Synchrony and maturation in loop-design grown microbial metacommunities

By using a looped mass transfer design which has 5 local communities (L1-L5) and regional pool (R), this study focuses on three phases: The insular phase (IS, in which no culture is recycled from the regional pool to the local communities), the RC10 phase (in which 10% of the culture is recycled from the regional pool to the local communities) and the RC50 phase (in which 50% of the culture is recycled from the regional pool to the local communities). Three rounds were established to explore how different combinations of insular and recycling phases contribute to maintaining stable community structures. Round 1 included the IS A, RC10 , RC50 and IS B phases, while Round 2 included the IS I and RC50 I, RC50 II , RC50 III phases. Round 3 immediately followed Round 2 and included the IS II, RC10 , RC50 and IS III. Each bioreactor was sampled, and each sample from the 3 rounds was processed for cytometric analysis to assess microbial community composition and dynamics. A total of 1,226 samples were analyzed, including 365 from a previously published data set (round 1), and 861 samples, which were collected from the bioreactors in this study in round 2 (513 samples) and round 3 (348 samples). Cells from Round 1, 2 and 3 were fixed with paraformaldehyde (PFA) and stored in 70% ethanol at −20 °C. with paraformaldehyde (PFA), stored at -20℃, A defined number of cells was and stained with 4',6-diamidino-2-phenylindole (DAPI) for subsequent measurement using a MoFlo Legacy Cell Sorter (Beckman Coulter, Brea, CA, USA). Cytometric data were collected as two-dimensional (2D) dot plots of DAPI fluorescence versus forward scatter, providing information on DNA content and cell size. After removing the noise and calibration beads by establishing the cell gate (SI 3.4), the cell clusters, identified as Gaussian distributions, were algorithmically defined as gates (G) using the multivariate Gaussian mixture models of FlowEMMi v2 (Bruckmann et al., 2022). A unified gate template containing 48 gates was generated based on emerging and subsequently implemented clusters. The gate template was applied to all samples, yielding 54,288 gates that stand for and were used to identify and track subcommunities consistently. Cell count per gate Absolute and relative cell numbers per gate over time were then calculated using the corresponding analytical functions within the FlowEMMi v2 module.