Single-Cell Analysis of RNA Virus Infection Identifies Multiple Genetically Diverse Viral Genomes within Single Infectious Units

Abstract

Genetic diversity enables a virus to colonize novel hosts, evade immunity, and evolve drug resistance. However, viral diversity is typically assessed at the population level. Given the existence of cell-to-cell variation, it is critical to understand viral genetic structure at the single-cell level. By combining single-cell isolation with ultra-deep sequencing, we characterized the genetic structure and diversity of a RNA virus shortly after single-cell bottlenecks. Full-length sequences from 881 viral plaques derived from 90 individual cells reveal that sequence variants pre-existing in different viral genomes can be co-transmitted within the same infectious unit to individual cells. Further, the rate of spontaneous virus mutation varies across individual cells, and early production of diversity depends on the viral yield of the very first infected cell. These results unravel genetic and structural features of a virus at the single-cell level, with implications for viral diversity and evolution.

Graphical abstract

Figure 1

Experimental Setup for Sequencing Single-Cell Bottlenecked Viruses BHK cells were inoculated with approximately 3 × 10⁴ PFUs of a VSV stock at an MOI of 0.3 PFU/cell, and individual cells were transferred to separate culture wells with a micromanipulator. After overnight completion of the viral infection cycle, supernatants were plated in solidified medium and single, isolated plaques (viral progeny) were picked and used for SOLiD massive parallel sequencing. The viral stock was subject to Illumina ultra-deep sequencing to detect polymorphisms present in the inoculum (parental sequence variants). Polymorphisms present in plaque sequences but not detected in the inoculum constitute likely spontaneous mutations arising after the single-cell bottleneck.

Figure 2

Parental Genetic Variants Delivered to Individual Cells (A) Parental sequence variants found after each single-cell bottleneck. Large gray circles depict individual cells labeled with numbers, and smaller circles correspond to the 7–10 plaques sequenced for each cell. The 36 parental variants are shown with colored dots inside each plaque. Colors are used for distinguishing different variants found among plaques derived from the same initial cell. Use of the same color in different cells is not meant to indicate a common sequence variant. (B) Distribution of the number of parental sequence variants found in the 7–10 plaques derived from single cells. (C) Distribution of the number of plaques derived from the same cell that contained a given parental variant. The list of parental sequence variants found in the inoculum is provided in Table S1 and the list of parental variants found in plaques derived from single cells is provided in Table S2.

Figure 3

Co-transmission of Sequence Variants BHK cells were co-infected (total MOI = 20 PFU/cell) with the wild-type (WT) and a MAR variant. After completion of the infection cycle, progeny viruses were used for plaque purification, and 20 plaques were picked. Each of these plaques was then assayed for its ability to escape antibody neutralization by plaque assays in the presence or absence of a monoclonal antibody. Red histogram bars show the relative proportion of MAR viruses within each plaque as determined by the neutralization assay. Plaques 1, 5, 6, 12, 18, and 20 showed a mix of MAR and WT phenotypes. Five plaques with fully WT phenotype (2, 3, 4, 9, and 10; marked with an asterisk) were further analyzed by RT, PCR, molecular cloning and Sanger sequencing of 15–19 clones each. Plaques 2 and 3 showed a mix of the WT sequence and MAR-conferring mutations, despite their fully WT phenotype. Plaques 1 and 18 were subsequently used for infecting fresh BHK cells at 0.02 PFU/cell and, after the entire culture was infected, 19–20 plaques were isolated and used to repeat the neutralization assay. Mixing of MAR and WT viruses was still evident after this additional transfer.

Figure 4

Genetic Diversity Found after Single-Cell Bottlenecks (A) Non-parental SNPs found after single-cell bottlenecks. Large gray circles depict individual cells indicated with numbers and smaller circles correspond to the 7–10 plaques derived from each cell. The 496 SNPs are shown with colored dots inside each plaque. Colors are used for distinguishing different SNPs found among plaques derived from the same cell. Use of the same color for different cells is not meant to indicate a common sequence variant. (B) Distribution of the number of non-parental SNPs found in the 7–10 plaques derived from each cell. (C) Distribution of the number of plaques derived from the same cell that contained a given non-parental variant. (D) Spectrum of nucleotide substitutions found after single-cell bottlenecks. (E) Correlation between the abundance of each type of substitution in single-cell-derived plaques and natural isolates. For natural isolates, 1,033 SNPs were extracted from the following available genome sequences of the Indiana VSV serotype: EU849003 (Mudd-Summers strain, used as reference), AF473865 (isolated from cattle in Colombia, 1985), AF473866 (isolated from cattle in Guatemala, 1994), AF473864 (isolated from horse in Colorado, 1998), EF197793 (unknown host), J02428 (unknown host), and NC_001560 (unknown host). The dots indicate the percentage of total SNPs of each type (A→G, U→C, and so on), in log scale. The list of non-parental SNPs is provided in Table S3.

Figure 5

Relationship between the Viral Yield per Cell and the Genetic Diversity Produced Immediately after Single-Cell Bottlenecks (A) Distribution of the viral yield among the 90 cells. (B) Correlation between the per-cell viral yield and the number of non-parental sequence variants found among the 7–10 plaques derived from each cell. The red dashed line indicates the linear regression line. The indicated p value rejects the null hypothesis of a null slope. (C) Box plot of the number of new variants found among the 7–10 plaques derived from each cell as a function of the per-cell viral yield with a 1,000 PFU/cell cutoff. Lines within each box mark the median, the boundaries of the box indicate the 25^th and 75^th percentiles, whiskers indicate the 10^th and 90^th percentiles, and dots show individual outlying points.