Measles virus glycoprotein complexes preassemble intracellularly and relax during transport to the cell surface in preparation for fusion

Abstract

Measles virus (MeV), a morbillivirus within the paramyxovirus family, expresses two envelope glycoproteins. The attachment (H) protein mediates receptor binding, followed by triggering of the fusion (F) protein, which leads to merger of the viral envelope with target cell membranes. Receptor binding by members of related paramyxovirus genera rearranges the head domains of the attachment proteins, liberating an F-contact domain within the attachment protein helical stalk. However, morbillivirus glycoproteins first assemble intracellularly prior to receptor binding, raising the question of whether alternative protein-protein interfaces are involved or whether an entirely distinct triggering principle is employed. To test these possibilities, we generated headless H stem mutants of progressively shorter length. Conformationally restricted H stems remained capable of intracellular assembly with a standard F protein and a soluble MeV F mutant. Proteolytic maturation of F, but not the altered biochemical conditions at the cell surface, reduces the strength of glycoprotein interaction, readying the complexes for triggering. F mutants stabilized in the prefusion conformation interact with H intracellularly and at the cell surface, while destabilized F mutants interact only intracellularly, prior to F maturation. These results showcase an MeV entry machinery that functionally varies conserved motifs of the proposed paramyxovirus infection pathway. Intracellular and plasma membrane-resident MeV glycoprotein complexes employ the same protein-protein interface. F maturation prepares for complex separation after triggering, and the H head domains in prereceptor-bound conformation prevent premature stalk rearrangements and F activation. Intracellular preassembly affects MeV fusion profiles and may contribute to the high cell-to-cell fusion activity characteristic of the morbillivirus genus.

Importance: Paramyxoviruses of the morbillivirus genus, such as measles, are highly contagious, major human and animal pathogens. MeV envelope glycoproteins preassemble intracellularly into tightly associated hetero-oligomers. To address whether preassembly reflects a unique measles virus entry strategy, we characterized the protein-protein interface of intracellular and surface-exposed fusion complexes and investigated the effect of the attachment protein head domains, glycoprotein maturation, and altered biochemical conditions at the cell surface on measles virus fusion complexes. Our results demonstrate that measles virus functionally varies conserved elements of the paramyxovirus entry pathway, providing a possible explanation for the high cell-to-cell fusion activity of morbilliviruses. Insight gained from these data affects the design of effective broad-spectrum paramyxovirus entry inhibitors.

FIG 1

MeV glycoprotein oligomerization models. Prior to receptor binding, the attachment protein (H) is shown in red in a hypothetical heads-down conformation. (A and C) As proposed in Jardetzky and Lamb (29), the head domains may physically impede interaction of the F₀ trimer with the H stalk, and intracellular H/F₀ complex preassembly is mediated through an alternative interface between the F and H head domains (A). Receptor binding induces an H heads-up conformation, allowing productive interaction of H with the F protein, resulting in F triggering (C). (B) Alternatively, F₀ may directly interact with the H stalk domain even if the H heads are in heads-down mode. While the model shown in panel A necessitates a protein-protein interface switch, the model shown in panel B raises the question of how premature F triggering is prevented. Note that evidence for different head arrangements was reported for MeV H (5); but the depicted heads-up and heads-down conformations are hypothetical, and cartoons are based on structural data obtained for NDV and PIV5 HN proteins and the subsequently proposed universal Paramyxovirinae fusion models (29).

FIG 2

Distinct effects of H stalk mutations and head truncation on H/F₀ and H/F₁₊₂ complex assembly. (A) Whole-cell lysates (WCL) of cells expressing MeV H_Flag mutants and F_HA were analyzed by immunoblotting (IB) using epitope tag-specific antibodies. (B) Cells from the experiment shown in panel A were subjected to coimmunoprecipitation (IP) of F with H or surface cross-linking using DTSSP to stabilize the H-F₁₊₂ interaction, in both cases followed by detection of coprecipitated F antigenic material. (C) Cells transfected with MeV H_Flag mutants and MeV or PIV3 F_HA were analyzed as described for panels A and B. (D) MeV receptor-negative MDCK cells were transfected with MeV F_HA and H_Flag mutants and analyzed for intracellular interaction as outlined above. α, anti.

FIG 3

MeV H/F₀ assembly is based on interaction of F with the H stalk domain. (A) Cell-to-cell fusion of cells transfected with MeV F and the newly generated GCN domain-stabilized H stem mutants of progressively shorter length. For comparison, the previously reported bioactive H-stem_122-GCN was included. In the case of cells expressing H-stem_110-GCN and F, arrowheads highlight individual syncytia. (B and C) Analysis of cell lysates and coprecipitated F material as outlined in the legend of Fig. 2A and B after pulldown with the specified H mutants in the presence or absence of DTSSP.

FIG 4

Conformational H stalk stabilization restores intracellular stalk interaction of headless H-stem_122-GCN with F. (A) Analysis of cell lysates and coprecipitated F material as outlined in the legend of Fig. 2A and B after pulldown with the specified H mutants in the presence or absence of DTSSP. (B) Schematic of the foldon domain-stabilized solF_HA-foldon mutant. (C) Immunodetection of solF_HA-foldon in whole-cell lysates (WCL) and culture supernatants after immunoprecipitation. (D and E) Analysis of the effect of H-stalk-stabilizing mutations D101C, T112C, and D113C from the constructs shown in panel A introduced into full-length H (D) or H-stem_122-GCN (E) on intracellular interaction with solF_HA-foldon.

FIG 5

DTSSP cross-linking stabilizes surface-exposed MeV glycoprotein complexes in prefusion conformation. (A) Cell-to-cell fusion of cells transfected with MeV H and F mutant F(L325D) or F(L457A-V459A). (B) F surface immunoprecipitation (SF IP) after incubation of intact cells transfected with MeV H and F(L325D) or F(L457A-V459A) with conformation-dependent antibodies directed against the F trimer in the prefusion (αPre F) or postfusion (αTrig F) conformation. (C) Analysis of cell lysates and coprecipitated F mutants F(L325D) and F(L457A-V459A) after pulldown with the specified H mutants in the presence or absence of DTSSP as outlined in the legend of Fig. 2A and B.

FIG 6

Uncleaved F₀ complexes remain stably associated with H on the cell surface. (A) Cell-to-cell fusion microphotographs of cells transfected with MeV H and F mutant F(K111N), disrupting the furin proteolysis motif. (B) Surface biotinylation of cells transfected with F(K111N) or standard F. The F₁-like material observed with F(K111N)-transfected cells represents nonproductive, most likely postextraction, F proteolysis. (C and D) Analysis of coprecipitated F(K111N) after pulldown with standard H (C) or the specified H mutants (D) in the presence or absence of DTSSP as outlined in the legend of Fig. 2A and B. (E) Fusion profiles of MeV H/F and H-stem122GCN/F pairs using a quantitative fusion assay based on the reconstitution of dual-split luciferase/enhanced green fluorescent protein fusion proteins. F-encoding plasmid DNA was kept constant for all samples, and the relative amount of H-encoding plasmid DNA was progressively reduced. Values represent mean relative luciferase units ± SEM of four independent experiments, each normalized (norm.) for a relative H-to-F ratio of 0.5. To determine the statistical significance of differences between the standard and mutant H-based data sets, unpaired two-tailed t tests were performed (*, P < 0.05; **P < 0.01; ***P < 0.001; NS not significant). SF bio, surface biotinylation.

FIG 7

(Top) Three-step safety-catch model of maturation and activation of the MeV entry machinery as the envelope proteins advance from the endoplasmic reticulum (ER) through the Golgi apparatus (Golgi) to the plasma membrane (PM). (Left) Hetero-oligomer preassembly. MeV H and F₀ proteins tightly assemble in the ER, engaging a protein-protein interface between the prefusion F head and H stalk domain. The H head domains in a hypothetical, prereceptor-bound heads-down conformation do not prevent assembly of the H/F₀ complex. (Middle) Readying the fusion complex. Proteolytic F maturation (symbolized by stars) in late Golgi compartments greatly reduces the strength of hetero-oligomer interaction, preparing the complexes for physical separation after triggering to accommodate F refolding. The H head domains in heads-down conformation act as a safety catch, preventing conformational changes of the H stalk and blocking premature F triggering. (Right) Triggering and hetero-oligomer separation. Receptor binding induces rearrangements of the H head domains (safety catch off), allowing unwinding of the central stalk section, which triggers the F refolding cascade leading to H and F₁₊₂ separation and, ultimately, membrane fusion.