Previous Article | Next Article ![]()
Journal of Virology, December 2008, p. 11979-11984, Vol. 82, No. 23
0022-538X/08/$08.00+0 doi:10.1128/JVI.00867-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Department of Virology, Faculty of Medicine, Kyushu University, Fukuoka 812-8582, Japan,1 Core Facility for Therapeutic Vectors, The Institute of Medical Science, The University of Tokyo, 4-6-1, Shirokanedai Minato-ku, Tokyo 108-8639, Japan,2 RNA and Biofunctions, PRESTO, Japan Science and Technology Agency, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan3
Received 24 April 2008/ Accepted 2 September 2008
|
|
|---|
|
|
|---|
16 kb that is classified in the genus Morbillivirus in the family Paramyxoviridae (12). The genome is encapsidated by the nucleocapsid (N) protein and associated with a viral RNA-dependent RNA polymerase composed of two subunits, phosphoprotein (P protein) and large (L) protein. The genome and these proteins form a helical ribonucleoprotein complex that acts as the active template for transcription and replication (12). The incidence and mortality of measles have been remarkably reduced in many countries as the vaccine coverage rate increases (12). Live attenuated MV vaccines have been generated via many rounds of passages of clinical isolates of MV in various cultured cells (12, 29). The Edmonston strain, isolated in 1954 (9), was used as a seed strain to generate live attenuated vaccines (17, 29). Although some of the molecular mechanisms by which Edmonston lineage vaccines have adapted to grow in various cultured cells have been elucidated (32, 33, 40), the mechanisms of their attenuation are poorly understood. In the present study, we generated various recombinant MVs with chimeric genomes of the Edmonston vaccine and wild-type strains of MV, and we found that viruses possessing the polymerase protein genes of the Edmonston strain exhibited attenuated MV gene expression and growth in cultured cells as well as in mice expressing a cellular receptor for MV, human signaling lymphocyte activation molecule (hSLAM) (21). These data provide a molecular basis for the avirulence of the Edmonston strain in vivo.
SLAM is a universal receptor for all MV strains, while CD46 functions as an additional receptor for vaccine strains but not for wild-type strains (41, 43). Thus, both wild-type and Edmonston strains replicate efficiently and produce high titers of progeny viruses in SLAM-expressing cells (10, 14). However, when plaque assays were performed, the Edmonston lineage strains produced smaller plaques than wild-type strains of MV (Fig. 1A). Wild-type strains of MV (JPN/50.98/1 and JPN/26.99/1) were isolated from patients with measles and passaged two or three times in B95a cells (15) to make stocks for experiments (kind gifts from H. Sakata). The Edmonston strain obtained from the American Type Culture Collection (ATCC) is designated the Ed-ATCC strain in this paper. Recombinant Edmonston tag (Ed-tag) (26) and wild-type IC323 (39) MV clones were recovered from cDNAs and grown in Vero/hSLAM cells (23). To clarify the genes responsible for the small-plaque-forming phenotype of the Edmonston strain, six recombinant MVs that each contained one of the Ed-tag strain genes in the backbone of the genome of the wild-type IC323 strain expressing enhanced green fluorescent protein (IC323-EGFP) (13) were prepared and designated IC/Ed-N-EGFP, IC/Ed-P-EGFP, IC/Ed-M-EGFP, IC/Ed-F-EGFP, IC/Ed-H-EGFP, and IC/Ed-L-EGFP. IC/Ed-N-EGFP, IC/Ed-P-EGFP, IC/Ed-F-EGFP, IC/Ed-H-EGFP, and IC/Ed-L-EGFP were reported previously and referred to as m2, m3, m7, m8, and m9, respectively (34). When the recombinant MVs were used to infect Vero/hSLAM cells, IC/Ed-P-EGFP and IC/Ed-L-EGFP produced smaller plaques than wild-type IC323-EGFP (Fig. 1B). These data indicate that the P and L genes of the Ed-tag strain contribute to the decreased plaque sizes produced by the Ed-tag strain. In addition to the P protein, the P gene encodes two accessory gene products, V and C (12). The V protein directly interferes with host interferon (IFN) induction and IFN signaling pathways, while the C protein modulates viral RNA synthesis to circumvent IFN induction (19). It should be noted that, unlike the V proteins of the Ed-ATCC and wild-type MV strains, the V protein of the Ed-tag strain is defective in counteracting IFN signaling pathways owing to tyrosine-to-histidine and cysteine-to-arginine substitutions at amino acid positions 110 and 272 (Y110H and C272R), respectively (5, 8, 11, 22). Substitutions at these positions are not unique to the Ed-tag strain and have been found in other vaccines and cultured cell-adapted MV strains, including the CAM-70 vaccine (4), the Changchun-47 vaccine (4), the chicken embryo fibroblast-adapted D-CEF strain (1), and the Vero cell-adapted 94YTV strain (containing the C272R substitution; our unpublished observation). These findings prompted us to carry out further experiments using the P gene of the Ed-tag strain.
![]() View larger version (34K): [in a new window] |
FIG. 1. Small plaque formed by the Edmonston strain and recombinant MVs with the P or L gene of the Edmonston strain on Vero/hSLAM cells. (A) Plaque assays were performed for various MV strains (three wild-type strains [IC323, JPN/50.98/1, and JPN/26.99/1] and two Edmonston lineage strains [Ed-ATCC and Ed-tag]). Monolayers of Vero/hSLAM cells on 12-well cluster plates were infected with 50 PFU of each virus and overlaid with Dulbecco's modified Eagle medium containing 2% fetal bovine serum and 1% methylcellulose. At 5 days p.i., the cells were stained with RTU Vectastain Elite ABC reagent (Vector Laboratories) using anti-MV H-protein monoclonal antibodies and a biotinylated secondary antibody. After high-resolution digital images were obtained, the sizes of all plaques were measured. The mean sizes ± standard deviations are shown in the bar graph. (B) Plaque assays were performed for wild-type IC323-EGFP and six recombinant MVs (each containing one of the Ed-tag strain genes in the backbone of the IC323-EGFP genome) on Vero/hSLAM cells, as described for panel A.
|
10-fold in IC/Ed-P-Luci- and IC/Ed-L-Luci-infected cells, compared with the levels in IC323-Luci-infected cells (Fig. 2B). The viral transcript levels were even further reduced in IC/Ed-PL-Luci-infected cells, being about 100-fold lower than those in IC323-Luci-infected cells (Fig. 2B). These data show that the small-plaque-forming phenotype of the Edmonston strain is at least partly caused by reduced viral gene expression levels in cells. We also analyzed the activities of the P and L proteins using a minigenome assay (18, 19). All combinations of the P- and L-protein-expressing plasmids of the wild-type IC323 and Ed-tag strains produced equivalent Renilla luciferase activities (Fig. 2C). The discrepancy between the data obtained with the minigenome assays and the data obtained using recombinant viruses may be attributed to differences in the gene expression systems. In minigenome assays, the components required for viral RNA synthesis (N, P, and L proteins) are continuously supplied in trans from expression plasmids, whereas in virus infections, MV has to synthesize these proteins by itself to continue viral RNA synthesis. Furthermore, in virus infections, but not in minigenome assays, MV produces the C, V, and matrix (M) proteins, which modulate viral RNA synthesis (3, 24, 27, 31, 42).
![]() View larger version (28K): [in a new window] |
FIG. 2. Attenuated gene expression by recombinant MVs with the P and/or L genes of the Edmonston strain. (A) Confluent monolayers of various cell lines (Vero/hSLAM, CV1/hSLAM, HeLa/hSLAM, CHO/hSLAM, and B95a) and suspensions of nonadherent MT2 cells cultured in 24-well cluster plates were infected with 2.5 x 103 PFU of IC323-Luci (circles), IC/Ed-P-Luci (squares), IC/Ed-L-Luci (triangles), and IC/Ed-PL-Luci (diamonds). After various intervals, the Renilla luciferase activities were measured. Data are means ± standard deviations for triplicate samples. RLU, relative light units. (B) Confluent monolayers of Vero/hSLAM cells in 6-cm culture plates were infected with 1.0 x 104 PFU of IC323-Luci (light gray bars), IC/Ed-P-Luci (black bars), IC/Ed-L-Luci (dark gray bars), and IC/Ed-PL-Luci (white bars) and cultured in the presence of a fusion-blocking peptide. At 18 h p.i., mRNAs were purified from the cells, and the levels of N, P, M, F, H, and L mRNAs were determined by reverse transcription-quantitative PCR. Data are means ± standard deviations for triplicate samples. (C) Minigenome assays. The method was described in detail elsewhere (19). Monolayers of CHO/hSLAM cells cultured in Opti-MEM on 24-well plates were infected with vTF7-3 at a multiplicity of infection of 0.5 and then transfected with 0.2 µg of p18MGFLuc01-wt-Le, 0.2 µg of pCAG-T7-IC-N, 0.3 µg of a P-protein expression plasmid (pCAG-T7-IC-P C or -Ed-P C), and 0.2 µg of an L-protein expression plasmid (pGEMCR-IC-L or -Ed-L) using Lipofectamine 2000 (Invitrogen). At 6 h p.i., the culture media were replaced with RPMI 1640 medium supplemented with 7.5% fetal bovine serum. At 48 h p.i., the firefly luciferase activities were measured. (-), L-protein expression plasmid was omitted. Data are means ± standard deviations for triplicate samples.
|
![]() View larger version (35K): [in a new window] |
FIG. 3. Neutral effects of substitutions in the leader region of the Edmonston strain on MV gene expression levels and plaque sizes. (A) Confluent monolayers of CV1/hSLAM, HeLa/hSLAM, and CHO/hSLAM cells cultured in 24-well cluster plates were infected with 2.5 x 103 PFU of IC323-Luci (circles) and IC/Ed-Le-Luci (triangles). After various intervals, the Renilla luciferase activities were measured. Data are means ± standard deviations for triplicate samples. RLU, relative light units. (B) Plaque assays were performed for the recombinant MVs, as described for Fig. 1A. The IC/Ed-Le-Luci, IC/Ed-LeP-Luci, IC/Ed-LeL-Luci, and IC/Ed-LePL-Luci genomes contain Ed-tag genome regions encoding Le alone, Le and P, Le and L, or Le, P, and L, respectively, in the backbone of the IC323-Luci genome.
|
![]() View larger version (19K): [in a new window] |
FIG. 4. Attenuated gene expression by the Edmonston strain in cultured cells and SLAM-knock-in mice. (A) Plaque assays were performed for IC323-Luci and Ed-tag-VH110Y/R272C-Luci on Vero/hSLAM cells, as described for Fig. 1A. (B) Confluent monolayers of HeLa/hSLAM, A549/hSLAM, CV1/hSLAM, B95a, and CHO/hSLAM cultured in 12-well cluster plates were infected with 2.5 x 103 PFU of IC323-Luci (black bars) and Ed-tag-VH110Y/R272C-Luci (gray bars). At 30 h p.i., the Renilla luciferase activities were measured. Data are means ± standard deviations for triplicate samples. Mock-infected samples are shown by white bars. RLU, relative light units. (C) IC323-Luci, IC/Ed-L-Luci, and Ed-tag-VH110Y/R272C-Luci (1.5 x 105 PFU) were injected into the peritoneal cavities of IFNAR1–/– SLAM-knock-in mice. At 5 days p.i., the spleens were removed and analyzed for their Renilla luciferase activities. Each symbol indicates the sample from a single mouse, and bars indicate the median values. RLU, relative light units. *, P < 0.05; **, P < 0.001.
|
10,000-fold lower than that in spleens from mice infected with wild-type IC323-Luci (P < 0.001; Fig. 4C). These data clearly indicate that the Edmonston strain is severely attenuated in IFNAR1–/– SLAM-knock-in mice. The median Renilla luciferase activity in spleens from IC/Ed-L-Luci-infected mice was
200-fold lower than that in spleens from mice infected with wild-type IC323-Luci (P < 0.001; Fig. 4C). These data indicate that the L gene of the Edmonston strain contributes to the in vivo attenuation of the Edmonston strain. Many viruses can be adapted to various cultured cells by passaging in the cells, and it is empirically known that these adaptations often reduce virus virulence in natural host animals (15). Previous studies indicated that MV can adapt to grow in some cultured cells by acquiring specific substitutions in the receptor-binding hemagglutinin (H) and/or M proteins (43). Although some mechanisms by which these changes in the H and M proteins may cause attenuation have been proposed (30, 33), their contributions in vivo remain to be determined in animal models. Consistent with the present data, our previous study using a wild-type MV strain and its Vero cell-adapted strain suggested that substitutions introduced into the polymerase protein genes (L and P) during passages in Vero cells caused MV attenuation by reducing the transcriptional activities of viral polymerase (35). However, we observed attenuated gene expression levels by the Edmonston strain only in virus infection analyses, not in minigenome assays. Bankamp et al. (2) also reported that the polymerase proteins of MV vaccine strains show higher transcriptional activities than those of wild-type MV strains when analyzed by minigenome assays. The detailed mechanisms of the attenuated gene expression induced by the polymerase protein genes of the Edmonston strain remain to be elucidated.
In conclusion, the present study demonstrates that the polymerase protein genes of the Edmonston strain contribute to its attenuated phenotype. Our data further show that assays using infectious recombinant viruses are crucial for understanding the contribution of each viral gene to virus replication and virulence and that the SLAM-knock-in mouse is a useful animal model for elucidating the attenuation mechanisms of MV vaccines.
This work was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology and the Ministry of Health, Labor and Welfare of Japan.
Published ahead of print on 17 September 2008. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»