A Structurally Unresolved Head Segment of Defined Length Favors Proper Measles Virus Hemagglutinin Tetramerization and Efficient Membrane Fusion Triggering

ABSTRACT

Paramyxoviruses include several insidious and ubiquitous pathogens of humans and animals, with measles virus (MeV) being a prominent one. The MeV membrane fusion apparatus consists of a receptor binding protein (hemagglutinin [H]) tetramer and a fusion (F) protein trimer. Four globular MeV H heads are connected to a tetrameric stalk through flexible linkers. We sought here to characterize the function of a 17-residue H-head segment proximal to the stalk that was unresolved in all five MeV H-head crystal or cocrystal structures. In particular, we assessed whether its primary sequence and length are critical for proper protein oligomerization and intracellular transport or for membrane fusion triggering. Extensive alanine substitutions had no effect on fusion triggering, suggesting that sequence identity is not critical for this function. Excessive shortening of this segment reduced or completely abrogated fusion trigger function, while length compensation restored it. We then characterized the mechanism of function loss. Mutated H proteins were efficiently transported to the cell surface, but certain alterations enhancing linker flexibility resulted in accumulation of high-molecular-weight H oligomers. Some oligomers had reduced fusion trigger capacity, while others retained this function. Thus, length and rigidity of the unresolved head segment favor proper H tetramerization and counteract interactions between subunits from different tetramers. The structurally unresolved H-head segment, together with the top of the stalk, may act as a leash to provide the right degree of freedom for the heads of individual tetramers to adopt a triggering-permissive conformation while avoiding improper contacts with heads of neighboring tetramers.

IMPORTANCE Understanding the molecular mechanism of membrane fusion triggering may allow development of new antiviral strategies. The fusion apparatus of paramyxoviruses consists of a receptor binding tetramer and a fusion protein trimer. Structural analyses of the receptor binding hemagglutinin-neuraminidases of certain paramyxoviruses suggest that fusion triggering is preceded by relocation of its head domains, facilitated by flexible linkers. Having noted a structurally unresolved 17-residue segment linking the globular heads to the tetrameric stalk of the measles virus hemagglutinin (H), we asked whether and how it may facilitate membrane fusion triggering. We conclude that, together with the top of the stalk, the flexible linker keeps H heads on a leash long enough to adopt a triggering-permissive conformation but short enough to limit roaming and improper contacts with heads of neighboring tetramers. All morbillivirus H-protein heads appear to be connected to their stalks through a “leash,” suggesting a conserved triggering mechanism.

INTRODUCTION

Although targeted for eradication, measles virus (MeV) still caused 120,000 deaths worldwide in 2014 alone (1, 2). Relaxed vaccination discipline has favored measles reemergence in Europe and North America, which now report costly epidemics: in 2013 the number of measles cases in the United States was triple that in previous years, in 2014 it was about 10-fold higher (3–6), and in 2015 a Disneyland-originated outbreak reminded the world of the immediate benefits of high measles vaccination coverage. Moreover, a recent retrospective study of the consequences of the introduction of measles vaccination 50 years ago indicated that elimination of measles-induced immune suppression significantly reduced child death due to opportunistic infections (7).

MeV is a negative-strand RNA virus of the family Paramyxoviridae (8), which includes deadly emerging viruses such as Hendra virus and Nipah virus and prevalent human pathogens such as mumps virus, parainfluenza virus, respiratory syncytial virus, and metapneumovirus. For cell entry, most Paramyxoviridae use a two-component fusion apparatus consisting of a receptor binding protein tetramer and a fusion (F) protein trimer. Those attachment proteins that bind sialic acid and have both hemagglutinin and neuraminidase activities are named HN, while those that bind specific proteins are named G or H (8, 9).

Paramyxovirus attachment proteins are type II transmembrane glycoprotein tetramers: four globular heads connect to a tetrameric stalk (10). The 617-amino-acid MeV H protein has an N-terminal 33-residue cytoplasmic tail followed by a transmembrane segment, a 96-residue stalk, and the globular head domain (11). Atomic structures of the H stalk are not available, but structures of the HN stalk alone or in the context of the whole ectodomain have revealed a four-helix bundle structure with a kink in the central region (12, 13).

Five atomic structures of the MeV H globular head have been determined, including those of H dimers covalently linked by Cys154 (14), H monomers (14, 15), and costructures in complex with three receptors: SLAM (16), nectin-4 (17), and CD46 (18). However, no structure is complete; some include a 17-residue gap (residues 167 to 183; PDB 2ZB5 and 2ZB6), while others start with residues 184 or higher and thus exclude the entire stalk-proximal region (PDB 2RKC, 3ALW, 3ALX, 3ALZ, 3INB, and 4GJT). On the other hand, complete HN ectodomain structures have revealed different arrangements of the heads about the stalk: “heads down,” in which the head dimers align sideways of the stalk; “heads up,” with all four heads postulated to be atop the stalk; and the intermediate “2 heads up/2 heads down” (13, 19).

Functional studies of the Paramyxoviridae receptor binding proteins have tested different models about how they initiate the membrane fusion process (20–22). A conserved paramyxovirus cell entry mechanism, with interesting genus-specific variations, is emerging (9, 10, 23, 24). In particular, it was shown that the MeV H tetramer can integrate signals from three different receptors that contact the heads through partially overlapping but substantially different binding surfaces (25–28). All these membrane-bound receptors may “pull” on head dimers, destabilizing them (11). The central segment of the H stalk then integrates and relays the triggering signals to the F trimer, which unfolds to fuse membranes (29–33).

Knowing that unstructured linkers connect the globular heads of other paramyxoviruses with their respective stalks (13), we assessed here whether the primary sequence and length of the 17-residue unstructured MeV H-head segment are critical for three functions: protein oligomerization, transport to the cell surface, and efficiency of membrane fusion triggering. We conclude that the unstructured segment may keep individual heads on a leash long enough to allow sufficient movement about the stalk for them to adopt a triggering-permissive conformation but short enough to limit improper contacts with heads of neighboring tetramers.

MATERIALS AND METHODS

Cells.

Vero (African green monkey kidney) cells and 293T (human embryonic kidney) cells were grown in Dulbecco's modified Eagle's medium (DMEM) (HyClone, South Logan, UT) supplemented with 10% fetal bovine serum (FBS). Chinese hamster ovary (CHO) cells were maintained in RPMI medium (Corning, Manassas, VA) supplemented with 10% FBS and 0.5 mg/ml of nonessential amino acids. Cell lines were incubated at 37°C with 5% CO₂.

Plasmids and mutagenesis.

All mutations were made in a vaccine lineage H-protein backbone (H-NSe) (34, 35). Mutations in the H stalk were introduced into the pCG-H expression plasmid by QuikChange site-directed mutagenesis (Agilent Technologies, Santa Clara, CA) with modifications to the manufacturer's instructions as described previously (36). The clones were verified by sequencing the H-protein gene in the vicinity of the mutation. At least two independent clones were tested for each mutation.

Cell-to-cell fusion assays.

The semiquantitative cell-to-cell fusion assay was performed as described previously (33). Briefly, 0.5 μg each of three plasmids, encoding the H protein, the F protein, and green fluorescent protein (GFP), were transfected into 1.0 × 10⁵ Vero cells in 24-well plates using Lipofectamine 2000 (Life Technologies, Grand Island, NY) according to the manufacturer's instructions. Twenty-four hours posttransfection, the extent of fusion was recorded in one field of view (about 2,000 cells) using the following criteria: 0, two or fewer syncytia with 4 to 5 nuclei (background); 1, three or more syncytia with 4 to 5 nuclei; 2, one to three syncytia with more than 10 nuclei; and 3, four or more syncytia with more than 10 nuclei.

The quantitative fusion assay was based on a dual-split-reporter assay adapted from that described previously (37). CHO cells (2 × 10⁵ in 12-well plates) were transfected with 0.3 μg each of the H and F expression plasmids and one of the dual-split-reporter luciferase expression plasmids (DSP_1–7, a kind gift of Z. Matsuda). As a control, only the H or the F plasmid was cotransfected with DSP_1–7. Vero cells (4 × 10⁵ in 6-well plates) were transfected with 1.5 μg of the other dual-split-reporter plasmid (DSP_8–11). Twenty-four hours posttransfection, Versene (0.48 mM EDTA in phosphate-buffered saline [PBS]) (Life Technologies) was used to gently detach the transfected CHO and Vero cells. Equal numbers of CHO (5 × 10⁴) and Vero cells were mixed and plated on black-wall 96-well plates. EnduRen (Promega) was added as the substrate to the culture medium (DMEM, 10% FBS) according to the manufacturer's instructions. Content mixing between CHO and Vero cells as a result of fusion was monitored at the indicated times using a Topcount NXT luminometer (Packard Instrument Company, Meriden, CT). Three replicates were used for each H construct.

Immunoblots.

To determine H-protein expression levels, Vero cells were transfected with plasmid DNA using Lipofectamine 2000 (Life Technologies) according to the manufacturer's instructions. H expression and oligomerization were documented as described previously (32). Briefly, the standard pCG-H (34) or mutated plasmids (2 μg) were transfected into 1.5 × 10⁵ Vero cells in 12-well plates. Thirty-six hours posttransfection, cytoplasmic extracts were made in lysis buffer (25 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, 1 mM EDTA, 5% glycerol) supplemented with cOmplete protease inhibitor cocktail (Roche, Basel, Switzerland) and 50 mM iodoacetamide (Sigma-Aldrich, St. Louis, MO) and the proteins separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (4 to 15% gradient) under reducing conditions. To determine H oligomeric states, nonreducing conditions were used. The H proteins were visualized on an immunoblot using an H cytoplasmic tail-specific polyclonal antibody (38) and a horseradish peroxidase (HRP)-conjugated secondary antibody.

FACS analysis.

To determine H-protein cell surface expression levels, 293T cells (8 × 10⁵ in a 6-well plate) were transfected with H-protein expression plasmids (3 μg) as described above. Thirty-six hours posttransfection, cells were washed with PBS and detached by incubating with Versene (Life Technologies) at 37°C for 10 min. The resuspended cells were washed twice with cold fluorescence-activated cell sorter (FACS) wash buffer (1× PBS, 2% FBS, 0.1% sodium azide) and then incubated with a blend of two anti-H monoclonal antibodies (1:100 dilution) (MAB8905; Millipore, Billerica, MA) for 1 h at 4°C. Cells were washed three times with cold FACS wash buffer and incubated with a phycoerythrin-conjugated secondary antibody (1:100 dilution) (115-116-146; Jackson ImmunoResearch, West Grove, PA) for 1 h at 4°C. After three washes with FACS wash buffer, cells were fixed in 4% paraformaldehyde and analyzed with a FACSCalibur (BD Biosciences, San Jose, CA) cytometer and FlowJo software (Tree Star Inc., Ashland, OR).

Analysis of H-protein oligomers at the cell surface.

Vero cells (3 × 10⁶ in a 10-cm dish) were transfected with 8 μg of plasmid DNA encoding MV H constructs as indicated. Twenty-four hours posttransfection, cells were washed in cold PBS and then incubated with 0.25 mg/ml sulfosuccinimidyl-2-(biotinamido)ethyl-1,3-dithio-propionate (Pierce) in PBS for 30 min at 4°C, followed by washing and quenching for 10 min at 4°C. Samples were lysed in lysis buffer (25 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, 1 mM EDTA, 5% glycerol) supplemented with cOmplete protease inhibitor cocktail (Roche) and 50 mM iodoacetamide (Sigma-Aldrich) for 30 min on ice with vortexing every 5 min. Lysates were cleared by centrifugation for 10 min at 10,000 × g and 4°C. Biotinylated proteins were adsorbed to immobilized streptavidin for 2 h at 4°C and then washed three times. Samples were boiled in Laemmli buffer for 5 min at 95°C and then subjected to 4 to 15% gradient SDS-PAGE and immunoblotted using an H cytoplasmic tail-specific antibody (38).

RESULTS

The unresolved head-stalk connecting linker (residues 167 to 183) is located between Cys154, which forms an intersubunit disulfide bond and defines the top of the stalk, and Cys188, which forms a disulfide bridge with Cys606, defining the globular head (Fig. 1A). Sequence alignment of MeV H with that of five other morbilliviruses reveals weak sequence conservation of the stalk-proximal residues, while the adjacent stalk sequence is strongly conserved (Fig. 1B).

FIG 1.

Schematic of MeV H protein, sequence alignment of morbillivirus H stalks and stalk-proximal head segment, and structure of a tetrameric stalk with a head dimer. (A) H linear structure. From left to right: C, cytoplasmic tail; T, transmembrane region; S, stalk; and β1 to -6, beta sheets 1 through 6. The two Cys residues that cross-link the H dimer (Cys139 and Cys154) and the two Cys residues that stabilize the H head (Cys188 and Cys606) are shown as gray lines. The 17-residue structurally unresolved head segment is shown in red. (B) Sequence alignment of the MeV H stalk and unresolved head segment with five other morbilliviruses. The alignment was made using Clustal Omega (43). Canine, canine distemper virus; Phocine, phocine distemper virus; Porpoise, porpoise morbillivirus; PPRV, peste-des-petits-ruminants virus; Rinderpest, rinderpest virus; Measles, measles virus. Top, shades of blue represent the degree to which the identity of an amino acid is conserved at a given position. Bottom, the conservation level of each residue is indicated by the height of the bars below each residue and the associated score below the bar (0 to 9). (C) Model of the tetrameric MeV stalk with a head dimer in the “down” position. The “2-heads-up/2-heads-down” HN crystal structure was used as the template to align the H-dimer crystal structure (PDB 2ZB5) in the “down” position relative to the stalk. Residues 185 to 607 of the two H monomers are shaded differently (green and cyan) for clarity. Residues 155 to 165 are shaded gray in both monomers. The structurally unresolved segment (residues 167 to 183), which constitutes a 17-residue gap in the H-head dimer crystal structure, is visualized as a red ribbon on the green subunit. These residues are not visible in the cyan subunit, as they are on the opposite face of the monomer. The two residues that flank the gap are shaded yellow. Cys154 at the N-terminal end of the H monomers is shaded magenta for the cyan subunit and orange for the green subunit. The stalk is illustrated with the aid of a structural model and schematics for regions which lack a structural template. Red, central fusion activation segment (residues 84 to 117). Green, spacer segment (residues 122 to 137). Blue, dimeric linker (residues 139 to 154).

The primary sequence of the structurally unresolved head segment is not critical for fusion triggering.

We used an alanine scan mutagenesis approach to assess whether the sequence of the unresolved 17-residue linker is critical for fusion triggering. Five mutants with blocks of 3 to 9 alanine substitutions were generated (Table 1, alanine substitutions). The fusion-triggering function of these mutants was tested by a semiquantitative cell-to-cell fusion assay (32) and found to be similar to that of the standard H protein (Table 1).

TABLE 1.

Sequences and fusion functions of unresolved head-segment mutants

Class	Name	Sequence	Fusion scorea
		167          183
Standard	H	VNSTLLETRTTNQFLAV	+++
Alanine substitutions	Ala_167–169	AAA--------------	+++
	Ala_167–172	AAAAAA-----------	+++
	Ala_167–175	AAAAAAAAA--------	+++
	Ala_173–178	------AAAAAA-----	+++
	Ala_176–183	---------AAAAAAAA	+++
Deletions	Del_167–169	ΔΔΔ--------------	+++
	Del_179–183	------------ΔΔΔΔΔ	+++
	Del_173–178	------ΔΔΔΔΔΔ-----	++
	Del_167–175	ΔΔΔΔΔΔΔΔΔ--------	0
	Del_167–183	ΔΔΔΔΔΔΔΔΔΔΔΔΔΔΔΔΔ	0
Insertions	Ins_167+SG	SG-----------------	++
	Ins_167+2×SG	SGSG-----------------	++
	Ins_167+4×SG	SGSGSGSG-----------------	+
	Ins_183+2×SG	-----------------SGSG	+++
Length compensation	Sub_167–175	SGSGSGSGS--------	+++
	Sub_167–183	SGSGSGSGSGSGSGSGS	++

^aFusion levels were determined and scored by a cell-to-cell fusion assay as described in Materials and Methods. +++, full fusion function; ++ and +, lower levels of fusion; 0, indicates no fusion.

To determine if altering the sequence interfered with the kinetics of fusion, we monitored fusion over time using a dual-split-reporter-based quantitative fusion assay (37). In this assay, content mixing between an effector cell expressing the MeV fusion complex and a target cell expressing the MeV vaccine lineage receptor CD46 allows the two halves of Renilla luciferase to functionally interact. Indeed two of the larger alanine substitution mutants, Ala_167–175 and Ala_173–178, exhibited similar levels and kinetics of fusion as the standard H protein (Fig. 2A). This confirms that the primary sequence of the structurally unresolved head segment is not critical for the fusion trigger function.

FIG 2.

Fusion kinetics of unresolved head segment mutants. Cell-to-cell fusion was monitored over a 20-hour period for two alanine scan mutants (A), two deletion mutants (B), two insertion mutants (C), and two length compensation mutants (D). The names of the mutants and symbols used to represent them are indicated in the insets. As negative controls, cells transfected with only one or the other standard glycoprotein were monitored. Only one control (F-only) is shown for clarity. Similar results were obtained for the H-only control.

The length of the structurally unresolved head segment is critical for efficient fusion triggering.

Next, we deleted blocks of 3 to 9 residues from different locations in the unresolved head segment (Table 1, deletions). In addition, we deleted the entire segment (Table 1, Del_167–183). While short deletions of 3 or 5 amino acids on either end of the segment had no impact on function (Table 1, Del_167–169 and Del_179–183), an intermediate deletion of 6 amino acids in the middle of the segment decreased fusion (Table 1, Del_173–178). Larger deletions of 9 and 17 amino acids completely abolished fusion function in the semiquantitative fusion assay (Table 1, Del_167–175 and Del_167–183). In the more sensitive dual-split-reporter-based quantitative fusion assay, low levels of fusion activity for these deletion mutants were detected: after 10 h, the deletion mutants had 5 to 10% of the fusion activity observed for the standard H protein (Fig. 2B). Thus, shortening the unresolved head segment interferes with efficient fusion function.

To assess whether elongating the unresolved head segment interferes with fusion triggering, we inserted 1, 2, or 4 serine/glycine pairs on its sides (Table 1, insertions). The functional assays indicate that 2-, 4-, or 8-residue insertions at the stalk-proximal side (before residue 167) reduced fusion function (Table 1). On the other hand, a four-amino-acid insertion on the stalk-distal side (after residue 183) had no impact on function (Table 1, Ins_183+2×SG). These results are corroborated by the dual-split-reporter-based quantitative fusion assay (Fig. 2C), which indicates slower fusion kinetics and overall lower levels of fusion for Ins_167+2×SG.

Finally, we sought to reverse the negative effect on function of the 9- and 17-residue deletion mutants. We compensated for these deletions in mutants Del_167–175 and Del_167–183 by inserting repeating units of serine and glycine (Table 1, length compensation). Indeed, both the semiquantitative (Table 1) and the split luciferase fusion (Fig. 2D) assays indicated full or partial functional compensation for the former and latter mutants, respectively. Thus, the length of the structurally unresolved head segment is critical for efficient fusion triggering.

All mutants but one are transported efficiently to the cell surface.

To characterize the mechanisms causing loss of function, we sought to document expression levels and intracellular transport of the mutants. Figure 3A documents by immunoblotting that all the key mutants are expressed at levels comparable to those for the standard H protein. Next, we measured the amount of H protein reaching the plasma membrane. 293T cells expressing the different H mutants or standard H protein were stained with a blend of two monoclonal antibodies that recognize the H ectodomain and analyzed by FACS. Figure 3B plots the mean fluorescence intensities of different MeV H mutants as percentages of that of the standard H protein at the cell surface. Only deletion mutant Del_167–175 reached the cell surface with less than 5% of the efficiency of standard H. The remaining mutants were present at the cell surface at levels 45 to 95% of that of standard H, which is sufficient for full fusion support function (32, 33). Mutant Ala_167–175 reached the cell surface at ∼65% efficiency and exhibited full fusion support function. On the other hand, 4- and 8-residue insertions of flexible Ser/Gly pairs at the membrane-proximal end of the unresolved head segment reached the cell surface at ∼90% efficiency (Fig. 3B, Ins_167+2×SG, and data not shown) but were impaired in fusion function (Table 1, Ins_167 2×SG and Ins_167+4×SG).

FIG 3.

Protein expression levels of unresolved head segment mutants. (A) Total protein expression. Cytoplasmic extracts of Vero cells transfected with the H expression plasmids indicated below the gel were separated by 4 to 15% SDS-PAGE under reducing conditions and immunoblotted with an anti-H (cytoplasmic tail) antibody. M, H-monomer band. A, β-actin was used as a loading control. (B) Cell surface protein expression. 293T cells transfected with the indicated H construct were stained with a blend of two anti-H ectodomain antibodies and phycoerythrin-conjugated secondary antibody. The mean fluorescence intensity of each mutant is presented as a percentage of the standard H-protein levels. Error bars indicate standard deviations from three experiments.

Certain mutants express nonstandard H oligomers.

We next sought to determine the oligomeric state of the mutants. Cytoplasmic extracts of cells transfected with H expression plasmids were separated on SDS-PAGE. Standard H protein exists almost exclusively as a stable disulfide-linked dimer under nonreducing conditions (Fig. 4, lane H). As documented in the immunoblot in Fig. 4, all the mutants expressed substantial levels of H dimer (Fig. 4, D bands).

FIG 4.

Oligomeric states of the mutants. (A) Cytoplasmic extracts of Vero cells transfected with the indicated H expression plasmid were separated by 4 to 15% SDS-PAGE under nonreducing conditions and immunoblotted with an anti-H (cytoplasmic tail) antibody. The different H forms are indicated. O, stable higher-order H oligomers; D, dimers; M, monomers. A, β-actin, used as a loading control.

The most striking observation was that stable higher-order oligomers were detected for six of the 11 mutants (Fig. 4, O bands). Oligomers were particularly prominent in deletion mutants Del_167–175 and Del_167–183, length compensation mutants Sub_167–175 and Sub_167–183, where repeating serine and glycine residues were used to restore the length of 9- or 17-amino-acid deletions, and insertion mutant Ins_167+2×SG, where 2 repeating serine and glycine pairs were used to extend the length of the unresolved head segment. Thus, certain mutations resulted in the generation of stable nonstandard higher-order H oligomers under nonreducing SDS-PAGE conditions.

Higher-order H oligomers of deletion mutants efficiently reach the cell surface but fail to trigger fusion.

Finally, we focused on two mutants with moderate or strong defects in membrane fusion function, which are the two deletion mutants Del_173–178 and Del_167–183, respectively. We analyzed their cell surface expression levels by protein biotinylation, and as controls we also assessed the cell surface expression levels of standard H and of the fully functional Del_167–169 mutant.

Figure 5 (left half) shows that dimeric and oligomeric species of the two functionally impaired mutants are transported to the cell surface about as efficiently as the corresponding species of the two functional H proteins. As expected, β-actin was detected only intracellularly (bottom panels, right five lanes). Thus, while higher-order H oligomers of mutants with larger deletions efficiently reach the cell surface, they fail to trigger fusion.

FIG 5.

Higher-order H oligomers of two deletion mutants are efficiently transported to the cell surface. Vero cells transfected with the indicated expression plasmids were lysed at 24 hours posttransfection. Biotin-labeled surface proteins (left) or total cell protein extracts (right) were separated on nonreducing SDS-PAGE and subjected to anti-H-protein immunoblotting (top gel). The same extracts were then blotted for β-actin (bottom gel). Molecular mass markers are indicated on the left side. O, stable higher-order H oligomers; D, dimers; M, monomers; A, β-actin.

DISCUSSION

We show here that a structurally unresolved 17-residue segment of the MeV H protein head must have a defined length to favor proper H tetramerization and promote efficient membrane fusion triggering. We propose that during H-tetramer maturation this segment would act as a “leash” long enough to allow the H heads to adopt a triggering-permissive conformation but short enough to limit their roaming and improper contacts with heads of neighboring tetramers.

Genus-specific variations of the conserved paramyxovirus membrane fusion mechanism are currently being characterized (9, 10, 23, 24). The paramyxovirus head repositioning model was suggested based on structural analyses of entire HN protein ectodomains, which crystallized in distinct conformations about the stalk: 4 heads down, 4 heads up, and 2 heads up/2 heads down (12, 13, 19). While in all these conformations the globular heads behave like rigid bodies maintaining similar structures, they might take advantage of flexible linkers to relocate from a metastable “down” to the “up” position, a movement that may trigger membrane fusion.

As to the H proteins of MeV and the other morbilliviruses, atomic structures of neither their entire ectodomains nor their isolated stalks are available. However, we and others have obtained data suggesting that morbilliviruses developed a variation of the head repositioning model accounting for intracellular H stalk to F-trimer complex formation without premature membrane fusion (24, 39). Namely, it was proposed (24) that in the “down” position the H heads contact an additional tetrameric “spacer” segment (32) located above the signal-transmitting central stalk segment, rather than contacting the central stalk segment directly like the HN heads (13). This arrangement provides room for morbillivirus F trimers to interact laterally with H tetramers already during intracellular transport, as is known to happen (40). A third variation on this theme may be played by the G proteins of henipaviruses, whose stalks also interact laterally with the respective F protein (9, 23).

Here, we observed that excessive shortening of the unresolved 167–183 H-head segment impacts proper H oligomerization and efficient fusion triggering. Specifically, deletion of 6 residues or more resulted in the accumulation of stable higher-order H oligomers under nonreducing SDS-PAGE conditions. In contrast, the standard H protein presents exclusively as a dimer under these conditions. Similarly, smaller deletions of 3 or 5 residues ran exclusively as dimers and maintained fusion function. Length compensation with Ser/Gly repeats fully restored the fusion-triggering function of a 9-residue deletion and partially restored the function of the entire 17-residue deletion. Thus, in combination with the dimeric top segments of the H stalk (32, 33), the unresolved 167–183 H-head segment may act as a leash, which must be long enough for the heads to move about the stalk and adopt the metastable “heads-down” conformation.

Interestingly, mutants with added Ser/Gly repeats at the N-terminal side of the 17-residue segment, as well as Ser/Gly compensation mutants, also accumulated stable higher-order H oligomers. In contrast, none of the alanine substitution mutants generated higher-order oligomers. The absence of a β carbon in Gly makes Gly-rich sequences more flexible than sequences of similar length without Gly (41). It was previously observed that very short or very long linkers promote “cross-folding” between subunits of oligomeric proteins, which can interfere with function (41, 42). Analogously, we propose here that H heads with longer or more flexible linkers have high propensity to roam, contact the heads of other tetramers, and cross-link. We note that, lacking a mutant with alanine substitution of the entire segment, we cannot discount some sequence dependence on proper H oligomerization. However, blocks of 3-amino-acid deletions across the entire length of the segment did no affect function (data not shown), further suggesting that sequence is not critical.

Finally, the presence of higher-order H oligomers is not diagnostic of fusion function. Two large deletions and a 4-amino-acid flexible insert at the beginning of the 17-residue segment resulted in an accumulation of stable higher-order H oligomers and impacted fusion function. On the other hand, restoring the length of the segment with a flexible Ser/Gly insert also resulted in the accumulation of stable higher-order oligomers, while maintaining function. This may be a reflection of the relative ratio of standard to nonstandard H oligomers generated by the different mutations. Mutations that allow the formation of sufficient amounts of the standard H tetramer may retain function. It is also likely that different types of higher-order H oligomers are generated. However, at this point we cannot discriminate between these H-oligomer types.

How the structurally unresolved 167–183 H-head segment favors proper tetramerization and membrane fusion triggering is not fully understood. However, residues 167 to 183 are part of a longer segment beginning with Cys154 and ending with Cys188. Cys154 forms a disulfide bridge with Cys154 of another H monomer within an H-tetramer subunit, defining the top of the stalk dimeric linker. Cys188 forms a long-range disulfide bridge with Cys606 of the same subunit, stabilizing the six-blade propeller globular head (14). Thus, it is possible that the entire 155–187 H-head segment rearranges during membrane fusion triggering. Indeed, we previously showed that cross-linking of H dimers through the introduction of Cys at position 161 reversibly blocks H-head movement and fusion triggering (11). Since sequence alignment of morbilliviruses indicates that residues 155 to 187 are poorly conserved (Fig. 1B, conservation level), we suggest that these residues, in combination with the dimeric top segments of the stalk, may act as leashes for the heads of all morbilliviruses.