2.1 Protein expression and purification.

The wide-type of full-length spfabF with fusion proteins including TRX, MBP and GST, and four chimeras as well, constructed on the basis of spfabF and its homologous genes Streptomyces albus J1074 fabF (safabF) (Fig. 1) were subcloned into the vector pET-28a with a TEV protease cleavage site following a six-His tag added at the N-terminus. Restriction sites BamHI and XhoI were used. The DNA sequence of the wide-type spfabF and safabF were synthesized by the company of Generay Biotech (Shanghai). The primers used for gene amplification are listed in Table 1.

Fig. 1Full-screen View

(Color online) The scheme of four designed chimeric FabFs. Cyan and green regions represent the sequence of saFabF and spFabF, respectively. The sequence identifies of the chimeric FabFs to sbFabF were shown in the right. Taking js₅₀FabF as an example, the first 50 amino acids are from the N-terminal 50 residues of J1074, the last 371 amino acids are from the C-terminal 371 residues of spFabF, and the sequence identity of js₅₀FabF to spFabF is 96.7%

The recombinant plasmids were transformed into E. coli BL21 (DE3) cells^[12]. Cells harboring this expression plasmid were grown in LB containing 50 μg/mL kanamycin to the mid-log phase at 37℃. Subsequently, the culture was induced by 0.1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and incubated for 14–16 hours at 16℃. The cell pellets were harvested, re-suspended, and lysed with the buffer consisting of 20 mM Tris pH 8.0, 10% (v/v) glycerol, 500 mM NaCl, 5 mM BME (β-mercaptoethanol), and 20 mM imidazole. The lysate was centrifuged. The supernatant was loaded on a 10-ml Ni-NTA Sepharose column (GE Healthcare). The column was washed with buffer A containing 20 mM Tris pH 8.0, 10% (v/v) glycerol, 500 mM NaCl, 5 mM BME, and 80 mM imidazole, until no further protein was detected by the Bio-Rad protein-assay kit, and then eluted with buffer B containing 20 mM Tris pH 8.0, 10% (v/v) glycerol, 500 mM NaCl, 5 mM BME, and 250 mM imidazole. The eluted fraction was concentrated and further loaded on the Superdex^TM 200 20/300 GL column (GE Healthcare) which is equilibrated in a gel filtration buffer consisting of 20 mM Tris pH 8.0, 100 mM NaCl, 10% (v/v) glycerol, 1 mM DL-Dithiothreitol (DTT). The eluted protein was concentrated to 10 mg/ml for crystallization.

2.2 Protein crystallization

Screening of the initial crystallization conditions was carried out using the sparse matrix approach with a series of kits including Index, Crystal Screen, Crystal Screen 2, PEG/Ion, PEG/Ion 2, SaltRx, SaltRx 2, PEGRx, PEGRx 2 (Hampton Research), and Wiz1-4 (Emerald BioSystems). Crystallization was performed at 4℃ and 20℃ using a sitting-drop vapour- diffusion method by mixing equal volumes of the protein solution (1 μL) and reservoir solution (1 μL). Once the crystal was grown, the vapour-diffusion method in hanging drops was utilized to optimize the condition. The perfluoropolyether PFO-X175/08 (Hampton Research)^[13] was used as a cryoprotectant for all of the crystals. The crystals were picked up with cryo-loops and flash-frozen in liquid nitrogen.

2.3 Data collection and processing (Table 2)

X-ray diffraction data were collected at 100 K on Beamline BL17U1 at SSRF ^[14], with the wavelength of 0.97923 Å and the crystal-to-detector distance of 250 mm. A total of 360 frames were collected with 0.5°oscillation and 1 s exposure time per frame. The data were processed with the HKL-2000 software package^[15]. Molecular replacement was performed with Phaser^[16] using the structure of Thermus thermophilus FabF (PDB code 1J3N)^[17] as a search model. The structure was refined with the Phenix code ^[18]. Using the Coot code ^[19], water molecules were fitted into to the initial F_o-F_c map.

3.1 Design of chimeric FabFs

The expression and purification of the wide-type spFabF with different fusion proteins in E. coli, as described in Section 2, led to the aggregates of the protein. A chimeric protein design strategy was subsequently applied. Sequence alignment suggested that saFabF might be of the highest sequence identity to spFabF (76.4%). The chimeric FabFs were thus constructed with the combination of spFabF and saFabF. In general, the N-terminal region of saFabF was used in the chimeric proteins because that its sequence was more identical than the C-terminal to spFabF. The first 200 amino acid residues of saFabF showed 81.5% sequence identity to the corresponding residues of spFabF. Four chimeras, namely js₅₀FabF, js₁₀₀FabF, js₂₀₀FabF, and sj₂₀₀FabF, were designed (Fig. 1). Taking js₂₀₀FabF as an example, the first 200 amino residues from the N-terminal of saFabF and the last 200 residues from the C-terminal of spFabF were combined (Fig. 1). The resulted js₂₀₀FabF had 91.2% sequence identity to the wide-type of spFabF.

The chimeric proteins were also expressed in E. Coli. The overexpressed proteins were purified by a Ni-NTA affinity chromatography together with a gel-filtration chromatography to obtain the homogeneous soluble proteins. The gel- filtration chromatographic experiment suggested that only js₂₀₀FabF might have proteins of low molecular weight, while the other three were aggregates. A further characterization of the low molecular weight samples with the analytical size exclusion chromatography showed that a single peak eluting at about 13.49 mL, which is usually corresponding to a protein of molecular weight around 90 kDa (Fig. 2). It is thus conjectured that js₂₀₀FabF was a comparatively stable dimer in solution. The SDS-PAGE showed that the proteins are quite pure (purity over 95%).

Fig. 2Full-screen View

Characterization of js₂₀₀FabF by the analytical size-exclusion chromatography and SDS-PAGE. js₂₀₀FabF was eluted with a peak at 13.49 mL of the Superdex^TM 200 20/300 GL column. In the SDS-PAGE, Lanes 0–4 represent the pellet resulted from the cell lysate, the flow-through from the NTA affinity column, the elute from the NTA affinity column with buffer A (containing 80 mM imidazole), the elute from the NTA affinity column with buffer B (containing 250 mM imidazole), and the concentrated sample used for the size-exclusion chromatography, respectively. Lane M represents molecular-weight markers labelled in kDa. Lanes 5–8 are corresponding to the fractions 5–8 from the size-exclusion chromatography.

3.2 Screening of crystallization conditions for js₂₀₀FabF

After the pure and dimeric js₂₀₀FabF were concentrated to 10 mg/mL, a crystallization condition screening was carried out using multiple screening kits. Rod crystals of the protein were obtained under conditions containing salts (the most), or PEGs (fewer) as a precipitating agent. However, these crystals had the low-resolution diffraction properties. The crystals from the condition No. 46 of Crystal Screen 2 consisting of 0.1 M NaCl, 0.1 M BICINE pH 9.0, 20% (v/v) PEG MME 550 (Fig. 3a) and the condition No. 30 of PEGRx consisting of 0.1 M HEPES pH 7.5, 12% PEG 3350 (Fig. 3b) were in bar-shapes. The crystal with the best diffraction was obtained using a reservoir solution consisting of 0.1 M Bis-Tris pH 5.5, 25% PEG 3350, 0.2 M ammonium sulfate (Fig. 3c). The crystals grown at 4℃ often showed better diffraction quality than those grown at 20 ℃. However, the crystal growth at 4℃ took about half a year before it was ready for XRD experiment. Besides, the six-His tag at the N-terminal of js₂₀₀FabF showed no apparent effects on crystallization and diffraction.

Fig. 3Full-screen View

(Color online) Crystals of js₂₀₀FabF grown with a reservoir containing 0.1 M sodium chloride, 0.1 M bicine pH 9.0, 20% (v/v) PEG MME 550 (a); 0.1 M HEPES pH 7.5, 12% PEG 3350 (b); and 0.1 M Bis-Tris pH 5.5, 25% PEG 3350, 0.2 M ammonium sulfate (c), respectively

3.3 Structure determination

Crystals of js₂₀₀FabF were flash-cooled in liquid nitrogen with a cryoprotectant and examined by XRD. An example of the XRD results is shown in Fig. 4. The data processing indicated that it belonged to the cubic space group P3₂21, with unit-cell parameters of a =160.27 Å, b =160.27 Å, and c = 95.57 Å. The calculated Matthews coefficient (V_M) was 2.32 ų/Da with 47.01% solvent content, assuming the presence of six molecules in the asymmetric unit ^[20]. The structure of js₂₀₀FabF was finally solved with the best resolution of 3.15 Å.

Fig. 4Full-screen View

A typical XRD result of the js₂₀₀FabF crystal.

PTM is a potent inhibitor of FabF from E. coli (ecFabF) and the crystal structures of the mutant ecFabF(C163Q and C163A) in complex with PTM are the only known structure of FabF bound with PTM ^[8,21]. Although the sequence identity of js₂₀₀FabF to ecFabF was 39.8%, the superimposition of two structures in Fig. 5(a) revealed that the topology of js₂₀₀FabF was similar to ecFabF. In addition, according to the complex structure of PTM bounded ecFabF(C163Q), the binding site for PTM in js₂₀₀FabF should be located at the C-terminal region, and the sequence was completely identical to the C-terminal of spFabF (Fig. 5b). This means that js₂₀₀FabF can be regarded as spFabF in studying the PTM binding to spFabF. Exploring the binding pocket of PTM in ecFabF(C163Q) and spFabF, we found that the three loops (residues 211-215, 277-281 and 407-411) adopted different conformations as compared with them in ecFabF(C163Q), resulting that the assumed binding pocket of PTM in spFabF was greatly occupied by these loop residues and the superimposed PTM had many bumping with the neighboring residues of spFabF (Figs. 5c and 5d). Therefore, the different conformations of three loops may account for the resistance of spFabF to PTM.

Fig. 5Full-screen View

(Color online) The superimposition of js₂₀₀FabF with ecFabF. (a) The cartoon representation of js₂₀₀FabF (green) and ecFabF (cyan) (PDB code 2GFW). (b) Molecular surface of js₂₀₀FabF. Orange and green represent the first 200 and the last 221 residues of js₂₀₀FabF, respectively. The position of PTM shown in yellow sticks was taken from the superimposed ecFabF(C163Q) in complex with PTM. (c) The binding pocket of PTM in ecFabF(C163Q). (d) The overlap of the superimposed PTM with the assumed binding pocket of PTM in js₂₀₀FabF