In plants, the xylem tissue functions to transport water and minerals unidirectionally, from roots towards the aboveground parts ^[1][2][1,2]. At the same time, the phloem serves as a highway for the bidirectional distribution of sugars, nutrients, and hormones, in a source-to-sink manner. Apparently, the microaerophilic environment of the phloem is thought to provide a suitable niche for plant-infecting viruses. In the past decades, considerable knowledge of the phloem structure and its role in virus colonization has been gained. It is widely accepted that plant viruses utilize the phloem as the main route for the long-distance spread within a host, along with photoassimilate translocation, and, thus, the virus–phloem interplay has been a hot area in plant virus research ^[3][4][5][3,4,5]. However, virus–xylem interactions have been largely overlooked. On the one hand, functional xylem cells are mainly dead cells, which lack the cellular machinery that the virus needs for the synthesis of its proteins, replication, and assembly of new particles. On the other hand, the movement of a virus from an adjacent cell containing a protoplast to the apoplastic compartment of the xylem requires passing across an intact plasma membrane and a thickened secondary cell wall. With that, the xylem seems to be a challenging territory for the plant viruses and an unlikely area to serve for virus occupation. Interestingly, a growing number of studies, especially those in more recent years, point to the importance of the relationship between the vascular xylem in the plant host–virus interactions and disease production ^[6][7][8][6,7,8]. Viruses do find ways out to blaze trails in the xylem, as they constantly evolve and adapt to environmental challenges. Some studies even suggest that, for specific viruses, active virus accumulation takes place in immature xylem-associated cells, rather than in phloem-associated cells. For instance, virions of Lettuce necrotic yellows virus (LNYV) were not observed in the thin sections of the phloem cells of tobacco (Nicotiana glutinosa L.). Instead, the virus particles were successfully visualized in juvenile xylem cells [9]. With evidence generated from these studies, it seems that at least some plant viruses can manipulate xylem cells to replicate and spread to distal plant parts.

Research on the virus-xylem interplay can be traced back to the 1930s when Johnson revealed the presence of Tobacco mosaic virus (TMV) in the xylem guttation from tomato (Solanum lycopersicum L.) [10]. Since guttation fluid is the exudation of xylem sap, finding virus particles in this exudate demonstrated that infectious TMV could spread via the xylem conduit. Later, the presence of several plant viruses was observed in the xylem of many herbaceous plant hosts. Virions of Beet necrotic yellow vein virus (BNYVV) were frequently found in the undifferentiated and mature xylem vessel elements and xylem parenchyma of sugar beet (Beta vulgaris L.) [11]. The infectious virions of Tomato mosaic virus (ToMV) and Pepper mild mottle virus (PMMV) were recovered and visualized from the guttation fluid of tomato (Lycopersicon esculentum Mill.) and green pepper (Capsicum annuum L.), respectively [12]. With such evidence on hand, French and colleagues tested the potential capability of a diverse group of viruses to move in the xylem sap. Strikingly, the respective results showed that plant viruses from at least ten genera could employ the xylem tissue as a conduit for their transportation [13]. Compared to those in the herbaceous hosts, studies on virus–xylem interactions in perennial woody plants are limited. The reason could be due largely to the fact that fewer viruses have been reported to infect the xylem of a tree. Some time ago, Blueberry shoestring virus (BSV) was shown to achieve the long-distance movement through the xylem, in addition to the phloem of a low woody shrub, highbush blueberry (Vaccinium corymbosum L.) [14]. Furthermore, a recent study on Citrus tristeza virus (CTV) reignited an interest in exploring virus–xylem interactions in perennial woody plants by demonstrating that CTV can invade and further develop a ‘clustered oasis’ of infected cells in the xylem tissue of Citrus macrophylla Wester [7]. It is possible that hijacking the xylem cells could be a strategy adapted by plant viruses to promote infection in a broad range of plant host.

The healthy and functional xylem of a plant is important to its vigorous growth. The invasion of certain vascular pathogens was shown to damage the xylem function and trigger disease syndromes. For instance, a xylem-residing Verticillium dahliae Klebahn, a filamentous fungus, could secret a toxic protein named xylanase 4 to degrade the xylem’s cell walls and induce necrosis. The destruction of the xylem vessels results in wilt syndrome in the infected cotton plants [15]. In contrast, the connection between the virus invasion of the xylem and disease development has not been clearly demonstrated, largely because the vascular xylem is not the main or the sole tissue that the virus colonizes. Other tissues, such as the epidermis, mesophyll, cortex, and phloem, provide more suitable niches for plant viruses, and the phenotypes induced by the virus in those tissues are usually more easily noticeable. For instance, while Brome mosaic virus (BMV) is able to invade xylem cells in oat (Arena sativa L.) and barley (Hordeum vulgare L.), leaf mosaic symptoms are attributed to the virus accumulation in the juvenile phloem and mesophyll cells [16]. In contrast, there is an exception where CTV triggers an obvious phenotype in the vascular tissues of several different citrus varieties, including the xylem. Remarkably, the invasion of CTV into the immature xylem treachery elements and ray parenchyma cells in the citrus plants was shown to account for the development of one of the major disease syndromes induced by this virus. Such invasion interrupts the differentiation of the xylem and phloem cells in the vasculature and eventually triggers the stem pitting disease, which is accompanied by the loss of tree vigor and production of unmarketable fruit. The manifestation of this syndrome in the citrus trees may emerge months or years after CTV inoculation ^[7][17][7,17]. Such phenotype in the CTV-colonized citrus stems and branches could be contingent upon the continuous growth of the xylem rings and the virus xylem invasion. Consequently, the severity of disease is enhanced with tree growth. It is possible that some xylem-invading viruses share similar mechanisms to trigger diseases in other hosts.

2. Examples of Xylem-Invading Viruses

To date, at least 39 plant viruses have been reportedly found in the xylem-associated cells of many different hosts, including herbaceous and perennial woody plants (Table 1). Interestingly, the majority of xylem-invading viruses (35/39) studied to date belong to the group of positive-sense single-stranded RNA (+ssRNA) viruses in the following families: Alphaflexiviridae, Benyviridae, Bromoviridae, Closteroviridae, Potyviridae, Secoviridae, Sobemoviridae, Tombusviridae, and Virgaviridae (Table 1) ^{[6][7][8][10][11][12][13][14][16][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36]}[6,7,8,10,11,12,13,14,16,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36]. Other viruses are three single-stranded DNA (ssDNA) viruses in the family of Geminiviridae [37] and a negative-sense RNA virus in the family Rhabdoviridae (Table 1) [9]. With such a broad range of virus families, the morphology of xylem-invading viruses is quite diverse. Virions produced by those viruses could be long flexuous or short rigid rod-like particles or have icosahedral, twinned-icosahedral, or bullet-like shape and range from approximately 10 nm to over 2000 nm (Table 2). The diversity of these viruses suggests that the xylem exploitation appears to be a widely adapted strategy for plant viruses, which might be an evolutionary consequence of the virus niche expansion.