Evaporation-induced hydrodynamics promote conjugation-mediated plasmid transfer in microbial populations

Conjugative plasmids bestow important traits to microbial communities, such as virulence, antibiotic resistance, pollutant biotransformation, and biotechnology-relevant functions. While the biological mechanisms and determinants of plasmid conjugation are well established, the underlying physical and ecological driving forces remain unclear. Microbial communities often inhabit unsaturated environments, such as soils and host surfaces (e.g., skin, teeth, leaves, roots), where water evaporation and associated small-scale hydrodynamic processes frequently occur at numerous air-water and solid-water interfaces. Here, we hypothesized that evaporation can induce water flows with profound effects on the spatial distribution and surface deposition of cells, and consequently on the extent of plasmid conjugation. Using droplet experiments with an antibiotic resistance-encoding plasmid, we show that evaporation-induced water flows reduce cell-cell distances and significantly increase the extent of plasmid conjugation. Counterintuitively, we found that evaporation results in lower expression levels of conjugation-related genes. This negative relationship between the extent of plasmid conjugation and the expression of conjugation-related genes could be attributed to increased conjugation efficiency during evaporation. This study provides new insights into the physical and ecological determinants of plasmid conjugation, with important implications for understanding the spread and proliferation of plasmid-encoded traits.


Supplementary Materials and Methods
Bacterial strains and growth conditions. We used Escherichia coli HB101 carrying plasmid RP4 as the donor strain for the bacterial mating experiment, where RP4 is an IncP α-type conjugative plasmid carrying ampicillin (Amp), kanamycin (Km), and tetracycline (Tet) resistance genes. We verified the three antibiotic resistance abilities of E. coli HB101 by confirming growth with lysogeny broth (LB; 10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl, pH 7.4) agar plates amended with 50 μg/mL Amp and 50 μg/mL Km. We routinely cultured E.
coli HB101 under the same conditions to maintain RP4 stability. We used E. coli K12 carrying plasmid pNW33n (NCBI: txid226614) as the recipient strain, where pNW33n carries the chloramphenicol (Chl) resistance gene and thus differentiates it from E. coli strain HB101. We verified the chloramphenicol resistance ability of E. coli K12 by confirming growth on LB agar plates amended with 25 μg/mL Chl. We routinely cultured E. coli K12 under the same conditions to maintain pNW33n stability.
Enumeration of transconjugants from colony plating. For the first experiment, we set the initial cell concentrations of the donor (E. coli HB101):recipient (E. coli K12) (1:1) mix to 9.6×10 5 , 9.6×10 4 , 9.6×10 3 and 9.6×10 2 CFU/μL, deposited them onto LB agar plates, and kept the temperature constant at 25°C. After 8 h mating time, we cut the agar containing the cell deposits, transferred them to a 50-ml tube containing 20 mL of PBS solution, and vortexed the tube for approximately 5 min to resuspend the cells. We then quantified E. coli K12 transconjugants for each experimental replicate (n = 3) three times by selective plating of the suspended cells on LB agar plates supplemented with 50 μg/mL Amp, 50 μg/mL Km and 25 μg/mL Chl. For the second experiment, we fixed the initial donor (E. coli HB101):recipient (E. coli K12) (1:1) cell concentration to 10 5 CFU/μL and set the mating temperature to 4, 16, 25, or 37°C. We manipulated evaporation conditions and enumerated transconjugants after an 8 h mating time as described for the first experiment. For the third experiment, we set the donor (E. coli HB101):recipient (E. coli K12) mating time to 0.5, 2, 4, 6, 8, 10, 12, or 24 h with a fixed initial bacterial concentration of 10 5 CFU/μL and a constant mating temperature of 25°C. We again manipulated evaporation conditions and enumerated transconjugants as described for the first and second experiments. For all experiments, we replicated each treatment three times (i.e., three droplets) and quantified transconjugants three times for each droplet. Thus, we have a total of nine datapoints (three droplets x three analytical replicates). We averaged the analytical replicates prior to all statistical analyses to maintain independence. We verified that drying did not impact cell viability at the end of the experiment (Supplementary Table S1).

Microscopy.
We centrifuged the overnight liquid cultures of donor (E. coli HB101) and recipient (E. coli K12) for 10 minutes at 6000 rpm with a high-speed centrifuge (H/T18MM, Herexi, China). We then discarded the supernatants and resuspended the cells in sterile phosphate-buffered saline (PBS). We next repeated the centrifugation and discarded the supernatants to eliminate any possible influence of extracellular secretions and residual culture medium. Finally, we resuspended the cells in sterile water with 1% polyethylene glycol (PEG) solution and adjusted the cell concentrations to obtain an optical density at 600 nm (OD600) of 1 (≈10 8 CFU/mL) with a UV/VIS double beam spectrophotometer (A590, Aoyi, China). After the droplets evaporated, we imaged the droplets with a Nikon A1RsiHD 25 confocal laser scanning microscope (CLSM) (Nikon, Tokyo, Japan) with an emission wavelength of 488 nm. We used the 10× objective lens to image the entire droplet and the 100× objective lens to track the movement of individual cells under non-evaporative condition.

Quantification of the expression of outer membrane and conjugation transfer-related genes.
For real-time quantitative PCR (qPCR), we cut the agar containing the cell deposits and transferred them to a 50-ml tube containing 20 ml of PBS solution. We then vortexed the tube for approximately 5 min and used tweezers to remove the remaining solid agar. We next collected the cells by low-speed centrifugation and resuspended them in 1 ml of PBS solution for RNA extraction. We extracted total RNA from the donor (E. coli HB101) and recipient (E.

Statistical analyses.
We performed all statistical analyses between the means of analytical replicates for each experimental replicate to maintain independence We performed all statistical analyses using core functions in R 3 and generated plots using the package ggplot2 4 .
For all datasets, we performed Analysis of Variance (ANOVA) with the number of transconjugants (CFU/μL) as a continuous response variable and evaporation conditions and initial cell density (first experiment), evaporation conditions and temperature (second experiment), and evaporation conditions and time point (third experiment) as categorical response variables. We performed repeated measures ANOVA for the latter to account for the repeated enumeration of transconjugants at different points in time. We tested for statistically significant differences between factor levels using the Welch two-sample twosided t-test, which is suitable for comparisons between means of factor levels with unequal variances. We used a confidence threshold of 95% and adjusted the P-values for multiple comparisons using the Holm-Bonferroni method. We tested for the influence of evaporation conditions on the expression of HGT-related genes using one-way ANOVA and compared factor level means as described above. Finally, we measured the effect sizes of evaporative conditions (EV) over Marangoni convection (MC) on the total number of transconjugants detected after 8h mating time using the Cohen's d coefficient. 5