|Home | About | Journals | Submit | Contact Us | Français|
The diversity of Rieske dioxygenase genes and short-term temporal variability in the abundance of two selected dioxygenase gene sequences were examined in a naphthalene-rich, coal tar waste-contaminated subsurface study site. Using a previously published PCR-based approach (S. M. Ní Chadhain, R. S. Norman, K. V. Pesce, J. J. Kukor, and G. J. Zylstra, Appl. Environ. Microbiol. 72:4078-4087, 2006) a broad suite of genes was detected, ranging from dioxygenase sequences associated with Rhodococcus and Sphingomonas to 32 previously uncharacterized Rieske gene sequence clone groups. The nag genes appeared frequently (20% of the total) in two groundwater monitoring wells characterized by low (~102 ppb; ~1 μM) ambient concentrations of naphthalene. A quantitative competitive PCR assay was used to show that abundances of nag genes (and archetypal nah genes) fluctuated substantially over a 9-month period. To contrast short-term variation with long-term community stability, in situ community gene expression (dioxygenase mRNA) and biodegradation potential (community metabolism of naphthalene in microcosms) were compared to measurements from 6 years earlier. cDNA sequences amplified from total RNA extracts revealed that nah- and nag-type genes were expressed in situ, corresponding well with structural gene abundances. Despite evidence for short-term (9-month) shifts in dioxygenase gene copy number, agreement in field gene expression (dioxygenase mRNA) and biodegradation potential was observed in comparisons to equivalent assays performed 6 years earlier. Thus, stability in community biodegradation characteristics at the hemidecadal time frame has been documented for these subsurface microbial communities.
Polycyclic aromatic hydrocarbons (PAHs), derived from both natural and anthropogenic sources, are toxic pollutants formed during the incomplete combustion of organic matter (58). Phylogenetically and physiologically diverse microbial populations contribute to biodegradation of PAHs in soil, sediment, and groundwater (22, 52). The most widely studied bacterial degradation pathways for naphthalene, the simplest and most soluble PAH, and a model compound for PAH metabolism, are initiated by the catalytic activity of an evolutionarily conserved naphthalene dioxygenase (NDO) system (22, 52). The two primary NDO-mediated degradation pathways are distinguished by conversion of naphthalene, via salicylate, to either catechol (e.g., nah genes) or gentisate (e.g., nag genes ). Many bacteria harbor well-characterized NDO systems, including gram-negative species (e.g., Pseudomonas spp., nah genes; Burkholderia spp., phn genes; and Ralstonia and Comamonas spp., nag genes) and gram-positive species (e.g., Mycobacteria spp., nid genes; and Rhodococcus spp., nar genes) (22, 44).
A variety of culture-independent, biomarker-based approaches for characterizing microbial assemblages and their activity in naturally occurring communities have been applied in PAH-contaminated environments. Analyses have ranged from metagenomic libraries (62), fluorescence in situ hybridization (56), and functional gene arrays (55) to PCR targeting both DNA (5, 9, 51) and expressed RNA transcripts (6). The catabolic importance of archetypal NDO systems (encoded by nah genes) in many PAH-contaminated environments has been successfully demonstrated by functional-gene-targeted molecular approaches (6, 12, 14, 18, 23, 57, 63, 65). Furthermore, broad-specificity PCR primers have enlarged the target range of molecular methods for environmental gene detection and uncovered previously unknown dioxygenase diversity in environmental samples (42, 49, 70).
Previous work at a coal tar waste-contaminated aquifer (South Glens Falls, NY) has provided substantial insight on microbially mediated in situ degradation of contaminants (2, 3, 4, 28, 43, 68, 69). Over more than a decade of monitored natural attenuation at the site, naphthalene concentrations within the contamination plume have diminished from high levels of ~1.6 ppm (13 μM) to persistent low ppb concentrations (48, 69a). Here, short-term variability in catabolic gene diversity and abundance were examined over a 9-month period. We further investigated long-term community characteristics by comparisons to previously conducted measures of in situ dioxygenase mRNA expression and community metabolism of naphthalene. The aim of the present study was to assess the short- and long-term stability of community genotypic (catabolic gene sequence diversity and in situ expressed dioxygenase mRNA) and phenotypic (metabolism of the dominant organic contaminant, naphthalene) traits in a subsurface field site undergoing documented natural attenuation (46, 47).
Methods for sampling groundwater and measuring its geochemical characteristics have been described previously (3, 69a; J. M. Yagi, J. Suflita, L. Gieg, C. O. Jeon, and E. L. Madsen, unpublished data). Groundwater samples were collected from a control well (well 60) upgradient of the source contamination and from two wells (wells 12 and 36) located within the contaminant plume (43, 48, 68). Groundwater naphthalene concentrations were determined in 1998 and 2005 as described previously (48, 69a). Subsequent naphthalene concentrations were determined by solvent extraction and gas chromatography-mass spectrometry (GC-MS) (see below). Groundwater from well 12 and well 36 was sampled for naphthalene biodegradation assays (August 2006) by pumping water directly into 40-ml sterile glass vials (I-CHEM, Rochester, NY). Vials were completely filled, sealed immediately with Teflon-lined butyl rubber stoppers, and placed on ice for transport to the laboratory. Groundwater samples obtained in August, November, and December 2006 were fixed on site by the addition of 2% formaldehyde for acridine orange direct counts (34). Groundwater biomass was collected for molecular analyses (in November 2005, August 2006, November 2006, December 2006, and May 2007) on Durapore filters (Millipore Corp., Bedford, MA). Total DNA and RNA were extracted separately from filter-collected groundwater biomass by using sodium dodecyl sulfate-boiling lysis procedures with phenol-chloroform extraction as described previously (68, 69a).
Laboratory naphthalene biodegradation assays were set up within 24 h of sample collection using replicate, sealed 40-ml serum bottles filled with groundwater on site (see above). For each sampling time, two (live) microcosms were amended with 500 μl of sterile deionized water, and a single poisoned (control) treatment received 500 μl of 0.25 M HgCl2 in 5% HCl. To ensure that naphthalene concentrations in all of the microcosms were in the range of detection by GC-MS, 50 μl of a concentrated stock of naphthalene dissolved in dichloromethane was added to each bottle, increasing the ambient concentration by 50 μg/liter. Vials were incubated at 10°C in the dark without shaking. At various times (0, 1, 3, 6, and 14 days), two live samples and one killed control were sacrificed for solvent extraction and GC-MS analysis of naphthalene concentrations. Then, 5 ml of groundwater was removed from each vial using a glass syringe, followed by addition of exactly 2 ml of dichloromethane and 6 g of NaCl. The vials were shaken vigorously for 10 s to achieve complete dissolution of salt and stored 2 days upside-down at 4°C to allow phases to settle. From each vial, ~1.5 ml of the organic layer was transferred to a tared 2-ml vial with Teflon-lined rubber septum and screw cap and shaken vigorously for 2 min with 0.2 g of anhydrous sodium sulfate. Next, 1 ml of the dried extract was transferred to a fresh, tared 2-ml vial, concentrated under a stream of nitrogen gas, weighed (to allow quantification of naphthalene present in each water sample), and stored at −20°C until analysis. Naphthalene concentrations were determined by GC-MS using a Hewlett-Packard Model 6890 Series II gas chromatograph (2).
Four clone libraries were generated for PCR-amplified Rieske gene sequences. Each was derived from groundwater DNA sampled from either well 36 (November 2005 and August 2006) or well 12 (August 2006 and November 2006). Rieske dioxygenase genes in groundwater DNA were amplified by using the primers Rieske_f and Rieske_r (Table (Table1)1) (49). Duplicate PCRs were performed for each clone library using 0.5 to 25 ng of DNA template in 25-ml volumes and previously described cycling conditions (49). The pooled 78-bp Rieske amplicons were purified by electrophoresis and gel extraction using the QIAquick gel extraction kit (Qiagen) and ligated into the vector pCR2.1 (TOPO-TA cloning; Invitrogen) according to the manufacturers' protocols. Clones with the correct insert were verified in 80 to 130 randomly picked colonies per library by PCR using M13 primers, and inserts were sequenced using an ABI 3730 automated sequencer (Life Sciences Core Laboratories Center, Cornell University). Rieske fragment sequences were manually checked for quality, edited to exclude primer binding sites using 4Peaks software (available at http://www.mekentosj.com/4peaks/), and grouped into clone families based on 95% sequence identity. Neighbor-joining trees were constructed by using CLUSTAL X.
The bacterial strains and plasmids used for competitive PCR method development and assay are listed in Table Table1.1. All strains were cultivated in standard R2A medium at 23°C except for Escherichia coli strains, which were grown in Luria-Bertani broth at 37°C. Two primer pairs were designed to specifically amplify the nah and nag gene clusters by targeting sequences upstream of genes encoding the NDO large subunit: known nag gene clusters, but not nah clusters, encode genes for salicylate-5-hydroxylase in this region (44). The 750-bp amplicon from nah gene clusters spanned the region from nahAa to nahAc, and the 888-bp amplicon from nag gene clusters spanned the region from nagH to nagAc (72). Deletion derivatives of the respective gene clusters carried on the pUC18 plasmid were used as internal standards for competitive PCRs (30). Shortened amplicons generated from PCR of these deletion derivatives during coamplification with site-derived target sequences were distinguished from wild-type amplicons by gel electrophoresis, allowing quantification of specific genotypes in field-sampled nucleic acid extracts (30). Standard curves for quantitative competitive PCR were constructed with a constant amount of target and serial dilutions of the competitor carried on a plasmid.
Primer pairs for competitive PCR were tested for group specificity on isolated strains of naphthalene degrading bacteria known to harbor particular naphthalene degradation operons. The nah gene amplicon but not the nag gene amplicon was observed from PCR amplification of total DNA from pure cultures of P. putida NCIB 9816-4 and P. putida G7, and only the nag gene amplicon was amplified from Ralstonia sp. strain U2 and P. naphthalenivorans strain CJ2 (data not shown). The competitive PCR assay was validated by using DNA extracted from pure cultures. When equimolar quantities of target and standard were added to PCRs, similar amplification efficiencies were observed over 3 orders of magnitude tested (data not shown). For instance, when either 50 or 500 copies of nag were amplified, the PCR plateau was attained after 34 or 36 cycles, respectively; quantitation was thus carried out after 32 or 34 cycles to fall within the exponential phase of amplification.
Total RNA extracts from well 36 and well 12 (August 2006) were treated with DNase I (Invitrogen) and converted to cDNA using random hexamer-primed reverse transcription with SuperScript III reverse transcriptase (Invitrogen), according to the manufacturer's instructions. PCRs (25 μl) were set up as described above, using the PCR primers Ac114F.1 and Ac596R.1 that target both nahAc and nagAc (28) (Table (Table2)2) and cDNA as a template, according to the reverse transcriptase manufacturer's recommendations. Gel-purified PCR amplicons were used to generate clone libraries as described above. Vector-specific primers used to screen plasmids for inserts and DNA amplicons for 18 and 16 clones from the well 12 clone library and the well 36 clone library, respectively, were digested with the restriction endonucleases HaeIII and HhaI. The restriction products were separated on agarose gels, and 11 representative sequences were obtained, as described above. CLUSTAL X was used for alignment of nucleic acid and deduced amino acid sequences and construction of neighbor-joining trees.
Nucleic acid sequences and deduced amino acid sequences were subjected to BLASTN and BLASTP searches, respectively, using the National Center for Biotechnology Information website to find the closest relatives.
Statistical analyses were carried out by using the programs Analytic Rarefaction 1.3 (24) and EstimateS (10). Good's estimate was used to calculate library coverage using the formula [1 − (n/N)] × 100, where n is the number of sequences represented by a single clone group only, and N is the total number of sequences analyzed (19, 33).
The nucleotide sequence data reported here have been submitted to GenBank under accession numbers FJ820292 to FJ820329.
A previously described degenerate PCR primer pair was used to broadly characterize the complement of both known and unknown aromatic oxygenase genes in groundwater. Sequences encoding the conserved Rieske domain of dioxygenase homologs associated with attack of nonpolar aromatic substrates were targeted (49). Four Rieske gene fragment clone libraries were generated representing well 36 (November 2005 and August 2006) and well 12 (August 2006 and November 2006) (Table (Table2).2). No Rieske gene amplification products were obtained from DNA extracted from the background control well. A total of 110 Rieske gene fragments were sequenced (28, 28, 27, and 27 from each library, respectively; Table Table2).2). Sequences with >95% nucleotide identity were classified as individual clone groups, designated by source well (e.g., well 36 = MW36) and an arbitrarily assigned clone number (Fig. (Fig.11).
Shannon-Wiener indices and Chao-1 indicators for the clone libraries suggested that clone group diversity and richness were lower for Rieske gene assemblages amplified from the more contaminated well 36 than for well 12, regardless of the sampling date (Table (Table2).2). This could be attributed to inhibitory aspects of either anaerobic physiological conditions or higher naphthalene concentrations in well 36 on NDO associated microbial populations. Classical estimates of library coverage (Good's estimate) and rarefaction curves for each library indicated that additional sampling would uncover greater Rieske gene diversity in the groundwater communities (Table (Table2,2, Fig. Fig.22).
In total, 35 clone groups were identified in the 4 clone libraries, with limited overlap between libraries from well 36 and well 12 regardless of sampling time: 21 clone groups were found exclusively in well 12 samples, and 12 clone groups were identified in well 36 only (Fig. (Fig.1,1, Table Table2).2). The dominant clone group from well 36 in November 2005, MW36-09 (46% of library total), was identical in sequence to nagAc from Polaromonas naphthalenivorans strain CJ2 (Fig. (Fig.1).1). Clone group MW36-06 (14% of library total) clustered with biphenyl dioxygenases (bphA). Clone group MW36-11 (11% of library total) showed no close BLAST matches but clustered with S25 from Ní Chadhain et al. (49). Clone group MW36-12 clustered with an uncharacterized ring hydroxylating dioxygenase gene amplified from intertidal sediment (42). Clone group MW36-13 was the only other group in the library similar to known genes, affiliating with genes from Mycobacterium spp. Surprisingly, nah-type sequences characteristic of Pseudomonas putida were not detected in the Rieske gene survey (Fig. (Fig.11).
The two clone libraries from August 2006 revealed differences in the distribution of aromatic oxygenase genes between well 36 and well 12, suggesting genotypic differences in respective functionally important aromatic degrading microbial populations (Table (Table2,2, Fig. Fig.1).1). Furthermore, the relative abundances of specific clone groups in well 36 from August 2006 were different from those observed for November 2005, indicative of temporal variability in the aromatic degrading populations (Fig. (Fig.1).1). The August 2006 sample from well 36 contained six unique clone groups and was dominated by clone group MW36-15. Six clones fell into clone group MW36-14, which included the nagAc gene from Ralstonia sp. strain U2. Three clones grouped with MW36-04, which had no close matches in the GenBank database but clustered weakly with clones (S24, S16, and S15) from the Ní Chadhain et al. study (49). In contrast, the well 12 library from August 2006 contained 14 clone groups. Clone groups MW12-14, MW12-09, and MW36-14 dominated the library (22%, 15%, 11%). Clone group MW12-16, represented by 1 clone, showed high sequence identity with a gene from Mycobacterium gilvum. The remaining clone groups from August 2006 had no close matches in the GenBank database.
The library from the November sample from well 12 (2006; 27 total; Table Table2,2, Fig. Fig.1)1) was dominated by clone group MW12-07 (26%). None of the clone groups represented in the library showed significant sequence similarity to known genes, but substantial overlap (13 of 27 unique clone groups) with the August sample from the same well was observed (Fig. (Fig.11).
Seven clone groups were more than 60% identical to nucleotide sequences in the GenBank database (Table (Table3).3). However, the majority of the putative Rieske sequences uncovered from groundwater DNA extracts were novel. A total of 36 of the 110 clones (33%, or 19 clone groups [55% of total]) formed an independent unclassified lineage (Fig. (Fig.1).1). These sequences were most closely related to each other and showed no close matches in the GenBank database. To validate the functional significance of this lineage, 13 deduced amino acid sequences were aligned and compared to Rieske sequences from known aromatic hydrocarbon degraders (see Fig. S1 in the supplemental material). Sequences for clone groups MW12-03, MW12-18, MW12-10, MW12-11, MW12-12, MW36-08, MW12-20, and MW12-04 encoded identical peptides. Despite the short length of the amplicon, an alignment of deduced amino acid residues showed that residues conserved in Rieske centers from known PAH degraders were in fact conserved in the unclassified clones (see Fig. S1 in the supplemental material). Although the aligned region lacks known catalytically important residues, the high level of sequence conservation compared to known dioxygenases, combined with inferred conservation of key amino acids in the flanking primer binding regions, suggests a functionally relevant novel lineage. The best match for clone group MW12-07, the most abundant sequence from this lineage was to clone family S4 identified by Ní Chadhain et al. (49) (Table (Table3).3). The closest matching characterized sequence was from Sphingomonas wittichii RW1 (67% amino acid identity). Indeed, several members of the Sphingomonas genus have PAH degradation capabilities (40).
Known nag gene sequences were abundant in three of the four Rieske clone libraries (Fig. (Fig.1).1). In total, nine clones (MW36-14) matched nag gene sequences from Comamonas and Ralstonia and 13 clones (MW36-09) matched nagAc from Polaromonas naphthalenivorans strain CJ2 (Fig. (Fig.1).1). Only one clone group, the unclassified clone group MW36-15, was represented by more clones than the nag lineage (16 total). In addition, seven clone types (36 clones, 33% of the total) appear to be associated with a gram-positive lineage including nar and nid genes (MW12-15, MW36-15, MW36-13, MW12-16, MW12-06, MW12-09, and MW12-14; Fig. Fig.11).
Only a small subset of genes sampled by the Rieske fragment survey (Fig. (Fig.1,1, Tables Tables22 and and3)3) are available in current databases as full sequences; of these, nag genes were most frequently detected in the survey (Fig. (Fig.1).1). We developed a set of quantitative competitive PCR assays to assess the temporal variability in nag gene abundance and to compare the abundance of nag genes and the expected very low abundance of archetypal nah genes (see above) in the groundwater microbial communities. PCR products of the expected size were amplified from total DNA extracted from samples collected in August, November, and December 2006 and May 2007 (Fig. (Fig.3).3). The numbers of nah and nag gene copies were determined and expressed as copies liter−1 groundwater. The nag gene abundance ranged from below detection to 3 × 105 copies liter−1. Because nah gene sequences were not obtained in the Rieske clone libraries, we expected to see relatively low copy numbers in the groundwater samples. Surprisingly, the nah genes were quite abundant, though copy numbers ranged from below detection to maximum copy numbers ~10-fold higher than observed for nag genes (3 × 106 copies liter−1; Fig. Fig.3).3). Furthermore, the nah genes were present in all of the samples from well 36.
Previous measurements of in situ gene expression and biodegradation potential enabled an assessment of the long-term stability of microbial communities at the site. Between 1998 and 1999, the ambient concentrations of naphthalene in wells 36 and 12 were 1,150 and 43 ppb, respectively. Approximately 6 years later, in 2005, concentrations had shown modest fluctuation—changing to 376 ppb (3 μM) and 107 ppb (0.8 μM), respectively, in wells 36 and 12. Key factors influencing microbial metabolism of naphthalene, especially in situ oxygen concentration (<0.3 mg liter−1) and the cell density (~104 to 105), also remained relatively constant (data not shown).
Based on structural gene abundances in the August 2006 samples of wells 12 and 36 (Fig. (Fig.3),3), we predicted that mRNA transcripts in the well 36 microbial community would include both nah and nag sequences and that well 12 cDNA would be dominated by nag sequences. Degenerate primers targeting the evolutionarily conserved large NDO (nah and nag) subunit gene were used to examine well 12 and well 36 cDNA from August 2006. Reverse transcription-PCR amplicons of the expected size were obtained from both samples. From two clone libraries, a total of 34 clones were screened by restriction fragment length polymorphism (RFLP) analysis and 11 clones representing the three observed RFLP types (MWnah01, MWnah08, and MWnah10; Fig. Fig.4)4) were sequenced. Phylogenetic analysis of the cloned sequences showed that they were, as expected, nag- and nah-type genes associated with members of the gram-negative beta- and gammaproteobacteria (e.g., Comamonas and Pseudomonas spp.; Fig. Fig.4).4). Consistent with results of the structural gene analysis (Fig. (Fig.3),3), all of the 18 clones screened from well 12 fell within the nag clade represented by MWnah01, a sequence closely related (98% amino acid sequence identity) to pahAc from Comamamonas testosteroni strain H. Eight of the sixteen cDNA clones screened from well 36 were also related to pahAc from C. testosteroni H. The remaining well 36 clones grouped with the nah clade, represented by nahAc from P. putida NCIB 9816-4 (99% amino acid sequence identity). Remarkably, these sequences were also detected 6 years previously in expressed mRNA pools at the study site (68).
To assess the physiological phenotype of the on-site microbial communities, laboratory-incubated microcosms were prepared using waters from wells 36 and 12 incubated at 10°C (Fig. (Fig.5).5). Samples from August 2006 were used in the biodegradation assays. Aside from a naphthalene amendment (50 ppb), no chemical or nutrient additions were otherwise made to the microcosms. Biodegradation occurred in samples from both wells (Fig. (Fig.5).5). Again, we found agreement between data reported here and those from site samples taken 6 years earlier (2). A lag phase of 3 days was observed prior to disappearance of naphthalene in the bottles from well 36 but not from well 12 (Fig. (Fig.5).5). The relative delay in biodegradation activity in well 36 samples suggests induction or enrichment of latent aerobic microbial activity; thus, it appears that both microaerobic and anaerobic physiological conditions have persisted waters from well 36 for 6 years (Fig. (Fig.5).5). The absence of a noticeable lag phase in biodegradation assays for well 12 samples (Fig. (Fig.5)5) matched the pattern from 6 years earlier and is consistent with the hypothesis that microsites in situ harbored active aerobic and microaerobic naphthalene-degrading microbial populations. Overall, within 14 days of incubation, more than 99% of naphthalene initially present was degraded by native microorganisms in viable treatments from both wells (Fig. (Fig.55).
This survey of Rieske gene diversity indicates: (i) spatial and temporal variability in the relative abundance of various oxygenase genes in two groundwater microbial communities within a single contaminant plume and (ii) the prevalence of nag genes compared to other known NDO systems at the site. It was expected that the generic nature of the PCR primers would enable detection of known NDO systems including nah, nag, phn, nid, and nar genes, as well as many other diverse sequences (49). Surprisingly, the archetypal nah genes were not detected in the degenerate primer PCR-based Rieske gene survey (Fig. (Fig.1),1), although present based on subsequent genotype-specific PCR-based quantification (Fig. (Fig.3)3) and expression (Fig. (Fig.4)4) data. The extent and nature of PCR bias introduced by the highly degenerate Rieske primer set is unknown; however, recent reports by other investigators also suggest that nah genes may occur at low frequency in contaminated soil and sediment habitats (12, 18, 55). The diversity of novel putative dioxygenase sequences revealed by our survey (Fig. (Fig.1)1) suggests that unrecognized PAH-degrading enzymes may be important in the functioning of the aromatic hydrocarbon degrading microbial communities at the South Glens Falls site; these will be investigated in the future.
Competitive PCR was chosen for gene quantification over real-time quantitative PCR because, when implemented properly (30, 38, 60), the former technique assures uniform amplification of both the targeted gene and the internal standard in complex environmental DNA extracts that may feature various levels of PCR inhibitors and/or materials influencing nonspecific competitive DNA-binding reactions. The nah and nag NDO gene abundances fluctuated substantially over a 9-month period in wells 36 and 12 (Fig. (Fig.3),3), suggesting that site-specific geochemical conditions lead to variable enrichment of particular hosts of different NDO genes. Quantitative PCR has been reported to demonstrate a positive correlation between PAH concentration and normalized PAH degradation gene copy numbers (9). However, DeBruyn et al. (12), using quantitative PCR, demonstrated substantial uncorrelated variation in nahAc gene abundance in contaminated sediment. We sought but could not find obvious geochemical parameters (e.g., concentrations of O2 or contaminants) that might be controlling the observed variation in nag and nah gene abundance at this study site. However, we speculate that in the presence of extremely low bioavailable PAH concentrations, the native microorganisms are co-utilizing several carbon compounds (31, 41). PAH bioavailability in groundwater may be decreased significantly by, for instance, the presence of dissolved organic matter (32). Furthermore, aromatic hydrocarbons in the environment are present, not singly, but as complex mixtures. Mixed substrate growth, including simultaneous utilization of mixtures of PAHs, is expected in natural environments (11, 27, 39, 67). Dissolved and particulate organic carbon are potential alternate substrates for aromatic hydrocarbon degrading groundwater microbes (45). Under carbon-limiting conditions, Ralstonia pickettii has been shown to utilize benzene in the presence of more easily utilizable substrate succinate (7) and P. putida was shown to simultaneously assimilate glucose and toluene (13). Substrate co-utilization may help to explain the variation in NDO-harboring microbial population densities in the absence of obvious changes in naphthalene concentration.
Recent work has shown that nah genes were expressed at detectable levels only at naphthalene concentrations of 30 μM and higher (26). However, reverse transcription-PCR results from the present study (Fig. (Fig.4)4) indicated that ndo genes were actively expressed under apparent suboptimal physiological conditions (low ambient levels of oxygen and naphthalene). In situ expression of nag genes in groundwater was detected with ambient naphthalene concentrations as low as 0.8 μM. Indeed, several prior reports have suggested that nag-type genes may be transcribed in response to low concentrations of substrate (12, 26, 59). In the present study, nag genes were more abundant than nah genes and expressed in well 12, associated with lower naphthalene concentrations than well 36. The mRNA transcript profile for well 36, however, suggests coexisting active microbial populations harboring both nah and nag genes.
Induction of aromatic compound degradation pathways usually occurs at micromolar ranges of substrates (64). However, constitutive expression of catabolic genes is considered an advantage in oligotrophic environments because of an associated increase in the capacity to react quickly to transiently available nutrients (27). Constitutive expression of nah genes has been observed in pure culture experiments (21). Our results (Fig. (Fig.4),4), however, do not enable a distinction between constitutive and substrate-induced expression of nah and nag genes at the study site. Naturally occurring PAH-degrading microbial populations may face low and fluctuating bioavailability of complex mixtures of substrates (1) and may utilize different degradation mechanisms depending on low or high available substrate concentrations (21). Substrate utilization capacity must be optimized in the context of myriad biotic and abiotic environmental stresses (54, 66). Especially at very low concentrations of substrate, microbes may simultaneously use multiple carbon sources, reducing threshold concentrations for metabolism of individual substrates (36, 39, 45).
The dynamics of microbial communities (those residing in laboratory settings and in field sites) have been examined using small subunit rRNA clone libraries, molecular fingerprinting (e.g., T-RFLP and ARISA ), meta-transcriptomics (53), selected mRNA pools (25), and physiological traits (43, 61). The stability of microbial communities has been examined extensively in bioreactors (see, for example, references 15 and 17). Comparatively few studies have examined the in situ dynamics of real-world field subsurface study sites (8, 50). In the present study, community dynamics were examined by quantification of dioxygenase genes (over a 9-month period; Fig. Fig.3)3) and assays for both the dioxygenase mRNA pool (Fig. (Fig.4)4) and community naphthalene biodegradation traits (Fig. (Fig.5).5). Despite evidence for short-term (9 month) shifts in dioxygenase gene copy number, we found remarkable agreement in field gene expression (dioxygenase mRNA) and biodegradation traits compared to the same assays performed 6 years earlier. The present study has shown that over a hemidecadal period genotypic (in situ expressed dioxygenase mRNA) and phenotypic (community metabolism of the dominant organic contaminant, naphthalene) traits appear to be stable in a subsurface field site undergoing natural attenuation. This type of stability bodes well for environmental cleanup and undoubtedly is a reflection of the relatively slow pace of subsurface geochemical and microbiological processes (see, for example, reference 35). Furthermore, longer-term (16-year) contaminant history data for this site suggest a trajectory in geochemistry, microbiology, and microbial processes toward community succession and the elimination of contamination (48, 69a).
Funding was provided by National Institute of Environmental Health Sciences grants 1-R21-ES012834 and NSF DEB 0841999 (to E.L.M.). J.M.Y. was funded in part by the NSF IGERT Program and NIEHS training grant 5-T32-ES00752-28 (to S. Bloom and A. Yen, Cornell University).
Christopher DeRito, Graham Pumphrey, Jack Liou, Buck Hanson, and James Doroghazi assisted with sample collection and chemical analyses. Nehemiah Smith and Andrew Ayre assisted with clone library construction and molecular analyses.
Published ahead of print on 21 August 2009.
†Supplemental material for this article may be found at http://aem.asm.org/.