The Use of a Combination of alkB Primers to Better Characterize the Distribution of Alkane-Degrading Bacteria

The alkane monooxygenase AlkB, which is encoded by the alkB gene, is a key enzyme involved in bacterial alkane degradation. To study the alkB gene within bacterial communities, researchers need to be aware of the variations in alkB nucleotide sequences; a failure to consider the sequence variations results in the low representation of the diversity and richness of alkane-degrading bacteria. To minimize this shortcoming, the use of a combination of three alkB-targeting primers to enhance the detection of the alkB gene in previously isolated alkane-degrading bacteria was proposed. Using this approach, alkB-related PCR products were detected in 79% of the strains tested. Furthermore, the chosen set of primers was used to study alkB richness and diversity in different soils sampled in Carmópolis, Brazil and King George Island, Antarctica. The DNA extracted from the different soils was PCR amplified with each set of alkB-targeting primers, and clone libraries were constructed, sequenced and analyzed. A total of 255 alkB phylotypes were detected. Venn diagram analyses revealed that only low numbers of alkB phylotypes were shared among the different libraries derived from each primer pair. Therefore, the combination of three alkB-targeting primers enhanced the richness of alkB phylotypes detected in the different soils by 45% to 139%, when compared to the use of a single alkB-targeting primer. In addition, a dendrogram analysis and beta diversity comparison of the alkB composition showed that each of the sampling sites studied had a particular set of alkane-degrading bacteria. The use of a combination of alkB primers was an efficient strategy for enhancing the detection of the alkB gene in cultivable bacteria and for better characterizing the distribution of alkane-degrading bacteria in different soil environments.


Introduction
Alkane-utilizing bacteria are widespread in marine and terrestrial environments [1], [2]. These bacteria generally possess the Alk enzyme system, which is involved in the metabolic pathway for the degradation of alkanes, the main compounds found in petroleum and its derivatives [3]. The functional Alk enzyme system comprises the transmembrane alkane monooxygenase AlkB (encoded by the alkB gene and involved in the initial activation step of aerobic aliphatic hydrocarbon metabolism) and two co-factors named rubredoxin (alkF) and rubredoxin reductase (alkG). These co-factors are responsible for transferring the electrons involved in alkane hydroxylation by AlkB [4]. In other studies, the alkB gene has been used as a biomarker for the determination of the abundance and diversity of alkane-degrading bacteria [527].
Bacteria that possess the Alk enzyme system are valuable in environmental bioremediation and biocatalysis for the synthesis of industrial compounds, including drugs, pravastatin, and other compounds [8]. The use of alkane-degrading bacteria in bioprocesses to produce valuable chemicals by transforming alkanes from hydrocarbon-contaminated samples is considered one of the most important biotechnological applications of these bacteria [9]. Therefore, many studies report the detection and further characterization of alkB genes in a wide variety of bacterial genera [4], [10], [11].
However, the alkB genes characterized thus far may only represent a small fraction of the diversity found in natural environments and a limited biotechnological potential of alkanedegrading bacteria. Kuhn and co-workers [5] found new alkB genes when contaminated and uncontaminated marine sediments in Admiralty Bay, King George Island, Antarctica were analyzed. Similar results were obtained for samples from the Timor Sea in Australia [12] and from chronically polluted, sub Antarctic coastal sediments [13].
Although all AlkB proteins share considerable sequence homology, the nucleotide sequences encoding the alkB gene vary widely within the alkB-containing bacteria. To overcome this limitation and study the alkB diversity in natural environments, different alkB primers have been described [5], [14], [15]. However, each set of primers is specific for primarily one group of bacteria [15217], and designing broad-ranging alkB primers is not an easy task. Consequently, the presence and diversity of alkB sequences have likely been underestimated in the environmental samples previously studied.
In this study, we propose to use a combination of primers, rather than designing new primers, to improve the identification of the alkB diversity in different environments. For this purpose, we retrieved several pairs of alkB-targeting primers from the literature and tested their specificities against alkane-degrading bacteria previously isolated by our research group. These bacteria are representatives of Proteobacteria, Actinobacteria, Flavobacteria and Firmicutes groups. After selecting the combination of alkBtargeting primers showing the broadest coverage among the bacterial strains tested, we constructed alkB clone libraries using an oil-contaminated semiarid soil from Carmópolis, located in Sergipe (SE), Brazil to test the approach proposed here in the environment. Moreover, this strategy was also used to understand the alkB diversity in three pristine soil samples and one dieselcontaminated soil sample from King George Island, Antarctica. The data obtained suggest that the alkB diversity in soil environments may be higher than previously described.

alkB-targeting Primers
The list of alkB-targeting primer pairs used is presented in Table 1. The PCR reactions were performed under the conditions previously described for each pair of primers (Table 1).
Bacterial strains, test of bacterial growth using heptadecane as the sole carbon source and bacterial identification.
The bacterial strains used (Table 2) were isolated and described in previous studies [18220]. The ability of these isolated strains to use heptadecane as the sole carbon source was determined as described by Alvarez et al. [19]. The 16S rRNA gene sequences were used for the identification of the bacterial strains. Genomic DNA was extracted using a protocol described by Pitcher et al. [21]. BOX-PCR was used to cluster the bacterial strains according to Versalovic et al. [22], and one representative strain from each BOX-PCR group was selected for 16S rRNA-based molecular identification. PCR amplification of the 16S rRNA coding gene and the molecular sequencing methodologies were performed as described in Alvarez et al. [19]. The partial 16S rRNA gene sequences (,800 bp) were identified using the BLAST-N tool (blast.ncbi.nlm.nih.gov) on the National Center for Biotechnology Information (NCBI) website using the GenBank non-redundant database.

Selection and Combined Use of alkB-targeting Primers to Study the alkB Diversity in Different Soils
The alkB-targeting primers were chosen based on the combination of primers that allowed for the detection of the alkB gene in as many bacterial strains tested as possible. Primers were combined manually in pairs and triplets, and the number of bacterial strains detected by the combination of primers used was determined.
To test the approach suggested here (the use of a combination of alkB-targeting primers) to better describe the alkB diversity in soil environments, we constructed clone libraries using individually the chosen primers and the DNA extracted from contaminated soil samples obtained in Carmópolis (denoted as sC throughout the manuscript), SE, Brazil, where the majority of the bacterial strains presented in Table 2 were isolated [18220]. Furthermore, this strategy was extended to study the alkB diversity present in dieselcontaminated soil (denoted as s3) and three uncontaminated (pristine) soil samples (sY, sI and sR) from King George Island, Antarctica (Fig. S1). These Antarctic soils were chosen because previous studies have suggested that the soils may contain an  underestimated diversity of genes coding for AlkB enzymes or even new alkB coding genes [20], [23]. Moreover, the soil samples were selected based on their different chemical and physicochemical properties ( Table 3). All samplings were performed in triplicate, and the soil samples were kept at 220uC until DNA extraction. All necessary permits were obtained for the soil samplings through the research projects funded by Petrobras (Carmópolis soil) and the Brazilian Antarctic Program (Antarctic soils).

DNA Extraction and Clone Library Construction
Total DNA was extracted directly from 0.5 g of soil using the Fast DNA Spin Kit for soil (QBIOgene, Carlsbad, CA) following the manufacturer's instructions. To improve the coverage of the results obtained here, DNA extractions and PCR reactions were performed in triplicate for each sampling site.
To construct the clone libraries, the fragments of the alkB gene were PCR amplified using DNA from the soils described above and with the primers (f), (e) and (d) described by Chénier et al. [24], Kloos et al. [14] and Kuhn et al. [5], respectively (Table 1). PCR amplification followed the conditions previously described for each pair of primers (Table 1), and the 25 ml-PCR reaction mix comprised 1 ml of template DNA (30-50 ng), 0.5 pmol of primers, 0.2 mM of each dNTP, 5 ml of 5X PCR buffer (100 mM Tris-HCl, pH 9.0, and 500 mM KCl), 1.5 mM MgCl 2 and 1 U of Taq DNA polymerase (Promega, Madison, WI, USA).
The PCR amplification products were used to construct the clone libraries. Before cloning procedures, the PCR products obtained from the different triplicates of each sampling site were pooled and purified using the Wizard SV Gel and PCR Clean-up System (Promega). Purified amplicons were then cloned using the InsTAclone PCR Cloning Kit (Fermentas, Maryland, USA) following the supplier's instructions. The insert-containing clones were sequenced using the forward primer M13F (59-GTA AAA CGA CGG CCA GT-39) in the vector pTZ57R/T on an ABI Prism 3100 automatic sequencer (Applied Biosystems Inc., CA, USA) using Macrogen (South Korea) facilities.

Sequence Analysis
The electropherogram files generated by sequencing were analyzed using the Phred program [25] for base calling and trimming of vector and low-quality (,20) sequences. Vector contamination and primer sequences were removed manually using Bioedit software (Ibis Biosciences Inc., CA, USA). To analyze only the overlapping fragments of the alkB gene amplified from all pairs of primer used, the obtained sequences were aligned using the package software Clustal X [26] and edited using Bioedit. Only the overlapping regions of alkB fragments were used in the following steps. The MOTHUR software [27] was used to classify the alkB genes into operational taxonomic units (OTUs) with 97% similarity. Next, the OTU-generated matrices were used to calculate the species richness using Chao1 estimators [28] and the Shannon-Weaver diversity index [29]. Coverage (C) was also calculated, where C equals 12n1/N, and n1/N is the ratio of clones that appeared only once (n1) to the total number of clones (N) [30]. Boneh's estimator was used to estimate the number of additional OTUs that would be observed if an additional sampling of clones would have been performed [31]. Finally, the diversity of OTUs and those found in common after the amplification with the different primers used were examined using rarefaction analysis and Venn diagrams.

Richness of alkB Phylotypes
The increase in alkB richness as a result of the use of the combination of alkB-targeting primers was calculated based on Venn diagram results. The calculation was performed as follows: (lR1+lR2)/hR, where the hR corresponds to the richness of alkB phylotypes detected only in the richest clone library, and (lR1+ lR2) represents the sum of the richness of alkB phylotypes observed in the other libraries (but not in hR).

Phylogenetic Analyses of alkB Phylotypes
The representatives of each OTU (at a distance level of 3%) obtained from the analyses of the clone libraries were taxonomically assigned using the BLAST-n tool on the NCBI website and the GenBank non-redundant database. A phylogenetic tree was constructed with representatives of each OTU found within the libraries and with closely related sequences that were recovered from the GenBank database. Sequence alignment was performed by Clustal X software [26], and the aligned sequences were then used to construct the phylogenetic tree using the neighbor-joining method in the MEGA 5 software [32]. Bootstrap analyses were performed with 1,000 repetitions, and only values higher than 99% are shown in the phylogenetic tree.
The nucleotide sequences were translated using the Transeq tool (http://www.ebi.ac.uk/Tools/emboss/transeq/) on the European Bioinformatics Institute website, and deduced amino acids were directly compared with the Protein Database using the algorithm BLASTP from NCBI to check if the sequenced alkB genes possess the conserved motifs common to AlkB proteins.

Clone Library Comparisons
The structure of the alkB phylotypes in each library was compared. For the comparison, a dendrogram describing the dissimilarity (1-similarity) among the clone libraries from each sampling site was clustered using the UPGMA algorithm and the Jaccard similarity coefficient based on the observed richness. To perform UniFrac-based library comparisons [33], neighborhoodjoining trees were constructed using the MEGA 5 software [32]. An unweighted UniFrac significance test was used to estimate whether the clone libraries corresponding to each soil sample were significantly different. One hundred permutations were performed, and the P-values were corrected for multiple comparisons using the Bonferroni correction [34].

Nucleotide Sequence Accession Numbers
The sequences obtained from the clone libraries were deposited in the GenBank database with the following accession numbers: KC733460-KC733713.

Growth and Identification of the Bacterial Strains Using Heptadecane as the Sole Carbon Source
From 85 oil-degrading bacteria previously described in von der Weid et al. [18], Alvarez et al. [19] and Jurelevicius et al. [20], 64 were able to use heptadecane as the sole carbon source (an example is shown in Fig. S2). BOX-PCR clustered these bacteria into 43 different groups (data not shown), and a representative strain of each BOX group was further identified through 16S rRNA gene sequencing ( Table 2). The results showed that the bacterial strains are distributed amongst Actinomycetales, Firmicutes, Flavobacteria and Alpha, Beta and Gamma proteobacteria ( Table 2).

Selection of the alkB-targeting Primers
The amplification range of eight different alkB-targeting primers previously described (Table 1) was determined by direct PCR amplification using the DNA of the alkane-degrading bacterial strains listed in Table 2. The use of each primer pair resulted in different amplification patterns. Using primers (a), (b), (c) and (g), only 18.6 to 20.9% of the bacterial strains tested were amplified. The broadest range of amplification (48.8% of the strains tested) Figure 1. Rarefaction analysis of alkB clone libraries using a distance level of 97% similarity. The clone libraries are denoted as follows: the letters (d), (e) and (f) correspond to the alkB-targeting primers described in Table 1, and sI, sR, sY, s3 and sC are the sampling site codes (described in Materials and Methods). doi:10.1371/journal.pone.0066565.g001 was observed with primer pair (e), and primers (h) and (f) produced amplification products in 44.2% of the strains tested (Table 2). Primer pair (d) resulted in the amplification of the alkB gene in 23.3% of the strains tested, and most of these strains were identified as Acinetobacter (Table 2). However, the results showed that none of the primers used were highly specific to one particular phylogenetic group. PCR amplification also varied among strains from the same species, depending on the set of primers used (Table 2). Finally, the alkB gene was not detected in 7 (16.3%) of the alkane-degrading bacterial strains tested. These strains belong to the genera Cupriavidus, Bacillus, Gordonia, Bosea and Rhizobium (Table 2).
To enhance the detection of the alkB gene in previously isolated alkane-degrading bacteria, the alkB-targeting primers were combined into pairs (28 combinations) or triplets (56 combinations). The combination of primers selected allowed for the detection of the alkB gene in as many bacterial strains tested as possible (Table 2). Therefore, primers (d), (e) and (f) were chosen as they covered 79% of the alkane-degrading bacterial strains tested. The addition of a fourth pair of primers did not improve the detection range achieved with the use of the three sets of primers ( Table 2).

Validation of the Use of the Selected alkB-targeting Primers to Study the alkB Diversity in Soil DNA Samples
The three selected alkB-targeting primers were used to amplify alkB genes present in an oil-contaminated soil sample from Carmópolis, SE, Brazil, which was the source of the majority of the strains tested. Additionally, the same strategy was used to amplify alkB genes in one diesel-contaminated site and three pristine soil samples from King George Island, Antarctica. alkB genes were detected in all sampling sites, except for site sY when primer pair (d) was used.
Clone libraries were used to describe the alkB phylotypes that resulted from the amplification by each pair of primers. Only the overlapping regions of alkB fragments amplified by the primers chosen were selected and used. The alkB genes were clustered into OTU groups (referred to as alkB phylotypes) that were defined as sequences with more than 97% similarity. Rarefaction curves indicated that the number of clones screened from all soil samples was sufficient to reveal the majority of alkB phylotypes within the community because the estimates of sequence types tended to reach a plateau using 97% sequence identity (Fig. 1). The data obtained from the statistical analyses showed that the clone libraries covered 65 to 95% of the alkB phylotypes that resulted from the PCR amplification using each pair of primers and the different soil samples (Table 4). Boneh's estimator showed that a new sequencing effort would increase the alkB phylotypes to a maximum of nine new alkB phylotypes (Table 4). Considering the results obtained from the sC sample, the highest richness and diversity of alkB phylotypes were detected by using primer pair (e) ( Table 4). Considering the soil sampling sites s3 and sR from King George Island, the results also showed that the highest richness of alkB phylotypes was observed in the clone libraries constructed with PCR amplification products obtained using primer pair (e), followed by the use of (f) and (d) ( Table 4). In the sampling site sI, the highest richness was observed in the clone libraries constructed with PCR amplification products obtained with the use of primer (d), followed by those from primers (f) and (e); the highest richness of alkB phylotypes in sampling site sY was observed with the use of primer pair (f) ( Table 4).

Improved Detection of the Richness of alkB Phylotypes
Chao richness-based Venn diagrams showed the shared presence of alkB phylotypes detected by each alkB-targeting primer and from each sampling site (Fig. 2). The results showed Figure 2. A Venn diagram of alkB clone libraries using a distance level of 97% similarity. The numbers represent the richness and the shared richness of each library using the Chao richness index [28]. The Venn diagram groups are denoted as follows: sI, sR, sY, s3 and sC represent the sampling sites, and the letters (d), (e) and (f) correspond to the alkB-targeting primers used in the PCR amplification of the alkB gene, as described in Table 2. doi:10.1371/journal.pone.0066565.g002 Figure 3. Phylogenetic tree of the alkB sequences obtained from sI, sR, sY, s3 and sC soil libraries and from the closely related alkB that the clone libraries generated by each primer pair shared low numbers of common alkB phylotypes (Fig. 2). Subsequently, the results from the Venn diagrams were used to calculate the gain in alkB phylotype richness by using the combined alkB-targeting primer strategy. The use of this strategy to describe the alkB diversity in sC soil resulted in a 47% increase in alkB richness (Fig. 2, Fig. S3). The same strategy used to study the alkB diversity in soils from King George Island, Antarctica, resulted in an increase of 45 to 139% in alkB richness (Fig. 2, Fig. S3).

Phylogenetic Analyses of alkB Genes
A total of 254 OTUs (referred to as alkB phylotypes) were obtained in this study. The phylogenetic analyses of the alkB phylotypes showed the detection of sequences sharing 64 to 100% identity with alkB sequences previously deposited in GenBank. However, from the 254 alkB phylotypes detected here, only 20% of the sequences showed identities 90% or higher with known sequences (35% of the sequences shared identities of 80% or lower with known sequences). Nucleotide translating analysis followed by deduced amino acid analysis using BLAST-P tools showed that only one alkB phylotype obtained did not encode a potentially functional AlkB. A stop codon was detected in the nucleotide sequence of this phylotype, and consequently this phylotype was excluded from further analyses. All other alkB phylotypes had conserved amino acid motifs found in functional AlkB enzymes.
Phylogenetic analyses of the alkB phylotypes showed that each of the alkB-targeting primers chosen was not specific to any monophyletic group of alkB genes because the primers were able to anneal to alkB gene sequences from diverse phylogenetic groups (Fig. 3A, B). In sampling site sC, the most representative alkB phylotypes (corresponding to 30% and 13.4% of the clone libraries) were related to alkB from Mycobacterium chubuense NBB4 (with 92% sequence identity) and to alkB from Acidisphaera sp. C197 (81% identity), respectively.
From the diesel-contaminated soil (s3) sampled in King George Island, the most abundant alkB phylotypes (29.7%) shared high identity (99%) with the alkB gene described in Rhodococcus sp. 28/ 19, an oil-degrading strain previously isolated from soils in Antarctica. Other alkB phylotypes obtained from the s3 sampling site were related to an alkB gene from non-cultivated bacteria detected in Arctic contaminated soils (99% identity) and to alkB described in a strain from the Acetobacteraceae family (89% identity). These alkB phylotypes correspond to 14.8% and 12.9% of the s3 clone library, respectively (Fig. 3A).
From the uncontaminated soils sampled in King George Island, most of the alkB phylotypes detected at sampling site sY (corresponding to 73.1% of the clones) were related to an alkB gene described in Pedobacter sp. MS245e (78 to 85% identity). Additionally, phylotypes related to alkB from Bacillus sp. MS238f (88% identity) corresponded to 23.3% of the sY clone library. In sampling site sI, the most abundant alkB phylotype (20.7%) shared 87% identity with alkB described in Rhodococcus opacus B4. Other phylotypes were associated with alkB genes from Nocardia brasiliensis ATCC 700358 (79% identity) and Acidovorax sp. KKS102 (73% identity), representing 13.6% and 9.5%, respectively, of the alkB phylotypes found at sampling site sI. In sample site sR, the most abundant alkB phylotype (15.1%) was related to an alkB gene from Rhodococcus erythropolis (77% identity). Other alkB phylotypes previously found in Pseudomonas aeruginosa PG201 and Pedobacter sp. MS245e represented 10.8% and 8.9%, respectively, of the clone library obtained from sampling site sR (Fig. 3A).

Comparison among Libraries
The alkB diversity was used to further analyze the structure of alkane-degrading bacteria present in each sampling site. In all cases, the different soil samples were clustered separately in the dendrogram analyses with less than 40% similarity (Fig. 4). Low numbers of alkB phylotypes were shared between two different sampling sites, and common alkB phylotypes were not found in more than two sampling sites (Table S1, Fig. 4 B, C). For example, the alkB phylotypes shared among sampling sites s3 and sI were related to an alkB gene from the Acetobacteraceae family (89% sequence identity) and represented only 12.9% and 2%, respectively, of the phylotypes found in these soils. Finally, qualitative corresponding colors for the different sampling sites are: red (sI), blue (sR), yellow (sY), black (s3) and purple (sC). (B) The same phylogenetic tree showing the distribution of the alkB phylotypes resulting from the amplification with the different targeting primers. The colors corresponding to the primers used are: yellow -(d), blue -(e) and red -(f). The color white was used for phylotypes that originated from the PCR amplification with more than one primer. doi:10.1371/journal.pone.0066565.g003 Figure 4. Dendrogram describing the dissimilarity (1-similarity) among the sampling sites (A). The groups were clustered using the UPGMA algorithm and the Jaccard similarity coefficient based on the observed richness. The clone libraries are denoted as follows: the letters (d), (e) and (f) correspond to the alkB-targeting primers described in Table 2, and the sampling sites (sI, sR, sY, s3 and sC) are described in the Materials and Methods. (B) and (C) Venn diagrams of all alkB phylotypes detected in each sampling site using a distance level of 97% similarity. doi:10.1371/journal.pone.0066565.g004 beta-diversity measures using unweighted UniFrac analysis confirmed the differences observed among the alkB clone libraries obtained by each sampling point studied here (p,0.001).

Discussion
The importance of the alkane monooxygenase AlkB in bioremediation processes of hydrocarbon-contaminated environments and in biocatalysis for the production of useful compounds justifies the worldwide interest in the distribution of the alkB gene in different environments [3], [35]. However, the current literature shows that the diversity of the alkB gene in environmental samples is still far from being well characterized [5], [6], [13], [14].
In this study, we show that the use of a single pair of primers for the PCR amplification of the alkB gene in soil environments as well as in isolated alkane-degrading bacteria limits the range of detection of this gene. However, the use of a combination of alkB-targeting primers results in a more reliable detection of the alkB gene in alkane-degrading bacteria from different soil samples. Even with considering the best combination of alkB-targeting primers, the presence of alkB genes was not detected in 21% of the bacterial strains tested. In a few of these strains, homologous alkB sequences have been previously detected by dot blot hybridization [20], [36]. Therefore, enzymes related to cytochrome P450 or others [10], [35] could be responsible for alkane degradation in the remaining strains. Nevertheless, if only the total number (79%) of strains detected with the combination of alkB-targeting primers is considered, our results indicate that 30.2% or more of alkBpossessing bacteria would be missed from environmental analyses with the use of individual alkB-targeting primers.
The use of different alkB-targeting primers to detect alkB genes in different bacterial isolates and also in environmental samples had been used before through dot blot hybridization and qPCR analyses [1], [6], [16], [37240]. However, no studies have focused on the coverage of each primer and the benefit of using the different primers pairs to increase the diversity and richness of alkB phylotypes. Our results demonstrated that the use of a combination of alkB-targeting primers resulted in up to 139% increase in the richness of alkB gene phylotypes obtained in the soils used here. Moreover, the analysis of the clone libraries showed that the richness of alkB phylotypes resulting from the amplification using each of the chosen primers was dependent on the sampling site. Although most of our results indicated that primer pair (e) designed by Kloos et al. [14] generated the most diverse clone libraries in the samples studied, the use of primer pair (d) [5] resulted in the highest diversity and richness at sampling site sI. In contrast, the lowest diversity and richness were detected at this site using primer pair (e). The chemical and physicochemical properties of the soils, such as pH, TPH, organic matter and/or plant litter, may have influenced the diversity and richness of alkB phylotypes, as suggested by other studies [17], [23], [39], [41].
Although the sampling sites used in this study have been studied before [20], [23], this study is the first to perform a broader alkB sequencing analysis in these soils. Interestingly, our results suggest that different alkB phylotypes were selected depending on the sampling site analyzed, and no phylotype was shared between more than two different King George Island sampling sites (Fig. 4 B, C). In contrast with the results obtained by Powell et al. [17] who studied the alkane-degrading bacteria present on sub-Antarctic Macquarie Island, the sampling sites used here (with their own chemical and physicochemical properties) were sufficient to determine the alkB-containing bacterial community structure (Table 3, Fig. 4 A, B, C). The observation that alkB clone libraries obtained from each sampling point were considerably different from each other (Fig. 4A) in addition to the results obtained from the qualitative beta-diversity measures UniFrac analysis (p,0.001) corroborate the above statement.
The low identities observed among the alkB genes from the different environments studied here (Fig. 3) may indicate an alkB gene diversity yet uncharacterized in natural environments, as suggested previously [5], [6], [13], [14], [23]. Also, the low specificity of the primers used in this study for any bacterial phylogenetic group may be explained by the occurrence of horizontal gene transfer among alkB-containing bacteria [10]. Finally, the alkB phylotypes detected here possess conserved amino acid motifs present in functional AlkB enzymes, suggesting their functionality in natural environments. As the properties of the enzymes encoded by these alkB genes are still uncharacterized in soils from Carmópolis and Antarctica, the isolation and characterization of bacteria harboring these genes are still necessary. Figure S1 Map of Antarctic continent adapted from Jurelevicius et al. [23]. King George Island, the biggest island of the South Shetland archipelago, is shown together with the sample sites (indicated by arrows). (TIF) Figure S2 Growth of some isolated strains using heptadecane as the sole carbon source. The columns represent different bacterial strains (1 to 5 -strains Br_O 3B, Cri_O 3, Ar_lB 45B, Bri_O 51 and Ar_lB 50B, respectively, and 6 -negative control), and the rows represent (A) the negative control where the strains were inoculated in mineral medium (Bushnell Haas) and (B) the strains were inoculated in mineral medium added with heptadecane (0.1% v/v) as the sole carbon source. (TIF) Figure S3 ichness of alkB phylotypes observed in each clone library and also in all clone libraries (sum) of each sampling site. The clone libraries are denoted as follows: the letters d, e and f correspond to the alkB-targeting primers as described in Table 1 and sI, sR, sY, s3 and sC correspond to the sampling sites as described in Materials and Methods. (TIF)