1. Structural Biology and Molecular Biophysics
Download icon

Structure of Mycobacterium tuberculosis cytochrome bcc in complex with Q203 and TB47, two anti-TB drug candidates

  1. Shan Zhou
  2. Weiwei Wang
  3. Xiaoting Zhou
  4. Yuying Zhang
  5. Yuezheng Lai
  6. Yanting Tang
  7. Jinxu Xu
  8. Dongmei Li
  9. Jianping Lin
  10. Xiaolin Yang
  11. Ting Ran
  12. Hongming Chen
  13. Luke W Guddat
  14. Quan Wang
  15. Yan Gao  Is a corresponding author
  16. Zihe Rao  Is a corresponding author
  17. Hongri Gong  Is a corresponding author
  1. State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy, Nankai University, China
  2. State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, China
  3. Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, China
  4. Bioland Laboratory (Guangzhou Regenerative Medicine and Health - Guangdong Laboratory), China
  5. School of Chemistry and Molecular Biosciences, The University of Queensland, Australia
  6. National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, China
  7. Laboratory of Structural Biology, Tsinghua University, China
Research Article
  • Cited 0
  • Views 89
  • Annotations
Cite this article as: eLife 2021;10:e69418 doi: 10.7554/eLife.69418

Abstract

Pathogenic mycobacteria pose a sustained threat to global human health. Recently, cytochrome bcc complexes have gained interest as targets for antibiotic drug development. However, there is currently no structural information for the cytochrome bcc complex from these pathogenic mycobacteria. Here, we report the structures of Mycobacterium tuberculosis cytochrome bcc alone (2.68 Å resolution) and in complex with clinical drug candidates Q203 (2.67 Å resolution) and TB47 (2.93 Å resolution) determined by single-particle cryo-electron microscopy. M. tuberculosis cytochrome bcc forms a dimeric assembly with endogenous menaquinone/menaquinol bound at the quinone/quinol-binding pockets. We observe Q203 and TB47 bound at the quinol-binding site and stabilized by hydrogen bonds with the side chains of QcrBThr313 and QcrBGlu314, residues that are conserved across pathogenic mycobacteria. These high-resolution images provide a basis for the design of new mycobacterial cytochrome bcc inhibitors that could be developed into broad-spectrum drugs to treat mycobacterial infections.

Editor's evaluation

The authors describe the cryo EM structures of cytochrome bcc from mycobacteria in complex with clinical drug candidates Q203 and TB47, which are being trialled for the treatment of tuberculosis. The cryo EM structures were obtained by expressing and purifying a hybrid super complex; the bcc part from M. tuberculosis and the cytochrome oxidase part from M. smegmatis. The cryo EM structures show how the Q203 and TB47 inhibitors bind and block the binding of quinone at the Qo site, thus inhibiting electron-transfer. The complex structures should facilitate the structure-based development of Q203 and TB47 drug compounds against tuberculosis.

https://doi.org/10.7554/eLife.69418.sa0

Introduction

Mycobacteria, which belong to the phylum Actinobacteria, have coevolved with humans over thousands of years (Chisholm et al., 2016). Approximately 200 species of mycobacteria have been identified that have diverse lifestyles, morphologies, and metabolic pathways (Tortoli et al., 2017). Mycobacteria can broadly be grouped into two categories: tuberculosis-causing mycobacteria and non-tuberculous mycobacteria (NTM). Mycobacterium leprae is often represented in a distinct genetic clade owing to its genetic and phenotypic differences compared to other mycobacterial species (Cole et al., 2001). Mycobacteria can be further classified into fast-growing and slow-growing species or species complexes; these assignments are according to the physiological, phenotypic and phylogenetic characteristics (Rastogi et al., 2001). Although nearly all mycobacteria are saprophytes or non-pathogenic to humans, a few species cause diseases, resulting in pulmonary and extra-pulmonary infections that can affect nearly all organs. Infections, which are caused by strict or opportunistic pathogenic mycobacteria (Figure 1), pose a sustained threat to human health. Of these, tuberculosis (TB), caused by Mycobacterium tuberculosis (M. tuberculosis), is the most serious, leading to ~1.2 million fatalities per year (World Health Organization, 2019). Infections involving other pathogenic mycobacteria, for example, Mycobacterium abscessus and Mycobacterium avium complex, are on the rise with some outnumbering those caused by M. tuberculosis in countries including the United States (Donohue, 2018; Johansen et al., 2020). These infections are notoriously difficult to treat due to intrinsic or emerging resistance to many common antibiotics, thus exacerbating the challenge to find suitable drug targets.

Sequence similarity comparison of M. tuberculosis QcrB with other pathogenic mycobacteria.

Oxidative phosphorylation (OXPHOS) has gained interest as a target space for antibiotic drug development (Cook et al., 2017; Hards et al., 2020). In OXPHOS, the protein complexes of the electron transport chain (ETC) establish a proton motive force (PMF) across a biomembrane that drives the synthesis of adenosine triphosphate (ATP) by ATP synthase (Mitchell, 1961). Maintenance of PMF and ATP homeostasis is required for the survival of both replicative and non-replicative (often referred to as dormant) mycobacteria, and its dissipation leads to a rapid loss of cell viability and cell death (Koul et al., 2008; Rao et al., 2008). The reliance on the PMF and ATP homeostasis thus highlights the importance of the mycobacterial proton-pumping cytochrome bcc-aa3 supercomplex, which consists of a bcc menaquinol reductase (complex III, CIII) and an aa3 oxidase (complex IV, CIV) that are tightly associated (Gong et al., 2018; Kim et al., 2015; Megehee et al., 2006). Several studies support the attractiveness of cytochrome bcc-aa3 for mycobacterial drug development (de Jager et al., 2020; Liu et al., 2019; Lu et al., 2019; Pethe et al., 2013; Scherr et al., 2018). Given the strict sequence conservation of this complex (Figure 1), broad-spectrum activity of a therapeutic within the pathogenic mycobacteria is likely (Lee et al., 2020b). Interestingly, all the cytochrome bcc-aa3 inhibitors published to date appear to target the QcrB subunit (Figure 1) of the cytochrome bcc complex and are likely bound to the menaquinol-binding (Qp) site of the QcrB subunit (Lee et al., 2020b). The most advanced of these are Q203 and TB47, which have been shown to clear infections due to M. tuberculosis (de Jager et al., 2020; Lu et al., 2019; Pethe et al., 2013) and Mycobacterium ulcerans (Liu et al., 2019; Scherr et al., 2018). Q203 has recently completed phase II clinical trials for TB treatment (ClinicalTrials.gov number, NCT03563599; de Jager et al., 2020). TB47 has also been evaluated in a preclinical study (http://www.newtbdrugs.org/pipeline/clinical).

To progress an understanding of cytochrome bcc and its interaction with new drug leads, here we have determined the atomic resolution cryo-electron microscopy (cryo-EM) structures of M. tuberculosis cytochrome bcc alone (2.68 Å resolution) and in complex with Q203 (2.67 Å resolution) and TB47 (2.93 Å resolution). These high-resolution structures will greatly accelerate efforts towards structure-guided drug discovery.

Results and discussion

Structure of M. tuberculosis cytochrome bcc

Considering that the hybrid supercomplex consisting of M. tuberculosis CIII and Mycobacterium smegmatis CIV can be stabilized as a functional assembly (Kim et al., 2015), we first chose to express and purify this complex to a high level of homogeneity (Figure 2—figure supplement 1). We confirmed this complex to be active with a turnover number of 23.3 ± 2.4 e-s–1 (mean ± SD, n = 4), in agreement with the previous study that shows M. tuberculosis CIII can functionally complement native M. smegmatis CIII and maintain the growth of M. smegmatis (Kim et al., 2015). The structure of this hybrid supercomplex was determined by cryo-EM to an overall resolution of 2.68 Å, allowing that the components to be clearly assigned in the density map (Figure 2—figure supplements 2 and 3, Table 1). The hybrid supercomplex including the M. tuberculosis CIII dimer forms a pseudo twofold symmetrical compact rod, but with a slight curvature in the membrane plane, as previously observed in the M. smegmatis CIII2CIV2 supercomplex (Gong et al., 2018; Wiseman et al., 2018; Figure 2A and B). The hybrid supercomplex is structurally similar to the supercomplex isolated from M. smegmatis (Gong et al., 2018; Wiseman et al., 2018; Figure 2—figure supplement 4), except that subunits LpqE and Unk (probably MSMEG_0987) (Wiseman et al., 2018) were not observed here. The absence of these two subunits may be due to their depletion during purification of the supercomplex or their map density signal was averaged to background noise during structural determination. As expected, the topology of M. tuberculosis cytochrome bcc and M. smegmatis CIV in the hybrid supercomplex is also similar to that of the equivalent subunits in the M. smegmatis CIII2CIV2 supercomplex (Figure 2C and D, Figure 2—figure supplements 4 and 5). As such, there is no notable non-native contacts that resulted from the hybrid assembly in the hybrid supercomplex. This is attributable to the high structural similarity between the M. tuberculosis cytochrome bcc and M. smegmatis cytochrome bcc. In M. tuberculosis cytochrome bcc, QcrA has three transmembrane helices (TMHs) and has a ‘U’-shaped structure (Figures 2D and 3A). The N-terminal region with TMH1/2 and the TMH3 make up the two arms of the ‘U’ structure. These arms are linked by the region located near the cytoplasmic side. Attached to QcrATMH3 is the C-terminal domain, which faces the periplasmic side of the membrane and holds the [2Fe-2S] cluster in place. QcrB has eight TMHs (Figures 2D and 3B). Four of these are responsible for burying two functionally important heme b cofactors (high potential heme bH and low potential heme bL). Heme bL and heme bH are located between TMH I/II and TMH III/IV, respectively. Heme bL is near the periplasmic side and heme bH near the cytoplasmic side. QcrC is a transmembrane protein with a C-terminal TMH located between QcrBTMH5 and QcrBTMH7 (Figures 2D and 3C). The N-terminal periplasmic portion of QcrC can be divided into two heme-containing cytochrome c domains designated D1 and D2. The D1 domain protrudes out of the core of CIII, whereas the D2 domain interacts extensively with QcrA and QcrB. In the M. smegmatis CIII2CIV2 supercomplex, the cytochrome cc head domain of QcrC adopts an ‘open’ or a ‘closed’ conformation (Gong et al., 2018; Wiseman et al., 2018). However, in this structure it is only the closed conformation (Figure 2—figure supplement 4). Considering the high topology similarity of the subunits and the arrangement of prosthetic groups compared to previous M. smegmatis CIII2CIV2 supercomplex, the mechanism of action of the hybrid supercomplex including M. tuberculosis cytochrome bcc is expected to be the same as that for the M. smegmatis CIII2CIV2 supercomplex (Gong et al., 2018).

Figure 2 with 5 supplements see all
Overall architecture of the hybrid supercomplex.

(A) Front view and (B) top view of the cryo-electron microscopy (cryo-EM) map of hybrid supercomplex at 2.68 Å resolution. QcrA, QcrB, and QcrC of the M. tuberculosis cytochrome bcc are colored teal, purple, and salmon, respectively. Other subunits of the hybrid supercomplex are in gray. (C) Cartoon representation of cytochrome bcc, using the same color scheme as above. The twofold symmetry of the dimer is depicted by the gray axis. The heme groups (bH, bL, cD1, and cD2) and menaquinone/menaquinol (MKP/MKN) are shown as stick models. The [2Fe-2S] clusters are shown as spheres. (D) A cross-sectional view (top) of the helices in the cytochrome bcc dimer.

Figure 2—source data 1

Oxygen consumption of the hybrid supercomplex measures using Clark-type oxygen electrode.

https://cdn.elifesciences.org/articles/69418/elife-69418-fig2-data1-v1.xlsx
Figure 3 with 3 supplements see all
Structure of the M. tuberculosis cytochrome bcc subunits.

Cartoon representation of the monomers of (A) QcrA, (B) QcrB, and (C) QcrC, with prosthetic groups. (D) The QP-binding site and (E) QN-binding site. The residues potentially involved in the binding of MK/MKH2 are shown with side chains in stick model representation. MK/MKH2 have their carbon atoms in green and are represented as stick models. The [2Fe-2S] and heme groups are shown as spheres and labeled.

Table 1
Cryo-electron microscopy data collection, refinement, and validation statistics of hybrid supercomplex.
Stateapo
Data collection
MicroscopeTitan Krios
Voltage (kV)300
Magnification29,000×
DetectorGatan K3
Data collection softwareSerialEM
Electron exposure (e2)60
Defocus range (μm)–1.2 to –1.8
Pixel size (Å)0.82
Data processing
Number of micrographs4141
Final particle images112,804
Symmetry imposedC1
Map resolution (Å)Fourier shell correlation 0.143 threshold
2.68
Refinement
Initial model used (PDB code)6ADQ
Map sharpening B factor (Å2)d FSC model (0.143) masked–65.32.5
Map correlation coefficient0.89
Mean CC for ligands0.78
Model composition
Non-hydrogen atoms42,279
Protein residues5122
Ligands
9Y0: 2
CDL: 17
9YF: 4
HEA: 4
HEC: 4
MQ9: 10
HEM: 4
PLM: 4
CU: 8
FES: 2
Root mean squared deviations
Bond lengths (Å)0.005
Bond angles (°)1.057
Validation
MolProbity score1.86
Clashscore7.26
Poor rotamers (%)0.05
Ramachandran plot
Favored (%)92.76
Allowed (%)6.97
Outliers (%)0.28
Cβ outliers (%)0.00

Quinone and quinone-binding pockets of M. tuberculosis cytochrome bcc

Quinone-binding sites are essential to the function of respiratory chain complexes and thus are good targets for drug discovery (Harikishore et al., 2021; Lee et al., 2020b). Structurally diverse quinones such as ubiquinone and menaquinone (MK) bind in the mitochondrial ETC (Zhang et al., 1998) and mycobacterial ETC (Gong et al., 2018; Wiseman et al., 2018), respectively, indicating that there are structural differences in the quinone-binding sites of the different species. Structural differences in the ubiquinone-binding site are also observed when the mitochondrial and Escherichia coli respiratory chain complex IIs are compared (Huang et al., 2021). Therefore, it is feasible that species-specific quinone-binding-site inhibitors could be designed. In the present structure, two canonical quinone-binding sites are identified (Figure 3D and E): (i) the quinol oxidation site (QP site) and (ii) the quinone reduction site (QN site). In addition, other quinone-binding sites with quinone bound are also identified as observed in our previous study (Gong et al., 2018; Figure 3—figure supplement 1). These quinones are remote from the canonical quinone-binding sites and into the membrane space, suggesting that they have a structural rather than functional role. The QP site responsible for menaquinol (MKH2) oxidation is near heme bL, whereas the QN site responsible for MK reduction is close to heme bH (Figure 3D and E, Figure 3—figure supplement 1). The QP site is at the center of an inverted triangle structure surrounded by helices (Figure 3D). One MK molecule was identified at this site with its naphthoquinone ring surrounded by hydrophobic residues, QcrBPhe158, QcrBTyr161, QcrBLeu180, QcrBIle183, QcrBMet310, and QcrBMet342. Its hydrophobic tail, which contains multiple isoprenoid groups, wraps around QcrBTMH6 down to its cytoplasmic end and in so doing interacts with QcrBMet187, QcrBAla339, QcrBLeu344, and QcrBVal347. The edge-to-edge distance from MK to heme bL is 15 Å, exceeding the 14 Å limit for efficient physiological electron transfer (Page et al., 1999). Furthermore, there are no observed hydrogen bonds to the carbonyl groups of MK, which are needed to help deprotonate MKH2. In addition, structural superposition with the inhibitor-bound bc1 complex also shows that the MK binds deeper into the QP pocket in the bc1 complex (Birth et al., 2014; Solmaz and Hunte, 2008) than in the currently described complex (Figure 3—figure supplement 2). Hence, the endogenous electron donor MKH2 should bind deeper inside the pocket and close to polar residues such as QcrBTyr161, QcrBThr313, and QcrBAsp314 (Figure 3D) to facilitate electron transfer. In addition, MK could also be modeled to fit deeper inside the pocket (Figure 3—figure supplement 3). Thus, we speculate that what is observed here is the oxidized product as it leaves the QP site. It is worth noting that all the reported inhibitors including Q203 (Pethe et al., 2013) and TB47 (Lu et al., 2019) are suggested to interact with this QP site. In addition, the QN site is formed largely by QcrBTMH1, QcrBTMH4, QcrBTMH5 and one loop region of QcrB (Figure 3E). The head group of MK is bound in this pocket interacting with QcrBPhe39, QcrBGlu49, QcrBLeu225, QcrBLeu232, QcrBTrp236, and QcrBPhe262, and its long hydrophobic tail extends along QcrBTMH1 towards the periplasmic side. MK/MKH2 are part of the Q-cycle hypothesis and essential for electron transfer in the cytochrome bcc complex (Gong et al., 2018). Given the crossspecies activity of this complex (Lee et al., 2020b) and high homology of the QcrB subunits across mycobacterial pathogens (Figure 1), these data open the way for the discovery of broad-spectrum mycobacterial agents based on rational, structure-based inhibitor design principles.

Q203 interactions in M. tuberculosis cytochrome bcc binding pocket

Q203 has recently been subjected to a phase II clinical study for M. tuberculosis treatment (de Jager et al., 2020). This compound has also been shown to be strongly bactericidal against M. ulcerans (Scherr et al., 2018). It is suggested to be an inhibitor that competes with endogenous substrate binding (QP site) of the cytochrome bcc complex (Pethe et al., 2013), but this hypothesis is yet to be verified by direct experimental evidence. To obtain atomic information on the mode of binding of Q203 to cytochrome bcc, we have determined the structure of the hybrid supercomplex as described above in the presence of Q203 by cryo-EM to an overall resolution of 2.67 Å (Figure 4—figure supplements 1 and 2, Table 2). We observe that close to the Qp-binding pocket within the membrane of each QcrB of cytochrome bcc there is density for Q203 (Figure 4A and B). All of the Q203 molecules fill each QcrB subunit binding deeply into the QP pocket and with identical binding modes. The key interactions that anchor Q203 are (i) a hydrogen bond between the hydroxyl oxygen of the side chain of QcrBThr313 and the amine in the carboxamide linker of Q203 (3.0 Å), (ii) a halogen bond between the chlorine atom of the heterocyclic group and an ordered water molecule that simultaneously forms a hydrogen bond with the hydroxyl oxygen of the side chain of QcrBTyr164 (Figure 4—figure supplement 3), (iii) a hydrogen bond between the side chain of QcrBGlu314 and the nitrogen atom in the imidazopyridine ring (3.0 Å), and (iv) a hydrogen bond between the side chain of QcrAHis375 and the nitrogen atom in the imidazopyridine ring (2.98 Å) (Figure 4C). In addition, the carbon atoms of Q203 make hydrophobic interactions with QcrBGly175, QcrBAla179, QcrBLeu180, QcrBThr184, QcrBSer304, QcrBPro306, QcrBMet310, QcrBAla317, and QcrBMet342. These extensive interactions are in agreement with the fact that the activity of the supercomplex is inhibited by Q203 in vitro according to the DMNQH2/oxygen oxidoreductase activity assay. After addition of Q203, the turnover number of the hybrid supercomplex reduces to 5.8 ± 2.4 e-s–1 from 23.3 ± 2.4 e-s–1 (mean ± SD, n = 4; Figure 4—figure supplement 4). In addition, functional studies have shown that substitution of QcrBThr313 to alanine confers Q203 resistance (Pethe et al., 2013). The binding of Q203 involves residues from both QcrA and QcrB. Due to the need to form stabilizing interactions between subunits, resistance may be more difficult to achieve here than if the binding site is within only one subunit. Consistently, the mapping of reported mutations in Q203-resistant M. tuberculosis reveals that they are positioned directly where Q203 binds in this structure (Lupien et al., 2020; Figure 4—figure supplement 5).

Figure 4 with 5 supplements see all
Cryo-electron microscopy (cryo-EM) structure of the hybrid supercomplex in the presence of Q203.

(A) Side (left) and top (right) views of the cryo-EM structure of the M. tuberculosis cytochrome bcc complex presented as a cartoon representation. Q203 (orange) is bound to the Qp site. (B) Visualization of densities for Q203. Hydrogen bonds are shown as dotted lines. (C) Plot of distances of various parts of Q203 to residues in the Qp site as determined using LIGPLOT (https://www.ebi.ac.uk/thornton-srv/software/LIGPLOT/).

Table 2
Cryo-electron microscopy data collection, refinement, and validation statistics for the Q203-bound hybrid supercomplex.
StateQ203
Data collection
MicroscopeTitan Krios
Voltage (kV)300
Magnification29,000×
DetectorGatan K3
Data collection softwareSerialEM
Electron exposure (e2)60
Defocus range (μm)–1.2 to –1.8
Pixel size (Å)0.82
Data processing
Number of micrographs3763
Final particle images106,770
Symmetry imposedC1
Map resolution (Å)Fourier shell correlation (FSC) 0.143 threshold
2.67
Refinement
Initial model used (PDB code)6ADQ
Map sharpening B factor (Å2)d FSC model (0.143) masked–70.02.6
Map correlation coefficient0.88
Mean CC for ligands0.76
Model composition
Non-hydrogen atoms42,695
Protein residues5132
Ligands
9Y0: 2
CDL: 17
9YF: 4
HEA: 4
HEC: 4
MQ9: 8
HEM: 4
PLM: 4
CU: 8
FES: 2
HUU (Q203): 2
Root mean squared deviations
Bond lengths (Å)0.003
Bond angles (°)0.659
Validation
MolProbity score1.84
Clashscore7.97
Poor rotamers (%)6.23
Ramachandran plot
Favored (%)93.07
Allowed (%)6.61
Outliers (%)0.31
Cβ outliers (%)0.00

TB47-binding mode of M. tuberculosis cytochrome bcc

TB47, also currently being evaluated in preclinical studies, has been suggested to target the QcrB of cytochromes bcc from M. tuberculosis (Lu et al., 2019) and M. ulcerans (Liu et al., 2019). The 2.93 Å cryo-EM map shows density for TB47 and confirms that it binds in the same location as Q203 (Figure 5A and B, Figure 5—figure supplements 1 and 2, Table 3). Three hydrogen bond interactions are observed involving the side chains of QcrBThr313, QcrBGlu314, and QcrAHis375. Similar interactions are also observed when Q203 binds (Figure 5C). QcrBTyr161, QcrBLeu171, QcrBGly175, QcrBAla179, QcrBLeu180, QcrBThr184, QcrBMet187, QcrBLeu194, QcrBSer304, QcrBGly305, and QcrBMet342 also contribute to TB47 binding, largely through hydrophobic interactions (Figure 5C). Unlike when Q203 binds (Figure 4—figure supplement 3), there is no interaction between QcrBTyr164 and TB47 (Figure 5—figure supplement 3). However, after addition of TB47, the values of turnover number for the hybrid supercomplex reduced to 5.1 ± 2.9 e-s–1 (mean ± SD, n = 4; Figure 5—figure supplement 4), a similar value that was observed when Q203 binds. Thus, the absence of this does not greatly diminish inhibition. A mutation in TB47-resistant M. smegmatis (M. tuberculosis: H195Y) is close to the Qp-binding site (Lu et al., 2019; Figure 5—figure supplement 5). As a result of the change in shape, it would appear to contribute to steric interference with the binding of TB47, thus accounting for the observed resistance.

Figure 5 with 5 supplements see all
Cryo-electron microscopy (cryo-EM) structure of the hybrid supercomplex in the presence of TB47.

(A) Side (left) and top (right) views of the cryo-EM structure of the M. tuberculosis cytochrome bcc complex presented as a cartoon representation. TB47 (green) and Q203 (orange) are bound to the Qp site. (B) Visualization of the density for TB47. Hydrogen bonds are shown as dotted lines. (C) Plot of distances of various parts of TB47 to residues in the Qp site were determined using LIGPLOT (https://www.ebi.ac.uk/thornton-srv/software/LIGPLOT/).

Table 3
Cryo-electron microscopy data collection, refinement, and validation statistics for the TB47-bound hybrid supercomplex.
StateTB47
Data collection
MicroscopeTitan Krios
Voltage (kV)300
Magnification29,000×
DetectorGatan K3
Data collection softwareSerialEM
Electron exposure (e2)60
Defocus range (μm)–1.2 to –1.8
Pixel size (Å)0.82
Data processing
Number of micrographs2698
Final particle images169,988
Symmetry imposedC1
Map resolution (Å)Fourier shell correlation (FSC) 0.143 threshold
2.93
Refinement
Initial model used (PDB code)6ADQ
Map sharpening B factor (Å2)d FSC model (0.143) masked–97.52.9
Map correlation coefficient0.90
Mean CC for ligands0.79
Model composition
Non-hydrogen atoms42,679
Protein residues5119
Ligands
9Y0: 2
CDL: 17
9YF: 4
HEA: 4
HEC: 4
MQ9: 8
HEM: 4
PLM: 4
CU: 8
FES: 2
HV0 (TB47): 2
Root mean squared deviations
Bond lengths (Å)0.005
Bond angles (°)0.739
Validation
MolProbity score1.87
Clashscore8.75
Poor rotamers (%)6.23
Ramachandran plot
Favored (%)92.39
Allowed (%)7.25
Outliers (%)0.36
Cβ outliers (%)0.00

Specificity of Q203 and TB47 for mycobacterial cytochrome bcc complex

The basis for the high specificity of Q203 and TB47 toward the Qp site of mycobacterial cytochromes bcc becomes apparent in the structural comparison between the QcrB subunit of M. tuberculosis and counterparts from other species (Figure 6). The highly conserved residues that are involved in the binding of these two molecules in this region (Figure 6—figure supplement 1) suggest a consistent overall fold and binding site exists in mycobacteria. This is also in agreement with the fact that Q203 and TB47 show antimycobacterial activity across many species (de Jager et al., 2020; Liu et al., 2019; Lu et al., 2019; Pethe et al., 2013; Scherr et al., 2018). It is worth noting that the structures of cytochrome bcc from M. tuberculosis and M. smegmatis have high similarity (Figure 2—figure supplement 5), and no steric hindrance is observed between the Q203 and M. smegmatis cytochrome bcc (Figure 6). This observation indicates that Q203 should have a similar binding mechanism and a similar effect on the activity of cytochrome bcc from M. smegmatis and M. tuberculosis. This is in good agreement with previous antimycobacterial activity data and inhibition data for the bcc complexes from M. smegmatis and M. tuberculosis (Gong et al., 2018; Lu et al., 2018). In contrast, in other prokaryotic, eukaryotic and human Qp-binding pockets, for example, from Saccharomyces cerevisiae (Lange and Hunte, 2002), Rhodobacter sphaeroides (Esser et al., 2008), and human (Guo et al., 2017), many of the observed interactions would not be possible (Figure 6). This suggests that Q203 and TB47 should have low-binding affinity toward its counterpart QcrB in non-mycobacterial bacteria and in eukaryotes. Even if there is some flexibility in the Qp-binding pocket that enables some level of binding, key residues that enable the binding of Q203 and TB47 in the mycobacteria are not present in other bacteria and eukaryotes (Figure 6—figure supplement 1). In addition, the binding free energies of Q203 between wild-type and mutant hybrid supercomplex were calculated and used to further elucidate the importance of some key residues that could play a role in determining the inhibitor specificity. According to a previous functional study (Pethe et al., 2013), structural comparison (Figure 6), and sequence alignment (Figure 6—figure supplement 1), QcrBThr313 and QcrBGlu314 of hybrid supercomplex were selected and mutated to their counterparts in the human supercomplex, namely residues Ala and Tyr, respectively. It is believed that Ala would decrease the affinity of Q203 (Figure 4; Pethe et al., 2013) and the Tyr could lead to steric hindrance of Q203 binding (Figure 6). 75 ns simulations were used for the calculation of binding free energy (Figure 6—figure supplement 2). The different values for the root mean squared deviation (RMSD) suggest that three mutants might have the different effects on the structural conformation of QcrB, and the dual mutant (Thr313Ala and Glu314Tyr) induced significantly structural changes compared to the wild-type complex. The 75 ns simulation was further equally divided into three phases, and the average values were used as the final relative binding free energy (Table 4), which was calculated as the binding free energy of Q203 in the mutant minus that of the wild-type systems. It is suggested that the wild-type system has stronger affinity than the mutant systems (Thr313Ala or Glu314Tyr). The relative free energy of binding calculation for Q203 in the dual mutant (Thr313Ala and Glu314Tyr) indicates that its binding affinity is significantly weaker compared to the wild-type or single-mutant systems. Therefore, the suggestions drawn from the calculation of binding free energy of Q203 provide further evidence for specificity of Q203 inhibition. In summary, these observations correlate with the observed low general antibacterial activity and low cytotoxicity of Q203 and TB47 (Liu et al., 2019; Lu et al., 2019; Pethe et al., 2013; Scherr et al., 2018).

Figure 6 with 2 supplements see all
Structural alignment between the M. tuberculosis Qp-binding pocket where Q203 or TB47 binds with homologous subunits from four other species.

These subunits are from (A) S. cerevisiae (pink, PDB: 1KYO), (B) R. sphaeroides (blue, PDB: 2QJP), (C) Homo sapiens (green; PDB: 5XTE), and (D) M. smegmatis (violet, PDB: 6ADQ). Residues causing steric clashes in the homologous subunits are labeled.

Table 4
Relative binding free energy (kcal/mol) for Q203 in three mutants of QcrB compared to the wild-type (WT).
Mutant25–50 ns50–75 ns75–100 nsAverageStandard deviation
T313A6.3211.018.118.482.37
T314Y4.546.008.076.201.77
T313A + E314Y10.2816.3013.6413.413.02

Implications of the Q203 and TB47 inhibitory mechanism

To gain further insights into the mechanism of action of Q203, we compared the structures of M. tuberculosis cytochrome bcc in the presence and absence of Q203 (Figure 7—figure supplement 1A). The structure of apo cytochrome bcc is almost identical with the Q203-bound structure (rmsd of 0.50 Å for all Ca atoms), which suggests that Q203 binding does not significantly affect the overall architecture of cytochrome bcc. A comparison of the Q203-bound and Q203-free cytochrome bcc structures shows residues involved in the binding pocket move outward, thus adapting to the shape of Q203 (Figure 7—figure supplement 1B). Specifically, the side chains of QcrBSer304, QcrBGlu313, QcrBGlu314, and QcrBMet342 undergo significant conformational changes to form hydrogen bonds with Q203. The binding of TB47 to M. tuberculosis cytochrome bcc also induces very similar conformational changes in the Qp-binding pocket to those seen for Q203 (rmsd of 0.454 Å for all Ca atoms) (Figure 7—figure supplement 1B). Differences in binding are due to the different ethyl group and methyl moieties in the head groups of Q203 and TB47. It is also important to note that one endogenous substrate molecule is also bound to the Qp site in the apo structure of cytochrome bcc, which potentially affects the evaluation of the conformational changes upon the binding of Q203 or TB47.

When analyzing the superimposed structures (Figure 7—figure supplement 1), it is apparent that Q203 and TB47 act competitively with the quinol binding as they almost completely occupy the Qp pocket. We therefore conclude that Q203 and TB47 are bona fide analogs of the substrate, and thus ultimately function by hindering the downstream synthesis of ATP (Figure 7). These two compounds are also highly bactericidal against M. ulcerans, almost certainly targeting the Qp-binding site (Liu et al., 2019; Scherr et al., 2018). In summary, the sequences of the QcrB subunits have high homology across pathogenic mycobacteria (Lee et al., 2020b) and the essential residues (QcrBGlu313 and QcrBGlu314) that are involved in hydrogen-bonding interactions with the inhibitors (Pethe et al., 2013; Scherr et al., 2018) are conserved across pathogenic mycobacteria (Figure 7—figure supplement 2).

Figure 7 with 2 supplements see all
Schematic of M. tuberculosis cytochrome bcc inhibition by Q203 and TB47.

The two monomers of M. tuberculosis cytochrome bcc are colored magenta and cyan, respectively. The binding of Q203 (orange spheres) or TB47 (green spheres) prevents substrate access (gray spheres).

Conclusions

We have determined the apo- and Q203 and Tb47-bound structures of a hybrid pathogenic M. tuberculosis/M. smegmatis cytochrome bcc complex. The study shows the structural features of M. tuberculosis cytochrome bcc and how it is specifically inhibited by Q203 and TB47. The extensive interactions between Q203 or TB47 and the Qp-binding pocket account for the highly specific binding of these two inhibitors to pathogenic M. tuberculosis cytochrome bcc compared to eukaryotic counterparts. Two conservative residues involved with the formation of hydrogen bonds are observed across the pathogenic mycobacteria. These structures provide a long-sought basis for rational, structure-based inhibitor design to accelerate the development of Q203 and TB47 analogs as drug leads for mycobacterial infections.

Materials and methods

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Gene (Mycobacterium tuberculosis)qcrCMycobrowserRv2194https://mycobrowser.epfl.ch/genes/Rv2194
Gene (M. tuberculosis)qcrAMycobrowserRv2195https://mycobrowser.epfl.ch/genes/Rv2195
Gene (M. tuberculosis)qcrBMycobrowserRv2196https://mycobrowser.epfl.ch/genes/Rv2196
Strain, strain background (Mycobacterium smegmatis)mc2 51Li et al., 2014
Genetic reagent (include species here)pVV16-QcrCAB-His10This paperConstruct contains the qcrCAB operon encoding three subunits
Commercial assay or kitClonExpress II One Step Cloning KitVazymeC112-01
Chemical compound, drugQ203MCEHY-101040Prepare stock solution in DMSO
Chemical compound, drugTB47Lu et al., 2019Prepare stock solution in DMSO
Chemical compound, drugLMNGAnatraceNG310
Chemical compound, drugDigitoninBIOSYNTHD-3200
Software, algorithmSerialEMMastronarde, 2003Version 3.6
Software, algorithmMotionCor2Zheng et al., 2017Version 1.2.1
Software, algorithmRELIONZivanov et al., 2019Version 3.03
Software, algorithmcryoSPARCPunjani et al., 2017Version 3.2.0
Software, algorithmPhyre2Kelley et al., 2015Version 2.0
Software, algorithmUCSF ChimeraPettersen et al., 2004Version 1.12
Software, algorithmCOOTEmsley et al., 2010Version 0.8.9
Software, algorithmPHENIXAdams et al., 2010Version 1.16
Software, algorithmPyMOLSchrödinger LLC, 2017Version 2.0
Software, algorithmChemDrawLi et al., 2004Version 19.0

Expression of hybrid supercomplex consisting of M. tuberculosis CIII and M. smegmatis CIV

Request a detailed protocol

The hybrid supercomplex was obtained according to a previous study (Kim et al., 2015) but with some modifications. The M. tuberculosis cytochrome bcc complex is encoded by three putative genes (Rv2194-2196). Genes were amplified from H37Rv genomic DNA by PCR using Phanta Max DNA polymerase (Vazyme), and two-step PCR was used to inset a 10× His tag at the C-terminus of the qcrB (Rv2196). Genes encoding the entire cytochrome bcc complex operon were then cloned into the pVV16 expression vector. The resultant plasmid was transformed into M. smegmatis mc2 51 (Li et al., 2014) cells whose qcrCAB operon encoding M. smegmatis cytochrome bcc had already been knocked out. The cells were cultivated in Luria-Bertani broth (LB) liquid media supplemented with 50 μg/mL hygromycin, 25 μg/mL streptomycin, and 0.1% Tween 80. Cell pellets were harvested by centrifugation when the cells were grown to an optical density (OD600 of 1.0–1.2) at 37°C (220 rpm). Harvested cells were frozen at –80°C until use.

Purification of the hybrid supercomplex

Request a detailed protocol

Cell pellets were thawed and resuspended in buffer A containing 20 mM 3-(N-morpholino)propanesulfonic acid (MOPS), pH 7.4, 100 mM NaCl, and then lysed by passing through a French Press at 1200 bar three times. Cell debris and non-lysed cells were removed by centrifugation at 14,000 rpm for 10 min at 4°C. The supernatant was collected and ultra-centrifuged at 36,900 rpm and 4°C for 2 hr. The membrane fraction was solubilized by addition of 1% (w/v) lauryl maltose neopentyl glycol (LMNG) in buffer A and incubated for 2 hr at 4°C with slow stirring. The suspension was ultra-centrifuged, and the supernatant was applied to Ni-NTA agarose beads (GE Healthcare) at 4°C. The beads were further washed in buffer A with 50 mM imidazole and 0.004% (w/v) LMNG. The buffer was exchanged to buffer B (20 mM MOPS, pH 7.4, 100 mM NaCl, and 0.1% [w/v] digitonin) and then washed in resin in batch mode. The protein was eluted from the beads with buffer B containing 500 mM imidazole. Protein was then concentrated and loaded onto a Superdex 6 increase (10/300GL, GE Healthcare) column equilibrated in buffer B. Peak fractions were pooled and concentrated to ~8 mg/mL for electron microscopy studies. The protein sample was analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and the bands were then identified through mass spectrometry.

Activity assays

Request a detailed protocol

2,3-Dimethyl-1,4-naphthoquinone (DMNQ, CAS 2197-57-1) was synthesized by WuXi AppTec. In order to obtained reduced DMNQH2, 20 mM DMNQ was ultrasonically dissolved in 1 mL ethanol with 6 mM HCl and reduced with sodium borohydride (NaBH4) in the ice bath. An appropriate amount of 12 N HCl was added into quench unreacted NaBH4 under the protection of argon. Oxidase activity was determined by a method described previously (Bonner et al., 1986; Kusumoto et al., 2000; Safarian et al., 2019). The oxygen consumption was monitored using the Clark-type oxygen electrode (Oxytherm+, Hansatech) in 1 mL reaction buffer (20 mM MOPS, pH 7.4, 100 mM NaCl, 0.004% LMNG) at 25℃, containing 62.5 nM hybrid supercomplex. The reaction was started by addition of 100 μM DMNQH2. The oxygen consumption curve was plotted using GraphPad Prime 8.0 software, from which an estimate of the oxygen-reduction rate (v0) was obtained (corrected for autoxidation). In the inhibition assay, 500 nM Q203 or TB47, which is approximately the value of the median inhibitory concentration (IC50) according to our previous study (Gong et al., 2018), was chosen and incubated with 62.5 nM hybrid supercomplex for 20 min at 25℃. Inhibition curves and oxygen-reduction rates (vi) with the different inhibitors were recorded. The inhibition rate (1 v0/vi) of Q203 and TB47 at 500 nM is reported. This assay was conducted using four groups of parallel experiments.

Cryo sample preparation and data collection

Request a detailed protocol

300-mesh Quantifoil R0.6/1.0 grids (Quantifoil, Micro Tools GmbH, Germany) were glow-discharged at H2/O2 atmosphere for 25 s. 3 μL aliquots of protein complex at a concentration of 10 mg/mL were applied to the grid and then blotted for 3 s with force 0 at 8°C and 100% humidity using a Vitrobot IV (Thermo). Images were collected using a Titan Krios 300 keV electron microscope (Thermo), equipped with K3 Summit direct electron detector camera (Gatan). Data were recorded at 29,000× magnification with a calibrated super-resolution pixel size 0.82 Å/pixel. The exposure time was set to 2.4 s with 40 subframes and a total dose of 60 electrons per Å2. All images were automatically recorded using SerialEM with a defocus range from 1.2 μm to 1.8 μm (Mastronarde, 2003). For the datasets of apo, Q203-bound and TB47-bound M. tuberculosis cytochrome bcc, a total of 4141, 3763, and 2968 images were collected, respectively.

Image processing

Request a detailed protocol

All dose-fractioned stacks were motion-corrected and dose-weighted using MotionCorr2 (Zheng et al., 2017) in RELION 3.03 (Zivanov et al., 2019). CTF estimation was conducted using cryoSPARC patch CTF estimation (Punjani et al., 2017). For the dataset of apo hybrid M. tuberculosis cytochrome bcc, 1,208,054 particles were picked automatically using EMD-9610 map as the template and extracted with a box size of rescaled 256 pixels (binned 2). 327,188 particles were selected after two rounds of 2D classification. 100,000 particles were used to perform ab initio reconstruction in four classes, and these four classes were used as 3D volume templates for heterogeneous refinement with all selected particles. 112,804 particles converged into one class with clear signals and then re-extracted with 512 pixels (binned 1). Next, this particle set was used to do homogeneous refinement and local refinement, yielding the final resolution 2.68 Å. For the dataset of Q203-bound and TB47-bound M. tuberculosis cytochrome bcc, the data processing was performed in a similar pipeline, resulting in the final reconstruction resolution at 2.67 Å and 2.93 Å, respectively (detailed parameters shown in supplementary figures).

Model building and validation

Request a detailed protocol

The M. smegmatis respiratory complex CIII2CIV2 (PDB code: 6ADQ) model (Gong et al., 2018) as rigid body was fitted into EM density maps using UCSF Chimera 1.12 (Pettersen et al., 2004). Next, the resultant atomic model was manually modified according to the subunit sequences of M. tuberculosis cytochrome bcc and refined in COOT 0.8.9.1 (Emsley et al., 2010), followed by real-space refinement in PHENIX (Adams et al., 2010). The smile strings of Q203 and TB47 were generated and copied from ChemDraw (Li et al., 2004) and defined in PHENIX elBOW. Q203 and TB47 were manually built into the corresponding EM densities. The local resolution map was calculated in cryoSPARC (Punjani et al., 2017). All reported resolutions were based on the gold-standard FSC 0.143 criteria (Rosenthal and Henderson, 2003).

MK9 docking study

Request a detailed protocol

The binding pose of MK9 with QcrB was generated by molecular docking using the Glide (Schrödinger, LLC). The spatially neighboring subunits QcrA/B were extracted from the hybrid supercomplex and used as the model. The protein structure was processed using the Protein Preparation Wizard module. In this process, hydrogens were added to heavy atoms and bond orders were assigned to each bond. The protonation state of each amino acid was predicted at pH 7.0 using the Epik algorithm (Shelley et al., 2007). All resolved waters were removed from the structure. At the same time, the structure of MK9 was prepared using the LigPrep module, and a low-energy conformation of MK9 was generated. The binding pocket of Q203 in QcrB was defined as the docking site using the Receptor Grid Generation module. The docking site was confined to an enclosed cube with side length of 20 Å, which was centered on the centroid of Q203. MK9 was then docked using the Ligand Docking module. Finally, 10 docking poses were generated for MK9 after post-docking minimization based on the force field of OPLS3 (Harder et al., 2016). These poses were ranked by the scoring function of GlideScore. The pose with the lowest GlideScore was chosen for binding mode analysis.

Binding free energies of Q203 between wild-type and mutant hybrid supercomplex

Request a detailed protocol

Binding free energies of Q203 between wild-type and mutant hybrid supercomplex were computed using the Molecular Mechanics/Poisson-Boltzmann Surface Area (MM/PBSA) method based on all-atom molecular dynamic simulations (Greene et al., 2019; Homeyer and Gohlke, 2012; Homeyer and Gohlke, 2015). This task can be divided into three phases: (i) simulation system construction, (ii) molecular dynamic simulation, and (iii) binding free energy computation.

Considering that QcrB is the core subunit interacting with Q203 in the hybrid supercomplex, the QcrB with Q203 bound was the only region extracted from the Q203-bound hybrid supercomplex for simulation system construction. Membrane Builder in CHARMM-GUI (Jo et al., 2007) was used to build the protein/membrane system solvated in water. PDB file containing the coordination information of Q203-bound QcrB was read while only protein residues and Q203 were retained. The CHARMM General Force Field (CGenFF) was applied to parameterize the Q203 (Vanommeslaeghe and MacKerell, 2012). DPPC lipid molecules were used to explicitly build the lipid bilayers and bulky water layers were added to the top and bottom sides of the membrane. In addition, 0.15 M KCl was added to neutralize the charge of system using the distance-based ion placing method. Next, the protein was inserted in the lipid bilayers and covered by the water layers despite exposure of some residues to the solvents. In order to generate topology and parameter files for subsequent AMBER dynamic simulations, the force fields of ff19SB (Tian et al., 2020) and Lipids17 (Lee et al., 2020a) were respectively applied to the parameterization of protein and lipid molecules. Four Q203-bound QcrB simulation systems were built, including wild-type QcrB, T313A mutant, E314Y mutant, and T313A/E314Y mutant.

Molecular dynamic simulation was carried out using AMBER 2020 (Case et al., 2005). Before dynamic simulation, 5000 steps of energy minimization were performed on waters to remove potential steric clashes between the solute and solvents. Harmonic restraints with a force constant of 10 kcal/mol Å2 and 2.5 kcal/mol Å2 were placed on protein and lipids in the minimization, respectively. Then, two stages of equilibrium simulations were carried out to relax the protein and lipid molecules by gradually decreasing the restraint force. The first stage was further divided into three short simulations each taking 375 ps with the time step set to 1 fs. The constant volume and temperature (NVT) ensemble was applied to this stage of simulation. The second one was a much longer simulation with the constant pressure and temperature (NPT) ensemble, which took 1.5 ns to completely relax the whole system with extremely low force. At this stage, the time step was increased to 2 fs while the SHAKE algorithm was used to fix the bonds with hydrogens (Bailey and Lowe, 2009). Subsequently, a 10 ns NPT simulation without any restraints was performed to further relax the system. Langevin dynamics was used to control the temperature at 303.15 K in NVT and NPT simulations (Grønbech-Jensen and Farago, 2014). In the NPT simulations, additional semi-isotropic pressure coupling and constant surface tension were applied to the simulation of lipid bilayers (Bennun et al., 2007). Once the equilibrium simulation was finished, a 100 ns NPT simulation free of restraints was carried out for each molecular system to produce the final trajectory used for energy calculation. Coordinates were printed every 100 ps so that in total 1000 frames were contained in the final trajectory. The CPPTRAJ module in AMBER was applied to analyze the trajectory (Roe and Cheatham, 2013).

MM/PBSA methods were employed to calculate the binding free energy of Q203 in wild-type and mutant QcrB (Homeyer and Gohlke, 2012). As shown in the equation (ΔG= EMM + Epol + Enp - TS), free energy can be calculated by combining the MM part (EMM) that represents the gas phase energy contribution as well as the solvation free energy components including polar (Epol) and non-polar (Enp) contributions. The gas phase free energy was calculated according to the force field. As for the solvation free energy, the polar part representing electrostatic contribution was calculated using the Poisson–Boltzmann (PB) equation based on implicit solvent model (Honig and Nicholls, 1995). In AMBER, the equation also provides an implicit membrane model to calculate the membrane-mediated electrostatic interactions. Here, geometric multigrid based on iterative solver was selected for PB equation calculation (Harris et al., 2013). On the other hand, the LCPO method was applied to calculate the non-polar part that represents the hydrophobic contribution (Weiser et al., 1999). The Python script MMPBSA.py (Miller et al., 2012) in AMBER was used for the computation based on previously produced simulation trajectories, from which one frame was extracted every 1 ns for the energy calculations. Ions, water, and lipid molecules were all striped from the molecular systems before calculation. For implicit membrane model, the membrane dielectric constant was set to 7.0, and membrane thickness was set to 42 Å. Entropy contributions (TS) could be neglected as the four molecular systems were assumed to have similar entropy changes due to the same protein-ligand structures except for the mutated residues.

Creation of figures

Request a detailed protocol

All the figures were created using UCSF Chimera (Pettersen et al., 2004) or PyMOL (Schrödinger LLC, 2017).

Data availability

All data generated or analysed during this study are included in the manuscript and supporting files. Source data files have been provided. The accession numbers for the 3D cryo-EM density map of apo, Q203-bound and TB47-bound hybrid supercomplex in present study are EMD-30943, EMD-30944 and EMD-30945, respectively. The accession numbers for the coordinates for the apo, Q203-bound and TB47-bound hybrid supercomplex in this study are PDB: 7E1V, PDB: 7E1W and PDB: 7E1X, respectively.

The following data sets were generated
    1. Zhou S
    2. Wang W
    3. Gao Y
    4. Gong H
    5. Rao Z
    (2021) Electron Microscopy Data Bank
    ID EMD-30943. Cryo-EM structure of apo hybrid respiratory supercomplex consisting of Mycobacterium tuberculosis complexIII and Mycobacterium smegmatis complexIV.
    1. Zhou S
    2. Wang W
    3. Gao Y
    4. Gong H
    5. Rao Z
    (2021) Electron Microscopy Data Bank
    ID EMD-30944. Cryo-EM structure of hybrid respiratory supercomplex consisting of Mycobacterium tuberculosis complexIII and Mycobacterium smegmatis complexIV in the presence of Q203.
    1. Zhou S
    2. Wang W
    3. Gao Y
    4. Gong H
    5. Rao Z
    (2021) Electron Microscopy Data Bank
    ID EMD-30945. Cryo-EM structure of hybrid respiratory supercomplex consisting of Mycobacterium tuberculosis complexIII and Mycobacterium smegmatis complexIV in presence of TB47.
    1. Zhou S
    2. Wang W
    3. Gao Y
    4. Gong H
    5. Rao Z
    (2021) RCSB Protein Data Bank
    ID 7E1V. Cryo-EM structure of apo hybrid respiratory supercomplex consisting of Mycobacterium tuberculosis complexIII and Mycobacterium smegmatis complexIV.
    1. Zhou S
    2. Wang W
    3. Gao Y
    4. Gong H
    5. Rao Z
    (2021) RCSB Protein Data Bank
    ID 7E1W. Cryo-EM structure of hybrid respiratory supercomplex consisting of Mycobacterium tuberculosis complexIII and Mycobacterium smegmatis complexIV in the presence of Q203.
    1. Zhou S
    2. Wang W
    3. Gao Y
    4. Gong H
    5. Rao Z
    (2021) RCSB Protein Data Bank
    ID 7E1X. Cryo-EM structure of hybrid respiratory supercomplex consisting of Mycobacterium tuberculosis complexIII and Mycobacterium smegmatis complexIV in presence of TB47.

References

    1. Emsley P
    2. Lohkamp B
    3. Scott WG
    4. Cowtan K
    (2010) Features and development of Coot
    Acta Crystallographica. Section D, Biological Crystallography 66:486–501.
    https://doi.org/10.1107/s0907444910007493
  1. Software
    1. Schrödinger LLC
    (2017)
    The PyMOL molecular graphics system, version 2.0
    Schrödinger LLC.
  2. Software
    1. World Health Organization
    (2019)
    Global Tuberculosis Report
    WHO.

Decision letter

  1. David Drew
    Reviewing Editor; Stockholm University, Sweden
  2. Volker Dötsch
    Senior Editor; Goethe University, Germany
  3. David Drew
    Reviewer; Stockholm University, Sweden

Our editorial process produces two outputs: i) public reviews designed to be posted alongside the preprint for the benefit of readers; ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "Structure of mycobacterial cytochrome bcc in complex with Q203 and TB47, two anti-TB drug candidates" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including David Drew as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Michael Marletta as the Senior Editor.

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

The main impact of this paper is the drug interaction. We think it is essential that both a mechanism of inhibition and selectivity for the Q203/TB47 drugs are addressed, i.e, an atomic model on its own is insufficient to reach these conclusions.

1) The authors need to demonstrate activity of the chimeric supercomplex (from oxygen consumption experiments or any other relevant activity assay) and validate that Q203 and TB47 inhibit as expected.

2) The affinity of Q203 (i.e., with ITC) in combination with mutagenesis and or computational approaches are required to elucidate the molecular basis how the selectivity of the drug compound to M. tuberculous is accomplished.

3) We require a deeper discussion regarding the super complex of the hybrid and differences in complex III in M. tuberculous compared to M. smegmatis to evaluate if the structure yields any new mechanistic insights.

Reviewer #1 Recommendations for the Authors:

1. A rigorous comparison, using mutagenesis and binding and functional assays and/or computational methods is essential to provide a meaningful model for establishing how drug specificity for M. tuberculosis cytochrome bcc is achieved by these compounds.

2. An MK molecule was identified, but it was located some 15 Å from the heme bL and is thought that the actual binding site for MK would be where the drugs are binding. What is the relevance for the bound MK molecule? Why was it not possible to model menaquinol in the proposed location and has this quinone site been structurally shown previously for other bcc complexes? Do the authors have a reason for this and why do quinone binding sites show species-specific differences? Why is it not possible to directly measure competition between the drugs and menaquinol?

3. The authors have not elaborated on any structural differences of previous cytochrome bcc complexes and how this might br relevant for understanding electron-driven proton transport in M. tuberculosis. Indeed, the Mtb cytochrome bcc is complex with cytochrome c oxidase from a different bacterial host M. smegmatis. What was the practical reason for expressing as a hybrid complex? How does the hybrid complex compare to the supercomplex from M. smegmatis? The paper would clearly benefit from more functional comparisons.

Reviewer #2 Recommendations for the Authors:

I find this work interesting, but certain shortcomings should be addressed before I can make a final recommendation.

1. Very little is said about the complex IV part of the resolved supercomplex. The authors should clearly explain how the hybrid supercomplex is affecting the structure, including possible artifacts arising from non-native contacts.

2. What is the activity of this chimeric supercomplex? What is the binding affinity of Q203 and TB47 and do they actually inhibit the activity?

Reviewer #3 Recommendations for the Authors:

1. Was the purified enzyme active? Please provide data from oxygen consumption experiments or any other relevant activity assay.

2.. To which degree was the purified enzyme susceptible to inhibition by Q203 and TB47?

3.. Lines 172 – 190 and Figure 6: the comparison of bcc/bc1 structures provides information regarding why Q203 does not inhibit homologies complexes from non-mycobacterial species. What do these structure data imply for M. smegmatis? This structure is shown in Figure 6 without steric clashes indicated, but is not referred to in main text. How do the structural data regarding Q203 binding to M. smegmatis bcc correlate to the sensitivity of this model strain for Q203?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Structure of mycobacterial cytochrome bcc in complex with Q203 and TB47, two anti-TB drug candidates" for further consideration by eLife. Your revised article has been reviewed by 2 peer reviewers and the evaluation has been overseen by Cynthia Wolberger as the Senior Editor and a Reviewing Editor.

The manuscript has been improved but there are a few remaining issues that need to be addressed, as outlined below.

Summary:

The authors describe the cryo EM structures of cytochrome bcc from mycobacteria in complex with clinical drug candidates Q203 and TB47, which are being trialled for the treatment of tuberculosis. The cryo EM structures were obtained by expressing and purifying a hybrid super complex; the bcc part from M. tuberculosis and the cytochrome oxidase part from M. smegmatis. The cryo EM structures show how the Q203 and TB47 inhibitors bind and block the binding of quinone at the Qo site, thus inhibiting electron-transfer. The complex structures should facilitate the structure-based development of Q203 and TB47 drug compounds against tuberculosis.

1. It would be helpful to give the timescale in the SI figure 6-suppl 2 rather than a frame number, which is cryptic for the general reader. The authors should also better describe what the different traces mean, for example what is the implication of the high RMSD for R313A+E314Y in panel B?

2. The binding free energies presented in Table 4 seem unphysically large (around -65 kcal/mol, which would result in binding constant of 10-(65/RT)), which could be misleading. Reporting relative values might be more appropriate here.

Reviewer #2 (Recommendations for the authors):

The authors have adequately addressed most of my comments, and this has resulted in a significantly improved manuscript. Importantly, the authors provide in the revised manuscript activity measurements, which show that the WT complex is active, and that the inhibitors significantly reduce the activity. The addition of missing residues in the figures, and explicitly showing the densities for selected residues and ligands have also improved the discussion. The authors could add a comment on what the biological function of binding so many MK molecules to the supercomplex might be. Although no experimental estimates of the binding affinities could be provided, the authors now present some computational simulations to estimate the binding effects. In this regard, it would be helpful to give the timescale in the SI figure 6-suppl 2 rather than a frame number, which is cryptic for the general reader. The authors should also better describe what the different traces mean, for example what is the implication of the high RMSD for R313A+E314Y in panel B. Also, the binding free energies presented in Table 4 seem unphysically large (around -65 kcal/mol, which would result in binding constant of 10^-(65/RT)!), which could be misleading. As far as I understand, the method estimates only relative binding affinities based on an approximate method. Reporting relative values might be more appropriate.

Reviewer #3 (Recommendations for the authors):

The authors have addressed my questions. I recommend accepting this paper.

https://doi.org/10.7554/eLife.69418.sa1

Author response

Essential revisions:

The main impact of this paper is the drug interaction. We think it is essential that both a mechanism of inhibition and selectivity for the Q203/TB47 drugs are addressed, i.e, an atomic model on its own is insufficient to reach these conclusions.

1) The authors need to demonstrate activity of the chimeric supercomplex (from oxygen consumption experiments or any other relevant activity assay) and validate that Q203 and TB47 inhibit as expected.

We have now shown that the purified chimeric supercomplex is a functional assembly with a turnover number of 23.3 +/- 2.4 e-s-1 (mean ± s.d., n = 4), in agreement with the previous study that shows M. tuberculosis CIII can functionally complement native M. smegmatis CIII and maintain the growth of M. smegmatis (Kim et al., 2015). The in vitro inhibitions of this enzyme by Q203 and TB47 was determined by means of an DMNQH2/oxygen oxidoreductase activity assay. In the assay, 500 nM Q203 or TB47 was chosen, which is close to the median inhibitory concentration (IC50) obtained from the menadiol-induced oxygen consumption in our previous study (Gong et al., 2018). After addition of Q203 and TB47, the values of turnover number of the hybrid supercomplex are reduced to 5.8 +/- 2.4 e-s-1 (Figure 4—figure supplement 4) and 5.1 +/- 2.9 e-s-1 (Figure 5—figure supplement 4) respectively, from 23.3 +/- 2.4 e-s-1. We have incorporated this new data into the text (lines 90-93, 187-189, 206-209).

2) The affinity of Q203 (i.e., with ITC) in combination with mutagenesis and or computational approaches are required to elucidate the molecular basis how the selectivity of the drug compound to M. tuberculous is accomplished.

It is not possible to detect the binding of Q203 and TB47 using isothermal titration calorimetry (ITC) or microscale thermophoresis (MST). This is due to the complicated identity of the multiple subunits of membrane protein complex, the limited solubility of the drugs and the limited yield of the hybrid supercomplex. In addition, the purified complex is bound to endogenous substrate at the drug-binding site, which also indirectly affects the accurate measurement of the affinity of the drugs. Therefore, an alternative approach is needed to measure the true binding affinity of the Q203 in the future.

Here we have been able to calculate the binding free energies of Q203 between wild-type and mutant hybrid supercomplex. This data has been used to further elucidate the importance of some key residues that could play a role in determining the inhibitor specificity. According to a previous functional study (Pethe et al., 2013), structural comparison (Figure 6) and sequence alignment (Figure 6—figure supplement 1), QcrBThr313 and QcrBGlu314 of hybrid supercomplex were selected and mutated to their counterparts in the human supercomplex, namely residues Ala and Tyr, respectively. It is believed that Ala would decrease the affinity of Q203 (Figure 4) (Pethe et al., 2013) and the Tyr could lead to steric hindrance of Q203 binding (Figure 6). 75 ns simulations were used for the calculation of binding free energy (Figure 6—figure supplement 2). The 75 ns simulation was further equally divided into three phases and the average values were used as the final binding free energy (Table 4). It is suggested that the wild-type system has stronger affinity than the mutant systems (Thr313Ala or Glu314Tyr). The free energy of binding calculation for Q203 in the dual mutant (Thr313Ala and Glu314Tyr) indicates that its binding affinity is significantly weaker compared to the wild-type or single-mutant systems. Therefore, the suggestions drawn from the calculation of binding free energy of Q203 provide further evidence for specificity of Q203 inhibition. We have incorporated discussion into the text (line 234-248).

3) We require a deeper discussion regarding the super complex of the hybrid and differences in complex III in M. tuberculous compared to M. smegmatis to evaluate if the structure yields any new mechanistic insights.

The hybrid supercomplex including the M. tuberculosis CIII dimer forms a pseudo 2-fold symmetrical compact rod, but with a slight curvature in the membrane plane, as previously observed in the M. smegmatis CIII2CIV2 supercomplex (Gong et al., 2018; Wiseman et al., 2018) (Figure 2A, B). The hybrid supercomplex is structurally similar to the supercomplex isolated from M. smegmatis (Gong et al., 2018; Wiseman et al., 2018) (Figure 2—figure supplement 4), except that subunits LpqE and Unk. (probably MSMEG_0987) (Wiseman et al., 2018) were not observed here. The absence of these two subunits may be due to their depletion during purification of the supercomplex or their map density signal was averaged to background noise during structural determination. As expected, the topology of M. tuberculosis cytochrome bcc and M. smegmatis CIV in the hybrid supercomplex is also similar to that of the equivalent subunits in the M. smegmatis CIII2CIV2 supercomplex (Figure 2C, D; Figure 2-figures supplement 4, 5). As such, there is no notable non-native contacts that resulted from the hybrid assembly in the hybrid supercomplex. This is attributable to the high structural similarity between the M. tuberculosis cytochrome bcc and M. smegmatis cytochrome bcc. In the M. smegmatis CIII2CIV2 supercomplex, the cytochrome cc head domain of QcrC adopts an ‘open’ or a ‘closed’ conformation (Gong et al., 2018; Wiseman et al., 2018). However, in this structure it is only the closed conformation (Figure 2—figure supplement 4). Considering the high topology similarity of the subunits and the arrangement of prosthetic groups compared to previous M. smegmatis CIII2CIV2 supercomplex, the mechanism of action of the hybrid supercomplex including M. tuberculosis cytochrome bcc is expected to be the same as that for the M. smegmatis CIII2CIV2 supercomplex (Gong et al., 2018). These new discussion points have been incorporated into the revised manuscript (lines 96-108, 120-127).

Reviewer #1 Recommendations for the Authors:

1. A rigorous comparison, using mutagenesis and binding and functional assays and/or computational methods is essential to provide a meaningful model for establishing how drug specificity for M. tuberculosis cytochrome bcc is achieved by these compounds.

See answer to Question 2 from the major Essential Revisions.

2. An MK molecule was identified, but it was located some 15 Å from the heme bL and is thought that the actual binding site for MK would be where the drugs are binding. What is the relevance for the bound MK molecule? Why was it not possible to model menaquinol in the proposed location and has this quinone site been structurally shown previously for other bcc complexes? Do the authors have a reason for this and why do quinone binding sites show species-specific differences? Why is it not possible to directly measure competition between the drugs and menaquinol?

We have now rephrased the corresponding sentences and discussed in depth. The edge-to-edge distance from MK to heme bL is 15 Å, exceeding the 14 Å limit for efficient physiological electron transfer (Page et al., 1999). Furthermore, there are no observed hydrogen bonds to the carbonyl groups of MK, which are needed to help deprotonate MKH2. In addition, structural superposition with the inhibitor-bound bc1 complex also shows that the MK binds deeper into the QP pocket in the bc1 complex (Birth et al., 2014; Solmaz and Hunte, 2008) than in the currently described complex (Figure 3—figure supplement 2). Hence, the endogenous electron donor MKH2 should bind deeper inside the pocket and close to polar residues such as QcrBTyr161, QcrBThr313, and QcrBAsp314 (Figure 3D), to facilitate electron transfer. In addition, MK could also be modeled to fit deeper inside the pocket (Figure 3—figure supplement 3). Thus, we speculate that what is observed here is the oxidized product as it leaves the QP site.

Quinone-binding sites are essential to the function of respiratory chain complexes and thus are good targets for drug discovery (Harikishore et al., 2020; Lee et al., 2020a). Structurally diverse quinones such as ubiquinone and menaquinone bind in the mitochondrial ETC (Zhang et al., 1998) and mycobacterial ETC (Gong et al., 2018; Wiseman et al., 2018), respectively, indicating there are structural differences in the quinone-binding sites of the different species. Structural differences in the ubiquinone-binding site are also observed when the mitochondrial and Escherichia coli respiratory chain complex IIs are compared (Huang et al., 2021). Therefore, it is feasible that species-specific quinone-binding-site inhibitors could be designed. We have incorporated above discussions into the revised text (lines 129-136, 147-156).

It is not possible to detect the binding of the drugs (Q203 and TB47) and menaquinol (native substrate, MK9) using isothermal titration calorimetry (ITC) or microscale thermophoresis (MST). This is due to the complicated identity of the multiple subunits of membrane protein complex, the limited solubility of the drugs and the limited yield of the hybrid supercomplex. In addition, the purified complex is bound to endogenous substrate at the drug-binding site, which also indirectly affects the accurate measurement of the affinity of the drugs and menaquinol. An alternative approach is needed to measure the true binding affinity of the drugs and menaquinol in the future.

3. The authors have not elaborated on any structural differences of previous cytochrome bcc complexes and how this might br relevant for understanding electron-driven proton transport in M. tuberculosis. Indeed, the Mtb cytochrome bcc is complex with cytochrome c oxidase from a different bacterial host M. smegmatis. What was the practical reason for expressing as a hybrid complex? How does the hybrid complex compare to the supercomplex from M. smegmatis? The paper would clearly benefit from more functional comparisons.

Information on the structural features and mechanism of action of hybrid complex including M. tuberculosis cytochrome bcc has been discussed above (See answer to Question 3 from the major Essential Revisions).

The reasons for expressing a hybrid complex are as follows. On the one hand, it is difficult to successfully express multiple subunits of a protein complex like mycobacterial CIII2CIV2 respiratory chain supercomplex, especially when the genes coding subunits of the complex are not clustered together. On the other hand, considering that the hybrid supercomplex has been confirmed as a stable and functional assembly and its acquisition can be operated practically (Kim et al., 2015), we therefore expressed and purified the hybrid supercomplex, and finally determined its structure.

Reviewer #2 Recommendations for the Authors:

I find this work interesting, but certain shortcomings should be addressed before I can make a final recommendation.

1. Very little is said about the complex IV part of the resolved supercomplex. The authors should clearly explain how the hybrid supercomplex is affecting the structure, including possible artifacts arising from non-native contacts.

See answer to Question 3 from the major Essential Revisions.

2. What is the activity of this chimeric supercomplex? What is the binding affinity of Q203 and TB47 and do they actually inhibit the activity?

See answer to Questions 1 and 2 from the major Essential Revisions.

Reviewer #3 Recommendations for the Authors:

1. Was the purified enzyme active? Please provide data from oxygen consumption experiments or any other relevant activity assay.

See answer to Question 1 from the major Essential Revisions.

2. To which degree was the purified enzyme susceptible to inhibition by Q203 and TB47?

See answer to Question 1 from the major Essential Revisions.

3. Lines 172 – 190 and Figure 6: the comparison of bcc/bc1 structures provides information regarding why Q203 does not inhibit homologies complexes from non-mycobacterial species. What do these structure data imply for M. smegmatis? This structure is shown in Figure 6 without steric clashes indicated, but is not referred to in main text. How do the structural data regarding Q203 binding to M. smegmatis bcc correlate to the sensitivity of this model strain for Q203?

It is worth noting that the structures of cytochrome bcc from M. tuberculosis and M. smegmatis have high similarity (Figure 2-figures supplement 5), and no steric hindrance is observed between the Q203 and M. smegmatis cytochrome bcc (Figure 6). This observation indicates that Q203 should have a similar binding mechanism and a similar effect on the activity of cytochrome bcc from M. smegmatis and M. tuberculosis. This is in good agreement with previous antimycobacterial activity data and inhibition data for the bcc complexes from M. smegmatis and M. tuberculosis (Gong et al., 2018; Lu et al., 2018a). This discussion has now been added into the text (line 220-227).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved but there are a few remaining issues that need to be addressed, as outlined below.

Summary:

The authors describe the cryo EM structures of cytochrome bcc from mycobacteria in complex with clinical drug candidates Q203 and TB47, which are being trialled for the treatment of tuberculosis. The cryo EM structures were obtained by expressing and purifying a hybrid super complex; the bcc part from M. tuberculosis and the cytochrome oxidase part from M. smegmatis. The cryo EM structures show how the Q203 and TB47 inhibitors bind and block the binding of quinone at the Qo site, thus inhibiting electron-transfer. The complex structures should facilitate the structure-based development of Q203 and TB47 drug compounds against tuberculosis.

1. It would be helpful to give the timescale in the SI figure 6-suppl 2 rather than a frame number, which is cryptic for the general reader. The authors should also better describe what the different traces mean, for example what is the implication of the high RMSD for R313A+E314Y in panel B?

We have provided a new SI figure 6-suppl 2 with the timescale as the X-axis. The different values for the root mean squared deviation (RMSD) suggest that three mutants might have the different effects on the structural conformation of QcrB, and the dual mutant (Thr313Ala and Glu314Tyr) induced significantly structural changes compared to the wild-type complex. New discussion points have been incorporated into the revised manuscript (line 244-247).

2. The binding free energies presented in Table 4 seem unphysically large (around -65 kcal/mol, which would result in binding constant of 10^-(65/RT)), which could be misleading. Reporting relative values might be more appropriate here.

We thank the reviewer for this suggestion. We have reported the relative value for each mutant compared to the wild-type in the updated Table 4.

Reviewer #2 (Recommendations for the authors):

The authors have adequately addressed most of my comments, and this has resulted in a significantly improved manuscript. Importantly, the authors provide in the revised manuscript activity measurements, which show that the WT complex is active, and that the inhibitors significantly reduce the activity. The addition of missing residues in the figures, and explicitly showing the densities for selected residues and ligands have also improved the discussion.

The authors could add a comment on what the biological function of binding so many MK molecules to the supercomplex might be.

These quinones are observed to be remote from the canonical quinone-binding sites and into the membrane space, suggesting they have a structural rather than functional role. We have incorporated this discussion into the text (line 140-141).

Although no experimental estimates of the binding affinities could be provided, the authors now present some computational simulations to estimate the binding effects. In this regard, it would be helpful to give the timescale in the SI figure 6-suppl 2 rather than a frame number, which is cryptic for the general reader. The authors should also better describe what the different traces mean, for example what is the implication of the high RMSD for R313A+E314Y in panel B.

See answer to Question 1 from the Summary.

Also, the binding free energies presented in Table 4 seem unphysically large (around -65 kcal/mol, which would result in binding constant of 10^-(65/RT)!), which could be misleading. As far as I understand, the method estimates only relative binding affinities based on an approximate method. Reporting relative values might be more appropriate.

See answer to Question 2 from the Summary.

https://doi.org/10.7554/eLife.69418.sa2

Article and author information

Author details

  1. Shan Zhou

    1. State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy, Nankai University, Tianjin, China
    2. State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review and editing
    Contributed equally with
    Weiwei Wang
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-5468-3538
  2. Weiwei Wang

    Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, Shanghai, China
    Contribution
    Data curation, Formal analysis, Methodology, Visualization, Writing – review and editing
    Contributed equally with
    Shan Zhou
    Competing interests
    No competing interests declared
  3. Xiaoting Zhou

    Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, Shanghai, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8238-8242
  4. Yuying Zhang

    State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
  5. Yuezheng Lai

    State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
  6. Yanting Tang

    State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8656-3220
  7. Jinxu Xu

    State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
  8. Dongmei Li

    State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy, Nankai University, Tianjin, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
  9. Jianping Lin

    State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy, Nankai University, Tianjin, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
  10. Xiaolin Yang

    Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, Shanghai, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-0992-8676
  11. Ting Ran

    Bioland Laboratory (Guangzhou Regenerative Medicine and Health - Guangdong Laboratory), Guangzhou, China
    Contribution
    Data curation, Formal analysis, Methodology
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1387-4634
  12. Hongming Chen

    Bioland Laboratory (Guangzhou Regenerative Medicine and Health - Guangdong Laboratory), Guangzhou, China
    Contribution
    Data curation, Formal analysis, Methodology
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8065-8333
  13. Luke W Guddat

    School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, Australia
    Contribution
    Formal analysis, Methodology, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8204-8408
  14. Quan Wang

    Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, Shanghai, China
    Contribution
    Formal analysis, Methodology
    Competing interests
    No competing interests declared
  15. Yan Gao

    Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, Shanghai, China
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    For correspondence
    gaoyan@shanghaitech.edu.cn
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-9665-8645
  16. Zihe Rao

    1. State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy, Nankai University, Tianjin, China
    2. State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, Tianjin, China
    3. Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, Shanghai, China
    4. National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Beijing, China
    5. Laboratory of Structural Biology, Tsinghua University, Beijing, China
    Contribution
    Funding acquisition, Resources, Supervision, Writing – review and editing
    For correspondence
    raozh@mail.tsinghua.edu.cn
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-9866-2384
  17. Hongri Gong

    State Key Laboratory of Medicinal Chemical Biology, College of Life Sciences, Nankai University, Tianjin, China
    Contribution
    Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Visualization, Writing – original draft, Writing – review and editing
    For correspondence
    gonghr@nankai.edu.cn
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-2596-7635

Funding

National Key Research and Development Program of China (2017YFC0840300)

  • Zihe Rao

National Key Research and Development Program (2020YFA0707500)

  • Zihe Rao

National Natural Science Foundation of China (81520108019)

  • Zihe Rao

National Natural Science Foundation of China (813300237)

  • Zihe Rao

National Natural Science Foundation of China (32100976)

  • Hongri Gong

Natural Science Foundation of Tianjin City (20JCQNJC01430)

  • Hongri Gong

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

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. Lei Li (College of Life Sciences, Nankai University) for their technical support on Clark-type oxygen electrode and oxygen consumption assay. We thank Prof. Kaixia Mi (CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of Microbiology, CAS) for sharing the strain M. smegmatis mc2 51. We would like to thank Prof. Gregory M Cook (School of Biomedical Sciences, University of Otago, New Zealand) and Prof. Xiaoyun Lu (School of Pharmacy, Ji′nan University, China) for kindly providing TB47 for this study. We would also like to thank the Bio-Electron Microscopy Facility of ShanghaiTech University.

Senior Editor

  1. Volker Dötsch, Goethe University, Germany

Reviewing Editor

  1. David Drew, Stockholm University, Sweden

Reviewer

  1. David Drew, Stockholm University, Sweden

Publication history

  1. Received: April 14, 2021
  2. Preprint posted: June 16, 2021 (view preprint)
  3. Accepted: November 5, 2021
  4. Version of Record published: November 25, 2021 (version 1)

Copyright

© 2021, Zhou et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 89
    Page views
  • 12
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Immunology and Inflammation
    2. Structural Biology and Molecular Biophysics
    Mohamed A Badawy et al.
    Research Article

    Human serum albumin (HSA) is the frontline antioxidant protein in blood with established anti-inflammatory and anticoagulation functions. Here we report that COVID-19-induced oxidative stress inflicts structural damages to HSA and is linked with mortality outcome in critically ill patients. We recruited 39 patients who were followed up for a median of 12.5 days (1-35 days), among them 23 had died. Analyzing blood samples from patients and healthy individuals (n=11), we provide evidence that neutrophils are major sources of oxidative stress in blood and that hydrogen peroxide is highly accumulated in plasmas of non-survivors. We then analyzed electron paramagnetic resonance (EPR) spectra of spin labelled fatty acids (SLFA) bound with HSA in whole blood of control, survivor, and non-survivor subjects (n=10-11). Non-survivor' HSA showed dramatically reduced protein packing order parameter, faster SLFA correlational rotational time, and smaller S/W ratio (strong-binding/weak-binding sites within HSA), all reflecting remarkably fluid protein microenvironments. Following loading/unloading of 16-DSA we show that transport function of HSA maybe impaired in severe patients. Stratified at the means, Kaplan–Meier survival analysis indicated that lower values of S/W ratio and accumulated H2O2 in plasma significantly predicted in-hospital mortality (S/W≤0.15, 81.8% (18/22) vs. S/W>0.15, 18.2% (4/22), p=0.023; plasma [H2O2]>8.6 mM, 65.2% (15/23) vs. 34.8% (8/23), p=0.043). When we combined these two parameters as the ratio ((S/W)/[H2O2]) to derive a risk score, the resultant risk score lower than the mean (< 0.019) predicted mortality with high fidelity (95.5% (21/22) vs. 4.5% (1/22), logrank c2 = 12.1, p=4.9x10-4). The derived parameters may provide a surrogate marker to assess new candidates for COVID-19 treatments targeting HSA replacements and/or oxidative stress.

    1. Cell Biology
    2. Structural Biology and Molecular Biophysics
    Carolina Franco Nitta et al.
    Research Article

    Crosstalk between different receptor tyrosine kinases (RTKs) is thought to drive oncogenic signaling and allow therapeutic escape. EGFR and RON are two such RTKs from different subfamilies, which engage in crosstalk through unknown mechanisms. We combined high-resolution imaging with biochemical and mutational studies to ask how EGFR and RON communicate. EGF stimulation promotes EGFR-dependent phosphorylation of RON, but ligand stimulation of RON does not trigger EGFR phosphorylation – arguing that crosstalk is unidirectional. Nanoscale imaging reveals association of EGFR and RON in common plasma membrane microdomains. Two-color single particle tracking captured formation of complexes between RON and EGF-bound EGFR. Our results further show that RON is a substrate for EGFR kinase, and that transactivation of RON requires formation of a signaling competent EGFR dimer. These results support a role for direct EGFR/RON interactions in propagating crosstalk, such that EGF-stimulated EGFR phosphorylates RON to activate RON-directed signaling.