Highlighting the potential utility of MBP crystallization chaperone for Arabidopsis BIL1/BZR1 transcription factor-DNA complex

Abstract

The maltose-binding protein (MBP) fusion tag is one of the most commonly utilized crystallization chaperones for proteins of interest. Recently, this MBP-mediated crystallization technique was adapted to Arabidopsis thaliana (At) BRZ-INSENSITIVE-LONG (BIL1)/BRASSINAZOLE-RESISTANT (BZR1), a member of the plant-specific BZR TFs, and revealed the first structure of AtBIL1/BZR1 in complex with target DNA. However, it is unclear how the fused MBP affects the structural features of the AtBIL1/BZR1-DNA complex. In the present study, we highlight the potential utility of the MBP crystallization chaperone by comparing it with the crystallization of unfused AtBIL1/BZR1 in complex with DNA. Furthermore, we assessed the validity of the MBP-fused AtBIL1/BZR1-DNA structure by performing detailed dissection of crystal packings and molecular dynamics (MD) simulations with the removal of the MBP chaperone. Our MD simulations define the structural basis underlying the AtBIL1/BZR1-DNA assembly and DNA binding specificity by AtBIL1/BZR1. The methodology employed in this study, the combination of MBP-mediated crystallization and MD simulation, demonstrates promising capabilities in deciphering the protein-DNA recognition code.

Figure 1

Crystallization and preliminary X-ray analysis of the unfused AtBIL1/BZR1 in complex with DNA. (a) Construction of the unfused AtBIL1/BZR1 used for crystallization. (b) DNA constructs successfully cocrystallized with unfused AtBIL1/BZR1. (c) Crystals of the unfused AtBIL1/BZR1 in complex with target DNA containing an imperfect G-box variant. The detailed sequence of the DNA is shown above the photograph. (d,e) SDS-PAGE analysis (d) and agarose gel electrophoresis analysis (e) of dissolved crystals. (f) X-ray diffraction image (3.0 Å at the edge) from a crystal.

Figure 2

Crystallization and structure determination of MBP-fused AtBIL1/BZR1 in complex with DNA. (a) Mutant MBP (mMBP)-fused AtBIL1/BZR1 constructs for crystallization screening with different linker lengths. (b) DNA constructs successfully cocrystallized with mMBP-fused AtBIL1/BZR1 via one alanine linker. (c,d) Crystals (c) and an X-ray diffraction image (2.0 Å at the edge) from a crystal (d) of mMBP-fused AtBIL1/BZR1 in complex with the G-box-containing DNA. (e) Electron density map (2F_o–F_c) of AtBIL1/BZR1-DNA assemblies with contours at 1.5 σ (blue meshes) in the asymmetric unit of the reported structure (PDB ID: 5ZD4) depicted by the COOT program. The ribbons display the main-chain trace of two AtBIL1/BZR1 dimer-DNA complexes. Different colors represent different chains.

Figure 3

Crystal packing of MBP-fused AtBIL1/BZR1 in complex with DNA. (a–d) Front and side views of the crystal packing of AtBIL1/BZR1-DNA assemblies 1 (a,b) and 2 (c,d), depicted by PyMOL viewer. The boundaries between the C-terminus of mutant MBP (mMBP) and the N-terminus of AtBIL1/BZR1 are indicated with blue arrows. (e,f) Close-up views of the packing between the mMBP (yellow or green surface model) and DNA ends (white surface model) or AtBIL1/BZR1 loops (cyan or magenta surface model). (g,h) Close-up views of the spatial positioning of mMBP and DNA recognition helices of BIL1/BZR1. Hydrogen bonds and salt bridges are indicated by dashed lines. The residues involved in van der Waals interactions are shown as sphere models.

Figure 4

MD simulations for the AtBIL1/BZR1-DNA complex. (a) The crystal structure and MD structures (every 100 ns (ns) up to 1000 ns) of the AtBIL1/BZR1-DNA complex (PDB ID: 5ZD4, chains C, D, G and H, assembly 2). The tilt angle between the DNA recognition helices is shown on each crystal structure. (b) Crystal structures of the AtMYC2-DNA complex (PDB ID: 5GNJ, chains A–D) and hBAL1-hCLOCK-DNA complex (PDB ID: 4H10, chains A–D). (c) Tilt angles between DNA recognition helices of AtBIL1/BZR-DNA every 1 ns up to 1000 ns (three independent runs). The MD structures (a) correspond to the results of Run 3.

Figure 5

The C₁A₂ base recognition mode of BZR TFs is distinct from that of typical bHLH TFs. (a,b) The essential hydrogen-bonding networks for C₁A₂ base recognition by the AtBIL1/BZR1-DNA complex (a) and AtMYC2-DNA complex (b), which are observed in the crystal structures. The residues with or without a prime mark (’) belong to different chains. ‘P_N’ represents a phosphate group at position N. Dashed lines and a red sphere represent hydrogen bonds (including salt bridges) and a water molecule, respectively. (c) Ratios of hydrogen bond (H-bond) formation in MD structures of the AtBIL1/BZR1-DNA complex. The results of both chains α and β, which correspond to chains C and D of AtBIL1/BZR1 (PDB ID: 5ZD4), respectively, are shown for 3 independent runs. Colors closer to red indicate a higher ratio. (d) The binding free-energy differences (ΔΔG) between different nucleobases (C₁ to T₁ and A₂ to G₂) in complex with AtBIL1/BZR1 or AtMYC2. Data are the means + standard deviations (SDs, n = 6 independent runs).