Crystal Structures of Wolbachia CidA and CidB Reveal Determinants of Bacteria-induced Cytoplasmic Incompatibility and Rescue

Wang, Haofeng; Xiao, Yunjie; Chen, Xia; Zhang, Mengwen; Sun, Guangxin; Wang, Feng; Wang, Lin; Zhang, Hanxiao; Zhang, Xiaoyu; Yang, Xin; Li, Wenling; Wei, Yi; Yao, Deqiang; Zhang, Bing; Li, Jun; Cui, Wen; Wang, Fenghua; Chen, Cheng; Shen, Wei; Su, Dan; Bai, Fang; Huang, Jinhai; Ye, Sheng; Zhang, Lei; Ji, Xiaoyun; Wang, Wei; Wang, Zefang; Hochstrasser, Mark; Yang, Haitao

doi:10.1038/s41467-022-29273-w

Download PDF

Article
Open access
Published: 25 March 2022

Crystal Structures of Wolbachia CidA and CidB Reveal Determinants of Bacteria-induced Cytoplasmic Incompatibility and Rescue

Haofeng Wang ORCID: orcid.org/0000-0001-8175-5621^1,2,3^na1,
Yunjie Xiao^1,2,3,4^na1,
Xia Chen¹^na1,
Mengwen Zhang ORCID: orcid.org/0000-0003-2703-211X^5,6,
Guangxin Sun⁵,
Feng Wang¹,
Lin Wang ORCID: orcid.org/0000-0003-2482-7638^2,3,
Hanxiao Zhang¹,
Xiaoyu Zhang^1,7,8,
Xin Yang⁴,
Wenling Li¹,
Yi Wei¹,
Deqiang Yao⁹,
Bing Zhang ORCID: orcid.org/0000-0001-8556-8049²,
Jun Li²,
Wen Cui^1,4,
Fenghua Wang¹,
Cheng Chen¹,
Wei Shen ORCID: orcid.org/0000-0002-3785-2245^2,3,
Dan Su¹⁰,
Fang Bai^2,3,
Jinhai Huang¹,
Sheng Ye¹,
Lei Zhang¹,
Xiaoyun Ji ORCID: orcid.org/0000-0002-0801-8825¹¹,
Wei Wang ORCID: orcid.org/0000-0002-2870-0733⁴,
Zefang Wang ORCID: orcid.org/0000-0002-0852-6483^1,8,
Mark Hochstrasser ORCID: orcid.org/0000-0002-1131-5484⁵ &
…
Haitao Yang ORCID: orcid.org/0000-0002-1875-3268^1,2,3,8

Nature Communications volume 13, Article number: 1608 (2022) Cite this article

4613 Accesses
10 Citations
3 Altmetric
Metrics details

Subjects

Abstract

Cytoplasmic incompatibility (CI) results when Wolbachia bacteria-infected male insects mate with uninfected females, leading to embryonic lethality. “Rescue” of viability occurs if the female harbors the same Wolbachia strain. CI is caused by linked pairs of Wolbachia genes called CI factors (CifA and CifB). The co-evolution of CifA-CifB pairs may account in part for the incompatibility patterns documented in insects infected with different Wolbachia strains, but the molecular mechanisms remain elusive. Here, we use X-ray crystallography and AlphaFold to analyze the CI factors from Wolbachia strain wMel called CidA^wMel and CidB^wMel. Substituting CidA^wMel interface residues with those from CidA^wPip (from strain wPip) enables the mutant protein to bind CidB^wPip and rescue CidB^wPip-induced yeast growth defects, supporting the importance of CifA-CifB interaction in CI rescue. Sequence divergence in CidA^wPip and CidB^wPip proteins affects their pairwise interactions, which may help explain the complex incompatibility patterns of mosquitoes infected with different wPip strains.

DNA glycosylases provide antiviral defence in prokaryotes

Article Open access 17 April 2024

Streptomyces umbrella toxin particles block hyphal growth of competing species

Article Open access 17 April 2024

DNA-targeting short Argonautes complex with effector proteins for collateral nuclease activity and bacterial population immunity

Article 15 April 2024

Introduction

Wolbachia pipientis is an intracellular bacterium infecting ~40% of all terrestrial arthropod species and certain filarial nematodes^1,2,3. They are primarily inherited maternally⁴. Their remarkable success in spreading in arthropod populations is in part driven by the ability of Wolbachia to increase the number of infected females⁵, often by acting as reproductive parasites that distort host sex ratios or reproductive outcomes⁶. The most common type of reproductive parasitism is cytoplasmic incompatibility (CI)⁶, which is a type of post-zygotic male sterility characterized by improper paternal chromatin condensation and separation in the first mitosis cycle^7,8,9. CI causes Wolbachia-infected male insects to be sterile in matings with uninfected females but not similarly infected females (called ‘rescue’), providing a selective advantage for infected females and thus driving Wolbachia into the host population^6,10. Insect species infected by two or more different Wolbachia strains can be bidirectional incompatible^11,12,13, which causes reproductive isolation of populations infected with different strains^11,14.

Based on the outcomes of crosses between Wolbachia-infected insects, it has been proposed that CI is governed by a modification-rescue system^15,16. Specifically, a modification activity from Wolbachia is postulated to modify sperm during spermatogenesis, and a rescue activity in the infected egg reverses or neutralizes the original sperm modification following fertilization. The existence of both unidirectional and bidirectional incompatibilities in infected insect populations has led to agreement that Wolbachia may carry multiple modification and rescue factors^12,13, but the molecular identity of these factors remained unknown for nearly half a century^17,18,19,20.

Recently, the key factors involved in CI induction and rescue were identified^21,22,23,24. They are known as CI factors or Cifs, which are encoded by pairs of linked genes, cifA and cifB. Transgenic expression of both cifA and cifB genes, or cifB alone, in the germline of male Drosophila melanogaster or Anopheles gambiae, induces sterility that is highly similar to CI induced by Wolbachia^{21,22,25,26,27}. This embryonic lethality can be rescued by crossing transgenic males with Wolbachia-infected females or those with a transgenic cifA gene expressed in the germline^22,23,24. The cif genes have diverged into at least five distinct phylogenetic groups, which are named types I-V^22,28,29,30. The degree of similarity and the presence/absence of cif gene homologs between Wolbachia strains correlates with known patterns of bidirectional incompatibility. For instance, Wolbachia strain wRi is able to rescue wMel-induced CI in same-species crosses, probably because these two strains share highly related type-I homologs of the cifA gene (99% amino acid identity); however, the reverse is not true: wMel cannot rescue wRi-induced CI, likely due to wRi also encoding a type II gene pair that is much more divergent^12,13,24. In general, the cif genes from bidirectionally incompatible Wolbachia pairs are highly divergent, with only 29–68% amino-acid identity²². Recent analysis of transgenic cif gene expression in D. melanogaster is largely consistent with this view³¹.

The cifA and cifB genes are coevolving^22,30,32. Protein-binding experiments and expression of cif genes in yeast have shown that CifA binds its cognate CifB specifically and rescues CifB-induced yeast growth defects^21,23. Although the specific roles of cifA and cifB in CI induction and rescue are still not fully understood, the co-divergence of the cifA and cifB genes^22,32,33 has been proposed to be partially responsible for bidirectional incompatibility, possibly by modulating CifA-CifB binding^34,35.

Among the five types of cif genes, only type I and certain type V cifB genes encode a deubiquitylase (DUB) domain²¹. (Hereafter, we will use “Cif” for the entire set of CI factors and “Cid” for the deubiquitylase family and primarily focus on the type I Cid protein family.) The cidA and cidB genes from Wolbachia strains wMel (hereafter cidA^wMel and cidB^wMel) and wPip (hereafter cidA^wPip and cidB^wPip) are among the most well studied cif genes (Fig. 1a). These Wolbachia strains are expected to be incompatible, as shown by crosses between trans-infected Aedes albopictus mosquitoes³⁶. Unlike the wMel strain, Wolbachia wPip strains found in different populations of Culex pipiens are highly diversified. Many of them carry several polymorphic copies of the cidA^wPip and cidB^wPip genes, which may function as independent modification and rescue factors and associate with the diverse CI phenotypes in wPip-infected C. pipiens^32,37. Thus, the Cid proteins can be used as a model to study how the diversification of cidA^wPip and cidB^wPip affects protein-protein interactions and how it is related to CI induction and rescue.

**Fig. 1: Crystal structures of CidA^wMel and CidB^wMel_DUB, and a model of CidB^wMel_ND1-ND2 reveal the molecular basis for CI.**

Here, we report crystal structures for CidA^wMel and the DUB domain of CidB^wMel and build a model of the CidA^wMel-CidB^wMel complex using AlphaFold-Multimer³⁸, which is validated and confirmed by comparison to a crystal structure of a homologous CidA^wPip-CidB^wPip complex. Substituting the CidA^wMel residues at the CidB-binding interfaces with those of CidA^wPip enables it to bind CidB^wPip and rescues CidB^wPip-induced yeast growth defects, which suggests a role of cid gene co-evolution in modulating CidA-CidB binding specificity. Analysis using the crystal structures of the CidA^wPip-CidB^wPip complex further shows that sequence divergence in Wolbachia wPip may result in complex binding patterns between Cid^wPip protein variants. These results provide a solid base for future investigations of CI mechanism and shed light on how co-evolution of CidA-CidB pairs modulates their interactions and bidirectional CI.

Results

Crystal structures of CidA^wMel and CidB^wMel _DUB, and a model of CidB^wMel _ND1-ND2

We determined the crystal structure of CidA^wMel to 2.75 Å resolution (Supplementary Table 1). CidA^wMel is made up largely of α-helices (Fig. 1b). Residues 111-154 and 158-165 are disordered and cannot be modeled, which divide CidA^wMel into an N-terminal domain (NTD, residues 1 to 161) (see below for why we include these disordered residues as part of the N-terminal domain) and a C-terminal domain (CTD, residues 162 to 422) (Fig. 1b). The C-terminal domain folds into a twisted set of six HEAT repeats, a structural motif that primarily mediates protein-protein interactions³⁹ (Fig. 1b).

CidB^wMel was expressed poorly in Escherichia coli, making direct structure determination difficult. We used a “divide-and-conquer” strategy to obtain structural information for each of its domains separately. CidB^wMel consists of a region, which is predicted to contain two PD-(D/E)XK (pseudo) nuclease domains (residue 1-751; hereafter CidB^wMel_ND1-ND2, ND: nuclease domain) and a C-terminal deubiquitylase (DUB) domain. An active DUB domain is required for inducing a CI-like phenotype in transgenic D. melanogaster for cid operons²¹. We determined the crystal structure of the CidB^wMel DUB domain to 1.85 Å resolution (Supplementary Table 1 and Fig. 1c). Similar to the structures of other proteases in the CE clan/Ulp1-like protease family^40,41, the DUB core consists of a five-stranded β sheet flanked by α helices on both sides. His957, Asp976, and Cys1016 form the catalytic triad (Fig. 1c). Three variable regions (VRs), a constant region (CR) and a C-terminal accessory domain may account for the S1 ubiquitin-binding interface⁴¹ (Supplementary Fig. 1a-d).

A model for CidB^wMel_ND1-ND2 (Fig. 1d) was built with the AlphaFold⁴² program. Although the predicted structural model has an average pLDDT score as high as 83.190, which confirms its quality (pLDDT > 70 indicates the backbone prediction is correct⁴³), we wanted to validate this model further.

CidB^wMel_ND1-ND2 shares sequence homology with other CifB molecules³⁴. Of note, the ND1-ND2 region of CidB^wPip_I(b/2) (a CidB^wMel ortholog from the Wolbachia strain wPip Tunis line³⁰, hereafter CidB^wPip(Tunis)_ND1-ND2) expressed well in E. coli and shares a sequence identity of 71% with CidB^wMel_ND1-ND2 (Supplementary Fig. 2). We successfully determined a crystal structure of CidB^wPip(Tunis)_ND1-ND2 in complex with CidA^wPip(Tunis) (specifically CidA^wPip_I(γ/2)³²) (Supplementary Table 1, Figs. 1e and 2a). The structure of CidB^wPip(Tunis)_ND1-ND2 could be used for comparison with CidB^wMel_ND1-ND2 modeled by AlphaFold⁴².

**Fig. 2: CidA and CidB interact through a large conserved tripartite interface.**

The structure of CidB^wPip(Tunis)_ND1-ND2 is very similar to the model for CidB^wMel_ND1-ND2 (Supplementary Fig. 3a). Moreover, the residues that are varied among CidB homologs are mainly located at the surfaces of the protein (Supplementary Fig. 4a), which should not affect protein folding. Thus, the model for CidB^wMel_ND1-ND2 is likely to be accurate. Since the cif genes from wMel have been well characterized, we will mainly focus on the structures and functional analysis of the CidA^wMel-CidB^wMel_ND1-ND2 pair in the following sections. The results from the CidA^wPip(Tunis)-CidB^wPip(Tunis) system will be mentioned when necessary.

CidB^wMel_ND1-ND2 consists of two NDs; ND1 comprises residues 1-391 and ND2 comprises residues 392-751 (Fig. 1d). The ND1 and ND2 domains of CidB^wMel interact through an extensive interface burying a surface area of ~2200 Å². This interface is highly similar to that of CidB^wPip(Tunis) (Supplementary Fig. 3b-c) and the residues involving in ND1-ND2 interaction are highly conserved among CidB homologs (Supplementary Fig. 3d-f). Interestingly, in addition to the (inactive) PD-(D/E)XK modules predicted by sequence homology analysis (residues 277-375 and 607-730)^21,28,29,30, the structural model of CidB^wMel shows that each ND has an additional more divergent PD-(D/E)XK module (residues 30-137 and 431-523) (Supplementary Fig. 5). None of these modules present the key canonical catalytic residues E-D-E-K²³ (the putative corresponding residues are P33, K77, S92, S94 in module I, K279, Y311, N330, T332 in module II, F435, K475, I491, D493 in module III and K609, G663, V680, G682 in module IV based on structural analysis (Supplementary Fig. 5)). These modules may play structural roles or support an alternative catalytic activity.

CidA^wMel binds CidB^wMel through a large distinct interface

We built a model for the CidA^wMel-CidB^wMel_ND1-ND2 complex using AlphaFold-Multimer³⁸ (Fig. 2b). The predicted end-to-end structure has a ranking score (TM-Score) as high as 0.8568 with a ptm score (intra-chain quality) of 0.86436 and an iptm score (interface score) of 0.85498. This indicates our prediction is highly reliable both for the protein themselves and the interface between them. The obtained end-to-end model was further optimized by Molecular Dynamics (MD) simulations. Two independent trajectories were performed and both achieved equilibrium within the first 100 ns and were then trapped in optima in the next 100 ns. The Root-Mean-Square Deviation (RMSD) for one of the simulation trajectories was recorded (Supplementary Fig. 6a). Trajectory cluster analysis was performed and the most stable binding conformation was extracted from the largest cluster of the trajectory as the final binding complex of CidA^wMel-CidB^wMel_ND1-ND2. The complex also predicts many hydrogen bonds and salt bridges between CidA^wMel and CidB^wMel_ND1-ND2 at the three interfaces (Fig. 2c–e) and only 0.57% of the residues are Ramachandran outliers (Supplementary Fig. 6b), which demonstrated the side chains of the predicted model have been reliably predicted. Therefore, we believe that the model of the CidB^wMel_ND1-ND2 and its complex with CidA^wMel is reasonable.

The interface between CidA^wMel and CidB^wMel_ND1-ND2 can be divided into three regions (Fig. 2b); each region mainly involves a pair of structural motifs. At the first region (Interface I), the helices consisting of the residues 93-158 of CidA^wMel interact with a loop in CidB^wMel_ND1-ND2 (residues 448-462) through a network of hydrogen bonds and salt bridges (Fig. 2c). Interestingly, the region of CidA^wMel which is disordered in the absence of CidB^wMel (residue 111-154 and 158-165) is modeled to form a helix bundle upon complex formation (Figs. 1b and 2c), consistent with its role in mediating CidA^wMel-CidB^wMel_ND1-ND2 interaction. These residues are part of the N-terminal domain. Interface II is formed by the helices at the N-terminus of CidA^wMel (residues 2-60) and helices from both NDs of CidB^wMel_ND1-ND2 (residues 337-353 and 395-418) (Fig. 2d). Interface III involves multiple HEAT repeats in CidA^wMel and a cross-cutting helix (residues 241-264) in CidB^wMel_ND1-ND2 that is bolstered by two β strands and their connecting loop (residues 365-387) to stabilize the interaction (Fig. 2e). Interestingly, although the model of CidA^wMel-CidB^wMel_ND1-ND2 complex is very similar to the structure of CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 (Fig. 2a, b), the residues involving in interaction at the three interfaces are very different between the two complexes (Fig. 2c–h), which could help explain their cognate-specific binding. (The residues at the interface are well resolved in CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 with good electron density (Supplementary Fig. 7)).

Pulldown experiments and yeast growth assays were used to validate these structural models. As a model system, it has been demonstrated that expression of cidB genes in the yeast Saccharomyces cerevisiae inhibits cell growth, but growth is at least partially restored if the cognate cidA gene is coexpressed, resembling CI induction and rescue in insects^21,23. Substituting the CidA^wPip(Tunis) residues at its CidB-binding interface with the corresponding ones from CidA^wMel abolished its ability to bind CidB^wPip(Tunis)_ND1-ND2 and neutralize the CidB^wPip(Tunis)-induced growth defect in yeast (Supplementary Fig. 8). Thus, the structure of the CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 complex accurately captures the CidA-CidB interaction mode and is biologically relevant. However, due to the poor expression of CidB^wMel_ND1-ND2 in E. coli and the low toxicity of CidB^wMel in yeast, we could not directly validate the model of the complex by pulldown or yeast growth analysis. To test the reliability of this model, we rationally designed a set of mutations based on the model to demonstrate that the interaction mode is conserved between the wPip and wMel CidA-CidB pairs (see below).

CidA-CidB binding specificity is determined by residues at their interfaces

Until now, only CifA and CifB proteins expressed from the same operon have been shown to interact^21,23,35. The specific interactions between CifA and CifB have been proposed to play a role in CI induction and/or rescue, depending on different CI models. We substituted multiple residues at the binding interface of CidA^wMel with the corresponding region of CidA^wPip(Tunis) (Fig. 3a–c) and investigated if the chimeric construct (named CidA^wMel(ST)) could now bind CidB^wPip(Tunis)_ND1-ND2. Their interaction, while detectable, was weak, probably because the residues substituted are not the major binding determinants for interaction between CidA^wPip(Tunis) and CidB^wPip(Tunis)_ND1-ND2 (Fig. 3d). Interestingly, CidA^wMel(ST) could bind a closely related CidB^wPip_ND1-ND2 from Wolbachia strain wPip(Pel) (hereafter CidB^wPip(Pel)_ND1-ND2) to an extent similar to the binding of the cognate CidA^wPip(Pel) (Fig. 3d, lane 2 vs. lane 4). By contrast, CidA^wMel had been shown previously to be unable to bind CidB^wPip(Pel) (Ref. 21). CidB^wPip(Pel)_ND1-ND2 only differs from CidB^wPip(Tunis)_ND1-ND2 by 30 residues (Supplementary Fig. 9). We used CidB^wPip(Pel) for further biochemical investigation.

**Fig. 3: Mutagenesis with binding and yeast growth assays reveal how residues at the three interfaces determine CidA binding specificity.**

Agreeing with the pulldown experiments, yeast growth defects induced by CidB^wPip(Pel) could be rescued by both CidA^wMel(ST) and CidA^wPip(Pel), but not CidA^wMel (Fig. 3e) even though the expression levels for CidA^wMel and CidA^wMel(ST) were similar (Supplementary Fig. 10). These data clearly show the importance of the interfacial residues (Fig. 3e) in determining binding specificity. They also support the accuracy of our CidA^wMel-CidB^wMel_ND1-ND2 model.

To locate the critical regions that determine binding specificity, the substituted interfacial residues of CidA^wMel(ST) were divided into nine groups/regions based on their location in the primary sequence and three-dimensional structure. Each region was individually reverted back to the original sequence in CidA^wMel (Fig. 3b, c), resulting in CidA^wMel(ST-1) to CidA^wMel(ST-9). (Regions 1, 2 are at Interface II. Regions 3, 4, 5, 6 are at Interface I. Regions 7, 8, 9 are at Interface III.) CidA^wMel(ST-4), CidA^wMel(ST-7) and CidA^wMel(ST-9) showed reduced binding affinity to CidB^wPip(Pel)_ND1-ND2 (Fig. 3d). They also failed to rescue CidB^wPip(Pel)-induced yeast growth defects (Fig. 3f). (These CidA variants were expressed at similar levels. Interestingly, however, co-expression of CidA^wMel variants with CidB^wPip(Pel) in yeast did strongly enhance levels of the latter protein but only if they were able to bind to it. (Supplementary Fig. 10)) Thus, these specific residues at the binding interface are particularly important in determining the binding specificity of CidA^wMel.

Crystal structure of CidA^wMel(ST)-CidB^wPip(Pel) _ND1-ND2 explains interaction specificity

We next determined the crystal structure of the CidA^wMel(ST)-CidB^wPip(Pel)_ND1-ND2 complex (Supplementary Table 1 and Fig. 3g). The structure of the complex was very similar to that of CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 (RMSD: 0.8 Å over 691 Cα atoms). This structure revealed the interacting residues at the CidA^wMel(ST)-CidB^wPip(Pel)_ND1-ND2 interface (Supplementary Fig. 11; electron density for representative residues shown in Supplementary Fig. 7) and explained why certain CidA^wMel(ST) variants show reduced binding to CidB^wPip(Pel)_ND1-ND2. For example, at region 4, residues R114 and K121 of CidA^wMel(ST) interact electrostatically with residues E460 and D459 of CidB^wPip(Pel)_ND1-ND2, respectively (Fig. 3h). In CidA^wMel(ST-4), residues 114 and 121 were reverted back to the corresponding residues of CidA^wMel, which are both glutamines (Fig. 3c). These residues are not charged and may lead to the reduced binding of CidA^wMel(ST-4) to CidB^wPip(Pel)_ND1-ND2 (Fig. 3d). Similarly, at region 7, residues R288 and D299 of CidA^wMel(ST) have charge complementarity with D246 and R252 of CidB^wPip(Pel)_ND1-ND2, respectively (Fig. 3i). In the CidA^wMel(ST-7) revertant, residues 288 and 299 are glutamine and asparagine, respectively (Fig. 3c), which can no longer form salt bridges with D246 and R252 of CidB^wPip(Pel)_ND1-ND2. Finally, at region 9, residue R393 of CidA^wMel(ST) interacts with residues E257 and D261 of CidB^wPip(Pel)_ND1-ND2 through salt bridges (Fig. 3i). In CidA^wMel(ST-9), residue 393 is glutamic acid (Fig. 3c), which cannot bind to E257 and D261 of CidB^wPip(Pel)_ND1-ND2 due to electrostatic repulsion. The crystal structure clearly explains the reduced binding of CidA^wMel(ST-4), CidA^wMel(ST-7) and CidA^wMel(ST-9) to CidB^wPip(Pel)_ND1-ND2 (Fig. 3d), which closely parallels their loss of function in yeast growth rescue assays (Fig. 3e, f).

Natural sequence variations affect CidA^wPip(Tunis)-CidB^wPip(Tunis) binding

The wPip cidA-cidB operon is duplicated and diversified extensively among CI-inducing wPip strains. Interestingly, however, only residues at certain positions are different among the CidA^wPip and CidB^wPip variants³² (Fig. 4a, b). If the varied residues are at the CidA^wPip-CidB^wPip binding interface, they may affect the interaction of these proteins and play a role in CI induction and/or rescue specificity. The wPip Tunis line belongs to group I of wPip (wPipI, grouping of wPip strains is based on phylogenetic analysis of seven Wolbachia genes)^32,44, so we focused on the CidA^wPip and CidB^wPip variants from this group (hereafter, CidA^wPip_I and CidB^wPip_I). (As mentioned before, CidA^wPip(Tunis) and CidB^wPip(Tunis) are CidA^wPip_I(γ/2) and CidB^wPip_I(b/2), respectively.)

**Fig. 4: Sequence variations modulate interactions between natural CidA^wPip and CidB^wPip alleles.**

Among the positions that show divergence (Fig. 4b), residues 118, 125, 149, 150, 341, 342, 399, and 400 of CidA^wPip and residues 242, 243, 253, 254, 257, 451, 452, 460, 461, and 462 of CidB^wPip map to the binding interface of the two proteins (Fig. 4c). These variations create distinct interaction surfaces for each of the four known CidA^wPip_I variants (Fig. 4a). Specifically, CidA^wPip variants I(α/1) and I(β/2) have the same residues A118, G125, G149 and G150; the corresponding residues in variants I(γ/1) and I(γ/2) are R118, K125, D149, and S150 (Fig. 4a). CidA^wPip variants I(α/1) and I(γ/1) both share residues G342, R343, K400, and N401, which correspond to residues Y341, T342, R399, and D400 in variants I(β/2) and I(γ/2). Similarly, each of the four CidB^wPip_I variants has a unique combination of residues at the CidA-binding interface (Fig. 4a). Pulldown experiments were used to investigate the interactions between the CidA^wPip variants and the CidB^wPip variants. His-tagged CidA variants I(α/1), I(γ/1), I(β/2), and I(γ/2) cannot pull down GST-tagged CidB variants I(b/2), I(b/1), I(a/2), and I(a/1), respectively (Figs. 4d, e), which supports the hypothesis that natural sequence variations affect binding interactions between CidA^wPip and CidB^wPip from different strains. If binding is required for CidA to rescue CidB-induced defects, this sequence diversity may at least partially explain the complex bidirectional incompatibility patterns in crosses between different C. pipiens lines (discussed below).

Discussion

Here, we have presented structural, biochemical and functional studies on the Wolbachia CI factors CidA^wMel and CidB^wMel, providing a solid foundation for investigation of the molecular functions of the Cid proteins in CI. These proteins are used here as a model to investigate the evolution of CidA-CidB interactions and their biological impact. Our structure-based mutagenesis has led to a clear demonstration of noncognate CifA-CifB binding, which suggests how sequence changes specifically at the interfaces between these proteins could lead to the complex incompatibility relationships that can be observed in the wild³².

The Cid systems from Wolbachia strains wMel and wPip are evolutionarily related, but these strains reside in different clades or supergroups (supergroups A and B, respectively). The two strains are expected to be incompatible based on crosses with trans-infected Ae. albopictus mosquitoes³⁶. Here, we show that the CidA-CidB pairs from these Wolbachia strains share the same interaction mode but have different residues at their binding interfaces. These residues are critical for determining their binding specificity. CidA^wMel(ST), a rationally designed chimeric construct containing the scaffold of CidA^wMel and the interfacial residues of CidA^wPip, binds CidB^wPip(Pel) and rescues CidB^wPip(Pel)-induced growth defects in yeast. It proves that CidA variants from different Wolbachia strains use the same pathway to carry out their function.

To further evaluate the conservation and evolutionary relationship of CidA and CidB found in different Wolbachia strains, the conservation scores of CidA and CidB residues were calculated by the ConSurf server and mapped onto their structures^45,46,47. The residues in the core of the proteins tend to be highly conserved. In contrast, the residues located at the surface are more variable (Supplementary Fig. 4a-b). Many of the varied residues are located at the CidA-CidB binding interfaces (Supplementary Fig. 4c). (Although several key residues of CidA at interface III are identical among its homologs, their interaction with CidB cannot be maintained due to residue variations at interface III of CidB.) This agrees with previous hypotheses that the CidA and CidB interfaces are co-evolving^32,35, which could modulate binding between CidA and CidB proteins and lead to incompatibility.

Wolbachia strains infecting C. pipiens usually contain more than one copy of the cidA and cidB genes³², which has been proposed to be responsible for the very complex incompatible patterns documented in C. pipiens infected with these strains. We showed here that the natural sequence variations in CidA^wPip and CidB^wPip proteins affects their binding specificity. The expression of multiple CidA^wPip and CidB^wPip variants in one Wolbachia strain could lead to a mix-and-match type of binding and may be one reason for the very complex crossing results. The presence of cidB_IV(2) variants was associated with the incompatible phenotype of mosquitoes infected with group IV wPip when crossing with those infected with wPip from other groups^32,37. Indeed, the CidB_IV(2) variants have unique residues at interface I (including residues 451, 452, 460, 461 and 462) compared to the CidB of the other groups. These differences may impair the binding of the CidB_IV(2) variants to the CidA variants of the other groups and lead to incompatibility.

The crystal structure of CidB^wMel_DUB provides a solid structural basis for further investigation of the molecular targets and function of CidB deubiquitylase family proteins. Given that bacteria do not possess a ubiquitin-conjugation system, the presence of a DUB domain in CidB strongly suggests that this Wolbachia protein functions within the host^21,40. Indeed, CidB proteins have been proposed to target nuclear-protein import and protamine-histone exchange factors⁴⁸. CidB^wMel_DUB has three variable regions and a C-terminal region different from other bacterial proteases in the CE-clan/Ulp1-like protease family, which could be responsible for its substrate preference for ubiquitin over other ubiquitin-like molecules and for Lys63-linked ubiquitin chains^21,40,41.

In different models that try to account for how Cif proteins induce and rescue CI, CifA-CifB interaction is suggested to play different roles. In one, which is named the “2-by-1” model, CifB is suggested to have an ancillary role by modulating CifA stability or activity in the male germline²⁵. By contrast, in the “toxin-antidote” model, specific CifA-CifB interaction in the egg has been proposed to be essential for CifA to rescue CifB-induced CI^32,35. Two major differences between these models are where CifA and CifB interact (in male or female) and whether the interaction is involved in CI induction or rescue. Our structures of CidA-CidB complexes and attempts to modulate their binding specificity provide a way to rationally design CifA and CifB mutants with desired binding attributes. Analysis of these variants will help distinguish between these and other CI models.

Methods

DNA manipulation

The genes coding for CidA, the variants of CidA, and CidB of wPip(Tunis), wPip(Pel) and wMel were synthesized at the Beijing Genomics Institute (BGI China). These sequences were codon-optimized for expression in E. coli by BGI. For crystallization, CidA^wPip(Tunis), CidB^wPip(Tunis)_ND1-ND2 (residues 1-761), CidB^wPip(Pel)_ND1-ND2 (residues 1-761) and CidA^wMel genes were subcloned into the pET-22b(+) vector using restriction sites NdeI and XhoI, encoding proteins with a C-terminal His-tag. CidB^wMel_DUB (residues 797-1128) gene was subcloned into the pGEX-6p-1 vector possessing an N-terminal glutathione-S-transferase (GST) tag using restriction sites BamHI and XhoI.

For in vitro pull-down experiments, CidA variants were the same as those used for crystallization. A pET28a-GST vector was made where the GST coding sequence was inserted between restriction sites NcoI and BamHI on pET28a. The CidB^wPip(Tunis)_ND1-ND2 and CidB^wPip(Pel)_ND1-ND2 genes were cloned into the pET28a-GST expression vector between restriction sites EcoRI and XhoI so that the proteins expressed had an N-terminal GST-tag.

For yeast growth analysis, DNA fragments were subcloned from E. coli vectors by restriction digest or PCR amplification and ligated into yeast expression vectors. The 2-micron plasmid pRS425GAL1 (LEU2) utilizing a GAL1 promoter was used for galactose-induced CifA expression in yeast and the low-copy vector pRS416GAL1 (URA3) was used for galactose-induced CifB expression.

Primers used to generate CifA or CifB expression plasmids are summarized in Supplementary Table 2. All plasmids were verified by sequencing (Sangon Biotech, China).

Protein expression and purification

All proteins were expressed in E. coli (BL21 (DE3) strain). Briefly, E. coli transformed with an expressing plasmid was cultured in Luria broth (LB) at 37 °C to an optical density (OD₆₀₀) of 0.6. Overexpression of the recombinant proteins was induced by adding isopropyl-β-D-thiogalactopyranoside (IPTG) to a final concentration of 0.5 mM at 16 °C for 16–18 h.

The harvested bacteria overexpressing CidA^wMel (or CidA variants) were resuspended in a lysis buffer (20 mM Tris-HCl, pH 8.0, 300 mM NaCl, 10 mM imidazole, 10% glycerol) and lysed via a high-pressure homogenizer at 4 °C. The lysate was centrifugated at 26500 × g for 30 min at 4 °C. After centrifugation, the supernatant was loaded onto a Ni-NTA column (GE Healthcare, USA). The column was washed using a lysis buffer supplemented with 50 mM imidazole and eluted using a lysis buffer supplemented with 500 mM imidazole. The eluted protein was diluted with buffer A (20 mM Tris-HCl, pH 8.0, 5 mM DTT) and further purified by anion-exchange chromatography (Hi Trap Q HP 5 mL, GE Healthcare, USA), using a linear gradient of 0%-40% mixture of buffer A and buffer B (20 mM Tris-HCl, pH 8.0, 1 M NaCl, pH 8.0, 5 mM DTT). Finally, the protein was purified by gel-filtration chromatography (Superdex 200 10/300 GL, GE Healthcare, USA), using a buffer containing 20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 5 mM DTT.

To obtain the CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 and CidA^wMel(ST)-CidB^wPip(Pel)_ND1-ND2 complexes, bacterial cells expressing each component of the specific complex were mixed and co-lysed in the lysis buffer containing 20 mM Tris-HCl, pH 8.0, 0.3 M NaCl, 10 mM imidazole, 10 μg/mL DNase 1, 10 μg/mL RNase A and 10% glycerol. The complex were further purified by running cleared lysate sequentially through Ni-NTA affinity, anion-exchange and size exclusion chromatography using the same columns and buffers as described for CidA^wMel.

The CidB^wMel_DUB protein was purified following a similar procedure, using affinity, anion-exchange and size exclusion chromatography. CidB^wMel_DUB has a GST-tag. Instead of using Ni-NTA affinity chromatography, GST affinity chromatography was used. The harvested bacteria overexpressing CidB^wMel_DUB were resuspended in a lysis buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 5% glycerol) and lysed via a high-pressure homogenizer at 4 °C. The lysate was centrifugated at 26500 × g for 30 min at 4 °C. After centrifugation, the supernatant was loaded onto a glutathione-Sepharose column (GE Healthcare, USA). The GST tag was removed by incubating the loaded column with PreScission protease overnight at 4 °C. The CidB^wMel_DUB protein was further purified through anion-exchange and size exclusion chromatography using the same columns and buffers as described above.

Selenomethionine (SeMet)-derivatized proteins were expressed by bacteria growing in M9 SeMet medium supplemented with 100 mg/L L-selenomethionine. The purification procedure of SetMet-derivated proteins was the same as mentioned above for the native.

Crystallization, data collection, and structure determination

All crystals were grown by the microbatch-under-oil method unless otherwise specified⁴⁹. CidA^wMel was crystallized at 16 °C by mixing 1 µL protein (5 mg/mL) with 1 µL crystallization buffer containing 0.2 M Sodium phosphate monobasic monohydrate, 20% w/v Polyethylene glycol 3350, pH 4.7. The crystals were cryoprotected by Parabar 10312 (previously known as Paratone oil, Hampton Research, USA). X-ray diffraction data were collected on beamline BL18U1 at the Shanghai Synchrotron Radiation Facility at 100 K and at a wavelength of 0.97852 Å. Data integration and scaling were performed using HKL3000⁵⁰. The structure was determined by SeMet single-wavelength anomalous dispersion (SAD) method with the AutoSol program in PHENIX⁵¹. The CidA^wMel model was initially built by the Autobuild program in PHENIX and subsequently subjected to iterative cycles of manual building in Coot⁵² and refinement in PHENIX.

The crystals of the CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 complex were grown at 16 °C from a mixture of 1 µL protein (5 mg/mL) and 1 µL crystallization buffer containing 0.1 M BICINE, pH 8.5 and 15% w/v Polyethylene glycol 1500. The crystals were cryoprotected by Parabar 10312. X-ray diffraction data were collected on beamline BL18U1 at the Shanghai Synchrotron Radiation Facility. The structure was determined by molecular replacement using the structure of the CidA^wPip(Pel)-CidB^wPip(Pel)_ND1-ND2 complex as the search model⁵³.

The crystals of the CidA^wMel(ST)-CidB^wPip(Pel)_ND1-ND2 complex were grown at 16 °C from a mixture of 1 µL protein (5 mg/mL) and 1 µL crystallization buffer containing 5% v/v (+/-)-2-Methyl-2,4-pentanediol, 0.1 M HEPES pH 7.5, 10% w/v Polyethylene glycol 10,000. The crystals were cryoprotected by Parabar 10312. X-ray diffraction data were collected on beamline BL18U1 at the Shanghai Synchrotron Radiation Facility. The structure was determined by molecular replacement using the structure of the CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 complex as the search model.

CidB^wMel_DUB was crystallized by hanging drop method at 18 °C, with the crystallization reservoir solution containing 10 mM Nickel (II) Chloride hexahydrate, 100 mM TRIS pH 8.5 and 20% w/v Polyethylene Glycol Monomethyl Ether 2000. The crystals were directly flash frozen in liquid nitrogen using reservoir solution supplemented with 10% glycerol as cryoprotectant. X-ray diffraction data were collected at beamline BL17U1 at the Shanghai Synchrotron Radiation Facility. The structure was determined by SeMet SAD method as described above.

Data collection and structure refinement statistics are summarized in Supplementary Table 1. All Molecular graphics were created using UCSF ChimeraX⁵⁴.

In vitro pull-down experiments

E. coli BL21(DE3) expressing either His-tagged CidA mutants or GST-tagged CidB_ND1-ND2 variants were harvested by centrifugation and resuspended separately in a buffer containing 20 mM Tris-HCl, pH 8.0, 300 mM NaCl, 1 mM PMSF and 5% glycerol. The expression levels of CidA variants were approximated by the BCA Protein Assay (Tiangen Biotech (Beijing) Co., Ltd., China). Equal amount of CidA variants were mixed with a large and fixed volume of CidB_ND1-ND2 variants to ensure that CidB_ND1-ND2 variants were in excess to CidA. The mixture was co-lysed by sonication. Cleared lysates were incubated with 50 μL Ni-NTA resin at 4 °C for 1 h. The resin was washed eight times with 800 μL buffer containing 20 mM Tris-HCl pH 8.0, 200 mM NaCl, 0.01% Tween-20, 50 mM imidazole each time. Proteins were eluted by adding 3 resin volumes of 20 mM Tris-HCl pH 8.0, 200 mM NaCl, 300 mM imidazole. The samples were analyzed using the SDS-PAGE and Coomassie stain.

Yeast growth assays

Growth was analyzed in the BY4741 strain as previously described⁵⁵. Briefly, yeast cultures were grown overnight at 30 °C in Yeast Extract Peptone Dextrose (YPD) media or synthetically defined (SD) raffinose media lacking uracil, leucine or both. Yeast were pelleted by centrifugation, washed with sterile water, and spotted in six-fold serial dilution from an initial OD₆₀₀ 0.2 concentration on solid minimal synthetic media containing either 2% galactose or glucose and lacking either uracil, leucine, or both. Plates were placed at 30 or 36 °C for 2 to 3 days.

Western blot analysis

For immunoblotting, co-expression culture in raffinose minimal medium (SD) lacking uracil and leucine were diluted to 0.2 OD₆₀₀ in galactose (inducing) minimal medium (SD) lacking uracil and leucine, kept 12–16 h at 30 °C until reaching 0.8-1.0 OD₆₀₀ at which point the equivalent of 2.5 OD₆₀₀ units of cells were harvested, washed and resuspended in 1 mL dH₂O followed by the addition of 200 μL dH₂O and 200 μL 0.2 M NaOH, incubated for 5 min at room temperature. Cells were vortexed intermittently for 20 s and pelleted at 10,000 x g, 1 min. Pellets were stored at −80 °C for at least 15 min, resuspended in 100 μL 1 x SDS-PAGE sample buffer and 4% β-mercaptoethanol then heated at 95 °C for 3 min, centrifuged and 20 μL supernatant were loaded in the 10% SDS-PAGE gel and transferred to PVDF Immobilon-P transfer membranes (0.45 μM pore size) (Sigma-Aldrich) under 70 V, 2.5 h used for immunoblot analyses. Antibodies used for immunoblotting were as following: mouse anti-FLAG M2 (Sigma, 1:10,000), secondary antibody used was sheep anti-mouse NXA931V (GE Healthcare, 1:5,000); and mouse anti-PGK1 (yeast phosphoglycerate kinase; Abcam, 1:10,000), secondary antibody used was sheep anti-mouse NXA931V (GE Healthcare, 1:10,000). All immunoblot analyses used 5% milk for blocking. All serial dilution and Western blot data are representative of at least two biological experiments. Proteins were visualized by HRP-based chemiluminescence⁵⁶.

AlphaFold modeling

In this study, AlphaFold was used to predict the monomer structure of CidB^wMel_ND1-ND2, and AlphaFold-Multimer was used to predict the binding complex of CidA^wMel-CidB^wMel_ND1-ND2 with multiple sequence alignments (MSA) set as the all genetics database used at CASP14. The prediction of complexes was run twice with different random seeds and 10 models were obtained. Beginning with visual inspection, four of them were selected to perform protein structural quality check for the side chain conformations using prime module of Schrödinger2021-3. Eventually, the one complex with the highest quality score was selected for further optimization with subsequent MD simulations.

Molecular dynamics (MD) simulations

MD simulations were performed by using Desmond package of Schrödinger2021-3⁵⁷ using the OPLS4⁵⁸ force field. The binding model of CidA^wMel-CidB^wMel_ND1-ND2 obtained in the last step was explicitly solvated with TIP3P⁵⁹ water molecules under cubic periodic boundary conditions for a 15 Å buffer region. The overlapping water molecules are deleted and 0.15 M KCl is added, and the systems were neutralized by adding K⁺ as counter ions. Brownian motion simulation was used to relax these systems into local energy minimum states separately. An ensemble (NPT) was then applied to maintain the constant temperature (300 K) and pressure (1.01325 bar) of the systems, and the simulations were started with different random initial velocities. The results were visually analyzed by using Maestro graphical interfaces and the RMSD was calculated based on C-alpha atoms. Produced trajectories were clustered using the Desmond trajectory cluster analysis panel. Finally, the most energetically stable binding complex from the largest cluster of conformation in MD trajectory was selected and minimized again with backbone constraints using the prime module of Schrödinger2021-3. The Ramachandran plot of the eventual model was generated with Schrödinger2021-3.

Evolutionary conservation analysis

The conservation score per amino acid of CidA and CidB was calculated using the ConSurf server (https://consurf.tau.ac.il/) based on their homologous sequences (specifically, CidA homologs from WOPip1, WOBol1-b, WOPipJHB, WOHa1, WOSol, WORecB, WOMelB, WOSuziB, and WORiB and CidB homologs from WOPip1, WOBol1-b, WOHa1, WOSol, WOMelB, WOMelPop, and WOSuziB). The structures of CidA^wPip(Tunis) and CidB^wPip(Tunis)_ND1-ND2 served as templates. “Maximun likelihood” and “best model” were selected as the calculation method and the evolutionary substitution model, respectively, for this analysis.

Reporting summary

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

Data availability

The data that support this study are available from the corresponding authors upon reasonable request. The coordinates for crystal structures of CidA^wMel, CidB^wMel_DUB, CidA^wPip(Tunis)-CidB^wPip(Tunis)_ND1-ND2 complex, and CidA^wMel(ST)-CidB^wPip(Pel)_ND1-ND2 complex have been deposited in the Protein Data Bank (PDB), with the accession codes 7FIT, 7FIU, 7FIV, and 7FIW, respectively. Source data are provided with this paper.

References

Werren, J. H. & Windsor, D. M. Wolbachia infection frequencies in insects: evidence of a global equilibrium? Proc. Biol. Sci. 267, 1277–1285 (2000).
Article CAS PubMed PubMed Central Google Scholar
Zug, R. & Hammerstein, P. Still a host of hosts for Wolbachia: analysis of recent data suggests that 40% of terrestrial arthropod species are infected. PloS One 7, e38544 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Weinert, L. A., Araujo-Jnr, E. V., Ahmed, M. Z. & Welch, J. J. The incidence of bacterial endosymbionts in terrestrial arthropods. Proc. Biol. Sci. 282, 20150249 (2015).
PubMed PubMed Central Google Scholar
Kaur, R. et al. Living in the endosymbiotic world of Wolbachia: A centennial review. Cell Host Microbe 29, 879–893 (2021).
Article CAS PubMed Google Scholar
Hoffmann, A. A., Turelli, M. & Harshman, L. G. Factors affecting the distribution of cytoplasmic incompatibility in Drosophila simulans. Genetics 126, 933–948 (1990).
Article CAS PubMed PubMed Central Google Scholar
Werren, J. H., Baldo, L. & Clark, M. E. Wolbachia: master manipulators of invertebrate biology. Nat. Rev. Microbiol. 6, 741–751 (2008).
Article CAS PubMed Google Scholar
Callaini, G. & Riparbelli, M. G. Fertilization in Drosophila melanogaster: centrosome inheritance and organization of the first mitotic spindle. Dev. Biol. 176, 199–208 (1996).
Article CAS PubMed Google Scholar
Lassy, C. W. & Karr, T. L. Cytological analysis of fertilization and early embryonic development in incompatible crosses of Drosophila simulans. Mech. Dev. 57, 47–58 (1996).
Article CAS PubMed Google Scholar
Reed, K. M. & Werren, J. H. Induction of paternal genome loss by the paternal-sex-ratio chromosome and cytoplasmic incompatibility bacteria (Wolbachia): a comparative study of early embryonic events. Mol. Reprod. Dev. 40, 408–418 (1995).
Article CAS PubMed Google Scholar
Shropshire, J. D., Kalra, M. & Bordenstein, S. R. Evolution-guided mutagenesis of the cytoplasmic incompatibility proteins: identifying CifA’s complex functional repertoire and new essential regions in CifB. PLoS Pathog. 16, e1008794 (2020).
Article CAS PubMed PubMed Central Google Scholar
Breeuwer, J. A. & Werren, J. H. Microorganisms associated with chromosome destruction and reproductive isolation between two insect species. Nature 346, 558–560 (1990).
Article ADS CAS PubMed Google Scholar
Poinsot, D., Bourtzis, K., Markakis, G., Savakis, C. & Mercot, H. Wolbachia transfer from Drosophila melanogaster into D. simulans: host effect and cytoplasmic incompatibility relationships. Genetics 150, 227–237 (1998).
Article CAS PubMed PubMed Central Google Scholar
Zabalou, S. et al. Multiple rescue factors within a Wolbachia strain. Genetics 178, 2145–2160 (2008).
Article PubMed PubMed Central Google Scholar
Brucker, R. M. & Bordenstein, S. R. Speciation by symbiosis. Trends Ecol. Evol. 27, 443–451 (2012).
Article PubMed Google Scholar
Werren, J. H. Biology of Wolbachia. Annu. Rev. Entomol. 42, 587–609 (1997).
Article CAS PubMed Google Scholar
Poinsot, D., Charlat, S. & Mercot, H. On the mechanism of Wolbachia-induced cytoplasmic incompatibility: confronting the models with the facts. Bioessays 25, 259–265 (2003).
Article PubMed Google Scholar
Laven, H. Crossing experiments with Cluex strains. Evolution 5, 370–375 (1951).
Article Google Scholar
Smith-white, S. & Woodhill, A. The nature and significance of non-reciprocal fertility in Aedes scutellaris and other mosquitoes. Proc. Linn. Soc. N. S. W. 79, 167–173 (1955).
Google Scholar
Ryan, S. L. & Saul, G. B. 2nd Post-fertilization effect of incompatibility factors in Mormoniella. Mol. Gen. Genet. 103, 29–36 (1968).
Article CAS PubMed Google Scholar
Yen, J. H. & Barr, A. R. The etiological agent of cytoplasmic incompatibility in Culex pipiens. J. Invertebr. Pathol. 22, 242–250 (1973).
Article CAS PubMed Google Scholar
Beckmann, J. F., Ronau, J. A. & Hochstrasser, M. A. Wolbachia deubiquitylating enzyme induces cytoplasmic incompatibility. Nat. Microbiol. 2, 17007 (2017).
Article PubMed PubMed Central Google Scholar
LePage, D. P. et al. Prophage WO genes recapitulate and enhance Wolbachia-induced cytoplasmic incompatibility. Nature 543, 243–247 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Chen, H., Ronau, J. A., Beckmann, J. F. & Hochstrasser, M. A. Wolbachia nuclease and its binding partner provide a distinct mechanism for cytoplasmic incompatibility. Proc. Natl Acad. Sci. 116, 22314–22321 (2019).
Article CAS PubMed PubMed Central Google Scholar
Shropshire, J. D., On, J., Layton, E. M., Zhou, H. & Bordenstein, S. R. One prophage WO gene rescues cytoplasmic incompatibility in Drosophila melanogaster. Proc. Natl Acad. Sci. 115, 4987–4991 (2018).
Article CAS PubMed PubMed Central Google Scholar
Shropshire, J. D. & Bordenstein, S. R. Two-By-One model of cytoplasmic incompatibility: synthetic recapitulation by transgenic expression of cifA and cifB in Drosophila. PLoS Genet. 15, e1008221 (2019).
Article CAS PubMed PubMed Central Google Scholar
Adams, K. L. et al. Wolbachia cifB induces cytoplasmic incompatibility in the malaria mosquito vector. Nat. Microbiol. 6, 1575–1582 (2021).
Article CAS PubMed PubMed Central Google Scholar
Sun, G., Zhang, M., Chen, H. & Hochstrasser, M. The Wolbachia CinB nuclease is sufficient for induction of cytoplasmic incompatibility. mBio 13, e03177-21 (2022).
Bing, X. L., Zhao, D. S., Sun, J. T., Zhang, K. J. & Hong, X. Y. Genomic analysis of Wolbachia from Laodelphax striatellus (Delphacidae, Hemiptera) reveals insights into its “Jekyll and Hyde” mode of infection pattern. Genome Biol. Evol. 12, 3818–3831 (2020).
Article CAS PubMed PubMed Central Google Scholar
Martinez, J., Klasson, L., Welch, J. J. & Jiggins, F. M. Life and death of selfish genes: comparative genomics reveals the dynamic evolution of cytoplasmic incompatibility. Mol. Biol. Evol. 38, 2–15 (2021).
Article CAS PubMed Google Scholar
Lindsey, A. R. I. et al. Evolutionary genetics of cytoplasmic incompatibility genes cifA and cifB in prophage WO of Wolbachia. Genome Biol. Evol. 10, 434–451 (2018).
Article CAS PubMed PubMed Central Google Scholar
Shropshire, J. D., Rosenberg, R. & Bordenstein, S. R. The impacts of cytoplasmic incompatibility factor (cifA and cifB) genetic variation on phenotypes. Genetics 217, 1–13 (2021).
Article PubMed Google Scholar
Bonneau, M. et al. Culex pipiens crossing type diversity is governed by an amplified and polymorphic operon of Wolbachia. Nat. Commun. 9, 319 (2018).
Article ADS PubMed PubMed Central Google Scholar
Lindsey, A. R. I. et al. Comparative genomics of the miniature wasp and pest control agent Trichogramma pretiosum. BMC Biol. 16, 54 (2018).
Article PubMed PubMed Central Google Scholar
Chen H., Zhang M., Hochstrasser M. The biochemistry of cytoplasmic incompatibility caused by endosymbiotic bacteria. Genes 11, 852 (2020).
Beckmann, J. F. et al. The toxin-antidote model of cytoplasmic incompatibility: genetics and evolutionary implications. Trends Genet. 35, 175–185 (2019).
Article CAS PubMed PubMed Central Google Scholar
Liang, X., Liu, J., Bian, G. & Xi, Z. Wolbachia inter-strain competition and inhibition of expression of cytoplasmic incompatibility in mosquito. Front. Microbiol. 11, 1638 (2020).
Article PubMed PubMed Central Google Scholar
Bonneau, M. et al. Variation in Wolbachia cidB gene, but not cidA, is associated with cytoplasmic incompatibility mod phenotype diversity in Culex pipiens. Mol. Ecol. 28, 4725–4736 (2019).
Article CAS PubMed PubMed Central Google Scholar
Evans R., et al. Protein complex prediction with AlphaFold-Multimer. Preprint at https://doi.org/10.1101/2021.10.04.463034, (2021).
Kobe, B. et al. Turn up the HEAT. Structure 7, R91–R97 (1999).
Article ADS CAS PubMed Google Scholar
Berk, J. M. et al. A deubiquitylase with an unusually high-affinity ubiquitin-binding domain from the scrub typhus pathogen Orientia tsutsugamushi. Nat. Commun. 11, 2343 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Pruneda, J. N. et al. The molecular basis for ubiquitin and ubiquitin-like specificities in bacterial effector proteases. Mol. Cell 63, 261–276 (2016).
Article CAS PubMed PubMed Central Google Scholar
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Tunyasuvunakool, K. et al. Highly accurate protein structure prediction for the human proteome. Nature 596, 590–596 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Atyame, C. M., Delsuc, F., Pasteur, N., Weill, M. & Duron, O. Diversification of Wolbachia endosymbiont in the Culex pipiens mosquito. Mol. Biol. Evol. 28, 2761–2772 (2011).
Article CAS PubMed Google Scholar
Ashkenazy, H. et al. ConSurf 2016: an improved methodology to estimate and visualize evolutionary conservation in macromolecules. Nucleic Acids Res. 44, W344–W350 (2016).
Article CAS PubMed PubMed Central Google Scholar
Ashkenazy, H., Erez, E., Martz, E., Pupko, T. & Ben-Tal, N. ConSurf 2010: calculating evolutionary conservation in sequence and structure of proteins and nucleic acids. Nucleic Acids Res. 38, W529–W533 (2010).
Article CAS PubMed PubMed Central Google Scholar
Celniker, G. et al. ConSurf: using evolutionary data to raise testable hypotheses about protein function. Isr. J. Chem. 53, 199–206 (2013).
Article CAS Google Scholar
Beckmann J. F., Sharma G. D., Mendez L., Chen H., Hochstrasser M. The Wolbachia cytoplasmic incompatibility enzyme CidB targets nuclear import and protamine-histone exchange factors. eLife 8, e50026 (2019).
Chayen, N. E. The role of oil in macromolecular crystallization. Structure 5, 1269–1274 (1997).
Article CAS PubMed Google Scholar
Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276, 307–326 (1997).
Article CAS PubMed Google Scholar
Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr., Sect. D: Biol. Crystallogr. 66, 213–221 (2010).
Article CAS Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr., Sect. D: Biol. Crystallogr. 66, 486–501 (2010).
Article CAS Google Scholar
Xiao, Y. et al. Structural and mechanistic insights into the complexes formed by Wolbachia cytoplasmic incompatibility factors. Proc. Natl. Acad. Sci. 118, e2107699118 (2021).
Pettersen, E. F. et al. UCSF ChimeraX: structure visualization for researchers, educators, and developers. Protein Sci. 30, 70–82 (2021).
Article CAS PubMed Google Scholar
Brachmann, C. B. et al. Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14, 115–132 (1998).
Article CAS PubMed Google Scholar
Mruk, D. D. & Cheng, C. Y. Enhanced chemiluminescence (ECL) for routine immunoblotting: an inexpensive alternative to commercially available kits. Spermatogenesis 1, 121–122 (2011).
Article PubMed PubMed Central Google Scholar
K. J. Bowers, et al. Scalable algorithms for molecular dynamics simulations on commodity clusters. SC ‘06: Proceedings of the 2006 ACM/IEEE Conference on Supercomputing, 43 (2006).
Lu, C. et al. OPLS4: improving force field accuracy on challenging regimes of chemical space. J. Chem. Theory Comput. 17, 4291–4300 (2021).
Article CAS PubMed Google Scholar
William, L., Jorgensen, J. C. & Madura, J. D. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. 79, 926–935 (1983).
Article ADS Google Scholar

Download references

Acknowledgements

We thank the staff from beamline BL17U1, BL18U1 and BL19U1 at Shanghai Synchrotron Radiation Facility (China). This work was supported by grants from the National Key Research Program of China (2016YFD0500300) to Lei Zhang, the National Natural Science Foundation of China (81772204), the Science and Technology Commission of Shanghai Municipality (YDZX20213100001556), the Lingang Laboratory (LG202101-01-07) and the Natural Science Foundation of Tianjin (18JCJQJC48000) to H.Y., the Natural Science Program of Chongqing Science and Technology Commission (cstc2019jcyj-msxmX0135), the Natural Science Program of Yuzhong District Science and Technology Commission (20190108) and the Science and Technology Research Program of Chongqing Municipal Education Commission (KJQN202000442) to W.W. and NIH grants GM136325 and GM046904 to M.H.

Author information

These authors contributed equally: Haofeng Wang, Yunjie Xiao, Xia Chen.

Authors and Affiliations

School of Life Sciences, Tianjin University, Tianjin, China
Haofeng Wang, Yunjie Xiao, Xia Chen, Feng Wang, Hanxiao Zhang, Xiaoyu Zhang, Wenling Li, Yi Wei, Wen Cui, Fenghua Wang, Cheng Chen, Jinhai Huang, Sheng Ye, Lei Zhang, Zefang Wang & Haitao Yang
Shanghai Institute for Advanced Immunochemical Studies and School of Life Science and Technology, ShanghaiTech University, Shanghai, China
Haofeng Wang, Yunjie Xiao, Lin Wang, Bing Zhang, Jun Li, Wei Shen, Fang Bai & Haitao Yang
Shanghai Clinical Research and Trial Center, Shanghai, China
Haofeng Wang, Yunjie Xiao, Lin Wang, Wei Shen, Fang Bai & Haitao Yang
Institute of Life Sciences, Chongqing Medical University, Chongqing, China
Yunjie Xiao, Xin Yang, Wen Cui & Wei Wang
Department of Molecular Biophysics & Biochemistry, Yale University, New Haven, USA
Mengwen Zhang, Guangxin Sun & Mark Hochstrasser
Department of Chemistry, Yale University, New Haven, USA
Mengwen Zhang
State Key Laboratory of Medicinal Chemical Biology, College of Pharmacy, Nankai University, Tianjin, China
Xiaoyu Zhang
Tianjin International Joint Academy of Biotechnology and Medicine, Tianjin, China
Xiaoyu Zhang, Zefang Wang & Haitao Yang
State Key Laboratory of Oncogenes and Related Genes, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
Deqiang Yao
State Key Lab of Biotherapy and Cancer Center, West China Hospital, Sichuan University and Collaborative Innovation Center for Biotherapy, Chengdu, China
Dan Su
State Key Laboratory of Pharmaceutical Biotechnology, Department of Biotechnology and Pharmaceutical Sciences, School of Life Sciences, Nanjing University, Nanjing, China
Xiaoyun Ji

Authors

Haofeng Wang
View author publications
You can also search for this author in PubMed Google Scholar
Yunjie Xiao
View author publications
You can also search for this author in PubMed Google Scholar
Xia Chen
View author publications
You can also search for this author in PubMed Google Scholar
Mengwen Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Guangxin Sun
View author publications
You can also search for this author in PubMed Google Scholar
Feng Wang
View author publications
You can also search for this author in PubMed Google Scholar
Lin Wang
View author publications
You can also search for this author in PubMed Google Scholar
Hanxiao Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Xiaoyu Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Xin Yang
View author publications
You can also search for this author in PubMed Google Scholar
Wenling Li
View author publications
You can also search for this author in PubMed Google Scholar
Yi Wei
View author publications
You can also search for this author in PubMed Google Scholar
Deqiang Yao
View author publications
You can also search for this author in PubMed Google Scholar
Bing Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Jun Li
View author publications
You can also search for this author in PubMed Google Scholar
Wen Cui
View author publications
You can also search for this author in PubMed Google Scholar
Fenghua Wang
View author publications
You can also search for this author in PubMed Google Scholar
Cheng Chen
View author publications
You can also search for this author in PubMed Google Scholar
Wei Shen
View author publications
You can also search for this author in PubMed Google Scholar
Dan Su
View author publications
You can also search for this author in PubMed Google Scholar
Fang Bai
View author publications
You can also search for this author in PubMed Google Scholar
Jinhai Huang
View author publications
You can also search for this author in PubMed Google Scholar
Sheng Ye
View author publications
You can also search for this author in PubMed Google Scholar
Lei Zhang
View author publications
You can also search for this author in PubMed Google Scholar
Xiaoyun Ji
View author publications
You can also search for this author in PubMed Google Scholar
Wei Wang
View author publications
You can also search for this author in PubMed Google Scholar
Zefang Wang
View author publications
You can also search for this author in PubMed Google Scholar
Mark Hochstrasser
View author publications
You can also search for this author in PubMed Google Scholar
Haitao Yang
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

M.H., and H.Y. conceived and supervised the project; H.W., Y.X., X.C., W.W., Z.W., M.H., and H.Y. designed the experiments; H.W., Y.X., X.C., Feng Wang, H.Z., X.Z., X.Y., W.L., and Y.W. cloned, expressed, purified and crystallized proteins; H.W., Y.X., X.C., H.Z., and Y.W. collected the diffraction data; Y.X., D.Y., B.Z., J.L., W.C., and W.W. solved the crystal structures; L.W. and F.B. performed the model prediction for CidB^wMel_ND1-ND2 and its complex with CidA^wMel. H.W., Y.X., M.Z., G.S., Feng Wang, and X.Z. performed mutagenesis of CidA and analyzed the mutants (CidA^wMel(ST) and CidA^wPip(Tunis)(ST), etc.) in in vitro pull-down experiments and yeast growth assays; Y.X., X.Y., and W.L. performed in vitro pull-down experiments of CidA^wPip_I and CidA^wPip_I variants; H.W., Y.X., X.C., M.Z., W.C., Fenghua Wang, C.C., W.S., D.S., J.H., S.Y., L.Z., X.J., W.W., Z.W., M.H., and H.Y. analyzed and discussed the data; H.W., Y.X., X.C., W.W., Z.W., M.H., and H.Y. wrote the manuscript.

Corresponding authors

Correspondence to Wei Wang, Zefang Wang, Mark Hochstrasser or Haitao Yang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Dahai Luo, Jonathan Pruneda and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

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

Supplementary information

Supplementary Information

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Wang, H., Xiao, Y., Chen, X. et al. Crystal Structures of Wolbachia CidA and CidB Reveal Determinants of Bacteria-induced Cytoplasmic Incompatibility and Rescue. Nat Commun 13, 1608 (2022). https://doi.org/10.1038/s41467-022-29273-w

Download citation

Received: 05 August 2021
Accepted: 03 March 2022
Published: 25 March 2022
DOI: https://doi.org/10.1038/s41467-022-29273-w

This article is cited by

Transgenic expression of cif genes from Wolbachia strain wAlbB recapitulates cytoplasmic incompatibility in Aedes aegypti
- Cameron J. McNamara
- Thomas H. Ant
- Steven P. Sinkins
Nature Communications (2024)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.