Contribution of historical herbarium small RNAs to the reconstruction of a cassava mosaic geminivirus evolutionary history

Abstract

Emerging viral diseases of plants are recognised as a growing threat to global food security. However, little is known about the evolutionary processes and ecological factors underlying the emergence and success of viruses that have caused past epidemics. With technological advances in the field of ancient genomics, it is now possible to sequence historical genomes to provide a better understanding of viral plant disease emergence and pathogen evolutionary history. In this context, herbarium specimens represent a valuable source of dated and preserved material. We report here the first historical genome of a crop pathogen DNA virus, a 90-year-old African cassava mosaic virus (ACMV), reconstructed from small RNA sequences bearing hallmarks of small interfering RNAs. Relative to tip-calibrated dating inferences using only modern data, those performed with the historical genome yielded both molecular evolution rate estimates that were significantly lower, and lineage divergence times that were significantly older. Crucially, divergence times estimated without the historical genome appeared in discordance with both historical disease reports and the existence of the historical genome itself. In conclusion, our study reports an updated time-frame for the history and evolution of ACMV and illustrates how the study of crop viral diseases could benefit from natural history collections.

Figure 1

Leaf of Manihot glaziovii specimen P04808771 collected in Bambari, Central African Republic, in June 1928 and preserved at the Herbarium of the Muséum national d'Histoire naturelle, Paris, France. The original annotation (hand-written in French on bottom left) states ”Leaf from a young diseased plant”. Typical symptoms of cassava mosaic disease such as chlorotic mosaic and deformation of the leaf can be distinguished.

Figure 2

Main characteristics of small RNA (sRNA) isolated from historical specimen P04808771 (Herbarium of the Muséum national d'Histoire naturelle, Paris, France). (a) Size distribution of all, cassava-mapping and ACMV (DNA-A & DNA-B) genome-mapping reads. (b) Proportion of reads mapping to cassava and ACMV reference genomes.

Figure 3

Reconstruction and authentication of historical ACMV genome. (a) Summary of mapping statistics to reference genomes for both ACMV DNA-A and DNA-B molecules. (b) Coverage plots (blue scale). Red arrays indicate regions that are not covered with siRNA reads (depth=0). Inner circle represents the genome and coding regions, as follows: AC1, AC3, AC3, AC4, AV1 and AV2 for DNA-A ; BC1 and BV1 for DNA-B; C: complementary strand; V: viral strand. Red cross symbolizes the geminivirus replication initiation site and grey ticks the SNPs between historical and reference sequences. (c) Post-mortem RNA damage patterns measured on historical (red) and modern ACMV sample isolated in 2017 (green). Straight and dotted lines represent C to U vs all other substitutions of the first 10 nucleotides from the 5’end, respectively.

Figure 4

Bayesian dated tree of 134 sequences of ACMV DNA-A built from 1081 non-recombining SNPs. The historical DNA-A sequence is highlighted in red. Node support values with posterior probabilities above 0.8 are displayed by black diamonds. Node bars cover 95% Highest Probability Density of node height. Tips are colored according to the sample's geographic origin, according to the map on top left. The node corresponding to the common ancestor of all Malagasy isolates is circled in purple. Both tests of temporal signal (top: linear regression of root-to-tip distance on year of sampling date and bottom: date-randomization test) are presented in the dotted box.

Figure 5

Bayesian estimations performed with or without including the historical genome. Substitution rate (a), MRCA of all (b) and from Madagascar (c) isolates, inferred with (red boxplot) and without (orange boxplot) the historical ACMV DNA-A component. ***p<0.001.