Mixed viral infection constrains the genome formula of multipartite cucumber mosaic virus

Abstract

Many plant viruses have a multipartite organization, with multiple genome segments packaged into separate virus particles. The genome formula describes the relative frequencies of all viral genome segments, and previous work suggests rapid changes in these frequencies facilitate virus adaptation. Many studies have reported mixed viral infections in plants, often resulting in strong virus–virus interactions. Here, we tested whether mixed infections with tripartite alfalfa mosaic virus (AMV) and monopartite potato virus Y (PVY) affected the genome formula of the tripartite cucumber mosaic virus (CMV), our experimental model. We found that the CMV titer was reduced in mixed infections with its tripartite Bromoviridae relative AMV and in triple infections with both AMV and PVY, indicating notable virus–virus interactions. The variability of the CMV genome formula was significantly lower in mixed infections (CMV and AMV, CMV and PVY, and CMV and AMV and PVY) than in single infections (CMV only). These observations led to the surprising conclusion that mixed infections with two distinct viruses constrain the CMV genome formula. It remains unclear how common these effects are for different combinations of virus species and strains and what the underlying mechanisms are. We, therefore, extended a simulation model to consider three putative scenarios in which a second virus affected the genome formula. The simulation results also suggested that shifts in the genome formula occur, but may not be widespread due to the required conditions. One scenario modeled—co-infection exclusion through niche differentiation—was congruent with the experimental data, as this scenario led to reductions in genome formula variability and titer of the multipartite virus. Whereas previous studies highlighted host–species effects, our results indicate that the genome formula is also affected by mixed infections, suggesting that there is a broader set of environmental cues that affect the genome formula.

Robust Approaches to the Quantitative Analysis of Genome Formula Variation in Multipartite and Segmented Viruses

“…Given the shortcomings of many methods for analyzing genome formula variation, we recently developed another approach, based on the genome formula distance metric [ 15 ], in combination with permutation-based statistical approaches [ 34 , 35 ]. Here, we build on this previous work by describing this metric in detail and considering some of its attributes, such as the range of values and its interpretation.…”

“…To compare two values of the genome formula, in a previous study, we proposed to consider the Euclidean distance between them [ 15 ]. We refer to this metric as the genome formula distance ( D ), such that for two genome formula observations a and b , the distance between them is as follows: …”

“…First, we compare these theoretical predictions to observed values of genome formula distance ( ). We obtain these observed values by re-analyzing genome formula data reported in three experimental studies in which the genome formula was measured in single leaves or whole plants [ 13 , 14 , 15 ]. For the tripartite RNA viruses AMV and CMV, we find that the observed values for the genome formula distance are below both of our reference values ( Table 3 ), as expected for systems that appear to converge on an equilibrium value.…”

“…For some multipartite and segmented viruses, variation in the frequency of genome segments has been observed [ 8 , 13 , 14 , 15 , 16 ]. The genome formula is the abundance of all virus genome segments, and it is typically described in one of two ways.…”

“…The result we work towards is having a robust, well-documented approach to analyzing genome formula data, which has been applied to various datasets, illustrating its applications and demonstrating its relevance. Here, we propose a simple and robust approach to genome formula analysis that relies on the genome formula distance metric [ 15 ]. We document this method in detail as a resource for the analysis of genome-formula data.…”