Population diversity of cassava mosaic begomoviruses increases over the course of serial vegetative propagation

Abstract

Cassava mosaic disease (CMD) represents a serious threat to cassava, a major root crop for more than 300 million Africans. CMD is caused by single-stranded DNA begomoviruses that evolve rapidly, making it challenging to develop durable disease resistance. In addition to the evolutionary forces of mutation, recombination and reassortment, factors such as climate, agriculture practices and the presence of DNA satellites may impact viral diversity. To gain insight into the factors that alter and shape viral diversity in planta, we used high-throughput sequencing to characterize the accumulation of nucleotide diversity after inoculation of infectious clones corresponding to African cassava mosaic virus (ACMV) and East African cassava mosaic Cameroon virus (EACMCV) in the susceptible cassava landrace Kibandameno. We found that vegetative propagation had a significant effect on viral nucleotide diversity, while temperature and a satellite DNA did not have measurable impacts in our study. EACMCV diversity increased linearly with the number of vegetative propagation passages, while ACMV diversity increased for a time and then decreased in later passages. We observed a substitution bias toward C→T and G→A for mutations in the viral genomes consistent with field isolates. Non-coding regions excluding the promoter regions of genes showed the highest levels of nucleotide diversity for each genome component. Changes in the 5' intergenic region of DNA-A resembled the sequence of the cognate DNA-B sequence. The majority of nucleotide changes in coding regions were non-synonymous, most with predicted deleterious effects on protein structure, indicative of relaxed selection pressure over six vegetative passages. Overall, these results underscore the importance of knowing how cropping practices affect viral evolution and disease progression.

Keywords: cassava mosaic begomoviruses; vegetative propagation; viral diversity.

Fig. 1.

ACMV and EACMCV nucleotide diversity in the Veg2 experiment. (a) Sliding window analysis of nucleotide diversity (π) of ACMV DNA-A and DNA-B and (c) EACMCV DNA-A and DNA-B. Red to green represents the nucleotide diversity across the genome of inoculated plants (p1) and two vegetative propagations (p2 and p3). Enhanced views of the nucleotide diversity of the AV1 and AC1 open reading frames during P1-P3 for ACMV-A (b) and EACMCV-A (d). Blue lines mark the locations of codons encoding functional motifs in the Rep protein, i.e. Rep C, Rep B, Walker B, Walker A [63], Motif 3 [62], GRS [64], Motif 2 [62] and Motif 1 [83]. The motifs are shown to scale. Genome coordinates (nt), the positions of open reading frames and their directions of transcription are shown below each graph.

Fig. 2.

Veg6 ACMV and EACMCV nucleotide diversity sliding windows. (a) Sliding window analysis of nucleotide diversity (π) of ACMV DNA-A and DNA-B and (c) EACMCV DNA-A and DNA-B. Red to pink represents the nucleotide diversity across the genome of inoculated plants (p1) and two vegetative propagations (p4 and p6). Enhanced views of the nucleotide diversity of the AV1 and AC1 open reading frames during P1, P4, and P7 for ACMV-A (b) and EACMCV-A (d). Blue lines mark the locations of codons encoding functional motifs in the Rep protein, i.e. Rep C, Rep B, Walker B, Walker A [63], Motif 3 [62], GRS [64], Motif 2 [62] and Motif 1 [83]. The motifs are shown to scale. Genome coordinates (nt), the positions of open reading frames and their directions of transcription are shown below each graph.

Fig. 3.

Tajima’s D analysis by passage and minimum variant frequency cutoff. Tajima’s D analysis passage (p) for (a) ACMV DNA-A and DNA-B and (b) EACMCV DNA-A and DNA-B in the Veg2 experiment. Tajima’s D by passage for (c) ACMV DNA-A and DNA-B and (d) EACMCV DNA-A and DNA-B in the Veg6 experiment. Dotted red horizontal lines represent thresholds for significant Tajima’s D values [59]. The vertical blue dotted line represents the minimum variant frequency of 3 % used in this study. The lines representing each passage are colour coded, as shown at the bottom.

Fig. 4.

Synonymous and non-synonymous codon changes in ACMV and EACMCV. Diagram of the locations of synonymous codon changes (grey), unassigned changes in overlapping open reading frames (ORFs; green), and non-synonymous changes for (a) ACMV (blue) and (b) EACMCV (red). The number of synonymous (c) and non-synonymous (d) codon changes normalized to the total number of codons in the ORF for ACMV (dark grey and blue) and EACMCV (light grey and red). The number if synonymous (grey) and non-synonymous (ACMV-blue, EACMCV-red) codon changes in overlapping coding sequences in ORFs overlapping within ACMV excluding AC5 (e), EACMCV (f), and regions within ACMV that overlap AC5 (g). The overlapping ORF is designated first and the ORF assessed for the codon change is designated second in bold.

Fig. 5.

ACMV common regions undergo sequence convergence. (a) Linear maps of ACMV DNA-A and ACMV DNA-B. The maps were linearized in the common region at the cleavage site in the top strand of the viral origin of replication. Red arrows mark the 3′-OH and the 5′-P of the nick site [101]. The common region upstream (yellow) and downstream (grey) of the nick site are marked. The open reading frames and directions of transcription are shown by the black arrows below. (b) ACMV DNA-A and ACMV DNA-B sequences showing their common regions in the circularized genomic form. The labelling is the same as in (a), with the nick site indicated by a red arrow and the upstream and downstream sequences marked by yellow and grey shading, respectively. The iterons (boxed) and the hairpin motif (stem: underlined; loop: dotted line) involved for the initiation of viral replication are marked. The TATA box (underlined) for complementary sense transcription is also labelled. SNPs showing convergence of the common region sequences of DNA-A and DNA-B are in black typeface, and other SNPs are in red typeface.

Fig. 6.

EACMCV common regions undergo sequence convergence. (a) Linear maps of EACMCV DNA-A and EACMCV DNA-B. The maps were linearized in the common region at the cleavage site in the top strand of the viral origin of replication. Red arrows mark the 3′-OH and the 5′-P of the nick site [101]. The common region upstream (yellow) and downstream (grey) of the nick site are marked. The open reading frames and directions of transcription are shown by the black arrows below. (b) EACMCV DNA-A and EACMCV DNA-B sequences showing their common regions in the circularized genomic form. The labelling is the same as in (a), with the nick site indicated by a red arrow and the upstream and downstream sequences marked by yellow and grey shading, respectively. The iterons (boxed) and the hairpin motif (stem: underlined; loop: dotted line) involved for the initiation of viral replication are marked. The TATA box (underlined) for complementary sense transcription is also labelled. SNPs showing convergence of the common region sequences of DNA-A and DNA-B are in black typeface, and other SNPs are in red typeface. Insertion is indicated by italics.