Pre-structured motifs in the natively unstructured preS1 surface antigen of hepatitis B virus

Abstract

The preS1 surface antigen of hepatitis B virus (HBV) is known to play an important role in the initial attachment of HBV to hepatocytes. We have characterized structural features of the full-length preS1 using heteronuclear NMR methods and discovered that this 119-residue protein is inherently unstructured without a unique tertiary structure under a nondenaturing condition. Yet, combination of various NMR parameters shows that the preS1 contains "pre-structured" domains broadly covering its functional domains. The most prominent domain is formed by residues 27-45 and overlaps with the putative hepatocyte-binding domain (HBD) encompassing residues 21-47, within which two well-defined pre-structured motifs, formed by Pro(32)-Ala(36) and Pro(41)-Phe(45) are found. Additional, somewhat less prominent, pre-structured motifs are also formed by residues 11-18, 22-25, 37-40, and 46-50. Overall results suggest that the preS1 is a natively unstructured protein (NUP) whose N-terminal 50 residues, populated with multiple pre-structured motifs, contribute critically to hepatocyte binding.

Figure 1.

¹⁵N-¹H HSQC spectrum of the 119-residue full-length preS1. All of the backbone ¹⁵N and amide protons are assigned except for Asp⁵⁴–Ile⁵⁶ and 21 prolines.

Figure 2.

Summary of interproton NOEs, ³J_HNHα, and chemical-shift index (CSI) of H_α protons for preS1 (1–119). Thickness of the bar represents the relative strength of NOEs. Gray bars indicate an ambiguity due to overlapping of NOEs. Black circles are drawn when ³J_HNHα is <6 Hz and white circles when ³J_HNHα is >8 Hz. The black squares above and below the horizontal line represent CSI values of +1 and −1, respectively.

Figure 3.

H_α chemical-shift deviation and temperature coefficient of preS1 (1–119). (A) Difference plot showing actual displacement of the measured H_α proton chemical shifts from random coil values. ΔH_α = H_α (measured) −H_α (random). (B) Difference plot showing actual displacement of the measured C_α chemical shifts from random coil values. ΔC_α = C_α (measured) −C_α (random). (C) Temperature coefficients of amide protons (Δ_NH) are shown in ppb/K. To highlight the regions with the temperature coefficients that deviate significantly from those of a random conformation, Δ_NH values of <4 ppb/K are displayed as black circles. The horizontal line indicates an average value.

Figure 4.

Plots of backbone ¹⁵N-¹H heteronuclear NOEs and ¹⁵N relaxation times and against the residue number of preS1 (1–119). (A) ¹⁵N-¹H heteronuclear NOEs, (B) T₁, (C) T₂ relaxation times. Horizontal lines in B and C indicate average values. To highlight the regions that show local structural ordering, heteronuclear NOE values of >0.4 in A and T₂ values of <0.14 sec in C are displayed as black circles.

Figure 5.

Summary of interproton NOEs, ³J_HNHα, and chemical-shift index (CSI) of H_α protons for preS1 (21–47) peptide. Thickness of bar represents relative strength of NOEs. Black circles are drawn when ³J_HNHα is <6 Hz and white circles when ³J_HNHα is >8 Hz. The black squares above and below the horizontal line represent CSI values of +1 and −1, respectively. Gray bars indicate an ambiguity due to overlapping of NOEs.

Figure 6.

Location of pre-structured regions (this work) and previously suggested functional domains in preS1. Pre-structured regions are shown above the amino acid sequence of preS1 N-terminal 60 residues and are indicated by black, gray, or hatched bars. The most prominently pre-structured region is formed by residues 27–45. The black bars indicate two well-defined helical turn motifs within this region. The gray bars are for residues showing an intermediate degree of structural ordering and the hatched bars for least pre-structuring residues. Shown below the amino acid sequence of preS1 are the previously reported functionally important domains; the putative hepatocyte receptor-binding domain with a green (Neurath et al. 1986) or a red bar (Barrera et al. 2005), hepatocyte attachment domains with blue bars (Glebe et al. 2005), and an immunogenic domain with a violet bar (Hu et al. 2005), respectively.