Analysis of the C-terminal membrane anchor domains of hepatitis C virus glycoproteins E1 and E2: toward a topological model

Abstract

The hepatitis C virus (HCV) glycoproteins E1 and E2 should be anchored in the viral membrane by their C-terminal domains. During synthesis, they are translocated to the endoplasmic reticulum (ER) lumen where they remain. The 31 C-terminal residues of the E1 protein and the 29 C-terminal residues of the E2 protein are implicated in the ER retention. Moreover, the E1 and E2 C termini are implicated in E1-E2 heterodimerization. We studied the E1 and E2 C-terminal sequences of 25 HCV strains in silico using molecular modeling techniques. We conclude that both C-terminal domains should adopt a similar and peculiar configuration: one amphipathic alpha-helix followed by a pair of transmembrane beta-strands. Several three-dimensional (3-D) models were generated. After energy minimization, their ability to interact with membranes was studied using the molecular hydrophobicity potentials calculation and the IMPALA procedure. The latter simulates interactions with a membrane by a Monte Carlo minimization of energy. These methods suggest that the beta-hairpins could anchor the glycoproteins in the ER membrane at least transiently. Anchoring could be stabilized by the adsorption of the nearby amphipathic alpha-helices at the membrane surface. The 3-D models correlate with experimental results which indicate that the E1-E2 transmembrane domains are involved in the heterodimerization and have ER retention properties.

FIG. 1.

(A) HCA plot of the HCV-BEINNX11 E1 glycoprotein. (B) Minimal (dashed line), mean (solid line), and maximal (dotted line) values of the Kyte and Doolittle hydrophobicity along the E1 sequences of 25 HCV strains obtained by using a seven-residue window. Amino acid numbering refers to the position of the amino acids in the mature proteins.

FIG. 2.

(A) HCA plot of the HCV-BEINNX11 E2 glycoprotein. (B) Minimal (dashed line), mean (solid line), and maximal (dotted line) values of the Kyte and Doolittle hydrophobicity along the E2 sequences of 25 HCV strains obtained by using a seven-residue window. Amino acid numbering refers to the position of the amino acids in the mature proteins.

FIG.3.

Sequence alignment of E1 (A) and E2 (B) C-terminal domains from 25 HCV strains. A dash indicates an amino acid identical to the amino acid of HCV-BEINNX11 strain in the alignment. Amino acids are indicated according to the following characteristics (13, 60). Hydrophobic amino acids (Val, Leu, Ile, Met, Phe, Tyr, and Trp) which are preferentially implicated in transmembrane stretches are hatched. Among these, the aromatic residues (Phe, Tyr, and Trp) prefer the water-membrane interface. Charged amino acids (Asp, Glu, Arg, and Lys) which are avoided in transmembrane stretches are highlighted on a black background. His, Asn, and Gln are polar amino acids and are generally located in the water phase or at the hydrophobic-hydrophilic interface. Pro, the well-known helix breaker, and Gly, Ser, and Thr are frequent in coil and turn structures. They are indicated in italic type. Ala can be both in the membrane and the water phase. Domains 1, 2, and 3 for E1 and domains 1, 2, 3, and 4 for E2 are discussed in the text. Positions of alanine inserted by Op De Beeck et al. (45) are indicated by arrows on the top of each alignment. Stars denote the positions of alanine which have a significant effect on heterodimerization.

FIG. 4.

Comparison of the HCA plots of the C-terminal domains of HCV-BEINNX11 E1 glycoprotein (A) and HCV-BEINNX11 E2 glycoprotein (B). (C) Schematic representation of the topological model. Putative transmembrane β-strands (1 and 2) and a putative amphipathic α-helix (3) are indicated.

FIG. 5.

HCV-BEINNX11 E1 putative amphipathic α-helix (Val142 to Ala158). (A) Minimized 3-D structure. (B) MHP around the peptide (orange = hydrophobic; green = hydrophilic). (C) Best configuration found by the IMPALA procedure. The center of the phospholipid bilayer is indicated by an orange line, the limit between the phospholipid acyl chains and polar heads is indicated by a yellow line, and the limit between the water phase and the phospholipids is indicated by a green line.

FIG. 6.

HCV-BEINNX11 E2 putative amphipathic α-helix (Val316 to Ile331). (A) Minimized 3-D structure. (B) MHP around the peptide (orange = hydrophobic; green = hydrophilic). (C) Best configuration found by the IMPALA procedure. The center of the phospholipid bilayer is indicated by an orange line, the limit between the phospholipid acyl chains and polar heads is indicated by a yellow line, and the limit between the water phase and the phospholipids is indicated by a green line.

FIG. 7.

HCV-BEINNX11 E1 putative transmembrane β-strands (Trp162 to Ala188). (A) Best minimized 3-D structure obtained by the stereoalphabet procedure. (B) MHP around the peptide (orange = hydrophobic; green = hydrophilic). (C) Best configuration found by the IMPALA procedure. The center of the phospholipid bilayer is indicated by an orange line, the limit between the phospholipid acyl chains and polar heads is indicated by a yellow line, and the limit between the water phase and the phospholipids is indicated by a green line.

FIG. 8.

HCV-BEINNX11 E2 putative transmembrane β-strands (Tyr335 to Ala359). (A) Best minimized 3-D structure obtained by the stereoalphabet procedure. (B) MHP around the peptide (orange = hydrophobic; green = hydrophilic). (C) Best configuration found by the IMPALA procedure. The center of the phospholipid bilayer is indicated by an orange line, the limit between the phospholipid acyl chains and polar heads is indicated by a yellow line, and the limit between the water phase and the phospholipids is indicated by a green line.

FIG. 9.

Topological models of the E1 and E2 C-terminal anchor domains.