Fig 1.
Phylogenetic analysis of HAstVs and delineation of the HAstV-VA1 capsid spike domain.
(A) Phylogenetic analysis of human and animal astroviruses. Complete capsid sequences were aligned using the MUSCLE Algorithm, and evolutionary analysis was done using MEGA 11 maximum Likelihood and JTT matrix-based model to yield the cladogram shown. Bootstrap values are shown next to the branches and were computed using 1,000 replicates. Turkey astrovirus 1–3 capsid sequences were used as an outgroup. (B) Schematic of HAstV-8 and HAstV-VA1 capsid structural domains and recombinant HAstV-VA1 capsid protein constructs. Question marks indicate the unknown termini of the spike domain and acidic region. (C) Coomassie-stained SDS-PAGE of limited proteolysis of HAstV-VA1 Capsid C-term with trypsin. Lane1: molecular weight marker, in kD (Biorad Precision Plus Protein Dual Color Standards). Lane 2: trypsin digestion products showing bands for the 43 kD capsid C-term, the 36 kD Spike trypsinized, and trypsin. (D) Superdex 200 size-exclusion chromatography trace of recombinant HAstV-VA1 spike dimer (orange) overlaid with the trace of gel filtration standards (black dotted line). (E) Structural alignment between the 1.46 Å-resolution crystal structure of HAstV-VA1 spike and the AlphaFold2-predicted HAstV-VA1 spike, which was predicted for the 15th Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction (CASP15). A structural alignment was performed using TM ALIGN software with a TM-Score of 0.85 and an RMSD of 2.90 Å across 264 residue pairs.
Table 1.
Crystallography data collection and refinement statistics.
Fig 2.
Comparison of the HAstV-VA1 spike structure to the classical and MLB HAstV spike structures.
(A) Structure of HAstV-VA1 spike dimer presented as a cartoon model with labeled features on one protomer colored rainbow from the N-terminus (blue) to the C-terminus (red) (top panel). Below, the dimer is presented as a surface model (red and grey) as the side view (middle panel) and top view (bottom panel). (B) Structure of classical HAstV-2 spike dimer (PDB: 5W1N) presented as a cartoon model and colored rainbow (top panel) and presented as a surface model (middle and bottom panels). (C) Structure of HAstV-MLB spike dimer (PDB: 7UZT) presented as a cartoon model and colored rainbow (top panel) and presented as a surface model (middle and bottom panels). Flexible residues that were not visible in each structure are drawn as dashed lines.
Fig 3.
In vitro HAstV infection competition assay.
Caco-2 cells were infected with HAstV-1 (red) or HAstV-VA1 (black) in the presence of the indicated concentration of recombinant HAstV-1 spike (squares) or VA1 spike (circles). Recombinant spike blocks infectivity by homologous HAstV but not heterologous HAstV. The data represent the HAstV infectivity in cells in the presence of each recombinant spike compared with infectivity in the absence of recombinant spike. The arithmetic means ± SEM from three independent experiments performed in duplicate are shown. ***p < 0.001.
Fig 4.
Processing and assembly of the mature HAstV-VA1 capsid protein.
(A) The complete amino acid sequence of the HAstV-VA1 capsid precursor protein (GenBank accession number YP_003090288.1) is shown. Amino acids highlighted in yellow were identified by mass spectrometry as peptides of VP33, and those highlighted in green correspond to peptides of the VP38. Amino acids that are not observed are highlighted in grey. The cleavage site that generates the amino-terminal end of VP38, previously determined by Edman degradation, is shown in bold. The blue boxes correspond to the N- and C-termini of the core domain, whereas the red boxes correspond to the N- and C-termini of the spike domain. The pink box corresponds to amino acid E296, the last amino acid detected in VP33. (B) Coomassie-stained SDS-PAGE analysis of CsCl-purified virus. Lane 1: molecular weight marker, in kD. Lane 2: two bands corresponding to the HAstV-VA1 capsid proteins VP33 and VP38. Bands were excised and utilized for proteolytic digestion, liquid chromatography, and tandem mass spectrometry to identify peptides. (C) Model of the mature HAstV-VA1 capsid protein. An AlphaFold2-predicted core domain structure, a linker region, and the spike domain crystal structure are shown, colored as in panel A. The second protomer of the spike domain is colored grey. The N-terminal amino acids 1–71 that are present in VP33 are not shown. Domain termini are labeled (R72 and P389 (core domain), and P408 and P680 (spike domain)). The location of the cleavage site that results in the N-terminus of VP38 (N347/T348) is indicated. The location of E296, the last observed amino acid of VP33, which is in a structurally similar site as a known trypsin cleavage site in classical HAstVs, is indicated. (D) Model of the mature classical HAstV-8 capsid protein. The crystal structure of the HAstV-8 core domain (PDB: 5IBV), a linker region, and the crystal structure of the HAstV-8 spike domain (PDB: 3QSQ) are shown. The second protomer of the spike domain is colored grey. The N-terminal amino acids 1–76 that are present in VP34 are not shown. Domain termini are labeled (R77 and L412 (core domain), and E424 and P645 (spike domain)). The location of the trypsin cleavage site that results in the N-terminus of VP27 (R393/Q394) is indicated. The location of R313, a trypsin cleavage site, is indicated.
Fig 5.
Comparison of HAstV-VA1gastro and HAstV-VA1neuro spike sequences and structures.
(A) Five representative HAstV-VA1gastro spike sequences (blue) and all five available HAstV-VA1neuro spike sequences (red) aligned using the MUSCLE algorithm. Sequence differences between HAstV-VA1gastro spikes are colored light pink, and sequence differences between HAstV-VA1neuro spikes are colored yellow. (B) Location of variations (yellow and light pink) mapped onto the the VA1 spike structure, presented as a surface model from side, top, and bottom views. (C) Structural alignment between the crystal structures of the HAstV-VA1gastro spike (blue)(accession no: YP_003090288.1) and the HAstV-VA1neuro spike (red)(accession no: ADH93577.1), presented as cartoon view.
Fig 6.
Antigenic analyses of HAstV-VA1gastro and HAstV-VA1neuro spikes.
(A) Coomassie-stained SDS PAGE and anti-HAstV-VA1gastro Western Blot analyses of recombinant HAstV-VA1gastro and HAstV-VA1neuro spike proteins. Lane 1: molecular weight marker, in kD (Biorad Precision Plus Protein Dual Color Standards). Lane 2: HAstV-VA1gastro spike. Lane 3: HAstV-VA1neuro spike. (B) Anti-HAstV-VA1gastro ELISA. Wells were coated with either HAstV-VA1gastro spike (blue) or HAstV-VA1neuro spike (red), and the immunoassay was performed against a serial dilution of anti-HAstV-VA1gastro rabbit serum. Error bars indicate the standard deviation of duplicates. (C) Anti-HAstV-VA1gastro BLI-ISA. After an initial baseline step, histidine-tagged HAstV-VA1gastro spike protein (blue) or HAstV-VA1neuro spike protein (red) or no protein (cyan) were loaded onto Anti-Penta-His biosensors, followed by another baseline step. Biosensors were then dipped into a 1:20 dilution of anti-HAstV-VA1gastro rabbit serum containing polyclonal antibodies (pAb) for 10 minutes. Biosensors were then dipped into buffer to evaluate pAb dissociation. Signal changes during the HAstV-VA1 spike loading step and during the polyclonal antibody association step were measured. (D) BLI-ISA average signal changes during the HAstV-VA1 spike loading step and during the polyclonal antibody association step. Bars represent the mean of three independent experiments, and error bars indicate the standard deviation. A two-tailed T-test was performed to evaluate significance.