Characterization of viruses associated with Grapevine Leafroll disease in Vitis rootstocks in South Africa

Abstract

One of the global and most economically significant viral disease of grape cultivation is grapevine-leafroll disease (GLD). Viruses associated with GLD are known as grapevine leafroll associated viruses (GLRaVs) and numbered in order of discovery, with the most important one being GLRaV-3. Viti- and Foveaviruses are often also often found in mixed infections with GLD, and speculation exists regarding a codependence of transmission of some Vitiviruses with GLRaV-3. As with most grapevine growing countries, many strains of GLRaVs, Viti- and Foveavirus are present in South Africa. A local certification scheme exists with the objective of providing grapevine scion and rootstock material free of these viruses to farmers. Rootstocks show no GLD symptoms thus making visual diagnostics impossible. Furthermore, it is generally accepted that GLRaV-3 is difficult to test for in rootstocks than in scions, possibly due to uneven distribution and low viral titre in rootstocks. It is however also possible that rootstocks select for specific strains of GLRaV-3 which may be less amenable to the detection technique. The difficulty in detection of GLRaV-3 could result in unwitting infection of healthy scion material by infected rootstock material during grafting. To gain greater insight into the detection of GLRaV-3 detection in rootstocks both rootstock and scion tissue were sampled separately from 95 grapevine individuals and tested using a broad spectrum PCR protocol directed at a highly conserved primer binding sequence flanking a highly variable region of the GLRaV-3 helicase gene. Amplicons of scion and rootstocks of 22 vines were subjected to Illumina next generation sequencing to determine the composition of the GLRaV-3 variant population. Poor detection (43% of samples) at low amplicon levels were obtained for GLRaV-3 in primarily R99 rootstock compared to the corresponding scions (GLRaV-3 detected in 93% of the scions samples) and corroborated the perceived poor detection of this virus in at least R99. We also determined that poor detection of GLRaV-3 is not due to the presence of a PCRinhibitory substance in rootstocks as we obtained high levels of amplicons of Vitiand Foveaviruses associated with the rootstocks of GLD infected vines. This was achieved using universal degenerate primer nested RT-PCR combined with Illumina next generation sequencing of the resulting amplicons. Confirmation tests were performed using Vitivirus specific primers to determine the identity of the unidentified Viti- and Foveaviruses found. Direct Sanger sequencing was used on 37 rootstock samples and 20 corresponding scion samples. Viti- and Foveaviruses were significantly less detected (61%) in rootstocks than corresponding scions (82%). The dominant component of the viral population was shown to differ between samples, with Grapevine virus B (GVB) and Grapevine rupestris stem pitting associated virus (GRSPaV) only found dominant in rootstocks. NGS data exhibited that the largest amount of the total reads belonged to GVB (44%), though no pattern could be differentiated between differing rootstocks. This is the most comprehensive study done on the viruses associated with GLD in rootstocks in South Africa, and will contribute to improve the certification scheme.