journal menu
research papers
 | STRUCTURAL BIOLOGY |
accessThe LH–DH module of bacterial replicative helicases is the common binding site for DciA and other helicase loaders
aUniversité Paris-Saclay, CEA, CNRS, Institute for Integrative Biology of the Cell (I2BC), 91180 Gif-sur-Yvette, France, and bSynchrotron SOLEIL, L'Orme des Merisiers, 91192 Gif-sur-Yvette, France
*Correspondence e-mail: helene.walbott@i2bc.paris-saclay.fr, sophie.quevillon-cheruel@i2bc.paris-saclay.fr
During the initiation step of bacterial genome replication, replicative helicases depend on specialized proteins for their loading onto oriC. DnaC and DnaI were the first loaders to be characterized. However, most bacteria do not contain any of these genes, which are domesticated phage elements that have replaced the ancestral and unrelated loader gene dciA several times during evolution. To understand how DciA assists the loading of DnaB, the of the complex from Vibrio cholerae was determined, in which two VcDciA molecules interact with a dimer of VcDnaB without changing its canonical structure. The data showed that the VcDciA binding site on VcDnaB is the conserved module formed by the linker helix LH of one monomer and the determinant helix DH of the second monomer. Interestingly, DnaC from Escherichia coli also targets this module onto EcDnaB. Thanks to their common target site, it was shown that VcDciA and EcDnaC could be functionally interchanged in vitro despite sharing no structural similarity. This represents a milestone in understanding the mechanism employed by phage helicase loaders to hijack bacterial replicative helicases during evolution.
Keywords: DnaB·DciA complex; crystal structure; helicase loaders; bacterial replicative helicases; interaction modules; replicative helicase hijacking; Vibrio cholerae.
PDB reference: VcDnaB·VcDciA, 8a3v
1. Introduction
The replication of the circular bacterial chromosome is an essential step for bacterial division. The initiator protein DnaA initiates replication by binding onto the origin DNA oriC and locally unwinds the double-stranded DNA by polymerization (Costa et al., 2013![[Costa, A., Hood, I. V. & Berger, J. M. (2013). Annu. Rev. Biochem. 82, 25-54.]](../../../../../../logos/arrows/d_arr.gif "Costa, A., Hood, I. V. & Berger, J. M. (2013). Annu. Rev. Biochem. 82, 25-54."){kind=small}
; Leonard & Méchali, 2013; Zawilak-Pawlik et al., 2017). The toroidal hexameric helicase DnaB is then loaded onto the locally open DNA duplex, with the help of a helicase loader, triggering the recruitment of the various proteins of the replisome (O'Donnell et al., 2013). ATP-dependent 5′-to-3′ translocation of the helicase ahead of the advancing replisome allows the unwinding of the DNA duplex into templates for new DNA synthesis (Strycharska et al., 2013).
Recruitment and loading of the replicative helicase depend on a loader protein, which has been characterized in the two model organisms Escherichia coli (Ec) and Bacillus subtilis (Bs), leading to the description of two loading strategies (see Table 1). In B. subtilis, the helicase loader DnaI assists the assembly of six monomers of the helicase to form an active hexameric ring around DNA according to a `ring-maker' scenario (Davey & O'Donnell, 2003; Velten et al., 2003). The GstDnaB·BsDnaI·GstDnaG prepriming complex exhibits a three-layered planar and dilated ring conformation with one hexameric helicase binding to three loader-protein dimers and three primase proteins (Liu et al., 2013). In the E. coli system, the helicase loader DnaC mediates the opening of the DnaB hexamer into a loading-competent cracked open ring according to a `ring-breaker' scenario (Arias-Palomo et al., 2019; Nagata et al., 2020). The EcDnaB·EcDnaC complex is dodecameric, with six subunits of each protein, and the complex assembles into a three-tier spiral. EcDnaC makes contact with EcDnaB through the first small α-helix (15 residues in length) of its extended N-terminal domain (NTD). This helix interacts with a DnaB module composed of the `linker helix' (LH) of one DnaB protomer and an antiparallel α-helix of the adjacent DnaB protomer, which we named the `determinant helix' (DH) in a previous study (Marsin et al., 2021) and which was named the `docking helix' in a recent review comparing the convergent functional mechanisms of EcDnaC and bacteriophage λ P loaders (Chase et al., 2022). This forms a three α-helix bundle which fixes the relative orientation of the two adjacent DnaB C-terminal domains (CTDs): the six DnaC molecules thus latched onto the DnaB hexamer adopt a spiral configuration that causes distortion of the helicase ring, resulting in its large opening as a means to allow single-strand DNA (ssDNA) to enter the helicase pore (Arias-Palomo et al., 2019; Nagata et al., 2020).
| ||||||||||||||||||||||||||||||||||||||||||
Despite this, the DnaC/I loader distribution is marginal in the bacterial domain. It was established phylogenetically that the dnaC/I genes are domesticated phage elements that have replaced the ancestral bacterial gene dciA several times during evolution (Brézellec et al., 2016). DciA and DnaC/I are not related in either their sequence or their structure (Marsin et al., 2021). While the DnaC/I CTD contains an AAA+ ATPase RecA-like domain (Koonin, 1992), the DciA NTD folds as a KH domain (Grishin, 2001; Marsin et al., 2021), also shared by domain I of DnaA and domain V of DnaX, which have both been described to interact with DnaB (Rajathei & Selvaraj, 2013; Jameson et al., 2014; Haroniti et al., 2004; Mann et al., 2017). Through multiple complementary approaches, we previously established that the disordered CTD of DciA can form transiently small helical structures (Chan-Yao-Chong et al., 2020) and is necessary for interacting with DnaB and stimulating the loading of DnaB onto DNA (Chan-Yao-Chong et al., 2020; Marsin et al., 2021). Direct interplay between the two proteins has also been demonstrated using the variation of intrinsic fluorescence of the conserved tryptophan residue located in the middle of the DH of DnaB in the presence of DciA (Marsin et al., 2021). It was therefore suspected that DciA interacts with the helicase near to the DH helix of its module.
To decipher the molecular interactions between DciA and DnaB, we solved the of the DnaB·DciA complex from Vibrio cholerae (Vc) together with ADP and Mg2+, forming a heterotetramer composed of the canonical VcDnaB dimer and two molecules of VcDciA. Interestingly, VcDciA interacts with VcDnaB through the LH–DH module, like EcDnaC and the phage λ P helicase loader on EcDnaB, suggesting a functional link between the different systems. Furthermore, we showed that VcDciA and EcDnaC are interchangeable for in vitro loading of the helicases from V. cholerae and E. coli, suggesting convergent evolution of both helicase-loader systems. However, the VcDnaB·VcDciA·ADP:Mg2+ structure also revealed that DciA binds to the periphery of the helicase CTD, in contrast to other known loaders that oligomerize and are positioned at the back of the DnaB CTD ring, leading to the presumption that its helicase-loading mechanism differs from those of DnaC/I and λ P.
2. Materials and methods
2.1. Protein-sample preparation
VcDnaB, EcDnaB, VcDciA, VcDciA(1–111) and EcDnaC were overexpressed in E. coli and purified as described in Marsin et al. (2021). EcDnaC(53-end) was purified in the same way as EcDnaC. Strains and plasmids are available upon request.
2.2. determination of the VcDnaB·VcDciA·ADP:Mg2+ complex
Purified VcDnaB was pre-incubated for 10 min at 4°C at a concentration of 0.115 mM (monomer) with 2 mM ADP and 5 mM MgCl2. Purified VcDciA was added to a final concentration of 0.138 mM (about seven monomers of DciA per helicase hexamer) before a second step of incubation. Native protein crystals were grown in sitting drops by mixing the protein solution with the reservoir solution in a 1:1 ratio. Rhombohedral crystals of the VcDnaB·VcDciA·ADP:Mg2+ complex appeared after five days at 18°C in 0.1 M sodium acetate pH 4.8–5.6, 0.7–0.9 M potassium/sodium tartrate. For derivatization, single crystals were then soaked for 2 h at 18°C in a solution containing 1 mM (Ta6Br12)2+ cluster (JBS Tantalum Cluster Derivatization Kit from Jena Bioscience GmbH, Jena, Germany). Native crystals cryoprotected with 25% glycerol or derivative crystals cryoprotected with a 50/50 Paratone/paraffin oil mixture were flash-cooled in liquid nitrogen.
Diffraction data-collection, phasing and are given in Table 2. Native and derivative crystallographic data were collected on the PROXIMA-2A and PROXIMA-1 beamlines, respectively, at the SOLEIL synchrotron, Saint-Aubin, France and were processed with XDS (Kabsch, 2010) through XDSME (Legrand, 2017). The strong diffraction anisotropy was corrected using the STARANISO server (https://staraniso.globalphasing.org; Tickle et al., 2018). The of the VcDnaB·VcDciA·ADP:Mg2+ complex was solved by (MR) with MOLREP (Vagin & Teplyakov, 2010) using the X-ray structures of the isolated NTD (22–175) and CTD (200–461) of VcDnaB·GDP:AlF4:Mg2+ as search models (PDB entry 6t66; Marsin et al., 2021). Two copies of each domain were correctly positioned. The initial model was then manually corrected and completed using Coot (Emsley et al., 2010). Significant extra electron density allowed the manual building of the isolated CTDs of two VcDciA monomers, the chains of which could be assigned using the 3D model of full-length VcDciA predicted by AlphaFold2 (Jumper et al., 2021) through the ColabFold server (Mirdita et al., 2022). Additional electron density allowed the manual positioning of the isolated NTDs of two VcDciA monomers using the NMR structure of VcDciA(1–111) solved by Marsin et al. (2021) (BMRB ID 27689). Finally, manual building of the linker regions connecting the NTDs and CTDs of VcDciA revealed domain swapping between symmetry-related molecules of VcDciA. The structure of the VcDnaB·VcDciA·ADP:Mg2+ complex was iteratively improved by manual building steps followed by cycles using native data to 2.9 Å resolution. Model was conducted with BUSTER (Bricogne et al., 2017) using 12 translation–libration–screw (TLS) motion groups, automated (NCS) restraints and local structure similarity restraints (LSSR) to the target models of the VcDnaB·VcDciA complex predicted by AlphaFold2 (Mirdita et al., 2022; Jumper et al., 2021) and RoseTTAFold (Baek et al., 2021).
‡Diffraction data collected were from one crystal which diffracted anisotropically to 3.43 Å resolution along 0.894a* − 0.447b*, to 3.43 Å resolution along b* and to 6.50 Å resolution along c*. | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
To avoid model bias, an experimental electron-density map was obtained at 3.7 Å resolution by single-wavelength anomalous diffraction (SAD) using derivative data collected at the tantalum peak wavelength. The (Ta6Br12)2+ cluster sites were initially found with SHELXD (Schneider & Sheldrick, 2002); the phases were then determined with Phaser (McCoy et al., 2007) and improved by density modification with Parrot (Cowtan, 2010) in the CCP4 suite (Winn et al., 2011). Superimposing the MR model on the experimental map confirmed its accuracy, except for the NTD of the second VcDciA monomer, for which no density was visible, likely due to too sharp solvent flattening. Crystals of the VcDnaB·VcDciA·ADP:Mg2+ complex were analyzed by SDS–PAGE and both VcDnaB and VcDciA were visualized on the gel after Coomassie Blue staining as full-length proteins without any proteolysis (Supplementary Fig. S1). BUSTER (Bricogne et al., 2017) calculated per-residue values for real-space correlation of the final refined model against the 2Fo − Fc map. The NTD of the second VcDciA molecule has an acceptable mean main-chain (RSCC) of about 0.74, although this is a little lower than the mean of about 0.84 for the NTD of the first VcDciA, which is very similar to the overall average of 0.83 for the whole model. This reflects a difference in flexibility between the NTDs of the two VcDciA monomers, which is likely to be due to crystal-packing and domain-swapping constraints.
The structure of the VcDnaB2·VcDciA2 heterotetramer forming the biological assembly can be reconstructed from the domains swapped between the symmetric VcDciA molecules. The resulting unswapped model consists of one VcDnaB dimer interacting with two VcDciA molecules, each formed by one NTD (Met1–Pro98) and one CTD (Glu122–Asp157) from two polypeptide chains related by a true crystallographic twofold rotation axis. Nevertheless, the conformation of the flexible hinge region (Glu99–Ser121) connecting the NTD and the CTD is only putative in these two reconstructed unswapped VcDciA molecules, as Pro98 and Glu99 are no longer linked in this model, and will therefore differ from that in the swapped Finally, the H32 symmetry of the crystal reconstitutes the VcDnaB6·VcDciA6 heterododecameric complex by the assembly of the heterotetramer with two neighboring symmetry mates related by a true crystallographic threefold rotation axis.
All structural figures were prepared using PyMOL (DeLano, 2002).
2.3. Protein interaction analysis by thermal shift assay and intrinsic fluorescence variation
As described in Marsin et al. (2021), intrinsic fluorescence changes of tryptophan (and, at a lower level, tyrosine) were recorded at 330 and 350 nm while heating the protein sample from 35 to 95°C at a rate of 3°C min−1. The emission profile of tryptophan is shifted to the red when it is released into the solvent during the thermal denaturing of the protein. We used Tycho analysis (Tycho NT.6, NanoTemper Technologies GmbH, Munich, Germany) to follow the interaction between VcDnaB or EcDnaB and VcDciA, VcDciA(1–111), EcDnaC or EcDnaC(53-end). Interactions were performed in 50 mM HEPES pH 7.5, 150 mM NaCl, 1 mM ATP, with 20 µM of each protein, in 10 µl capillary tubes. Three to five replicates were obtained to increase the confidence in the results. To detect binding, we compared the 350/330 nm fluorescence ratio of the complex with the predicted ratio that would be obtained in the absence of interaction by additivity of the fluorescence of the proteins alone (Sample at 350 nm/Sample at 330 nm).
2.4. Measurement of protein–DNA interaction by biolayer interferometry (BLI)
BLI experiments were conducted on an Octet RED96e system (Pall ForteBio, Fremont, California, USA) using streptavidin (SA) biosensors. BLI monitors the wavelength shifts (in nanometres) resulting from changes in the optical thickness of the sensor surface during association or dissociation of the analyte. All BLI experiments were performed at 30°C while stirring at 1000 rev min−1. The streptavidin biosensor was hydrated in a 96-well plate containing phosphate-buffered saline (PBS; Bio-Rad) for at least 10 min before each experiment. The 3′-biotinylated oligonucleotide oso13 [50-nucleotide ssDNA at 10 nM; GCAGGCTCGTTACGTAGCTGTACCG(dT)25-biotin] was immobilized in PBS onto the surface of the SA biosensor through a cycle of baseline (120 s), loading (300 s) and baseline (120 s). Association interactions were then monitored for 300 s in wells containing 200 µl sample at 100 nM VcDnaB or EcDnaB with different ratios of the indicated loader in HNATM1 buffer (50 mM HEPES pH 7, 150 mM NaCl, 1 mM ATP, 0.1% Tween 20, 1 mM MgCl2). At the end of each binding step, the sensors were transferred into protein-free HNATM1 binding buffer to follow the dissociation kinetics for 600 s. The sensors can be recycled by dipping them into 0.08% SDS for 10 s. The experiments were carried out in duplicate; only one is presented.
3. Results and discussion
3.1. The of the VcDnaB·VcDciA complex forms a heterotetramer with 2:2 stoichiometry
We have previously demonstrated by functional studies that DciA from V. cholerae increases the loading of VcDnaB onto DNA, resulting in an increased unwinding activity of the helicase (Marsin et al., 2021). To understand the molecular interplay between the two proteins, we determined the of the VcDnaB·VcDciA·ADP:Mg2+ complex (deposited as PDB entry 8a3v). The structure was solved by (see Section 2 and Table 2) using the VcDnaB·GDP:AlF4:Mg2+ (PDB entry 6t66; Marsin et al., 2021), the VcDciA(1–111) NMR structure (BMRB ID 27689; Marsin et al., 2021) and the full-length VcDciA model predicted by AlphaFold2 (Jumper et al., 2021), and was refined to 2.9 Å resolution. The accuracy of the final model was further verified by superimposition on an experimental electron-density map obtained at 3.7 Å resolution by single-wavelength anomalous diffraction (SAD) using derivative data from a (Ta6Br12)2+-cluster-soaked crystal (see Section 2 and Table 2).
The of the crystal contains two molecules of VcDnaB and two molecules of VcDciA. The of the VcDnaB·VcDciA complex revealed domain swapping between symmetry-related molecules of VcDciA (Fig. 1a). The NTD and CTD (N-terminal and C-terminal domains) of two VcDciA molecules connected by an extended hinge region (residues Glu99–Ser121) are exchanged between neighboring molecules related by a true crystallographic twofold rotation axis. It is not known at present whether this domain swapping is due to a crystal artifact or whether it is biologically relevant. However, it is known that replication is bidirectional and therefore two helicases must be recruited to the replication fork (Chodavarapu & Kaguni, 2016; Hayashi et al., 2020). The ability of DciA to link helicases together via this domain swapping could therefore improve the recruitment of two helicases to oriC and thus optimize the replication-initiation step.
![[Figure 1]](jb5052fig1thm.gif) | Figure 1 Crystal structure of the VcDnaB·VcDciA·ADP:Mg2+ complex. (a) Domain-swapped heterooctamer. The of the VcDnaB2·VcDciA2 complex revealed domain swapping between symmetry-related molecules of VcDciA. Left: schematic representation of domain swapping. The NTD and CTD of the two VcDciAs, connected by an extended hinge region (dark blue lines), are exchanged between neighboring molecules related by a true crystallographic twofold rotation axis (red dashed line). The four molecules of VcDnaB are in two shades of blue and green and the four molecules of VcDciA are in pink, purple, orange and brown. The four protein chains of the symmetry mate are marked with a prime. Right: ribbon representation of the heterooctameric structure using the same color code. (b) Structure of the VcDnaB·VcDciA·ADP:Mg2+ heterotetrameric complex forming the unswapped biological assembly, reconstituted from swapped VcDciA domains. Left: schematic representation. The hinges encompassing residues 99–121 of the two VcDciAs are putative in this model (dark blue dotted lines). Right: ribbon representation of the heterotetrameric structure. The VcDnaB dimer is in the same colors as in (a); the two VcDciA molecules are in magenta and yellow. The dark blue dotted rectangle frames the enlarged view shown in (c). (c) Enlargement of the interface region forming a five-helix bundle, with the superimposed experimental electron-density map from SAD phasing after solvent flattening (gray mesh, contoured at 1σ). |
The VcDnaB·VcDciA heterotetramer structure forming the unswapped biological assembly can be reconstructed from the domains swapped between symmetric VcDciA molecules (see Section 2 and Fig. 1b) and exhibits two VcDciA monomers fixed on a VcDnaB dimer. The VcDnaB dimer in the heterotetrameric complex is bound to ADP:Mg2+ and is almost identical to one VcDnaB dimer of the GDP-bound VcDnaB hexamer structure (PDB entry 6t66; Marsin et al., 2021), with an overall r.m.s.d. of 1.66 Å for 858 aligned residues (all-atom r.m.s.d.s of 1 Å for 302 aligned residues of the NTDs and 1.31 Å for 556 aligned residues of the CTDs; Supplementary Fig. S2a). The NTP:Mg2+ binding site in the CTDs is very similar and the P-loop is in the same conformation (Supplementary Fig. S2b). Therefore, formation of the complex with VcDciA does not alter the overall canonical architecture of the VcDnaB dimer or the NTP binding site. Nevertheless, the binding of VcDciA onto the LH–DH module of VcDnaB slightly accentuates the maximum gap between the Cα atoms of the LH and DH helices of about 5 Å relative to the module in the free VcDnaB hexamer structure (Supplementary Fig. S2a). The two VcDciA molecules of the heterotetrameric complex are practically identical to each other (all-atom r.m.s.d.s of 0.5 Å for 78 aligned residues of the NTDs and 0.19 Å for 35 aligned residues of the CTDs), and the NTD of VcDciA exhibits a KH-like fold very similar to that of the VcDciA(1–111) structure (BMRB ID 27689; Marsin et al., 2021) that we obtained by NMR (all-atom r.m.s.d. of 1.2 Å for 78 aligned residues; Supplementary Fig. S2c). However, the first long α1 helix is straight in the first VcDciA molecule but is kinked with an angle of 50° at residue His24 in the second molecule (Supplementary Fig. S2c). This kink can be explained because of with the extended hinge of the neighboring symmetric VcDciA molecule with which it is engaged in domain swapping (VcDciA molecules B and B′ in Fig. 1a). This possibility of bending the long α1 helix of VcDciA was predicted by previous molecular-dynamics analyses (Chan-Yao-Chong et al., 2020). Interestingly, the CTD tail of VcDciA, which was observed to be disordered in solution by SAXS (Marsin et al., 2021), folds into a small helix hairpin in contact with VcDnaB (Supplementary Fig. S2d), again in agreement with our previous molecular-dynamics analyses (Chan-Yao-Chong et al., 2020). The structure of the extended hinge region consisting of the last helix of the NTD (α3 in Supplementary Fig. S2c) and the proline-rich flexible linker, which connects the NTD to the CTD of VcDciA, is only putative in this unswapped biological model reconstructed from the swapped VcDciA domains (see Section 2 and Fig. 1b).
The two VcDciA molecules interact `in trans' with the VcDnaB dimer at the periphery of its CTDs (Fig. 1b). The CTD hairpin helix of one VcDciA molecule stacks entirely on the LH–DH module of VcDnaB (shown in magenta in Fig. 1c). The kinked α2 helix of the NTD of the second VcDciA (shown in yellow in Fig. 1c) interacts with the first helix of the CTD hairpin of the first VcDciA (approximately 238 Å2 `in trans' interaction interface between the CTD and NTD from two different VcDciA molecules, as measured by the PDBePISA server; Krissinel & Henrick, 2007) and also with the DH helix of VcDnaB. The assembly forms a five-helix bundle (Fig. 1c). In addition to the kinked α2 helix, the β3 strand of the VcDciA NTD also interacts at the periphery of the VcDnaB CTD, particularly with the DH helix. This DH helix (determinant/docking helix) is thus at the heart of the interaction between VcDciA and VcDnaB, as proposed by our previous Tycho experiments (Marsin et al., 2021). The overall interaction interface of a VcDciA molecule with a VcDnaB dimer is about 1264 Å2, with about 515 Å2 for the NTD of VcDciA and about 750 Å2 for its CTD (as measured by the PDBePISA server; Krissinel & Henrick, 2007). The structure of the complex therefore confirmed the essential role played by the CTD of VcDciA in the interaction with VcDnaB as shown by our previous ITC experiments (Chan-Yao-Chong et al., 2020). Finally, it should be noted that the NTD of the VcDciA molecule with a kinked α1 helix is rather poorly defined in the density, which can be explained by the fact that its interaction interface with the CTD of VcDnaB is only about 347 Å2, compared with 515 Å2 for the VcDciA NTD with a straight α1 helix. This binding difference between the NTDs of the two VcDciAs is likely to be caused by different steric constraints due to crystal packing and/or related to domain swapping.
3.2. Both the VcDciA and EcDnaC loaders target the conserved LH–DH module of DnaB helicases and are functionally interchangeable in vitro
The LH–DH module is conserved in DnaB helicases (Chase et al., 2022). However, we have previously identified a residue in the DH helix that discriminates DciA helicases from DnaC/I helicases (a serine and a glycine, respectively; Marsin et al., 2021; Fig. 2a). Moreover, the VcDciA and EcDnaC loaders have no sequence or structural similarity. Yet, the two loaders target the same binding site on the helicase: the conserved LH–DH module of DnaB, albeit with differences in the interaction interfaces (Fig. 2b). This interaction involves a single small α-helix in the NTD of EcDnaC, which forms a three-helix bundle, whereas in VcDciA the CTD helix hairpin and the elbow of the kinked α2 helix of NTD both participate in forming a five-helix bundle. We may wonder at this stage whether this common target on DnaB, the LH–DH module, allows cross-talk between the two helicase-loader systems, despite their specificities.
![[Figure 2]](jb5052fig2thm.gif) | Figure 2 VcDciA and EcDnaC target the same binding site on DnaB helicases. (a) Sequence alignment of the LH and DH helices of VcDnaB and EcDnaB (generated by the EMBL–EBI Clustal Omega server; https://www.ebi.ac.uk/Tools/msa/clustalo/; Sievers et al., 2011 ) and displayed using the ESPript 3.0 server (https://espript.ibcp.fr; Robert & Gouet, 2014). The conserved residues are in white on a red background. An orange asterisk marks the conserved tryptophan residues, while a black asterisk marks the specific serine/glycine residues in the DH helix. (b) Close-up view of the interaction interface forming a three- or five-helix bundle between the two-helix LH–DH module of DnaB (blue and green) and VcDciA (top; magenta and yellow; PDB entry 8a3v; this study) or EcDnaC NTD (bottom; magenta; PDB entry 6qel), respectively. The tryptophans whose intrinsic fluorescence variation was measured in the Tycho experiments are shown in orange sticks. (c, d) Tycho NT.6 analysis. The emission profile of a tryptophan is shifted to the red when it is released to the solvent during thermal denaturing of the protein. The 350/330 nm ratios measured for the different helicase-loader mixtures are reported in red and the predicted ratio in the absence of interaction is reported in black. The ratio comparisons are reported for each helicase-loader couple indicated, namely VcDnaB (c) or EcDnaB (d) with two constructs of VcDciA or EcDnaC. The curves correspond to the mean ± SEM of three to five analyses. The Tycho interaction analysis between VcDnaB and VcDciA has previously been published (Marsin et al., 2021) but is reproduced here (indicated by *) for easy evaluation with other helicase-loader couples. |
Using Tycho nano-DSF technology, the fluorescence variation of tryptophan residues can be followed in real time along a thermal ramp (see Section 2). Ideally, a conserved tryptophan is located in the DH helix, namely Trp291 in VcDnaB and Trp294 in EcDnaB (Figs. 2a and 2b; Marsin et al., 2021). A second conserved tryptophan is located in the globular head of the NTD domain of DnaB (positions 45 and 48 in VcDnaB and EcDnaB, respectively), but is buried and cannot participate in any protein–protein interactions. VcDciA does not contain any tryptophans in its sequence. We have previously demonstrated the binding of VcDciA in the proximity of the LH–DH module of VcDnaB using this technology (Marsin et al., 2021), showing that Trp291 is inaccessible to solvent when VcDnaB interacts with VcDciA. We further showed that VcDciA with its CTD deleted [VcDciA(1–111)] can no longer bind VcDnaB (Fig. 2c, left), as would be expected if the CTD of VcDciA covers the DH α-helix of VcDnaB (Fig. 2b). On the other hand, EcDnaB contains a third nonconserved tryptophan residue (Trp457) that is solvent-exposed in its CTD and is at the interface with the loader in the structure of the EcDnaB·EcDnaC complex (PDB entry 6qel; Arias-Palomo et al., 2019). EcDnaC also encloses three solvent-exposed tryptophan residues in its sequence, with one in its NTD extended end at the interface with the helicase in the structure of the EcDnaB·EcDnaC complex (Trp32 in Fig. 2b; PDB entry 6qel; Arias-Palomo et al., 2019) and two in its CTD. All of these tryptophan residues could therefore participate in protein–protein interactions. Indeed, Tycho allowed confirmation of the interaction between EcDnaB and EcDnaC by showing a lower initial 350/330 nm ratio for the measured curve (in red; Fig. 2d, left) compared with the theoretical curve representing the absence of solvent protection (in black; Fig. 2d, left). In addition, EcDnaC with its NTD deleted [EcDnaC(53-end)] can no longer bind EcDnaB (Fig. 2d, left), as would be expected if the α1 helix of EcDnaC forms a three-helix bundle with the LH–DH module of EcDnaB (Fig. 2b). These findings are in agreement with the 3D structures of the two complexes: the interaction takes place at the DH of DnaB and requires the CTD for DciA and the NTD for DnaC. We further carried out crossover experiments using the heterologous VcDnaB/EcDnaC and EcDnaB/VcDciA systems. We showed efficient noncognate helicase-loader interactions, which also require the NTD of EcDnaC (Fig. 2c, right) and the CTD of VcDciA (Fig. 2d, right).
We then investigated whether these interactions are relevant for the stimulation of helicase loading onto DNA by the loaders. We attached a 3′-biotinylated 50-mer ssDNA to a streptavidin-coated probe to measure interactions using biolayer interferometry (BLI; see Section 2). We monitored interactions between immobilized ssDNA and the helicases in real time in the presence of different concentrations of loaders (Fig. 3). The BLI experiments confirmed the results previously observed by SPR (Marsin et al., 2021), namely that the loading stimulation of both helicases increases with the concentration of added cognate loader, VcDciA or EcDnaC (Figs. 3a and 3b, left), at concentrations for which the response for loaders alone is negligible (Supplementary Fig. S3). Moreover, under the same conditions EcDnaC is able to efficiently load VcDnaB onto the ssDNA, and VcDciA is able to load EcDnaB, although less effectively than EcDnaC (Figs. 3a and 3b, right). The cross-talk is verified in vitro and there seems to be a functional convergence between the two systems. This could explain why the replacement of DciA by DnaC/I has occurred at least seven times during evolution, and how phage loaders have been able to hijack bacterial replicative helicases efficiently (Brézellec et al., 2016).
![[Figure 3]](jb5052fig3thm.gif) | Figure 3 VcDciA and EcDnaC loaders are functionally interchangeable in vitro. Biolayer interferometry (BLI) analysis using a biotinylated oligonucleotide (50 nucleotides) immobilized onto the surface of an SA-coated probe by its 3′ extremity (see Section 2 ). Association was performed with the indicated helicase at a concentration of 100 nM during 300 s in a buffer solution containing ATP and MgCl2. Dissociation was assessed in the same buffer for 600 s. Increasing loader concentrations (from 0 to 200 nM in subunits; blue to red) were analyzed. The experiments were carried out in duplicate; only one is presented. (a) VcDnaB binding on ssDNA in the presence of VcDciA (left) or EcDnaC (right). (b) EcDnaB binding on ssDNA in the presence of EcDnaC (left) or VcDciA (right). |
3.3. VcDciA binds to the periphery of the VcDnaB CTD, in contrast to other loaders, which are positioned at the back of the helicase CTD ring
Our previous SEC–SAXS and SEC–MALS experiments showed that a complex between the VcDnaB hexamer and VcDciA is formed in solution under specific in vitro conditions with a predominant 6:3 stoichiometry (Marsin et al., 2021). However, this complex is not stable and tends to dissociate, showing that it is dynamic in solution, probably due to rapid molecular exchanges leading to a mixture of several conformational states (Marsin et al., 2021). In the current study, the experimental conditions were optimized to compare the V. cholerae and E. coli helicase-loader systems and the BLI curves did not reach a saturation plateau even at a ratio of two VcDciA molecules to one VcDnaB molecule (Fig. 3a, left). Interestingly, the H32 symmetry of the crystal reconstitutes a VcDnaB6·VcDciA6 heterododecamer complex by the assembly of the heterotetramer with two neighboring symmetry mates related by a true crystallographic threefold rotation axis (Fig. 4a). Of course, a crystallographic structure captures only one of the possible conformational intermediates, with a 2:2 or 6:6 stoichiometry in our case, and it is not certain that the current structure is that of an active state. Further work remains to be performed to determine whether or not this heterododecameric structure is biologically relevant, but in the meantime we can compare it with other helicase-loader structures in the literature. Three 3D structures of helicase-loader complexes are currently known (see Table 1): EcDnaB·EcDnaC (PDB entry 6qel; Arias-Palomo et al., 2019), EcDnaB·λP (PDB entry 6bbm; Chase et al., 2018) and GstDnaB·BsDnaI (PDB entry 4m4w; Liu et al., 2013). Strikingly, VcDciA binds to the periphery of the VcDnaB CTD, in contrast to the other three loaders, which oligomerize and are positioned at the back of the helicase CTD ring (Fig. 4b). This discrepancy leads us to consider that the loading mechanism used by DciA, which has still to be elucidated, could differ from those previously described for the other three loaders. However, it is possible that a partner, a protein or a nucleic acid, remains to be discovered in order to fully decipher the function of DciA.
![[Figure 4]](jb5052fig4thm.gif) | Figure 4 DciA binds to the periphery of the DnaB CTD. (a) Structure of the VcDnaB6·VcDciA6 heterododecameric complex reconstituted by the crystal H32 symmetry. The color code is the same as in Fig. 1 (b). Left: schematic representation. The heterododecameric ring is reconstituted by assembly of the heterotetramer with two neighboring symmetry mates (hatched textures) related by a true crystallographic threefold rotation axis (red dashed lines). The hinges encompassing residues 99–121 of the VcDciA molecules are putative in this unswapped model (dark blue dotted lines). Right: ribbon representation of the heterododecameric model. (b) Structural comparison of four helicase-loader complexes: VcDnaB·VcDciA (PDB entry 8a3v, this study), EcDnaB·EcDnaC (PDB entry 6qel), EcDnaB·λP (PDB entry 6bbm) and GstDnaB·BsDnaI (extracted from PDB entry 4m4w). The DnaB hexamers are represented as surfaces (blue and green) and the helicase loaders as magenta sticks. Unlike DnaC, λP and DnaI, which cover the back of the DnaB CTD ring, DciA leaves it free by positioning itself at the periphery of the helicase. |
4. Conclusions
The genes coding for DciA and DnaC/I are unrelated and are mutually exclusive in bacterial genomes (Brézellec et al., 2017; Brézellec et al., 2016). However, like DnaC and the bacteriophage λ P helicase loader (Chase et al., 2018, 2022; Arias-Palomo et al., 2019), DciA interacts with the two-helix LH–DH module of DnaB. It is not yet known whether DnaI also targets the LH–DH module of the replicative helicase. In the available low-resolution of the GstDnaB·BsDnaI complex (PDB entry 4m4w; Liu et al., 2013), several parts of the NTD end of DnaI, which have already been identified to interact with DnaB (Loscha et al., 2009; Tsai et al., 2009), as well as the LH of the DnaB module (which is also conserved in GstDnaB; PDB entries 2r6a and 4esv; Bailey et al., 2007; Itsathitphaisarn et al., 2012), are not visible, preventing us from definitively concluding that there is a common binding site for the various helicase loaders. However, AlphaFold-Multimer (Evans et al., 2022; Mirdita et al., 2022) predicts an interaction between the NTD of DnaI and the LH–DH module of DnaB (Supplementary Fig. S4), making a universal site of interaction plausible. This `binding module', which is conserved in the bacterial replicative helicases, as well as the fact that cross-talk reactions are efficient in vitro between the DciA and DnaC systems, suggest convergent evolution of the different helicase-loader systems. Nevertheless, the structural data on the VcDnaB·VcDciA·ADP:Mg2+ complex provided here do not permit us to postulate that the loading mechanism used by DciA will be of the `ring-breaker' or `ring-maker' type, or even of a third type which remains to be elucidated. This conformational state may be an inactive intermediate state before its activation by a third partner that has yet to be discovered. A recent computational evolutionary study showed that DciA homologs exhibit a tremendous diversity of domain architectures across bacterial phyla (Blaine et al., 2022). Notably, one group of DciA homologs only encodes a KH-like fold and no other N- or C-terminal extensions, suggesting that an additional partner may indeed be required for DciA function. Thus, future investigations will probably uncover a cofactor, a protein or a nucleic acid, that is necessary for DciA to complete its helicase-loading cycle.
Supporting information
PDB reference: VcDnaB·VcDciA, 8a3v
Supplementary Figures. DOI: https://doi.org/10.1107/S2059798323000281/jb5052sup1.pdf
Acknowledgements
X-ray diffraction data were collected on the PROXIMA-2A and PROXIMA-1 beamlines at the SOLEIL synchrotron, Saint-Aubin, France. We thank the beamline staff for their assistance and advice during data collection. This work has benefited from the expertise of the Macromolecular Interaction Measurements Platform and the Crystallization Platform of I2BC, Gif-sur-Yvette, France and we thank Magali Aumont-Nicaise and Stéphane Plancqueel, respectively, for their assistance. We are grateful to Dominique Durand for her ongoing support. We thank Jean-Luc Ferat for the initiating discussions.
Funding information
This work was supported by the French Infrastructure for Integrated Structural Biology (FRISBI) ANR-10-INBS-05, by funds from the Centre National de la Recherche Scientifique (CNRS) and by grants from the Region Ile de France. CC was supported by a PhD fellowship from the French Ministry of Education.
References
Arias-Palomo, E., Puri, N., O'Shea Murray, V. L., Yan, Q. & Berger, J. M. (2019). Mol. Cell, 74, 173–184. CAS PubMed Google Scholar
Baek, M., DiMaio, F., Anishchenko, I., Dauparas, J., Ovchinnikov, S., Lee, G. R., Wang, J., Cong, Q., Kinch, L. N., Schaeffer, R. D., Millán, C., Park, H., Adams, C., Glassman, C. R., DeGiovanni, A., Pereira, J. H., Rodrigues, A. V., van Dijk, A. A., Ebrecht, A. C., Opperman, D. J., Sagmeister, T., Buhlheller, C., Pavkov-Keller, T., Rathinaswamy, M. K., Dalwadi, U., Yip, C. K., Burke, J. E., Garcia, K. C., Grishin, N. V., Adams, P. D., Read, R. J. & Baker, D. (2021). Science, 373, 871–876. Web of Science CrossRef CAS PubMed Google Scholar
Bailey, S., Eliason, W. K. & Steitz, T. A. (2007). Science, 318, 459–463. CrossRef PubMed CAS Google Scholar
Blaine, H. C., Burke, J. T., Ravi, J. & Stallings, C. L. (2022). J. Bacteriol. 204, e00163-22. CrossRef PubMed Google Scholar
Brézellec, P., Petit, M. A., Pasek, S., Vallet-Gely, I., Possoz, C. & Ferat, J.-L. (2017). Genome Biol. 9, 1561–1566. Google Scholar
Brézellec, P., Vallet-Gely, I., Possoz, C., Quevillon-Cheruel, S. & Ferat, J.-L. (2016). Nat. Commun. 7, 13271. PubMed Google Scholar
Bricogne, G., Blanc, E., Brandl, M., Flensburg, C., Keller, P., Paciorek, W., Roversi, P., Sharff, A., Smart, O. S., Vonrhein, C. & Womack, T. O. (2017). BUSTER. Global Phasing Ltd, Cambridge, United Kingdom. Google Scholar
Chan-Yao-Chong, M., Marsin, S., Quevillon-Cheruel, S., Durand, D. & Ha-Duong, T. (2020). J. Struct. Biol. 212, 107573. PubMed Google Scholar
Chase, J., Berger, J. & Jeruzalmi, D. (2022). Trends Biochem. Sci. 47, 620–630. CrossRef CAS PubMed Google Scholar
Chase, J., Catalano, A., Noble, A. J., Eng, E. T., Olinares, P. D., Molloy, K., Pakotiprapha, D., Samuels, M., Chait, B., des Georges, A. & Jeruzalmi, D. (2018). eLife, 7, e41140. CrossRef PubMed Google Scholar
Chodavarapu, S. & Kaguni, J. (2016). Enzymes, 39, 1–30. CrossRef CAS PubMed Google Scholar
Costa, A., Hood, I. V. & Berger, J. M. (2013). Annu. Rev. Biochem. 82, 25–54. CrossRef CAS PubMed Google Scholar
Cowtan, K. (2010). Acta Cryst. D66, 470–478. Web of Science CrossRef CAS IUCr Journals Google Scholar
Davey, M. J. & O'Donnell, M. (2003). Curr. Biol. 13, R594–R596. CrossRef PubMed CAS Google Scholar
DeLano, W. L. (2002). PyMOL. https://www.pymol.org. Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. (2010). Acta Cryst. D66, 486–501. Web of Science CrossRef CAS IUCr Journals Google Scholar
Evans, R., O'Neill, M., Pritzel, A., Antropova, N., Senior, A., Green, T., Žídek, A., Bates, R., Blackwell, S., Yim, J., Ronneberger, O., Bodenstein, S., Zielinski, M., Bridgland, A., Potapenko, A., Cowie, A., Tunyasuvunakool, K., Jain, R., Clancy, E., Kohli, P., Jumper, J. & Hassabis, D. (2022). bioRxiv, 2021.10.04.463034. Google Scholar
Grishin, N. V. (2001). Nucleic Acids Res. 29, 638–643. Web of Science CrossRef PubMed CAS Google Scholar
Haroniti, A., Anderson, C., Doddridge, Z., Gardiner, L., Roberts, C. J., Allen, S. & Soultanas, P. (2004). J. Mol. Biol. 336, 381–393. CrossRef PubMed CAS Google Scholar
Hayashi, C., Miyazaki, E., Ozaki, S., Abe, Y. & Katayama, T. (2020). J. Biol. Chem. 295, 11131–11143. CrossRef CAS PubMed Google Scholar
Itsathitphaisarn, O., Wing, R. A., Eliason, W. K., Wang, J. & Steitz, T. A. (2012). Cell, 151, 267–277. Web of Science CrossRef CAS PubMed Google Scholar
Jameson, K. H., Rostami, N., Fogg, M. J., Turkenburg, J. P., Grahl, A., Murray, H. & Wilkinson, A. J. (2014). Mol. Microbiol. 93, 975–991. CrossRef CAS PubMed Google Scholar
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., Bridgland, A., Meyer, C., Kohl, S. A. A., Ballard, A. J., Cowie, A., Romera-Paredes, B., Nikolov, S., Jain, R., Adler, J., Back, T., Petersen, S., Reiman, D., Clancy, E., Zielinski, M., Steinegger, M., Pacholska, M., Berghammer, T., Bodenstein, S., Silver, D., Vinyals, O., Senior, A. W., Kavukcuoglu, K., Kohli, P. & Hassabis, D. (2021). Nature, 596, 583–589. Web of Science CrossRef CAS PubMed Google Scholar
Kabsch, W. (2010). Acta Cryst. D66, 125–132. Web of Science CrossRef CAS IUCr Journals Google Scholar
Koonin, E. V. (1992). Nucleic Acids Res. 20, 1997. CrossRef PubMed Google Scholar
Krissinel, E. & Henrick, K. (2007). J. Mol. Biol. 372, 774–797. Web of Science CrossRef PubMed CAS Google Scholar
Legrand, P. (2017). XDSME: XDS Made Easier.https://github.com/legrandp/xdsme. Google Scholar
Leonard, A. C. & Méchali, M. (2013). Cold Spring Harb. Perspect. Biol. 5, a010116. CrossRef PubMed Google Scholar
Liu, B., Eliason, W. K. & Steitz, T. A. (2013). Nat. Commun. 4, 2495. CrossRef PubMed Google Scholar
Loscha, K. V., Jaudzems, K., Ioannou, C., Su, X. C., Hill, F. R., Otting, G., Dixon, N. E. & Liepinsh, E. (2009). Nucleic Acids Res. 37, 2395–2404. CrossRef PubMed CAS Google Scholar
Mann, K. M., Huang, D. L., Hooppaw, A. J., Logsdon, M. M., Richardson, K., Lee, H. J., Kimmey, J. M., Aldridge, B. B. & Stallings, C. L. (2017). PLoS Genet. 13, e1007115. CrossRef PubMed Google Scholar
Marsin, S., Adam, Y., Cargemel, C., Andreani, J., Baconnais, S., Legrand, P., Li de la Sierra-Gallay, I., Humbert, A., Aumont-Nicaise, M., Velours, C., Ochsenbein, F., Durand, D., Le Cam, E., Walbott, H., Possoz, C., Quevillon-Cheruel, S. & Ferat, J.-L. (2021). Nucleic Acids Res. 49, 6569–6586. CrossRef CAS PubMed Google Scholar
McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C. & Read, R. J. (2007). J. Appl. Cryst. 40, 658–674. Web of Science CrossRef CAS IUCr Journals Google Scholar
Mirdita, M., Schütze, K., Moriwaki, Y., Heo, L., Ovchinnikov, S. & Steinegger, M. (2022). Nat. Methods, 19, 679–682. Web of Science CrossRef CAS PubMed Google Scholar
Nagata, K., Okada, A., Ohtsuka, J., Ohkuri, T., Akama, Y., Sakiyama, Y., Miyazaki, E., Horita, S., Katayama, T., Ueda, T. & Tanokura, M. (2020). J. Biochem. 167, 1–14. CrossRef CAS PubMed Google Scholar
O'Donnell, M., Langston, L. & Stillman, B. (2013). Cold Spring Harb. Perspect. Biol. 5, a010108. PubMed Google Scholar
Rajathei, D. M. & Selvaraj, S. (2013). Comput. Biol. Chem. 47, 156–166. PubMed Google Scholar
Robert, X. & Gouet, P. (2014). Nucleic Acids Res. 42, W320–W324. Web of Science CrossRef CAS PubMed Google Scholar
Schneider, T. R. & Sheldrick, G. M. (2002). Acta Cryst. D58, 1772–1779. Web of Science CrossRef CAS IUCr Journals Google Scholar
Sievers, F., Wilm, A., Dineen, D., Gibson, T. J., Karplus, K., Li, W., Lopez, R., McWilliam, H., Remmert, M., Söding, J., Thompson, J. D. & Higgins, D. G. (2011). Mol. Syst. Biol. 7, 539. Web of Science CrossRef PubMed Google Scholar
Strycharska, M. S., Arias-Palomo, E., Lyubimov, A. Y., Erzberger, J. P., O'Shea, V. L., Bustamante, C. J. & Berger, J. M. (2013). Mol. Cell, 52, 844–854. CrossRef CAS PubMed Google Scholar
Tickle, I. J., Flensburg, C., Keller, P., Paciorek, W., Sharff, A., Vonrhein, C. & Bricogne, G. (2018). STARANISO. Global Phasing Ltd, Cambridge, United Kingdom. Google Scholar
Tsai, K.-L., Lo, Y.-H., Sun, Y.-J. & Hsiao, C.-D. (2009). J. Mol. Biol. 393, 1056–1069. CrossRef PubMed CAS Google Scholar
Vagin, A. & Teplyakov, A. (2010). Acta Cryst. D66, 22–25. Web of Science CrossRef CAS IUCr Journals Google Scholar
Velten, M., McGovern, S., Marsin, S., Ehrlich, S. D., Noirot, P. & Polard, P. (2003). Mol. Cell, 11, 1009–1020. Web of Science CrossRef PubMed CAS Google Scholar
Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G. W., McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A. & Wilson, K. S. (2011). Acta Cryst. D67, 235–242. Web of Science CrossRef CAS IUCr Journals Google Scholar
Zawilak-Pawlik, A., Nowaczyk, M. & Zakrzewska-Czerwińska, J. (2017). Genes, 8, 136. PubMed Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.
| STRUCTURAL BIOLOGY |
access
