Deep-learning-based removal of autofluorescence and fluorescence quantification in plant-colonizing bacteria in vivo.

Fluorescence microscopy is common in bacteria-plant interaction studies. However, strong autofluorescence from plant tissues impedes in vivo studies on endophytes tagged with fluorescent proteins. To solve this problem, we developed a deep-learning-based approach to eliminate plant autofluorescence from fluorescence microscopy images, tested for the model endophyte Azoarcus olearius BH72 colonizing Oryza sativa roots. Micrographs from three channels (tdTomato for gene expression, green fluorescent protein (GFP) and AutoFluorescence (AF)) were processed by a neural network based approach, generating images that simulate the background autofluorescence in the tdTomato channel. After subtracting the model-generated signals from each pixel in the genuine channel, the autofluorescence in the tdTomato channel was greatly reduced or even removed. The deep-learning-based approach can be applied for fluorescence detection and quantification, exemplified by a weakly expressed, a cell-density modulated and a nitrogen-fixation gene in A. olearius. A transcriptional nifH::tdTomato fusion demonstrated stronger induction of nif genes inside roots than outside, suggesting extension of the rhizosphere effect for diazotrophs into the endorhizosphere. The pre-trained convolutional neural network model is easily applied to process other images of the same plant tissues with the same settings. This study showed the high potential of deep-learning-based approaches in image processing. With proper training data and strategies, autofluorescence in other tissues or materials can be removed for broad applications.

Global expression analysis of the response to microaerobiosis reveals an important cue for endophytic establishment of Azoarcus sp. BH72.

The endophyte Azoarcus sp. BH72, fixing nitrogen microaerobically, encounters low O2 tensions in flooded roots. Therefore, its transcriptome upon shift to microaerobiosis was analyzed using oligonucleotide microarrays. A total of 8.7% of the protein-coding genes were significantly modulated. Aerobic conditions induced expression of genes involved in oxidative stress protection, while under microaerobiosis, 233 genes were upregulated, encoding hypothetical proteins, transcriptional regulators, and proteins involved in energy metabolism, among them a cbb3 -type terminal oxidase contributing to but not essential for N2 fixation. A newly established sensitive transcriptional reporter system using tdTomato allowed to visualize even relatively low bacterial gene expression in association with roots. Beyond metabolic changes, low oxygen concentrations seemed to prime transcription for plant colonization: Several genes known to be required for endophytic rice interaction were induced, and novel bacterial colonization factors were identified, such as azo1653. The cargo of the type V autotransporter Azo1653 had similarities to the attachment factor pertactin. Although for short term swarming-dependent colonization, it conferred a competitive disadvantage, it contributed to endophytic long-term establishment inside roots. Proteins sharing such opposing roles in the colonization process appear to occur more generally, as we demonstrated a very similar phenotype for another attachment protein, Azo1684. This suggests distinct cellular strategies for endophyte establishment.