Sequential conformational changes in the morbillivirus attachment protein initiate the membrane fusion process

Abstract

Despite large vaccination campaigns, measles virus (MeV) and canine distemper virus (CDV) cause major morbidity and mortality in humans and animals, respectively. The MeV and CDV cell entry system relies on two interacting envelope glycoproteins: the attachment protein (H), consisting of stalk and head domains, co-operates with the fusion protein (F) to mediate membrane fusion. However, how receptor-binding by the H-protein leads to F-triggering is not fully understood. Here, we report that an anti-CDV-H monoclonal antibody (mAb-1347), which targets the linear H-stalk segment 126-133, potently inhibits membrane fusion without interfering with H receptor-binding or F-interaction. Rather, mAb-1347 blocked the F-triggering function of H-proteins regardless of the presence or absence of the head domains. Remarkably, mAb-1347 binding to headless CDV H, as well as standard and engineered bioactive stalk-elongated CDV H-constructs treated with cells expressing the SLAM receptor, was enhanced. Despite proper cell surface expression, fusion promotion by most H-stalk mutants harboring alanine substitutions in the 126-138 "spacer" section was substantially impaired, consistent with deficient receptor-induced mAb-1347 binding enhancement. However, a previously reported F-triggering defective H-I98A variant still exhibited the receptor-induced "head-stalk" rearrangement. Collectively, our data spotlight a distinct mechanism for morbillivirus membrane fusion activation: prior to receptor contact, at least one of the morbillivirus H-head domains interacts with the membrane-distal "spacer" domain in the H-stalk, leaving the F-binding site located further membrane-proximal in the stalk fully accessible. This "head-to-spacer" interaction conformationally stabilizes H in an auto-repressed state, which enables intracellular H-stalk/F engagement while preventing the inherent H-stalk's bioactivity that may prematurely activate F. Receptor-contact disrupts the "head-to-spacer" interaction, which subsequently "unlocks" the stalk, allowing it to rearrange and trigger F. Overall, our study reveals essential mechanistic requirements governing the activation of the morbillivirus membrane fusion cascade and spotlights the H-stalk "spacer" microdomain as a possible drug target for antiviral therapy.

Fig 1. Inhibition of CDV-mediated viral-cell and cell-cell fusion by the anti-CDV-H mAb-1347.

(A) Syncytium formation assay. Cell-to-cell fusion activity in Vero-cSLAM cells triggered by co-expression of CDV H-wt and CDV F-wt (A75/17 strain) in the presence of absence of mAb αH-1347. Representative fields of view of cell-cell fusion induced 24h post-transfection are shown. (B) Quantitative fusion assay. Vero-cSLAM cells (target cells) were infected with MVA-T7 (MOI of 1). In parallel, a population of Vero cells (effector cells) was transfected with the F-wt and H-wt-expressing vectors and a plasmid containing the luciferase reporter gene under the control of the T7 promoter. Twelve hours after transfection, effector cells were mixed with target cells and seeded into fresh plates. After 2.5 h at 37°C, fusion was indirectly quantified by using a commercial luciferase-measuring kit. For each experiment, the value obtained for the standard F/H combination was set to 100%. (C) Virus neutralization assay. A total of 100 infectious units of recA75/17^red was incubated with the indicated dilution of antibody for 1 h at 37°C. The virus-antibody mixtures were then added to 3h on Vero cells, overlaid with agar-containing medium and further incubated for 72 h at 37°C. Cell entry efficiency was determined by counting the number of red fluorescent syncytia induced by recA75/17^red. (D) Effect of mAbs on H/SLAM binding efficiency. Vero cells were transfected with H-wt. Prior to treatment with soluble HA-tagged cSLAM molecules, mAbs were added as indicated, and SLAM-binding activity was calculated as the ratio of mean fluorescence intensities obtained with an anti-HA polyclonal Ab values (staining for sol. cSLAM) normalized to the levels obtained with the anti-FLAG mAb (staining for H). Values recorded for H-wt/cSLAM-binding efficiency in the absence of the mAb were set at 100%. Wt: wild type, α: monoclonal antibody, sSLAM: soluble version of cSLAM. Means ± S.D. of data from three independent experiments in triplicate are shown.

Fig 2. mAb αH-1347 epitope mapping.

(A) Schematic representation of the full length CDV-H-protein (H-wt) and a derivative soluble form (sH-ecto). The main functional regions of the stalk are color-coded. CT: cytosolic tail; TM: transmembrane domain. (B) Schematic representation of soluble Hstalk-RFP chimeric proteins and derivative deleted mutants (1–5). Del: deletion; RFP: red fluorescent protein. Drawings are not to scale. (C, D and F) Immunoprecipitation experiments performed with the indicated mAb of soluble H forms derived from the supernatant of transfected 293T cells. Antigenic materials were detected using polyclonal anti-GCN4 (upper panel) or anti-His antibody (lower panel). IP: immunoprecipitation; IB: immunoblotting. (E) Upper part: Schematic representation of the deletion 5 soluble Hstalk-RFP. Lower part: alignment of the primary amino acid sequence of the CDV H-stalk “spacer” section and the derivative triple alanine-scan mutants. Alanines are highlighted in red. (G) Homology structural model of CDV-F and positions (in red) where the mAb αH-1347 epitope has been inserted. Lower part: primary amino acid sequence of the CDV H-stalk “spacer” section with the encompassed F-transferred mAb-1347 peptide (highlighted in red). (H) Reactivities of F-wt-tagFLAG or F-wt-tag1347 with the conformation-insensitive anti-FLAG, the F prefusion state-specific anti-F-Pre (4941) [61, 63] or the anti-CDV-H 1347 mAbs, probed 24 hours post-transfection in receptor-negative Vero cells. After addition of the secondary antibody, MFI values were recorded by flow cytometry analyses. Means ± S.D. of data from three independent experiments in triplicate are shown.

Fig 3. Inhibition of headless H/F-mediated membrane fusion by mAb αH-1347.

(A, C and D) Syncytium formation assay. Cell-to-cell fusion activity in Vero or Vero-cSLAM cells triggered by co-expression of CDV H-wt or headless H variants with CDV F-wt or CDV F-V447T (A75/17 strain) [62] in the presence (+) of absence of mAb αH-1347 (-). Representative fields of view of cell-cell fusion induced 24h post-transfection are shown. (B) F-triggering assay. The CDV F-V447T mutant was expressed in Vero cells together with the indicated Hstalk variants. One day post-transfection, the conformation of the FLAG-tagged F-V447T mutant was monitored by probing its reactivity with an anti-F prefusion-specific mAb (4941; green histograms), triggered-specific mAb (3633; red histograms) [61, 63]. Secondary antibodies were added at 4°C, and to record quantitative values, mean fluorescence intensities were monitored by flow cytometry. (E) Quantitative fusion assays were performed as described in the legend of Fig 1B in the presence (light grey histograms) or absence (dark grey histograms) of mAb αH-1347. Means ± S.D. of data from three independent experiments performed in duplicates are shown.

Fig 4. Receptor-induced conformational change in H detected by mAb αH-1347.

(A) Reactivity to mAb αFLAG and αH-1347 of standard (H-wt) or headless (Hstalk) attachment proteins in the presence or absence of F-wt. After addition of the secondary antibody, MFI values were recorded by flow cytometry. H/F co-immunoprecipitation with different anti-CDV-H mAbs. (B) Cell surface assessment of H protein interaction with cleaved F-proteins. To stabilize H/F complexes, transfected Vero cells were treated with the membrane non-permeable cross-linker DTSSP and subsequently lysed with RIPA buffer. Complexes were then immunoprecipitated (IP) with anti-CDV-H mAb 3734 [61] and protein G-Sepharose beads treatment. Proteins were boiled and subjected to immuoblotting using a polyclonal anti-CDV-F antibody [71] to detect F antigenic materials (coIP). Co-IP F proteins were detected in comparison with F proteins present in cell lysates prior to IP by immuoblotting using the same anti-F antibody (TL, total lysate; F₀, uncleaved F protein; F₁, cleaved membrane-anchored F subunit). The specific mAb used for the immunoprecipitation step is indicated at the top of the gel. (C) Similar to (B) but with cell extracts obtained from Vero cells transfected with headless H and F. (D) Effect of receptor treatment on mAb αH-1347’s H-binding activity (at 4°C or 37°C). Vero cells expressing H-wt were co-cultured with Vero-SLAM, Vero-Nectin-4 or Vero cells together with mAb αH-1347. After addition of the secondary antibody, MFI values were recorded by flow cytometry. Means ± S.D. of data from three independent experiments performed in triplicates are shown.

Fig 5. Most alanine H mutants spanning the “spacer” CDV H-stalk region are fusion-promotion impaired.

(A) Schematic representation of the main functional domains of the full length CDV H-wt protein. The detailed primary sequence of the “spacer” section and the residues defining mAb αH-1347 epitope are shown. The H stalk-section scanned by alanine mutagenesis is also highlighted. (B) Syncytium formation assay. Cell-to-cell fusion activity in Vero-cSLAM cells triggered by co-expression of CDV H-wt or derived alanine variants with CDV F-wt. Representative fields of view of cell-cell fusion induced 24h post-transfection are shown. (C) Characterization of H mutants. Standard and derivative H mutants were expressed in Vero cells. One day post-transfection, cell surface expression (recorded by anti-FLAG staining followed by flow cytometry analyses), SLAM binding efficiency (calculated as described in the legend of Fig 1D, but without addition of mAbs before SLAM treatments) and fusion activity (monitored as described in the legend of Fig 1B) were determined. All values were normalized to H-wt. (D) Reactivity of standard and derivative H mutants to mAb αH-1347. MFI values were recorded by flow cytometry 24h post-transfection in Vero cells. (E) Effect of SLAM treatment on mAb αH-1347’s H-binding activity (37°C). The assay and values were determined as described in the legend of Fig 4D. Means ± S.D. of data from four independent experiments performed in triplicates are shown. To determine the statistical significance of differences between the standard and mutant H (127 and 138) data sets, unpaired two-tailed t tests were performed (*, P < 0.05; **P < 0.01).

Fig 6. Functional and biochemical characterization of stalk-elongated and shortened engineered H variants.

(A) Upper part: schematic representation of the main functional domains of the full length CDV H-wt protein. Middle part: schematic representation of the H-elongated construct with the precise position of the insertion indicated. The 11 residues inserted are derived from the N-terminal stalk “contact” section 112–122 that putatively assume a helical fold with an 11-mer repeat. Lower part: schematic representation of the H-shortened version with the precise position of the deletion indicated. (B) Immunoblotting of standard and size-modulated H proteins. Antigenic materials ran in an 8% SDS-Page gel under reducing conditions were detected using a polyclonal anti-H antibody. (C) Syncytium formation assay. Fusion activity in Vero-cSLAM cells triggered by co-expression of CDV H (or size-modulated mutants) and CDV F in the presence (+) of absence (-) of mAb αH-1347. Representative fields of view of cell-cell fusion induced 24h post-transfection are shown. (D) Reactivity of H-wt and size-modulated H variants to mAbs αFLAG and αH-1347. Vero cells were transfected with the different H plasmids and, after addition of the secondary antibody, MFI values were recorded 24h post-transfection. (E) SLAM-binding activity of H mutants. Standard and H mutants were expressed in Vero cells. One day post-transfection, cell surface expression (CSE; recorded by anti-FLAG staining followed by flow cytometry analyses) and SLAM binding efficiency (calculated as described in the legend of Fig 1D, but without addition of mAbs before SLAM treatments) were determined. All values were normalized to H-wt. (F) Co-IPs, performed as described in the legend of Fig 4B, were obtained 24 hours post-transfection of cell extracts of Vero cells transfected with standard H or derivative mutants and F-wt-expressing vectors. (G) Semiquantitative assessment of F/H avidity of interactions. To quantify the avidities of F₁-H interactions, the signals in each F₁ and H bands were quantified using the AIDA software package. The avidity of F₁-H interactions is represented by the ratio of the amount of coimmunoprecipitated (coIP) F₁ over the product of F₁ in the cell lysates divided by the ratio of the amount of immunoprecipitated H over the product of H in the cell lysate ((coIP F₁/TL F₁)/(IP H/TL H)). Subsequently, all ratios were normalized to the ratio of the wild-type F-H interactions set to 100%. Averages represent at least two independent experiments. (H) Effect of SLAM treatment on mAb αH-1347’s H-binding activity. The assay and values were determined as described in the legend of Fig 4D (37°C). Means ± S.D. of data from two independent experiments performed in triplicates are shown.

Fig 7. Investigation of the mAb αH-1347 to H stoichiometry required for membrane fusion inhibition.

(A) Cartoon representation of the CDV H-wt tetramer and derived “headless” variant in a putative pre-F-triggering state. The reported I98A mutation shown to abrogate F-triggering without impairing H/F interaction is also shown in the full length H-wt protein [17, 64]. (B) Representation of the different hetero-oligomeric assemblies that may emerge from H-98A and H-stalk co-expressing cells. (C-E) Syncytium formation assay. Fusion activity in Vero-cSLAM cells triggered by co-expression of CDV H-98A and F-wt, or headless H and F-wt, or H-I98A and headless H and F in the presence (+) of absence (-) of mAb αH-1347. Representative fields of view of cell-cell fusion induced 24h post-transfection are shown. (F-H) Quantitative fusion assay. The fusion promotion efficiency of each H/F combinations was determined as described in the legend of Fig 1B. Means ± S.D. of data from three independent experiments performed in triplicates are shown.

Fig 8. Models of paramyxovirus membrane fusion activation.

(A) Summary of the putative mechanism of membrane fusion activation for morbilliviruses (carrying the “spacer” module in the H-stalk). In the first panel, the cartoon represents one H-wt tetramer (in blue) assuming the auto-repressed state not yet bound to its cognate receptor (i.e. SLAM; represented by black ovals anchored in the target plasma membrane). The associated trimeric fusion protein (F) in the pre-fusion conformational state is represented in green. Because the “4-heads-down” configuration determined for NDV HN [36] contains elements that fit with the proposed pre-receptor-bound locked H-structure, we arbitrarily illustrated H in this conformation. In this state, the two “lower” H-heads are partially covering the “spacer” stalk microdomain (green box) where the mAb αH-1347 epitopes also locate. These “head-to-spacer” contacts are critical in stabilizing the auto-repressed H state. Because the H-stalk “spacer” module maps membrane-distal from the candidate F-binding/activation regulatory segment (red box), H/F assembly can occur even prior to receptor binding. In the second panel, both accessible “upper” heads engage with the receptor. Conversely, distance and/or physical constraints may prevent efficient receptor binding to the “lower” head units. In the third panel, because both monomeric head units within each dimers may assemble into structurally stable complexes (or may require some adjustment prior to achieve stable dimeric units [75]), the contact of the “upper” heads with the receptor leads to a rearrangement of the dimers that automatically relocates the “lower” heads “away” from the stalk (achieving a putative “heads away” structural intermediate). This step thus disrupts the critical “head-to-spacer” contacts and leads to the “de-activation” of the auto-repressed state. In the fourth panel, the stalk is now free to refold into the F-triggering competent state. The latter includes “structural flexibility” or “opening” of the central section. In the fifth panel, upon H-mediated activation, F achieves the pre-hairpin intermediate structural state bridging the viral envelope and the target cell plasma membranes (color-coded in yellow and orange, respectively). Basics of this model was recently hypothesizes for CDV [68] and MeV [48] (referred to as “safety catch” in the latter study). (B) Summary of the putative mechanism of membrane fusion activation of PIV5 and NDV (expressing attachment protein with short “spacerless” stalks). In the first panel, the pre-receptor-bound state of HN is represented in the “4-heads-down” configuration not yet bound to its cognate receptor (i.e. sialic acid; represented by red spheres attached to membrane bound molecules). As a major difference with morbillivirus attachment proteins, HN-stalks do not carry the analogous “spacer” segment that locate membrane-distal to the F-binding/activation sites (red box). Consequently, the two backfolding dimeric head units directly cover the F-binding/activation sites, which prevent HN/F assembly prior to receptor engagement. In the second panel, both accessible “upper” heads engage with the receptor. In the third panel, the contact of the HN-Heads with the receptor triggers a large-scale conformational change that switches the heads from the “down” to the “up” configuration. The latter structural state implies a tetrameric assembly of the heads positioned above the stalk. Consequently, the F-binding/activation sites are unmasked, which in turn allow for HN/F interactions and subsequent “induced fit” mechanism that ultimately lead to F-triggering. In the fourth panel, F achieves the pre-hairpin intermediate structural state bridging the viral envelope and the target cell plasma membranes (color-coded in yellow and orange, respectively). This model is referred to as the “stalk-exposure/induce fit” model [40, 46].