Morbilliviruses use signaling lymphocyte activation molecules (CD150) as cellular receptors

Abstract

Morbilliviruses comprise measles virus, canine distemper virus, rinderpest virus, and several other viruses that cause devastating human and animal diseases accompanied by severe immunosuppression and lymphopenia. Recently, we have shown that human signaling lymphocyte activation molecule (SLAM) is a cellular receptor for measles virus. In this study, we examined whether canine distemper and rinderpest viruses also use canine and bovine SLAMs, respectively, as cellular receptors. The Onderstepoort vaccine strain and two B95a (marmoset B cell line)-isolated strains of canine distemper virus caused extensive cytopathic effects in normally resistant CHO (Chinese hamster ovary) cells after expression of canine SLAM. The Ako vaccine strain of rinderpest virus produced strong cytopathic effects in bovine SLAM-expressing CHO cells. The data on entry with vesicular stomatitis virus pseudotypes bearing measles, canine distemper, or rinderpest virus envelope proteins were consistent with development of cytopathic effects in SLAM-expressing CHO cell clones after infection with the respective viruses, confirming that SLAM acts at the virus entry step (as a cellular receptor). Furthermore, most measles, canine distemper, and rinderpest virus strains examined could any use of the human, canine, and bovine SLAMs to infect cells. Our findings suggest that the use of SLAM as a cellular receptor may be a property common to most, if not all, morbilliviruses and explain the lymphotropism and immunosuppressive nature of morbilliviruses.

FIG. 1

Cell tropism of CDV strains. CHO, Vero, and B95a cells were infected with the HA7 strain or Onderstepoort strain of CDV at an MOI of 0.1. Cells were observed at 24 h after infection.

FIG. 2

CHO cell clones stably expressing SLAMs of various species. (A) CHO.Neo, CHO.B95aSLAM, and CHO.SLAM cells were stained with IPO-3 (solid profile) or mouse control IgG antibody (open profile), followed by staining with FITC-labeled goat anti-mouse IgG. (B) CHO.Neo, CHO.DogSLAMtag, and CHO.CowSLAMtag cells were stained with anti-influenza virus HA epitope MAb 12CA5 (solid profile) or mouse control IgG antibody (open profile), followed by staining with FITC-labeled goat anti-mouse IgG.

FIG. 3

CDV infection of marmoset SLAM-expressing CHO and B95a cells. CHO.Neo, CHO.B95aSLAM, and B95a cells were either untreated or treated with IPO-3 and then infected with the HA7 strain of CDV at an MOI of 0.5 (0.1 for B95a cells). Cells were observed at 24 h after infection.

FIG. 4

Predicted amino acid sequences of canine, bovine, and human SLAMs. Amino acid sequences of canine, bovine, and human SLAMs are aligned. Residues having similarity are shaded (dark shading, identical residues; light shading, conservative changes). The predicted signal peptides of respective SLAMs and transmembrane domain of human SLAM are underlined. Potential N-linked glycosylation sites are circled. Cysteine residues predicted to make disulfide bonds in the Ig C2 domain are indicated by asterisks. Tyrosine-based signaling motifs are boxed. Spaces (indicated by dashes) were introduced for optimal comparison.

FIG. 5

CDV and RPV infections of SLAM-expressing CHO cell clones. CHO.Neo, CHO.DogSLAMtag, CHO.SLAM, and CHO.CowSLAMtag cells were infected with the HA7 or Onderstepoort strain of CDV or Ako strain of RPV at an MOI of 0.1. Cells were observed at 24 h after infection with CDV strains or at 12 h after infection with the RPV strain.

FIG. 6

Infectivities of pseudotype viruses on CHO cells expressing canine, bovine, or human SLAM. The indicated CHO cell clones were infected with VSVΔG*-G, VSVΔG*-OPHF, VSVΔG*-HA7HF, VSVΔG*-EdHF, VSVΔG*-KAHF, or VSVΔG* (A) or with VSVΔG*-G or VSVΔG*-KOHF (B), and infectious titers were measured by counting the number of GFP-expressing cells.