Structural insights into the methyl donor recognition model of a novel membrane-binding protein UbiG

UbiG is a SAM-dependent O-methyltransferase, catalyzing two O-methyl transfer steps for ubiquinone biosynthesis in Escherichia coli. UbiG possesses a unique sequence insertion between β4 and α10, which is used for membrane lipid interaction. Interestingly, this sequence insertion also covers the methyl donor binding pocket. Thus, the relationship between membrane binding and entrance of the methyl donor of UbiG during the O-methyl transfer process is a question that deserves further exploration. In this study, we reveal that the membrane-binding region of UbiG gates the entrance of methyl donor. When bound with liposome, UbiG displays an enhanced binding ability toward the methyl donor product S-adenosylhomocysteine. We further employ protein engineering strategies to design UbiG mutants by truncating the membrane interacting region or making it more flexible. The ITC results show that the binding affinity of these mutants to SAH increases significantly compared with that of the wild-type UbiG. Moreover, we determine the structure of UbiG∆165–187 in complex with SAH. Collectively, our results provide a new angle to cognize the relationship between membrane binding and entrance of the methyl donor of UbiG, which is of benefit for better understanding the O-methyl transfer process for ubiquinone biosynthesis.

structural information, the methyl donor recognition model of UbiG remains unclear. Furthermore, the significance of the membrane-binding ability of UbiG in the O-methyl transfer process for ubiquinone biosynthesis is still worth exploring.
Here, we construct an UbiG mutant (UbiG∆ 165-187 ) by deleting the sequence insertion that covers the methyl donor binding pocket. The binding affinity of UbiG∆ 165-187 to SAH is approximately 58-fold higher than that of wild-type UbiG. Moreover, both wild-type UbiG bound to liposome and UbiG mutants that weaken the interaction of this sequence insertion with the core component show an enhanced binding ability toward SAH. Finally, we solve the crystal structure of UbiG∆ 165-187 complexed with SAH at 2.10 Å. Taken together, our results uncover  (compound 4). ITC profile of SAH titrated against wild-type UbiG (B) and liposome-bound UbiG (C). The upper panels showed the raw ITC data for injection of ligands into the sample cell containing wild-type UbiG or liposome-bound UbiG. The peaks were normalized to the ligand: protein molar ratio, and were integrated as shown in the bottom panels. Solid dots indicated the experimental data, and their best fit was obtained from a nonlinear least squares method, using a one-site binding model depicted by a continuous line. the methyl donor diffusion mechanism of UbiG, and reveal that the membrane association of UbiG may regulate the entrance of methyl donor, which suggests an inextricable relationship between membrane anchoring and O-methyl transfer reaction in the ubiquinone biosynthesis pathway.

Results and Discussion
UbiG bound with liposome displays an enhanced binding ability toward SAH. Our previous results have reported the crystal structure of UbiG from E. coli, and identified the residues vital for membrane binding. Interestingly, these residues mainly locate in helix α 9 and loop α 9/α 10, a region that covers the possible methyl donor binding pocket 15 . Moreover, to gain insight into the methyl donor recognition model of UbiG, we tried to determine the complex structure of UbiG with SAH. However, we failed to obtain the complex structure by either co-crystallization or crystal soaking. To investigate whether the membrane association of UbiG influences the diffusion of methyl donor, we compare the binding affinity of wild-type UbiG and liposome-bound UbiG to SAH ( Table 2). The ITC experiments show that wild-type UbiG bound SAH with a K d of 104.43 ± 17.21 μM (Fig. 1B), whereas the affinity of liposome-bound UbiG to SAH (K d = 9.63 ± 2.10 μM) increased ≈ 11-fold (Fig. 1C), indicating that the membrane association promotes UbiG interacting with SAH.

Structure of UbiG∆ 165-187 in complex with SAH.
To disclose the accurate recognition pattern of SAH, we crystallized UbiG∆ 165-187 in complex with SAH at a resolution of 2.10 Å. The details of the data collection and refinement statistics are summarized in Table 1. The final model contains one molecule of UbiG∆ 165-187 and one molecule of SAH, with a stoichiometry of 1:1. Due to the insufficient electron density, the N-terminal 9 residues could not be traced. UbiG∆ 165-187 displays a similar fold as wild-type UbiG (Fig. 3A). The overall main-chain root-mean squared deviation (RMSD) between UbiG∆ 165-187 and wild-type UbiG is 0.397 Å for 215 comparable Cα atoms. Comparison with the structure of wild-type UbiG, helix α 1 of UbiG∆ 165-187 moves toward the SAH binding pocket and forms extensive hydrophobic interactions with the carbon-skeleton of SAH (Fig. 3B). In addition, due to the lack of the hydrophobic packing with helix α 8, the β 6 and β 7 of UbiG∆ 165-187 move away from the core structure (Fig. 3B).
The electron density for the SAH is well defined in the final model of UbiG∆ 165-187 and the SAH is bound via an extensive hydrogen bond network and hydrophobic interaction. In light of the structure, we easily identify the SAH binding sites. The adenine ring of SAH is located in a hydrophobic pocket constituted by residues Val 12 , Ile 17 ,  (Fig. 3C). The ribosyl moiety is anchored via hydrogen bonds from the O2′ and O3′ hydroxyl groups to the side chain of Asp 85 (Fig. 3C). The SAH carboxyl is locked by the side chain of Arg 44 , whereas the corresponding SAH amine is anchored to the main chain carbonyl oxygen atoms of Gly 64 and Met 129 via hydrogen bonds (Fig. 3C).
The methyl donor binding model and diffusion mechanism of UbiG. Superimposition of the structures of wild-type UbiG and UbiG∆ 165-187 in complex with SAH, we map the SAH binding model of wild-type UbiG. As shown in Fig. 4(A), SAH is situated in the central cavity of the Rossmann-fold domain of UbiG. The interaction between UbiG and SAH can be divided into three parts in accordance to the moieties of SAH. For the adenine moiety, hydrophobic residues Met 86 , Met 131 , Val 135 , Met 180 , Val 181 and Pro 182 make extensive van der Waals interactions with the adenine ring (Fig. 4A). For the ribosyl moiety, the side chain of Asp 85 forms two hydrogen bonds with the O2′ and O3′ hydroxyl groups (Fig. 4A). The interaction between UbiG and the homocysteine moiety of SAH is dominated by four hydrogen bonds. The side-chain of Arg 44 contributes to two hydrogen bonds with the amino group of the homocysteine (Fig. 4A). The carboxyl group of the homocysteine makes another two hydrogen bonds with the main-chain carbonyl oxygen atoms of Gly 64 and Met 129 , respectively (Fig. 4A). Then, we used the program CAVER to analyse the diffusion pathway of the methyl donor, which revealed a tunnel gated by residues Met 86 , Thr 111 , Glu 113 , Pro 136 , Asp 137 , Ser 140 , and Pro 182 (Fig. 4B). This gate seems much narrow compared with that of most other class I SAM-MTases, such as catechol O-methyltransferase COMT (PDB code 1VID) 16 , rebeccamycin sugar 4′ -O-Methyltransferase RebM (PDB code 3BUS) 17 , and 2-methoxy-6-polyprenyl-1, 4-benzoquinone 5′ -C-methyltransferase Coq5 (PDB code 4OBX) 18 , in which the methyl donor binding pocket is uncovered.
Combining with the ITC results mentioned above, we conclude that in the membrane-unbound state, the diffusion of methyl donor of UbiG is greatly affected by the narrow gate constituted by the membrane binding region. When UbiG associates with the membrane, strong hydrophobic driving forces may loosen the interaction of this membrane binding region with the core structure, and cause a relatively open channel for the diffusion of methyl donor during the O-methyl transfer process for ubiquinone biosynthesis (Fig. 4C). Association of membrane-bound proteins with the surface of cellular membranes usually plays a necessary role for a large variety of cellular functions. For example, the cytoskeleton uses the lipid-binding domain for directly anchoring to the membrane surface 19 . Bin-Amphiphysin-Rvs (BAR) domain containing proteins bind to the membrane surface to act as membrane shapers 20 . The attaching of alpha-toxin to membrane surface pushes the opening of the active center, which is help for hydrolysis of membrane phospholipids 21,22 . As we known, the O-methyl transfer reaction for ubiquinone biosynthesis catalyzed by UbiG is membrane associate in vivo 14 . Obviously, the membrane anchoring ability of UbiG is of benefit for sequestering substrates located in the lipid bilayer. In this study, we find surprisingly that the membrane association of UbiG also regulates the entrance of methyl donor, thus activating the O-methyl transfer reaction for ubiquinone biosynthesis. Our results provide much insight into the role of membrane association in regulating the enzyme activity of UbiG, and enhance our better understanding of the O-methyl transfer process for ubiquinone biosynthesis in vivo.

Materials and Methods
Cloning, expression and purification. Full-length UbiG from E. coli was expressed and purified as described previously 23 . UbiG mutants was generated by PCR with the MutanBEST Kit (TaKaRa) using the parent expression plasmid pET28a-UbiG (1-240) as template. The mutant plasmids were confirmed by DNA sequencing (Invitrogen). Plasmids containing the confirmed UbiG mutations were then transformed into E. coli BL21 (DE3) strain (Novagen), and the corresponding overproduced recombinant mutant proteins were purified as described for the wild-type UbiG 23 .
Crystallization, data collection and processing. Crystallization trials were conducted using the hanging drop vapour diffusion method at 287 K. The protein UbiG∆ 165-187 was concentrated to approximately 16 mg/ml. The UbiG∆ 165-187 -SAH complex was prepared by mixing UbiG∆ 165-187 with SAH at a 1:3 molar ratio. Diffraction quality crystals of UbiG∆ 165-187 -SAH complex were obtained with 0.1 M citric acid pH 5.0 and 20% (v/v) 2-Methyl-2,4-pentanediol. For data collection, the crystals were cryo-protected using 25% (v/v) glycerol supplemented with crystallization solution, and flashed cool in liquid nitrogen. Diffraction data sets for the UbiG∆ 165-187 -SAH complex were collected on beamline 19U of the Shanghai Synchrotron Radiation Facility (SSRF) using a CCD detector. All frames were collected at 100 K using a 1° oscillation angle with an exposure time of 0.2 s per frame. The crystal-to-detector distance was set to 250 mm. The complete diffraction datasets were subsequently processed using HKL-2000 24 and programs in CCP4 package 25 .
To capture an open state of UbiG, we prepared UbiG-phosphatidylglycerol (PG) complex by mixing 16 mg/ml protein with PG in a molecular ratio of 1:3 ∼ 1:10. Crystallization screens were performed with a Mosquito liquid-handling robot (TTP LabTech) using the vapour-diffusion method in 96-well crystallization plates at 289 K. We also tried to screen UbiG for other crystal morphologies as an alternative. However, both of these attempts were failed.
Structure determination and refinement. The complex structure of the UbiG∆ 165-187 -SAH was solved using the molecular replacement method in Molrep 26 , using the structure of the full-length UbiG from E. coli K12 (PDB code 4KDC) as the search model. The model was refined at 2.10 Å resolution using Refmac5 27 and COOT 28 by manual model correction. The structure factors refinement were converged to an R-factor of 17.63% and R-free of 21.52%. These final models were both evaluated with the programs MOLPROBITY 29 and PROCHECK 30 . The data collection and structure refinement statistics were listed in Table 1. All structure figures were created using the program PyMol (DeLano Scientific LLC).
Liposome preparation. The total lipid extract of E. coli (Avanti Polar Lipids, Inc) was used to generate liposomes that mimic the component of the E. coli plasma membrane. For liposome preparation, the total lipid extract were dissolved in chloroform in a glass tube and then was evaporated under a stream of nitrogen for 20 minutes. Next, the lipid films were dried with a vacuum pump overnight and then were hydrated at room temperature with constant mixing in buffer (20 mM Tris-HCl, 50 mM NaCl, pH 7.5). After hydration, lipid vesicles were subjected to freeze-thaw cycles in liquid nitrogen and a room temperature water bath, and then sized using Mini-Extruder Set (Avanti) with 100 nm polycarbonate filters.

Isothermal titration calorimetry (ITC) experiments.
The ITC binding studies were performed using an ITC200 (GE) at room temperature with 0.04 ml of 1 mM SAH in the injector cell and 0.26 ml of 2 mg/mL (75 mM) UbiG, UbiG mutants and liposome-bound UbiG in the sample cell, respectively. The protein and ligands were kept in a buffer consisting of 20 mM Tris-HCl (PH 7.5) and 50 mM NaCl. Five group experiments were conducted: for the first four groups, proteins (wt-UbiG and three UbiG mutants, respectively) were titrated with SAH directly, and for another group, wt-UbiG was titrated after the incubation with liposome. For the preparation of UbiG and liposome complex, 400 μg liposome was incubated with UbiG at 4 °C for 30 min. Twenty microliters injection volumes were used for all experiments. Two consecutive injections were separated by 2 min to reset the baseline. The control experiment, consisting of titration of SAH against buffer, was performed and substracted from each experiment to adjust for the heat of dilution of ligands. ITC data was analyzed with a single-site fitting model, using Origin 8.6 (OriginLab Corp).