Screening and discovery of symbiotic and antagonistic microbial networks using microwell recovery arrays

Abstract

Understanding the dynamic interactions among soil and plant rhizosphere microbiomes is critical for predicting community function and developing improved probiotic and biocontrol agents for plant growth, biofuel production, and human health. However, uncovering these interactions is a grand challenge in microbiology due to the lack of experimental tools suitable for discovery and characterization. This research develops high throughput microwell recovery arrays (MRA) combined with advanced bioinformatics techniques to screen and detect microbial interactions across soil/root microbial communities to uncover bacteria species with an important function. The first part of this thesis describes developing a novel, light-responsive, step-polymerized poly(ethylene glycol) hydrogel membrane to retrieve cells from MRAs with a high degree of spatial control. The utility of microwell arrays, particularly in screening applications, could be significantly expanded if cells of interest could be removed from individual wells for subsequent genetic and phenotypic characterizations. The photodegradability of the membrane permits exchange of nutrients and waste products and seals motile bacteria within microwells and enables individual wells of interest to be opened using a patterned UV light for selective release and retrieval. The second part of the thesis demonstrates the unique application of the MRA platform to discover multi-membered consortia that generate emergent outcomes. The platform was initially developed to discover dual-species co-culture and interactions between two well-characterized interaction pairs, Agrobacterium tumefaciens and Pseudomonas aeruginosa. After investigating the on-chip co-culture using this pair, Populus trichocarpa rhizosphere microbiome was screened for strains affecting the growth of Pantoea sp. YR343, an indole-3-acetic acid (IAA) producing, plant growth-promoting bacteria isolated from Populus deltoides rhizosphere. The third chapter of the thesis uses this approach for enhancing the survival and colonization of commercial nitrogen-fixing, plant growth-promoting bacteria, Azospirillum brasilense, into maize roots to improve crop yield. Diazotrophs such as Azospirillum brasilense function as biofertilizers by colonizing plant roots and enhancing plant productivity through symbiotic interactions within the rhizosphere. Using the MRAs, new isolates showing that promote A. brasilense growth were extracted and identified by 16S sequencing as Serratia mercescens, Serratia nematodiphila, Serratia urelytica, Pantoea agglomerans, Enterobacter tabaci, and Acinetobacter bereziniae, and the interactions were validated off-chip in 96 well plate reader. Also, the growth enhancement and the improvement of the survival and colonization of A. brasilense in Zea mays roots were validated in plant growth chamber experiments, demonstrating the potential to apply the interactions found in vitro towards in vivo systems of agricultural relevance. In the final chapter of the thesis, the screening capabilities using MRAs were further extended towards screening non-pathogenic Agrobacterium isolates for the growth inhibition of pathogenic A. tumefaciens, which is a key plant biotechnology tool and also the causative agent of Crown Gall disease. MRAs were used to combine fluorescently labeled A. tumefaciens sp.15955 with non-pathogenic Agrobacterium isolates collected from native plant roots at the Konza Prairie Biological Station (Manhattan, KS) to uncover several candidates for inhibiting A.tumefaciens sp. 15955 growth. The discovery of such growth-inhibiting isolates will help improve plant productivity by using them as reliable biocontrol agents that prevent Crown Gall disease, and further demonstrates the unique capability of the MRA platform to screen natural isolate collections to discover bacteria capable of inhibiting pathogens.