Methods

Inoculation and Growth:

An initial 1 L seed culture was prepared to condition cells by inoculating 1 L of SN media with 10 ml of a growing Crocosphaera watsonii culture obtained from the Waterbury laboratory (WH 8501) and incubating at 27.5 ^oC under standard cool white light.  Cells were monitored for growth via epifluorescence microscopy.  After 10 days of growth, the 1 L culture was concentrated to ca. 12 ml and transferred to a fresh 20 L of SN media and incubated under the same conditions.  20 L of SN media with no cells added was also incubated under these conditions alongside the culture to serve as a media-only control.  A suite of measurements was taken during growth in this large-volume format to monitor the growth and organic carbon production of the culture as compared to the media-only control.  Subsamples were taken from the culture and control for analysis of the following parameters as described in ancillary measurements: nutrients, particulate organic carbon (POC) and particulate organic nitrogen (PON), DOC and dissolved organic nitrogen (DON), dissolved organic phosphorus (DOP), chlorophyll, and direct cell counts via epifluorescence microscopy.  After ca. 8 days of growth with periodic gentle mixing, cells reached stationary phase and the culture was harvested to remove cell biomass and extract extracellular DOC material as described in the extraction protocol.

Ancillary measurement protocols for the diazotrophic cyanobacterium, Crocosphaera watsonii (WH 8501):

All subsamples for ancillary measurements were taken and processed using combusted glassware.  Measurements were taken immediately prior to and after inoculation as well as 2, 4, 5, 6, 7 and 8 days post-inoculation.  One final suite of measurements was also taken after removal of all cell biomass, prior to DOC extraction after 8 days of growth.  Samples for nutrient analysis were processed by filtering ca. 30 ml of sample through a pre-rinsed 0.45 µm capsule filter (Millipore Sterivex-HV filter unit) and collecting in acid cleaned polyethylene bottles prior to freezing.  Nutrient samples are analyzed at the Woods Hole Oceanographic Institution Nutrient Analytical Facility for ammonium, orthophosphate, silicate, nitrite and nitrate.  Samples for POC and PON analysis were taken by vacuum filtering ca. 50 ml of sample onto a combusted 25 mm 0.7 µm glass fiber filter (Whatman GF/F).  Filters were then placed inside a combusted glass petri dish, wrapped in foil, and immediately frozen.  This process was repeated to obtain duplicates of each sample and a blank (combusted filter only) was also taken during each sampling.  Filters were then thawed in batches and put in a drying oven overnight before placement in concentrated hydrochloric acid (HCl) fumes.  After 24 hr, filters were folded and placed into 9x10 mm tin capsules in a 96-well plate and shipped to the University of California Davis Stable Isotope Facility for analysis.  Samples for DOC and DON analysis were taken by transferring ca. 30 ml of the POC/PON filtrate to a combusted glass vial and adding 75 µl of 25% phosphoric acid before refrigeration.  Samples are analyzed for DOC and DON by high-temperature combustion using a Shimadzu TOC-V_CSH with platinum catalyst and an ASI-V auto sampler and TNM-1 total nitrogen detector at the Woods Hole Oceanographic Institution.  Samples are run alongside blanks and reference standards for low carbon water and deep seawater obtained from the Hansell laboratory at the University of Miami.  Samples for DOP analysis were taken by transferring ca. 50 ml of the POC/PON filtrate to an acid cleaned polyethylene bottle, which was then frozen.  Samples are analyzed for DOP using a novel iron oxide XAD resin extraction method and characterized by high-resolution mass spectrometry.  Samples for chlorophyll analysis were taken by vacuum filtering ca. 50 ml of sample onto a 25 mm 0.7 µm glass fiber filter (Whatman GF/F).  Filters were then wrapped in combusted foil and immediately frozen.  This process was repeated to obtain duplicates of each sample.  Samples are analyzed for chlorophyll-a concentration as a proxy for phytoplankton biomass using a fluorometer (Turner Designs).  Cells were monitored for growth via epifluorescence at 100X magnification on a Zeiss light microscope.

Harvest & Extraction

Extraction Protocol for the diazotrophic cyanobacterium, Crocosphaera watsonii (WH 8501)

Removal of Cell Material:

After ca. 8 days of growth with periodic gentle mixing, cells reached stationary phase and the culture was processed to remove cell material and to store spent media for future isolation of DOC exudates.  Cells were removed by centrifugation at 7,000 RPM for 15 min, followed by 0.2 µm filtration (Whatman Polycap 36 TC capsule filter).  Cell pellets from the centrifugation were combined and frozen for future chemical analyses.  The 0.2 µm filtrate was then frozen until solid-phase extraction of DOC compounds.  The media-only control was processed and stored alongside the culture in an identical fashion.

Extraction of Extracellular DOC:

Initially, a 1 L subsample of both the culture and control sample was processed using the same protocol listed below to test methodology and analysis procedures.  Afterwards, the remaining volume of both the culture and control was thawed and filtered through a combusted 47 mm 0.7 µm glass fiber filter (Whatman GF/F) to remove any flocculent formed during the freeze/thaw process.  The filtrates were then acidified by adding ca. 13 ml of trace metal grade concentrated hydrochloric acid before loading onto a column (300x7.8 mm) of purified octadecyl (C₁₈) functionalized silica gel (Sigma-Aldrich) at a flow rate of 5 ml/min.  Silica gel was purified using a soxhlet extractor with 24 hr rinses of HPLC-grade methylene chloride, methanol, and ultrapure water.  Samples were taken of what went into and came out of the column for DOC analysis to assess the ability of the silica gel to retain the DOC compounds in each sample.  Columns were then washed with ca. 20 column volumes of acidified ultrapure water (pH 2-3) at a flow rate of 1 ml/min to remove salts before eluting with ca. 10 column volumes of acidified HPLC-grade methanol (pH 2-3) at a flow rate of 1 ml/min.  Salt removal was confirmed using a silver nitrate solution and methanol elutions were collected in combusted flasks and then concentrated using a rotary evaporator.  Samples were centrifuged to remove any flocculent and dried completely using purified nitrogen gas.   A subsample of each elution was placed onto a combusted 25 mm 0.7 µm glass fiber filter (Whatman GF/F) for POC analysis prior to drying in an effort to further quantify the amount of organic carbon recovered via solid-phase extraction.  A subsample was also removed prior to drying for chemical analysis as described in the LC-MS Protocol.  All dry samples are stored in combusted amber vials at 4^oC in the dark.

Chromatographic Methods

Samples library were acquired using an Agilent 1200 series liquid chromatograph comprised of a G1379B degasser, G1312A binary pump, G1367C automatic liquid sampler and F1315C diode array detector coupled to an Agilent 6130 mass spectrometer all controlled and processed with Agilent Chemstation version B.03.02.

The mobile phases were A = 0.1% formic acid in water and B = 0.1% formic acid in methanol. Samples were dissolved in 100% B, and 25 ul (corresponding to 250 ml of culture or medium) was injected onto an Agilent ZORBAX SB-C18 column (3.5um 4.6x150mm P/N: 863953-902). The column flow was 1.0 ml min-1 starting with 100% A, ramping to 80% B at 25 min, ramping to 100% B at 35 min and holding until 55 min, ramping to 0% B at 65 min and holding until 75 min.
Full scan absorbance data were acquired from 210 to 800 nm with a 2.0 nm step and 4 nm slit width.

Mass spectrometry used the atmospheric electrospray ionization ("API-ES") source. Source conditions were: drying gas at 11.5 l min-1, nebulizer at 60 psig, drying gas temperature at 300 ºC, capillary voltages at + or - 4000 V. Acquisition ranges were from 100-2000 Da in the positive mode and 230 - 2000 Da in the negative mode, with each mode taking 50% of the 1.56 sec total cycle time. Both positive and negative data used a fragmentor at 4.0, threshold at 150 and step size of 0.1.

Optical data were processed within the Agilent Chemstation software by subtracting specific wavelength traces (254 or 290 nm) of the medium from the same wavelength trace of the culture sample.

Mass spectral data were processed by converting the native Agilent ".ms" format files to the industry standard ".cdf" (common data format) files using ChemSW's "GCMS File Translator Pro". The .cdf files were then converted to Thermo Xcalibur ".raw" format using the Xcalibur "File Converter" tool. Once in the .raw format, the Xcalibur QualBrowser "Backround Subtract" tool was used to subtract the medium file from the culture file. The resulting medium subtracted culture files are displayed as the total ion current ("TIC") or topological view data in the Analytical Results. These medium subtracted culture files are the starting point for all data extraction using the freely available metabolomics software "mzMine" (http://mzmine.sourceforge.net/).  In mzMine, significant "features" (local maxima in a plot of intensity vs. mass vs. time, such as are seen in the topological views) from several cultures can be compared to find those that are common or unique to each culture. We also looked for features indicative of molecular ions, by employing a filter for only those features with 22 amu higher parallel features at the same time, and with a given intensity ratio. (22 amu is the difference between the atomic weight of sodium adduct and a hydrogen adduct, both of which are commonly expected to be seen for molecular ions under electrospray conditions.) These potential molecular ions are then the starting point for our further analysis by high resolution mass spectral techniques (such as Fourier Transform -MS) to determine their accurate molecular weight and elemental composition, or by MS/MS techniques using an ion trap or triple quadrupole MS to determine their molecular structure.