Table 1.
Summary of all synthesized CD81 and HCV E2 peptides.
Fig 1.
Sequence alignment of human, chimpanzee, and rat CD81s.
The sequence alignments of full-length human and chimpanzee CD81s display 100% identity, whereas that of human and rat CD81s show over 93% identity (97% similarity). The sequence identity of the LEL between humans and rats is 84% (93% similarity) and that of the transmembrane domain is 99%, indicating that the major differences between human and rat CD81s are in the LEL. The transmembrane domain at a.a. 1–112 and 202–236 is indicated with a green rod, whereas the LEL region at a.a. 113–201 is indicated with a red rod. The largest differences between the CD81s located within 160–190 a.a are within the black–framed box.
Fig 2.
Comparison of surface charge and lipophilicity distributions between human and rat CD81 ectodomains.
The structures of the resolved human CD81 (gray) and the homology-modeled rat CD81 (pink) shown in the ribbon were superposed. The major differences between the two CD81 structures are at the flexible loops from 173 to 186 a.a. (green: human CD81; red: rat CD81). (B) The surface charge distributions of the two CD81s show that the rat CD81 is more positively charged than human CD81 at the flexible loop region marked with a dashed line (blue: positive charge; red: negative charge). (C) The lipophilicity maps do not show much difference between the human and rat CD81s in the loop region (blue: hydrophilic; green: lipophilic).
Fig 3.
Molecular docking of human and rat CD81s to HCV E2 protein.
Preferable sites of HCV E2 binding with human and rat CD81s are shown in A to C. (A) The HCV E2-site1 loop could bind to human and rat CD81s with similar RDOCK scores (human: −18.3 kcal/mol; rat: −16.2 kcal/mol). (B) The HCV E2-site2 loop was able to dock to human and rat CD81s, but the RDOCK score for human CD81 was more than twice as low as rat CD81 (human: −14.7 kcal/mol; rat: −6.2 kcal/mol). (C) HCV E2 bound to human CD81 with both E2-site1 and E2-site2 loops (RDOCK score: -19.1 kcal/mol). Green: E2-site1; blue: E2-site2; pink: the binding loops of human and rat CD81s.
Fig 4.
MM/PBSA binding free energy calculations for human and rat CD81s to HCV E2 protein.
(A) For different HCV E2 sites (E2-site1, E2-site2, and E2-both sites) binding to human and rat CD81s, the binding free energies of human CD81 to HCV E2 were lower than those of rat CD81. HCV E2-site2 bound to human CD81 with the lowest binding free energy (H-E2-S2). (B) The detailed analysis of the components of binding free energies showed that the major difference for HCV E2-site2 binding to human and rat CD81s lies in the electrostatic interactions (H-E2-S2 and R-E2-S2). VDW dominates the binding of HCV E2-site1 to human CD81 (H-E2-S1). The figure represents the following. For H-E2-S1: the E2-site1 binding to human CD81; for H-E2-S2: the E2-site2 binding to human CD81; for H-E2-both: E2-both sites binding to human CD81; for R-E2-S1: the E2-site1 binding to rat CD81; and for R-E2-S2: the E2-site2 binding to rat CD81.
Fig 5.
Surface charge and lipophilicity distributions for HCV E2 binding to human CD81.
The complex structure is presented as a ribbon (orange: HCV E2; gray: human CD81). (A) and (B) are the surface lipophilicity distributions on HCV E2-site1 and human CD81 at the binding interface. In the figures, blue represents the hydrophilic part and green the hydrophobic part. The hydrophobic residues around the binding interface are labelled and presented as sticks. (C) and (D) are the surface charge distributions on HCV E2-site2 and human CD81 at the binding interface mapped according to the Poisson-Boltzmann equation. Blue and red correspond to positive and negative electrostatic potential, respectively. Charged residues around the binding interface are labelled and presented as sticks.
Fig 6.
SPR measurements for the interactions between peptides derived from HCV E2 and CD81s.
(A) SPR responses when HCV p_E2-site1 peptide in various concentrations was flowed over the immobilized human CD81 peptide (left). The equilibrium KD of 7.96 ± 1.7 μM for p_E2-site1 binding to human CD81 peptide was determined by steady-state interaction isotherm (right). (B) SPR responses measured when p_E2-site1 in various concentrations was tested on rat CD81 peptide (left). The equilibrium KD of 13.85 ± 2.46 μM for the p_E2-site1 binding to rat CD81 peptide was given by the binding isotherm (right). (C) SPR response when peptide p_E2-site2 in different concentrations was flowed over the immobilized human CD81 peptide (left). The binding isotherm gives the equilibrium KD of 1.07 ± 0.09 μM for p_E2-site2 binding to human CD81 peptide (right). (D) SPR responses when peptide p_E2-site2 in various concentrations was tested on rat CD81 peptide. The response only increased slightly as the concentration of the peptide increased (left). The equilibrium KD of 6.38 ± 0.58 μM for the p_E2-site2 binding to rat CD81 peptide was calculated from the steady-state binding isotherm (right). The KD values shown are the averages of three measurements. Errors for KD are standard deviations.
Fig 7.
Flow cytometry of E2 peptides binding to human and rat cells and inhibitions of anti-CD81 antibody/cell binding by E2 peptides.
(A) and (B) show the fluorescence intensity of Huh-7 cells treated with fluorescent p_E2-site1 (E2-s1) and p_E2-site2 (E2-s2), respectively, at different concentrations; (C) and (D) are the fluorescence intensity measurements of the rat PC12 cells treated with fluorescent p_E2-site1 or p_E2-site2, respectively; and (E) and (F) are the inhibitions of fluorescent anti-CD81 antibodies targeting Huh 7 cells by HCV E2 peptides. In this experiment, untreated cells were used as negative controls, and the cells treated with fluorescent-labelled anti-CD81 antibodies were used as positive controls and set as 100% binding. p_m_E2-site1 (m-E2-s1) and p_m_E2-site2 (m-E2-s2) in (A), (B), (C) and (D) are mutant peptides of p_E2-site1 and p_E2-site2 indicated in S1 Table.
Fig 8.
Putative model of the HCV E2/CD81 binding process.
The initial HCV E2/CD81 binding process can be divided into two steps. Step 1: E2 initially recognizes and approaches CD81 with the E2-site2 region. Step 2: The orientation of E2 changes to a more preferable binding pose with the E2-site1 region auxiliary binding to execute the processes that follow. The E2-site1 region, E2-site2 region, and CD81 binding loop are presented in green, blue, and pink, respectively.