Strains

Six clones each of the open ocean Synechococcus strain WH8102 and the coastal Synechococcus strain CC9311 were obtained by dilution to extinction in SN media [1]. The parent cultures of each organism were obtained from the National Center for Marine Algae (Boothbay Harbor, Maine) and were axenic upon receipt. Six clones of Alteromonas sp. strain EZ55 and Prochlorococcus MIT9312 were also previously obtained and cryopreserved at -80 °C [2]. The EZ55 clones used in our Synechococcus co-cultures were the same 6 clones used in our previous transcriptomic study of MIT9312 [2] in order to maximize the comparability of results between that study and the present study. Co-cultures were initiated by mixing each of the six clones of CC9311 and WH8102 with one of the EZ55 clones.

Culture conditions

Synechococcus cultures were grown under similar conditions to those described in our previous experiment with Prochlorococcus [2]. Briefly, all cultures were prepared in acid-washed conical-bottom glass centrifuge tubes containing 13 mL of artificial seawater (ASW) amended with nutrient stocks [1] and with acid and/or base to control pCO2. ASW (per L: 28.41 g NaCl, 0.79 g KCl, 1.58 g CaCl2*2H2O, 7.21 g MgSO4*7H2O, 5.18 g MgCl2*6H2O) was sterilized in acid-washed glass bottles, amended with 2.325 mM (final concentration) of filter-sterilized sodium bicarbonate, then bubbled with sterile air overnight. Synechococcus cultures were grown in SEv (per L: 32 μM NaNO3, 2 μM NaH2PO4, 20 μL SN trace metal stock, and 20 μL F/2 vitamin stock). The primary differences between this medium and the PEv medium used in our earlier Prochlorococcus study are the nitrogen source (NO3- vs. NH4+, with molar concentration of N and N:P ratios identical to PEv) and the addition of F/2 vitamins [1]. Carbonate chemistry of each media batch was determined prior to pCO2 manipulations by measuring alkalinity and pH by titration and colorimetry, respectively [2, 3] and then using the oa function in seacarb package in R to determine how much hydrochloric acid and bicarbonate (for 800 ppm pCO2) or sodium hydroxide (for 400 ppm pCO2) was needed to achieve desired experimental conditions [4]. Acid and base amendments were introduced immediately prior to inoculation. Cultures were grown in a Percival growth chamber at 21º C under 150 μmol photons m-2 s-1 on a 14:10 light:dark cycle. Synechococcus cultures were grown on a rotating tissue culture wheel at approximately 60 rpm.

Growth experiments

The transcriptomes of all six clonal replicates of each Synechococcus strain along with their EZ55 partners were assessed under approximately 400 (based on atmospheric pCO2 measured at Mauna Loa in 2015, when the experiment was planned) or 800 ppm (i.e., approximate predicted year 2100 pCO2 under IPCC scenario A2) pCO2. Prior to RNA extraction, each culture was acclimated to experimental conditions for three transfer cycles (approximately 14 generations). Growth was tracked by flow cytometry using a Guava HT1 Flow Cytometer (Luminex Corporation, Austin, TX). EZ55 cell concentrations were determined by dilution onto YTSS agar (per L, 4 g tryptone, 2.5 g yeast extract, 15 g sea salts, 15 g agar). Whenever Synechococcus cell densities reached 2.6 x 105 cells mL-1, cultures were diluted 26-fold into fresh media. Preliminary experiments revealed that this cell concentration was low enough that growth was not limited by nutrients and pH and pCO2 were not significantly impacted by cyanobacterial carbon concentrating mechanisms. In the final transfer cycle, each culture was split into 5 identical subcultures to increase the biomass available for RNA extraction; all 5 subcultures of each clone were then pooled and collected on a single 0.2 mm polycarbonate filter by gentle syringe filtration, then flash-frozen in liquid nitrogen and stored at -80o C prior to RNA extraction. For WH8102 cultures, an average of 4.04 x 107 WH8102 cells and 3.91 x 108 EZ55 cells were collected per filter, and for CC9311 cultures, an average of 5.47 x 107 CC9311 and 7.33 x 108 EZ55 cells were collected per filter.

RNA library preparation and sequencing

RNA extraction was performed separately for each replicate culture with the RNeasy Mini Kit (Qiagen, Valencia, CA, USA) with a small modification of the lysis step [2]. rRNA was removed with the Ribo-Zero rRNA Removal Kit for Bacteria (Illumina, San Diego, CA, USA) [7]. Following rRNA removal, samples were purified and concentrated with a RNeasy MiniElute cleanup kit (Qiagen). Quantity and quality of post-digestion RNA were assessed with an Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA, USA). mRNA library preparation for Illumina Hi-seq 2500 paired-end sequencing (PE100) used TruSeq RNA sample prep kit v2 (Illumina, San Diego, CA, USA). DNA fragment length was 100 bp, paired ends were non-overlapping, and the insert size was approximately 300 bp. Individual barcode sequences were added to sequence reads for multiplex sequencing which were run in a single lane at the Sulzberger Columbia University Genome Center (CUGC) (New York, NY, USA).

EZ55 growth experiments with photorespiration metabolites

We investigated the ability of EZ55 to grow on metabolic intermediates in the photorespiration pathway as their sole carbon source. Glycine, glycolate, glucose, and glyoxylate stock solutions (concentrations of 20%, 4%, 4% and 10% W/V, respectively) were filter sterilized using a 0.2 μM filter. The pH of glycolate and glyoxylate stocks was adjusted to approximately 7 using 10 M NaOH. EZ55 clones were inoculated into ASW supplemented with Pro99 nutrients [1] and 0.1% (W/V) glucose [5] and acclimated for 24 hours at 28 ℃ with orbital shaking at 120 rpm. The detection of intracellular H2O2 was performed according to Lu et al. [6], with slight modification. Briefly, 1.5 ml of culture was centrifuged at 8000 rpm for 5 min, the supernatant was removed, and the pellet was resuspended in 1 ml phosphate buffered saline (pH=7.4, Fisher). 5 μl of 1 mM 2′,7′-dichlorodihydrofluorescein diacetate was added to the resuspension and vortexed for 5 seconds and then incubated for 1 h on a shaker (120 rpm) in the dark. The suspension was centrifuged at 8000 rpm for 5 min, the pellet was washed twice with PBS, and finally resuspended in 200 μl PBS. Fluorescence was measured by flow cytometry at excitation/emission wavelengths of 485/535 nm.

Detection of glycolate utilization genes (see related dataset "Pipelines for transcriptome analyses" https://www.bco-dmo.org/dataset/881942)

Several genes involved in the bacterial glycolate utilization pathway (glycolate/lactate oxidase, the 3 subunits of glycolate dehydrogenase, and tartronate semialdehyde reductase) were not annotated in the reference genomes for our organisms so we specifically sought to detect them using a reciprocal BLAST analysis. We retrieved any sequences from each of the four reference genomes with high similarity (E-value < 0.001) to the relevant genes from Escherichia coli and/or Synechococcus elongatus using blastp [7] and then back-matched each retrieved sequence to the E. coli or S. elongatus reference genome. If the reciprocal match was the same gene used in the original BLAST search, we considered the match significant.

Problems/Issues
In some of our earliest cultures, too few daily flow cytometry measurements were collected to calculate robust exponential growth rates, because it was impossible to confirm at least three data points corresponding to the logarithmic growth phase of the culture. For these cultures, only malthusian growth rates are reported.