Biophysical characterization of the cetacean morbillivirus haemagglutinin glycoprotein

Highlights

• •Cetacean morbillivirus (CeMV) recombinant haemagglutinin ectodomain (rH-ecto) is designed.
• •CeMV rH-ecto is purified as soluble, globularly folded and glycosylated protein.
• •CeMV rH-ecto is thermally stable and exists in solution in the form of homo-dimers.
• •CeMV rH-ecto molecular architecture features morbilliviral H structural properties.

Abstract

Cetacean morbillivirus (CeMV) is an enveloped, non-segmented, negative-stranded RNA virus that infects marine mammals, spreading across species and causing lethal disease outbreaks worldwide. Among the eight proteins encoded by the CeMV genome, the haemagglutinin (H) glycoprotein is responsible for the virus attachment to host cell receptors. CeMV H represents an attractive target for antiviral and diagnostic research, yet the elucidation of the molecular mechanisms underlying its role in infection and inter-species transmission was hampered thus far due to the unavailability of recombinant versions of the protein. Here we present the cloning, expression and purification of a recombinant CeMV H ectodomain (rH-ecto), providing an initial characterization of its biophysical and structural properties. Sodium dodecyl sulphate - polyacrylamide gel electrophoresis (PAGE) combined to Western blot analysis and periodic acid Schiff assay showed that CeMV rH-ecto is purifiable at homogeneity from insect cells as a secreted, soluble and glycosylated protein. Miniaturized differential scanning fluorimetry, Blue Native PAGE and size exclusion chromatography coupled to multiangle light scattering revealed that CeMV rH-ecto is globularly folded, thermally stable and exists in solution in the oligomeric states of dimer and multiple of dimers. Furthermore, negative stain electron microscopy single particle analysis allowed us to delineate a low-resolution molecular architecture of the CeMV rH-ecto dimer, which recapitulates native assemblies from other morbilliviral H proteins, such as those from measles virus and canine distemper virus. This set of experiments by orthogonal techniques validates the CeMV rH-ecto as an experimental model for future biochemical studies on its structure and functions.

1. Introduction

Cetacean morbillivirus (CeMV) is an enveloped virus with non-segmented, negative sense single-stranded RNA (ssRNA) genome that belongs to the genus Morbillivirus (species Morbillivirus ceti) of the family Paramyxoviridae (subfamily Orthoparamyxovirinae) within the order Mononegavirales (Seki and Takeda, 2022). CeMV infects a wide range of marine mammal species, is globally widespread and evolutionarily distinct into two lineages and six recognized strains, named after the species in which they were first identified, including the dolphin morbillivirus (DMV), the porpoise morbillivirus (PMV), the beaked whale morbillivirus (BWMV), the pilot whale morbillivirus (PWMV), the Guyana dolphin morbillivirus (GDMV) and the recently characterized Fraser´s dolphin morbillivirus (FDMV) (Jo et al., 2018; West et al., 2021). All CeMV strains are deemed as highly pathogenic, causing an acute respiratory disease that often results in fatal outcome due to the induced profound immunosuppression, or may otherwise evolve into chronic infection with late neurological manifestations (Van Bressem et al., 2014). Moreover, CeMV is one of the major non-anthropogenic stressors that may contribute to the decline of marine mammal populations, and its detection in species already endangered such as the Guiana dolphin (Sotalia guianensis) (Groch et al., 2018), the southern right whale (Eubalaena australis) (Groch et al., 2019), the Mediterranean monk seal (Monachus monachus) (Osterhaus et al., 1997; 1998; van de Bildt et al., 1999; 2001; Petrella et al., 2021) and the killer whale (Orcinus orca) (Groch et al., 2020) implies that CeMV represents a looming threat to their survival on a large scale. Favored by the gregarious behavior of susceptible hosts, CeMV horizontal transmission is thought to occur after inhalation of aerosolized virus from the exhaled breath of infected individuals (Groch et al., 2021), whereas vertical transmission may also occur during pregnancy or lactation (Fernández et al., 2008; Di Guardo et al., 2011; West et al., 2015). CeMV primarily infects lymphocytes, spreading then systemically through epithelia, with the virus utilizing the signaling lymphocytic activation molecule (SLAM; also known as cluster of differentiation 150, CD150) and the poliovirus receptor‐like 4 (PVRL‐4; commonly known as nectin‐4) as lymphocytic and epithelial cell receptors, respectively (Delpeut et al., 2014; Ohishi et al., 2019). Among the products encoded by the eight genes of the CeMV genome, the one deputed to viral attachment to either the SLAM or the nectin-4 receptor is the haemagglutinin (H) glycoprotein (Blixenkrone-Møller et al., 1996; Takeda et al., 2020). Comparative analysis of the measles virus (MeV) H structure with the amino acid sequence of orthologs from other morbilliviruses has shown that also CeMV H is a type II integral membrane protein, which comprises an N-terminal endodomain facing either the cytoplasm or the virion lumen, a transmembrane domain and a C-terminal ectodomain decorating the outer of the virus envelope (Colf et al., 2007). While binding of CeMV H to nectin-4 has been poorly investigated, its interaction with SLAM appears to be characterized by a broad host tropism and tendency for cross-species transmission. In fact, as suggested by in silico homology modeling using as template the structure of the ortholog MeV H in complex with marmoset (Saguinus oedipus) SLAM, CeMV H is predicted to bind with low specificity to a cluster of 35 residues within the SLAM immunoglobulin-like variable (V) domain, whose sequence varies among cetacean species (Ohishi et al., 2010; Shimizu et al., 2013; Beffagna et al., 2017). Moreover, functional studies on Vero cells expressing the dolphin or the seal receptor have shown that CeMV H can indifferently bind to the SLAM of both cetaceans and pinnipeds, while being unable to interact with the same human receptor (Jo et al., 2018; Seki et al., 2020). The H gene has been the target of several works aimed at deciphering CeMV molecular epidemiology (Centelleghe et al., 2016; Mazzariol et al., 2016, 2017; Rubio-Guerri et al., 2018) and evolutionary dynamics (van de Bildt et al., 2005; Banyard et al., 2011; Bellière et al., 2011; Jo et al., 2018), whereas the H protein was used as antigenic marker for the diagnosis of CeMV infection during post mortem examination of cetaceans showing no signs of viral disease (Domingo et al., 1992; Domingo et al., 1995). Additionally, together with the one encoding for the fusion (F) protein, the H gene was evaluated for the immunogenicity properties of its encoded product during the development of an experimental DNA vaccine against CeMV (Vaughan et al., 2007). Nevertheless, despite the significant progress made towards a better understanding of CeMV cross-species transmission and evolutionary dynamics, the molecular mechanisms through which CeMV H exerts its function during virus entry, along with its role in the pathogenesis of infection and experimental data about its structure are currently lacking, which at least in part is due, to the best of our knowledge, to the unavailability of a recombinant version of the protein (Zinzula et al., 2022). With the aim of filling this knowledge gap, we report herein the cloning, expression and purification in a heterologous system of a recombinant CeMV H ectodomain (rH-ecto), together with a preliminary characterization of its biophysical properties through a set of experiments by orthogonal techniques. As a novel resource suitable to perform functional and structural in vitro studies, as well as for diagnostic and antiviral research, rH-ecto is a valuable biochemical tool that offers an array of new opportunities to get insights into CeMV molecular biology and pathogenesis of viral infection.

2. Materials and methods

2.1. Construct design and molecular cloning

The molecular evolution of CeMV H within the Morbillivirus genus was analyzed by performing multiple sequence alignment (MSA) and building a phylogenetic tree with Clustal Omega server (Sievers and Higgins, 2021). Results were visualized with FigTree (http://tree.bio.ed. ac.uk/software/figtree/) and GraphPad Prism v.9.4.1 (https://www.graphpad.com/scientific-software/prism/) software. For the analysis, RefSeq amino acid sequences were retrieved from the NCBI Protein database (Canine distemper virus, CDV, GenBank: NP_047206.1; CeMV, GenBank: NP_945029.1; Feline morbillivirus, FeMV, GenBank: YP_009512963.1; Gierle apodemus virus, GaV, GenBank: UQM99546.1; Longquan Berylmys bowersi morbillivirus 1, LBbMV-1, GenBank: UBB42344.1; Myotis bat morbillivirus, MbMV, GenBank: UBB97712.1; MeV, GenBank: NP_056923.1; Phocine distemper virus, PDV, GenBank: YP_009177603.1; Porcine morbillivirus, PoMV, GenBank: QWQ56142.1; Rinderpest virus, RPV, GenBank: YP_087125.2; Raton olivaceo morbillivirus, RoMV, GenBank: DAZ91188.1, Small ruminants morbillivirus, SRMV, GenBank: YP_133827.2; Wufeng Niviventer fulvescens morbillivirus 1, WNfMV-1, GenBank: UBB42351.1). CeMV rH-ecto truncation was designed by comparing intrinsic disorder and structural domain boundaries predictions between CeMV, MeV and CDV H amino acid sequences, using IUPred3 (Erdős et al., 2021) software. CeMV H cDNA sequence (GenBank: NC_005283.1) was obtained from a codon-optimized synthetic preparation (BioCat, Germany). Gene truncation comprising amino acid residues 149-604 (GenBank: NP_945029.1) was amplified by PCR and subcloned by sequence and ligation independent cloning (SLIC) into the pCoofy-64 vector, a modified pFastBac-derived plasmid for the expression in baculovirus-infected insect cells of secretory proteins which, to ensure high solubility and yield, are fused downstream to an N-terminally hexahistidine (His₆)-tagged, synthetic small ubiquitin-like modifier (SUMO)-3 (GenBank: AVL_26008.1) (Scholz et al., 2013).

2.2. Protein expression and purification

Protein expression was performed in High Five™ (H5) cells (ThermoFisher Scientific) cultured in EX-CELL® 420 serum free medium (Sigma-Aldrich) for 72 h at 121 rpm and 26 °C. After cell harvesting by centrifugation for 30 min at 4300 g and 4 °C, the medium supernatant was collected and loaded by gravity-flow for affinity purification onto an Econo-Pac® chromatographic column (BioRad) packed with 2 mL Ni-Sepharose High Performance affinity resin (Cytiva) equilibrated in buffer A (25 mM Tris, pH 8.0; 300 mM NaCl; 1% Glycerol; 20 mM Imidazole). After washing and elution with buffer A supplemented with 40 mM and 600 mM Imidazole, respectively, protein was concentrated up to ∼1 mL final volume by centrifugation at 3200 g and 4 °C with an Amicon® Ultra-15 centrifugal filter unit 30 kDa molecular weight cutoff (Merck Millipore), then loaded onto a Superdex 200 10/300 GL column (Cytiva) connected to an ÄKTA Pure fast protein liquid chromatography (FPLC) system (Cytiva) for purification by size exclusion chromatography (SEC) in buffer C (25 mM Tris, pH 7.4; 150 mM NaCl). Peak fractions were pooled, analyzed by sodium dodecyl sulphate (SDS) - polyacrylamide gel electrophoresis (PAGE) and liquid chromatography mass spectrometry (data not shown), then concentrated at ∼ 2.5 mg mL^-1, flash-frozen in liquid nitrogen and stored at -80 °C for subsequent experiments.

2.3. Western blotting (WB) – periodic acid Schiff (PAS) assay

For WB, purified CeMV rH-ecto (∼15 μg) was mixed with NuPAGE™ LDS Sample Buffer (ThermoFisher Scientific) in a 3:1 ratio and loaded onto a NuPAGE™ 4–12% Bis-Tris mini protein gel, with PageRuler™ Prestained Protein Ladder 10–180 kDa as molecular weight marker (ThermoFisher Scientific), then run at 180 V constant current in NuPAGE™ MES SDS Running Buffer (ThermoFisher Scientific). Subsequently, gel was sandwiched between pads soaked in NuPAGE™ Transfer Buffer (ThermoFisher Scientific) for protein blotting onto a 0.2 μm Invitrolon™ PVDF membrane (ThermoFisher Scientific) pre-wet in methanol, by running for 90 min at 30 V on a Trans-Blot semi-dry electrophoretic transfer cell (BioRad). Following transfer, CeMV rH-ecto bands were detected by probing the protein N-terminal His₆-tag using the SuperSignal™ West HisProbe Kit (ThermoFisher Scientific) according to manufacturer's instructions. For glycosylation assessment by PAS assay, ∼10 μg of CeMV rH-ecto and an equal amount of horseradish peroxidase and soybean trypsin inhibitor proteins used as glycosylated and non-glycosylated controls, respectively, were run by electrophoresis and transferred onto a 0.2 μm nitrocellulose membrane (BioRad) in the same conditions as for WB. Glycosylated protein was visualized on the membrane by using the Pierce™ Glycoprotein Staining Kit (ThermoFisher Scientific) following manufacturer's instructions. Briefly, after fixation in 3% acetic acid, the membrane was treated with periodate solution to oxidize cis-diol sugar groups on glycoproteins, after which the resulting aldehyde groups on the glycosylated protein reacted with a reducing solution, allowing detection through the formation of Schiff-base bonds that produced a shift to magenta of their band color.

2.4. Blue native (BN)-PAGE

Purified CeMV rH-ecto (∼40 μg) was mixed with NuPAGE™ NativePAGE Sample Buffer and 5% G-250 Sample Additive (ThermoFisher Scientific) in a 3:1 ratio, and loaded onto a NuPAGE™ 4–16% NativePAGE mini gel, with NativeMark™ Unstained Protein Standard as molecular weight marker (ThermoFisher Scientific). Electrophoresis was run at 4 °C and 150 V constant current for 180 min, in NuPAGE™ NativePAGE Running Buffer and Light-blue Cathode Buffer containing 0.002% G-250 additive according to manufacturer´s instruction (ThermoFisher Scientific). Protein bands on the gel were visualized by Coomassie blue staining with InstantBlue™ protein stain (Expedeon).

2.5. SEC - multiangle light scattering (SEC-MALS)

Purified CeMV rH-ecto (∼100 μg) was loaded by auto-injection onto a Superdex 200 10/300 GL gel filtration column (GE Healthcare) connected to a HPLC system with variable UV absorbance detector set at 280 nm (Agilent Technologies, 1100 series), coupled in line with a mini–DAWN TREOS MALS detector, and followed by an Optilab rEX refractive-index detector (Wyatt Technology, 690 nm laser). Run was performed at 20 °C and 0.5 mL min⁻¹ flow rate in buffer C. Protein absolute molecular mass was calculated with ASTRA 6 software (Wyatt Technology) with the dn dc⁻¹ value set to 0.185 mL g⁻¹ and using bovine serum albumin (ThermoFisher Scientific) as calibration standard. Graph plotting was performed using GraphPad Prism v.9.4.1 (https://www.graphpad.com/scientific-software/prism/) software.

2.6. Miniaturized differential scanning fluorimetry (Nano-DSF)

Purified CeMV rH-ecto was diluted to ∼1 mg mL⁻¹ in buffer C, loaded on standard capillaries and tested over a linear 20–95 °C thermal gradient at 0.5 °C min ^-1 rate in a Prometheus NT.48 Nano-DSF instrument (NanoTemper Technologies). Temperature-dependent shifts in the protein intrinsic fluorescence at emission wavelengths of 330 and 350 nm were measured, and inflection points in the fluorescence transition profile corresponding to the melting temperature (Tm) values were determined as first derivative maxima of the fluorescence intensity ratio at the aforementioned wavelengths (F330/F350). Samples were tested in quadruplicates or triplicates for averaging of Tm values from at least three independent measurements. Data processing and graph plotting were performed using the PR.Therm Control software (NanoTemper Technologies) and the GraphPad Prism v.9.4.1 (https://www.graphpad.com/scientific-software/prism/) software, respectively.

2.7. Negative stain electron microscopy (EM), single particle analysis (SPA) and molecular modelling

Dilution of purified CeMV rH-ecto to ∼0.04 mg mL⁻¹ in buffer C was applied for 2 min at RT to glow-discharged, carbon-coated 400 mesh nickel grids (Electron Microscopy Sciences), which were stained three times (30 s each) with 2% Uranyl formate, blotted with 11.0 μm pore-size filter paper (Whatman) and air-dried. Collection of the micrographs dataset for SPA was performed using a Tecnai F20 FEI transmission electron microscope (TEM) equipped with a BM-Eagle 4 K CCD camera (FEI) operated at 200 kV. Single images were acquired at 62,000 × magnification, corresponding to a calibrated physical pixel size of 1.78 Å, by applying a defocus value range from -2.50 to -3.50 μm. Image processing was performed by using RELION-4.0 software (Kimanius et al., 2021), through which 30,758 particles belonging to consistent classes were selected after automatic picking, extraction and two-dimensional (2D) classification. Of these, 10,818 particles from 8 classes were chosen for the generation of a 3D initial model, which in turn served as template for subsequent 3D classification and assignment of angular orientations. The best 3D class, composed by 21,838 particles, 71 % of the selected ones, was then subjected to 3D-refinment, leading to a volume with ∼20 Å estimated final resolution. The whole atomic model of CeMV H head ectodomain was predicted by AlphaFold2 multimer using the ColabFold server (Mirdita et al., 2022) with default settings for template-based modelling of homo-dimeric complexes. The top five ranked outputs were visually inspected and the most reliable one was fitted to the experimental volume to prepare the figures. Crystal structures of dimeric MeV H, either isolated (PDB: 2ZB6) or complexed to SLAM (PDB: 3ALZ) or nectin-4 (PDB: 4GJT), and dimeric isolated CDV H (PDB: ZNY), were used for comparative structural analysis. Molecular graphics were produced using the Chimera v.1.15 (Pettersen et al., 2004) software.

3. Results and Discussion

3.1. CeMV rH-ecto is glycosylated, globularly folded and thermally stable

Comparison of the CeMV H amino acid sequence with those from other members in the Morbillivirus genus showed that the CeMV protein is phylogenetically related most closely to the orthologs from RPV and SRMV (sequence identity of 46.1 % and 47.2 %, respectively) (Fig. 1a). However, MeV and CDV are the closest relatives for which atomic structures of the H protein have been determined so far (Hashiguchi et al., 2007; Santiago et al., 2010; Hashiguchi et al., 2011; Zhang et al., 2013; Kalbermatter et al., 2023). We therefore relied on the information available from MeV and CDV to derive the CeMV H structural organization, and to map the boundaries between its cytoplasmic (or virion-luminal) tail endodomain (aa 1-37), transmembrane domain (aa 38-58) and ectodomain (aa 59-604), the latter being further organized into the stalk (aa 59-154), neck (aa 155-183) and head (aa 184-604) subdomains (Fig. 1b). Moreover, aiming at designing a construct that would comprise CeMV H antigenically and functionally important regions - i.e., the most exposed ones and those involved in the attachment to cognate receptors SLAM and nectin-4 receptors - and that could be purifiable as soluble in non-denaturing conditions, we focused our study on the portion spanning residues 151–604, namely the rH-ecto, which solely includes the neck and head subdomains of the CeMV H ectodomain (Fig. 1b). In addition, to prevent solubility issues for such truncation and to warrant high protein yield for downstream applications, we subcloned rH-ecto into the pCoofy64, a plasmid vector available in our library for the production of secreted recombinant proteins, which is designed to express the gene of interest as N-terminally fused to a His₆-tagged SUMO-3 (Scholz et al., 2013). Noteworthy, when compared to native full-length CeMV, MeV and CDV H for intrinsic disorder prediction, the presence of the SUMO-3 fusion partner was not predicted to impact the overall protein folding beyond the local level (Fig. 1c). In such chimeric configuration, CeMV rH-ecto was then purified at homogeneity from the culture medium of transfected insect cells (Fig. 1d). Moreover, by means of SDS-PAGE and WB analysis, purified CeMV rH-ecto displayed one major protein band whose ∼70 kDa apparent molecular weight (MW) was slightly higher than the one theoretically expected for a monomer (MW 63.5 kDa), together with other faint bands showing even higher apparent molecular masses (Fig. 1e). Noteworthy, discrepancy between experimental and theoretical MW was also reported for a recombinant H protein from RPV expressed in insect cells using a baculovirus system, and hypothesized to be the result of glycosylation (Naik and Shaila, 1997). In line with such prediction, we also interpreted the observed pattern of migration into the gel as the result of post-translational modifications acquired by the recombinant protein during its ectopic expression in a eukaryotic system. In order to verify this hypothesis, we performed a PAS assay for the specific detection of protein glycosylation. As shown by WB analysis, both the major band and the upper ones displayed by CeMV rH-ecto tested positive to PAS staining, thereby confirming the glycosylated state of the protein (Fig. 1f). Next, we sought to evaluate the structural integrity and thermal stability of the rH-ecto preparation by miniaturized differential scanning fluorimetry (Nano-DSF), a technique that allows to monitor protein unfolding and conformational changes from the shifts in intrinsic fluorescence emission during a controlled heat denaturation (Raynal et al., 2021). Over a 20–95 °C thermal gradient, CeMV rH-ecto intrinsic fluorescence profile showed two very neat inflection points, the first derivative of which revealed corresponding melting temperature (Tm) values of 52.0 ± 0.1 °C (Tm₁) and 64.0 ± 0.1 °C (Tm₂), respectively (Fig. 1g). The measured Tm values are consistent with a globular folding of the protein, and are indicative of relatively high stability at physiological conditions. Nevertheless, the presence of such distinct shifts in intrinsic fluorescence is also reminiscent of sudden conformational changes occurring during thermal unfolding, which in the case of a multimeric protein might be in turn correlated to dissociation events of oligomeric species into their protomers. Of note, a similar profile of thermal unfolding was also observed by differential scanning calorimetry and fluorimetry for the glycoprotein (haemagglutinin-neuraminidase) of Sendai virus, also a paramyxovirus and therefore related to CeMV at the family level, which showed distinct inflection points and relative Tm values reminiscent of the presence of two independently folded units (Manfrinato et al., 2001). In summary, results from this set of experiments demonstrated that the purified His₆-tagged, Sumo3-fused CeMV rH-ecto is soluble, globularly folded and thermally stable, carrying also post-translational modifications typical of a glycoprotein.

Fig. 1.

Design and biophysical characterization of CeMV rH-ecto. (a) Molecular evolutionary analysis of the H proteins from members of the Morbillivirus genus based on a maximum-likelihood phylogenetic tree and an amino acid sequence identity heatmap, both built from a MSA of the NCBI Protein RefSeq GenBank IDs: CDV, NP_047206.1; CeMV, NP_945029.1; FeMV, YP_009512963.1; GaV, UQM99546.1; LBbMV-1, UBB42344.1; MbMV, UBB97712.1; MeV, NP_056923.1; PDV, YP_009177603.1; PoMV, QWQ56142.1; RPV, YP_087125.2; RoMV, DAZ91188.1, SRMV, YP_133827.2; WNfMV-1, UBB42351.1; the target of this work and orthologs with available protein structure are highlighted as circled in regal-blue (CeMV), bright-red (MeV) and lilac (CDV) color in the tree. (b) CeMV genome organization (upper panel), with H gene highlighted in dark gray, and structural layout of encoded H protein (lower panel) with tail (champagne) and transmembrane (tacao) domain and stalk (link-water), neck (cornflower) and head (astral) subdomains highlighted in color. (c) IUPre3 comparative prediction of intrinsic disorder and structural domain organization profile between CeMV (regal-blue), CeMV rH-ecto (downy, cream-brulee, astral), CDV (lilac) and MeV (bright red) H proteins. (d) schematic representation of the structural organization of pCoofy64-SUMO3-CeMV-rH-ecto construct; His₆-tag and SUMO-3 fusion protein are represented as cylinders colored in downy and cream-brulee, respectively. (e) Blue Coomassie-stained SDS-PAGE (left), and anti-His₆ WB analysis (right) of purified CeMV rH-ecto (M, molecular weight marker). (f) WB coupled to PAS glycosylation assay of CeMV rH-ecto; magenta-colored bands are indicative of the Schiff bonds presence and of the protein glycosylated state (M, molecular weight marker; C+ and C-, positive and negative controls, respectively). (g) Nano-DSF analysis of purified CeMV rH-ecto; thermal stability and conformational profile is shown by superimposed F330/350 ratio (azure) and first derivative (outrageous-orange); conformational transitions corresponding to inflections in intrinsic fluorescence and related Tm₁ (52.0 ± 0.1 °C) and Tm₂ (64.0 ± 0.1 °C) are indicated as dashed and dotted lines, respectively. Data points represent averages of three independent experiments. (h) BN-PAGE analysis of purified CeMV rH-ecto (M, molecular weight marker). (i) SEC-MALS analysis (left panel) of purified CeMV rH-ecto; relative absorbance (azure) and absolute molecular mass (outrageous-orange) are indicated as continuous and dashed colored lines, respectively; the main peak (1) at which the average absolute mass (147.6 kDa) and its stoichiometry with respect to expected monomeric mass (2.3 folds) were calculated is indicated; a table (right panel) summarizing the theoretically expected monomeric mass and the experimentally determined oligomeric one is shown.

3.2. CeMV rH-ecto exists in solution as dimer and multiple of dimers

Previous structural works on MeV elucidated that the H ectodomain is a monomer in solution and in the crystallographic asymmetric unit in the absence of the region comprising a cysteine at position 154 (Cys154), which is responsible for the formation of a disulfide bridge (Colf et al., 2007; Zhang et al., 2013). Conversely, when the neck was included in the construct, MeV H was shown, either alone or in complex with SLAM and CD46 receptors, to form homo-dimers and even to weekly assemble into tetrameric form of dimers-of-dimers (Hashiguchi et al., 2007; Santiago et al., 2010; Hashiguchi et al., 2011). More recently, a structure of the H ectodomain from CDV was determined by cryo-electron microscopy (cryo-EM), revealing that in the presence of the full-length stalk such homo-dimers are associated in a stable tetrameric complex (Kalbermatter et al., 2023). Given that Cys154 is highly conserved among morbilliviral H antigens and was included in our construct, we wanted to characterize the CeMV rH-ecto oligomerization profile in solution, and to this aim we subjected the purified protein to BN-PAGE and orthogonal SEC-MALS analysis. As shown by its migration in the gel under non-denaturing conditions, rH-ecto displayed one major band with the apparent MW of a homo-dimer, another band compatible with a homo-tetramer, and one faint band of undefinable multimeric order (probably an octamer) (Fig. 1h). Consistently, the SEC-MALS profile of CeMV rH-ecto revealed one major peak with average absolute molecular mass of 147.6 kDa, which corresponds to 2.3 folds the theoretical mass of the monomeric protein, thereby indicating a homo-dimeric species (Fig. 1i). In addition, a second peak of higher oligomeric order was present in the chromatogram, albeit barely noticeable and with an intensity signal that was too weak for a reliable calculation of its absolute molecular mass. Therefore, results from our experiments showed that CeMV rH-ecto is a homo-dimer in solution and can further assemble into higher-order oligomers as multiples of the dimeric unit, which is in line with previous observations on H ectodomains from closely related morbilliviruses.

3.3. Low-resolution molecular architecture of CeMV rH-ecto recapitulates assembly of morbilliviral glycoproteins

As shown by MeV and CDV structures, the head of morbilliviral H ectodomain is of cubic shape and folds as a β-propeller of six-blades (β1-β6) surrounding a central cavity, with each blade consisting of four antiparallelβ-strands (Colf et al., 2007; Hashiguchi et al., 2007; Kalbermatter et al., 2023). Besides, two heads undergo dimerization via interactions between their β1 and β2 blades, whereas at each head a concave groove between blades β4-β5 and a portion of the β6 blade are involved in binding to CD46, SLAM and nectin-4 receptors (Santiago et al., 2010; Hashiguchi et al., 2011; Zhang et al., 2013). Hence, given the sequence similarity between CeMV, MeV and CDV H, we were interested in understanding if, in the homo-dimeric complex, the two CeMV H ectodomains would spatially arrange as their orthologs from MeV and CDV, and we therefore sought to obtain a preliminary low-resolution molecular architecture of CeMV rH-ecto by negative stain EM SPA. In the micrographs, the negatively stained CeMV rH-ecto appeared as a monodisperse sample with elongated particles of ∼ 1.2 nm in length (Fig. 2a), of which two-dimensional (2D) classification and subsequent 3D volume reconstruction revealed a density map of ∼ 20 Å estimated resolution, which well resembled the peanut-like shape observed in the homo-dimeric MeV and CDV H ectodomains (Fig. 2b). Moreover, not only the volume of the experimental density map was very similar with those obtained by downfiltering MeV and CDV H dimeric structures to the same low-resolution value, but also an AlphaFold2 homology-modelled structure of the homo-dimeric CeMV H head ectodomain could be reliably fitted into it (Fig. 2c). Furthermore, mapping in the modelled structure the positions where SLAM and nectin-4 would interact with CeMV H, as inferred from the known binding sites in the MeV counterpart (Hashiguchi et al., 2011; Zhang et al., 2013), allowed us to confirm that these regions are both solvent-exposed in the volume surface of the experimentally determined molecular architecture, and therefore to conclude that the CeMV rH-ecto retains the structural requirements for the attachment of its cognate receptors (Fig. 2d). Taken together, our findings demonstrate that CeMV rH-ecto recapitulates the characteristic modality of assembly and the spatial arrangement of native morbilliviral haemagglutinin antigens.

Fig. 2.

Low-resolution molecular architecture of CeMV rH-ecto. (a) negative stain EM micrograph of purified CeMV rH-ecto (62,000 ×; scale bar, 50 nm), with three representative single particles in the zoomed insets. (b) selected 2D class averages (left panel) of dimeric CeMV rH-ecto particles, and 3D volume reconstruction (right panel) of dimeric CeMV rH-ecto, shown in three orientations. (c) comparative volumetric analysis (upper panel) of 3D-reconstructed dimeric CeMV rH-ecto with the volumes of dimeric MeV and CDV H ectodomains obtained by downfiltering to 20 Å the respective crystal structures. Atomic model (lower panel) of dimeric CeMV rH-ecto modelled with ColabFold AlphaFold2 and fitted into the SPA 3D-reconstructed volume, shown in two orientations; its spatial arrangement with respect to the remaining portions of the H proteins and the viral envelope is shown in the cartoon. (d) Schematic view of the 3D-reconstructed volume and fitted atomic model of dimeric CeMV rH-ecto (regal-blue) complexed to SLAM (reef-green) and nectin-4 (chestnut-rose), shown in two orientations, obtained by structural alignment with corresponding crystallographic structures of MeV in complex with the two receptors; SLAM and nectin-4 are purposedly shown simultaneously to highlight partial overlap of their putative binding sites.

4. Conclusions

During the last thirty-five years, CeMV epizootics have caused mass mortality outbreaks among cetaceans and pinnipeds worldwide (Jo et al., 2018), and the sudden emergence of this virus in species already endangered by extinction, such as Hawaiian and Mediterranean monk seals (Baker et al., 2017; Petrella et al., 2021), or southern resident killer whales (Weiss et al., 2020), could irreparably jeopardize their recovery, thwarting decades of conservation efforts. Moreover, while vaccination remains an aleatory and logistically challenging hypothesis, surveillance of circulating viral strains in free-ranging individuals and diagnosis of infection in those stranded ashore, represent the only deployable preventive measures at present. Whatever the strategy pursued to mitigate the effects of CeMV spread among cetacean and pinniped populations, there is a compelling need of validating molecular targets for diagnostic and therapeutic countermeasures. We previously elucidated how CeMV packages viral genome by determining the cryo-EM structure of its nucleoprotein-RNA complex (Zinzula et al., 2021), thereby providing the first experimental biomolecular structure from this pathogen. As part of the same effort towards a more detailed understanding of the CeMV molecular biology, we have provided here a preliminary biophysical characterization of its haemagglutinin glycoprotein. This experimental model could serve - with the caveat of a different glycosylation profile from the one obtainable upon expression in mammalian cells - as a more specific tool for seromonitoring compared to diagnostic surrogates from related morbilliviral species (e.g., CDV), and even for developing therapeutic antibodies. In this regard, it is worth noting that recombinant H proteins from RPV produced in insect cells were proven to fully retain their antigenic and immunogenic properties, eliciting neutralizing antibodies and conferring protection against live-attenuated RPV by producing both humoral and cell-mediated immune response in cattle (Naik and Shaila, 1997; Naik et al., 1997; Sinnathamby et al., 2001a, 2001b; Rahman et al., 2003). In addition, the recombinant CeMV H herein described could pave the way for future structural and functional in-depth studies on CeMV cross-species transmission and pathogenesis of viral infection. More in detail, since it was reported that especially striped dolphins (Stenella coeruleoalba) may develop a peculiar, “brain-only” form of CeMV infection sharing pathological similarities with “subacute sclerosing panencephalitis” (SSPE), a neurologic sequela that is rarely encountered in MeV-infected patients (Di Guardo, 2023), the present study could also provide the basis for the design of targeted diagnostics and ad hoc assays in cellulo. Ultimately, this could allow a better pathogenetic characterization of SSPE-like lesions in striped dolphins, before drawing conclusion whether they mirror (or not) their human counterpart phenotype.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Luca Zinzula: Conceptualization, Investigation, Methodology, Validation, Data curation, Formal analysis, Visualization, Supervision, Project administration, Writing – original draft. Judith Scholz: Investigation, Methodology, Data curation, Writing – review & editing. István Nagy: Investigation, Methodology, Formal analysis, Writing – review & editing. Giovanni Di Guardo: Conceptualization, Supervision, Writing – review & editing. Massimiliano Orsini: Conceptualization, Investigation, Methodology, Validation, Data curation, Formal analysis, Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in the paper.

Data availability

• Data will be made available on request.