Enrichment cultures were set up on various carbon and energy sources as shown in Table T3. A total of 461 basalt enrichment cultures and 25 sediment cultures were started on board, and after ~1 month, growth was observed in 366 of the basalt enrichments and 17 of the sediment cultures (Table T3).
Methane was produced in 12 out of 25 basalt enrichment cultures with a methanogenic medium containing H2 + CO2 as energy and carbon sources (Table T4). No methane was detected in sediment enrichment cultures or in basalt enrichment cultures to which acetate or trimethylamine was added as energy and carbon sources. Reduction of ferric iron to ferrous iron was observed in four of the 78 media designed for IRB and inoculated with basalt; all of these were grown on the Fe-TSB medium. Iron reduction was not detected in IRB media inoculated with sediment (Table T4). Sulfide production was not observed in any of the enrichment cultures set up for sulfate reducing bacteria, either with basalt or with sediment as inoculum (Table T4).
The DNA band profiles of bacterial populations from native samples of basalt, sediment, and seawater differed (Fig. F1). Figure F1 shows a representative selection of samples of basalt, sediment, and seawater from different sites and depths to illustrate similarities and differences in the bacterial community structures. Results from sequencing analysis also confirmed that the dominant bacterial population in basalt was different from that of sediment and seawater (Figs. F2, F3, F4, F5, F6).
The DGGE analysis showed a relatively high diversity of bacteria in native basalt samples, which was visualized as several DNA bands on the gel (Fig. F1). However, as seen in Figure F1, each basalt sample had only one or a few dominant bands, indicating a limited number of dominant members of the population in these samples. Three dominant populations (DNA bands 1-3 in Fig. F1) were common to most of the basalt samples. The sequencing results of the common DNA bands found in the basalt samples in lanes 7, 9, and 10 (DNA band 1 in Fig. F1), corresponding to the OTU sequence named ODP-1161A-1162A-1163A in Figure F2, were closely related (99% similarity) to the uncultured gamma Proteobacterium clone 33-PA57B00 (accession number AF469304), previously isolated from subseafloor habitats associated with a deep-sea volcanic eruption (B. Huber et al., unpubl. data). The common DNA bands found in basalt samples in lanes 2, 3, 4, 7, 9, and 10 (band 2 in Fig. F1), corresponding to the OTU sequences named ODP-1162B-1163A and ODP-1160B-1161A-1162A in Figure F2, were related (97%-99% similarity) to Acinetobacter junii, a gamma Proteobacterium associated with both bacterial infections (Higgins et al., 2001) and aquatic environments (Guardabassi et al., 1999). One dominant band (band 3 in Fig. F1) was present in most basalt samples and was also found in the sediment sample. The sequencing results (Figs. F2, F5) showed that the bands in lanes 3, 4, 5, 6, 8, and 11, corresponding to OTU ODP-1157B-1160B-1161A-1162A in Figure F2 and the sequence ODP-sediment-1154B-644 in Figure F5, were 99% similar to the gamma Proteobacterium Pseudoalteromonas flavipulchra, previously isolated from seawater (Ivanova et al., 2002). The DNA band in lane 2, named ODP-1160B-600 in Figure F2, seemed to be in the same position on the gel but was only 87% similar to P. flavipulchra. Two DNA bands in lanes 1 and 10 were also in the same position on the DGGE gel, but these bands were not sequenced.
The 16S rDNA sequences retrieved from microbes in native basalt samples indicated phylogenetic affiliation with six main groups of the domain Bacteria: Actinobacteria, green nonsulfur bacteria, the Cytophaga/Flavobacterium/Bacteroides (CFB) group, the Bacillus/Clostridium group, and the beta and gamma subclasses of Proteobacteria (Fig. F2). The most abundant phylogenetic group present in the basalt samples, based on the number of partial DNA sequences retrieved, was the gamma Proteobacteria. The closest known relatives to these sample sequences were P. flavipulchra (Ivanova et al., 2002), the metal reducer Shewanella (Bowman et al., 1997), the marine bacterium Marinobacter (Coates et al., 2002), and representatives of several unidentified gamma Proteobacteria isolated from subseafloor habitats associated with a deep-sea volcanic eruption (B. Huber et al., unpubl. data).
Results from the DNA analysis of microbes from enrichment cultures inoculated with basalt showed phylogenetic affiliation with seven main groups of Bacteria: Actinobacteria, the CFB group, the Bacillus/Clostridium group, and the alpha, beta, gamma, and epsilon subclasses of the Proteobacteria (Fig. F3A, F3B). As in the case of native basalt, the majority of these partial sequences are grouped with the gamma Proteobacteria (Fig. F3B). The DNA analysis revealed that few of the microbes from native basalt samples could be cultured (Fig. F4). The sequences affiliated with the green nonsulfur bacteria, found in native basalt samples, were not found in basalt enrichment cultures. Sequences grouping within the alpha and epsilon subclass of the Proteobacteria were found in basalt enrichment cultures but not in native basalt samples.
The sequences obtained from both basalt and seawater with the highest similarity were ODP-1157B-614 (Fig. F2) and ODP-seawater-1162-651 (Fig. F5), which were 96% similar. All other sequences, when aligned, showed 92% or less similarity. The sequences from microbes from native basalt and sediment, with the exception of the sequences clustering with P. flavipulchra, showed 93% or less similarity (Figs. F2, F5).
DNA analysis of microbes from sediment enrichment cultures showed phylogenetic affiliation with four main groups of bacteria: the CFB group, the Bacillus/Clostridium group, and the alpha and gamma subclasses of the Proteobacteria (Fig. F6). The sequences obtained from enriched basalt and sediment with the highest similarity were Cultured-ODP-1155B-18 (Fig. F3B) and Cultured-ODP-sediment-1154B-663 (Fig. F6), which were 100% similar, and the sequences clustering with the uncultured alpha Proteobacterium (Figs. F3A, F6), with 95%-96% similarity. All other sequences showed 93% or less similarity.
The sepiolite (drilling mud) sample did not yield a PCR product, and DNA sequences have therefore not been obtained from this sample.
The fluorescent microsphere data for the igneous rock samples showed that microspheres were detected in wash water from the exterior of all rock samples examined (Table T5). In many of the thin sections, microspheres were found located both inside fractures and on the thin section surfaces. However, six of the thin sections lacked microspheres. In the two sediment cores examined, fluorescent microspheres were not found (Table T5).