Grapevine–virus–environment interactions: an intriguing puzzle to solve

Grapevine (Vitis spp.) is one of the major and most economically important fruit crops worldwide. Unlike other crop species, grapes and wine have ancient historical connections with the development of human culture and with the socio-cultural background of grape-growing areas, particularly in the Mediterranean Basin. Grape cultivation has very old origins, with the earliest evidence of wine production dating to 7000–7400 yr bp (before present) and the discovery of seeds of domesticated grapes dating to 8000 yr bp in the Middle East (This et al., 2006). Grapevine potentially can host many viral entities, and the cohabitation between grapevine and viruses likely has lasted for centuries, since the first indirect evidence of viral diseases in grapevine dates back to Roman times (Martelli, 2015; see Maree et al., 2013, for a historical outline of Grapevine Leafroll Disease). Viruses are obligate intracellular parasites that live and multiply in host cells. They possess a small genome with limited coding capacity – often no more than a dozen proteins. In order to complete their life cycle and spread in plant tissues, viruses have evolved a very complex process that involves both virus and host factors (Wang, 2015). Grapevine is also known to host a large number of pathogenic agents, including 65 viruses, five viroids and eight phytoplasmas (Martelli, 2014). The expansion of high-throughput sequencing technologies in recent years has given a great impetus to the discovery of unknown grapevine-infecting viruses (e.g. Zhang et al., 2011; Giampetruzzi et al., 2012; Al Rwahnih et al., 2013). Many reports have described the genetic diversity of virus populations and symptoms linked to virus diseases in grapevine. However, few studies have explored the molecular, metabolic and physiological events resulting from viral infections in grapevines. Only recently has interest in this field emerged, because it is becoming increasingly evident that viruses induce several changes that potentially affect the general responses of the grapevine to its environment, and vice versa. Virus infections are common under field conditions and multiple infections are routinely detected, as suggested, for example, by a study identifying the presence of eight viruses and two viroids in a single vineyard by use of an in silico approach (Jo et al., 2015). Since all studies conducted on grapevine are likely to be performed on plants infected by one or more viruses, an update on grapevine–virus interactions discussed in the light of the emerging phytobiome concept (Dicke, 2016) could be useful for researchers working on grapevine and for plant biology more generally. Although currently 65 different viral species have been identified in grapevine, only about half are recognized as agents or putative agents of ‘major diseases’ in terms of economic relevance (Martelli, 2014). The major diseases infecting grapevine are: the ‘infectious degeneration’ or ‘decline’ caused by Grapevine fanleaf virus (GFLV), Arabis mosaic virus (ArMV) and other genus Nepovirus members; the ‘grapevine leafroll disease’ (GLD), caused by five distinct viruses known as Grapevine leafroll-associated viruses that belong to the genera Ampelovirus, Closterovirus and Velarivirus; and the ‘rugose wood complex’, which consists of four distinct diseases associated with different etiological agents (Martelli, 2014). The impacts of major virus diseases on grapevine include reductions in plant vigour, yield and fruit quality, which lead to large economic losses to the whole grapevine agribusiness (including the wine industry) and justify the recent burst of studies about profit-maximizing disease-control strategies (Fuller et al., 2015; Ricketts et al., 2015). Recently, the economic impact of GLD was estimated at $25 000–$40 000 per hectare during a 25-yr lifespan of a ‘Cabernet franc’ vineyard (Atallah et al., 2012) and c. $47 000 per hectare over 20 yr on ‘Sauvignon blanc’ and ‘Merlot’ vineyards in New Zealand (Nimmo-Bell, 2006). Also in Californian grape-growing areas the economic cost of GLD was estimated to range from $29 902 to $226 405 per hectare depending on percentage of infection and on management strategies (Ricketts et al., 2015). These reports suggest that global financial losses caused by grapevine virus diseases involve millions of dollars annually. The impact of viral diseases is amplified by the fact that grapevine is a vegetatively propagated woody species, and many grapevine cultivars have been distributed as vegetative cuttings worldwide, with little regard for the presence of possible intracellular pathogenic agents (Naidu et al., 2015). Against viruses, only preventative measures can be adopted to limit the infection risks, such as chemical treatments against vectors or the application of sanitation techniques to remove infection (Maliogka et al., 2015). Over the past 20 yr different strategies have been applied to virus elimination in order to produce certified material, free of the most dangerous viral pathogens. The methodologies commonly employed in the sanitation programmes in grapevine, such as in vitro and/or in vivo thermotherapy, meristem culture, chemotherapy and cryotherapy, show differential success depending on the viral species (Maliogka et al., 2015). Some viruses, such as Grapevine rupestris stem pitting-associated virus (GRSPaV), and viroids are particularly recalcitrant to elimination (Gribaudo et al., 2006; Gambino et al., 2011). Only somatic embryogenesis has proven very effective against a large number of viruses and viroids, although this is not a commonly-used technique due to the risk of inducing somaclonal variation (Gambino et al., 2006, 2011). GLRaV-3 is a grapevine virus whose molecular interactions with the host have been extensively studied. GLRaV-3 is one of the most widespread leafroll viruses and is associated with GLD, which is characterized by severe symptoms at the end of the growing season, including leaf deformation, reddish or yellowing of leaves, chlorosis, and detrimental effects on sugar and anthocyanin accumulation in berries and on yield (Fig. 1a) (Guidoni et al., 1997; Martelli, 2014). Phenotypic alterations are presumably linked to source–sink balance alterations of sugar levels, since GLRaV-3 spreads through the phloem, associated, in turn, with molecular events linked to changes in metabolic pathways in infected tissue. This virus induces changes in transcript profiles in a wide range of cellular leaf processes, including functions related to protein turnover, transport (e.g. sugar transporters), metabolism, transcription and defence. Several down-regulated grapevine genes code for proteins associated with photosystems and photosynthetic pigments linked to a decrease in photosynthetic activity (Espinoza et al., 2007b). In addition, grapevine responds to GLRaV-3 infection by activating the senescence programme in the leaf: several virus-induced genes (i.e. senescence-associated genes (SAGs), pathogenesis-related (PR) proteins) are expressed at elevated levels during physiological senescence in healthy grapevines (Espinoza et al., 2007a). Another interesting effect of GLRaV-3 is the differential regulation of the flavonoid pathway in different organs. Anthocyanin biosynthetic genes are up-regulated in GLRaV-3-infected leaves, contributing to the reddish-purple colour of symptomatic grapevine leaves of red/dark berried cultivars (Gutha et al., 2010), whereas they are down-regulated in berries during ripening (Vega et al., 2011) in association with altered expression of genes involved in sugar transport and hormone responses (Fig. 1a). In grapevine, it is known that anthocyanin levels increase concomitantly with sugar accumulation (Deluc et al., 2009). The repression of genes involved in the anthocyanin pathway by GLRaV-3 infection is probably linked to the lower sugar content in infected berries (Vega et al., 2011), while the opposite situation occurs in infected leaves, in which high sugar accumulation (Espinoza et al., 2007b) takes place and induces de novo synthesis of some anthocyanins (Gutha et al., 2010). Although no unequivocal cause/effect explanation can be attributed to molecular changes reported in these works, their potential contribution to the grapevine response to a virus should be carefully considered to elucidate the complex biological mechanisms responsible for symptom development in plants. Viruses are not necessarily detrimental in all situations; indeed, in specific cases they could show a mutualistic relationship depending on their host (Roossinck, 2011) and the environment (Hily et al., 2016). An interesting example of this phenomenon is related to GRSPaV, one of the most prevalent viruses that infects grapevines, usually found in Vitis vinifera cultivars in a latent state, without the development of macroscopic phenotypic alterations (Fig. 1b). In V. rupestris, the symptomatology attributable to GRSPaV infection is associated generally with ‘rupestris stem pitting’, a disorder of the ‘rugose wood complex’ (Martelli, 2014). However, symptom development varies according to GRSPaV variant, Vitis species and probably co-infection with other viral entities that are not always known (Meng et al., 2006). A global gene expression analysis of GRSPaV-infected V. vinifera plants showed some unexpected responses that have not been reported previously for other plant–virus interactions. In field conditions, GRSPaV-infected plants show: a very moderate decrease in physiological efficiency and yield performance associated with an increase in transcripts involved in photosynthesis and CO2 fixation; a reduction in some defence mechanisms; and an overlap with water and salinity stress responses (Fig. 1b) (Gambino et al., 2012). These considerations suggest some types of interaction between GRSPaV and abiotic stress in grapevine. In glasshouse conditions, this virus alters grapevine responses to drought: GRSPaV-infected plants show higher rates of photosynthesis and stomatal conductance; low hydraulic resistance to water transport; and increased ability to extract water from the soil (Pantaleo et al., 2016). In addition, GRSPaV also alters the profiles of several micro RNAs (miRNAs), which in some cases appear to be important players in the association between GRSPaV and drought in grapevine (Fig. 1b). The physiological modifications observed in infected plants could be linked to the specific modulation of some miRNAs and their targets (i.e. miR396, miR164, miR156, miRC121, miRs409712_2) (Pantaleo et al., 2016). Interestingly, another example of a positive interaction between a virus and grapevine has been shown recently. Grapevine infected with GLRaV-3 showed an increase in resistance to the causal agent of grapevine downy mildew (the fungus Plasmopara viticola) (Repetto et al., 2012). Field-grown grapevines exhibited a significant reduction in downy mildew secondary infections in GLRaV-3-infected plants. The authors suggested that extremely complex regulatory mechanisms occurring in virus-infected plants (e.g. alteration of some physiological parameters, accumulation of resveratrol and up-regulation of some defence-related genes) could partially explain the observed lowered susceptibility to P. viticola attack (Repetto et al., 2012). In agreement with the observations that the origins of the major groups of GLRaV-3 may have occurred before worldwide cultivation of grapevine (Maree et al., 2015), we can consider the partnership between grapevine and its viruses to be an ancient cohabitation that, in this case, led to the co-evolution of the responses to the fungal pathogen. These examples represent two ‘peculiarities’ compared to the many works that demonstrate the pathogenicity of viral agents in grapevine. They are very interesting since they illustrate a further level of complexity in the grapevine–virus relationship and bring new findings to better understand these interactions that deserve further attention. These examples suggest that many variables might influence these relationships. For example: (1) virus strains, different isolates can induce different symptoms in grapevine (Maree et al., 2013). (2) Grapevine genotype, genetic and biological differences in the host may determine different reactions in response to the same virus. ‘Cabernet Franc’, ‘Pinot noir’, ‘Cabernet Sauvignon’ and ‘Barbera’ are cultivars very sensitive to GLRaV-3, and may be used as indicator hosts in the biological indexing to screen grapevine material for this virus (Maree et al., 2013). (3) Co-infection with different viruses. Vegetative propagation, use of infected rootstocks and vector transmission can contribute to multiple infection events in vineyards. In these situations, some eventual relationships established among different viruses could result in synergistic or antagonistic (competitive) interactions (Syller & Grupa, 2016). (4) Human selection. The grapevine cultivars currently available are the results of centuries of human selection. During the clonal selection, combinations of specific grapevine genotypes with viral populations, involving particular viral species and particular variants within these species, may have been selected because they conferred particularly interesting traits. Only recently have plant biologists widened their point of view and demonstrated that plants and the phytobiome (or holobiont, composed of the microbiome and the macrobiome) interact with each other and with the environment, modulating plant responses to biotic and abiotic factors (Dessaux et al., 2016; Dicke, 2016). A further level of complexity, the concept of pathogenesis has also evolved in the pathobiome, that takes into account the phytobiome and its pathogenic agents, both of which are integrated within the environment affected by factors such as human activities and pollutants (Vayssier-Taussat et al., 2014). In light of this, and owing to the advance of ‘meta-omics’ techniques, grapevine–virus interactions can be addressed not only from the point of view of a classical binary relationship (host–pathogen), but also by considering the plant and its endophytic microorganisms as a single organism or microecosystem adapting to the external environment. Here, we have reported some important molecular effects induced in grapevine by two widespread viruses with different pathogenicities and economic impacts: GLRaV-3, which causes a disease leading to large economic losses, and GRSPaV, which can be considered latent in V. vinifera. The positive interaction between viruses and abiotic stress has already been reported in herbaceous species infected with several viruses (Xu et al., 2008; Hily et al., 2016). The strains of GRSPaV and GLRaV-3 analysed (Repetto et al., 2012; Pantaleo et al., 2016) may represent other interesting examples of plant viruses with potential mutualistic interactions in specific conditions. It will be necessary to explore further and confirm these findings in vineyards with different cultivars, several rootstock–scion combinations and different virus strains. These findings could provide a new way to study plant–microorganism interactions and also a new awareness to study plant biology, since many viruses are asymptomatic and difficult to identify visually. Because of the availability of the grapevine genome sequence (Jaillon et al., 2007) and new high-throughput ‘-omics’ technologies, grapevine is becoming a model species for fruit trees, as shown by a continuously growing number of publications on all aspects of grapevine biology. The use of these very sensitive technologies emphasizes the need for closer scrutiny of the starting material; for example, the study of the ripening processes in berries can give significantly different results if considering virus-free or viruses-infected grapevines (Vega et al., 2011; Gambino et al., 2012). Moreover, in light of the ongoing climate change phenomenon and the increasing necessity for the adoption of more sustainable viticulture approaches to preserve environmental resources, these emerging studies could be important for directing future viticulture practice. These considerations pertaining to grapevine could be extended to other crop species, such as forest trees and other vegetatively propagated woody species, since vegetative propagation is one of the most efficient ways of virus dissemination. The study of virus–host interactions in several plant species could reveal interesting findings to help to advance our understanding of the adaptability and evolution of plants. The authors thank Marco Martini for Fig. 1. This research was funded by the Fondazione Cassa di Risparmio di Cuneo, Project ‘SAFEGRAPE: sustainable grapevine protection against fungal pathogens’. I.P., W.C., G.G. conceived the idea and wrote the manuscript and P.B. contributed to the writing.