Grid partitioning image analysis for bacterial cell aggregates

Abstract

Bacterial cell aggregation plays a fundamental role in surface colonization, stress tolerance, and interspecies metabolite exchange; yet quantitative analysis remains challenging. Here, we introduce grid partitioning image analysis (GPIA), a simple workflow that quantifies the compositional heterogeneity of bacterial aggregates. Confocal laser scanning microscopy (CLSM) images of fluorescently labeled Acinetobacter sp. Tol 5, which exhibits a self-aggregative nature through its cell surface protein AtaA, were partitioned into 2-{micro}m square grids. Grids containing one or no cells were classified as dispersed, whereas those containing multiple cells were classified as aggregates, and the proportion of EGFP-labeled cells within each grid was recorded. Reference images representing dispersed cells, homo-aggregates, and hetero-aggregates produced characteristic EGFP-ratio histograms that matched binomial predictions. When AtaA production in one cell type was decreased, the histogram changed from a symmetric unimodal histogram with the peak at 40-60 % EGFP-ratio to a skewed distribution, indicating that GPIA can detect differences in cell-to-cell affinity. Using the same procedure, we examined six in-frame deletion variants of AtaA. The deletion of the N-terminal head domain alone prevented co-aggregation with full-length AtaA, suggesting that homophilic recognition by this domain mediated self-aggregation, whereas deletions in all other regions had no measurable effect. GPIA, therefore, offers a simple and rapid approach for quantitative studies on bacterial cell aggregation, bridging the gap between qualitative microscopy and quantitative but technically demanding single-cell analysis. GPIA will accelerate research on cell-cell interactions, which are the foundational processes that drive biofilm formation and the assembly of microbial consortia.