Natural Hosts and Genetic Diversity of the Emerging Tomato Leaf Curl New Delhi Virus in Spain

Abstract

Knowledge about the host range and genetic structure of emerging plant viruses provides insights into fundamental ecological and evolutionary processes, and from an applied perspective, facilitates the design and implementation of sustainable disease control measures. Tomato leaf curl New Delhi virus (ToLCNDV) is an emerging whitefly transmitted begomovirus that is rapidly spreading and inciting economically important diseases in cucurbit crops of the Mediterranean basin. Genetic characterization of the ToLCNDV Mediterranean populations has shown that they are monophyletic in cucurbit plants. However, the extent to which other alternative (cultivated and wild) hosts may affect ToLCNDV genetic population structure and virus prevalence remains unknown. In this study a total of 683 samples from 13 cultivated species, and 203 samples from 24 wild species from three major cucurbit-producing areas of Spain (Murcia, Alicante and Castilla-La Mancha) from five cropping seasons (2012-2016) were analyzed for ToLCNDV infection. Except for watermelon, ToLCNDV was detected in all cultivated-cucurbit species as well as in tomato. Among weeds, Ecballium elaterium, Datura stramonium, Sonchus oleraceus, and Solanum nigrum were identified as alternative ToLCNDV plant hosts, which could act as new potential sources of virus inoculum. Furthermore, we performed full-genome deep-sequencing of 80 ToLCNDV isolates from different hosts, location and cropping year. Our phylogenetic analysis supports a Mediterranean virus population that is genetically very homogeneous, with no clustering pattern, and clearly different from Asian virus populations. Additionally, D. stramonium displayed higher levels of within-host genetic diversity than cultivated plants, and this variability appeared to increase with time. These results suggest that the potential ToLCNDV adaptive evolution occurring in wild plant hosts could serve as a source of virus genetic variability, thereby affecting the genetic structure and spatial-temporal dynamics of the viral population.

Keywords: ToLCNDV; begomovirus; genetic diversity; host range; molecular epidemiology.

FIGURE 1

Plant images showing virus-like symptoms of tomato leaf curl New Delhi virus (ToLCNDV) infection in cultivated and wild plant species. The top row shows symptomatic cultivated Zuchinni (A), Cucumber (B), and Melon (C) plants, where common symptoms include curling, yellowing and severe mottle mosaic in young leaves with plants that appear stunted in growth. The bottom row shows symptomatic wild Ecballium elaterium (D), Datura stramonium (E), and Solanum nigrum (F) plants, comprising curling leaves, chlorotic mottling and mosaic in young leaves, including yellow spots and pale green tissues.

FIGURE 2

Phylogenetic trees of the 80 ToLCNDV consensus nucleotide sequences. The phylogenetic analysis was inferred by Maximum Likelihood, using the full-length genome A (left) and B (right) of the ToLCNDV isolates of this study, including a collection of 50 full-genome ToLCNDV sequences retrieved from NCBI-GenBank (referenced according to their accession number), with also full-length ToLCNDV sequences from Spain available to date. Each ToLCNDV isolate from this study was referenced according to its location (MU, Murcia; AL, Alicante; CLM, Castilla-La Mancha), year (2012–2016), host (ZU, Zucchini; ME, Melon; PU, Pumpkin; CU, Cucumber; DA, D. stramonium), as well as the plot and sample number. The bootstrap values corresponded to 1,000 replications, and the scales provide genetic distances.

FIGURE 3

Inference of the heterogeneity within-population of ToLCNDV averaged by the mutation frequency (mutations based on nucleotide substitutions and indels per nucleotide) that was estimated for each viral genomic component A (left, white-bars) and B (right, black-bars), grouping ToLCNDV populations by collection year (A,B; 2012 to 2016), and also by infected host (C,D; Cucumber, Melon, Pumpkin, Zucchini, and D. stramonium). Bars represent the mean value with the 95% CI error bar.

FIGURE 4

Principal Component Analysis (PCAs) of both genomes for the ToLCNDV isolates. All isolates were clustered based on the correlation matrix of the genetic variables after normalization data. ToLCNDV isolates are colored according to the cluster, and each isolate is named after its location (MU, Murcia; AL, Alicante; CLM, Castilla-La Mancha), year (2012–2016), and host (ZU, Zucchini; ME, Melon; PU, Pumpkin; CU, Cucumber; DA, D. stramonium). Supplementary Tables S3 and S4 show the significances and variable responsible for the group separation. In genome A (left), the six groups consisted of isolates that originated from the same year, whereas the three groups that originated in genome B (right) were fitted according to time and host.