Are lichens potential natural reservoirs for plant pathogens?

Most ecological studies on plant‐pathogenic bacteria have focused, perhaps understandably, on the agricultural environment. Interest is increasing, however, in the occurrence of plant‐pathogenic bacteria in habitats outside of agriculture. The seminal work by Morris et al. (2008), for instance, has shown that Pseudomonas syringae occupies a wide range of niches linked with the water cycle, including alpine lakes, streams and snow. Moreover, it is becoming clear that traits that are linked to adaptation to biotic and abiotic stress in the non‐agricultural environment can have a secondary function as virulence factors in plants (Morris et al., 2009). Indeed, adaptation to non‐host environments has been suggested to have played a non‐trivial role in the evolution of P. syringae phytopathogenicity and to have affected its epidemic potential (Bartoli et al., 2015).

A reservoir sensu Haydon et al. is defined as ‘one or more epidemiologically connected populations or environments in which the pathogen can be permanently maintained and from which infection is transmitted to the defined target population’ (Haydon et al., 2002). Although, in some cases, plant pathogens are known to survive and persist in plant debris left on the soil surface, in and on seeds, in soil, in association with perennial hosts, in water, on or inside insects, or even on inorganic objects, such as machinery and packaging material, for the most part, the natural reservoirs of most plant pathogens remain unidentified.

Lichens are symbiotic associations of fungi, algae and bacteria that are present in most environments, often quite prominently. Many lichen species are found in close association with plants, growing, for example, on tree trunks or on grassland or heath scrub soils, albeit rarely with an apparent detrimental effect on plant health. Nevertheless, recent metagenomic, amplicon‐based and culture‐based investigations have indicated that some plant pathogens, or their very close relatives, are present in lichen‐associated microbiomes, in some cases in significant numbers (Cardinale et al., 2008; Sigurbjörnsdóttir et al., 2014). Should we perhaps regard these unassuming and generally welcome members of the environmental vegetation as potential reservoirs for established or emerging plant diseases?

According to the traditional textbook definition, lichens are bipartite symbiotic associations of a mycobiont (usually an Ascomycete or a Basidiomycete) and a photobiont (a green alga or a cyanobacterium). The mycobiont provides the structural bulk of the lichen thallus, whereas the photobiont subsists in colonies or layers sheltered beneath fungal peripheral layers. In many species, either or both partners produce secondary metabolites with antibacterial activity, thus providing a somewhat selective environment in which complex bacterial communities thrive, some in biofilm‐like surface communities (Grube et al., 2015) and others endothallically, below the surface of extracellular polysaccharides (Cardinale et al., 2008; Grube et al., 2009). Thus, lichens are hosts to large populations of bacteria whose identity and role in the lichen symbiotic association have been gradually emerging in recent years (Hodkinson and Lutzoni, 2009; Sigurbjörnsdóttir et al., 2015). Although there is considerable inter‐species and inter‐study variability, it has become clear that most lichens harbour a sizeable, complex, Proteobacteria‐dominated microbiota, with dominant families typically comprising the Rhizobiaceae, Methylobacteraceae, Sphingomonadaceae, Acetobacteraceae, Rhodospirillaceae, Comamonadaceae and Burkholderiaceae (Aschenbrenner et al., 2014; Sigurbjörnsdóttir et al., 2015). Typically, taxa more commonly associated with phytopathogens, such as members of the Xanthomonadaceae, Pseudomonadaceae and Enterobacteriaceae, are also present, albeit in lower numbers. Although many of the bacteria present in lichen thalli have doubtless never been cultured, members of many or even most of the dominant lichen‐associated taxa have been isolated and grown in pure culture (Cardinale et al., 2006; Sigurbjörnsdóttir et al., 2014). Among the reported bacterial isolates from lichens, phylogenetic analysis has revealed several close relatives of known plant pathogens, including P. syringae (Liba et al., 2006) and Burkholderia glathei (Cardinale et al., 2006). Plant pathogenicity has, however, not been established, or indeed experimentally addressed, for these or any other lichen‐associated isolates to our knowledge. Among the bacteria indicated by metagenomic or amplicon‐based analysis to be present in lichen‐associated microbiomes that have not been cultured to our knowledge are a few plant pathogens. For example, both functional genes and 16S rDNA from Xanthomonas and its close relative Xylella were observed in the membranous dog lichen (Peltigera membranacea) metagenome (Sigurbjörnsdóttir et al., 2015), but have not, as yet, been cultured from that species. The finding of Xanthomonas‐like organisms in lichens is interesting because Xanthomonas spp. are known to survive poorly when unprotected by host tissues. Further culturing efforts are thus called for.

Our recent analysis of a partial shotgun metagenome of the microbiome associated with Pe. membranacea (Sigurbjörnsdóttir et al., 2015) yielded a number of multiple hits on several genes involved in lichen secondary metabolite resistance, inorganic phosphate mobilization, biopolymer degradation and several other potentially important functions in thallus colonization and symbiosis, as well as several genes likely to play a role in plant pathogenicity. Among the plant virulence genes found in the metagenome are several homologues of type III and type IV secretion systems from diverse bacteria, including xanthomonads, burkholderiae, sphingomonads and acidobacters. However, it should be borne in mind that some genes identified by blast searches as virulence factors might just as well be involved in mutualistic interactions. As so much of the previous work has been performed on pathogenic relations, the blast results are expected to be biased and to underestimate positive interactions. Indeed, bacteria closely related to known plant growth‐promoting bacteria have been observed in a number of studies on lichen‐associated bacteria (Cernava et al., 2015; Hodkinson and Lutzoni, 2009; Sigurbjörnsdóttir et al., 2015). Detailed data mining of completed and ongoing shotgun metagenome sequencing data is likely to yield a more complete picture of the plant‐pathogenic potential of members of the lichen‐associated microbiome. A curated lichen‐associated metagenome database would facilitate such work.

Given the size and complexity of both the endothallic and epithallic bacterial communities, as well as the restrictive environment, one might surmise a considerable scope for the evolutionary acquisition of plant pathogenicity traits to be present. Adaptation to the fungal/algal environment is likely to select for traits useful for plant cohabitation and, indeed, the production of phytohormones and volatile organics has been observed in several lineages of lichen‐associated bacteria (Aschenbrenner et al., 2014; Cernava et al., 2015). Further, the high abundance of associated bacteria yields a significant gene pool that may be available for lateral genetic exchange, possibly facilitating virulence gene acquisition. The lichen‐associated genetic mobilome has, as yet, barely been studied at all. Only a few genomes of lichen‐associated bacteria have thus far been published, all sphingomonads. As in other sphingomonads, lichen‐associated Sphingomonadaceae strains appear to be hosts to several prophages and other mobile elements (Aylward et al., 2013), suggesting high capacity for lateral genetic exchange. To assess the potential of the lichen‐associated microbiome for the interspecific exchange of genetic material, we call for a thorough and systematic push for whole‐genome sequencing of lichen‐associated bacteria, obtaining several genomes from each major lineage of lichen‐associated taxa, as well as a more extensive database of shotgun metagenomic data.

The potential for microbial population exchange between lichens and plants should also not be discounted offhand. Both arboricolous and several terricolous lichens are found in quite intimate proximity to plants, and vertical transmission of bacterial populations on lichen propagules has been observed for the arboricolous lung lichen, Lobaria pulmonaria (Aschenbrenner et al., 2014). Whether and, if so, to what extent interkingdom exchange of bacterial symbionts occurs between lichens and plants has not been investigated to our knowledge.

Although the idea of lichens as potential plant pathogen reservoirs is, in many ways, attractive, we should hasten to point out that there are striking dissimilarities between plants and lichens as microbial habitats, calling for some notable differences in selection pressures and adaptation mechanisms of plant‐associated versus lichen‐associated bacteria. For instance, temporal variation in humidity can be expected to be much more dramatic in the lichen environment, with periodic droughts occurring frequently within the lichen thallus. This environmental variability calls for an environmental adaptability not necessarily seen in plant pathogens, which normally are r‐strategic (that is, they rely on fast growth and rapid substrate exploitation for selective advantage). Survival experiments of plant pathogens in lichen thalli, such as the spraying and subsequent monitoring of green fluorescent protein (GFP)‐tagged plant pathogens onto lichens in a field setting, are therefore needed to assess the survivability of plant pathogens in the lichen environment. Conversely, lichen‐associated bacteria closely related to known plant pathogens may, of course, be adapted lichen symbiovars with no plant‐pathogenic traits at all, and thus need to be tested for their growth in plant models.

In summary, we feel that lichens ought to be considered as potential non‐host reservoirs for phytopathogenic bacteria, including P. syringae, B. glathei and xanthomonads, which have been repeatedly observed through molecular or, in some cases, culture‐based observations to be present in significant numbers in or on lichen thalli. Much remains to be done, however, before the role of lichens in the ecology and evolution of these pathogens can be reliably elucidated. Among the tasks to be tackled are the systematic data mining of extant and emerging metagenomes for genes encoding virulence factors, the identification and quantification of prophages and other mobile genetic elements within lichen‐associated microbiomes, the assessment of population exchange potential between lichens and plants, and the rigorous characterization of potentially plant‐pathogenic isolates.