A) Microbe-microbe and host-microbe communication in host tissues (Ariane Kemen, Daniel Gómez)
We have previously shown for obligate biotroph fungi (that are fungi depending on a living plant host) to secrete amyloid proteins capable of fibril formation in the interaction zone between host and microbe (Kemen et al. 2013) where they are likely mediating the interaction. Expanding our analyses to plant colonizing protists, particularly the genus Albugo, a causal agent of white rust on Arabidopsis thaliana, we have identified numerous amyloid-like proteins secreted into the apoplast. A previous study we had conducted on field grown plants combined with laboratory reconstitution experiments revealed Albugo sp. to be capable of structuring microbial communities through reducing diversity and benefiting certain bacterial species (Agler et al. 2016).
This was particularly imminent to endophytic bacterial communities where diversity was negatively correlated with Albugo sp. abundance (Agler et al. 2016). To identify specificity of different Albugo sp. varying in their amyloid repertoire and relative secretion, we initialized a high throughput screen of plant associated epiphytic and endophytic bacteria versus apoplastic fluid of plants colonized by two different Albugo species (Ruhe, unpublished). We could identify a broad range of sensitivity by different bacteria to the individual Albugo sp. . Particularly epiphytic bacteria that we exclusively identified on leaf surfaces were strongly supressed by plant apoplastic fluid from A. candida colonized plants, while some endophytic species showed even enhanced growth.
In general, numerous peptides and proteins have been found to possess an inherent tendency to convert from native functional states into amyloid aggregates. This phenomenon has bene intensely studied in medical research as it is associated with a range of common human disorders, including Alzheimer and Parkinson diseases (Chiti & Dobson 2017). Substantial evidence implicates oligomeric intermediates of amyloids as toxic species in the pathogenesis of these chronic disease states (Kagan et al. 2012). Such oligomers are not only toxic to their native organism but have been found to possess antimicrobial activity. Besides protegrin-1, which kills microbes by a channel forming mechanism (Kagan et al. 2012), a recent study demonstrates that amyloid‐β (Aβ) can function as an antimicrobial peptide, and additional data show that bacteria and yeast can seed Aβ deposition into amyloid (Golde 2016). Seeding amyloid formation is a general mechanism used by numerous microbes for their advantage. Recent insights into bacterial biofilm matrix structures, for example, have induced a paradigm shift toward the recognition of amyloid fibres as common building block structures that confer stability to the exopolysaccharide matrix (Taglialegna, Lasa & Valle 2016). Amyloid fibres and therefore oligomeric intermediates have been shown to be present from prokaryotes to eukaryotes and are besides biofilm formation integral to many biological functions such as cell-cell contacts, scaffolding and substrate adhesion (Garcia-Sherman et al. 2014). In fact, microbial amyloids are not only structural components but modulate host-microbe- (Gilchrist et al. 2012; Bois et al. 2013) and likely microbe-microbe-interactions. Many microbes utilize functional amyloids to specifically attach to and colonize the host (Garcia-Sherman et al. 2014).
In summary, current data reveals a bivalent character of amyloid forming proteins as beneficial and as toxic compounds in microbe-microbe-interaction and raises two fundamental questions: (1) How do microbes in complex microbial communities protect themselves from various toxic amyloids they and others produce and (2) how and where is amyloid fibril formation seeded in microbe-microbe and host-microbe interactions?
B) The genetic effect of host genes on hub microbes (Ariane Kemen)
Previous studies from our lab and others revealed complex and highly connected microbial networks that structure the leave microbiome (Agler et al. 2016). These networks are centred on microbial hubs that are postulated to act as mediators between the host plant and the microbiome. We could show in field experiments that both abiotic factors and plant genotype influence microbiome composition. Presence of a gene conferring resistance to the hub species Albugo, the causal agent of white rust on Brassicacea, has been shown to significantly impact microbiome structure. We therefore hypothesize, hub species of the genus Albugo transmit host gene phenotypes to the whole microbial community via interfering with microbe-microbe interactions in the network and therefore impacting host fitness.
How different resistance mechanisms against hub species like Albugo influence the structure of the microbial network and therefore microbiome composition is yet unstudied. To address this question we quantify epi- and endophytes in different susceptible and resistant Arabidopsis ecotypes under controlled conditions. In parallel we use synthetic microbial communities in gnotobiotic systems to test our hypothesis and reconstruct microbial networks with reduced complexity. Our synthetic communities span two kingdoms (bacteria and eukaryotes) inspired by data from natural sites and enable us to study cross-kingdom interactions between host and microbes and microbes and microbes. We use quantitative PCR to follow dynamics of community shifts in the presence and absence of hub species and relevant host genes. Under such controlled conditions we could identify a significant genotype effect on microbiome composition mediated via hub microbes. We will use this system to dissect mechanisms and evolutionary aspects of host-microbe and microbe-microbe interactions in dynamic host associated mixed microbial communities. Data will be relevant to build synthetic communities that can benefit plants and therefore human nutrition under natural conditions as up to now biocontrol agents based on living microbes are rarely conferring durable resistances.