Previous   |    Next

doi:10.2204/iodp.proc.329.102.2011

Microbiology

Research conducted during scientific ocean drilling continues to reveal significant information regarding subseafloor microbial habitats, the diverse microbial communities that reside in those habitats, and the biogeochemical implications of their metabolic activities. Using a combination of shipboard and shore-based analyses, we gain knowledge of the distribution of their biomass, metabolic activities, and phylogenetic relationships. Expedition 329 is dedicated to assessing the factors that control distribution, activities, and composition of microbial life in organic-poor sediment and the underlying basalt. During the expedition, microbial abundance was quantified, experiments were initiated to quantify activities using multiple stable isotope and radioisotope tracers, and samples were taken for postexpedition characterization of communities by a variety of molecular techniques targeting DNA and RNA at both single-cell and community levels. Additional information on the types of microorganisms that inhabit these subseafloor environments will be obtained by examining microbes isolated by a wide suite of culturing methods. In order to ensure that we are indeed analyzing indigenous microbial life and its activities, samples were routinely taken for assessment of contamination during coring and sampling.

Core handling and sampling

Drilling

Microbiological sampling depends on careful and appropriate sample handling techniques. Because the samples were retrieved from very stable sedimentary environments, subseafloor microorganisms are expected to be sensitive to chemical and physical change, in particular to changes in oxygen concentration and temperature. Consequently, all cores to be sampled for microbiological studies were transferred from the drilling platform to the Hold Deck refrigerator (~7°–10°C) as quickly as possible and kept as whole-core sections until processed. As initiated during ODP Leg 201 (D’Hondt, Jørgensen, Miller, et al., 2003) in order to avoid intermittent warming of retrieved cores, the conventional core handling procedure was modified. Once a core was retrieved, it was immediately transferred to the catwalk for labeling and cutting of sections before the next core barrel was deployed. Efforts were also made to obtain advanced piston corer (APC) cores even when this led to an increase in core recovery time, as APC cores are generally much less disturbed than extended core barrel (XCB) cores.

While drilling cores for microbiology, potential for contamination from surface-derived microbes is high. Consequently, quantification of potential contamination and avoidance of core disturbance are critical for microbiological studies. To check for potential intrusion of drill water from the periphery toward the center of cores and, thus, to confirm the suitability of core material for microbiological research, samples were routinely taken for contamination tests using solutes (perfluorocarbon tracer [PFT]) and, for basalt cores, cell-sized particles (fluorescent microspheres). Furthermore, the freshly collected cores were visually examined for cracks and other signs of disturbance by observation through the core liner. Core sections observed to be disturbed before or after subsampling were not analyzed further.

Whole-round core sampling in the core refrigerator (cold room)

It is important to emphasize that the different analyses, experiments, and cultivation attempts that fall under the rubric “microbiological methods” have widely different requirements concerning handling and storage. Keeping samples cool and processing times short and minimizing contamination were the key criteria for determining how the core sections were processed. Therefore, all core sections sampled for microbiology or geochemistry were transferred to the core refrigerator on the Hold Deck for most microbiological sampling (the only microbiological samples collected on the catwalk were for cell counts and PFT analyses). To minimize changes in microbial populations, all core processing for whole-round core microbiological samples took place in the refrigerated core room on the Hold Deck of the JOIDES Resolution at ~7°–10°C. This room was equipped with a workbench, working space for two to four persons, and all sampling tools, including core cutters, chain vises, sample containers, label maker, and so forth. Because the core liner is not sterile and the outer surface of the core is contaminated during drilling, subsampling of whole-round cores excluded the sediment next to the core liner. Where appropriate, handling and subsampling were performed under anoxic conditions using an anaerobic glove box in the cold room.

Normally, all 1.5 m sections from all cores were held in the cold room in the event section(s) intended for microbiological sampling were found during sampling to be disturbed. In cases where cores were determined to be disturbed, sections from other cores were sampled. The core liner was cut by the standard IODP core cutter and cut with an ethanol-wiped spatula. Some whole-round core sections were immediately capped with new end caps and stored at 4°C for shore-based analysis, whereas others were immediately subsampled into sterile 5, 10, 20, or 60 cm³ tip-cut syringes. Samples (whole-round cores or syringe subcores) for molecular analysis were transferred to ultralow-temperature freezers (–80°C) on board.

Basaltic core sampling

During Expedition 329, basaltic rock of different ages was recovered from three sites using the rotary core barrel (RCB) coring system. In addition, small pieces of the uppermost altered basalt were recovered from sediment/basalt interfaces at most drilling sites using standard APC coring. Those interface samples were subsampled aseptically in the cold room for microbiological and biomineralogical investigations. The basalt cores were brought to the cold room after recovery and stored there until mineralogical and microbiological subsampling. To maintain the moisture of the cores, wet sponges were placed in the core liner. For microbiological subsampling, pieces of basalt or altered basalt were cracked with a flame-sterilized hammer and chisel and suspended in a 3% NaCl solution to remove surface contaminants and cuttings. Some portions of the “contaminated” suspension were also subsampled as a reference for subsequent analyses. If there was enough material for multiple microbiological experiments, a piece of the surface part was briefly flamed and crushed on a flame-sterilized tray for aseptic subsampling. The aseptically prepared basalt pieces were then further powdered with a frame-sterilized percussion mortar, and the powdered samples were used for cultivation, cell enumeration, and shore-based molecular analyses.

Quantifying potential contamination

To evaluate the extent to which contaminating cells may have penetrated a sample, the shipboard science party (1) continuously injected PFT into the drilling fluid (in sediment and basalt) and (2) introduced fluorescent microspheres at the drill bit (in basalt). Contamination will be further assessed by postcruise genomic comparison of the basalt or sediment sample with drilling fluid from the time of coring. The two shipboard techniques have been successfully used on multiple ODP legs and during Expedition 301 (Smith et al., 2000a, 2000b; House et al., 2003; Lever et al., 2006). The third technique is common for studies of seafloor samples, in which contamination is ubiquitous and genomic signatures of the contaminating material are subtracted from those of the seafloor sample. Previous shipboard PFT studies have consistently shown that the interior of cores are much less contaminated than core peripheries (by factors of 3–300) and that APC cores are generally not significantly contaminated (Smith et al. 2000a; House et al., 2003). The uppermost 1.5 m section of an APC core tends to be more contaminated than deeper sections (Lever et al., 2006). Contamination tends to be much greater in RCB cores (e.g., on surfaces of cored basalt) and XCB cores than in APC cores. In all categories of core, potential contamination varies considerably from sample to sample. Consequently, to avoid contamination of microbiological results, contamination tests must be conducted on each core or hard rock sample that is used for a microbiological experiment (D’Hondt, Jørgensen, Miller, et al., 2003).

Perfluorocarbon tracer

The PFT used on the JOIDES Resolution during Expedition 329 was perfluromethylcyclohexane. This compound is volatile, with a boiling point of 76°C, and is chemically inert. PFT was introduced into the drilling fluid with a high-pressure liquid chromatography (HPLC) pump at a constant concentration of 1 mg/L as a proxy for potential contamination of the core material by nonindigenous microbes. It is important to work quickly but carefully with the core material; because the PFT is volatile, it is possible to contaminate internal core material with the PFT (and microbes) from the relatively closed atmosphere of the cold room or the outside of the core section.

For PFT measurement of sediment cores, 3 cm³ subsamples were taken with a cut-off syringe and directly extruded into a gas chromatograph sample vial. During Expedition 329, we tested a new PFT measurement approach using iso-octane as a solvent. Fifteen milliliters of iso-octane (2,2,4-trimethylpentane) was added to the vial, which was then quickly sealed. Sample vials were placed on the wrist shaker (Burrell) and shaken for a minimum of 5 min. One microliter of the supernatant was injected into a gas chromatograph (Agilient 6890) with an electron capture detector (ECD). A calibration curve of PFT in the solvent was prepared from concentrations of 10^–10 to 10^–6 by serial dilution. Each standard was injected into the gas chromatograph, and a curve of response factors was developed.

Using the new method as described above, the sediment subcore did not break down into the iso-octane but remained as a solid piece. There was very little improvement after 24 h of vigorous shaking. The low efficiency hindered the release of PFT into the iso-octane phase and hence a reliable quantity of PFT could not be extracted (see “Microbiology” in the “Site U1365” chapter [Expedition 329 Scientists, 2011]). Therefore, as an alternative method of PFT measurement, we prepared PFT samples using the previously established method (Smith et al. 2000a; House et al., 2003; Lever et al., 2006), slightly modified by taking 3 cm³ sediment samples on the catwalk immediately after core recovery and placing them into gas chromatograph vials with 2 mL of 18.2 MΩ water. Each vial was quickly sealed and stored upside down at 4°C for later analysis. The analyses will be done postexpedition.

Fluorescent microspheres for basalt contamination test

Fluorescent microspheres of a size similar to indigenous microorganisms (0.5 µm) have long been used in drilling operations to assess dispersal and transport of microbe-sized objects (Harvey et al., 1989). Yellow to green fluorescent microspheres (Fluoresbrite carboxylate microspheres; Polysciences, Inc., 15700) with a diameter of 0.52 ± 0.01 µm were used as a particulate tracer during basalt coring. These microspheres are highly fluorescent (458 nm excitation; 540 nm emission). They appear bright green when observed by epifluorescence microscopy through a blue filter set (Zeiss filter set 05 or 08) (Smith et al., 2000b).

Microspheres were only deployed on cores that recovered basalt samples for microbiological cultivation and molecular biological analyses. The concentration of microspheres was set at 10¹⁰ spheres/mL (Smith et al., 2000b; House et al., 2003). Microspheres were deployed in plastic bags containing 40 mL of microsphere suspension in 18.2 MΩ water (2 × 10¹¹ microspheres in a 40 mL bag). The bag was then heat-sealed and placed into an additional plastic bag that was open at each end. By attaching the loose plastic ends with cord, the bag was wedged into a shim above the core catcher and stretched across the throat of the core barrel. The bags were ruptured as core entered the barrel, releasing the microspheres.

Concentrations of fluorescent microspheres in core samples were quantified using a Zeiss Axiophot epifluorescence microscope fitted with a mercury lamp (HBO 100 W), a blue filter set, and a 100× Plan-NEOFLUAR oil-immersion objective. Nonfluorescent immersion oil was used for all observations.

For hard rock samples, aliquots (100 µL) of the crushed rock were suspended in 10 mL of 0.2 µm filtered Tris-buffered saline (TBS; pH 7.4) solution and filtered onto black, 25 mm diameter polycarbonate filters (0.2 µm pore size) in a filtration tower. The filters were then mounted on microscope slides with a drop of nonfluorescent immersion oil and covered with a coverslip. The microspheres on the filter were counted using epifluorescence microscopy as described above. Microsphere abundance on the filter was determined by averaging the total number seen in at least 20 randomly selected fields of view. However, it is worth noting that we were not able to assess the contamination level quantitatively based on the number of fluorescent beads because the bag is crushed immediately at the initiation of RCB coring and the drilling fluid (i.e., seawater) subsequently dilutes the in situ concentration of fluorescent beads during drilling.

Assessment of potential contamination sources

RCB drilling requires huge amounts of surface seawater to be pumped into the borehole. This water is the major source of microbial contamination to cores collected for microbiological analyses. In order to check for this contamination (independent of the tracer tests), drilling fluid, surface seawater, and bottom seawater above the mudline were collected. Microorganisms present in these fluids were collected on 0.2 µm pore filters. The filters were frozen (–80°C) and will be analyzed postcruise for 16S rRNA. In some cases, filters were divided into pieces using a sterile scalpel for microbiological and biomineralogical investigations.

Cell detection and enumeration

Sediment

One of the fundamental objectives of subseafloor biosphere research is to understand the global distribution of microbial populations and their activities. Previous microbiological efforts of scientific ocean drilling expeditions have mainly focused on organic-rich continental-margin sediments. Hence, the distribution of organisms in organic-poor subseafloor environments, such as the subseafloor in the South Pacific Gyre, remains largely unknown. Previous study of shallow (≤9 mbsf) sediment from the South Pacific Gyre showed that cell abundance is several orders of magnitude lower than in any other marine sedimentary environment studied to date (D’Hondt et al., 2009). Even in samples from the sediment/water interface, microbial abundance was only ~10⁵ cells/cm³, decreasing with depth to values of 10³ to 10⁴ cells/cm³ by ~9 mbsf. These results suggest that microbial populations in deeper sediment realms might be extremely low, requiring very sensitive technology to obtain accurate biomass estimates.

Because extremely low cell abundances were expected in the South Pacific Gyre, cells were extracted from the sediment prior to counting according to the protocol of Kallmeyer et al. (2008) with slight modifications as described below. The extracted cells were stained by SYBR Green I fluorescent dye and enumerated by direct counting with epifluorescence microscopy.

Samples close to the sediment/water interface may harbor cell abundances that are high enough to be counted directly without cell extraction. In order to obtain a measure of the extraction efficiency of the cell extraction, direct (nonextracted) cell counts were performed on a few samples from the upper part of the core. Because 100× to 500× less sediment can be used for counting than for cell extraction, the minimum detection limit increases accordingly from ~10³ cells/cm³ (extracted samples) to ~5 × 10⁴ cells/cm³ (nonextracted samples).

During Expedition 329, 2 cm³ sediment plug samples for cell enumeration were taken using 3 cm³ tip-cut syringes, either directly at the catwalk or in the cold room, from 5 cm whole-round core. Only the 2 cm³ syringe samples were used for the onboard analysis. The 5 cm whole-round cores were stored at –80°C for additional shore-based analyses. The 2 cm³ sediment plug was transferred into a sterile 15 mL centrifuge tube containing 8 mL of 0.2 µm filtered 2.5% (w/v) NaCl solution with 2% (v/v) formalin as a fixative and thoroughly shaken to form a homogeneous suspension. The slurried formalin-fixed samples were subjected to cell detachment and cleaning steps using the following protocol:

1. Depending on the expected cell abundance, between 50 and 500 µL of the (1:5) slurry was used for the extraction. When sample volumes were <500 µL, 2.5% NaCl solution was added to adjust the sample volume to 500 µL. Then, 50 µL each of detergent mix (100 mM disodium EDTA dihydrate, 100 mM sodium pyrophosphate decahydrate, and 1 vol% Tween 80) and methanol were added. The sample was placed on a vortex shaker for 30 min to 1 h. In cases of extremely low cell abundances, the volumes of slurry and all reagents were increased.

2. The sample was underlain by a cushion of ~500 µL Nycodenz (50% w/v) and centrifuged at 3250× g for 20 to 40 min.

3. The supernatant was carefully removed by siphoning it off with a small syringe and retained. The pellet was then resuspended in NaCl solution to a total volume of 500 µL.

4. The resuspended sample was treated as described in Step 1. After the vortex mixing, the sample was subjected to ultrasonic treatment with an ultrasonic probe (Model UH-50, SMT Co. Ltd., Tokyo, Japan) at 20 W for 5 × 10 s with 20 s breaks. The samples were placed in a cold water bath to minimize heating.

5. The sample was underlain with Nycodenz (set to ~1.2 g/mL), centrifuged, and the second supernatant was removed as described and added to the first one.

6. A volume of 100 µL of 1% HF was added to the pooled supernatant and left for 20 min.

7. To remove HF from the supernatant, 120 µL of a solution containing 0.5 M each of CaCl₂ and sodium acetate was added. Visible precipitation of the highly insoluble CaF₂ occurred immediately. Alternatively, we added 100 µL of 1.5 M TBS solution to neutralize the solution pH without CaF₂ precipitation (see Morono et al., 2009).

8. After a vortex mix, the sample was underlain by a cushion of Nycodenz and centrifuged (10 min at 2500× g) in order to remove the CaF₂.

9. The sample was then prepared for cell enumeration. For direct cell enumeration using epifluorescence microscopy, the pooled supernatant was filtered on 0.2 µm black polycarbonate filters (Whatman), stained with SYBR Green I, and counted directly on a Zeiss Axiophot epifluorescence microscope using 63× or 100× magnification.

10. For samples that contained carbonate minerals, the slurry was first treated with 10× the sample volume of a carbonate dissolution mix containing 20 mL/L glacial acetic acid and 35 g/L sodium acetate. This solution has a high acidity but a moderate (4.6) pH, avoiding excess stress on the cells. The slurries were placed in 5 mL round-bottom culture tubes with ventilated caps in order to avoid buildup of CO₂. After all carbonate was dissolved, the samples were centrifuged and the supernatants removed and kept for cell enumeration. The pellet was washed twice with Tris-EDTA (TE) buffer and treated like a normal sample as described in Step 1.

Estimation of the blank and minimum detection limit (MDL) was very important because of the very low cell abundances. Samples were processed in batches of six plus one additional blank sample, in which sterile-filtered TE buffer replaced the sediment slurry. The blank was calculated as the average of all blanks processed during the expedition. The MDL was set to be the blank value plus three times the standard deviation of the blank.

For quality assurance and control, all glassware used for cell count was combusted at 500°C for at least 1 h to remove any remaining cells. Because of higher turnover, the filter towers were soaked in 5% sodium hypochloride solution for at least 15 min, rinsed with 18.2 MΩ water, and finally rinsed with 100% technical-grade ethanol and flamed with a blow torch. All reagents were filtered through 0.45 or 0.2 µm pore filters. An aliquot of every batch of every reagent was placed on a filter, stained, and checked for contamination using epifluorescence microscopy.

We also experimented with flow cytometry (FCM) for cell counting. This was the first use of FCM on the JOIDES Resolution. Because the use of this technique for enumeration of cells in sediments is still experimental, the FCM results and methodological developments are presented in a separate chapter as a specialty paper rather than in the site reports (see Morono et al., 2011). If FCM proves to be successful for onboard cell enumeration, it’s advantages will be

• High sample throughput and continuous cell enumeration,

• Better counting statistics because of a much larger numbers of cells being counted from individual samples, and

• Consistency of criteria for cell identification.

Consequently, the use of a high-throughput FCM cell count technique may ultimately result in higher data resolution and more consistent results than counts using epifluorescence microscopy.

Basalt

During Expedition 329, we enumerated cell abundance for basaltic rock samples without the cell extraction steps used for sediment samples. First, the surface of basalt cores was washed twice with 25 mL of 3% NaCl solution in a sterile plastic bag. After the washing step, the rock surface was briefly flamed with a propane torch. The flamed rock was cracked using a flame-sterilized hammer and chisel, and small pieces from the interior were powdered using a flame-sterilized mortar and pestle. The powdered basalt (1–2 mL) sample was fixed with 5–10 mL of TBS solution and then fixed with 4% paraformaldehyde at 4°C overnight. The fixed basalt suspension was subjected to cell counting using the following protocol:

1. After vigorous stirring of the parformaldehyde-fixed slurry, 100 µL of the suspension was dispensed in 10 mL of filtered 1× phosphate-buffered saline (PBS) buffer (pH = 7.4)

2. After vigorous stirring, 1 mL of the second suspension (made in Step 1) was diluted again in 9 mL of filtered 1× PBS buffer and sonicated for 30 s.

3. The sonicated suspension was filtered through a 25 mm diameter, 0.2 µm pore size black polycarbonate filter. A cellulose filter (25 mm diameter; 0.45 µm pore size) was placed beneath the polycarbonate filter as support.

4. Cells on the polycarbonate filter were stained with filtered 1× SYBR Green I in filtered TE buffer for 5 min.

5. The stained filter was washed twice with 5 mL of filtered TBS solution.

6. The washed filter was mounted on a slide with one drop of immersion oil and covered with a coverslip.

7. The SYBR Green I–stained cells were directly observed by using a Zeiss Axiophot epifluorescence microscope with a band-path filter slit (excitation at 470 nm, fluorescence above 515 nm) at 63× magnification.

8. On each filter, 300 microscopic fields of views were observed, and cell-shaped forms that produced bright green fluorescence were enumerated as cells.

The MDL was estimated by counting blank filter samples that were treated with 100 µL of filtered TBS solution with 4% paraformaldehyde as described above. Cell number in the blank filters was calculated as the average of all blanks processed at each site. The MDL was set to be the average blank value plus three times the standard deviation of the blanks. For negative control samples (i.e., cell free), crushed basalt pieces were combusted at 500°C for 3 h before preparation as described above.

Fluorescent in situ hybridization

Fluorescent in situ hybridization (FISH) is a powerful molecular ecological technique, enabling us to detect metabolically active microbial cells using a specific fluorescent-labeled oligonucleotide probe for microscopic analysis. The probe will hybridize with 16S rRNA (or mRNA of the targeted functional gene) in the metabolically active cells. Thus, appropriate sample processing (i.e., cell fixation) using freshly collected samples is necessary. During Expedition 329, we prepared fixed slurries for shore-based FISH analyses. Using sterile tip-cut syringes, 10 cm³ plug samples were collected from the innermost part of whole-round cores and placed in four times the sample volume of paraformaldehyde solution (4% paraformaldehyde in PBS buffer). The slurry was vigorously shaken until no visible clumps were observed. It was then incubated for several hours at 4°C. Following incubation, the slurry was pelleted in a centrifuge at 2500× g for 10 min, washed twice with cold PBS buffer, and stored in 10 mL of 1:1 (v/v) PBS:ethanol solution at –20°C.

Enumeration of viruslike particles

During Expedition 329, the number of virus-like particles (VLPs) in sediment was evaluated on board by direct counting using epifluorescence microscope. For each sample analysis, 1 cm³ of a sediment sample was added to a mixture of 3.5 mL of 0.02 µm filter-sterilized 18.2 MΩ H₂O and 1 mL sodium pyrophosphate (55 mM; 0.02 µm filtered). The slurry was shaken until it formed a homogeneous suspension. Samples were sonicated for 3 min with 30 s breaks after each minute and vortex mixed at maximum speed during the breaks. After centrifugation for 10 min at 3250× g, the supernatant was removed and filtered through a 0.45 µm pore size polycarbonate filter. A volume of 4 mL of TE buffer (10 mM) was added, followed by resuspension of the sediment pellet and centrifugation as described before. The second supernatant was also filtered through a 0.45 µm pore size filter. Both the extracted solutions were pooled and processed immediately or temporarily stored at 4°C until further use. Volumes of 1 to 4 mL of the extracts were filtered onto 0.02 µm pore size membrane filters (Anodisc 13, Whatman) and washed with TE buffer (Tris 10 mM; EDTA 10 mM; pH 7.7). Filters were stained with SYBR Gold (40 µL; 2.5×; Invitrogen) and stained for 1 h in the dark. Stained filters were mounted to microscope slides with 5 µL of mounting solution. VLPs were directly counted in randomly chosen fields using an epifluorescence microscope (Zeiss Axioplan 2 imaging microscope) with a blue filter set.

Preparation of fixed samples for combined microbiological and mineralogical analyses

Mineralogical, textural, and isotopic studies will be used to evaluate the extent of microbial mineral alteration (Boyd and Scott, 2001; Rouxel et al., 2003). For example, the habitability of aged subseafloor basalt is a vital factor to accurately elucidate the potential extent of Earth’s subsurface biosphere (Bach and Edwards, 2003). One of the primary potential electron donors in basalt is Fe[II]. If oxidation of Fe[II] is microbially mediated, a likely result is coating of the cell periphery with Fe[III] oxyhydroxides. Understanding of microbe-mineral relationships requires characterization of (1) phylogenetic affiliations of microbial populations that mediate Fe[II] oxidation and (2) crystallographic features of nanometer-scale redox features. Especially, crystallographic evolution of oxidation products over 10–120 Ma needs to be clearly defined in order to discern the modern and past occurrence of Fe[II] oxidizers and to constrain dissolution kinetics of rock-forming minerals.

Scanning transmission X-ray microscopy analysis

Scanning transmission X-ray microscopy analysis is a powerful tool for characterization of organic materials (DNA, extracellular polysaccharide [EPS], etc.) because this method can provide nanometer-scale information on light element (e.g., C and N) distributions in environmental samples (Tsuji et al., 2010). Consequently, it is very helpful for the study of biomineralization. Because dehydration can cause serious changes in some molecular species, sampling was conducted immediately to prevent samples from drying. Samples for scanning transmission X-ray microscopy analysis were soaked in filter-sterilized artificial seawater stored at 4°C.

Electron energy–loss spectroscopy and transmission electron microscopy analysis

Biogeochemical mineral transformations associated with microbial Fe reduction-oxidation play significant roles in controlling redox balance and carbon cycling (Nealson and Saffarini, 1994), as well as in mineral dissolution and neoformation (Kim et al., 2004). Combined use of electron energy–loss spectroscopy and transmission electron microscopy is capable of measuring mineral transformations mediated by microbial activities at micro- and nano-scales (Buatier et al., 2004). From the measured electron energy–loss spectroscopy spectra, the energy-loss ratio of Fe-L₃ to Fe-L₂ lines will be determined to estimate the Fe oxidation state using the universal curve as a function of the integral ratio of L₃/L₂ versus ferric Fe concentration as determined by van Aken et al. (1998). During Expedition 329, we stored sediment and basalt samples immediately in the freezer at –80°C for shore-based electron energy–loss spectroscopy-transmission electron microscopy investigations.

Micro X-ray absorption fine structure coupled with catalyzed reporter deposition-fluorescence in situ hybridization analysis

Micro X-ray absorption fine structure (µ-XAFS) and bulk XAFS will be used to directly characterize minerals (mineral phase, size, crystallinity, etc.) in basalt samples. Coupling of µ-XAFS with catalyzed reporter deposition-fluorescence in situ hybridization analysis is a powerful approach to study microbial associations with minerals. For this approach, basalt samples were fixed with TBS buffer (pH 7.2–7.6) containing 3.7% formaldehyde. After the fixation, samples were washed three times with 1:1 (v/v) TBS buffer:ethanol and stored at 4°C. For bulk XAFS analysis, the crushed basalt samples were immediately put in plastic bottles and stored in vacuum-sealed bags at 4°C without further treatment.

Cultivation experiments

Cultivation-dependent studies will provide particularly useful information on physiology, potential biogeochemical function, diversity, and habitat range of low-activity subseafloor microorganisms. During Expedition 329, onboard microbiologists initiated cultivations with a wide range of culture media, targeting various physiological types of microbes (Table T8). Generally, the inoculum slurries used for onboard cultivation experiments were individually prepared by individual shipboard scientists using subsamples taken from whole-round cores with cut-off syringes of various sizes. In addition, some sediment samples or inoculum slurries, as well as the sample-inoculated test tube cultures, were stored at 4°C or appropriate conditions for shore-based cultivation experiments.

Inoculum and culture media

To inoculate sediment and basalt samples into the culture media, we prepared inoculum slurries with 2–10 cm³ subcore samples mixed with either sterileculture media or artificial seawater (Suess et al., 2004). Some inoculum slurries were prepared anaerobically with continuous flushing of the headspace with N₂. The test tubes or serum bottles for anaerobic media were sealed with thick butyl rubber stoppers. No reducing reagents such as Na₂S or titanium solution were used for the inoculum slurry preparation, except for samples targeting methanogens and other types of strictly anaerobic microbes. We attempted to cultivate various types of microorganisms with various combinations of electron donors and acceptors as well as various carbon and nitrogen sources under multiple incubation conditions (e.g., psychrophilic and mesophilic microbes such as aerobic or microaerobic heterotrophs, fermenters, organotrophs, autotrophic and heterotrophic nitrate/nitrite reducers, sulfate reducers, sulfur oxidizers, autotrophic or mixotrophic ammonium oxidizers, nitrite oxidizers, aerobic and anaerobic hydrogenotrophic autotrophs, iron oxidizers/reducers, and manganese oxidizers/reducers) (Table T8). Cultivations were performed with both semisolid-plate and liquid-culture media prepared by individual shipboard microbiologists. An autoclave or oven in the Microbiology Laboratory on the JOIDES Resolution was used for sterilizing media and equipment. Nitrogen, N₂:CO₂ (80:20 [v/v]) and/or H₂:CO₂ (90:10 [v/v]) gas cylinders were used to prepare headspace gas for anaerobic media. Inoculations were performed using sterile needles and syringes for liquid media or a spreader for solid plates. Because in situ temperature in sediment and basalt of the South Pacific Gyre is low (a few degrees Celsius), most cultures were incubated at ~10°C.

Most probable number cultivation method

To evaluate most probable numbers for cultivable microbial populations from the South Pacific Gyre sediments, we performed parallel cultivation assays using a serial dilution method of liquid media in 96-well microtiter plates or test tubes. The most probable number media will continue to be incubated in shore-based laboratories, with the goal to quantify specific physiological types of microorganisms and subsequently to isolate pure cultures.

Experiments with stable and radioactive isotopes

A major objective of Expedition 329 is to elucidate the energy and nutrient substrates that may sustain microbial populations in subseafloor environments of the South Pacific Gyre. For example, the list of potential electron donors includes photosynthetically derived organic matter, reduced metal, and hydrogen produced in situ by radiolysis (D’Hondt et al., 2009). Thus, experiments were performed to detect specific aerobic respiration pathways and to track potential routes of C-, N-, and P-fixation and their turnover. During Expedition 329, freshly collected samples were fixed for (1) determination of tracer turnover and (2) further processing in shore-based laboratories. Samples were also taken to isolate cell material for single-cell approaches, such as nano-scale secondary ion mass spectrometry (NanoSIMS) analysis combined with various types of in situ hybridization techniques.

The use of isotope-labeled compounds is a critical and sensitive tool for the analysis of microbial activities in subseafloor environments. In subseafloor sediment, microbial processes, such as oxic respiration, nutrient uptake, oxidation of reduced substrates, and so on, take place at rates that are subnanomolar per cubic centimeters per day or even subpicomolar per cubic centimeters per day, which is more than six orders of magnitude lower than microbial activities on Earth’s surface habitats (D’Hondt et al., 2002). Interstitial water gradients of dissolved compounds, such as oxygen, nitrate, or dissolved inorganic carbon, and sediment physical properties can be used to calculate in situ rates of microbial processes. However, such calculations generally do not provide highly specific insight into physiological pathways of substrate turnover. Detailed studies of these pathways typically require incubation studies with chemical tracers. It is often not possible to detect the small concentration changes in dissolved chemical concentrations during incubations of experimentally realistic durations (days to months). Using radiotracers, however, rate determinations in incubated sediment samples may become >10,000 times more sensitive than when measuring changes in dissolved chemicals in incubated samples. However, microbial metabolic processes in subseafloor sediments are typically so slow that even radiotracer methods operate at the limit of their detectability when studying million year–old sediments. Consequently, the amounts of radioactivity applied must be higher than normally used in studies of near-surface sediments.

Pulse-chase experiments constitute a second application of isotope-labeling methods. In these experiments, a radioactive or stable isotope–labeled compound is added to a sample and the isotope’s incorporation into cells or specific biomolecules is tracked. Also known as stable isotope probing or radioisotope probing, these methods have the potential to link microbial activities with specific organisms or communities.

It is critical for all experimental activity measurements that the sediment or rock samples remain as intact as possible; otherwise data become unrealistic. This requires

• Fast handling of cores on deck to avoid warming of cold sediments. In cold deep-sea sediment, microbes may be sensitive to warming and could be killed at temperatures above 10°–15°C. In warmer sediment, this problem is not critical;

• Aseptic subsampling by sterile and anaerobic techniques from whole-round cores in which the sediment or rock has not been exposed to atmospheric oxygen by splitting of the core; and

• Starting radiotracer experiments as soon as possible after coring because the sample’s chemistry and microbiology gradually change once the sediment has been brought up on deck. Experiments started after the cruise are still very useful for some categories of study (e.g., factors regulating process rates), but the absolute rates become less trustworthy with time.

In general, procedures for rate measurements with radiotracers involve subsampling of sediment or basalt into glass vials or other small sterile containers. These still-unlabeled containers are then incubated at in situ temperature in the Isotope Isolation Van for a day until thermal equilibrium is reestablished. Only then is radiotracer or stable isotope tracer injected into the sample (few microliters) and incubation continues in the Isotope Isolation Van for up to a month or more; if necessary, cooled samples still incubating may be shipped back to the shore-based laboratories by airfreight. Incubations are stopped at different times in order to check whether process rates are constant over time. The terminated samples are fixed in a manner that kills the microorganisms and preserves the isotope-labeled substrate and product. In this state, the samples are stable and can be handled without risk of contamination or risk of altering the incubation results. At the end of the expedition, these fixed (and frozen) samples were transported, either cooled or frozen, to shore-based laboratories for further processing.

Contamination control

Laboratory contamination was controlled through standard radiochemical safety procedures, including monitoring by wipe tests before and after sample handling. Trays and absorbent paper were used. Care was taken to restrict radionuclide use to the back end of the Isotope Isolation Van. Access to the Isotope Isolation Van was restricted to trained personnel. Laboratory outerwear (lab coats and gloves) and footwear (or shoe covers) remained in the Isotope Isolation Van. Wipe tests were performed on all material prior to their removal from the Isotope Isolation Van.

Carbon isotope–labeled compounds pose the greatest danger for long-term contamination of shipboard surfaces because of the long half-life of ¹⁴C (5780 y). Bicarbonate may react with water to form volatile CO₂ compounds, but all of the procedures use seawater and alkaline-buffered solution, so degassing of radioactive ¹⁴CO₂ should be minimal (expected specific activity of the experiments is 25 MBq/mM). ¹⁴C-acetate at various pH conditions encountered in seawater exists primarily as soluble acetate anion and should not form a volatile compound. The small amounts of CO₂ resulting from oxidation reactions form bicarbonate and are trapped in the alkaline seawater and buffered solutions. Likewise, the small amounts of CO₂ resulting from oxidation of ¹⁴C-labeled leucine to form bicarbonate should remain in solution. ³³P-labeled compounds have a short half-life (25.3 days) and have no volatile forms. ³⁵S-sulfate has a relatively short half-life (87.4 days) and is not volatile. Nevertheless, opening of all radioactive tracer stock solution occurred in the hood in the Isotope Isolation Van on the JOIDES Resolution.

Although stable isotope tracers are not considered to pose any health hazard, they are potential sources of contamination for geological and oceanographic materials. Therefore, during Expedition 329 all activities with open stable isotope tracers were treated in the same manner as the radioactive tracers. Use of stable isotopes was restricted to the Isotope Isolation Van. Users also followed the same protocols concerning laboratory apparel and shoes as for radioisotope work.

Combined radioactive and stable isotope tracer experiments

Slurry samples of sediment from the South Pacific Gyre subseafloor environments were prepared for radioactive and stable isotope–tracer experiments. Sediment whole-round cores selected for incubation were stored either in the cold room of the Microbiology Laboratory or core refrigerator on the Hold Deck of the JOIDES Resolution before processing. Sediment slurry was prepared from 10 cm whole-round cores. For all steps of slurry preparation, sediment was processed in the laminar flow hood and quickly returned to the cold room. For each sediment interval, 1 L of slurry was prepared by adding 200 cm³ of sediment to 800 mL of sterile-filtered (0.2 µm) surface seawater from Site U1368. The exteriors of the sediment sections were trimmed, and the interior pieces were added to chilled (9°C) and filtered seawater. Slurries were stirred constantly for 12 h. Subsequently, 40 mL of slurry was transferred to 51 mL serum vials, closed with butyl rubber stoppers, and sealed with aluminum caps. Two replicate serum vials were prepared for each sediment sample. After tracer addition to both serum vials, one was frozen immediately as the control sample. The others were incubated at 10°C. Incubations were planned to continue beyond the duration of the expedition. Samples were prepared for refrigerated shipment to the home institutes where further incubation, processing, and analyses are to follow.

Incubation experiments were performed with radioisotope-labeled compounds, stable isotope–labeled compounds, or a combination of both. The labeled compounds were added in the hood in a compartment of the Isotope Isolation Van that was cooled to 9°–10°C. In general, autotrophy radioactive experiments were performed with ¹⁴C-labeled bicarbonate and heterotrophy experiments were performed with ¹⁴C-labeled acetate. To evaluate nitrification, ¹⁵N-labeled ammonium was added. For nitrogen fixation, ¹⁵N–double labeled molecular nitrogen (N₂) gas (1 mL) was injected into sealed containers. Samples from these experiments for isolation of cell material were taken for single-cell approaches such as halogenated in situ hybridization–secondary ion mass spectrometry analysis. Phosphorus uptake was evaluated by use of ³³P-labeled phosphate or ³³P-labeled ATP. Incubations with ¹⁴C-labeled compounds (ATP, leucine, and thymidine) for cell viability were initiated on board. Incubations with ¹⁸O-labeled water were used to quantify enzyme mediated isotopic exchange with phosphate. The incubation experiment on nitrate reduction coupled to autotrophic C uptake was performed by adding ¹⁵N-nitrate and ¹³C-bicarbonate. For other stable isotope incubation experiments (i.e., ¹³C- and/or ¹⁵N-labeled substrates) with sediment and basalt samples, see “Single-cell analyses of carbon and nitrogen assimilation rates of subseafloor autotrophic and heterotrophic microbial community” and “Detection of metabolically active microorganisms in basaltic rocks,” respectively. Tracer amounts, specific activities, and isotope labeling percentages are listed in Table T9. Radiotracer amounts for the cell viability assays are listed in Table T10.

Stable isotope probing of subseafloor microbes with ¹³C-labeled tracers

With the ¹³C-labeled tracer approach, the objective is to cultivate marine subsurface microbes capable of utilizing benzoate, methanol, or acetate as a carbon source. The primary interest is to test for the presence of and identify aerobic and anaerobic benzoate degraders and methylotrophs. To encourage the growth of these organisms, aerobic and anaerobic microbial communities were incubated with low-nutrient conditions at close to in situ temperature. If these organisms are present and physiologically active (or even dormant in the case of anaerobes) in the South Pacific Gyre sediment, they are likely to be activated on the above-mentioned ¹³C-labeled carbon sources. Even very limited growth will yield partially or completely (minimum five generations) ¹³C-labeled DNA. This “heavy” DNA can be extracted from the microcosm, purified, and separated from nonlabeled, “light” ¹³C-DNA. The experiment will be conducted in time series, in which subsamples will be collected at selected intervals to monitor population change with time.

Crimp-sealed headspace vials (20 mL) were used to prepare the incubation samples. Sediment cores (4–5 cm³, except for Site U1365 where 3–4 cm³ was used) were subsampled from whole-round cores with cut-off syringes under sterile conditions and were immediately placed in the headspace vials in a stream of nitrogen and crimp sealed with a butyl rubber septums. All vials were baked at 450°C for minimum of 6 h to ensure complete combustion of residual organic carbon that might interfere with the postcruise analyses. Media composed of filter-sterilized (0.22 µm pore size membrane filter) surface seawater amended with specific ¹³C-stable isotope tracers were prepared for the experiment. Aliquots of these media were introduced to the respective incubation samples through the septum to ensure 50% inoculum. Overpressure in the samples was removed with a sterile needle filter that was briefly inserted through the septum. For anaerobic incubations, sample handling was performed in under a stream of N₂ gas and media were additionally amended with FeCl₂ and Na₂S. Large inocula were used in all incubations to increase the concentration of viable cells. At the end of the expedition, all incubation samples were transferred to an onshore laboratory where they will be incubated in similar conditions for as long as 6 months for aerobic cultivations or 12 months for anaerobic cultivations. These shore-based incubation experiments will be subsampled periodically to monitor population change over time. In the case of aerobic cultivations, subsampling will be performed from the master slurry, whereas anaerobic cultures were quadrupled to facilitate study with four time points. ¹³C-labeled heavy DNA and nonlabeled light ¹³C-DNA will be extracted and separated by CsCl gradient ultracentrifugation.

Single-cell analyses of carbon and nitrogen assimilation rates of subseafloor autotrophic and heterotrophic microbial community

Previous studies of subseafloor microbial communities in continental margin sediments demonstrated that heterotrophs that utilize organic matter derived from photosynthesis are the predominant microbes in organic-rich sediments. Conversely, the sedimentary habitat in the South Pacific Gyre contains very little organic matter because of (1) the low primary photosynthetic production in the water column and (2) the extremely low sedimentation rates, which leave organic matter exposed at the seafloor for tens of thousands to hundreds of thousands of years. A list of potential electron donors ranged from the standard photosynthetically derived organic matter inputs such as reduced carbon and ammonium to the proposed radiolytic in situ production of hydrogen (D’Hondt et al., 2009). The rates of oxygen consumption through any of these processes may be so low as to be disguised by diffusion. Thus, experiments were performed to detect specific aerobic respiration pathways and to track potential routes of carbon and nitrogen fixation and turnover (also see “Experiments with stable and radioactive isotopes”). In using sensitive isotope tracer incubation methods on selected samples, such experimentation required immediate and proper sampling as well as access to the shipboard isotope facilities. Furthermore, we have interests in experimentally elucidating types of microbial metabolism in these ultra-oligotrophic sediments in the South Pacific Gyre.

To identify autotrophic and heterotrophic microbial populations, as well as their potential substrate uptake rates, sediment samples were incubated with various stable isotope–labeled substrates, including ¹³C-glucose, ¹³C-acetate, ¹³C-pyruvate, ¹³C-bicarbonate, ¹³C-¹⁵N-amino acids mix, ¹³C-methane, and ¹⁵N-ammonia (Table T11). Given oxygen concentrations in interstitial water during Expedition 329, we set oxygen concentration in headspace of vials at ~4% (v/v) by adding filter-sterilized air. Individual samples of sediment (15 cm³) were collected with tip-cut 30 mL syringes, placed in 50 cm³ sterile glass vials sealed with rubber stoppers and screw caps, and stored at 10°C. The labeled substrates were injected (15 µM of ¹³C-labeled substrates and 1.5 µM of ¹⁵N-labeled ammonium, dissolved in 50–100 µL of sterile water) into each subcore sample and incubated at 10°C in a refrigerator in the Isotope Isolation Van and onshore after the expedition. All reagents and gas components, including air used for sample preparation, were filtered through 0.45 µm polycarbonate membranes. At each of four given time points (~2 weeks, 1 month, 6 months, and 12 months after starting incubation), vials were opened and sediment samples were fixed with 4% paraformaldehyde in PBS solution (or 2% formalin) and frozen for onshore single-cell analysis by NanoSIMS coupled with molecular ecological techniques.

Detection of metabolically active microorganisms in basaltic rocks

To detect the presence of metabolically active microorganisms in the basaltic basement and sediment/basalt interface in the South Pacific Gyre, the incorporation of stable isotope–labeled tracers (¹⁵N-labeled nitrate and ¹³C-labeled acetate and bicarbonate) into cells was investigated by incubation under microaerobic conditions, to be followed by postexpedition analysis with NanoSIMS. All experiments were duplicated, and a negative control was prepared with autoclaved samples. The experimental protocol was as follows:

1. Crushed basalt pieces (3–5 mL) and sterilized bottom seawater (20 mL) were placed in 67 mL wide-mouth glass vials for incubation under microaerobic conditions (~4% O₂). Headspace gas was replaced with N₂, and 11 mL of filtered air (0.2 µm pore diameter) was injected in the vial.

2. ¹⁵N-labeled sodium nitrate solution (100 µL; 20 mM) was added to the vial. The final concentration of ¹⁵N-nitrate was 100 µM.

3. For incubation with ¹³C-labeled sodium acetate or ¹³C-labeled sodium bicarbonate as the potential carbon substrate, 100 µL each of 20 mM substrate-stock solution was added.

4. At given time points (~4 weeks, 6 months, and 2 y after starting incubation), vials will be opened and basalt pieces will be fixed by placement in 4% paraformaldehyde in TBS solution or by freezing for subsequent molecular analyses coupled with NanoSIMS.

Molecular analyses of subseafloor microbial communities

Culture-independent molecular ecological techniques have shown that subseafloor sediment harbors phylogenetically diverse microbial components that are generally distinct from isolates with known physiology. Previous studies of organic-rich sediments from the continental margins (e.g., offshore Peru, ODP Leg 201) suggest that geological and geochemical characteristics of subseafloor microbial habitats (such as lithology, physical properties, and interstitial water chemistry) strongly impact microbial diversity and the community structure. However, microbial communities that inhabit organic-poor open-ocean sediment, which covers near 50% of the seafloor, remain uncharacterized.

During Expedition 329, sediment and basalt samples were collected for a variety of molecular ecological analyses. All of these samples for culture-independent molecular analyses were stored at –80°C after recovery. Bulk environmental DNA and RNA will be directly extracted from contaminant-free or relatively uncontaminated material. If the extracted DNA or RNA concentration is not high enough for molecular analyses, the extracted DNA may be further amplified by multiple displacement amplification with phi29 polymerase (e.g., Forschner et al., 2008).

Spatial distribution of microbial community composition and the structure will be intensively studied by statistical analysis of polymerase chain reaction (PCR)-amplified fragments of 16S rRNA gene sequences (e.g., Inagaki et al., 2006). A powerful new approach using 454 Life Sciences high-throughput sequencing technologies will allow us to determine (1) microbial diversity, (2) community structure, and (3) phylogenetic affinities based on large numbers of 16S rRNA gene-tagged PCR fragments (e.g., Sogin et al., 2006). RNA sequencing studies, such as 16S rRNA and mRNA, will document the composition and functional taxonomy of the actively growing and metabolizing fraction of the microbial community. These high-throughput molecular studies will quantify bacterial and archaeal diversity richness and evenness in this little explored subseafloor environment and enable detection of rare members that cannot be sampled using conventional capillary sequencing approaches.

A functional gene survey for key metabolic pathways is a powerful molecular ecological approach for identification of the potential function of biogeochemically significant microbial metabolisms. Because the South Pacific Gyre subsurface is very different from previously explored subseafloor microbial habitats in terms of nutrient and energy constraints, physiological characteristics of its microbial communities are also expected to be different. For example, its most deeply buried microbial communities may primarily rely on inorganic nutrient and energy sources (e.g., radiolytic H₂). Given this hypothesis, sediment of the South Pacific Gyre subsurface may harbor (micro)aerobic chemolithoautotrophic or chemoorganotrophic mesophiles, including hydrogenotrophic aerobes and nitrate or nitrite reducers. Using DNA from the South Pacific Gyre, we will attempt PCR detection and quantification (if possible) of some specific 16S rRNA and functional genes, such as carbon fixation pathways and key genes for aerobic respiration in shallow to deep sediment and the underlying basalt.

Top of page   |    Previous   |    Next