Amino acid substitutions in VP2 residues contacting sialic acid in low-neurovirulence BeAn virus dramatically reduce viral binding and spread of infection

Abstract

Theiler's murine encephalomyelitis viruses (TMEV) consist of two groups, the high- and low-neurovirulence groups, based on lethality in intracerebrally inoculated mice. Low-neurovirulence TMEV result in a persistent central nervous system infection in mice, leading to an inflammatory demyelinating pathology and disease. Low- but not high-neurovirulence strains use sialic acid as an attachment factor. The recent resolution of the crystal structure of the low-neurovirulence DA virus in complex with the sialic acid mimic sialyllactose demonstrated that four capsid residues make contact with sialic acid through noncovalent hydrogen bonds. To systematically test the importance of these sialic acid-binding residues in viral entry and infection, we mutated three VP2 puff B amino acids proposed to make contact with sialic acid and analyzed the consequences of each amino acid substitution on viral entry and spread. The fourth residue is in the VP3-VP1 cleavage dipeptide and could not be mutated. Our data suggest that residues Q2161 and G2174 are directly involved in BeAn virus attachment to sialic acid and that substitutions of these two residues result in the loss of or reduced viral binding and hemagglutination and in the inability to spread among BHK-21 cells. In addition, a gain of function-revertant virus was recovered with the Q2161A mutation after prolonged passage in cells.

FIG. 1.

Stick model showing the binding site on sialic acid (SAL) for VP2 puff B residue Q2161. Carbon atoms are off white, nitrogen blue, and oxygen red, and hydrogen bonds are yellow. (A) The Q2161 side chain nitrogen forms a hydrogen bond with a sialic acid hydroxyl group (OH1). Residues A2163 and G2174 near the binding site are labeled. LAC, lactose moiety of sialyllactose bound in the crystal structure. (B) Substitution of the glutamate side chain with that of the alanine eliminates the group that binds sialic acid in this site. (C) Substitution of the glutamate side chain with that of arginine is too bulky to allow binding of sialic acid. (D) Substitution of the glutamate side chain with that of tryptophan is too bulky to allow binding to sialic acid.

FIG. 2.

Viral replication in BHK-21 cells after electroporation of parental and mutant virus RNA transcripts. (A) BHK-21 cells were seeded on plastic, and monolayers were fixed and stained with an immunoperoxidase conjugate to detect BeAn virus antigens at 15 and 48 h after electroporation. Photomicrographs were taken at ×400. Results are representative of three independent experiments. (B) BHK-21 cells were seeded at 10⁴ cells/well in 96-well microplates after electroporation, and cell viability was determined by MTT assay in eight replicate samples from three independent experiments. Results are mean number of viable cells (± standard deviation) at 18 and 48 h. OD560, optical density at 560 nm.

FIG. 3.

BHK-21 cells were electroporated with either parental or mutant BeAn RNA transcripts, seeded in 100-mm² wells, labeled with l-[³⁵S]methionine at 5 h, harvested at 24 h, and run on 20 to 70% sucrose gradients. (A) Gradient fractions (0.5 ml) were assessed for radioactivity by scintillation counting, and the first 16 fractions were plotted; 160 S peaks were seen in fraction 5. (B) An aliquot of each peak fraction was analyzed by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis. Molecular weight markers are indicated on the left, and capsid proteins are indicated on the right. Lanes: 1, parental BeAn; 2, Q2161A; 3, Q2161R; 4, Q2161W; 5, G2174F; 6, G2174W; 7, purified stock BeAn virus.

FIG. 4.

(A) Viral binding to BHK-21 cells in suspension was performed at 4°C by using peak fractions of the sucrose gradient (5 × 10³ cpm) shown in Fig. 3. (B) Viral binding to untreated and neuraminidase (NA)-treated (50 mU) BHK-21 cells was carried out with stock BeAn virus labeled with l-[³⁵S]methionine. Cell-associated radioactivity (means ± standard deviations of three samples) at 1 and 30 min was determined by scintillation counting as described in Materials and Methods.

FIG. 5.

Hemagglutination analysis of stock BeAn virus (lysate) and parental and mutant BeAn viruses from sucrose gradient peak fractions shown in Fig. 3. (A) Human type O erythrocytes (0.5%) were pretreated with either 50 mU neuraminidase or PBS at 24°C for 1 h, or BeAn virus was preincubated with glycophorin A and used in a hemagglutination inhibition assay with BeAn stock virus. (B) Sucrose gradient peak fractions (5 × 10³ cpm) of parental and mutant BeAn viruses were compared in the hemagglutination assay.

FIG. 6.

CPE after passage of lysates from BHK-21 cells electroporated with parental and mutant BeAn virus RNA transcripts. When extensive CPE developed in the parental BeAn virus-infected monolayers at 24 h postinfection, cells were fixed and stained with crystal violet stain and photographed. The lack of staining of monolayer C indicates complete CPE. (A) Uninfected control; (B) ΔBeAn; (C) parental BeAn; (D) Q2161A; (E) Q2161R; (F) Q2161W; (G) G2174F; (H) G2174W; (I) A2163F.

FIG. 7.

Characteristics of the Q2161A revertant virus. (A) The mutant virus plaque size was compared to the small plaque size of parental BeAn and the large plaque size of the high-neurovirulence GDVII virus. (B) Hemagglutination was abolished by neuraminidase (NA) treatment of erythrocytes (RBC), whereas the mutant Q2161A virus did not hemagglutinate (Fig. 5B). (C) MTT cell viability assay in sialic acid-deficient Lec-2 cells after infection with GDVII, parental BeAn, and Q2161A revertant viruses is shown. OD560, optical density at 560 nm.