A number of bacterial and fungal plant pathogens have evolved to inhabit plant xylem vessels. For example, species of the Gram-negative bacterium Ralstonia and fungi Verticillium and Fusarium cause what are commonly referred to as wilt diseases. Xylella fastidiosa (Xf), a xylem-limited Gram-negative bacterium, causes diseases such as coffee leaf scorch, citrus variegated chlorosis, almond leaf scorch, and Pierce's disease (PD) in grapevine (Hopkins and Purcell, 2002). For all of these organisms pathogenesis is, or was, thought to result from their accumulation within xylem vessels leading to vascular occlusion and water deficit. In all cases, this hypothesized mechanism of pathogenesis has been the subject of controversy, usually coming into conflict with the hypothesis that disease symptoms result instead from ‘phytotoxins’ acting directly on plant tissue or eliciting a plant response. In reference to Fusarium, Beckman (1987) referred to this as ‘the great debate’.
During Xf pathogenesis, vascular occlusion is primarily attributed to Xf bacteria and their associated gums (reviewed in Hopkins, 1989; Purcell and Hopkins, 1996). In general, the vascular occlusion hypothesis predicts a positive correlation between symptom severity and pathogen concentrations. Earlier studies correlated high virulence of Xf strains, or high susceptibility of host plants, with high Xf populations at the point of inoculation (107–108 CFU g–1) and Xf movement away from the point of inoculation (Hopkins, 1985b; Fry and Milholland, 1990; Hill and Purcell, 1995). In a comparative study of grape hybrids, Krivanek and Walker (2005) correlated Xf concentration with susceptibility, as predicted. However, Xf lives as a harmless endophyte in most plant species, and many resistant hosts (including species of grapevine) harbour high concentrations and support systemic movement of Xf (Baumgartner and Warren, 2005; Wistrom and Purcell, 2005). Thus, although in susceptible cultivars of grapevine Xf infection results in a suite of symptoms that include leaf scorch and senescence, petiole ‘matchsticks’, incomplete periderm development (green islands), and eventually vine death (Stevenson et al., 2005), no symptoms occur in the presence of high concentrations of Xf in some plant hosts. Together these studies demonstrate that high populations of Xf, along with movement, are not sufficient for pathogenesis in many plant species. A similar conclusion has been reached for other vascular pathogens (Grimault and Prior, 1993; McGarvey et al., 1999).
The vascular occlusion hypothesis of pathogenesis depends on the development of water deficits and, despite their central role in the Xf literature, plant water status has seldom been measured. Inoculation with Xf in both grapevine (Goodwin et al., 1988b) and Parthenocissus quinquefolia (McElrone et al., 2003) resulted in leaf-scorch symptoms that were correlated with reductions in hydraulic conductance, stomatal conductance, and leaf water potential. Thorne et al. (2006) did not observe reduced leaf water potential in infected and symptomatic grapevines, although stomatal closure may have obscured differences in plant water status. More importantly, that study clearly demonstrated that there are qualitative and quantitative differences between the visual symptoms resulting from experimentally imposed water deficits and Xf inoculation. Although water deficits are clearly a component of wilt diseases, their roles in other vascular diseases such as citrus variegated chlorosis and PD, where wilting is not generally observed, are less clear. Furthermore, the presence of low leaf water potentials is not always sufficient to conclude the cause of the low water status. Biotic stressors in general induce premature leaf senescence (reviewed in Guo and Gan, 2005), a process during which there are also decreases in leaf hydraulic conductance and leaf water potential (reviewed in Sack and Holbrook, 2006). Thus, it is possible that changes in the water relations of symptomatic leaves do not result from vascular occlusion by the pathogen per se.
Our knowledge of the nature of the plant–pathogen interaction and the plant immune system has evolved, and it is now understood that plants recognize and respond to pathogens in a variety of ways (reviewed in Dangl and Jones, 2001; Jones and Dangl, 2006). Many molecules that were described previously as ‘phytotoxins’ are now recognized to bind specific plant receptors, inducing the defence response (Deboer et al., 1989; Meyer and Dubery, 1993). This knowledge has guided the research of some xylem pathogens to a new nuanced understanding of the mechanisms of pathogenesis. For example, in Verticillium, numerous small molecules have been identified and isolated that contribute to pathogenicity and evoke many of the observed wilt symptoms (Nachmias, 1985; Meyer and Dubery, 1993; Davis et al., 1998; Chu et al., 1999; Wang et al., 2004). In the case of Ralstonia, yet to be characterized elicitors certainly play a role in pathogenesis (Pfund et al., 2004).
Resolving this debate is important because Xf pathogenesis has become synonymous with vessel occlusion in the literature about PD (e.g. Hopkins, 1989; Fry and Milholland, 1990; Newman et al., 2003; Krivanek and Walker, 2005) without there being a single report that demonstrates a strong and positive correlation between symptoms and concentration of bacteria. Understanding the nature of the pathogenesis clearly has important implications in designing screens for resistance or other approaches to ameliorating the effects of the disease. This is illustrated in Verticillium and Ralstonia where resistance is strongly correlated with the absence of disease symptoms, but not with reduced accumulation or movement of the pathogen (Mace et al., 1981; Grimault and Prior, 1993). The studies of Xf-resistant hosts discussed above suggest that the same situation may be present for Xf, and a recent study provides some evidence of an alternative mode of Xf pathogenicity involving toxins (Reddy et al., 2007). In this study, the aim was to determine the relationship between Xf concentration and the development of leaf-scorch symptoms during PD. To this end, a novel, robust quantitative PCR (qPCR) assay was developed to quantify Xf in planta. Xf populations were quantified and related to symptom development in both artificially inoculated, greenhouse-grown and naturally inoculated, field-grown Chardonnay (a PD-susceptible variety) grapevines. This study provides the first detailed look at Xf populations across leaf lamina and among leaves exhibiting various severities of symptoms, and demonstrates that there is little or no correlation between localized Xf concentrations and symptom development.