Christopher Davis, Helen J Harris, Ke Hu, Heidi E Drummer, Jane A McKeating, Jonathan G L Mullins and Peter Balfe
Cellular Microbiology, 14: 1892-1903, 2012.
Cellular Microbiology website
| Get a PDF file of this paper
back to publications page
Although the molecular interaction of viruses with individual receptor proteins has been elucidated in great detail, the more dynamic process of multi-protein engagement has proven a greater challenge. The sequential engagement of receptors provides a coordinated process to regulate virus internalization but is largely unstudied to date.
Antibodies targeting the CD81 second extracellular loop (ECL2) have been shown to inhibit HCV infection in vitro and in vivo (Farquhar et al 2012). In contrast, there is limited evidence for tight junction protein association with HCV, suggesting an indirect role for claudin-1 and occludin in virus internalization (Meredith et al., 2012). We (Harris et al., 2008, Harris et al., 2010) and others (Cukierman et al., 2009, Kovalenko et al., 2007) have shown a direct interaction of claudins with tetraspanins, supporting a model where CD81-claudin-1 receptor complexes define HCV internalization.
We investigated the molecular interface between CD81 and claudin-1 using a combination of bioinformatic sequence based modelling, site-directed mutagenesis and Fluorescent Resonance Energy Transfer (FRET). Although the crystal structure of human CD81 ECL2 is known (PDB 1G8Q), no X-ray or NMR structural data exists for any member of the claudin family. We generated a homology model for claudin-1 ECL1 which we used to guide the genesis and in vitro screening of a panel of claudin-1 and CD81 mutants. We demonstrated an essential role for CD81 residues T149, E152 and T153 in the association of CD81 with claudin-1. Importantly, these mutations had no impact on protein conformation or HCV E2 binding but ablated HCV entry, highlighting the novel function of these CD81 residues. These observations increase our understanding of CD81/claudin-1 association and suggest a new approach for the computational design of peptides and small molecules to disrupt these receptor complexes and inhibit viral infection.
Hepatitis C virus (HCV) entry is dependent on host cell molecules tetraspanin CD81, scavenger receptor BI and tight junction proteins claudin-1 and occludin. We previously reported a role for CD81/claudin-1 receptor complexes in HCV entry, however, the molecular mechanism(s) driving association between the receptors is unknown. We explored the molecular interface between CD81 and claudin-1 using a combination of bioinformatic sequence based modelling, site-directed mutagenesis, and Fluorescent Resonance Energy Transfer (FRET) imaging methodologies. Structural modelling predicts the first extracellular loop of claudin-1 to have a flexible beta conformation and identifies a motif between amino acids 62-66 that interacts with CD81 residues T149, E152 and T153. FRET studies confirm a role for these CD81 residues in claudin-1 association and HCV infection. Importantly, mutation of these CD81 residues has minimal impact on protein conformation or HCV glycoprotein binding, highlighting a new functional domain of CD81 that is essential for virus entry.
Fig.1: Structural modelling of claudin-CD81 association.
Ribbon models of the ECL1 domain of native claudin-1 (A) or I32M (B), E48K (C) and I32M/E48K (D) mutants with CD81 ECL2 (PDB: 1G8Q). Claudin-1 is depicted according to its predicted secondary structure (alpha helices in red, beta turn in white) with CD81 in green, key interacting residues are labelled. The interacting regions include claudin-1 residues 33-35 and 63- 66, and CD81 residues K148, T149, E152 and T153. I32M and E48K claudin-1 mutations alter key inter-residue distances and interactions with CD81.
Fig.2: Effect of claudin-1 mutations on protein localization, CD81 association and HCV infection.
(A) 293-T cells were transfected to express wild type (WT) AcGFP- claudin-1 and mutants and their localization assessed by confocal microscopy. (B) Cell surface expression of AcGFP-claudin mutants was determined using the Zeiss profiling function to trace cell boundaries, the data shown is the average MFI of 10 cell profiles, the data are representative of three transfection experiments. (C) %FRET between AcGFP-claudin-1 mutants and DsRED-CD81 in transfected 293-T cells. None of the mutants in group I (W30A, I32A, D38A, G49A and W51A) showed any significant interaction with CD81. The mean % FRET values are derived from triplicate estimates within a single experiment and are representative of two further experiments. The mean plus 2SD of non-specific FRET values observed in parental non-transfected is indicated by the dashed line and represents the threshold for the assay. The diminution in %FRET signal for the members of group I relative to the other 2 groups was significant (Kruskal Wallis test, p< 0.01) (D) Infectivity of HCVpp in parental 293-T cells and in those transfected to express AcGFP-claudin-1 WT and mutants. 293-T cells expressed comparable levels of both WT and claudin-1 mutants.
Fig.3: Effect of CD81 mutations on protein localization, association and HCV infection.
(A)HepG2 cells were transduced to express wild type (WT) and mutant DsRED-CD81 proteins and their localization assessed by confocal microscopy. (B) %FRET between DsRED-CD81 mutants and AcGFP-CD81 in transfected 293-T cells. (C) %FRET between AcGFP-claudin-1 and DsRED-CD81 and mutant proteins expressed in 293-T cells, the threshold for the association is indicated by the dashed line. The mean values are derived from triplicate estimates which are representative of three further experiments. (D) Effect of CD81 mutations on HCV infection. HepG2 cells were transduced to express WT DsRED-CD81 and mutant proteins and evaluated for their ability to support HCVpp or MLVpp infection. The dashed line represents the mean plus 2 SD of HCVpp infection of CD81 negative parental HepG2 cells; levels of infection below this threshold are considered negative. The data presented are from a single experiment and are representative of two further independent experiments.
This paper identifies a new class of CD81 residues which modulate HCV entry independent of viral glycoprotein-receptor interaction. The experiments performed substantiate the key role of CD81-claudin-1 complexes in HCV internalization. The CD81/claudin-1 interface provides a new conserved target for anti-viral drug design and will allow the rational design of small molecule and peptide mimetics targeting the receptor complex.