Species interactions constrain adaptation and preserve ecological stability in an



The S. cerevisiae strain used to find the evolution experiment was a haploid, non-recombining derivative of YJM978 (Mata, ho::HygMX, ura3::KanMX) [37, 38]. The L. plantarum strain used in this study is a novel strain, LPKH, isolated from a sourdough bread culture.

Growth media

Evolution experiments and all growth experiments were carried out in CSM, which is comprised of Complete Supplement Mixture (CSM, Sunrise Science), 20 g/L glucose, 6.7 g/L nitrogen base, and here was supplemented with tryptophan (0.05 mg/mL) and uracil (0.02 mg/mL). Modifications of this media for specific assays are described below.

Growth assays

The performance of ancestral strains was assayed by measuring OD600 and cell density after 48 hours of growth in either fresh growth media, or “spent” growth media. To generate spent medium for growth assays with L. plantarum, single clones of ancestral S. cerevisiae YJM978 were grown to saturation in 3 mL of CSM at 28oC over 48 hours (n = 3), the cells were removed by centrifugation and filtration using a syringe filter with a pore size of 0.2 μm. To generate spent medium for the growth of S. cerevisiae, we cultured clones of ancestral L. plantarum in separate wells in a 96-well plate, containing 132 μL of fresh media overlaid with a foil seal (n = 72). Cells were removed from overnight cultures via centrifugation (10,000 rpm for 2 min) and the supernatant retained. Cells were washed in phosphate-buffered saline via centrifugation before being resuspended in 3 mL of phosphate-buffered saline. The supernatant was filter-sterilised through a syringe filter with a pore size of 0.2 μm. We compared the growth of both species in both spent and fresh medium. To do this, filtered spent medium was added to a 96-well plate, with 128 μL per well, as well as separate wells for fresh media. About 4 μL of resuspended cells were then added to spent medium and fresh media and grown for 48 hours under experimental conditions (n = 64). To obtain time-course measurements, replicates were destructively sampled every three hours to measure optical density (n = 4 at each timepoint). After 48 hours, the final optical density was measured, and maximum growth rate was calculated over three time points.

Glucose utilisation assay

Glucose utilisation in our L. plantarum– and S. cerevisiae evolved strains was measured using a colorimetric enzyme-based Glucose Assay Kit (Sigma-Aldrich, catalogue number MAK263). Single colonies of ancestral (n = 4), monoculture (n = 4), and co-culture (n = 4) strains of both species were grown to saturation separately in 132 μL of CSM in a foil-sealed 96-well plate for 48 hours. After 48 hours, 10 μL of media was extracted from each well and diluted 100-fold in PBS to obtain measurements within the glucose standard-curve measurements. Measured samples were blanked against a glucose standard containing 0 ng/μL of glucose, then compared against a standard curve to estimate that the remaining amount of glucose is spent media of L. plantarum or S. cerevisiae strains.

Evolution experiment

To establish the evolution experiment, single clones of L. plantarum and S. cerevisiae were grown to saturation in Complete Supplement Mixture (CSM) modified with additional amino acids tryptophan and uracil. Co-culture populations were mixed at an initial ratio of 1:1 L. plantarum:S. cerevisiae and diluted 32-fold into 132 μL of supplemented CSM media, with 96 replicate co-culture populations distributed across a single 96-well plate. The 48-replicate monoculture treatment populations for each species were founded by the same procedure, but with S. cerevisiae and L. plantarum kept separate. We took a number of measures to reduce the chance of contamination. Co-culture and monoculture populations were propagated on separate 96-well plates and all dilutions were performed using the Hamilton96 liquid Handler with the CO-RE Probe Head, encased in a HEPA-filtered chamber. In addition, all cultures were foil-sealed before incubation, without shaking, at 28 oC. Transfers of cultures into fresh media occurred every 2 days in a 32-fold dilution (4 μL into 128 μL), resulting in 5 generations per transfer. After every 50 generations, populations were mixed with 50 μL of 75% glycerol and archived at −80 oC. The cultures were propagated for around 925 generations of growth and dilution at 28 oC, under non-shaking and oxygen limiting conditions. Six populations from each condition were chosen to track colony forming units (CFU) throughout the course of the experiment. Every 50 generations, CFU counts were taken for 6 populations from each condition. In all, 10 μL of culture was diluted 105-fold into 1x PBS and plated onto selective YPD agar–cycloheximide to select for L. plantarum and G418 to select for S. cerevisiae for each dilution. Plates with a suitable number of colonies were recorded and used to calculate CFU/mL.

Fitness assays

Fitness assays for both evolved S. cerevisiae and L. plantarum were conducted for every population in monoculture (n = 48) and co-culture (n = 96) by comparing population density after 48 hours of growth. Strains coming from the −80 oC freezer were passaged by diluting 32-fold in a 96-well plate to re-acclimatise over 48 hours of growth. Populations were passaged into fresh media again, but separated by species using selective media for L. plantarum (cycloheximide) and S. cerevisiae (G418) in separate 96-well plates and grown under experimental conditions for 48 hours. Growth plates were then shaken at 1000 rpm and then measurements of optical density (600 nm) were taken.

Whole-genome sequencing

DNA was extracted using GenFind v3 kit (Beckman Coulter). Modifications for the enzymatic lysis were made as follows: S. cerevisiae lyticase (Sigma, 1000 U/ml) for 60 min. at 30 oC, cells were pelleted and resuspended in proteinase K and RNAse A (100 mg/mL) and incubated at 50 oC for 60 min. For L. plantarum lysozyme (Sigma, 100 mg/ml) for 60 min. at 37 oC followed by proteinase K and RNAse A as above. DNA concentrations were measured on a Quantus analyser (Promega) before library preparation using Illumina DNA Prep kit (Illumina) using a modification manufacturer’s instructions to reduce reaction volumes to 25% of the recommended amount. Libraries were sequenced on the NextSeq 550 (Illumina) platform, using a Mid Output Kit v2.5 (150 cycles) [39]. An appropriate reference for the S. cerevisiae strain YJM978 was available (Project accession number: PRJNA189874; [40]): however, de novo assembly of the L. plantarum, short reads revealed no available reference sequences appropriate for use in reference-guided assembly. Therefore, to generate a reference sequence for the L. plantarum ancestor, DNA was prepared for long-read sequencing with the MinION device (Oxford Nanopore Technologies) according to protocols previously outlined [41]. Long reads were de-multiplexed and base-called according to Wick et al. [42]. The Unicycler v0.4.4 and Bandage v0.8.1 software packages, respectively, were used for hybrid assembly of short and long reads [39], and visualisation thereof [43]. The presence of a series of long repetitive regions in the bacterial chromosome precluded its complete assembly, but resulted in the contiguous assembly of 96% (3,029,606 bp) of the chromosome, with the remainder predicted to represent a triplicated region of 40,495 bp (triplicate based on mean read depth relative to that of the long contig). The sequences of two potential plasmids (9211 and 3493 bp in length) carried by the ancestral L. plantarum stain were also resolved as circular replicons, although it cannot be ruled out that these forms extended repeat regions within the LPKH chromosomal genome (Accession number PRJNA749634). Annotation of the hybrid assembly was completed using RAST [44]. Open-reading frames from RAST annotation underwent BLAST searches to look for homologous gene names; if a gene homologue was found with 100% sequence similarity in either Escherichia coli or Bacillus subtilis, then the gene name was amended to the RAST annotation. The completed L. plantarum sequence was denoted strain LPKH. Both the nearly assembled sequence and the short and long reads from sequencing were uploaded to NCBI databases under the BioProject number PRJNA749634.

After 925 generations, 10 evolved populations from each…

Read More:Species interactions constrain adaptation and preserve ecological stability in an

Notify of
Inline Feedbacks
View all comments