Figure 1.
Conformational states of the dengue virus E protein and sequences of the membrane-proximal “stem”.
(A) Structure of E in the dimeric conformation present on the virion surface prior to low-pH exposure. The view is tangential to the viral membrane (gray stripe). The sE component (residues 1–395) of the ectodomain is in ribbon representation, with domains I, II, and III in red, yellow, and blue, respectively. The stem (residues 396–447) is shown as a helix-loop-helix, modeled from a cryoEM reconstruction [7]. The transmembrane anchor is an α-helical hairpin. (B) E in the trimeric conformation it adopts following a low-pH induced conformational change. Dashed lines indicate the likely position of the stem segments. (C) Linear representation of the E polypeptide chain, illustrating the regions that fold into the structures shown in A and B. The stem is shown “magnified”, together with stem sequences from dengue serotypes 1–4 and from West Nile virus (WNV). The designations of “helix 1”, “conserved region”, and “helix 2” follow.
Figure 2.
Binding and affinities of stem-derived peptides with DV2 sE(1–395) and inhibition of DV2 infectivity.
(A) Binding of a representative stem-peptide, DV2419–447, and of a peptide with the same composition but a scrambled amino-acid sequence, determined by fluorescence anisotropy. The fraction bound is plotted as a function of sE protein concentration (in µM). sE2 and sE3 refer to dimer and trimer, respectively. Trimer-binding curves were corrected for low levels of binding to the UDM present in the sE3 preparations. (B) The dissociation constants for various stem peptides were estimated by the concentration of half-maximal change in fluorescence polarization (FP IC50), as described in Methods. The effect on viral infectivity is shown as “IC90”, the concentration of peptide that reduced viral yield (of DV2 on BHK cells) to 10% of the control. N.B., no binding; N.A., no activity. Each FP IC50 is an average of at least three independent experiments, each done in triplicate. Each IC90 is from an assay set up in duplicate and plaqued in triplicate.
Figure 3.
Inhibition of fusion of virus and liposomes.
(A) Effect on fusion-pore formation (content mixing) of preincubating virions with DV2419–447 or DV2419–447(scram). Virions and peptides were incubated with liposomes encapsulating trypsin and acidified to pH = 5.5. Following back-neutralization and incubation for 1hr at 37C, samples were prepared for SDS-PAGE by TCA precipitation and immunoblotted with anti-C antibody. Fusion leads to exposure of capsid protein to trypsin and loss of the corresponding band. (B) Effect on the hemifusion step of preincubating pyrene-labeled virions with DV2419–447 or DV2419–447(scram). Hemifusion was measured by the decrease in pyrene excimer intensity when virions were acidified to pH = 5.5 in the presence of liposomes. Loss of excimer fluorescence following dissolution of the viral membrane by TX100 was taken as 100%. Data represent three independent experiments.
Figure 4.
Inhibition and reversibility of viral infectivity by DV2419–447.
(A) Influence of the time of peptide addition on viral yield. DV2419–447 (or DMSO carrier, for the control) was added to virus at 37°C, 45 minutes (preincubation) or immediately (coinfection) before applying the inoculum to cells, or at various time points postinfection (as shown by the time ramp). For the 0′ time point, peptide was added promptly after adding the inoculum. (B) Effect of separating peptide and virus prior to infection. DV2419–447 (10 µM) was preincubated with virus for 15′ or 45′ at 37*C, and unbound peptide was removed by rapid passage of the inoculum through a size-exclusion spin-column prior to addition to cells. Passage of virus (plus DMSO carrier) through the column reduced titre by about 10-fold. The flow-through fraction (F.T.) of medium plus peptide had no inhibitory effect on virus added subsequently, consistent with retention of all unbound peptide by the matrix. (C) Reversibility of stem peptide inhibition. Viral inoculum was preincubated with DV2419–447 for 10′ at 37C. Liposomes were then added to the peptide∶inoculum, in the molar excess of lipid molecules to peptides shown and incubated for an additional 45′. The inoculum was added to cells and harvested 24hrs later. An inoculum preincubated with liposomes alone has no loss in viral titre.
Figure 5.
Temperature dependence of peptide-virus interaction.
(A) Inhibition of infectivity after incubating virus with DV2419–447 at various temperatures, as indicated. RT, room temperature. (B) Association of biotinyl-DV2419–447 (DV2B419–447) with purified dengue virions at 4°C and 37°C (see Methods).
Figure 6.
Stem peptides interact with the dengue virus membrane.
Decrease in pyrene-excimer fluorescence intensity, as a function of the concentration of stem peptide added to pyrene-labeled virions, normalized to its complete loss upon addition of TX100.
Figure 7.
Model for inhibition of dengue-virus fusion by stem-derived peptides.
The stem peptide (magenta-dotted line) may bind in an initial, protein-specific interaction to the edge of E, domain II, dynamically exposed at neutral pH. The yellow arrow shows a likely fluctuation in the orientation of domain II, based on observed variation in the domain I-domain II hinge angle. Alternatively, a two-step model may obtain, in which the stem peptide has an initial non-specific membrane interaction that anchors it through its C-terminus in the viral membrane (magenta-solid line) at neutral pH. As fusion within an endosome proceeds toward the final “zipping up” step, with collapse of the extended intermediate, the peptide, carried into the endosome by association with the viral membrane, could then make a high-affinity, protein-specific interaction with trimer-clustered domain II, blocking hemifusion and pore formation by preventing close approach of the fusion peptide and transmembrane anchor (later images of panel).