Gauging genetic diversity of generalists: A test of genetic and ecological generalism with RNA virus experimental evolution

Abstract

Generalist viruses, those with a comparatively larger host range, are considered more likely to emerge on new hosts. The potential to emerge in new hosts has been linked to viral genetic diversity, a measure of evolvability. However, there is no consensus on whether infecting a larger number of hosts leads to higher genetic diversity, or whether diversity is better maintained in a homogeneous environment, similar to the lifestyle of a specialist virus. Using experimental evolution with the RNA bacteriophage phi6, we directly tested whether genetic generalism (carrying an expanded host range mutation) or environmental generalism (growing on heterogeneous hosts) leads to viral populations with more genetic variation. Sixteen evolved viral lineages were deep sequenced to provide genetic evidence for population diversity. When evolved on a single host, specialist and generalist genotypes both maintained the same level of diversity (measured by the number of single nucleotide polymorphisms (SNPs) above 1%, P = 0.81). However, the generalist genotype evolved on a single host had higher SNP levels than generalist lineages under two heterogeneous host passaging schemes (P = 0.001, P < 0.001). RNA viruses' response to selection in alternating hosts reduces standing genetic diversity compared to those evolving in a single host to which the virus is already well-adapted.

Keywords: dsRNA virus; generalist; genetic diversity; phage evolution.

Figure 1.

Average frequency of host A mutations among lineages. S/P in blue, G/P in beige, G/PT in orange, G/PE in maroon. Data collected every 10-days during the 30-day passaging, points are average values of four lineages, error bars are standard deviations.

Figure 2.

Heat shock tolerance mutation frequency of all lineages. At the top, heat shock tolerance of four S/P lineages (blue); below is the heat shock tolerance of four G/P lineages (beige), four G/PT lineages (orange), four G/PE lineages (maroon). Data collected every 10 days during the 30-day passaging. The horizontal gray line is the heat shock survival of the S ancestor (top) and G ancestor (bottom).

Figure 3.

Number of SNPs above 1 per cent across all lineages of all passaging schemes. Each data point is labeled with the raw SNP count, and the position of the data point is normalized among samples sharing the same MiSeq sequencing run. This means that some raw SNP count numbers will appear lower on the y-axis than a smaller raw SNP count number from a different MiSeq run (and vice versa). The four colors indicate four different sequencing runs—the first replicate of each passaging scheme was run in one MiSeq flow cell (colored red), as were the second replicates (colored purple), the third replicates (colored black), and the fourth replicates (colored gray).

Figure 4.

Top ten SNPs from each Day 30 population, four replicate populations for each treatment. Each square is a SNP placed according to its genomic position on the segmented phi6 genome. Color indicates frequency of the SNP. Vertical dashed lines show parallel SNPs by connecting the squares in the identical location from two populations. From left to right, with uppercase letters for amino acids within the indicated protein, and lowercase letters for non-coding region nucleotides, the lines are at P5 G101C on the Small segment; a830g, t869c, P6 S166G, P3 Q130R, P13 T65A on the Medium segment. All other specific SNPs shown are listed in Supplementary Files S1–S28.

Figure 5.

Relative fitness of populations on P. Both lysates grown from the Day 0 ancestors are squares (specialist S is black, generalist G is gray). The Day 29 populations are diamonds and Day 30 populations are circles. S/P populations are blue, G/P populations are beige, G/PT populations are orange, G/PE populations are maroon. Values are means of six replicates (with the exception of G/PT lineage one at Day 30 and G/PE lineage four at Day 30 having five replicates), error bars are standard deviations.

Figure 6.

Relative fitness of populations on T and E. The lysates grown from the Day 0 generalist G ancestor is shown as the gray square, which serves as the origin for the axes (the scale of each axis is relative fitness compared to the appropriate common competitor). The Day 29 populations are diamonds and Day 30 populations are circles. G/P populations are beige, G/PT populations are orange, G/PE populations are maroon. Values are means of six replicates, error bars are standard deviations.