Flow cytometry combined with viSNE for the analysis of microbial biofilms and detection of microplastics

Biofilms serve essential ecosystem functions and are used in different technical applications. Studies from stream ecology and waste-water treatment have shown that biofilm functionality depends to a great extent on community structure. Here we present a fast and easy-to-use method for individual cell-based analysis of stream biofilms, based on stain-free flow cytometry and visualization of the high-dimensional data by viSNE. The method allows the combined assessment of community structure, decay of phototrophic organisms and presence of abiotic particles. In laboratory experiments, it allows quantification of cellular decay and detection of survival of larger cells after temperature stress, while in the field it enables detection of community structure changes that correlate with known environmental drivers (flow conditions, dissolved organic carbon, calcium) and detection of microplastic contamination. The method can potentially be applied to other biofilm types, for example, for inferring community structure for environmental and industrial research and monitoring.


Supplementary Figure 2
Similarity between scattering and fluorescence properties of the single species.

Supplementary Figure 3
Similarity between viSNE submaps of single species. The maximum mean discrepancy (MMD) was calculated for each pair of viSNE submaps (Figure 1a) and is presented as a heatmap going from white (identity) to dark red (maximum discrepancy in this set). For Diatoma, two subcultures of the same species were used in both viSNE and MMD calculations.

Supplementary Figure 4
Tracking temporal shifts of subpopulations in artificial microbial communities. Two (Achnanthidium minutissimum and Nitzschia palea, (A+N)) or four (A, N, Botryococcus braunii and Chamaesiphon polonicus, (A+N+B+C)) single species were mixed (n = 3) and assessed by flow-cytometry immediately (t 0 ) after mixing and after one week (t 1 ). Single species cultures were used as control (n = 3). At t 0 the community was planktonic and composed of the single species, whereas at t 1 species and their subpopulations have segregated into planktonic and bottom fractions.  Figure 3a. The maximum mean discrepancy (MMD) was calculated for each pair of viSNE submaps and is presented as a heatmap going from white (identity) to dark red (maximum discrepancy in this set). The mean MMD between the different time points was 5.18 ± 1.44, which was larger than the MMD between the five biological replicates (3.55 ± 1.23) and that of the three technical replicates per sample (2.20 ± 0.95).

Supplementary Figure 17
Similarity between viSNE submaps presented in Figure 5a. The maximum mean discrepancy (MMD) was calculated for each pair of viSNE submaps and is presented as a heatmap going from white (identity) to dark red (maximum discrepancy in this set). The mean MMD between the different sampling sites was 13.78 ± 4.21, which was larger than the MMD between the three biological replicates (9.73 ± 4.34) and that of the three technical replicates per sample (5.22 ± 3.22). Figure 4b, normalized to total number of particles for each of biological replicate (n=3) from each sampling site A-F.

Supplementary Figure 19
Biplots of the RDA based on the fraction of particles in the subpopulations in Figure 4b constrained by field forward selected physico-chemical parameters (Supplementary Tables 3, 5). Dots/grey tones: specific sampling sites, stars/coloured: centroids of the subpopulations (MA1-MA10). Explained variation for the first two constraint axes is given. The relation (r 2 ) of chloride (Cl), potassium (K), sodium (Na) and silicate (H 4 SiO 4 ) concentrations as well as electrical conductivity (cond) and pH to the subpopulation structure according to general additive models (GAM; as amount of deviance accounted for) and the corresponding significance    Table 3 Presence/absence of organisms identifiable by light microscopy in samples from natural stream microbial communities after temperature increase. d0 d7 d14 d21 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Chamaesiphon polonicus x x Each sample was subsampled three times for microscopy, 300 intact cells per subsample were analysed. Results from subsamples are summed for better readability. x: organism present, 1: single cell found in all three subsamples.

Supplementary Table 4
Coordinates, water temperature (T), flow rate, pH, electrical conductivity (CD), and description of the six sampling sites on Mönchaltorfer Aa, Canton Zurich, Switzerland. Measurements were made a few centimeters above the ground near the surface of the sampled stones. * Waste-water treatment plant (WWTP) Gossau. A B C D E F 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 Phormidium autumnale

Supplementary
x x x Phormidium spec x x x Achnanthidium spec x x x x 1* Each sample was subsampled three times for microscopy, 300 intact cells per subsample were analysed. Results from subsamples were summed for better readability. x: organism present; 1: single cell found in all three subsamples; *Thecamoeba