Skip to main content
PMC home page
Nature Communications logoLink to Nature Communications
. 2021 Jul 30;12:4621. doi: 10.1038/s41467-021-24924-w

Cryo-EM structure of mycobacterial cytochrome bd reveals two oxygen access channels

Weiwei Wang 1,2,3,4,#, Yan Gao 1,#, Yanting Tang 2, Xiaoting Zhou 1,3,4, Yuezheng Lai 2, Shan Zhou 2, Yuying Zhang 2, Xiuna Yang 1, Fengjiang Liu 1,3,4, Luke W Guddat 5, Quan Wang 1,✉, Zihe Rao 1,2,6,7,✉, Hongri Gong 2,✉
PMCID: PMC8324918  PMID: 34330928

Abstract

Cytochromes bd are ubiquitous amongst prokaryotes including many human-pathogenic bacteria. Such complexes are targets for the development of antimicrobial drugs. However, an understanding of the relationship between the structure and functional mechanisms of these oxidases is incomplete. Here, we have determined the 2.8 Å structure of Mycobacterium smegmatis cytochrome bd by single-particle cryo-electron microscopy. This bd oxidase consists of two subunits CydA and CydB, that adopt a pseudo two-fold symmetrical arrangement. The structural topology of its Q-loop domain, whose function is to bind the substrate, quinol, is significantly different compared to the C-terminal region reported for cytochromes bd from Geobacillus thermodenitrificans (G. th) and Escherichia coli (E. coli). In addition, we have identified two potential oxygen access channels in the structure and shown that similar tunnels also exist in G. th and E. coli cytochromes bd. This study provides insights to develop a framework for the rational design of antituberculosis compounds that block the oxygen access channels of this oxidase.

Subject terms: Cryoelectron microscopy, Cryoelectron microscopy

Cytochromes bd oxidase (Cyt-bd) catalyzes the reduction of oxygen to water and is the terminal oxidase in the respiratory chain of prokaryotes. Here, the authors present the 2.8 Å cryo-EM structure of Mycobacterium smegmatis Cyt-bd and identify two potential oxygen access channels in the structure, which is of interest for the development of novel antituberculosis drugs.

Introduction

Respiratory oxygen reductases (terminal oxidases) comprise a series of structurally distinct enzymes that are widely distributed across all kingdoms of life. The heme-copper oxidases (HCO) and bd-type oxidases (cytochromes bd) are two well-known types of membrane-integrated terminal oxidases1,2. They catalyze the reduction of molecular oxygen (O2) to water by the respiratory substrate, cytochrome c or quinol, coupled to the generation of a proton motive force utilized for adenosine triphosphate (ATP) synthesis3,4. Compared to the well-characterized HCOs, cytochromes bd have not been widely studied. These cytochromes are only present in prokaryotes, which include many human pathogens, and thus belong to an evolutionarily distinct oxidase family4.

Cytochrome bd oxidases possess a high affinity for oxygen5,6, which facilitates bacterial survival under O2-poor environments7,8. Apart from this, cytochromes bd also endow bacteria with a number of vitally important physiological functions including enhancing tolerance to nitrosative stress9, contribute to resistance to hydrogen peroxide10, suppress extracellular superoxide production11, and confer the ability to defend against antibacterial agents12. It is likely that these properties of cytochromes bd promote virulence in a number of bacterial pathogens that cause serious infectious diseases to humans, such as Mycobacterium tuberculosis13, Brucella abortus14, as well as Salmonella Typhimurium15,16, Bacteroides7, and Listeria monocytogenes17. Since cytochrome bd is a key enzyme for the survival of prokaryotes and is absent in mammals, it is a promising therapeutic target for the development of antibacterial agents4,18.

To date, two structures of the bd oxidases have been reported. One is from Geobacillus thermodenitrificans (G. th)19 and the other is from Escherichia coli (E. coli)20,21. These enzymes possess two key subunits CydA and CydB but vary in the numbers of additional small subunits that are associated with them. G. th bd oxidase contains an association subunit called CydS and E. coli bd oxidase includes two association subunits named CydX and CydH (or CydY). Common features in these complexes are two b-type hemes (low-spin heme b558 and high-spin heme b595) and one d-type heme which are arranged in a triangular manner but with different relative positions in the two structures. These observations suggest that homologous bd oxidases share a similar architecture but can vary in mechanistic detail21–24. Therefore, more structural information is needed to obtain a comprehensive knowledge of the structure and function of bd oxidases. The properties of several mycobacterial cytochromes bd have been investigated12,18,24,25, but given they share only a low sequence identity (the lowest similarity score is 20%) with the two previously reported cytochrome bd structures it is difficult to rationalize the function of these cytochromes bd18. In the present study, we have determined a 2.8 Å cryo-electron microscopy (cryo-EM) structure of a dimeric bd oxidase from Mycobacterium smegmatis (Msm). In so doing, we have identified two potential oxygen access channels, which could be excellent targets for anti-tuberculosis drug discovery.

Results and discussion

Overall structure of Mycobacterium smegmatis bd oxidase

Msm bd oxidase was recombinantly expressed and purified to homogeneity (Supplementary Fig. 1a–d). The purified Msm bd enzyme is a stable and functional assembly with a turnover number of 21.6 ± 2.8 e− s−1 (Supplementary Fig. 1e). A 2.8 Å resolution structure was determined in lipid nanodiscs using cryo-EM. Details for data collection and model statistics are provided (Supplementary Fig. 2 and Supplementary Table 1). Although the construct design, expression, and purification of Msm bd oxidase were performed according to a procedure described previously for G. th19 and E. coli20,21 bd oxidases, only CydA and CydB were observed. No additional small subunit similar to CydS/X or CydH/Y found in the G. th19 and E. coli20,21 bd oxidases were observed in the Msm complex, noting that in the G. th and E. coli studies, the genes coding for the associated subunits CydS/X of G. th19 bd oxidase and CydH/Y of E. coli20,21 bd oxidase were not included in the corresponding expression plasmid. Nonetheless, these two small subunits were observed to co-elute upon purification in both complexes. Thus, we believe that the Msm bd oxidase contains only two core subunits, CydA and CydB (Fig. 1a, b, and Supplementary Figs. 3, 4). This hypothesis is further supported by a BLAST search of the mycobacterial genomes which did not identify any subunits homologous to CydS/X and CydH/Y (Supplementary Fig. 5). In addition, Msm bd oxidase in the absence of associated subunits is active, whereas the E. coli bd oxidase is not active in the absence of its associated subunits26. In summary, mycobacterial bd oxidase appears to only consist of the two core subunits and no other units. Notably, there is a high structural similarity between the G. th and E. coli counterparts compared to Msm core subunits, except for a little difference in the topology of CydB subunits between Msm and E. coli (Supplementary Fig. 6).

Fig. 1. Overall structure of the Msm bd oxidase.

Fig. 1

Open in a new tab

a The bd oxidase cryo-EM density map at 2.8 Å resolution. MSP, membrane scaffold protein. b Cartoon representation of the bd oxidase consisting of subunits CydA (red) and CydB (cyan). c Structural superposition of CydA and CydB.

In terms of secondary structure, Msm bd oxidase, CydA, and CydB possess a common fold each consisting of nine transmembrane helices (TMHs). These two subunits are related by an approximate two-fold rotational axis of symmetry. The TMH domains can be divided into two four-helix bundles and one additional peripheral helix (Fig. 1c). The domains of the two subunits superpose well with a root mean square deviation of 3.6 Å for all the Ca atoms (Fig. 1c) and thus suggest that CydA and CydB evolved by a gene duplication event. The three heme groups (b558, b595, and d) are clearly visible in the CydA subunit (Fig. 1b and Supplementary Fig. 3). Given the two other bd oxidases have associated subunits in their structures, our structure suggests that their presence is species-dependent.

Q-loop of M. smegmatis bd oxidase

The bd oxidases are divided into the S (short)- and the L (long)-subfamilies4. This is according to the length of the region of polypeptide referred to as the Q-loop (quinol-binding domain). Mycobacterial cytochromes bd belong to the S-subfamily20. The Q-loop of CydA has a water-exposed domain (segment 257–339), connecting the 6 and 7 in the hydrophilic extracellular space19–21. It has two parts, one associated with the N-terminal domain (QN) and the other with the C-terminal domain (QC). The QN-loop plays a functional role in the binding and oxidation of the quinol27,28 and the QC-loop is needed for the assembly/stability of the enzyme22,23,25.

In the present study, the densities corresponding to the Q-loop without and with aurachin D (a quinone analog inhibiting mycobacterial cytochrome bd) bound are not completely resolved (Supplementary Figs. 2 and 7). A comparison of these two structures does not show any conformational changes which may be a function of the resolution of the data or the fact that the region around the aurachin D is inherently disordered. Aurachin D is bound to the QN-loop20, potentially stabilizing the Q-loop, and has the ability to inhibit the activity of Msm cytochrome bd29. The structure of the QN-loop in the E. coli enzyme is also not fully resolved20,21. Therefore, these structural data suggest that the QN-loop is intrinsically flexible. Its flexibility may be required for the rapid binding and release of the quinols. The remaining segments of the Q-loop in the Msm bd oxidase structure are well resolved. At the periplasmic side of TMH 6, there is a short horizontal helix, Qh1, that includes the highly conserved residues Lys260 and Glu265 (Lys252 and Glu257 in G. th and E. coli)19–21, critical for quinol substrate binding and electron transfer28 (Fig. 2a). The structure of Qh1 is conserved with respect to the G. th and E. coli enzymes19–21 (Fig. 2b). It is noteworthy that the QC-loop here adopts a rigid secondary structure with a horizontal Qh2 (residues 317–327), that emerges from the flexible QN-loop part and covers the periplasmic surface of CydA. Site-directed mutagenesis and whole-bacteria assays, performed on the Mycobacterium tuberculosis (Mtb) bd oxidase, demonstrated that the regions corresponding to the Qh2 and the equivalent residues Tyr323, Phe327, and Tyr332 nearby the Qh2 stretch are essential for the function of the oxidase25 (Supplementary Fig. 8). In Msm bd oxidase, we observe that these three residues are involved in the interactions with Qh1 and the periplasmic loop between TMH 8/9, located at the periplasmic surface of heme groups b558 and b595, and as a result potentially affect the cofactor and quinol binding (Fig. 2a). Tyr323 forms the van der Waals interaction with Pro258 in the Qh1 region, and Phe327 and Tyr332 form the stacking interactions with Trp404 in the loop TM8/9 region. So the observed structural features are in agreement with the previous functional characterization of the Mtb enzyme25. Additionally, according to the structural superposition between Msm and E. coli CydA subunits (Supplementary Fig. 6), the corresponding residues of Tyr323, Phe327, and Tyr332 in Msm are His314, Tyr353, Ser379 in E. coli, respectively. His314 and Tyr353 also form van der Waals interactions and a stacking interaction, respectively, with Thr251 in the Qh1 region and Trp451 in the loop TM8/9 region. Hence, these interactions in the Msm bd oxidase are very similar to those observed in the E. coli complex. It is also worth noting that the folding of the QC-loop region is different (Fig. 2b), compared to the other bd oxidases, though the Q-loop domain of Msm bd oxidase should be remembered these belong to the short Q-loop class (Supplementary Fig. 9)20. This array of differences suggests that the Q-loop domain could be used as a marker for evolutionary analysis of bd oxidases in prokaryotes (Supplementary Fig. 10).

Fig. 2. The Q-loop in bd oxidases.

Fig. 2

Open in a new tab

In CydA, a hydrophilic region between the transmembrane helices 6 and 7 harbors the quinol-binding site and has thus been named the Q-loop. a The N-terminal and C-terminal Q loops are rigid and well-ordered helical segments, but the linking region between them is not resolved in the maps. The residues labeled are important for stability or activity. b The bd oxidases from G. th, E. coli, and Msm are superimposed. The Q-loop domains from G. th, E. coli, and Msm are shown in red, blue, and cyan, respectively.

Electron transfer in Msm bd oxidase

The three heme groups (b558, b595, and d) unambiguously identified in Msm bd oxidase are organized in a triangulated arrangement near the periplasmic side of CydA (Fig. 3a). In this structure, the low-spin b558 is within the transmembrane core of subunit CydA, adjacent to the QN-loop segment. Its axial ligands are conserved residues His185 and Met346 (His186 and Met325 in G. th; His186 and Met393 in E. coli)19–21 (Fig. 3b). Heme b595 is located closer to the periplasmic side and is ligated by Glu398 (Glu378 in G. th; Glu445 in E. coli)19–21 (Fig. 3b). There is a conserved Trp394A between heme groups b595 and b558 that may mediate electron transfer19–21,30. The third cofactor heme d, the site of oxygen binding and reduction, is positioned at the center of CydA and the invariant His18 (His21 in G. th; His19 in E. coli) appears to be ligated to the heme iron on one side19–21.

Fig. 3. Cofactor organization in Msm cytochrome bd.

Fig. 3

Open in a new tab

a Triangular arrangement of the heme cofactors in CydA. Heme edge-to-edge distances are indicated by numbers in the parentheses. b Axial amino acid ligands of the heme cofactors. c Heme superposition between E. coli, G. th, and Msm. d The superimposed heme d sites from Msm and E. coli.

Although the bd oxidases of G. th and Msm are from the same S-subfamily, the relative arrangement of the three heme groups are strikingly different19 (Fig. 3c). The organization of the redox center in Msm cytochrome bd is similar to that reported for the L-subfamily as exemplified by E. coli bd oxidase20,21. A structural superimposition (Fig. 3c) shows the two heme groups are located in the same position in the CydA, the distances between the central iron atoms are consistent as the orientations of the heme planes relative to the membrane plane. Therefore, the cofactor location is a conserved feature of the three respiratory bd oxidases. On the opposite side of the heme d, Glu98 (Glu101 in G. th; Glu99 in E. coli) acts as the axial ligand. Glu98 here is 6.2 Å from the central iron atom. The equivalent distances in G. th and E. coli are 2.1 and 6.0 Å, respectively19–21. This voluminous cavity is suggested to be used for the binding of substrates such as oxygen, roofed by the hydrophobic Ile143 and Phe103 (Ile144 and Phe104 in E.coli)20,21. The location and surrounding environment of the cofactors in the bd oxidase are highly conserved between Msm and E. coli (Fig. 3d). Collectively, it is suggested that a sequential electron transfer from heme b558 via heme b595 to heme d also exists in the mycobacterial bd oxidases20,21.

Two oxygen access channels in Msm bd oxidase

Cytochromes bd play a role in energy metabolism with high O2 affinity under hypoxic conditions4,12, conditions often encountered by the microorganisms in their natural habitats. The dioxygen has to bind to heme d and then is reduced at this position19–21. In E. coli bd oxidase, the O2-channel acts as a pathway for direct oxygen diffusion from the membrane interior to the heme d reaction site. It is formed by a small direct hydrophobic channel, which starts above Trp63 (Trp67 in Msm) at the membrane interface between TMH1 and TMH9 of CydB and extends further to heme d on CydA20,21. The corresponding residue Trp has been demonstrated to be essential for bd activity in Mtb25. Noteworthy, in the Msm structure, the channel is also formed by a conserved structural topology and residues according to a comparison between the bd oxidases from E. coli and Msm (Supplementary Fig. 11), which suggests that the oxygen here may also access the active site through this conserved O2-channel (identified as channel 1) (Fig. 4). Intriguingly, there is an additional accessible channel directly connecting to the protein surface and extending to the heme b595, which is also identified in the G. th enzyme19. This channel has also been proposed for the oxygen entry site in G. the bd oxidase19,20, which is blocked by the single-transmembrane subunit CydH in the E. coli enzyme20,21. In addition, the heme d in this structure is buried deeper inside the subunit CydA and the penetration of dioxygen from this cavity into heme d is also blocked by heme d itself21. However, the channel is accessible in the Msm enzyme and previous studies have reported that the high-spin heme b595 could be the second reaction site for O25,31. Therefore, this channel (channel 2) is very likely to be an alternative pathway to guide dioxygen to heme b595, which may further sustain energy metabolism in the bacterial cell and enhance mycobacterial survivability in the host. It has been reported that the threshold pO2 (O2 tension) of the growth medium for the induction of cyd gene cluster in M. smegmatis (ca. 1% air saturation)32 is significantly lower than that of E. coli (10% air saturation)33, suggesting that the Msm bd oxidase has a different functional or kinetic range with respect to oxygen availability compared to that of E. coli32. Overall, therefore, the two oxygen channels in Msm bd oxidase are reasonably proposed based on these structural features and the previously described studies. Future investigations are needed to determine whether these two catalytic reactions take place at the same time.

Fig. 4. Proton/gaseous substrate channels in the Msm bd oxidase.

Fig. 4

Open in a new tab

Subunits CydA/B are shown in cartoon representation. The putative dioxygen and proton channels are labeled. The heme groups are shown as stick models. The proton transfer step from heme b595 to heme d is identified by a black dashed path.

Although the bd oxidases do not pump protons from the cytoplasmic side to periplasmic side, producing the proton motive force across the membrane, the pathway for proton uptake from the cytoplasm is crucial to reduce dioxygen to water4. Two proton pathways in subunits CydA and CydB from the cytoplasmic side to the active site have been proposed in G. th19 and E. coli enzymes20,21. These studies indicate two hydrophilic channels for proton transfer. Based on the superimposition and structural analysis between the bd oxidases from G. th and E. coli19,20, the relatively conserved hydrophilic residues in our model, along the canonical CydA and CydB pathways, are His125.A, Gln36.A, Glu106.A, Ser107.A, Ser139.A to Glu98.A, and Asp25.B, Asp62.B, and Asn64.B (Fig. 4, Supplementary Fig. 11). Given the conserved identity of the proton pathways in mycobacterial enzymes, they are also likely to facilitate proton transfer for dioxygen reduction at the heme d site. In addition, in terms of the dioxygen reduction at the heme b595 site, there must be an additional proton transfer step from heme d to heme b595 in order to deliver protons to the oxygen reduction site (Fig. 4), which may be potentially similar to that of the G. th enzyme19. According to the current structure, the heme propionate of heme d is in a hydrophobic environment without any charge compensation. It is thus very likely protonated and supplies protons for unresolved water molecules here that connect heme b595 to heme d. These protons would be replenished via the CydA/B proton pathways.

Overall, the electron released from the quinol bound at the quinone-binding site is transferred, in turn, to the prosthetic groups heme b585, heme b595 to heme d. At the same time, the oxygen molecule that is diffused to the heme b595 and/or heme d sites is reduced to water, a process that is involved in conducting protons to the oxygen-binding site through the CydA/B pathways (Fig. 5).

Fig. 5. A schematic diagram showing the electron/proton transfer pathway in Msm bd oxidase and the relevant oxygen entry pathway.

Fig. 5

Open in a new tab

The heme groups are shown as in stick models. The putative dioxygen and proton channels are labeled. Electron transfer directions are shown in black arrows.

Cytochromes bd is ubiquitous among prokaryotes (but not present in eukaryotes) and is now attracting attention as promising targets for next-generation antibacterials. Here, we have determined the 2.8 Å cryo-EM structure of Msm bd oxidase. The overall fold is similar to the two other previously reported bd oxidases but exhibits several different features, including the fold of the Q-loop and the number of associated subunits. In addition, we have identified two potential oxygen access channels that look to be also present in G. th and E. coli cytochromes bd. The quinol-binding site located in the Q-loop has been proposed to be a target for drug discovery. However, the structure of the Q-loop has not been fully determined, thus posing a challenge for the design of quinol-type inhibitors. The two oxygen-conducting O2-channels could be alternative targets for the discovery of anti-tuberculosis drugs.

Methods

Bacteria strain and culture

The cydAB gene was cloned into pMV261 plasmid with a 10x His tag at the C-terminus of cydB. The primers are listed in Supplementary Table 2. Expression was achieved by electroporation of the plasmid into strain Msm mc2 5134. A volume of 1 mL strain stock was added to 24 mL LB broth (0.1% Tw80, 50 μg/mL kanamycin, and 20 μg/mL carbenicillin) and cultured overnight at 37 °C and 220 rpm. Next, 4 mL pre-culture aliquots were transferred to 1 L LB broth (rubber plug, 0.1% Tw80, 50 μg/mL kanamycin, and 20 μg/mL carbenicillin) and cultured at 37 °C and 220 rpm. When the OD600 reached 0.8, 5 mL of 40% acetamide was added to induce the expression of the target protein over 3 days at 25 °C and 220 rpm.

Protein purification and characterization

The purification procedure followed a previous study but with a few modifications30. Membranes of the cells were extracted in buffer (20 mM HEPES, pH 7.4, 100 mM NaCl), and then stirred slowly at 4 °C for 2 h with 1% (w/v) dodecyl-beta-d-maltoside (DDM). The supernatant after centrifugation was loaded onto a Ni-NTA column and the eluted fraction including the protein of interest was loaded onto a Superdex 200 (GE Healthcare) column equilibrated in a buffer containing 20 mM HEPES, pH 7.4, 100 mM NaCl, and 0.02% (w/v) DDM. The peak fractions were analyzed by SDS–PAGE (sodium dodecyl sulfate–polyacrylamide gel electrophoresis), then pooled and concentrated to 6 mg/mL.

Preparation of reduced quinol substrate

2,3-Dimethyl-1,4-naphthoquinone (DMNQ, CAS 2197-57-1) was synthesized by WuXi AppTec. DMNQ reduction was performed following previously published protocols with some modifications35. To prepare the reduced quinol, DMNQH2, 20 mM DMNQ was ultrasonically dissolved in 1 mL ethanol with 6 mM HCl. A few grains of sodium borohydride (NaBH4) were then added to obtain a fully reduced, colorless solution in the ice-bath. An appropriate amount of HCl was used to quench the mixture under the protection of argon. The quinol solution was stored at −80 °C.

Oxygen consumption assay

Oxidase activity was determined according to the previous studies20,36,37. Oxygen consumption was monitored with a Clark-type oxygen electrode (Hansatech Chlorolab 2) in the buffer 20 mM HEPES, pH 7.4, 100 mM NaCl, 0.04% DDM, and 10 mM DTT at room temperature. To begin the assay, 480 μL buffer was first added until the oxygen equilibrium. 20 μL DMNQH2 was then added and the substrate autoxidation rate was recorded. The reaction was started by the addition of 0.8 μM bd complex. The time course for oxygen consumption was curved with GraphPad prime 6.0 software, from which an estimate of the observed pseudo-first-order rate constant (kobs) is obtained (corrected for autoxidation). This assay included four groups of parallel experiments.

Reconstruction of cytochrome bd into nanodiscs

MSP1D1 was used to reconstruct the nanodisc and the purification and reconstruction followed the reported study38. Briefly, cytochrome bd, MSPD1, and POPC were mixed with a stoichiometry of 1:4:160 and incubated at 4 °C for one hour. Next, 200 μL of resuspended Bio-Beads (0.5 g/mL) were added twice with an interval of 30 min to remove the detergents. After 12 h of incubation, the supernatant was applied to a Superdex 200 (GE Healthcare) column equilibrated in 20 mM HEPES, pH 7.4, 100 mM NaCl buffer. The peak fraction was collected and concentrated.

Cryo-EM sample preparation and data collection

Aliquots (4 μL) of reconstructed nanodisc-cytochrome bd at a concentration of 1 mg/mL were applied to glow-discharged Quantifol Cu 1.2/1.3 (mesh 300) grids. For cytochrome bd and aurachin D complex, 0.35 mM aurachin D was added and incubated with nanodisc-cytochrome bd for half an hour before sample vitrification. Glow discharge was accomplished by adding an H2 and O2 mixture in the Gatan Solarus 950 for 25 s. After blotting for 3 s with a blot force of −2, grids were flash-frozen in liquid ethane cooled by liquid nitrogen using an FEI Vitrobot operated at 8 °C and 100% humidity. For cytochome bd complex without aurachin D, data collection was achieved using the Titan Krios electron microscopy operated at 300 kV with a Gatan K3 detector at a magnification of SA ×29,000. Images were recorded in super-resolution mode binned to a pixel size of 0.82 Å/pixel. Data acquisition was achieved by using serialEM39. Images were collected with 40 frames and a total dose of 60 e−/Å2. The defocus range was set to 1.2–1.8 μm. For cytochrome bd–aurachin D complex, images were recorded using an FEI Titan Krios electron microscopy operating at 300 kV with a Gatan K2 detector at a magnification of ×165,000 with an energy filter, corresponding to a pixel size of 0.82 Å/pixel. Images were collected with 40 frames and a total dose of 60 e−/Å2 with a defocus range between 1.2 and 1.8 μm.

Image processing

Dose-fractioned images were motion-corrected and dose-weighted by MotionCor2 software40. CTF estimation was performed by cryoSPARC41. 1,578,284 particles were automatically picked and extracted with a box size of 256 pixels41. 2D classification and 3D classification and refinement were all performed in cryoSPARC41. 50,000 particles were used to generate three classes in ab-initio reconstruction. The classes were used as templates for heterogeneous refinement with all selected particles. After a few rounds of heterorefinement, 270,938 particles converged into one class with a 3.4 Å initial map. These particles were used to perform a homogeneous refinement and local refinement to obtain a final resolution of 2.79 Å.

Model building and refinement

The final map was sharpened automatically using a B-factor of 121.3 Å2 in cryoSPARC. The atomic model was manually built in Coot42 (version 0.8.9.1) using the crystal structure of G.th cytochrome bd (PDB: 5doq)19 as a template. Real-space refinement and validation of the final model were performed in Phenix (version 1.14)43. The local resolution map was calculated with ResMap44. All reported resolutions were based on the gold-standard FSC 0.143 criteria45. FSCwork and FSCtest were conducted to check for over fitting46.

All figures were created using UCSF Chimera47 or PyMOL48.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Supplementary information

Supplementary Information (8.2MB, pdf)
Reporting summary (197.4KB, pdf)

Acknowledgements

We thank Dr. Chao Peng of the Mass Spectrometry System at the National Facility for Protein Science in Shanghai (NFPS), Zhangjiang Lab, SARI, China for data collection and analysis and Prof. Tongbu Lu and Yali Bai (School of Materials Science and Engineering, Tianjin University of Technology) for their technical support on Clark-type oxygen electrode and oxygen consumption assay. We would like to thank the Bio-Electron Microscopy Facility of ShanghaiTech University, and we are grateful to Dr. Qianqian Sun for her help with cryo-EM technical support. This work was supported by Grants from the National Key Research and Development Program of China (Grant No. 2017YFC0840300), the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDB37030201, XDB37020203), and the National Natural Science Foundation of China (Grant Nos. 81520108019, 813300237).

Source data

Source Data (251KB, xlsx)

Author contributions

Z.R. and H.G. conceived and supervised the study. W.W. purified the M. smegmatis bd complex. Y.T. and Y.L. performed activity identification of the purified M. smegmatis bd complex. W.W. and Y.G. collected and processed cryo-EM data and built the structure model; H.G., W.W., Y.G., Q.W., X.Z., Y.T., S.Z., Y.Z., Y.L., X.Y., F.L., L.W.G., and Z.R. analyzed the structure and discussed the results and the manuscript was written by H.G., W.W., L.W.G. and Z.R.

Data availability

The accession numbers for the 3D cryo-EM density map of Msm bd oxidase without and with bound AD in present study are EMD-30582 and EMD-31302, respectively. The accession number for the coordinates for the Msm bd oxidase without bound AD in this study is PDB: 7D5I. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Footnotes

Peer review information Nature Communications thanks Tim Rasmussen and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Weiwei Wang, Yan Gao.

Contributor Information

Quan Wang, Email: wangq@shanghaitech.edu.cn.

Zihe Rao, Email: raozh@tsinghua.edu.cn.

Hongri Gong, Email: gonghr@nankai.edu.cn.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-021-24924-w.

References

  • 1.Michel, H. Terminal oxidases of the heme-copper and bd oxidase type, a structural and functional comparison. Biochim. Biophys. Acta1859, e3 (2018).
  • 2.Borisov VB, Siletsky SA. Features of organization and mechanism of catalysis of two families of terminal oxidases: heme-copper and bd-type. Biochemistry. 2019;84:1390–1402. doi: 10.1134/S0006297919110130. [DOI] [PubMed] [Google Scholar]
  • 3.Mitchell P. Coupling of phosphorylation to electron and hydrogen transfer by a chemiosmotic type of mechanism. Nature. 1961;191:144–148. doi: 10.1038/191144a0. [DOI] [PubMed] [Google Scholar]
  • 4.Borisov VB, et al. The cytochrome bd respiratory oxygen reductases. Biochim. Biophys. Acta. 2011;1807:1398–1413. doi: 10.1016/j.bbabio.2011.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.D’Mello R, Hill S, Poole RK. The cytochrome bd quinol oxidase in Escherichia coli has an extremely high oxygen affinity and two oxygen-binding haems: implications for regulation of activity in vivo by oxygen inhibition. Microbiology. 1996;142:755–763. doi: 10.1099/00221287-142-4-755. [DOI] [PubMed] [Google Scholar]
  • 6.Belevich I, et al. Cytochrome bd from Azotobacter vinelandii: evidence for high-affinity oxygen binding. Biochemistry. 2007;46:11177–11184. doi: 10.1021/bi700862u. [DOI] [PubMed] [Google Scholar]
  • 7.Baughn AD, Malamy MH. The strict anaerobe Bacteroides fragilis grows in and benefits from nanomolar concentrations of oxygen. Nature. 2004;427:441–444. doi: 10.1038/nature02285. [DOI] [PubMed] [Google Scholar]
  • 8.Shi L, et al. Changes in energy metabolism of Mycobacterium tuberculosis in mouse lung and under in vitro conditions affecting aerobic respiration. Proc. Natl Acad. Sci. USA. 2005;102:15629–15634. doi: 10.1073/pnas.0507850102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Mason MG, et al. Cytochrome bd confers nitric oxide resistance to Escherichia coli. Nat. Chem. Biol. 2009;5:94–96. doi: 10.1038/nchembio.135. [DOI] [PubMed] [Google Scholar]
  • 10.Forte E, et al. Cytochrome bd oxidase and hydrogen peroxide resistance in Mycobacterium tuberculosis. mBio. 2013;4:e01006–e01013. doi: 10.1128/mBio.01006-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Huycke MM, et al. Extracellular superoxide production by Enterococcus faecalis requires demethylmenaquinone and is attenuated by functional terminal quinol oxidases. Mol. Microbiol. 2001;42:729–740. doi: 10.1046/j.1365-2958.2001.02638.x. [DOI] [PubMed] [Google Scholar]
  • 12.Mascolo L, Bald D. Cytochrome bd in Mycobacterium tuberculosis: a respiratory chain protein involved in the defense against antibacterials. Prog. Biophys. Mol. Biol. 2020;152:55–63. doi: 10.1016/j.pbiomolbio.2019.11.002. [DOI] [PubMed] [Google Scholar]
  • 13.Kalia NP, et al. Exploiting the synthetic lethality between terminal respiratory oxidases to kill Mycobacterium tuberculosis and clear host infection. Proc. Natl Acad. Sci. USA. 2017;114:7426–7431. doi: 10.1073/pnas.1706139114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Endley S, McMurray D, Ficht TA. Interruption of the cydB locus in Brucella abortus attenuates intracellular survival and virulence in the mouse model of infection. J. Bacteriol. 2001;183:2454–2462. doi: 10.1128/JB.183.8.2454-2462.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Duc, K. M. et al. The small protein CydX is required for Cytochrome bd quinol oxidase stability and function in Salmonella enterica Serovar Typhimurium: a Phenotypic Study. J. Bacteriol.202, e00348-19 (2020). [DOI] [PMC free article] [PubMed]
  • 16.Wilson RP, et al. STAT2 dependent Type I Interferon response promotes dysbiosis and luminal expansion of the enteric pathogen Salmonella Typhimurium. PLoS Pathog. 2019;15:e1007745. doi: 10.1371/journal.ppat.1007745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Corbett, D. et al. Listeria monocytogenes has both Cytochrome bd-type and Cytochrome aa (3)-type terminal oxidases, which allow growth at different oxygen levels, and both are important in infection. Infect. Immun.85, e00354-17 (2017). [DOI] [PMC free article] [PubMed]
  • 18.Lee BS, Sviriaeva E, Pethe K. Targeting the cytochrome oxidases for drug development in mycobacteria. Prog. Biophys. Mol. Biol. 2020;152:45–54. doi: 10.1016/j.pbiomolbio.2020.02.001. [DOI] [PubMed] [Google Scholar]
  • 19.Safarian S, et al. Structure of a bd oxidase indicates similar mechanisms for membrane-integrated oxygen reductases. Science. 2016;352:583–586. doi: 10.1126/science.aaf2477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Safarian S, et al. Active site rearrangement and structural divergence in prokaryotic respiratory oxidases. Science. 2019;366:100–104. doi: 10.1126/science.aay0967. [DOI] [PubMed] [Google Scholar]
  • 21.Theßeling A, et al. Homologous bd oxidases share the same architecture but differ in mechanism. Nat. Commun. 2019;10:5138. doi: 10.1038/s41467-019-13122-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Goojani HG, et al. The carboxy-terminal insert in the Q-loop is needed for functionality of Escherichia coli cytochrome bd-I. Biochim. Biophys. Acta Bioenerg. 2020;1861:148175. doi: 10.1016/j.bbabio.2020.148175. [DOI] [PubMed] [Google Scholar]
  • 23.Theßeling A, et al. The long Q-loop of Escherichia coli cytochrome bd oxidase is required for assembly and structural integrity. FEBS Lett. 2020;594:1577–1585. doi: 10.1002/1873-3468.13749. [DOI] [PubMed] [Google Scholar]
  • 24.Harikishore, A. et al. Targeting the menaquinol binding loop of mycobacterial cytochrome bd oxidase. Mol. Divers.25, 517–524 (2021). [DOI] [PubMed]
  • 25.Sviriaeva E, et al. Features and functional importance of key residues of the Mycobacterium tuberculosis Cytochrome bd oxidase. ACS Infect. Dis. 2020;6:1697–1707. doi: 10.1021/acsinfecdis.9b00449. [DOI] [PubMed] [Google Scholar]
  • 26.Hoeser J, et al. Subunit CydX of Escherichia coli cytochrome bd ubiquinol oxidase is essential for assembly and stability of the di-heme active site. FEBS Lett. 2014;588:1537–1541. doi: 10.1016/j.febslet.2014.03.036. [DOI] [PubMed] [Google Scholar]
  • 27.Matsumoto Y, et al. Mass spectrometric analysis of the ubiquinol-binding site in cytochrome bd from Escherichia coli. J. Biol. Chem. 2006;281:1905–1912. doi: 10.1074/jbc.M508206200. [DOI] [PubMed] [Google Scholar]
  • 28.Mogi T, et al. Probing the ubiquinol-binding site in cytochrome bd by site-directed mutagenesis. Biochemistry. 2006;45:7924–7930. doi: 10.1021/bi060192w. [DOI] [PubMed] [Google Scholar]
  • 29.Lu P, et al. The anti-mycobacterial activity of the cytochrome bcc inhibitor Q203 can be enhanced by small-molecule inhibition of cytochrome bd. Sci. Rep. 2018;8:2625. doi: 10.1038/s41598-018-20989-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Gong H, et al. An electron transfer path connects subunits of a mycobacterial respiratory supercomplex. Science. 2018;362:eaat8923. doi: 10.1126/science.aat8923. [DOI] [PubMed] [Google Scholar]
  • 31.Rothery RA, Houston AM, Ingledew WJ. The respiratory chain of anaerobically grown Escherichia coli: reactions with nitrite and oxygen. J. Gen. Microbiol. 1987;133:3247–3255. doi: 10.1099/00221287-133-11-3247. [DOI] [PubMed] [Google Scholar]
  • 32.Kana BD, et al. Characterization of the cydAB-encoded cytochrome bd oxidase from Mycobacterium smegmatis. J. Bacteriol. 2001;183:7076–7086. doi: 10.1128/JB.183.24.7076-7086.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Becker S, et al. Regulatory O2 tensions for the synthesis of fermentation products in Escherichia coli and relation to aerobic respiration. Arch. Microbiol. 1997;168:290–296. doi: 10.1007/s002030050501. [DOI] [PubMed] [Google Scholar]
  • 34.Li, X. et al. Draft Genome sequence of mc2 51, a highly hydrogen peroxide-resistant Mycobacterium smegmatis mutant strain. Genome Announc.2, e00092-14 (2014). [DOI] [PMC free article] [PubMed]
  • 35.Bennett MC, et al. Chronic in vivo sodium azide infusion induces selective and stable inhibition of cytochrome c oxidase. J. Neurochem. 1996;66:2606–2611. doi: 10.1046/j.1471-4159.1996.66062606.x. [DOI] [PubMed] [Google Scholar]
  • 36.Bonner WD, Clarke SD, Rich PR. Partial purification and characterization of the quinol oxidase activity of Arum maculatum mitochondria. Plant Physiol. 1986;80:838–842. doi: 10.1104/pp.80.4.838. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Kusumoto K, et al. Menaquinol oxidase activity and primary structure of cytochrome bd from the amino-acid fermenting bacterium Corynebacterium glutamicum. Arch. Microbiol. 2000;173:390–397. doi: 10.1007/s002030000161. [DOI] [PubMed] [Google Scholar]
  • 38.Ritchie TK, et al. Chapter 11—reconstitution of membrane proteins in phospholipid bilayer nanodiscs. Methods Enzymol. 2009;464:211–231. doi: 10.1016/S0076-6879(09)64011-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Mastronarde DN. Automated electron microscope tomography using robust prediction of specimen movements. J. Struct. Biol. 2005;152:36–51. doi: 10.1016/j.jsb.2005.07.007. [DOI] [PubMed] [Google Scholar]
  • 40.Zheng SQ, et al. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat. Methods. 2017;14:331–332. doi: 10.1038/nmeth.4193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Punjani A, et al. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods. 2017;14:290–296. doi: 10.1038/nmeth.4169. [DOI] [PubMed] [Google Scholar]
  • 42.Emsley P, et al. Features and development of Coot. Acta Crystallogr. D. 2010;66:486–501. doi: 10.1107/S0907444910007493. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Adams PD, et al. Advances, interactions, and future developments in the CNS, Phenix, and Rosetta structural biology software systems. Annu. Rev. Biophys. 2013;42:265–287. doi: 10.1146/annurev-biophys-083012-130253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Kucukelbir A, Sigworth FJ, Tagare HD. Quantifying the local resolution of cryo-EM density maps. Nat. Methods. 2014;11:63–65. doi: 10.1038/nmeth.2727. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Rosenthal PB, Henderson R. Optimal determination of particle orientation, absolute hand, and contrast loss in single-particle electron cryomicroscopy. J. Mol. Biol. 2003;333:721–745. doi: 10.1016/j.jmb.2003.07.013. [DOI] [PubMed] [Google Scholar]
  • 46.Brown A, et al. Tools for macromolecular model building and refinement into electron cryo-microscopy reconstructions. Acta Crystallogr. D. 2015;71:136–153. doi: 10.1107/S1399004714021683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Pettersen EF, et al. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. 2004;25:1605–1612. doi: 10.1002/jcc.20084. [DOI] [PubMed] [Google Scholar]
  • 48.The PyMOL molecular graphics system, version 1.3r1. (Schrödinger, LLC. 2010).

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information (8.2MB, pdf)
Reporting summary (197.4KB, pdf)

Data Availability Statement

The accession numbers for the 3D cryo-EM density map of Msm bd oxidase without and with bound AD in present study are EMD-30582 and EMD-31302, respectively. The accession number for the coordinates for the Msm bd oxidase without bound AD in this study is PDB: 7D5I. Source data are provided with this paper.

Articles from Nature Communications are provided here courtesy of Nature Publishing Group

RESOURCES