Subcellular localization and topology of the p7 polypeptide of hepatitis C virus

Abstract

Although biological and biochemical data have been accumulated on most hepatitis C virus proteins, the structure and function of the 63-amino-acid p7 polypeptide of this virus have never been investigated. In this work, sequence analyses predicted that p7 contains two transmembrane passages connected by a short hydrophilic segment. The C-terminal transmembrane domain of p7 was predicted to function as a signal sequence, which was confirmed experimentally by analyzing the translocation of a reporter glycoprotein fused at its C terminus. The p7 polypeptide was tagged either with the ectodomain of CD4 or with a Myc epitope to study its membrane integration, its subcellular localization, and its topology. Alkaline extraction studies confirmed that p7 is an integral membrane polypeptide. The CD4-p7 chimera was detected by immunofluorescence on the surface of nonpermeabilized cells, indicating that it is exported to the plasma membrane. However, pulse-chase analyses showed that only approximately 20% of endoglycosidase H-resistant CD4-p7 was detected after long chase times, suggesting that a large proportion of p7 stays in an early compartment of the secretory pathway. Finally, by inserting a Myc epitope in several positions of p7 and analyzing the accessibility of this epitope on the plasma membrane of HepG2 cells, we showed that p7 has a double membrane-spanning topology, with both its N and C termini oriented toward the extracellular environment. Altogether, these data indicate that p7 is a polytopic membrane protein that could have a functional role in several compartments of the secretory pathway.

FIG. 1.

Sequence analyses of p7. (A) Position of p7 in the HCV polyprotein. p7 numbering is according to genotype 1a HCV H strain consensus cDNA (26) (accession number AF009606). Solid diamonds indicate cleavages of the HCV polyprotein precursor by an ER signal peptidase(s), and arrows indicate cleavages by the NS2-3 and NS3/4A proteases. The open diamond indicates further processing of the capsid protein by a cellular protease. (B) Alignment of p7 sequences representative of the principal HCV subtypes of clades 1 to 6. p7 numbering has been normalized from 1 to 63. The EMBL accession number of each sequence is indicated in parentheses. The consensus amino acid sequence for the 13 selected sequences is indicated on the top of the panel. Amino acids identical to those in the consensus sequence are represented by a dash. (C) Repertoire of amino acids per position in 289 HCV isolates of various genotypes. Amino acids are listed in decreasing order of observed frequency, from top to bottom. Amino acids within the box correspond to residues observed in more than 10% of the 289 sequences. Amino acids observed at a given position in fewer than two distinct sequences (<0.3%) were not taken into consideration. (D) Histogram showing the hydropathic character of residues at each p7 position. The height of each box in each bar indicates the number of sequences observed with a given residue at a given position. The boxes are presented in order of decreasing hydrophobicity, from bottom to top, according to the hydrophobicity scale of Black and Mould (F, I, W, Y, L, V, M, P, C, A, G, T, S, K, Q, N, H, E, D, R). Each box is colored according to the hydrophobic character of the residue: dark grey for hydrophobic (F, I, W, Y, L, V, M, P, C, and A), light grey for neutral (G, T, and S), and white for hydrophilic (K, Q, N, H, E, D, and R). (E) Consensus hydropathic pattern deduced from the data in panel D. o, hydrophobic residue; n, neutral residue; i, hydrophilic residue; v, variable residue; +, fully conserved positively charged residues. The black boxes indicate predicted minimal transmembrane segments deduced by various prediction methods (see Materials and Methods section). The grey boxes indicate that the corresponding amino acid positions were predicted to be membranous but not for all the HCV isolates.

FIG. 2.

p7 is an integral membrane protein. HepG2 cells were coinfected with vTF7-3 and the appropriate vaccinia virus recombinant at a multiplicity of infection of 5 PFU/cell. At 5 h postinfection, membrane fractions were prepared as described in Materials and Methods and treated with 0.1 M sodium carbonate, pH 11.3. After separation of membrane-bound (M) and soluble (S) proteins, the presence of p7NT or p7CT in these fractions was revealed by Western blotting with the anti-Myc MAb. Sizes (in kilodaltons) of protein molecular mass markers are indicated on the left.

FIG. 3.

CD4-p7 protein is exported to the cell surface. HepG2 cells were coinfected with vTF7-3 and a vaccinia virus recombinant expressing the ectodomain of CD4 in fusion with p7 (A) or a truncated form of HCV polyprotein (CE1E2p7CT) (B) at a multiplicity of infection of 3 PFU/cell. Cells were fixed with paraformaldehyde at 6 h postinfection, permeabilized or not with Triton X-100, immunostained with the anti-CD4 MAb OKT4 (A) or anti-Myc antibody (B), and analyzed by immunofluorescence.

FIG. 4.

Analysis of the endo H sensitivity of CD4-p7. HepG2 cells were coinfected with vTF7-3 and a vaccinia virus recombinant expressing CD4 or CD4-p7 at a multiplicity of infection of 5 PFU/cell. At 4.5 h postinfection, infected cells were pulse-labeled for 10 min and chased for the indicated times (in hours). Cell lysates were immunoprecipitated with the anti-CD4 MAb and treated or not with endo H. Proteins were separated by SDS-PAGE (10% polyacrylamide). Endo H-resistant proteins are indicated by asterisks. The size (in kilodaltons) of a protein molecular mass marker is indicated on the left.

FIG. 5.

Identification of a signal sequence in the C-terminal half of p7. (A) Sequence analysis of the C-terminal transmembrane domain of the p7 used in this work (HCV H strain), indicating that this domain has the characteristic structural features of a signal peptide (54). A typical signal peptide is composed of an N-terminal region (n-domain) encompassing between one and three positively charged residues (K33 and R35 here), a hydrophobic core region (h-domain) forming an α-helix (segment 41 to 57), and a more polar, flexible region (c-domain) containing the signal peptidase cleavage site (segment 58 to 63). Residues at positions −1 and −3 relative to the cleavage site are small neutral residues (A63 and A61) and form the recognition site for signal peptidase (53). Furthermore, an α-helix-destabilizing residue is frequently located at position −6 (P58) and/or in the middle of the h-domain (P49). The black box indicates the predicted minimal transmembrane segment (see Fig. 1E). (B) Schematic representation of the proteins used to identify the signal sequence function of the C-terminal half of p7. Sp1E1 corresponds to E1 with its signal sequence, and SpNS2E1 corresponds to the C-terminal half of p7 (residues 781 to 809 on the polyprotein) followed by the ectodomain and transmembrane domain of E1. (C) The C-terminal transmembrane half of p7 is a signal sequence. HepG2 cells were coinfected with vTF7-3 and the appropriate vaccinia virus recombinant at a multiplicity of infection of 5 PFU/cell. At 4.5 h postinfection, cells were labeled for 1 h with ³⁵S-Protein Labeling Mix. Cell lysates were immunoprecipitated with MAb A4 (anti-E1) and treated or not with endo H. Samples were analyzed by SDS-PAGE (12% polyacrylamide) and autoradiography. Sizes (in kilodaltons) of protein molecular mass markers are indicated.

FIG. 6.

Determination of the topology of p7 by epitope tagging. HepG2 cells were coinfected with vTF7-3 and the appropriate vaccinia virus recombinant at a multiplicity of infection of 3 PFU/cell and analyzed by indirect immunofluorescence. Cells were fixed with paraformaldehyde at 6 h postinfection, permeabilized or not with Triton X-100 (A) or digitonin (B), and immunostained with the anti-Myc MAb. A schematic representation of the topology of the different p7 polypeptides tagged with a Myc epitope is presented on the left.

FIG. 7.

Structural prediction for transmembrane domains of p7. (A) Repertoire of amino acids per position in 248 HCV isolates of clade 1 (genotypes 1a, 1b, and 1c). Amino acids are listed in decreasing order of observed frequency, from top to bottom. Amino acids within the box correspond to residues observed in more than 10% of the 248 sequences. Residues observed in fewer than three sequences (<0.8%) are not presented. (B) Consensus hydropathic pattern deduced from the data in panel A. o, hydrophobic residue; n, neutral residue; i, hydrophilic residue; v, variable residue; + and −, positions where fully conserved positively or negatively charged residues, respectively, were observed. The black boxes indicate the predicted transmembrane segments deduced by various prediction methods (see Materials and Methods section). (C) Ideal α-helix projection of the putative p7 transmembrane segments TM1 (left) and TM2 (right). The variability of residues at each position is included, according to the data in panel A. The larger characters indicate the most frequently observed residues. Outlined, italic, and bold letters correspond to neutral, hydrophilic, and hydrophobic residues, respectively (for details, see legend to Fig. 1D).

FIG. 8.

Conservation and comparison of amino acid sequences of p7 of the four main types of pestiviruses. The repertoires of residues per position for BVDV1, BVDV2, classical swine fever virus (CSFV), and border disease virus (BDV) (pestivirus types 1, 4, 2, and 3, respectively [1]) were deduced from the sequences of natural variants available in the EMBL database. Note that pestivirus variants exhibiting p7 sequences with a higher or lower number of residues were not taken into account in the reported repertoires. The number of sequences analyzed is indicated in parentheses. Amino acids are listed in decreasing order of observed frequency, from top to bottom. The bottom panel summarizes the consensus features for the four types of pestivirus. The hydropathic pattern is deduced from the repertoire analyses: o, hydrophobic residue; n, neutral residue; i, hydrophilic residue; v, variable residue (for details, see legend to Fig. 1D). Positions where fully conserved positively or negatively charged residues were observed are indicated by + and −, respectively. The boxes indicate possible transmembrane segments deduced from sequence analysis by various prediction methods (see Materials and Methods section). Dark grey sections show the consensus hydrophobic clusters of residues assumed to be involved in transmembrane segments. The shaded section indicates that the corresponding amino acid positions were predicted as membranous by some prediction methods and not for all the isolates.