Abstract
Tissue-like 3-dimensional (3D) microbial communities called biofilms colonize a wide variety of biotic and abiotic surfaces and, in aggregate, constitute a major component of bacterial biomass on earth. As such, biofilms have a tremendous impact on the biogeochemistry of our planet and the biochemistry of higher living organisms. However, how macroscopic biofilm properties, such as its tolerance up to 1000 times higher concentrations of antibiotic drugs, its mechanical adhesion/cohesion and its biochemical metabolism, emerge from the collective actions of individual bacteria remains unclear.
There are two critical barriers to study single cell behaviors within thick 3D biofilms in a non-invasive manner. First, conventional imaging modalities are not able to non-invasively resolve individual cells within thick 3D biofilms. Second, accurate cell detection and cellular shape measurements in densely packed biofilms are challenging because of the limited resolution and low signal to background ratios (SBRs) in fluorescence microscopy images. The focus of the research described in this dissertation is to solve these problems.
To image bacterial biofilms with cellular/subcellular resolution, we used lattice light-sheet microscopy (LLSM), a new imaging technology that effectively combines low photo-toxicity and high spatiotemporal resolution. To enable growing and imaging biofilms, especially pathogenic species, at high resolution, we designed a flow chamber system that is compatible with LLSM. To accurately segment and classify single bacterial cells in 3D fluorescence images, we developed Bacterial Cell Morphometry 3D (BCM3D), an image analysis workflow that combines deep learning with mathematical image analysis. Compared to state-of-the-art bacterial cell segmentation approaches, BCM3D consistently achieves higher segmentation accuracy and further enables automated morphometric cell classifications in multi-population biofilms. The accurate segmentation results from BCM3D provide precise single-cell observables, including cell positions, orientation, morphologies, volumes and fluorescent intensities. We developed a multi-cell tracking method by utilizing these cell observables to associate the same cells imaged at different time points.
The integrated workflow, namely non-invasive imaging biofilms with subcellular resolution, accurate segmentation and classification of single bacterial cells in 3D fluorescence images and tracking multi-cell in the segmentation results are applied to study the diffusive behavior of individual cells in Shigella flexneri biofilms.
