Identification of proteins and selection of peptides

We performed a literature search to identify proteins involved in psB₁₂ production (Cob/Cbi proteins) and utilization (MetH and NrdJ). BLASTp was used to search for homologues of these Synechococcus sp. WH8102 proteins in other available marine Synechococcus genomes using the GenBank nr database. Proteins were aligned using COBALT (Papadopoulos & Agarwala, 2007) and digested in silico with trypsin to reveal conserved tryptic peptides. Sequences were introduced in the Unipept Tryptic Peptide Analysis search (Gurdeep Singh et al., 2019) to determine if the peptide is present in a significant number of the Synechococcus strains of interest. Peptides for CobO, cob(I)alamin adenosyl transferase, and MetH, methionine synthase, were selected for analyses as they met the peptide selection guidelines for SRM targeted assays described in Hoofnagle et al. (2016) (Hoofnagle et al., 2016) and were largely conserved across Synechococcus clades. Peptides are reported in Table S1. We found no suitable peptide with adequate coverage across Synechococcus clades for NrdJ.

Culture experiments to measure Synechococcus peptides and pseudocobalamin quotas

Synechococcus sp. WH8102 (CCMP2370) obtained from the National Center for Marine Algae and Microbiota (NCMA, Bigelow Laboratory, USA) was grown axenically in plastic, vented 250 mL flasks (10861–576, VWR) with modified Synthetic Ocean Water media (Dupont et al., 2008) void of any added cobalamin, under ~20 μmol photons m⁻² s⁻¹ supplied through white LED lights (LEDMO-EZ550). Irradiance was measured by a PAR irradiance sensor (QSL2101, Biospherical Instruments) and set at a light:dark cycle of 12:12 h (Table 1). Throughout culturing, we monitored growth via relative fluorescence units (RFU) measured in a 10-AU Fluorometer (Turner Designs, San Jose, CA), and cell counts using a BD Accuri™ C6 flow cytometer as described below. Cultures were also tested for axenicity once a week using SYBR Green II dye (Invitrogen) followed by flow cytometry screening. Specific growth rates (d⁻¹) were calculated using linear regression of ln cell counts during exponential growth.

For strain Synechococcus sp. WH8102 grown in batch culture, we harvested samples at both exponential and stationary phases (Figure S1). For the early exponential phase, we harvested 35 mL for protein samples, 17 mL for metabolite samples, and 8 mL for POC samples at the three points in the diel cycle. In the stationary phase, we harvested 20 mL for protein samples, 10 mL for metabolite samples, and 5 mL for POC samples at the three points in the diel cycle. The three points of the diel cycle evaluated were: 1.5 hours after the lights went on (9:30), 1 hour before the lights went off (18:00), and 3 hours after the lights went off (22:00).

Synechococcus sp. WH8102 culture cell counts

Culture cell density was measured on an Accuri C6 flow cytometer (BD Biosciences) equipped with a Red (640 nm) and Blue (488 nm) laser, after labelling with1:1000 SYBR Green II dye (Invitrogen). Single-cell events were recorded on forward side scatter and side scatter plots and distinguished using optical filters of phycoerythrin (488 nm laser, filter = 572/28 nm) and chlorophyll a (488 nm laser, filter = 675/30 nm). Accuri run limits were set to a maximum of 500,000 events or 50 μL, with a flow rate of 35 μL/min and a core size of 16 μm. The gating strategy for Synechococcus populations was developed based on the Acccuri application note provided by BD Biosciences.

POC measurements

Samples were acidified for 6 h in a glass desiccator with an open bottle of concentrated hydrochloric acid (37% HCl, 250 mL) and later dried overnight at 45°C. Filters were packed in tin capsules and analysed for particulate carbon on an elemental analyser (Elementar Vario microcube) coupled to an IRMS (Isoprime 100). The samples were flash combusted at 1150°C to convert particulate carbon into CO₂ gas. These gaseous components were then analysed by the Isoprime 100. The values were blank-corrected and divided by cell number to give per cell quota.

Environmental sample collection

We collected environmental samples in the fall of 2016 onboard an AZMP (Atlantic Zone Monitoring Program) cruise aboard CCGS Hudson from 15 September 2016 to 6 October 2016 (HUDSON 2016027). Samples were taken at four stations on the Halifax line (HL 01, HL 02, HL 05.5 and HL 06). At each of the stations, we collected 1 L of water at each depth for particulate metabolite analysis. The sample was gently filtered through a 0.2 μm Nylon filter in the dark to reduce photodegradation. For protein samples, 10 L of water was pre-filtered (330 μm) from Niskin bottles and then filtered and gentled through 3 and 0.2 μm polycarbonate filters via peristaltic pumping. Parallel 1.8 mL seawater samples for analytical flow cytometry were fixed with 1% paraformaldehyde. Flow cytometry samples and filters were then frozen immediately at −80°C until analysis.

Environmental cyanobacterial cell counts

Preserved samples were thawed at room temperature, and prefiltered using 35 μm cell strainers into a 5 mL fluorescence-activated Cell Sorting (FACS) tube. The samples were then collected and analysed on a Novocyte 3000 with NovoExpress software (Agilent, USA) for natural fluorescence cell counts. The Novocyte 3000 is equipped with three lasers (red [640 nm], blue [488 nm], and violet [405 nm]), and the optical filters of interest are chlorophyll-a (488 nm laser, filter = 675/30 nm), divinyl chlorophyll a (405 nm laser, filter = 675/30 nm), phycocyanin (640 nm, filter = 675/30 nm) and phycoerythrin (488 nm laser, filter = 572/28 nm). The samples were collected with a flow rate of 120 μL/min and a threshold of chlorophyll >450 nm. The gating strategy was developed based on the Novocyte application note provided by Agilent and the Prochlorococcus gate was constructed with the aid of pure Prochlorococcus culture, an example of an environmental flow cytogram is provided (Figure S2).

Particulate metabolites extraction

Metabolites were extracted from both culture and environmental samples following the methodology described by Heal et al. (2017) with minor modifications. Briefly, 0.2 mL of each 100 μm and 400 μm silica beads were added to the bead beater tubes containing sample filters. One microlitre of ice-cold solvent mixture (40:20:20 acetonitrile:methanol:MQ water) was added and the mixture was agitated using a bead beater (MP Biomedicals) in 3 × 40-s pulses at 1800 rotations per minute (RPM) over 20 min. Heavy-CN-B₁₂ (Cambridge Isotopes Labs, CLM-9770-E) was used as an internal standard and was added to a final concentration of 3 nM in all metabolite samples prior to extraction. The solvent was removed under vacuum (Eppendorf, Mississauga, ON) at room temperature. Culture samples were re-suspended in buffer A (20 mM ammonium formate, 0.1% formic acid, 2% acetonitrile) to generate the equivalent of extracts from 12,100 cells/μL. Environmental samples were resuspended with 100 μL of buffer A. All samples were then vortexed and centrifuged at 10,000g for 3 min at 4°C and then diluted twofold with buffer in conical polypropylene HPLC vials (Phenomenex, Torrance, CA) prior to mass spectrometry analysis.

Metabolite quantification

HPLC–MS was used to quantify cobalamins using a Dionex Ultimate-3000 LC system coupled to the electrospray ionization source of a TSQ Quantiva triple-stage quadrupole mass spectrometer in SRM mode. Details of mass spectrometry conditions are reported in supplement methods and the transition list is reported in Table S2.

For metabolite samples, quality control (QC) samples were prepared by mixing equal volumes of each sample within its respective group (one QC for culture samples and one QC for environmental samples). Calibration curves with authentic cobalamin standards were prepared using the QC sample as a matrix and triplicate injections were performed for 0, 0.1, 5, 10, 50, and 100 fmol of each authentic cobalamin standard on analytical column with linearity up to 1000 fmol. All cobalamins were normalized by heavy-CN-B₁₂ internal standard to account for variability introduced by matrix effects, during extraction and instrument analysis. Limits of quantitation and limits of detection were calculated as 10x and 3x the variation in the QC sample without spike, respectively, and recorded in Table S3.

Given that commercial psB₁₂ standards are not available, the following assumptions were made for psB₁₂ quantification using associated cobalamin analogue standards: (1) B₁₂ and psB₁₂ ionize with the same efficiency, (2) B₁₂ and psB₁₂ fragments are generated similarly, (3) B₁₂ and psB₁₂'s different elution times do not greatly affect the response.

Protein extraction and digestion

Protein sample filters were submerged in 750 μL of SDS extraction buffer (2% SDS, 0.1 M Tris/HCl pH 7.5, 5% glycerol, 5 mM EDTA) and incubated for 10 min on ice before being heated for 15 min at 95°C in a ThermoMixer C (Eppendorf, Mississauga, ON) at 350 RPM. Samples were then sonicated on ice for 1 min with a Q125 Sonicator (Qsonica Sonicators, Newton, CT) at 50% amplitude and 125 W (pulse 15 s ON, 15 s OFF) then kept at room temperature for 30 min and vortexed every 10 min. The filter was then removed, and samples were centrifuged at 15,000 × g for 30 min at room temperature. 4x volume of ice-cold acetone was added to the supernatant for overnight precipitation of proteins at −20°C before washing with 3 × 400 μL of ice-cold acetone and 1 × 400 μL with ice-cold methanol. Following each wash, samples were centrifuged at 15,000g for 30 min at room temperature and the supernatant was removed. The pellet was dried down under vacuum (Eppendorf, Mississauga, ON) for 15 min at room temperature. Extracted protein was resuspended in 20 μL of 8 M urea for 10 min at room temperature then80 μL of freshly made 50 mM ammonium bicarbonate was gradually added to get a final solution of 1.6 M urea and 40 mM ammonium bicarbonate.

Aliquots of 50 μg of protein were removed from each sample and diluted to 100 μL with 50 mM ammonium bicarbonate for protein digestion. Proteins were reduced with 5 mM dithiothreitol (DTT) at 56°C for 30 min in a ThermoMixer, then alkylated with 15 mM iodoacetamide at room temperature in the dark. Residual iodoacetamide was quenched with a second addition of DTT to give a final concentration of 10 mM DTT. Protein was digested with 1 μg trypsin (Thermo Scientific, Waltham, MA) at 37°C overnight. To halt the digestion, 0.5 μL of formic acid was added. A modified protein Micro BCA kit (Thermo Scientific—p/n 23,235)was conducted to confirm the protein concentration of each sample before injection using a trypsin-digested BSA (bovine serum albumin) standard to make the calibration curve.

Peptide quantification

Targeted mass spectrometry was performed using a Dionex Ultimate 3000 UPLC system interfaced to a TSQ Quantiva triple-stage quadrupole mass spectrometer (MS) (Thermo Scientific, Waltham, MA), fitted with a heated, low-flow capillary ESI probe (HESI-II). Isotopically labelled, heavy internal standard versions of each peptide were synthesized by Thermo Scientific™ at >95% purity. Heavy peptide standard was added to samples to yield 20 fmol on the column for every 1 μg total protein injected. Details of mass spectrometer conditions are reported in supplement methods. Each sample was analysed via triplicate injections using the transition list of peptides that are found in Table S1.

Statistical analysis

Statistical analyses were performed using R (RStudio Team, 2020). A one-way ANOVA followed by a post-hoc Tuckey's Test was used to determine the differences in Synechococcus culture experiments. A two-way ANOVA followed by a post-hoc Tuckey's Test was used to determine the effects of the growth phase and diel cycle on the same variables. Differences were considered significant when a p-value <0.05 was observed.