Diversity and abundance of n-alkane degrading bacteria in the near surface soils of a Chinese onshore oil and gas field

Introduction Conclusions References


Introduction
Hydrocarbon microseepage is a widely distributed natural phenomenon in the geochemical carbon cycle (Etiope and Ciccioli, 2009).Driven by reservoir pressure, some volatile components from oil and gas reservoirs can vertically pen-etrate the cover above and rise to the surface of the earth.These gaseous and volatile hydrocarbons will affect the distribution and growth of the microbial community in the nearsurface soil (Klusman and Saeed, 1996).The technology of microbial prospecting for oil and gas (MPOG) is based on this theory to forecast the existence of oil and gas deposits.In recent years, the microbial prospecting method was integrated with geological and geophysical methods to evaluate the hydrocarbon prospect of an area and to prioritize the drilling locations, thereby reducing drilling risks and achieving higher success in petroleum exploration (Rasheed et al., 2012;Wagner et al., 2002).
In general, the indicator bacteria for MPOG can be classified into two major groups: methane-oxidizing bacteria (methanotrophs) and C 2+ alkane-degrading bacteria.In the last few years, the abundance, distribution and community composition of methane-oxidizing bacteria have been relatively well studied in various methane seeps (Deutzmann et al., 2011;Håvelsrud et al., 2011;Kadnikov et al., 2012).Despite that methane is by far the most abundant hydrocarbon gas associated with petroleum in the reservoir, bacterial oxidation of C 2+ alkanes is of infinitely more value in petroleum prospecting than methane oxidation (Muyzer and van der Kraan, 2008).Since methane is also a common product of the anaerobic digestion of organic matter, it is obvious that the presence of methanotrophs in a soil sample is less likely to be indicative of leakage from a subsurface reservoir than the presence of the more specific C 2+ alkane-degrading bacteria (Shennan, 2006).
Therefore, recently, the ecological characteristics of alkane-degrading bacteria at various hydrocarbon macroseeps (active seeps with large concentrations of migrated hydrocarbons) have been surveyed, such as marine hydrocarbon seeps (Redmond et al., 2010;Wasmund et al., 2009) and oil spill zones (Valentine et al., 2010;Wang et al., 2011).These studies revealed that the presence of high concentrations of hydrocarbons significantly affects the biogeographical distribution and phylogenetic diversity of alkane-degrading bacteria in the near-surface waters and sediments.Nevertheless, the knowledge on the ecological characteristics of alkane-degrading bacteria at hydrocarbon microseeps (passive seeps with low concentrations of migrated hydrocarbons) is still lacking to date.It remains to be shown whether there is any difference in alkane-degrading community between the oil and gas field and the background soils, which is a critical issue for the success of MPOG.
Recent advances in microbial molecular biology have significantly improved our knowledge of the genes and enzymes associated with alkane metabolism as well as the microbiology of C 2+ alkane degraders.Alkane hydroxylase (alkB) is one of the key enzymes of the degradation of aliphatic alkanes under aerobic conditions (van Beilen and Funhoff, 2007).This enzyme is highly relevant and representative in aerobic oil-degrading bacteria (Rojo, 2009;Smits et al., 1999).In this study, by using alkB gene as the n-alkane-degrading biomarker, we chose a Chinese onshore oil and gas field as a hydrocarbon microseep model system in order to investigate the n-alkane-degrading community composition in oil and gas field soils and compare it to the compositions of the communities in the background soils by a combination of clone library and terminal restriction fragment length polymorphism (T-RFLP) analysis.The abundance of n-alkane degraders was determined by using quantitative real-time polymerase chain reaction (PCR).

Study site and sampling
Soil samples were collected from the Shaozhuang oil and gas field (37 • 7 N, 118 • 14 E) within the Shengli area, Shandong, China, in June 2011.The soil is classified based on the FAO (Food and Agriculture Organization of the United Nations) system as a stagno-fluvic Gleysol on loamy-sandy sediments over gley.The pH of surface soil (0-200 cm) in water (1 : 2) is in the range of ∼ 8.09-9.36.The underground gas and oil reservoir, covering 3.7 km 2 , is ∼ 1100-1270 m beneath the earth's surface.The oil-and gas-bearing bed is approximately 15 m thick.Soils were collected in pre-sterilized whirl pack bags under aseptic conditions from a depth of 60 cm.The detailed sample collection scheme is shown on the geological map (Fig. 1).Soil pH, water content, electrical conductivity, total headspace n-alkanes (C 2+ ) and total adsorbed n-alkanes (C 2+ ) are shown in Table 1.All of these chemical  properties were determined as previously described (Schumacher, 1996;Xu et al., 2010).

DNA extraction
DNA was extracted in duplicate from the soil samples using a modified procedure (Xu et al., 2010) and FastDNA Spin kit for soil (MP Biomedicals) according to the manufacturer's instructions.The products of duplicate extractions were pooled and used for further PCR amplification.

Cloning, sequencing and phylogenetic analysis
Three clone libraries of partial alkB genes were constructed to provide a better resolution for differentiating the individual T-RFs as phylogenetic lineages.Partial alkB genes were amplified using the forward primer alkBf (5 -AAYACNGCNCAYGARCTNGGNCAYAA-3 ) and the reverse primer alkBr (5 -GCRTGRTGRTCNGARTGNCGYTG-3 ) (Kloos et al., 2006).This primer set yields a PCR product of approximately 550 bp.A Mastercycler nexus PCR cycler (Eppendorf) with a TaKaRa PCR kit was used for the DNA amplification.The reaction mixture contained, in a total volume of 50 µL, 5 × PCR buffers, 4.0 mM MgCl 2 , 0.2 mM of each deoxyribonucleotide triphosphate (dNTP), 0.5 mM each primer, and 2 U of Ex-Taq DNA polymerase, and 1 µL (10-20 ng) of the genomic DNA was added as the template.The PCR program consists of an initial 5 min denaturation step at 95 • C, 30 cycles of repeated denaturation at 94 • C for 45 s, annealing at 53 • C for 45 s, and extension at 72 • C for 1 min, followed by final extension step of 10 min at 72 • C. Amplicons were gel-purified using a gel extraction kit (TaKaRa) according to the manufacturer's instructions.Ligation into pGEM-T Easy Vector (Promega) and transformation into Escherichia coli JM109 were performed according to the manufacturer's instructions.Clones were selected randomly and sequenced by ABI 377 DNA sequencer (Applied Biosystems).AlkB sequences were aligned manually with alkB sequences obtained from the GenBank database and checked for chimeras by bisecting and drawing two sub-phylogenetic trees from the bisects of each sequence.The sequences that showed different topologies among the two sub-trees were regarded as chimeric and removed from the libraries.Sequences are available under the GenBank accession numbers JX276475-JX276506.Nucleotide sequences from PCR products and reference strains were analyzed using the MEGA software (Tamura et al., 2011).The neighbor-joining method, with complete deletion of gaps and missing data and Poisson correction for multiple substitutions, was used to calculate the distances and to construct phylogenic trees.

Terminal restriction fragment length polymorphism analysis
For terminal restriction fragment (T-RF) profiles, partial alkB genes were amplified with 5 fluorescently labeled forward primer (alkBf-labeled with 6-carboxyfluorescein) and a reverse primer (alkBr).Two independent 25 µL PCRs were performed for each sample, and the products were combined and purified with gel extraction kit (TaKaRa).Aliquots of the purified amplicons were then digested in triplicate with 10 U of Msp I (TaKaRa) for 3 h at 37 • C. Each 200 µL tube contained 16 µL of amplicons, 2 µL of the incubation buffer, and 2 µL of restriction enzyme made up to a total volume of 20 µL.The digested amplicons were mixed with GeneScan 500 ROX size standards (Applied Biosystems) and analyzed by capillary electrophoresis.After electrophoresis, the sizes of the 5 -terminal restriction fragments (T-RF) and the intensities of their fluorescence emission signals were automatically calculated by the GeneScan analysis software (Applied Biosystems).Signals with a peak area that was less than 1000 relative fluorescence units were regarded as background noise and excluded from the analysis.The relative abundance of a detected terminal restriction fragment (T-RF) within a given T-RFLP pattern was calculated as the respective signal height of the peak divided by the peak height of all peaks of the T-RFLP pattern.The size of each n-alkanedegrading species T-RF peak corresponded to the value for that species determined by in silico analysis of clone library with EditSeq software (by searching for the first restriction enzyme site "C/CGG" of Msp I) of the DNASTAR package.All predicted and observed T-RF matches were within 2 bp from each other.Both the presence/absence and relative abundance of T-RFs were considered in data analysis.

Ordination analyses of T-RFLP fingerprints
The ordination analyses of T-RFLP profiles were performed using CANOCO 4.5 software (Microcomputer Power) as previously described (Rui et al., 2009).The detrended correspondence analysis (DCA) was firstly run to estimate the gradient length of variables.It was found that the longest gradient was shorter than 3.0.Thus, the principal component analysis (PCA) was chosen for analysis, because it performed better than the unimodal approaches under such conditions according to the CANOCO manual (Ter Braak and Smilauer, 1998).The settings of CANOCO 4.5 are as follows: inter-sample distance scaling, no post-transformation of scores, log data transformation (no offset), and center by species.

Real-time quantitative PCR
Copy numbers of partial alkB genes of n-alkane-degrading bacteria were quantified with the real-time quantitative PCR analysis (primer set alkBf-alkBr To evaluate the abundance of n-alkane-degrading bacteria relative to total bacteria, the percentages of alkB genes in proportion to 16S rRNA were also calculated.Bacterial 16S rRNA genes were quantified as previously described (Xu et al., 2009) with slight modification by using primer set 515f (5 -GTGCCAGCMGCCGCGG-3 ) and 907r (5 -CCGTCAATTCMTTTRAGTTT-3 ).The standard curves of 16S rRNA genes were made in a similar manner to alkB genes as described above.Real-time quantitative PCR was performed in triplicate with 25 µL reaction mixture containing 12.5 µL SYBR Premix Ex Taq, 0.5 µM of each primer, 11 µL of H 2 O and 1.0 µL of DNA template.

Data analysis
Statistical analysis of data was performed by ANOVA, with differences determined by the method of least significant differences at the 5 % level (P < 0.05).All statistical analyses were run with STATISTICA 6.0 (StatSoft).

Phylogenetic diversity of n-alkane-degrading bacteria
In order to assess the diversity of alkB at the phylogenetic level, three clone libraries representative of samples from oil field, gas field and the reference site were constructed (for sampling sites see Fig. 1).Approximately 83-95 % of the more abundant n-alkane-degrading phylotypes in the soil libraries were identified from S Chao1 and S ACE richness estimators (Fig. S1).Phylogenetic analyses of 136 clones revealed that the obtained alkB genes in this study exhibited 69-99 % similarity at the amino acid level with sequences retrieved from the GenBank database.More than half of clones showed highest affinity to either Acinetobacter strains or γproteobacterial marine hydrocarbon-degrading bacteria, such as Alcanivorax dieselolei and Marinobacter aquaeolei VT8.Interestingly, only a small part of clones were grouped in a large cluster that were most closely related to Gram-positive actinobacterial Rhodococcus, Mycobacterium, and Nocardia, which are well known for their ability to degrade shortchain and gaseous alkanes (Pérez- de-Mora et al., 2011;Shennan, 2006).A of three OTUs (operational taxonomic units) from the study zone formed a separate cluster (cluster A; Fig. 2) with very little sequence identity to alkB sequences from known organisms (only 71 % with Burkholderia cepacia) and other unknown uncultured bacteria (79 % identity with uncultured bacterium DQ288068).

n-alkane-degrading community structure above oil and gas reservoirs
In this study, we used T-RFLP fingerprinting based on alkB genes to characterize the n-alkane-degrading community structure above oil and gas reservoirs.Only those with an Ap (percentage abundance) higher than 1 % in at least one profile were selected as the signature T-RFs for further analyses.This cutoff was adopted from a previous study in order to minimize the amount of data noise (Noll et al., 2005).Seven fragments were detected as major peaks in the T-RFLP profiles (Fig. 2 As shown in Fig. 3, T-RFLP profiles revealed that longterm and continuous hydrocarbon supply considerably influenced the structure of n-alkane-degrading community.The most notable differences were the relative abundances rather than the fragment lengths.In the background soil, the 133 bp T-RF showed a high abundance (up to ∼ 32-45 %), indicating that the actinobacterial Mycobacterium and Rhodococcus were a predominant group on alkB gene level.By contrast, in the oil and gas field soils, the Actinobacteria became less frequent, and the niche vacated by this dominant group was gradually filled with proteobacterial Alcanivorax and Acinetobacter.The replacement of Proteobacteria was more pronounced in oil field soil as compared with gas field soil (Fig. 3).Comparing the T-RFLP results to alkB gene copy numbers determined by qPCR (Fig. S2), it also indicated that Proteobacteria are becoming more frequent in the oil and gas field (Table S1).
Principal component analyses (PCA) furthermore showed that the samples from oil field and gas field were clearly separated from those from reference site, although the difference between oil field and gas field was indiscernible (Fig. 4).Peaks 133 (putative actinobacterial Mycobacterium and Rhodococcus) had large loadings in the direction in which the non-oil and gas reference sample cluster was localized.On the contrary, in oil and gas cluster, Peaks 70 (putative Alcanivorax) and 74 (putative γ -proteobacterial Acinetobacter and Marinobacter) seemed to be positively correlated with the trace and continuous hydrocarbons.
We speculate that the trace amounts of hydrocarbon migrated from oil and gas reservoirs cause a shift of the nalkane-degrading bacterial community from Gram-positive bacteria to Gram-negative genotypes.As mentioned above, Gram-positive bacteria possess the ability to utilize shortchain and gaseous alkanes (C 1 -C 5 ) in contrast to other hydrocarbon-degrading bacteria (Pérez-de-Mora et al., 2011;Shennan, 2006).However, dominance of Actinobacteria has also been previously reported in hydrocarbon-unaffected soils (Luz et al., 2004;Margesin et al., 2003), possibly owing to their metabolic versatility, which includes n-alkanes and a wide range of organic compounds (Rojo, 2009).That is why Gram-positive Actinobacteria can be detected in both oil and gas field and background zones in the present study (Fig. 3).On the contrary, medium-and long-chain hydrocarbons (C 13+ ) have long been considered to be difficult to penetrate the cap rock of oil deposits and rise to the surface of the earth, thereby being ignored in most surface geochemical surveys.Recently, however, W. L. Gore & Associates, Inc. has successfully detected thermogenic hydrocarbon compounds from C 2 to C 20 in the near-surface soils of oil and gas reservoirs with a highly sensitive passive diffusion module (Anderson, 2006), indicating that volatile heavy hydrocarbons are able to migrate to the surface on a geologic time scale.Interestingly, Gram-negative Acinetobacter and Marinobacter are well-known for their ability to degrade medium-and long-chain n-alkanes ranging from C 13 to C 30   ( Doumenq et al., 2001;Sakai et al., 1994).The predominance of Gram-negative bacteria over Gram-positive bacteria was also observed in other alkane-rich habitats, such as marine hydrocarbon macroseeps (Wasmund et al., 2009) and heavy hydrocarbon-contaminated soils (Kaplan and Kitts, 2004).Therefore, Gram-negative n-alkane degraders seemed to be a good indicator population for MPOG.

Quantitative analysis of n-alkane-degrading bacteria
Real-time quantitative PCR was used to estimate the abundance of n-alkane-degrading bacteria at the different locations.In order to minimize environmental interference, such as noise caused by variations in soil type, we normalized the abundances of alkB genes against the total abundance of 16S rRNA genes, sample-by-sample.The biogeographical distribution of the relative abundances of n-alkane-degrading bacteria in soil samples collected from east-west direction survey line is shown in Fig. 5 (the absolute numbers of alkB gene were shown in Fig. S2).In the oil and gas field soil samples, the relative abundances of alkB genes were significantly higher than in all other soil samples (4.7 % ± 0.3 % and 7.4 % ± 1.1 %, respectively).In the reference samples, the ratios were 2.2-to 3.4-fold lower, in the range from 0.7 % to 4.6 %.The high value area was basically consistent with the areas of the oil and gas accumulation (Fig. 5).However, to our surprise, hydrocarbon concentrations in soils above oil and gas fields were relatively low or even undetectable (Table 1).We speculate that efficient biodegradation of hydrocarbons was occurring and that this degradation removed detectable quantities of hydrocarbons before they could be measured by geochemical analyses.A similar effect has been previously reported in a study of marine hydro-    carbon macroseeps (visible seeps) (Wasmund et al., 2009).Therefore, the quantification of alkB gene copy numbers in soils provides an insight into the microbial response to the microseepage of hydrocarbons and acts as a useful complementary tool for understanding this habitat response to trace hydrocarbons in addition to geochemical measurements (Wasmund et al., 2009).
In the preliminary study, we tried to isolate and enumerate n-alkane-degrading bacteria using standard plate count method (data not shown).However, quantitative distribution of culturable species was confused and cannot be used to forecast oil and gas reservoirs.This phenomenon might be caused by (i) the existence of a large quantity of uncultured alkane degraders in soils (Kloos et al., 2006)   contradiction between short-term culture (days) in laboratory under high-concentration alkanes and long-term acclimation (years) to trace alkanes in natural environments.Therefore, culture-independent approaches using alkB gene to detect the presence and abundance of n-alkane-degrading bacteria taken directly from soil samples without cultivation can significantly improve the accuracy rate of MPOG.

Conclusions
In this study, the diversity and abundance of n-alkanedegrading bacterial community in the near-surface soils of a onshore oil and gas field were investigated using molecular techniques.The determination of alkB gene-based T-RFLP profiles and subsequent affiliation to clone sequences and PCA ordination showed that trace amounts of hydrocarbons migrated from oil and gas fields influenced not only the quantity but also the structure of n-alkane-degrading bacterial community.The predominance of Gram-negative Proteobacteria (Alcanivorax and Acinetobacter) over Gram-positive Actinobacteria (Mycobacterium and Rhodococcus) was observed in oil and gas field soils.Real-time PCR results furthermore showed that the abundance of alkB genes increased substantially in the surface soils above oil and gas reservoirs even though only low concentrations of hydrocarbons were measured in these soils.

Figure 1 .
Figure 1.Geological base of Shaozhuang Oil and Gas Field and sampling scheme for spatial analysis of n-alkane degrading bacterial community.Sampling positions for clone library analyses of alkane hydroxylase (alkB) gene diversity at three different sites are marked with green circles.Small blue circles indicate the soil samples for T-RFLP and qPCR analysis.The thinner black lines and related negative 4 digits number are contour lines.Soils were collected at a depth of 60 cm.

Fig. 1 .
Fig. 1.Geological base of Shaozhuang oil and gas field and sampling scheme for spatial analysis of n-alkane-degrading bacterial community.Sampling positions for clone library analyses of alkane hydroxylase (alkB) gene diversity at three different sites are marked with green circles.Small blue circles indicate the soil samples for T-RFLP and qPCR analysis.The thinner black lines and related negative 4-digit numbers are contour lines.Soils were collected at a depth of 60 cm.

Figure 2 .
Figure 2. Phylogenetic relationship of deduced alkB sequences (182 amino acids) generated from different soil samples.The scale bar represents 10% sequence divergence; values at the nodes are the percentages of 1000 bootstrap replicates supporting the branching order; bootstrap values below 50% are not shown.Representative clones obtained in this study are shown in different colors which represent different n-alkane degrading groups, and the in silico T-RF size is given in square brackets.A similarity cutoff of 97% is used to define an OTU (Operational Taxonomic Unit).The number of clones for each representative OTU is given in parentheses (reference soil/oilfield surface soil/gasfield surface soil).Reference sequences are shown in black.

Fig. 2 .Figure 3 .
Fig. 2. Phylogenetic relationship of deduced alkB sequences (182 amino acids) generated from different soil samples.The scale bar represents 10 % sequence divergence; values at the nodes are the percentages of 1000 bootstrap replicates supporting the branching order; bootstrap values below 50 % are not shown.Representative clones obtained in this study are shown in different colors, which represent different n-alkane-degrading groups, and the in silico T-RF size is given in square brackets.A similarity cutoff of 97 % is used to define an OTU (operational taxonomic unit).The number of clones for each representative OTU is given in parentheses (reference soil/oil field surface soil/gas field surface soil).Reference sequences are shown in black.

Figure 4 .
Figure 4. PCA ordination plot for the samples and T-RFs based on alkB T-RFLP data from different soil samples.Symbols: red arrows = T-RFs (the size in base pairs of forward T-RFs by Msp I); green triangles, blue squares, black circles and associated numbers indicate oilfield surface soils, gasfield surface soils, background soils and sample serial numbers, respectively.

Fig. 4 .
Fig. 4. PCA ordination plot for the samples and T-RFs based on alkB T-RFLP data from different soil samples.Symbols: red arrows = T-RFs (the size in base pairs of forward T-RFs by Msp I); green triangles, blue squares, black circles and associated numbers indicate oil field surface soils, gas field surface soils, background soils and sample serial numbers, respectively.

Figure 5 .
Figure 5. Biogeographical distribution of the percentages of alkB gene copy numbers (representing the number of n-alkane degrading bacteria) in proportion to 16S rRNA gene copy numbers (representing the number of total Eubacteria) of soil samples collected from east-west direction survey line of Shaozhuang Oil and Gas Field (n=3).Soils were collected at a depth of 60 cm.

Fig. 5 .
Fig. 5. Biogeographical distribution of the percentages of alkB gene copy numbers (representing the number of n-alkane-degrading bacteria) in proportion to 16S rRNA gene copy numbers (representing the number of total Eubacteria) of soil samples collected from eastwest direction survey line of Shaozhuang oil and gas field (n = 3).Soils were collected at a depth of 60 cm.
such as the 142 and 340 bp T-RFs in our case (Fig.2) and/or (ii) the

Table 1 .
General soil properties, headspace and adsorbed n-alkanes content of research area * .
* Soils were collected at a depth of 60 cm.