Phenotyping of growth rate and copper tolerance
Culture density was routinely measured using in vivo chlorophylla (chl a ) fluorescence (referred to as RFU [relative fluorescence units] from here on) using a VarioscanTM Flash Multimode Reader (ThermoScientific). During re-inoculation, individual culture densities were normalized and reduced to low enough densities to maintain exponential growth until the subsequent dilution, as described in (Andersson et al., 2020). In tandem with seeding the artificial evolution experiment (described below), growth curves (N=4) were collected on individual strains in one mL media on 24-well plates (Polystyrene, FalconTM) for four days with daily RFU measurements midway into the diel light cycle. Microscopic cell counts and dimension measurements of all strains (N=58) were collected to convert strain-specific RFU measurements to cell density. Three hundred cells of each strain were counted in Sedgewick-Rafter cell counting chamber (Wildlife Supply CompanyVR, U.S.A.) using an inverted microscope (Axiovert 135, Zeiss), at ×200 magnification. To compute the strain-specific cellular surface-to-volume of strains (Hillebrand et al., 1999), the dimensions of 20 cells per strain were measured at ×400 magnification.
In conjunction with the start of assays, signs of disruption of photosynthetic activity was monitored in all strains using the maximum quantum yield of photosynthesis (FV/FM), measured using pulse amplitude modulated chl a fluorescence (PAM) on a Phyto-PAM (Schreiber, 1998). Cultures were dark-adapted for 10 min to relax any quenching of chl a fluorescence and establish minimal fluorescence level (F0). After 10 min in darkness, a saturating pulse (200 ms of 10,000 µmol photons m-2 s-1) was used to elicit FM. The maximum quantum yield of photosynthesis (FV/FM) was calculated as (FM – F0)/FM using the 630 nm excitation channel.
Acute toxic 72-hrs dose-response curves were used to measure individual strains tolerance to copper. The method was based on the OECD standard (OECD, 2006), and adapted for S. marinoi as described in Andersson et al. (2020), with modification to the range of copper concentrations to improve the resolution. Specifically, this study used twelve 1.08-fold dilution steps, covering the range between 5.1 and 13 µM CuSO4, with triplicate replicates around the 5% (7.6 µM), 50% (8.9 µM), and 95% (10 µM) growth inhibitory concentrations established for S. marinoi strain RO5AC in pilot experiments. Throughout the experiments we employ strain RO5AC as a mono-clonal reference control to control for experimental artifacts, development of chronic toxic stress and phenotypic plasticity. The copper-amended medium was mixed from 100 µM CuSO4 stocks ~3 hrs before adding cells. Each strain was treated on a single 24-well plate with the 18 copper treatments, five controls with standard f/2 media (containing f/2 amended levels of 0.04 µM Cu), and an internal plate control containing strain RO5AC in standard f/2. Each replicate was inoculated at 0.003 RFU (~900 cells mL-1). After 72-hrs exposure, the RFU was measured, and the specific growth rate (day-1) was calculated as follows:
Eq. 1: µ =\(\frac{\text{LN}\left(\text{RFU}_{72h}\right)\ -\ LN\left(\text{RFU}_{0h}\right)}{3}\)
For the dose-response curves, inhibition of growth rate was computed as the growth rate in a copper treatment divided by the average growth rate across the five control replicates. Some strains exhibited strong inhibition of growth in the controls relative to the low copper exposure, which generated normalization artifacts. To mitigate this,t -tests were performed on all dose-response curves, and whenever the control replicates had a significantly (p <0.05) lower growth rate than the three lowest copper treatment concentrations, normalization was done using the average of these values instead of the control. From the dose-response curves, inhibitory concentrations (EC) of 5, 50, and 95 percent were calculated using the drc-package version 3.0-1 (Ritz et al., 2015) in R (R Core Team 2018) with the Weibull type II function [W2.2]. The copper response range was computed as the EC95/EC05. The absolute concentration of copper in the treatment was measured (ICP-OES) several times during the experiments, including before and after cells had grown in the media for three days. Across all samples (N=8), the nominal concentration (Cun) correlated well (R2=0.992) with the observed concentration (Cua), and all concentrations presented have been corrected to the absolute µM values using the relationship Cua=Cun × 0.703 – 0.142.