Abstract
Methanotrophic bacteria play an important role in global cycling of carbon and co-metabolism of contaminants. Methanotrophs from pristine regions of the Snake River Plain Aquifer (SRPA; Idaho, USA) were studied in order to gain insight into the native groundwater communities' genetic potential to carry out TCE co-metabolism. Wells were selected that were proximal to a TCE plume believed to be undergoing natural attenuation. Methane concentrations ranged from 1 to >1000 nM. Carbon isotope ratios and diversity data together suggest that the SRPA contains active communities of methanotrophs that oxidize microbially produced methane. Microorganisms removed from groundwater by filtration were used as inocula for enrichments or frozen immediately and DNA was subsequently extracted for molecular characterization. Primers that specifically target methanotroph 16S rRNA genes or genes that code for subunits of soluble or particulate methane monooxygenase, mmoX and pmoA, respectively, were used to characterize the indigenous methanotrophs via PCR, cloning, RFLP analysis, and sequencing. Type I methanotroph clones aligned with Methylomonas, Methylocaldum, and Methylobacter sequences and a distinct 16S rRNA phylogenetic lineage grouped near Methylobacter. The majority of clone sequences in type II methanotroph 16S rRNA, pmoA, and mmoX gene libraries grouped closely with sequences in the Methylocystis genus. A subset of the type II methanotroph clones from the aquifer had sequences that aligned most closely to Methylosinus trichosporium OB3b and Methylocystis spp., known TCE-co-metabolizing methanotrophs.
1 Introduction
Methanotrophic bacteria are found in numerous environments where they utilize methane as their sole source of carbon and energy [1]. Through the oxidation of methane, a significant greenhouse gas, they play an important role in the global cycling of carbon. Furthermore, they play a significant role in contaminant co-metabolism [2]. Their ability to co-metabolize chlorinated ethenes is important because these compounds are persistent in soils and aquifers that have been exposed to them, for example, as a result of container leaks or inappropriate disposal practices.
In recent years various methods have been devised to stimulate microbial co-metabolism of TCE. Approaches involve the use of indigenous methanotrophs that can be purposefully enriched in the subsurface to take advantage of their ability to co-metabolize TCE [3–5] or by natural attenuation [6]. For monitored natural attenuation to be fully accepted as a treatment technology the specific processes that are responsible for the loss of the contaminant must be known [7,8].
Methane monooxygenase (MMO), the enzyme used by methanotrophs to oxidize methane to methanol, is also capable of co-metabolic oxidation of chlorinated compounds including TCE [9–12]. MMO exists either as a soluble cytoplasmic form (sMMO) or a membrane-associated particulate form (pMMO) [1]. Although co-metabolism occurs with both forms of MMO, sMMO appears to be capable of higher rates of co-metabolism of halogenated solvents than pMMO [13,14]. Methanotrophs are taxonomically classified into two groups: type I methanotrophs belong to the γ-proteobacteria and assimilate carbon via the RuMP pathway, while type II methanotrophs are members of the α-proteobacteria and use the serine carbon assimilation pathway [1]. Both types co-metabolize chlorinated compounds [1]. Type I methanotrophs generally have only pMMO, while type II methanotrophs typically have both sMMO and pMMO [1], but there are exceptions [12,15–19].
Identification of methanotrophs from various habitats has been accomplished using both phylogenetic (16S rRNA) and functional (sMMO and pMMO) genes. The active site-containing polypeptide of pMMO is coded for by pmoA. The gene sequence for PmoA has been observed to diverge among methanotrophs and has been used by several groups as a molecular tool to assess methanotroph diversity [15,20–23]. Parallel phylogenetic analyses of 16S rRNA gene and pmoA nucleotide sequences suggest that environmental sequences of pmoA are useful indicators of phylogenetic position [15,21,22]. The mmoX[15,24] gene that codes for the α subunit of the sMMO hydroxylase component is also often targeted for diversity studies because it is the most conserved of the sMMO subunits. The genes for sMMO appear to diverge less than pmoA and have been less useful in assessing diversity of type II methanotrophs [25]. This may be due in part to the limited number of mmoX sequences in public databases.
The goals of this study were to assess both the phylogenetic and the functional (MMO) gene diversity of methanotrophs in the basaltic SRPA [6]. We were interested in the genetic attributes of methanotrophs in the aquifer and thus their potential role in the observed attenuation of a plume of TCE in the SRPA. While this study did not sample groundwater within the contaminant plume, presumably the microbial population examined would have the same attributes as the initial population within the plume prior to extended alteration in the presence of TCE. By studying the microorganisms from uncontaminated wells, insight into the process of natural attenuation within the plume may be gained.
2 Materials and methods
2.1 Site description
The SRPA is a large, freshwater, semi-confined aquifer that exists beneath the semi-arid high desert in southeastern Idaho, USA. In situ aquifer temperatures are typically constant, 12–13 °C, and the water has saturating levels of dissolved oxygen (7.3–9.7 mg L−1) and low dissolved organic carbon (<1 mg L−1) [26]. The stratigraphic sequence in the subsurface consists of multiple basalt flows interbedded with thin sedimentary zones comprised of silts and clays with sporadic sand and gravel layers [27,28]. The top of the aquifer at the study sites varies between 69 m below land surface (mbls) at ANP 9 and 235 mbls at USGS 8.
Pristine groundwater was collected from four steel-encased wells (ANP 9, USGS 8, USGS 17, and USGS 112) for molecular and culture-based diversity analyses of methanotrophs and for aquifer geochemistry measurements.
2.2 Geochemical measurements
For each well at each sampling time, groundwater pH, temperature, and dissolved oxygen were determined using the MiniSonde® 4a (Hydrolab, Austin, TX). Three well volumes (one volume equivalent to the standing water in the well) were purged from each well prior to collection of water samples. MiniSonde® 4a measurements were stable at the time they were recorded. Wells were sampled in August 2000, January 2001, June 2001, and May 2002.
For dissolved organic carbon (DOC) measurements, 100-mL aliquots of water were collected in triplicate and stored in combusted (550 °C for 4.5 h), teflon-sealed amber vials. Water was immediately frozen on dry ice prior to storage at −80 °C. DOC was measured (Alchem Laboratories Inc., Boise, ID) following standard protocols for total organic carbon combustion or oxidation (EPA Method 415.1).
To measure dissolved methane concentrations, groundwater was collected in tared glass bulbs (Ace Glass Co., Vineland, NJ) according to the method described by Chapelle et al. [29] with modifications. At the wellhead, bulbs were filled in triplicate, flushed for at least 5 min to remove bubbles, sealed, and stored at 12 °C. Bulbs were weighed and one bulb from each well was injected through the septum with 6 mL of dry ultra-high purity (UHP) nitrogen. The bulbs were gently shaken at 12.5 °C for at least 1 h to equilibrate the nitrogen with the aqueous phase. The bubble was then extracted and injected directly into a gas chromatograph (GC) with flame ionization detector (FID) (Trace Analytical, Sparks, MD). The methane concentration in the original groundwater samples was calculated from GC readings according to Henry's Law using a constant value of H= 3.17×104 atm. mol−1 fraction at T= 12.5 °C.
For carbon isotopic analysis of the dissolved methane and inorganic carbon, the remaining bulbs were injected with 75 mL of dry nitrogen and equilibrated as described previously. The bubbles were extracted and injected into 35-mL serum vials which had been sealed with thick butyl rubber stoppers, then flushed with UHP nitrogen and evacuated to 2.7 kPa three times prior to use. Vials were connected to the bulbs and allowed to fill by pressure equilibration, after which any remaining gas was injected into the vials, except for 6 mL that was used for methane concentration analysis. Vials were stored at 4 °C prior to isotope ratio analyses. The 13C/12C measurements of methane (δ13CH4) and carbon dioxide (δ13CO2) were made by Continuous Flow-Isotope Ratio Mass Spectrometry [30]. After partitioning the vial gases by GC, the CH4 was converted to CO2 and H2O by oxidation in a high temperature Cu–Pt–Ni micro-combustion tube (900 °C). Combustion H2O was removed using a semi-permeable membrane NafionTM tube. The purified CO2 in the He carrier gas stream was directed to a splitter where a predetermined fraction was “sipped” by a controlled capillary leak into the source of the IRMS for separation and detection of m/z 44, 45, and 46. Values for CO2 were measured in a similar manner, but without the combustion stage. The isotope ratios were calculated from these results and reported in the standard δ-notation vs. V-PDB reference material. The precision is typically ± 0.3% for δ13CH4 and ± 0.2% for δ13CO2.
2.3 Direct microscopic counts
Total cells in the groundwater were enumerated in triplicate from aliquots that were fixed with formaldehyde (1.5%) upon collection and filtered onto polycarbonate track-etched membranes (Osmonics Inc., Westborough, MA). Cells on membranes were prepared and enumerated by the acridine orange direct count (AODC) method [31].
2.4 Sample collection and preparation for culture-independent and culture-based methanotroph diversity analyses
To obtain biomass for molecular analysis, cells in groundwater (100 L) were collected by direct inline filtration onto a 142-mm Supor® (Pall Corporation, Ann Arbor, MI) membrane filter (0.2-μm pore size). Filters were placed in sterile Whirl-Pak® bags (Nasco, Fort Atkinson, WI) and frozen immediately for transport to the laboratory where they were stored at −80 °C. Genomic DNA was extracted from the filters by crushing the frozen filters within the collection bags and processing in three bead-beating tubes from an UltraClean Water DNA Kit (MoBio, Solana Beach, CA) according to the manufacturer's instructions with the modification that the final elution was in 1 mL per tube. DNA from each filter was pooled (3 mL final volume) for each well and stored at −20 °C prior to molecular characterization.
Enrichments were prepared to determine if methanotrophs could be cultured from the groundwater and also to facilitate detection of methanotroph genes that might be undetectable by culture-independent molecular characterization. One liter of groundwater was filtered through a 47-mm Supor® filter (0.2 μm) that was subsequently used to inoculate methanotroph enrichments. Serum vials (125 mL) containing 50 mL of modified nitrate minimal salts (NMS) medium (copper-limited) [32] at either 1× (full-strength) NMS or 0.2× (one-fifth strength) NMS concentrations were used to enrich for type I and type II methanotrophs, respectively [33]. Cultures were incubated in the dark at 28 °C with shaking at 100 rpm. The headspace atmosphere was a mix of 90% air, 9% methane, and 1% CO2 in all inoculated vials (duplicates from each well) and in control vials without cells. Control vials with cells had a methane-free atmosphere (99:1; air:CO2). The methane atmosphere was maintained by periodic (every 3–4 days) flushing of the headspace. After two weeks of enrichment, aliquots of each culture were removed by syringe, boiled for 15 min to lyse the cells to obtain template for molecular analysis, and stored at −20 °C.
2.5 PCR amplification
PCR primers that specifically target the 16S rRNA gene(s) in type I and type II methanotrophs or functional genes that code for the α-subunit of the hydroxylase component of the sMMO (mmoX) or the 27-kDa peptide of the pMMO (pmoA) were used to characterize the diversity of microbial populations (Table 1). All PCR products were observed with UV illumination following electrophoresis through agarose gel and ethidium bromide staining.
Primer sequences used for the molecular characterization of methanotroph populations derived from the Snake River Plain Aquifer
| Primera | Sequence (5′ to 3′) | Target | References |
| mmoX f882 | GGCTCCAAGTTCAAGGTCGAGC | mmoX | [35] |
| mmoX r1403 | TGGCACTCGTAGCGCTCCGGCTCG | mmoX | [35] |
| pmof1 | GGGGGAACTTCTGGGGITGGAC | pmoA | [34] |
| pmor | GGGGGRCIACGTCITTACCGAA | pmoA | [34] |
| MethT1bR 988–1006 | GATTCYMTGSATGTCAAGG | Type I-specific | [33] |
| MethT1dF 84–102 | CCTTCGGGMGCYGACGAGT | Type I-specific | [33] |
| MethT2R 997–1017 | CATCTCTGRCSAYCATACCGG | Type II-specific | [33] |
| 27F | GAGAGTTTGATCMTGGCTCAG | Universal bacterial 16S rRNA gene | [48] |
| 8F | AGAGTTTGATCCTGGCTCAG | Universal bacterial 16S rRNA gene | [49] |
| 1492R | TACGGYTACCTTGTTACGACTT | Universal bacterial 16S rRNA gene | [50] |
| Primera | Sequence (5′ to 3′) | Target | References |
| mmoX f882 | GGCTCCAAGTTCAAGGTCGAGC | mmoX | [35] |
| mmoX r1403 | TGGCACTCGTAGCGCTCCGGCTCG | mmoX | [35] |
| pmof1 | GGGGGAACTTCTGGGGITGGAC | pmoA | [34] |
| pmor | GGGGGRCIACGTCITTACCGAA | pmoA | [34] |
| MethT1bR 988–1006 | GATTCYMTGSATGTCAAGG | Type I-specific | [33] |
| MethT1dF 84–102 | CCTTCGGGMGCYGACGAGT | Type I-specific | [33] |
| MethT2R 997–1017 | CATCTCTGRCSAYCATACCGG | Type II-specific | [33] |
| 27F | GAGAGTTTGATCMTGGCTCAG | Universal bacterial 16S rRNA gene | [48] |
| 8F | AGAGTTTGATCCTGGCTCAG | Universal bacterial 16S rRNA gene | [49] |
| 1492R | TACGGYTACCTTGTTACGACTT | Universal bacterial 16S rRNA gene | [50] |
Numbers indicate nucleotide position of primer in the gene sequence: mmoX numbering corresponds to Methylococcus capsulatus (Bath) and Methylomonas trichosporium OB3b; 16S rRNA gene numbering corresponds to E. coli.
Primer sequences used for the molecular characterization of methanotroph populations derived from the Snake River Plain Aquifer
| Primera | Sequence (5′ to 3′) | Target | References |
| mmoX f882 | GGCTCCAAGTTCAAGGTCGAGC | mmoX | [35] |
| mmoX r1403 | TGGCACTCGTAGCGCTCCGGCTCG | mmoX | [35] |
| pmof1 | GGGGGAACTTCTGGGGITGGAC | pmoA | [34] |
| pmor | GGGGGRCIACGTCITTACCGAA | pmoA | [34] |
| MethT1bR 988–1006 | GATTCYMTGSATGTCAAGG | Type I-specific | [33] |
| MethT1dF 84–102 | CCTTCGGGMGCYGACGAGT | Type I-specific | [33] |
| MethT2R 997–1017 | CATCTCTGRCSAYCATACCGG | Type II-specific | [33] |
| 27F | GAGAGTTTGATCMTGGCTCAG | Universal bacterial 16S rRNA gene | [48] |
| 8F | AGAGTTTGATCCTGGCTCAG | Universal bacterial 16S rRNA gene | [49] |
| 1492R | TACGGYTACCTTGTTACGACTT | Universal bacterial 16S rRNA gene | [50] |
| Primera | Sequence (5′ to 3′) | Target | References |
| mmoX f882 | GGCTCCAAGTTCAAGGTCGAGC | mmoX | [35] |
| mmoX r1403 | TGGCACTCGTAGCGCTCCGGCTCG | mmoX | [35] |
| pmof1 | GGGGGAACTTCTGGGGITGGAC | pmoA | [34] |
| pmor | GGGGGRCIACGTCITTACCGAA | pmoA | [34] |
| MethT1bR 988–1006 | GATTCYMTGSATGTCAAGG | Type I-specific | [33] |
| MethT1dF 84–102 | CCTTCGGGMGCYGACGAGT | Type I-specific | [33] |
| MethT2R 997–1017 | CATCTCTGRCSAYCATACCGG | Type II-specific | [33] |
| 27F | GAGAGTTTGATCMTGGCTCAG | Universal bacterial 16S rRNA gene | [48] |
| 8F | AGAGTTTGATCCTGGCTCAG | Universal bacterial 16S rRNA gene | [49] |
| 1492R | TACGGYTACCTTGTTACGACTT | Universal bacterial 16S rRNA gene | [50] |
Numbers indicate nucleotide position of primer in the gene sequence: mmoX numbering corresponds to Methylococcus capsulatus (Bath) and Methylomonas trichosporium OB3b; 16S rRNA gene numbering corresponds to E. coli.
The 16S rRNA gene primers MethT1bR and MethT1dF were designed to be specific for type I methanotrophs [33], while MethT2R [33] and bacterial-specific 27F target type II methanotrophs. PCR mixtures (25 μL) contained 200 μM each deoxynucleotide triphosphate (dNTP; Roche Molecular Biochemicals, Indianapolis, Indiana), 0.4 μM each specified primer, 1.25 U Taq2000TM DNA polymerase (Stratagene, La Jolla, CA), 1× polymerase buffer, and DNA template (15 μL from filters or 5 μL from enrichments). The thermocycler program included a 5-min denaturation at 95 °C, 35 cycles at 94 °C (1 min), 45 °C (1 min), and 72 °C (2 min) and final elongation at 72 °C (15 min).
To amplify 16S rRNA genes directly from groundwater a nested PCR approach was used because of the low abundance of methanotrophs, even after concentration of cells by filtration. Bacterial 16S rRNA genes were amplified in PCR mixtures (25 μL) containing 200 μM each dNTP, 0.8 μM each primer (8F and 1492R), 1.25 U of AmpliTaq® DNA polymerase LD (Applied Biosystems, Foster City, CA), 1× polymerase buffer, 10 μg BSA (Roche Molecular Biochemicals, Indianapolis, IN), and 2 μL DNA template. The thermocycler program included a denaturation at 94 °C (5 min), 33 cycles at 94 °C (1 min), 50 °C (1 min), and 72 °C (2 min) and final elongation at 72 °C (15 min). Nested PCR was carried out subsequently using 2 μL of the initial PCR diluted 40×, with methanotroph type I- and type II-specific 16S rRNA gene primers. PCRs were conducted as explained previously for the methanotroph-specific 16S rRNA gene amplification with the exception that SureStartTM DNA polymerase (1.25 U; Stratagene, La Jolla, CA) was used and activated by a 10-min hot-start.
The diversity of pMMO in groundwater methanotrophs was assessed using primers specific for pmoA[34]. The primer pair pmof1 and pmor amplifies about half of pmoA and is one of the two sets of primers reported in the literature that do not also amplify the ammonia monooxygenase gene (amo A) from ammonia oxidizers [34]. PCR amplifications of pmoA (25 μL) contained 200 μM each dNTP, 0.1 μM each designated primer, 1.25 U Taq2000TM DNA polymerase, 1× polymerase buffer, 10 μg BSA, 2% DMSO, and 15 μL DNA extract from groundwater filters. The amplification reaction was the same as described above for the methanotroph 16S rRNA gene-specific primers except that the elongation step was 1 min.
Primer pair mmoX f882 and mmoX r1403 that target mmoX was used to assess the diversity of sMMO in groundwater methanotrophs [35]. Lysed cells from enrichments were used as template (2 μL) and the reaction performed as for pmoA with the exception that SureStartTM DNA polymerase was used and activated by a 10-min hot-start.
2.6 Amplicon cloning and screening by RFLP analysis
For each specific amplified DNA preparation, 3–4 independent PCR products were combined, concentrated, and purified by agarose gel electrophoresis (Qiagen PCR Purification Kit and QIAquick Gel Extraction Kit, respectively; Valencia, CA). Individual clone libraries were constructed from amplified DNA using the pCR4-TOPO vector and TA cloning (Invitrogen, Carlsbad, CA). Recombinant plasmids were recovered from Escherichia coli TOP10 cells following transformation and selection on S-Gal LB agar plates (Sigma, St. Louis, MO) with kanamycin (50 μg mL−1).
To identify unique amplified methanotroph 16S rRNA, sMMO, or pMMO genes, clones from each well were screened by restriction fragment length polymorphism (RFLP) analysis. Clones maintaining plasmid inserts were lysed at 99 °C for 15 min and inserts were amplified in reaction mixtures using 0.4 μM each of primers M13 forward and M13 reverse (flanking the insertion site of interest), 0.75 U Taq DNA polymerase (Promega, Madison, WI), 1× polymerase buffer, 1.8 mM MgCl2, and 200 μM each dNTP. Thermocycler conditions included a 4-min denaturation at 94 °C, 33 cycles of 94 °C (20 s), 53 °C (20 s) and 72 °C (30 s for pmoA and 75 s for the 16S rRNA gene), and final extension at 72 °C (10 min). The mmoX gene was amplified using 25 cycles at 94 °C (1 min), 53 °C (1 min), and 72 °C (2 min) and final extension at 72 °C (5 min). Amplified inserts were digested with 0.4 U of each Hin P1I and Msp I (New England Biolabs, Beverly, MA), for 4.5 h at 37 °C. Clones were screened by separating DNA fragments through 3% MetaPhor Agarose (BioWhittaker Molecular Applications, Rockland, ME) in Tris–borate–EDTA buffer. Unique restriction patterns were identified, and representative clones from each library were selected for sequencing.
2.7 Sequencing and phylogenetic analysis
DNA sequencing of amplified, cloned sequences was performed on an ABI Prism Model 3700 (Applied Biosystems, Inc., Foster City, CA) using BigDye version 3 chemistry. A consensus sequence was generated based on a minimum of four independent sequence determinations with each primer.
Electropherograms were edited using the Chromas freeware (version 1.45; School of Health Science, Griffith University, Gold Coast Campus, Southport, Qld., Australia) and sequences were assembled, aligned, and analyzed with the DNAStar Lasergene software package (Madison, WI). 16S rRNA gene sequences were checked with Chimera Check from the Ribosomal Database Project II (RDP) [36]. Closest relatives were initially identified from sequences aligned against known sequences in the GenBank database using the gapped BLAST tool and the RDP databases for 16S rRNA genes [37]. 16S rRNA gene alignments were performed against known taxonomic sequences obtained from the RDP and GenBank using the BioEdit Sequence Alignment Editor freeware (version 5.0.9; Department of Microbiology, North Carolina State University, Raleigh, NC). Sequences were manually corrected using the MacClade software (version 3.0; Sinauer Associates, Inc., Sunderland, MA) to ensure only homologous nucleotides were compared between sequences. Sequences were considered to be unique if they differed by one base pair or more. GenBank and SwissProt databases were used to obtain sMMO and pMMO gene nucleotide and amino acid sequences.
The edited alignments were evaluated with maximum parsimony, maximum likelihood, and distance methods using the PAUP package (version 4.0b10; Sinauer Associates, Inc., Sunderland, MA). Trees generated with all three methods were congruent, with only minor rearrangements in branching order. Phylogenetic inference and evolutionary distance calculations were generated using the Jukes–Cantor distance model (γ parameter equal to 2.0). Bootstrap analysis replicates (10,000) were used to obtain confidence estimates for the phylogenetic trees.
3 Results and discussion
3.1 Groundwater characteristics for SRPA wells
The water from each well was near neutral pH and oxic (Table 2), properties that are characteristic of SRPA wells in general [26]. Dissolved organic carbon (DOC) ranged from <1.0 mg L−1 in USGS 17 to 2.6 mg L−1 in ANP 9. These values are similar to DOC levels in groundwater across the SRPA, especially where the aquifer has not been impacted by anthropogenic activities [38–40]. Both the low DOC and the high dissolved oxygen concentrations that are typical of the SRPA indicate that it is largely an oligotrophic system. Values for methane dissolved in the aquifer waters that were tested ranged from 1 to 1250 nM (Table 2). Considerable variability in the dissolved methane concentrations was evident in the same wells sampled at different times. Groundwater at equilibrium with the atmosphere under SRP conditions would have a methane concentration of about 3 nM. The measured values therefore suggest a methane source within the aquifer.
Characteristics of Snake River Plain Aquifer groundwater for wells ANP 9, USGS 8, USGS 17 and USGS 112
| Well | Groundwater characteristicsa | ||||||||||
| Methane (nM)b | Dissolved organic carbon (mg L−1) | Cell number (cells mL−1) | Dissolved oxygen (mg L−1) | pH | Temperature (°C) | Stable isotopesc | |||||
| Ave | SD | n | Range | δ13CCH4 | δ13CCO2 | ||||||
| ANP 9 | 72 | 51 | 6 | 33–155 | 1.5–2.6 | 1.6 × 103 | 7.5–8.4 | 7.2–7.9 | 14.1–14.5 | −61.7 to −62.0 | −18.3 to −19.2 |
| USGS 8 | 287 | 369 | 12 | 74–1250 | 1.6 | 1.4 × 103 | 7.0–8.8 | 7.0–7.9 | 11.4–11.6 | −62.9 to −69.0 | −18.9 to −19.1 |
| USGS 17 | 86 | 31 | 7 | 49–134 | <1.0 | 2.4 × 103 | 9.5–9.7 | 7.3–8.1 | 12.8–13.2 | −48.0 to −58.7 | −20.4 to −21.0 |
| USGS 112 | 19 | 22 | 8 | 1–52 | 1.0–1.2 | 2.1 × 103 | 8.6–9.0 | 7.1–7.8 | 13.1–13.9 | −24.4 to −24.7 | −20.2 to −21.3 |
| Well | Groundwater characteristicsa | ||||||||||
| Methane (nM)b | Dissolved organic carbon (mg L−1) | Cell number (cells mL−1) | Dissolved oxygen (mg L−1) | pH | Temperature (°C) | Stable isotopesc | |||||
| Ave | SD | n | Range | δ13CCH4 | δ13CCO2 | ||||||
| ANP 9 | 72 | 51 | 6 | 33–155 | 1.5–2.6 | 1.6 × 103 | 7.5–8.4 | 7.2–7.9 | 14.1–14.5 | −61.7 to −62.0 | −18.3 to −19.2 |
| USGS 8 | 287 | 369 | 12 | 74–1250 | 1.6 | 1.4 × 103 | 7.0–8.8 | 7.0–7.9 | 11.4–11.6 | −62.9 to −69.0 | −18.9 to −19.1 |
| USGS 17 | 86 | 31 | 7 | 49–134 | <1.0 | 2.4 × 103 | 9.5–9.7 | 7.3–8.1 | 12.8–13.2 | −48.0 to −58.7 | −20.4 to −21.0 |
| USGS 112 | 19 | 22 | 8 | 1–52 | 1.0–1.2 | 2.1 × 103 | 8.6–9.0 | 7.1–7.8 | 13.1–13.9 | −24.4 to −24.7 | −20.2 to −21.3 |
Sampling and analysis procedures are described in Section 2.
Ave, average; SD, standard of deviation; n, sampling number.
Characteristics of Snake River Plain Aquifer groundwater for wells ANP 9, USGS 8, USGS 17 and USGS 112
| Well | Groundwater characteristicsa | ||||||||||
| Methane (nM)b | Dissolved organic carbon (mg L−1) | Cell number (cells mL−1) | Dissolved oxygen (mg L−1) | pH | Temperature (°C) | Stable isotopesc | |||||
| Ave | SD | n | Range | δ13CCH4 | δ13CCO2 | ||||||
| ANP 9 | 72 | 51 | 6 | 33–155 | 1.5–2.6 | 1.6 × 103 | 7.5–8.4 | 7.2–7.9 | 14.1–14.5 | −61.7 to −62.0 | −18.3 to −19.2 |
| USGS 8 | 287 | 369 | 12 | 74–1250 | 1.6 | 1.4 × 103 | 7.0–8.8 | 7.0–7.9 | 11.4–11.6 | −62.9 to −69.0 | −18.9 to −19.1 |
| USGS 17 | 86 | 31 | 7 | 49–134 | <1.0 | 2.4 × 103 | 9.5–9.7 | 7.3–8.1 | 12.8–13.2 | −48.0 to −58.7 | −20.4 to −21.0 |
| USGS 112 | 19 | 22 | 8 | 1–52 | 1.0–1.2 | 2.1 × 103 | 8.6–9.0 | 7.1–7.8 | 13.1–13.9 | −24.4 to −24.7 | −20.2 to −21.3 |
| Well | Groundwater characteristicsa | ||||||||||
| Methane (nM)b | Dissolved organic carbon (mg L−1) | Cell number (cells mL−1) | Dissolved oxygen (mg L−1) | pH | Temperature (°C) | Stable isotopesc | |||||
| Ave | SD | n | Range | δ13CCH4 | δ13CCO2 | ||||||
| ANP 9 | 72 | 51 | 6 | 33–155 | 1.5–2.6 | 1.6 × 103 | 7.5–8.4 | 7.2–7.9 | 14.1–14.5 | −61.7 to −62.0 | −18.3 to −19.2 |
| USGS 8 | 287 | 369 | 12 | 74–1250 | 1.6 | 1.4 × 103 | 7.0–8.8 | 7.0–7.9 | 11.4–11.6 | −62.9 to −69.0 | −18.9 to −19.1 |
| USGS 17 | 86 | 31 | 7 | 49–134 | <1.0 | 2.4 × 103 | 9.5–9.7 | 7.3–8.1 | 12.8–13.2 | −48.0 to −58.7 | −20.4 to −21.0 |
| USGS 112 | 19 | 22 | 8 | 1–52 | 1.0–1.2 | 2.1 × 103 | 8.6–9.0 | 7.1–7.8 | 13.1–13.9 | −24.4 to −24.7 | −20.2 to −21.3 |
Sampling and analysis procedures are described in Section 2.
Ave, average; SD, standard of deviation; n, sampling number.
The dissolved methane measured in ANP 9 and USGS 8 have δ13CH4 values between −60‰ and −70‰ (Table 2). This strongly suggests that the methane is of microbial origin, likely by the CO2 reduction pathway [41]. In contrast, the methane from USGS 17 and USGS 112 is distinctly enriched in 13C (Table 2). These heavier δ13CH4 values follow a trend consistent with methane, formed by methanogenesis, then subsequently undergoing methanotrophic oxidation, most probably in the aquifer. The kinetic isotope effect associated with this microbial methane oxidation (ηc ca. 10 [41]) preferentially removes 12CH4 over 13CH4 producing the observed 13C-enrichment in the remaining non-oxidized methane in the aquifer. Methane oxidation also adds 12C-rich carbon dioxide to the aquifer. This shifts the δ13CO2 to heavier values. However, due to high natural concentrations of dissolved inorganic carbon in these arid region aquifers, it is difficult to observe the addition of 12CO2 from methane oxidation and hence δ13CO2 shift. Biogenically produced methane and oxidation of methane has been demonstrated in other aquifers [42]. Although our research has focused on methanotrophic communities in the SRPA, methanogens have also been found in the aquifer [43] and these cells likely account for the origin of the methane that sustains the methanotrophs. (see [42] for review of subsurface microbial methane cycling).
3.2 Total microbial cells in SRPA groundwater
Total counts of microbial cells in each of the well water samples determined by AODC were similar between wells and slightly greater than 103 cells mL−1 (Table 2). These values are approximately 10-fold lower than direct counts of microbial cells in the SRPA from past studies [40,44]. The lower values for the direct counts observed in the current study may be due to the isolation of the wells that were sampled compared to the other studies in which the wells were near existing facilities.
3.3 Methanotroph 16S rRNA gene clone libraries
Because biomass from each well was low, cells were concentrated via filtration onto membranes to about 107–108 cells mL−1 and the DNA was extracted. Using the DNA extracted from the groundwater of each well as template, 16S rRNA genes were amplified using universal bacterial domain-specific primers (data not shown). However, when methanotroph-specific primers (type I and type II) were used, 16S rRNA genes could be amplified only from DNA obtained from USGS 8. Assuming equal success for amplification of DNA from all wells, these results suggest that USGS 8 had a larger natural methanotroph population. This result is reasonable since this well had the highest measured methane concentration of the wells analyzed.
To further overcome the methanotroph biomass limitation, a nested PCR approach using internal methanotroph-specific (type I and type II) 16S rRNA gene primers was used subsequent to PCR with universal bacterial 16S rRNA gene primers. The nested approach was successful in amplifying methanotroph sequences from each well.
Methanotrophs were enriched from groundwater, with methane as the only carbon source in 1× or 0.2× NMS medium, to detect methanotroph genes otherwise undetectable using culture-independent molecular characterization. Previously Wise et al. [33] suggested that type I methanotrophs could be enriched in 1× NMS medium, that type II methanotrophs could be enriched preferentially in 0.2× NMS medium, and that lower methane levels (9%) increased diversity. To favor isolation of type II methanotrophs from each of the samples, 0.2× NMS medium with copper omitted was used as the growth medium. 16S rRNA genes were amplified from DNA from two-week-old enrichment cultures using type I and type II methanotroph-specific primer pairs. More 16S rRNA gene product was amplified from the 1× NMS medium using type I primers than with the type II primers, whereas amplification from the 0.2× NMS medium resulted in more PCR product with the type II primers (data not shown). Consistent with Wise et al. [33], the results suggest that 1× NMS medium tended to enrich mostly type I methanotrophs and 0.2× NMS medium enriched type II methanotrophs.
In addition, sequence data showed that 0.2× NMS medium with limited copper enriched type II methanotrophs. Full strength (1×) NMS medium with limited copper enriched type I methanotrophs and type II methanotrophs. Although copper was not provided, the enrichment media may have provided sufficient copper for some type I methanotroph growth particularly in the low-methane atmosphere and during the low-biomass stages of growth, where the copper-to-biomass ratio was high [45,46]. The diversity of methanotroph sequences recovered from enrichments may have been underestimated due to the use of boiling as a means of lysing cells to release DNA for PCR amplification.
3.4 Methanotroph 16S rRNA gene sequence diversity
Sequence analysis of 16S rRNA gene clones showed that both type I and type II methanotrophs could be enriched from all wells (Figs. 1 and 2). The greatest number of unique 16S rRNA gene sequences was obtained from USGS 8 groundwater, the well that had the highest average methane concentration (287 nM).
Phylogenetic relationships among partial 16S rRNA gene sequences retrieved from water samples and enrichment cultures (as indicated by the codes explained below) using type I methanotroph-specific primers (884 base positions were considered). Scale bar indicates 0.005 nucleotide substitutions per site. Phylogenetic tree was constructed using PAUP based on maximum-distance analysis with the Jukes–Cantor correction. Bootstrap probability (percentage) values greater than 50% are represented at the nodes (10,000 replicates). Unique clones obtained from wells ANP 9, USGS 8, USGS 17, and USGS 112 in the SRPA are indicated by boldface type. Clones are named according to the following scheme: numerical identification of well (9, 8, 17, and 112); amplification (groundwater (GW), nested-groundwater (nGW), and enrichment (E)); clone designation. Accession number and fraction of clones from library are indicated in parentheses. The vertical bars indicate groupings of clones from combined libraries along with the percentage of clones that reside within that group.
Phylogenetic relationships among partial 16S rRNA gene sequences retrieved from water samples and enrichment cultures (as indicated by the codes explained in Fig. 1) using type II methanotroph-specific primers (914 base positions were considered). Scale bar indicates 0.005 nucleotide substitutions per site. Analysis was performed and clones named in the same manner as described for Fig. 1.
Of the type I clone sequences identified, about 70% were most closely related to Methylomonas spp., however, Methylobacter-related species were well represented by the clone libraries (14%). A novel group (SRPA-1) of sequences diverged from the closest related genus, Methylobacter, by 4–12%. The SRPA-1 group associated most closely with the Methylobacter genus and branched with the Methylobacter, from the other type I methanotrophs with a high bootstrap value (97%) supporting this branching order. Each of these clones was amplified from groundwater, without the need for enrichment. Considering the oligotrophic nature of the SRPA environment, methanotroph sequence divergence from previously cultivated methanotrophs may be indicative of selection of novel organisms by the environment and/or the inability of such methanotrophs to grow on conventional methanotroph media.
Phylogenetic analysis of the 16S rRNA gene type II methanotroph sequences indicated that most of the clones are closely related to the Methylocystis and the Methylosinus genera (Fig. 2). Each well was represented by a single predominant type II 16S rRNA phylotype: for ANP9, 9E-T2A1; USGS8, 8E-T2A1; USGS17, 17E-T2A11; and USGS 112, 112E-T2A12 (Fig. 2). These four phylotypes together comprise almost 75% of the total type II 16S rRNA gene clones. Although the sequences were not identical they all clustered closely with representative sequences from the genus Methylocystis and presumably speciate within that genus. Each of these sequences diverged <0.5–1.5% from each other. Although each of these clones was obtained from enrichments (1× and 0.2× NMS enrichment clone libraries), the clone from USGS 8 was also detected directly in the groundwater and represented half of the clones in that library. Results suggested that the organism from which this sequence was derived might have been an important member in the natural environment for USGS 8. Even though we were unable to amplify 16S rRNA genes directly from the remaining groundwater samples, it is likely that Methylocystis species dominate the pristine groundwater type II methanotrophs, since 94% of the phylotypes (SRPA-2) branched with the Methylocystis genus. Our analysis may underestimate the true diversity of type II methanotrophs present, since Heyer et al. [24] have noted the large number of mismatches between the primer pair used here and sequences for Methylosinus sporium.
Five phylotypes representing less than 1% of the type II methanotroph libraries clustered apart from other type II methanotroph sequences recovered in this study, and some of these sequences may not represent methanotroph sequences at all given their branching positions in the tree shown in Fig. 2.
3.5 Methanotroph pmoA and mmoX sequence diversity
The pmoA gene, but not mmoX, could be amplified directly from groundwater in this study. This may be due, in part, to the pmoA PCR being more efficient than that for mmoX. Phylogenetic analysis of the pmoA sequences recovered (Fig. 3) suggests that virtually all type II methanotrophs present in the groundwater fall within the genus Methylocystis, while the diversity of type I pmoA sequences mirrors the 16S rRNA gene diversity (Fig. 1), with sequences branching among Methylomonas and Methylobacter (the exception being two pmoA clones from USGS 17 that were most closely related to the thermotolerant type I methanotroph, Methylocaldum[47]). The mmoX gene sequences recovered (Fig. 4) were only obtained after amplification of mmo X genes from DNA from enrichment cultures. The association of these sequences with Methylocystis parallels the results obtained with both the pmoA sequence analysis as well as that of type II methanotroph 16S rRNA gene sequences (Fig. 2). Interestingly, only one type I mmoX clone (Fig. 4, 17E-sS122) was obtained in our study and was recovered from well USGS 17.
Phylogenetic relationships among partial pmoA gene sequences retrieved from water samples (as indicated by the codes explained in Fig. 1) using pmoA-specific primers (286 base positions were considered). Scale bar indicates 0.05 nucleotide substitutions per site. Analysis was performed and clones named in the same manner as described for Fig. 1.
3.6 Summary
The low total microbial biomass along with the observed dissolved organic carbon and dissolved oxygen concentrations are consistent with an oligotrophic environment within the SRPA. Stable isotope geochemistry of the dissolved methane and dissolved inorganic carbon indicate that the microbially produced methane in the groundwater is undergoing oxidation by the methanotrophs in the pristine aquifer.
In the four wells studied, we detected both type I and type II methanotroph sequences using cultivation-based and cultivation-independent analyses. Based on gel analysis of the amount of amplicon generated using the methanotroph type-specific primers, the growth of type II methanotrophs was favored in 0.2× NMS medium, whereas the growth of type I methanotrophs was favored in 1× NMS medium. However, diversity did not necessarily correlate with abundance. Clones obtained from amplification with type II methanotroph-specific 16S rRNA gene targeted primers yielded phylotypes similar to Methylocystis and Methylosinus genera that include TCE degraders. Clones resulting from amplification with type I methanotroph-specific 16S rRNA gene targeted primers yielded phylotypes similar to Methylomonas spp. and Methylobacter spp., and a novel lineage within the genus Methylobacter. Amplification with mmoX-specific primers yielded clones with phylotypes most similar to Methylocystis spp. and one type I Methylomonas-like clone. The pmoA clones were associated with type I and type II methanotroph genera, largely paralleling those seen in the respective phylogenetic trees generated from 16S rRNA gene sequences. Although some common organisms appear to be present in each well, RFLP and sequence analyses also suggest the presence of unique methanotrophs. Because of limited sampling and observed variability of dissolved methane concentrations measured in these wells, we were unable to determine whether the differences in the methanotroph communities were dictated by any particular property of the well. We did note that the well with the highest dissolved methane concentration (USGS 8) yielded the highest 16S rRNA gene sequence diversity. Only water from USGS 8 could be used to directly detect type I and type II methanotroph-specific 16S rRNA genes using PCR without a preliminary enrichment for methanotrophs or a bacterial DNA amplification step. This suggests that USGS 8 may possess a more abundant methanotroph population than other wells.
The finding of diverse methanotrophs in the basaltic Snake River Plain Aquifer is consistent with other investigations of deep subsurface environments, especially those habitats that foster the juxtaposition of oxic and anoxic zones that permit a coupling of methanotrophic communities with a source of methane [42]. Based on the molecular characterization of the methanotroph communities in the SRPA, our data suggest that the observed natural attenuation of the TCE plume in the aquifer at the Test Area North [6] could, in part, be due to the presence of methanotrophs in the aquifer sustained by methane as their primary energy source. Only pmoA nucleotide sequences could be amplified directly from DNA from groundwater without enrichment or the use of the nested PCR approach, and so extrapolation of the extent of methanotroph diversity in the aquifer is best made from examining those sequences. In that analysis the distribution of clones is almost equally divided between type II (Methylocystis; 47%) and type I (Methylomonas, Methylobacter, and Methylocaldum; 53%) methanotrophs. Future research on these subsurface methanotrophs will clarify the actual rates of methane oxidation and TCE co-metabolism under growth-limiting conditions consistent with the conditions typical of an aquifer. Our aim is to determine whether the inherent rates of methane oxidation and co-metabolism of TCE by these microorganisms can fully account for the demonstrated natural attenuation.
Acknowledgements
This work was supported with funding provided by the US Department of Energy, Office of Environmental Management, to the Idaho National Engineering and Environmental Laboratory, operated by Bechtel BWXT Idaho, LLC, under Contract DE-AC07–991D13727, and an NSERC Discovery Grant (M.J.W.). Specific support from the Office of Environmental Management, Environmental Systems Research, and the Environmental Management Science Program is appreciated.
References
Author notes
Authors contributed equally to manuscript.
