1 Introduction
Biological invasions represent major threats to biodiversity as invasive species can replace native species. Evolutionary and ecological studies are of major importance for understanding the patterns and causes of biological invasions and plant and animal invasions have been extensively studied since the 19th century (Burns, Murphy, & Zheng, 2018; Cadotte, Campbell, Li, Sodhi, & Mandrak, 2018; MacDougall, Gilbert, & Levine, 2009). Evolutionary and ecological approaches aim at elucidating the dynamics of invasions as well as the advantages of invaders that allow them to outcompete resident species (Facon et al., 2006; Gladieux et al., 2015). Although microbial invasions can represent important threats to ecosystems and human societies, they have been much less studied than invasive plants and animals (Gladieux et al., 2015; Kinnunen et al ., 2016; Litchman, 2010; Mallon, Elsas, & Salles, 2015; Perkins, Leger, & Nowak, 2011).
In this work, we used evolutionary and ecological approaches for understanding the patterns and causes of the establishment of the bacterial pathogen Dickeya solani in potato agrosystems and the maintenance of the resident, related competitor D. dianthicola . There are two major clades of potato bacterial pathogens in Europe,Pectobacterium and Dickeya, two sister genera among the order of Enterobacterales (Gamma-proteobacteria). While somePectobacterium species had settled since before the 1970s in potato agrosystems in Europe (Pérombelon, 2002), Dickeya species invaded later, in two successive waves. Mostly isolated from ornamental plants such as Dianthus, Dahlia and Begonia , D. dianthicola has been reported in Europe as early as the 1950s. Isolation of D. dianthicola from Solanum tuberosum dates back to the 70’s (Toth et al., 2011; Parkinson et al., 2009). Nowadays,D. dianthicola is considered an endemic pathogen of potato plants in Europe. It is present in all continents, including North-America, where it was associated with disease outbreaks in potato farms in 2015 (Oulghazi et al., 2017; Patel, Baldwin, Patel, Kobayashi, & Wyenandt, 2019; Sarfraz et al., 2018; Toth et al. , 2011; Wright et al., 2018).
Dickeya solani has emerged in potato production systems in Europe much more recently, in the early 2000s (Toth et al., 2011; van der Wolf et al., 2014). Aside from potatoes, D. solani has been isolated from ornamental plants such as Hyacinthus , Iris andMuscari , but the primary host(s) of that species remain(s) uncertain (Chen, Zhang, & Chen, 2015; Slawiak et al., 2009; van der Wolf et al., 2014). Population genomics revealed little genetic variation in D. solani , in agreement with a recent spread of this species following a bottleneck during invasion (Golanowska et al., 2018; Khayi et al., 2015). An analysis of 20 D. solani genomes identified a non-synonymous single nucleotide polymorphism (SNP) in each of the two fliC and fliN genes that are involved in bacterial motility (Khayi et al., 2015). Horizontal gene transfer events from D. dianthicola to D. solani have been documented, adding or replacing genomic fragments (Khayi et al., 2015).
The D. solani invader, as well as the D. dianthicola andPectobacterium spp. residents, are necrotrophic pathogens causing similar symptoms, resulting in blackleg disease on stems and soft-rot disease of tubers (Charkowski 2018; Toth et al., 2011). All these bacteria secrete plant cell-wall macerating enzymes (including the pectate lyases) and proliferate by assimilating the plant cell remains, including iron by producing the chrysobactin and achromobactin siderophores. Pectate lyases, i.e., PelA, PelD and especially PelE, play a key role in the initiation of the plant cell wall maceration. Using pectin as a substrate, these enzymes release oligosaccharides that elicit expression of other plant-cell wall macerating enzymes in a positive feedback loop (Duprey, Nasser, Léonard, Brochier-Armanet, & Reverchon, 2016). Noticeably, the pelA gene is truncated, hence non-functional, in D. dianthicola (Duprey et al., 2016; Raoul des Essarts et al., 2019). The secretion of these virulence factors is a costly process that is tightly regulated, being induced by the perception of chemicals that inform the pathogen on plant host physiology (mainly plant sugars) and on the size of its own population (quorum-sensing signals) (Leonard, Hommais, Nasser, & Reverchon, 2017).Dickeya pathogens synthetize and sense two types of quorum-sensing signals, the N-acylhomoserine lactones and Vfm compounds, which activate virulence gene expression when a threshold concentration of signals (reflecting the number of cells in the environment) is reached (Crépin et al., 2012; Nasser et al., 2013; Potrykus, Golanowska, Hugouvieux-Cotte-Pattat, & Lojkowska, 2014; Potrykus, , Hugouvieux-Cotte-Pattat, & Lojkowska, 2018). By modulating the synthesis of virulence factors, the quorum-sensing signals participate in the transition from an oligotroph lifestyle in soils and surface waters, where Dickeya and Pectobacterium bacteria are rare, to a copiotroph lifestyle on a plant host, where they increase in number by several orders of magnitude (Laurila et al., 2008; Potrykus et al., 2016). Functional differences such as metabolic capacities have been identified between D. solani and D. dianthicola by comparative genomics, transcriptomics and biochemical approaches (Bellieny-Rabelo et al., 2019; Raoul des Essarts et al., 2019), but fitness differences within host plants remain only partially investigated (Czajkowski et al., 2013; Czajkowski, de Boer, Velvis, & van der Wolf, 2010; Raoul des Essarts et al., 2019; Shyntum et al., 2019). Whether the potato host could be differentially exploited byD. dianthicola and D. solani and whether the two pathogens stably coexist by using different ecological niches or compete directly when exploiting this host remain unsolved questions, despite their importance for understanding the D. solani invasion.
Here, we combined epidemiological records in potato fields, experimental inoculations and population genomics to contribute to our understanding of the patterns and ecological determinants of D. solaniinvasion. Epidemiological records over one decade in France allowed to compare the relative abundances of the D. solani invader, theD. dianthicola resident and Pectobacterium spp . Using experimental populations and plant assays in greenhouse, with a diversity of strains and of ecological conditions, we then focused onD. solani and D. dianthicola to determine the ecological traits that could facilitate or constrain D. solaniestablishment, including competition or ecological niche differences between the two Dickeya species. Finally, we obtained the genomes of 56 D. solani isolates that we analyzed together with 20 genomes previously published (Khayi et al., 2015), revealing novel variations in D. solani population. We focused on the quorum-sensing gene vfmB to investigate whether its variation, which has not been studied so far, could be associated to invasion traits in D. solani .