STRUCTURAL BIOLOGY

Structure of a bd oxidase indicatessimilar mechanisms for membrane-integrated oxygen reductasesSchara Safarian,1 Chitra Rajendran,1* Hannelore Müller,1 Julia Preu,1

Julian D. Langer,1† Sergey Ovchinnikov,2 Taichiro Hirose,3 Tomoichirou Kusumoto,3

Junshi Sakamoto,3 Hartmut Michel1‡

The cytochrome bd oxidases are terminal oxidases that are present in bacteriaand archaea. They reduce molecular oxygen (dioxygen) to water, avoiding theproduction of reactive oxygen species. In addition to their contribution to the protonmotive force, they mediate viability under oxygen-related stress conditions andconfer tolerance to nitric oxide, thus contributing to the virulence of pathogenicbacteria. Here we present the atomic structure of the bd oxidase from Geobacillusthermodenitrificans, revealing a pseudosymmetrical subunit fold. The arrangementand order of the heme cofactors support the conclusions from spectroscopicmeasurements that the cleavage of the dioxygen bond may be mechanistically similarto that in the heme-copper–containing oxidases, even though the structures arecompletely different.

Living in a reducing atmosphere more than3 billion years ago, the ancestors of the cy-anobacteria found away to extract electronsfrom water via oxygenic photosynthesis.The released oxygen can form reactive ox-

ygen species, and partly reduced oxygen is highlytoxic. To cope with this threat, microorganismshad to survive in anaerobic niches and removeoxygen by reducing it to water. Several enzymaticmachineries are employed for the reduction ofoxygen to water. Oxygen-tolerant methanogenicarchaea use a soluble di-iron flavoprotein (1),whereas most other organisms use membrane-integrated enzymes, primarily members of theheme-copper–containing oxidase (HCO) super-family. DifferentHCOsuse cytochromes or quinolsas electron donors providing the electrons fromthe extracellular side,whereas theprotons requiredfor water formation are delivered from the intra-cellular side. This vectorial delivery of oppositelycharged substrates also generates the electro-chemical proton gradient used for energy-requiringprocesses such as adenosine triphosphate (ATP)synthesis.The cytochrome bd oxidases form the other

known family of membrane-spanning oxygenreductases of bacteria and archaea, unrelated

to HCOs or the membrane-attached cyanide-resistant alternative oxidases (AOXs). The bdoxidases use quinols as electron donors fromthe extracellular side and protons from the in-tracellular side like HCOs, but do not pump pro-tons (2). They possess a high oxygen affinity andthus play a role in protection against oxidativestress (3, 4) and the colonization of O2-poor en-vironments by pathogenic bacteria (5, 6). The bdoxidases also rapidly dissociate gaseous inhibi-tory nitrogen ligands and confer tolerance undernitric oxide stress conditions (7). This ability iscrucial for the viability of pathogenic bacteriaupon host infection (5, 8). The hypersensitivityof a bd oxidase deletion strain ofMycobacteriumtuberculosis to bedaquiline, a tuberculosis drugthat inhibits the F1Fo-ATP synthase (9, 10), high-lights the potential of bd oxidases as drug targets.Crystal structures of representative members

of all threeHCO families (A, B, and C) are known,yet the structure of bd oxidases has remainedelusive. The bd oxidases contain two individualprotein subunits of ~45 kDa (CydA) and ~35 kDa(CydB) (11–13); those of the proteobacteria con-tain a third subunit consisting of a single trans-membrane helix of ~4 kDa, which plays a crucialrole in the activity of the enzyme (14, 15). The bdoxidases contain a low-spin hexacoordinated hemeB (b558) and a high-spin pentacoordinated hemeB (b595) and a chlorin-type heme D (2, 16, 17) aselectron-transferring prosthetic groups. The quinolto be reduced is initially bound via a conservedwater-exposed loop (a Q loop) located in the hy-drophilic extracellular space connecting trans-membrane helices 6 and 7 of CydA. Subsequently,the quinol is oxidized and heme b558 acts as aprimary electron acceptor, whereas two protonsare released to the extracellular side. The elec-trons are transferred to the b595/d high-spin activesite, where oxygen binding and the reduction of

oxygen to water take place (2, 18). No Cu cofactoris present in the active site of the bd oxidases.Here we present the crystal structure of the

bd oxidase from G. thermodenitrificans K1041,using crystals diffracting anisotropically to 3.1to 4 Å resolutions (19). The prominent struc-tural feature of the bd oxidase is the presenceof 19 transmembrane helices in a nearly oval ar-rangement when viewed down to the membrane(Fig. 1A). CydA and CydB possess the same fold,sharing a nine-transmembrane helix topology,each composed of two four-helix bundles andan additional peripheral helix. CydA and CydBare related by an approximate twofold rotationaxis perpendicular to the membrane. A struc-tural alignment of both subunits yields a rela-tively low root mean square deviation value of3.1 Å (Fig. 1B and fig. S2). Such a pseudosym-metrical structural arrangement is most likelythe result of a gene duplication of a single an-cestral gene that encoded a homodimeric proteinand subsequent mutations. The situation resem-bles that of the photosynthetic reaction centers,where the core is constituted by two structurallysimilar protein subunits, only one of them beingused for electron transfer (20, 21). The hetero-dimer formation is mediated by interactions ofhydrophobic residues at the interface betweena3, a4, and a9 of CydA and the symmetry-relatedhelices a3, a4, and a9 of CydB. The Q loop, a hy-drophilic region between transmembrane helices6 and 7 of CydA (segments 249 to 319) facing theextracellular space, possesses a largely irregularprotein fold apart from a short helical stretchleading to an antiparallel b sheet (Fig. 1A). Apotentially accessible pocket for the binding ofthe substrate quinol was identified near Glu257

and Lys252 (Glu257 and Lys252 in Escherichia coli)in the Q loop (Fig. 2C); both residues, which arefound in the short helix (segments 249 to 258),have already been associated with the bindingand oxidation of quinol by site-directed muta-genesis experiments (22). Lys252 is in reasonabledistance to form a polar contact with the O1/2Apropionate group of heme b558.The HCOs generally exhibit a more compact

and globular packing within the membranethan the bd oxidase. The conserved catalyticcore subunits (subunit 1 or CcoN) of the HCO-family enzymes generally contain 12 transmem-brane helices with a threefold rotational symmetricarrangement, resulting in the generation of threepores that are formed by semicircular arrange-ments of four helices each (23). Neither theoverall topology of CydA with its nine transmem-brane helices nor the heme organization indi-cates any evolutionary relation of any functionalmotif. Comparison with other structures in theProtein Data Bank (PDB) (24) reveals the uni-queness of the overall protein structure and theredox active site organization. The four-helixbundle motif as an individual structural unit isfound in a few other bioenergetically importantmembrane proteins, such as cytochrome b651(PDB accession number 4O7G) and the trans-membrane part of a polysulfide reductase (PsrC)(PDB accession number 2VPX).

SCIENCE sciencemag.org 29 APRIL 2016 • VOL 352 ISSUE 6285 583

1Department of Molecular Membrane Biology, Max PlanckInstitute of Biophysics, Max-von-Laue-Straße 3, D-60438Frankfurt/Main, Germany. 2Department of Biochemistry,University of Washington, Seattle, WA, USA. 3Departmentof Bioscience and Bioinformatics, Kyushu Institute ofTechnology, Kawazu 680-4, Iizuka, Fukuoka-ken820-8502, Japan.*Present address: Faculty of Biology and PreclinicalMedicine, University of Regensburg, Universitätsstrasse 31,D-93051 Regensburg, Germany. †Present address:Department of Molecular Membrane Biology, Max PlanckInstitute for Brain Research, Max-von-Laue-Straße 4,D-60438 Frankfurt/Main, Germany. ‡Corresponding author.Email: hartmut.michel@biophys.mpg.de

RESEARCH | REPORTSon M

arch 27, 2021

http://science.sciencemag.org/

Dow

nloaded from

A third subunit, which we designate CydS, con-sisting of only one single transmembrane helix,was identified in the electron density map. CydS,a short peptide of 33 amino acid residues, couldbe purified and sequenced by matrix-assistedlaser desorption/ionization mass spectrometry(MALDI-MS) (fig. S3). We located its gene im-mediately downstream of the cbdB gene in theGeobacillus genome. Positioned at the peripheralinterface of a5-6 of CydA, CydS may stabilizethe heme b558 harboring a four-helix bundle(a5-8 of CydA) during potential structural re-arrangements of the Q loop upon binding andoxidation of quinol (Fig. 1A). Although conservedin the order of Bacillales (fig. S5), there is no se-quence similarity between CydS and the proteo-bacterial single transmembrane helix subunits.However, because a recent de novo blind struc-ture prediction for the E. coli bd oxidase proposesan almost identical location of CydX, CydS andCydX might play analogous roles (25).The three hemes are found in a triangular ar-

rangement (Fig. 2A) and not in the form of alinear chain in the order b558, b595, and then das expected (2). The low-spin heme b558 is lo-cated within themembrane core, with its planetilted by ~110° to the membrane plane. The ver-tical distance from the heme iron to the extra-cellular surface is ~18 Å. It is in close proximityto the Q loop (Fig. 2, A and C). As inferred frommutagenesis experiments and electron para-magnetic resonancedata (16,26),His186 andMet325

(His186 and Met393 in E. coli) are the axial ligandsof heme b558 (Fig. 2C). The terminal electron ac-ceptor heme d is located closer to the extracellularmembrane surface, with its plane tilted by ~60°to the membrane plane. The vertical distancefrom the heme iron to the P side of the mem-brane is ~13.5 Å. Heme b595 is coordinated by theinvariant axial ligand His21 (His19 in E. coli) onone side. On the opposite side, Glu101 (Glu99 inE. coli) appears to be ligated to the heme iron.Originally proposed to be ligated by Glu101 (27),heme d seems to possess another glutamateligand, namely Glu378 (Glu445 in E. coli), whichhas been postulated to be a key residue takingup a charge-compensating proton, when the high-spin hemes become reduced (Fig. 2D) (28). OurMALDI-MS analyses indicate that heme d ispresent as a cis–heme d hydroxychlorin g–spirolactone in the Geobacillus bd oxidase (fig.S4). This result is supported by the electron den-sity map, which moreover implies a bent con-formation of heme d.Despite their close proximity, the arrangement

of the two high-spin hemes does not supportthe existence of a binuclear active center (seethe supplementary materials). The hemes arein van der Waals contacts, with edge-to-edgedistances around 3.5 Å and a distance betweenthe central Fe atoms of 11.6 Å. The highly con-served (>99%) Trp374 residue may be impor-tant for the electron transfer between hemeb558 and heme d. Conserved Trp residues areoften found between intermediate electron ac-ceptors in biological electron transfer chains,such as in the active branch of the photosyn-

thetic reaction centers from purple bacteria andof photosystem II between a (bacterio)phaeophytinand a quinone acceptor (20), or between the high-spin heme and the low-spin heme of the HCOs.In addition, Trp residues themselves can donateelectrons and form Trp radicals. In particular,Trp374 could donate an electron to compound I(F+* state)—a catalytic intermediate of heme dforming an oxoferryl-porphyrin radical duringthe splitting of the dioxygen bond by a singlefour-electron transfer reaction (29)—and thenreceive an electron from heme b558.The triangular arrangement of the hemes

and the distances between them suggest a directelectron transfer from heme b558 to heme d, be-cause the edge-to-edge distance and the Fe-Fedistance between these hemes are significantlyshorter (5.9 and 15.2 Å) than the respective dis-tances between the two b hemes (8.5 and 19.4 Å).Finally, the electron would equilibrate betweenheme d and heme b595. For the breakage of thedioxygen bond, four electrons are required. Inthe reduced enzyme, two of them could be sup-plied by the heme d iron by the formation of anoxoferryl state, the third one fromheme b595, and[according to (29)] the fourth one from the mac-rocycle of heme d. Therefore, the bd oxidases andthe HCOs appear to use the same principle toavoid the formation of superoxide anions and ofperoxides (29), namely a very rapid or simulta-

neous transfer of four electrons onto dioxygen,leading to breakage of the dioxygen bond. Themain functional difference appears to be thatone electron is provided by an amino acid sidechain (tyrosine) in the case of the HCOs and bythe heme d macrocycle in the case of the bd oxi-dases (29). The analogy goes further: The oxygen-binding heme (d or a3 for the canonical hemeaa3 HCOs) is located between a low-spin heme(b558 versus a) and the donor of the third electron(heme b595 versus CuB). There is even a Trp resi-due between the low-spin heme and the dioxygen-binding high-spin heme in both the bd oxidasesand the HCOs.Although the bd oxidase doesn't pump pro-

tons across the membrane, the existence of pro-ton transfer pathways from the cytoplasm iscrucial for the access of protons to the oxygen-binding site and potentially also for protons in-volved in compensating for the negative chargeof the electrons used for heme reduction. Weidentified two potential pathways throughwhichprotons could pass from the cytoplasm to thehigh-spin heme site. One pathway is formed in-side of the four-helix bundle a1-4 of CydA andthe second one in the symmetry-related a1-4four-helix bundle of CydB. Therefore, we callthese potential routes for proton transfer CydAand CydB pathways (Fig. 3 and fig. S7). The lo-cation of Glu108 (Glu107 inE. coli) in our structure,

584 29 APRIL 2016 • VOL 352 ISSUE 6285 sciencemag.org SCIENCE

Fig. 1. Overall structure of the bd oxidase from G. thermodenitrificans. (A) Ribbon model of thebd oxidase consisting of the subunits CydA (blue), CydB (red), and CydS (beige). (B) Structuralsuperposition of CydA and CydB. The dashed line indicates the border between the hydrophobic andhydrophilic layers of the membrane.

RESEARCH | REPORTSon M

arch 27, 2021

http://science.sciencemag.org/

Dow

nloaded from

together with previousmutagenesis experimentsand Fourier transform infrared spectroscopy data,supports the proposal that this glutamate resi-due is a redox state–dependent mediator of pro-

ton transfer to a charge compensation site (30).With the CydA pathway leading to Glu101, thisresiduemight be the hypothetical protonatablegroup used for charge compensation upon heme

b595 reduction (28). It remains to be investigatedwhether protons entering the CydA pathway canbe transferred from Glu101 to Glu378 for compen-sating the negative charge of a second electronused to reduce the two high-spin hemes. Proto-nation of Glu378 could alternatively also be ac-complished by a proton accessing from theextracellular side. Currently it remains unclearwhether the CydA pathway is solely providingprotons for charge compensation or whetherGlu108 can be a branching point that is able topass protons via the heme b595 propionates tothe oxygen-binding site. With both channelsleading to heme b595, there must be an addi-tional proton transfer step from heme b595 toheme d in order to deliver protons to the oxy-gen reduction site (Fig. 3). Based on the currentstructure, we can exclude the possibility of di-rect proton transfer between the two high-spinhemes. It seems more plausible that the finalproton transfer from heme b595 to heme d isfacilitated by connecting unresolved water mol-ecules. The heme propionates of heme b595are in a hydrophobic environment without anycharge compensation. They are thus most likelyprotonated and could supply protons for waterformation. These protons would be replenishedvia the proton pathways (Fig. 3 and fig. S7).HCOs contain oxygen channels lined by hy-

drophobic residues leading from the membraneinterior to their binuclear heme-Cu active site

SCIENCE sciencemag.org 29 APRIL 2016 • VOL 352 ISSUE 6285 585

Fig. 2. The prosthetic groups in the bd oxidase. (A) Picture of the heme groups represented by stick models bound to CydA (blue). Heme b558 and b595are depicted in yellow and heme d in brown. (B) Surface representation of CydA showing a pocket formed between a1 and a8 that could enable dioxygen todirectly access heme d. (C) The initial electron acceptor heme b558 in proximity to the Q loop. (D) The high-spin heme site composed of hemes b595 and d.

Fig. 3. The potential proton transfer pathways in the bd oxidase.

RESEARCH | REPORTSon M

arch 27, 2021

http://science.sciencemag.org/

Dow

nloaded from

(31). Our bd oxidase structure does not indicatethe presence of such voluminous pathways;instead, molecular oxygen may access heme dlaterally from the alkyl chain interface withthe membrane over a short distance (Fig. 2B).Hence, oxygen dissolved in the membrane couldrapidly bind to heme d without traveling throughany tunnel-like protein cavity.

REFERENCES AND NOTES

1. H. Seedorf et al., FEBS J. 274, 1588–1599 (2007).2. V. B. Borisov, R. B. Gennis, J. Hemp, M. I. Verkhovsky, BBA

Bioenergetics 1807, 1398–1413 (2011).3. S. E. Edwards et al., FEMS Microbiol. Lett. 185, 71–77 (2000).4. R. K. Poole, S. Hill, Biosci. Rep. 17, 303–317 (1997).5. L. Shi et al., Proc. Natl. Acad. Sci. U.S.A. 102, 15629–15634 (2005).6. A. D. Baughn, M. H. Malamy, Nature 427, 441–444 (2004).7. M. G. Mason et al., Nat. Chem. Biol. 5, 94–96 (2009).8. A. Giuffrè, V. B. Borisov, M. Arese, P. Sarti, E. Forte, BBA

Bioenergetics 1837, 1178–1187 (2014).9. M. Berney, T. E. Hartman, W. R. Jacobs Jr., mBio 5, e01275–14

(2014).10. P. Lu et al., Scientific Rep. 5, 10333 (2015).11. M. J. Miller, R. B. Gennis, J. Biol. Chem. 258, 9159–9165 (1983).12. K. Kita, K. Konishi, Y. Anraku, J. Biol. Chem. 259, 3375–3381 (1984).13. J. Sakamoto, A. Matsumoto, K. Oobuchi, N. Sone, FEMS

Microbiol. Lett. 143, 151–158 (1996).14. R. J. Allen et al., BMC Genomics 15, 946 (2014).15. J. Hoeser, S. Hong, G. Gehmann, R. B. Gennis, T. Friedrich,

FEBS Lett. 588, 1537–1541 (2014).16. F. Spinner et al., Biochem. J. 308, 641–644 (1995).17. J. Sun et al., Biochemistry 35, 2403–2412 (1996).18. F. G. M. Wiertz et al., FEBS Lett. 575, 127–130 (2004).19. Materials and methods are available as supplementary

materials on Science Online.20. J. Deisenhofer, O. Epp, K. Miki, R. Huber, H. Michel, Nature 318,

618–624 (1985).21. J. P. Allen, J. C. Williams, FEBS Lett. 438, 5–9 (1998).22. T. Mogi et al., Biochemistry 45, 7924–7930 (2006).23. S. Iwata, C. Ostermeier, B. Ludwig, H. Michel, Nature 376,

660–669 (1995).24. H. M. Berman et al., Nucleic Acids Res. 28, 235–242 (2000).25. S. Ovchinnikov, L. Kinch, H. Park, Y. Liao, J. Pei, D. E. Kim,

H. Kamisetty, N. V. Grishin, D. Baker, eLife Sci. 4, e09248 (2015).26. H. Fang, R. J. Lin, R. B. Gennis, J. Biol. Chem. 264, 8026–8032

(1989).27. T. Mogi, S. Endou, S. Akimoto, M. Morimoto-Tadokoro,

H. Miyoshi, Biochemistry 45, 15785–15792 (2006).28. I. Belevich et al., Proc. Natl. Acad. Sci. U.S.A. 102, 3657–3662 (2005).29. A. Paulus, S. G. Rossius, M. Dijk, S. de Vries, J. Biol. Chem. 287,

8830–8838 (2012).30. K. Yang et al., Biochemistry 46, 3270–3278 (2007).31. V. M. Luna, Y. Chen, J. A. Fee, C. D. Stout, Biochemistry 47,

4657–4665 (2008).

ACKNOWLEDGMENTS

This work was supported by the Max Planck Society and theDeutsche Forschungsgemeinschaft (Cluster of ExcellenceMacromolecular Complexes Frankfurt), by a Grant-in-Aid forScientific Research (C) (25440050 to J.S.) from the Japan Societyfor the Promotion of Science, and by the National Institutes ofHealth (grant R01GM092802). We thank the staff of beamlineX10SA of the Swiss Light Source for assistance. Parts of theexperiments were performed on beamlines ID29, ID23.1, andID23.2 at the European Synchrotron Radiation Facility, Grenoble,France. We are grateful to beamline scientists for providingassistance; D. Baker for support during model building; andR. Gennis, U. Ermler, R. Murali, and A. Resemann for scientificdiscussion and expertise. Coordinates and structure factors havebeen deposited under PDB accession codes 5DOQ and 5IR6.

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/352/6285/583/suppl/DC1Materials and MethodsSupplementary TextFigs. S1 to S7Table S1References (32–59)

13 January 2016; accepted 28 March 201610.1126/science.aaf2477

MICROBIOME

Durable coexistence of donor andrecipient strains after fecalmicrobiota transplantationSimone S. Li,1,2 Ana Zhu,1 Vladimir Benes,3 Paul I. Costea,1 Rajna Hercog,3 Falk Hildebrand,1

Jaime Huerta-Cepas,1 Max Nieuwdorp,4,5,6 Jarkko Salojärvi,7,8 Anita Y. Voigt,1,9,10

Georg Zeller,1 Shinichi Sunagawa,1*WillemM. de Vos,7,11,12* Peer Bork1,10,13,14*

Fecal microbiota transplantation (FMT) has shown efficacy in treating recurrent Clostridiumdifficile infection and is increasingly being applied to other gastrointestinal disorders, yet thefate of native and introduced microbial strains remains largely unknown.To quantify the extentof donor microbiota colonization, we monitored strain populations in fecal samples from arecent FMTstudy on metabolic syndrome patients using single-nucleotide variants inmetagenomes.We found extensive coexistence of donor and recipient strains, persisting 3months after treatment. Colonization success was greater for conspecific strains than for newspecies, the latter falling within fluctuation levels observed in healthy individuals over a similartime frame. Furthermore, same-donor recipients displayed varyingdegrees ofmicrobiota transfer,indicating individual patterns of microbiome resistance and donor-recipient compatibilities.

Fecalmicrobiota transplantation (FMT), whichentails the transfer of a microbial commu-nity from a healthy donor to a patient (of-ten after bowel lavage), has emerged as apromising treatment option for a range of

chronic disorders (1–3). However, despite having asuccess rate of over 90% for recurrent Clostridiumdifficile infection (R-CDI) (1), there are indica-tions that the therapy as currently performed isnot as effective in treating other diseases (4). This,in addition to inherent risks (5), has created a needto characterize and understand the full effects ofFMT on the microbiome of the recipient, acrossdifferent diseases.Recent studies have indicated that establish-

ment of donor-specific species is possible alongsideresident microbiota of the recipient (6–9) andthat they can still be detected 70 days after FMT

(7). However, the definitive origin of these newlyobserved species—whether they are indeed fromthe donor or fromother sources (for example, dietor environment) or had simply been below detec-tion levels in the recipient—is uncertain. More-over, in the majority of these studies, the FMTprocedurewas preceded by a course of antibiotics,which is known to alter the gut flora (10, 11) andmay or may not facilitate implantation of exoge-nous microorganisms.The full extent to which donor microbiota col-

onize in a recipient host has not yet been mea-sured, as studies to date have been restricted togenus- and species-level comparisons and havenot distinguished or elucidated the fate of donorand recipient strains of the same species (6–9, 12).Genetic variants of bacterial species are knownto coexist in the gut (13), and strain-level differenceshave recently been shown to have functional andclinically relevant consequences (13–15). However,other studies in humans andmodel systems havefound thatnewly introducednonpathogenic strainsareunable topersist inanestablishedgutecosystem—especially if the species was already present—bethey probiotics (16), xenomicrobiota (17), or evenmodified strains isolated from the same healthyindividual (18).Here, we quantify and describe the extent of

changes to the population structure of the gutmicrobiome after FMT, at both species and strainlevel, using shotgun metagenomic data. Donorstrains established extensively in the recipientand persisted over the 3-month observationalperiod, either replacing or, more strikingly, exist-ing alongside indigenous strains. Outcomes variedacross donor-recipient pairs, indicating thatmicro-biome compatibility (the likelihood that donorstrain populations are able to coexist or even re-place microbial strains in the recipient) is a fac-tor that could provide a rationale formore targetedmicrobiota-based therapies.

586 29 APRIL 2016 • VOL 352 ISSUE 6285 sciencemag.org SCIENCE

1Structural and Computational Biology Unit, EuropeanMolecular Biology Laboratory, 69117 Heidelberg, Germany.2School of Biotechnology and Biomolecular Sciences,University of New South Wales, 2052 Sydney, Australia.3Genomics Core Facility, European Molecular BiologyLaboratory, 69117 Heidelberg, Germany. 4Department ofVascular Medicine, Academic Medical Center, 1105 AZAmsterdam, Netherlands. 5Diabetes Center, Vrije UniversityMedical Center, 1018 HV Amsterdam, Netherlands.6Wallenberg Laboratory, University of Gothenburg, 41345Gothenburg, Sweden. 7Department of Veterinary Biosciences,University of Helsinki, 00014 Helsinki, Finland. 8Departmentof Biosciences, University of Helsinki, 00014 Helsinki,Finland. 9Department of Applied Tumor Biology, Institute ofPathology, University Hospital Heidelberg, 69120 Heidelberg,Germany. 10Molecular Medicine Partnership Unit, Universityof Heidelberg and European Molecular Biology Laboratory,69120 Heidelberg, Germany. 11Laboratory of Microbiology,Wageningen University, 6703 HB Wageningen, Netherlands.12Immunobiology Research Program, Department ofBacteriology and Immunology, University of Helsinki, 00014Helsinki, Finland. 13Max Delbrück Centre for MolecularMedicine, 13125 Berlin, Germany. 14Department ofBioinformatics, Biocenter, University of Würzburg, 97074Würzburg, Germany.*Corresponding author. Email: bork@embl.de (P.B.); willem.devos@wur.nl (W.M.dV.); sunagawa@embl.de (S.S.)

RESEARCH | REPORTSon M

arch 27, 2021

http://science.sciencemag.org/

Dow

nloaded from

reductasesStructure of a bd oxidase indicates similar mechanisms for membrane-integrated oxygen

Tomoichirou Kusumoto, Junshi Sakamoto and Hartmut MichelSchara Safarian, Chitra Rajendran, Hannelore Müller, Julia Preu, Julian D. Langer, Sergey Ovchinnikov, Taichiro Hirose,

DOI: 10.1126/science.aaf2477 (6285), 583-586.352Science 

, this issue p. 583; see also p. 518Scienceoxidases, despite serving a similar function.and triangular arrangement of its heme cofactors bear little structural resemblance to those of other membrane-spanningcytochrome bd oxidase from a thermophilic bacterium (see the Perspective by Cook and Poole). The overall structure

were able to resolve the three-dimensional structure ofet al.has eluded high-resolution crystallography. Safarian bd oxidases serve this role and also protect pathogenic bacteria from nitric acid; however, this class of enzymes so far

Microorganisms have evolved a number of enzymes to reduce oxygen and prevent oxidative stress. CytochromePeering into a membrance oxidase

ARTICLE TOOLS http://science.sciencemag.org/content/352/6285/583

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2016/04/27/352.6285.583.DC1

CONTENTRELATED http://science.sciencemag.org/content/sci/352/6285/518.full

REFERENCES

http://science.sciencemag.org/content/352/6285/583#BIBLThis article cites 58 articles, 13 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2016, American Association for the Advancement of Science

on March 27, 2021

http://science.sciencem

ag.org/D

ownloaded from