind to the reduced heme b 595 under turnover conditions.This binding would, in turn, inhibit the electron transfer from heme b 558 to heme d that occurs through heme b 595 , leading to the inhibition of the enzyme-catalyzed O 2 consumption.Second, it was reported that in the bd-II oxidase, the axial heme d ligand is His 19 of the AppC subunit [66,67].However, there is no consensus on the nature of the axial ligand of heme d in the bd-I oxidase.It is either His 19 (of the CydA subunit, homolog to AppC) [64] or Glu 99 of CydA [65].Notably, His 19 and Glu 99 are located on opposite sides of the plane of the porphyrin macrocycle.If the axial heme d ligand in the bd-I and bd-II enzymes in bacterial cells is indeed different, this may also underlie the difference in sensitivity of the two proteins to CO.The fact that heme d in the bd-II oxidase has a much higher midpoint redox potential than that in the bd-I oxidase (+440 vs. +258 mV) [28]) would be consistent with the different nature of its axial ligand.Third, the bd-II protein incorporated into amphipols was shown to be mainly in the form of a dimer, while the bd-I enzyme exists only as a monomer [66].This difference might also contribute to the observed difference in the sensitivity of these two bd oxidases to CO.In this regard, it is also worth noting that Cupriavidus necator H16, like E. coli and some other bacteria, has two different operons encoding cytochrome bd, cydA1B1 and cydA2B2.Interestingly, the expression of only one of the two (cydA2B2) seems to enable cell growth in the presence of CO under heterotrophic conditions [100].The deletion of cydA2B2 had a detrimental effect on CO resistance, and plasmid-based expression of cydA1B1 did not improve CO tolerance [64].These data are consistent with our observation that, of the two bd oxidases in E. coli cells, only cytochrome bd-I promotes aerobic respiration and growth in the presence of CO.
The possible molecular mechanisms of the inhibition of the catalytic activity of the E. coli cytochromes bo 3 , bd-II, and bd-I by CO are shown in Figures 8-10, respectively (for more detail, see the legends to the figures).
The possible molecular mechanisms of the inhibition of the catalytic activity of the E. coli cytochromes bo3, bd-II, and bd-I by CO are shown in Figures 8-10, respectively (for more detail, see the legends to the figures).[101].CO binds to heme o3 2+ in the R species, forming the o3 2+ -CO complex that prevents the binding of O2 to heme o3 2+ , and, as a consequence, leads to the inhibition of the enzyme activity.  where d* 4+ =O 2 is a ferryl porphyrin π-cation radical [102,103]), and F (b558 3+ b595 3+ d 4+ =O 2− ) are shown.In the O 1 species, an electron is probably distributed between the three hemes.The reaction of CO with O 1 stabilizes the electron on heme d, and the d 2+ -CO complex The possible molecular mechanisms of the inhibition of the catalytic activity of the E. coli cytochromes bo3, bd-II, and bd-I by CO are shown in Figures 8-10, respectively (for more detail, see the legends to the figures).[101].CO binds to heme o3 2+ in the R species, forming the o3 2+ -CO complex that prevents the binding of O2 to heme o3 2+ , and, as a consequence, leads to the inhibition of the enzyme activity.is generated.This, in turn, prevents the binding of O2 to heme d 2+ and, as a consequence, leads to the inhibition of the enzyme activity.Furthermore, CO could also bind to heme b595 2+ in the A 3 species, producing the b595 2+ -CO complex.This would stabilize heme b595 in the reduced state and would not allow heme b595 2+ to rapidly donate an electron to O2 bound to heme d 2+ to perform concerted fourelectron reduction of O2 to 2H2O.  2 where d* 4+ =O 2 is a ferryl porphyrin π-cation radical [66,67]), and F (b558 3+ b595 3+ d 4+ =O 2− ) are shown.CO reacts with the O 1 species, yielding the d 2+ -CO complex, as suggested for cytochrome bd-II.However, at variance with cytochrome bd-II, in the case of cytochrome bd-I, due to the unusually high off-rate [69], CO does not bind with high affinity to heme d 2+ and is rapidly ejected from the enzyme.As a result, CO does not affect its catalytic activity much.Also, in cytochrome bd-I, heme b595 2+ does not bind CO to any significant extent [92].These features thus presumably make cytochrome bd-I relatively insensitive to CO.
In this work, we have used mutant strains, and this may present limitations, as mutations can potentially influence other processes in the cell that might go unnoticed and remain unaccounted for.However, the data on the wild-type cells and membranes are consistent with those obtained with the mutant strains.This gives us reason to believe that our conclusions are correct.As we mentioned above, the addition of CO triggers shortterm alterations in the E. coli transcriptome [86], but a complete picture of the impact of CO on bacterial bioenergetics is lacking.Further studies are needed to shed light on longterm adaptation effects and fully uncover the extent of CO resistance in vivo.

Materials, E. coli Mutant Strains and Growth Conditions
The CO and N2 gases were purchased from Linde (Danbury, CT, USA) and Air Liquide (Air Liquide Italia Spa, Milano, Italy), respectively.Other chemicals were purchased from Merck KGaA (Darmstadt, Germany).A stock solution of CO or N2 was prepared by equilibrating degassed water with the pure gas at 1 atm and room temperature, yielding 1 mM CO and 0.7 mM N2 in solution.E. coli respiratory mutant strains TBE025 (MG1655 ΔcydB nuoB appB::kan), TBE026 (MG1655 ΔcydB nuoB cyoB::kan), and TBE037 (MG1655 ΔappB nuoB cyoB::kan), which respectively express cytochrome bo3, cytochrome bd-II, or cytochrome bd-I as the sole terminal oxidase, were used [77].E. coli cultures were grown in Luria-Bertani (LB) medium supplemented with 30 μg/mL kanamycin, at 37 °C and 200 rpm.In the case of growth studies, after inoculation at an OD600 of 0.15 ± 0.03, the These features thus presumably make cytochrome bd-I relatively insensitive to CO.
In this work, we have used mutant strains, and this may present limitations, as mutations can potentially influence other processes in the cell that might go unnoticed and remain unaccounted for.However, the data on the wild-type cells and membranes are consistent with those obtained with the mutant strains.This gives us reason to believe that our conclusions are correct.As we mentioned above, the addition of CO triggers short-term alterations in the E. coli transcriptome [86], but a complete picture of the impact of CO on bacterial bioenergetics is lacking.Further studies are needed to shed light on long-term adaptation effects and fully uncover the extent of CO resistance in vivo.

Materials, E. coli Mutant Strains and Growth Conditions
The CO and N 2 gases were purchased from Linde (Danbury, CT, USA) and Air Liquide (Air Liquide Italia Spa, Milano, Italy), respectively.Other chemicals were purchased from Merck KGaA (Darmstadt, Germany).A stock solution of CO or N 2 was prepared by equilibrating degassed water with the pure gas at 1 atm and room temperature, yielding 1 mM CO and 0.7 mM N 2 in solution.E. coli respiratory mutant strains TBE025 (MG1655 ∆cydB nuoB appB::kan), TBE026 (MG1655 ∆cydB nuoB cyoB::kan), and TBE037 (MG1655 ∆appB nuoB cyoB::kan), which respectively express cytochrome bo 3 , cytochrome bd-II, or cytochrome bd-I as the sole terminal oxidase, were used [77].E. coli cultures were grown in Luria-Bertani (LB) medium supplemented with 30 µg/mL kanamycin, at 37 • C and 200 rpm.In the case of growth studies, after inoculation at an OD 600 of 0.15 ± 0.03, the cells were grown in 50 mL rubber-stoppered flasks in 16 mL of air-equilibrated LB at 37 • C for 60 min; when OD 600 reached 0.4 ± 0.08, they were bubbled for 30 s with either CO (pure gas at 1 atm and room temperature) or N 2 (pure gas at 1 atm and room temperature) as a control to yield a final concentration of both gases of ~20%.concentration values measured in the degassed chamber of a high-resolution respirometer (Oxygraph-2k, Oroboros Instruments GmbH, Innsbruck, Austria) of non-fluxed and fluxed E. coli cultures with CO or N 2 , respectively.The growth of the E. coli cultures was then monitored via a standard method using optical density measurements in an Eppendorf BioSpectrometer basic at 600 nm every 30 min.When the OD 600 was above 1, cultures were diluted before reading.

Investigation of the Effect of CO on Respiration of Wild-Type E. coli Cells
To assess the effect of CO on the respiration of wild-type E. coli cells (strain MG1655), we investigated aerobic cultures in which a change in oxidase expression from cytochrome bo 3 to the bd-type cytochromes is expected to occur during cell growth, following a progressive reduction in the availability of O 2 in the medium, taking into account the striking difference between the oxidases in sensitivity to sodium sulfide, according to Forte et al. [77].When cells grown in Luria-Bertani (LB) medium are assayed in an early growth phase of the culture (OD 600 < 0.8), most of the respiration (50-70%) is sensitive to 50 µM sodium sulfide, indicating a prevalent expression of cytochrome bo 3 .In contrast, in a late growth phase of the culture (OD 600 > 2.5), when oxygen is limiting, sodium sulfide causes only marginal effects on respiration, indicating a prevalent expression of the bd-type cytochromes.

Isolation of Membranes from E. coli Respiratory Mutants
To isolate membranes from E. coli respiratory mutants expressing only one terminal quinol oxidase (cytochrome bd-I or cytochrome bd-II or cytochrome bo 3 ), cell cultures were grown in LB medium until an OD 600 of about 2 was reached.The cells were then pelleted by centrifugation at 10,000 rpm for 10 min and washed twice in 20 mM TRIS pH 8.3 containing 0.5 mM EDTA and 5 mM MgSO 4 .The cells were resuspended in the same buffer containing 1 mg/mL lysozyme and incubated for two hours on ice.A spatula tip of DNAse and RNase was then added, and the cells were lysed by sonication.Cell debris was removed by centrifugation at 15,000 rpm for 10 min.Membrane fractions were collected and stored at −80 • C. The protein content was determined using the Bradford method, using the Bradford reagent (Sigma) with bovine serum albumin as standard.

Oxygraphic Measurements
Oxygraphic measurements were performed at 25 • C in 50 mM K/phosphate pH 7.0, using a high-resolution respirometer (Oxygraph-2k, Oroboros Instruments GmbH, Innsbruck, Austria) equipped with two 1.5-mL chambers.O 2 consumption of E. coli cells was followed with endogenous reductants.In the case of isolated membranes, O 2 consumption was measured in the presence of 2.5 mM dithiothreitol (DTT) and 2.5 µM 2,3-dimethoxy-5methyl-6-(3-methyl-2-butenyl)-1,4-benzoquinone (Q 1 ), an artificial reducing couple specific for quinol oxidases.The catalytic O 2 -consuming activity was obtained by subtracting the non-enzymatic O 2 consumption (Q 1 /DTT autoxidation in the absence of the membranes) from the O 2 consumption rate measured after the addition of membranes.

Spectroscopic Measurements
UV-visible absorption spectra were recorded to estimate the amount of each terminal oxidase present in cell suspensions.For this purpose, an Agilent Cary 60 UV-Vis or a Varian Cary 300 Bio UV-Visible spectrophotometer was used.The amount of oxidase present in each strain was estimated from the dithionite-reduced-minus-ferricyanyde-oxidized difference absorption spectrum of sonicated cells using ∆ε 561-580 of 21 mM −1 cm −1 (bd-I and bd-II) [92] and 16.3 mM −1 cm −1 (bo 3 ) [88].

Data Analysis
Data analysis was carried out using software packages Origin v7.0 (OriginLab Corporation, Northampton, MA, USA) and GIM (Scientific Graphic Interactive Management System developed by A.L. Drachev in Lomonosov Moscow State University).The percentage inhibition of O 2 consumption of cell suspensions (i%) was calculated using the equation i% = ((V 0 − V 0,i )/V 0 )•100, where V 0 and V 0,i are the initial rates in the ab- sence and in the presence of the inhibitor (CO), respectively.The apparent IC 50 values of cytochromes bd-II and bo 3 for CO were estimated by plotting i% as a function of CO concentration added ([CO] 0 ).The data were fitted to the standard hyperbolic equation i% = i max %•[CO] /(IC 50 + [CO] 0 ) using a built-in approximation function ('Hyperbola function') in 'Advanced Fitting tool' in the Origin program.The i max % parameter is a theoretical maximum percent inhibition.The inhibition constants (K i ) of cytochromes bd-II and bo 3 for CO were estimated by plotting, as a function of [O 2 ]/K m (O 2 ), the IC 50 values measured at different O 2 concentrations and fitting the data to the equation IC 50 = (K i •[O 2 ]/K m (O 2 )) + K i [91], assuming O 2 competitive inhibition.The apparent K m (O 2 ) values 2 µM (for cytochrome bd-II) and 6 µM (for cytochrome bo 3 ) were taken from [90] and [70], respectively.
Statistical analyses were performed using an unpaired t-test for comparisons between two strains or conditions.

Conclusions
In this study, we show that of the two E. coli bd-type oxidases, only bd-I promotes growth in the presence of toxic concentrations of CO, giving the bacterium resistance to the gas.Our results are relevant for clarifying the discrepancies present in the literature on the effect of CO on the O 2 consumption of E. coli oxidases and expanding the knowledge of cytochromes bd, a protein family of increasing interest due to their unique functional and structural features and their importance to pathogens.These findings are also important in the field of microbial physiology as they contribute to a better understanding of how different terminal complexes participate in the respiratory chain under various growth conditions.Moreover, they may provide a basis for biotechnological applications in which an increased bacterial resistance to CO is needed [100].Finally, we believe that these findings could have a biomedical significance.These data should be taken into account when next-generation antimicrobials, whose mechanism of action is based on the release of CO that blocks the aerobic respiration and growth of pathogenic bacteria, are studied for potential use in medicine.Therapeutic CO delivery in humans and animals by such CO-releasing molecules would most likely be less efficient if the target pathogen possesses a terminal oxidase similar to the E. coli bd-I enzyme.

Figure 1 .
Figure 1.Typical O2 consumption traces showing the effect of CO on E. coli aerobic respiration in cell suspensions of mutant strains expressing cytochrome bd-I (A), cytochrome bd-II (B), or cytochrome bo3 (C) as the sole terminal oxidase.Here, 96.3 μM CO was added at [O2] = 100 μM.The O2 consumption rates (nM O2/s) measured prior to and following the addition of CO are shown adjacent to each trace.Concentrations of bd-I, bd-II, and bo3 oxidases were 75, 52, and 47 nM, respectively.

Figure 1 .
Figure 1.Typical O 2 consumption traces showing the effect of CO on E. coli aerobic respiration in cell suspensions of mutant strains expressing cytochrome bd-I (A), cytochrome bd-II (B), or cytochrome bo 3 (C) as the sole terminal oxidase.Here, 96.3 µM CO was added at [O 2 ] = 100 µM.The O 2 consumption rates (nM O 2 /s) measured prior to and following the addition of CO are shown adjacent to each trace.Concentrations of bd-I, bd-II, and bo 3 oxidases were 75, 52, and 47 nM, respectively.

Figure 3 .
Figure 3. CO inhibition of aerobic respiration of cell suspensions of E. coli mutant strain expressing cytochrome bd-II as the sole terminal oxidase at different O2 concentrations.(A,B,C) Determination

Figure 3 .Figure 3 .
Figure 3. CO inhibition of aerobic respiration of cell suspensions of E. coli mutant strain expressing cytochrome bd-II as the sole terminal oxidase at different O2 concentrations.(A,B,C) DeterminationFigure 3. CO inhibition of aerobic respiration of cell suspensions of E. coli mutant strain expressing cytochrome bd-II as the sole terminal oxidase at different O 2 concentrations.(A-C) Determination of apparent IC 50 .Square symbols are experimental data points.Measurements were performed at increasing [CO] added at [O 2 ] = 50 µM (A), 100 µM (B), and 200 µM (C).(D) Determination of K i .It was estimated using the depicted IC 50 values (circle symbols), assuming competitive inhibition.The IC 50 value obtained at [O 2 ] = 150 µM (187.1 ± 11.1 µM CO) was taken from [88].The apparent K m (O 2 ) value (2 µM) was taken from [90].Concentrations of bd-II oxidase in experiments in which CO was added at 50, 100, and 200 µM O 2 were 53 ± 11, 51 ± 8, and 51 ± 11 nM, respectively.Values represent the mean (n = 3) ± standard deviation.

Figure 4 .
Figure 4. CO inhibition of aerobic respiration of cell suspensions of E. coli mutant strain expressing cytochrome bo 3 as the sole terminal oxidase at different O 2 concentrations.(A-C) Determination of apparent IC 50 .Rhombus symbols are experimental data points.Measurements were performed at increasing [CO] added at [O 2 ] = 50 µM (A), 100 µM (B), and 200 µM (C).(D) Determination of K i .It was estimated using the depicted IC 50 values (circle symbols), assuming competitive inhibition.The IC 50 value obtained at [O 2 ] = 150 µM (183.3 ± 13.5 µM CO) was taken from [52].The apparent K m (O 2 ) value (6 µM) was taken from [70].Concentrations of bo 3 oxidase in experiments in which CO was added at 50, 100, and 200 µM O 2 were 58 ± 17, 56 ± 9, and 57 ± 9 nM, respectively.Values represent the mean (n = 3) ± standard deviation.As the inhibition of respiration of bd-II-only and bo 3 -only E. coli mutants by CO was significant, we were able to obtain the apparent half-maximal inhibitory concentration values, IC 50 , for CO added at different O 2 concentrations.At [O 2 ] = 50, 100, and 200 µM, the respective IC 50 values were 88.6 ± 9.3, 170.4 ± 15.0, and 230.2 ± 12.0 µM CO for bd-II-only cells (Figure 3A-C) and 66.5 ± 10.0, 130.6 ± 14.0, and 330.1 ± 19.6 µM CO for bo 3 -only cells (Figure 4A-C).In view of the insignificant inhibition, it was not possible to obtain IC 50 for bd-I-only cells.With the IC 50 values at different O 2 concentrations, we were in a position to estimate the inhibition constants (K i ) for CO in the case of bd-II-only and bo 3 -only E. coli cells.To achieve this goal, the IC 50 values acquired at different [O 2 ] values were plotted as a function of [O 2 ]/K m (O 2 ).The data were fitted to the appropriate equation assuming a competitive mode of inhibition[91].For analysis purposes, the previously reported K m (O 2 ) values, 2 µM (for cytochrome bd-II[90]) and 6 µM (for cytochrome bo 3[70]), were used.As a result of this analysis, we obtained the following K i values for CO: 2.5 ± 0.2 µM for bd-II-only cells (Figure3D) and 8.4 ± 0.7 µM for bo 3 -only cells (Figure4D).We also investigated the effect of CO on the aerobic respiration of wild-type E. coli cells.Transcriptomic studies on wild-type E. coli cultured at different O 2 availabilities, coupled

Figure 5 .
Figure 5. Traces of O2 consumption of cell suspensions showing the effect of CO on ae tion of the E. coli wild-type (wt) strain MG1655 at high OD (blue line) and low OD (r cating the prevalent expression of the bd oxidases and cytochrome bo3, respectively.H CO was added at [O2] = 100 μM.The O2 consumption rates (nM O2/s) measured prior ing the addition of CO are shown adjacent to each trace.Additions: 1 mL of cells at line), 0.2 mL of cells at 3 OD (blue line).Inset: Percent activity after CO addition to re type cells.Asterisks denote statistically significant differences between wild-type cells and low OD (**, p < 0,01; t-test).

Figure 5 .
Figure 5. Traces of O 2 consumption of cell suspensions showing the effect of CO on aerobic respiration of the E. coli wild-type (wt) strain MG1655 at high OD (blue line) and low OD (red line), indicating the prevalent expression of the bd oxidases and cytochrome bo 3 , respectively.Here, 96.3 µM CO was added at [O 2 ] = 100 µM.The O 2 consumption rates (nM O 2 /s) measured prior to and following the addition of CO are shown adjacent to each trace.Additions: 1 mL of cells at 0.48 OD (red line), 0.2 mL of cells at 3 OD (blue line).Inset: Percent activity after CO addition to respiring wild-type cells.Asterisks denote statistically significant differences between wild-type cells grown at high and low OD (**, p < 0,01; t-test).

Figure 6 .
Figure 6.Traces of O2 consumption showing the effect of CO on aerobic respiration of isolated membranes from E. coli mutant strains expressing cytochrome bd-I (A), cytochrome bd-II (B), or cytochrome bo3 (C) as the sole terminal oxidase.(D) Percent of O2 reductase activity after CO additions to respiring membranes.Asterisks denote statistically significant differences with respect to bd-I activity (**, p < 0,01; t-test).Here, 96.3 μM CO was added at [O2] = 100 μM.The O2 consumption rates (nM O2/s) measured prior to and following the addition of CO are shown adjacent to each trace.Additions: 0.3 mg/mL of bd-I-containing isolated membranes (black line), 0.5 mg/mL of bd-II-containing isolated membranes (blue line), 0.6 mg/mL of bo3-containing isolated membranes (red line).

Figure 6 .
Figure 6.Traces of O 2 consumption showing the effect of CO on aerobic respiration of isolated membranes from E. coli mutant strains expressing cytochrome bd-I (A), cytochrome bd-II (B), or cytochrome bo 3 (C) as the sole terminal oxidase.(D) Percent of O 2 reductase activity after CO additions to respiring membranes.Asterisks denote statistically significant differences with respect to bd-I activity (**, p < 0,01; t-test).Here, 96.3 µM CO was added at [O 2 ] = 100 µM.The O 2 consumption rates (nM O 2 /s) measured prior to and following the addition of CO are shown adjacent to each trace.Additions: 0.3 mg/mL of bd-I-containing isolated membranes (black line), 0.5 mg/mL of bd-II-containing isolated membranes (blue line), 0.6 mg/mL of bo 3 -containing isolated membranes (red line).

Figure 7 .
Figure 7. Effect of CO on E. coli cell growth.Cell growth of E. coli mutant strains expressing cytochrome bd-I (A), cytochrome bd-II (B), or cytochrome bo 3 (C) as the sole terminal oxidase assayed in the presence of either ~20% CO ('closed symbols') or ~20% N 2 ('open symbols').The arrow shows the time (60 min) at which cells were subjected to the gas flushing treatment for 30 s. Values represent the mean (n = 3) ± standard deviation.Asterisks denote statistically significant differences between CO-and N 2 -treated cells (*, p < 0,05; **, p < 0,01; t-test).

Figure 8 .
Figure 8. Proposed mechanism for the inhibitory effects of CO on the catalytic cycle of cytochrome bo3.Catalytic intermediates O (o3 3+ -OH CuB 2+ ), R (o3 2+ CuB + ), A (o3 2+ -O2 CuB + ), P (o3 4+ =O 2-CuB 2+ -OH), and F (o3 4+ =O 2-CuB 2+ ) are shown.Only chemical protons are shown.Pumped protons are not shown for clarity.The two ferryl intermediates P and F probably differ in the presence of an aromatic amino acid radical in P, as in the PM species of aa3-type cytochrome c oxidase[101].CO binds to heme o3 2+ in the R species, forming the o3 2+ -CO complex that prevents the binding of O2 to heme o3 2+ , and, as a consequence, leads to the inhibition of the enzyme activity.

Figure 8 .
Figure 8. Proposed mechanism for the inhibitory effects of CO on the catalytic cycle of cytochrome bo 3 .Catalytic intermediates O (o 3 3+ -OH Cu B 2+ ), R (o 3 2+ Cu B + ), A (o 3 2+ -O 2 Cu B + ), P (o 3 4+ =O 2− Cu B 2+ -OH), and F (o 3 4+ =O 2− Cu B 2+ ) are shown.Only chemical protons are shown.Pumped protons are not shown for clarity.The two ferryl intermediates P and F probably differ in the presence of an aromatic amino acid radical in P, as in the P M species of aa 3 -type cytochrome c oxidase [101].CO binds to heme o 3 2+ in the R species, forming the o 3 2+ -CO complex that prevents the binding of O 2 to heme o 3 2+ , and, as a consequence, leads to the inhibition of the enzyme activity.

Figure 8 .
Figure 8. Proposed mechanism for the inhibitory effects of CO on the catalytic cycle of cytochrome bo3.Catalytic intermediates O (o3 3+ -OH CuB 2+ ), R (o3 2+ CuB + ), A (o3 2+ -O2 CuB + ), P (o3 4+ =O 2-CuB 2+ -OH), and F (o3 4+ =O 2-CuB 2+ ) are shown.Only chemical protons are shown.Pumped protons are not shown for clarity.The two ferryl intermediates P and F probably differ in the presence of an aromatic amino acid radical in P, as in the PM species of aa3-type cytochrome c oxidase[101].CO binds to heme o3 2+ in the R species, forming the o3 2+ -CO complex that prevents the binding of O2 to heme o3 2+ , and, as a consequence, leads to the inhibition of the enzyme activity.
The gas percentage was assessed by taking into account the different O 2 concentrations, present in equal volumes of LB equilibrated with air or flushed with CO/N 2 for 30 s, and calculated according to the following equation: Gas % = ([O 2 ] air equilibrated − [O 2 ] gas flushed )/[O 2 ] air equilibrated × 100.[O 2 ] air equilibrated and [O 2 ] gas flushed are the O 2