Isolations from seafloor sediment were performed as described above prior to attempting subseafloor enrichments. Seafloor sediment was plated in duplicate, and one plate of each set was incubated at either 2° or 10°C. Colonies appeared within 3 days at 10°C but required up to 17 days at 2°C, and similar morphologies were seen on media incubated at each temperature. Three colony morphologies were seen, and samples of these were restreaked from plates at 2°C to homogeneity to yield five pure cultures.
Analysis of 16S rRNA gene sequences from these five cultures revealed five distinct sequences (Fig. F1). Isolate 1230sf1, which had formed large colonies within the first 3 days of incubation at 2°C during the initial plating, was most related to the deep-sea benthic bacterium Photobacterium profundum. Isolate 1230sf2 was most closely related to Halomonas boliviensis. Isolates 1230sf3, 1230sf4, and 1230sf5, which had very similar colony morphologies, were all closely related to previously cultivated deep-sea strains of Shewanella benthica (Fig. F1).
The abilities of these five isolates to grow at different temperatures and produce extracellular enzymes is summarized in Table T1. All isolates grew well between 2° and 20°C. Only isolate 1230sf2, which had an optimal growth temperature between 20° and 35°C, was able to grow at 37°C. In addition to its high 16S rDNA similarity to the two other described isolates of P. profundum, isolate 1230sf1 also had similar growth characteristics with a doubling time of 11 hr at 2°C and 2.5 hr. at 15°C (Nogi et al., 1998). Isolate 1230sf1 also had a genome size and piezotolerance (tolerant to 40 MPa, the highest pressure tested) (F. Lauro, pers. comm., 2003) that is consistent with its identification as a P. profundum strain (Bidle and Bartlett 1999; Nogi et al., 1998). Similar to isolate 1230sf1, isolates 1230sf3, 1230sf4, and 1230sf5 also did not grow at temperatures above 30°C and had optimal growth between 18° and 25°C. All isolates can therefore be characterized as psychrophilic according to the definition of Neidhardt et al. (1990). All isolates grew slowly anaerobically at low temperatures, taking 6 months to form isolated colonies on anaerobic plates at 10°C (Table T1).
All isolates, except for 1230sf2, produced extracellular degradative enzymes that were active on the screening media at temperatures between 2° and 18°C (Table T1). The Shewanella related species produced proteases and esterases that continued to be active at temperatures up to 30°C. Chitinase activity was not seen for any of the isolates tested. The frequency of protease and esterase production may reflect the substrate availability of deep seafloor areas where proteins and lipids are present as substrates and possible nutrients (Boetius and Lochte, 1994; Luna et al., 2004).
Based on results from the seafloor isolation experiments, we attempted to cultivate similar microbes farther down the sediment column at 0.67 mbsf (Section 201-1230C-1H-1; as reported, the contamination in this sediment was nonexistent or low) (D'Hondt, Jørgensen, Miller, et al., 2003). DMB media were inoculated with sediment and incubated aerobically at 10°C. After 3 days of incubation, the culture was examined microscopically. Numerous cell morphologies were observed, with the most striking being numerous spirochetes and spirilla-like cells. Plating of this culture revealed a single colony morphology, so molecular methods were used to explore the additional population members. Genomic DNA was extracted from the culture, and the intergenic spacer regions between rRNA genes were amplified using both bacterial and archaeal primers. RISA showed a diverse bacterial population in this sample (Fig. F2). Surprisingly, archaeal primers amplified a single band at 500 bp from this enrichment. Purification and sequencing of this archaeal PCR product of RISA fingerprinting showed that it was from a group classified as uncultivated benthic Crenarchaea from low-temperature environments (Fig. F3) (Bowman and McCuaig, 2003; Vetriani et al., 1998). Additional research showed that this archaeal fingerprint was only detected in the enrichment culture until day 10, after which it could not be detected.
The diverse bacterial populations indicated by these RISA fingerprints were also examined through the construction of 16S rDNA libraries (Fig. F4). Following enrichment for 3 days, the majority of ARDRA- determined ribotypes in the16S rDNA library were related to Vibrio spp. and Halomonas spp., with the Halomonas spp. being represented by three separate ribotypes after digestion with RsaI. This high representation of up to 70% of the ribotypes by possible Halomonas spp., which are rod-shaped cells, suggests that the cells observed microscopically with the distinctive morphologies of spirochete and spirilla-like cells were either not lysed or their rRNA genes were poorly amplified by general bacterial primers.
Incubation of these cultures was continued and observed microscopically after 15 days. The 15-day cultures appeared similar to the 3-day cultures both microscopically and in the banding patterns of the bacterial RISA (Fig. F2). Because many of the members of the population, such as spirilla, spirochetes, Vibrio spp., and Halomonas spp., are known to be facultative anaerobes, the culture was cycled through aerobic and anaerobic conditions, at 2°C, in an attempt to enrich for facultative anaerobes. The final culture examined had been incubated anaerobically for a month, making the total liquid cultivation time 7 months. Microscopic observation of the population showed a large number of rod-shaped cells, and analyses by both RISA fingerprinting (Fig. F2) and 16S rDNA libraries (Fig. F4) showed that it was populated by Vibrio spp. and a Shewanella sp.
Further subcultivation of this mixed population was performed on agar media and yielded isolated colonies similar to the dominant colony morphology observed initially, designated 12301H1, related to Vibrio diazotrophicus (Fig. F1). This isolate had an optimal growth temperature at 37°C, grew well at colder temperatures, and produced amylases for starch hydrolysis at all temperatures tested (Table T1). The 16S rDNA sequences of this Vibrio-like sp. and the Vibrio-like sp. previously detected in the mixed culture were identical. Other colonies contained mixed cell types that could not be subcultured to homogeneity and were not identified. Additional undescribed isolates have been obtained through similar methods of aerobic plating of samples from 12, 30, and 100 mbsf (data not shown).
Because spirochetes have been previously reported in enrichment cultures from 4.15 mbsf at the Nankai Trough (ODP Site 1173) (Toffin et al., 2004), the spirochetes observed after the 3-day incubation were particularly interesting. We attempted to isolate these by enriching for their growth through traditional methods such as adding antibiotics to the media, incubating anaerobically, serially diluting the sample to decrease the number of competitors, incubating in soft agar with gradient conditions, and filtering cultures. We also used a more nontraditional method of flow cytometry to separate individual cells from the population. However, none of these methods increased the number of spirilla or spirochete-like morphologies seen in enrichment.
Several enrichments were designed to cultivate methanogens from samples taken throughout the sediment column at Site 1230 (1, 30, 102, and 258 mbsf) (D'Hondt, Jørgensen, Miller, et al., 2003). Four different potential carbon sources were added individually (H2/CO2, methanol, acetate, and trimethylamine) to anaerobic marine salts media, DGH. After sediment addition, the enrichments were incubated anaerobically at 4°C. Because growth of a psychrophilic methanogen could be slow and might not yield turbidity that would be detectable in the presence of the sediment, the headspace was analyzed for the presence of methane as an indication of methanogenic activity. Methane levels were measured by gas chromatography over a period of 2 yr (Table T2); however, no increase above background levels was found.
In situ levels of methane in the sediment were determined on board the ship (D'Hondt, Jørgensen, Miller, et al., 2003) and would have yielded an 8 to 10-ppm background level in our experiments after methane was desorbed from the sediment. We also confirmed these in situ background levels in selected enrichments by adding 1 mL of 3-M sodium hydroxide to 10 mL of enrichment culture to release the methane absorbed on the sediment (D'Hondt, Jørgensen, Miller, et al., 2003). In these tests, background methane levels from sediment added to each bottle equaled ~7 ppm, which is in reasonable agreement with shipboard measurements considering the loss of some methane during storage. In this cultivation experiment, growth of methanogens could have been detected by a signal above background (even if some of the gas had absorbed onto the sediment used as inoculum). As a control to determine whether the media could support growth of methanogens if they had been present in the samples, Methanosarcina acetivorans cells were inoculated into the media and incubated. The culture became turbid within 2 days and produced gas, demonstrating that the medium could support the growth of a known methanogen.
In addition to examining the cultures for methane production, these anaerobic enrichments were examined under phase contrast microscopy after 1 yr of incubation, and extracted DNA was analyzed by PCR amplification using archaeal primers for both the 16S rDNA and RISA. No evidence for methanogens was found by any of these methods. In an additional experiment, we attempted to detect a possible extreme minority population of methanogens by using the GenomiPhi kit (Amersham), which amplifies total DNA nonspecifically, to increase the overall amount of extracted DNA. No archaeal RISA signatures were detected even after PCR amplification of the GenomiPhi-treated DNA extracted from the methanogen enrichments. During the course of these enrichments, the cultivation of a deep-sea methanogen, Methanoculleus submarinus, was reported from Nankai Trough sediment (Mikucki et al., 2003). At this time, we replicated the conditions under which M. submarinus was isolated using sediment from 1.02, 30.4, 65.9, and 102.3 mbsf because our original enrichments would not have supported growth of this organism. In these new enrichments, incubated at 37°C, no evidence for methanogenic cells was found, although bacterial growth occurred in cultures from 30.4 and 102.3 mbsf.