Correspondence: Jørn Einen, Department of Biology, University of Bergen, Jahnebakken 7, 5007 Bergen, Norway. Tel.: +47 55 58 81 87; fax: +47 55 58 96 71; e-mail: email@example.com
A SYBR Green real-time quantitative PCR (Q-PCR) assay for the detection and quantification of Bacteria and Archaea present in the glassy rind of seafloor basalts of different ages and water depths is presented. Two sets of domain-specific primers were designed and validated for specific detection and quantification of bacterial and archaeal 16S rRNA genes in DNA extracted from basaltic glass. Total cell numbers were also estimated by fluorescence microscopy analysis of SYBR Gold-stained samples. The results from the two different approaches were concurrent, and Q-PCR results showed that the total number of cells present in basalts was in the range from 6 × 105 to 4 × 106 cells g−1 basaltic glass. Further, it was demonstrated that these cells were almost exclusively from the domain Bacteria. When applying the same methods on samples of different ages (22 years–0.1 Ma) and water depths (139–3390 mbsl), no significant differences in cell concentrations or in the relative abundance of Archaea and Bacteria were detected.
The ocean crust covers c. 60% of the Earth's surface, and it is continuously regenerated as lava erupts at the midocean spreading ridges, forming pillow- and sheath-flow structures. The rapid cooling of the lava after eruption results in the formation of an outer c. 1-cm-thick amorphous glass rind that constitutes c. 10% (vol.) of the lava (Staudigel & Hart, 1983). Subsequent eruptions bury older lava structures, resulting in an ocean crust lava layer with a depth up to 500 m. The igneous ocean crust is the largest endolithic biotope on Earth, and the high porosity and fluid exchange suggest a high productivity. Most of the volcanic basement of the ocean crust is buried by sediments, but near the midocean spreading ridges there is globally >1 000 000 km2 of exposed basalt (Edwards et al., 2005). This is accessible without drilling and provides a window into the ocean crust endolithic microbial community.
Previous results from DNA-specific staining and chemical analysis indicate that microorganisms present in 6 Ma sediment buried basalts are associated with the altered glassy rind (Giovannoni et al., 1996; Torsvik et al., 1998). Furthermore, cells and pit marks apparently formed by microorganisms have been observed in association with the altered glass in recent (<1 Ma) seafloor-exposed basalts using scanning electron microscopy (SEM) (Thorseth et al., 2001, 2003). DNA-based techniques have also demonstrated that the glassy rind of young basalt is colonized rapidly by a diverse and unique microbial community (Thorseth et al., 2001; Lysnes et al., 2004a), and that there is an alteration of the microbial community during aging of the formed crust and burial by deep-sea sediments (Lysnes et al., 2004a, b). Although the knowledge regarding microbial populations and communities inhabiting pillow lava basalt is starting to accumulate, the enumeration and amount of biomass of ocean crust basalt has been sparsely studied.
SEM analysis has however demonstrated a high concentration of cells attached to seafloor basaltic glass, and based on organic carbon content cell concentration has been estimated to be 109 cells g−1. Parts of this carbon are most likely derived from extracellular material and fossilized cells containing partially degraded organic matter (Kruber, 2007). In a study of terrestrial hyaloclastic basalt from Hawaii, sampled at a depth of 1400 m below sea level (mbsl), a cell number of <105 cells g−1 rock was estimated based on amino acid concentrations. All 16S rRNA gene sequences retrieved from these samples belonged to organisms from the domain Archaea whereas no DNA was amplified with the Bacteria-specific primers (Fisk et al., 2003).
The objective of our study was to gain more knowledge about the biomass of the microbial communities in seafloor basalts. In order to study this we developed a SYBR Green-based quantitative-PCR (Q-PCR) assay to estimate cell numbers of Bacteria and Archaea present in the glassy rind of recent seafloor lavas. To evaluate the impact of variable environmental parameters on the microbial communities, samples of different ages and from different water depths were investigated and compared. In order to strengthen our findings, the samples were also investigated with fluorescence microscopy and SEM analysis.
Material and methods
Samples of seafloor basalts were collected from the Knipovich (SM00-52, -54, -55) and the Mohns (SM00-47) Ridges by dredging during two cruises in 2000 and 2001 with R/V Håkon Mosby to the Norwegian/Greenland Sea (Table 1). The water depth at the sampling sites varied between 2500 and 3390 mbsl, and the bottom seawater temperature was −1 °C. In addition three samples were dredged from a shallow marine area (139–564 mbsl) off Jan Mayen, where the bottom seawater temperature was 0 °C. These shallow samples originated from lavas that flowed into the ocean during eruptions from the Beerenberg volcano on Jan Mayen in 1970 (SM01-4 and SM01-5) and 1985 (SM01-60). The seafloor in this area was colonized by benthic fauna belonging to the phylum Echinodermata. All samples from the Knipovich Ridge and the Jan Mayen area had black, glassy surfaces with signs of incipient weathering in form of rust-colored iron hydroxides. From the degree of weathering, samples from the Knipovich Ridge were assumed to be of similar age as those from Jan Mayen. Based on the degree of oxidation and lack of sediment coverage, the sample from the Mohns Ridge was estimated to be 0.1 Ma old. Shipboard, the glassy rind of each sample was collected, crushed in a flame sterilized mortar, transferred to 50 mL Falcon tubes and immediately frozen at −20 °C for onshore DNA extraction. In addition, samples of crushed glass were stored in 95% ethanol for fluorescence microscopy enumeration. For SEM investigations, fragments of the basaltic glass were air dried and sputter coated with Au–Pd.
Table 1. Sample location and depth
Sampling depth (mbsl)
72°44.5′N–3°56.3E′ to 72°44.9′N–3°57.8E′
76°47.7′N–7°22.6′E to 76°48.4′N–7°22.7′E
76°48.1′N–7°24.6′E to 76°48.9′N–7°24.7′E
76°48.2′N–7°25.0′E to 76°49.9′N–7°25.3′E
Jan Mayen area
71°07.74′N–7°53.82′W to 71°07.91′N–7°55.28′W
Jan Mayen area
71°07.15′N–7°55.12′W to 71°07.14′N–7°56.19′W
Jan Mayen area
71°09.87′N–7°59.28′W to 71°09.90′N–7°58.62′W
DNA was extracted from 1–2 g of basaltic glass, by applying a freeze–thaw step to lyse the cells (Ueda et al., 1995), in combination with the BIO101 Soil DNA extraction kit (Qbiogen) (Borneman et al., 1996), with the following modifications. The glass was placed in a FastDNA tube, and 1 mL of 0.1 M Na-phosphate buffer (pH 9) containing 5% sodium dodecyl sulfate was added. The relatively high pH was used to facilitate the release of DNA from the glass fragments. The samples were then subjected to three freeze–thaw cycles using an ethanol/dry ice mixture with a water bath at a temperature of c. −60 and 65 °C. Further steps in the DNA extraction procedure followed the protocol described by the manufacturer.
In order to calculate the DNA extraction efficiency, basaltic glass was made DNA-free by heating at 450 °C for 24 h. Four samples of 1 g DNA-free basaltic glass were weighed at three digits accuracy to the BIO 101 tubes. An actively growing Escherichia coli culture was cooled on ice and diluted to 1 : 100 in cold fresh media. The total cell number of the culture was determined by counting in a Thoma chamber. Each of the four samples were then spiked with 8.7 × 106 cells in 100 μL media and DNA extraction was started immediately. DNA was extracted from the samples as described above, and the 16S rRNA genes from the extracted DNA were quantified by Q-PCR. Quantification of 16S rRNA genes was also performed directly on 8.7 × 106E. coli cells. All analyses were performed in triplicate. A duplicate dilution series of genomic E. coli DNA was used to create a standard curve. The DNA extraction efficiency was calculated by dividing the number of genes found in the extracted DNA from the number of genes found when using whole cells as template. The DNA extraction efficiency was found to be 1.6±0.5% (SD).
A range of primer combinations was screened in silico using the RDP II database (04/06). High domain specificity and the ability to bind to a large number of known 16S rRNA genes were used as selection criteria for the individual primers (Table 2). For the selected primer combinations, sensitivity and domain specificity were checked by Q-PCR as described below using 1 ng and 0.0001 ng DNA from E. coli, Flaviramulus basaltis, Micrococcus lysodeicticus, Clostridium perfringens, Archaeoglobus fulgidus, Methanosaeta concilii, Methanobacterium bryantii, Thermococcus litoralis and Sulfolobus sulfataricus as template.
Table 2. Sequence of primers used in this study and their specificity
Hits/total searched in RDP II v 9, allowing two mismatches. Domain Bacteria*
Hits/total searched in RDP II v 8.1, allowing two mismatches. Domain Archaea*
Search in RDP II database was performed only on those entries that had sequence information in the relevant area; this is why numbers of total search sequences varies slightly.
The numbers refer to relative positions in Escherichia coli 16S rRNA.
ACT CCT ACG GGA GGC AGC AG
ATT ACC GCG GCT GCT GG
AGG AAT TGG CGG GGG AGC A
BGG GTC TCG CTC GTT RCC
Q-PCR was performed using the QuantiTect SYBR® Green PCR Kit (Qiagen). The reactions were run on an MJ research Opticon 2 (Bio-Rad) Q-PCR machine. Concentrations of Bacteria 16S rRNA genes in the extracted DNA were estimated using 0.5 μM of the primers 338f and 518r (Ovreas & Torsvik, 1998), with the following thermal program: 95 °C for 15 min, 45 cycles of denaturing (15 s at 94 °C), annealing (30 s at 61 °C), extension (30 s at 72 °C), plate read (1 s at 72 °C) and plate read (1 s at 80 °C). The cycling was followed by a final extension at 72 °C for 7 min, and a melting curve analysis from 65–95 °C with a plate read every 0.5 °C. Samples were run in triplicate; PCR quality water and 0.75 ng of genomic A. fulgidus DNA were used as negative controls. A duplicate tenfold dilution series of genomic E. coli DNA, ranging from 7.5 × 10−5 to 0.75 ng, was used to generate a standard curve. The standard curve for the bacterial Q-PCR reaction had an R2 value of 0.931 and an amplification efficiency estimated to 104%.
Concentrations of Archaea 16S rRNA genes in the samples were estimated using 0.5 μM of the primers 931f (Jackson et al., 2001) and m1100r (Table 2). The following thermal program was used: 95 °C for 15 min, 45 cycles of denaturing (15 s at 94 °C), annealing (30 s at 64 °C), extension (30 s at 72 °C), plate read (1 s at 72 °C) and plate read (1 s at 81 °C). The cycling was followed by a final extension at 72 °C for 7 min, and a melting curve analysis from 65 to 95 °C with a plate read every 0.5 °C. Samples were run in triplicate, PCR quality water, 0.75 ng and 7.5 ng of genomic E. coli DNA were used as negative controls. A duplicate 10-fold dilution series of genomic A. fulgidus DNA, ranging from 1.5 × 10−5 to 1.5 ng, was used to generate a standard curve. The standard curve for the archaeal Q-PCR reaction had an R2 value of 0.993 and an amplification efficiency estimated to 109%.
Calculation of biomass
To calculate the number of 16S rRNA genes from E. coli and A. fulgidus genomic DNA, the following formula was used: 16S rRNA genes per nanogram genomic DNA=[1/(genome size × 660 Da × 1.661 × 10−15 g)] × number of 16S rRNA genes per genome. The average molecular mass for a base pair is 660 Da; 1 Da is 1.661 g × 10−15 g in the metric system. The genome sizes used were 4 639 221 bp for E. coli (Blattner et al., 1997) and 2 178 400 bp for A. fulgidus (Klenk et al., 1997). The numbers of 16S rRNA genes per genome used for E. coli and A. fulgidus were seven and four, respectively. These values were collected (01/2008) from the ribosomal RNA operon copy number database (http://ribosome.mmg.msu.edu/rrndb/index.php) (rrndb) (Klappenbach et al., 2001).
To calculate the number of cells g−1 basaltic glass from the number of 16S rRNA genes, the following formula was used: Cells g−1 basaltic glass=[(ng DNA measured by Q-PCR g−1 basalt)/(average number of 16S rRNA genes for the taxa)]/DNA extraction efficiency. The values for the average number of 16S rRNA genes used were 3.9 for Bacteria and 1.8 for Archaea; these values are from rrndb.
Fluorescent staining and microscopy
Finely crushed basaltic glass fixated in ethanol were shaken vigorously and large particles were allowed to settle. Then, 500 μL of the supernatant was removed, diluted in phosphate-buffered saline (PBS) and filtered onto 0.2 μm Anodisc filters (Whatman). The filters were stained with SYBR® Gold (Molecular Probes) using the protocol for SYBR® Green I as described by Noble & Fuhrman (1998) with slight modifications. Briefly, the filters were stained with SYBR® Gold (× 25 concentrated) for 15 min, surplus stain was removed and the filters were dried overnight at 40 °C. The filters were then mounted in PBS/glycerol 1 : 1 with 0.1%p-phenylenediamine as an antifading reagent. The cells were visualized and counted, using a fluorescence microscope (Axioplan, Zeiss).
Results and discussion
The Q-PCR assay worked well on the selected controls; all reactions yielded a single product of expected size. The archaeal primer set amplified all archaeal strains and none of the bacterial strains, and vice versa for the bacterial primer set. However, when using the Bacteria specific primers, a low background signal was seen in the negative controls and the nontemplate control; this background signal was two to three orders of magnitude lower than the signals in the samples. No background signals were observed when using the archaeal primers.
Using Q-PCR, Bacteria cell numbers were found to be in the range from 6 × 105 cells g−1 glass (SM00-54) to 4 × 106 cells g−1 glass (SM00-52); both samples were from the young (≤30 years) lava flows at the Knipovich Ridge. These values are effectively also the total cell numbers found using Q-PCR because the Archaea cell numbers in the samples are several orders of magnitude lower (Fig. 1). The total cell numbers as determined by SYBR® Gold staining and fluorescent microscopy were found to be in the range between 1 × 106 (SM00-55) and 2 × 107 cells g−1 glass (SM00-52) (Fig. 1). Both the samples were from the young (≤30 years) lava flows at the Knipovich Ridge. The similar results obtained by the two different methods strengthen our estimation for cell numbers in basalt.
No significant differences in cell numbers could be observed between the group of shallow and deep samples (P=0.05), nor was there any significant difference in cell numbers between young (15–30 years) and old (100 000 years) samples (P=0.05). A previous study has, on the other hand, revealed that aging of basalt leads to a qualitative change in the composition of the prokaryotic community (Lysnes, 2004).
A striking feature seen in our results was the very small Archaea population present in the basaltic glass. Archaeal 16S rRNA genes were found in all samples, except in SM00-54. The highest number of Archaea cells, 1 × 103 cells g−1 basaltic glass, was found in sample SM00-52, which represents <0.02% of the total prokaryotic community. The low numbers of Archaea in the basalt distinguished this biotope from the surrounding seawater (L. Øvreås, F.L. Daae, B.O. Steinsbu & J. Einen, unpublished data) and from deep-sea sediments where Archaea constitutes up to 22% and 2–3% of the prokaryotic community (Vetriani et al., 1999). Archaea probably play a minor role in seafloor basalts as they are found in very low cell numbers. This is also suggested in a study by Lysnes (2004), in which PCR amplification of archaeal genes proved difficult, most likely because of low template concentrations in basalts.
When comparing our data with those obtained from deep Hawaiian basalts (Fisk et al., 2003), some differences were found. The seafloor basalts from the Mohns and Knipovitch Ridges revealed cell numbers at least one order of magnitude higher than the Hawaiian basalts, and whereas no Bacteria were present in the deep subsurface Hawaiian basalts, organisms from this domain accounted for more than 99.9% of the prokaryotic community in the Mohns and Knipovich Ridge basalts. This discrepancy could be due to differences in lithology and/or microbial community responses during the burial process.
During sampling, the pillow lavas are dredged through thousands of meters with seawater. During this process, there is a potential for planktonic microorganisms in the seawater to attach to the surface of the basalts. However, in a previous study where samples were collected in the same manner, it was shown that DNA extracted from the basalt differed significantly from that found in the surrounding seawater (Lysnes et al., 2004a, b). This in addition to the lower cell concentrations found in seawater, typically 5 × 104–5 × 105 cells mL−1 (Whitman et al., 1998), compared with our findings in the basalt, most likely makes seawater intrusion into the samples of little importance in this study.
Several factors in the Q-PCR analysis have a potential to over- or underestimate the ‘real’ values. Each of the primer sets used in this study aligned with c. 90% of the target sequences in the RDP II database when allowing for two mismatches. The failure of the primers to anneal with 100% of the target sequences will lead to an underestimation of the total cell concentration.
Our calculation of cell numbers in basalt is based on one genome per cell. In an active microbial community the correct value must inevitably be higher. This will lead to an overestimation of cell numbers. Another factor that may lead to an overestimation of cell numbers is the presence of extracellular DNA in the environment. Dell'Anno & Danovaro (2005) reported that the majority of the DNA found in deep-sea sediments are of extracellular origin. Because the organic carbon load is much lower in basalts than in sediments, we expect that the concentration of extracellular DNA in basalt is also lower. As this is not verified, it remains one of the big uncertainties in analyses based on gene enumerations. Despite the uncertainties in both microscopic enumeration and Q-PCR, the consistency in the obtained results from the two different methods strongly supports that cell numbers in basaltic glass are in the range of c. 106 cells g−1.
Our estimate of c.106 cells g−1 basaltic glass is significantly lower than previously reported by Kruber (2007) where cell density, based on the organic carbon content, was estimated to be c.109 cells g−1 basaltic glass. A significant portion of this measured carbon is thus probably from accumulated fossilized cells and extracellular material as also discussed by Thorseth et al. (2001, 2003), and as seen in Fig. 2.
The cell enumerations reported here are one to two orders of magnitude higher than those usually reported from deep seawater (Whitman et al., 1998). If cell numbers of the same order of magnitude are found to be present throughout the pillow lava layer, the total numbers of microorganisms present in the oceanic crust exceed the number of microorganisms present in seawater. However, as basaltic glass alters during aging, the porosity and hence the seawater circulation is reduced. This, together with burial by sediments, will probably lead to a decrease in biomass as one moves away from the spreading ridge axis.
The crew and participants of the SUBMAR 2000 and 2001 cruises, and especially the cruise leader Rolf Birger Pedersen are thanked for well-executed cruises. Steffen Leth Jørgensen and Frida Lise Daae are thanked for advice and discussions regarding the Q-PCR assay. Bjørn Olav Steinsbu is thanked for invaluable discussions concerning the decision of primers used in this study. This work was supported by the Norwegian research council through projects SUBMAR and BIODEEP, and the EU funded project MIRACLE.