Occurrence and diversity of naphthalene dioxygenase genes in soil microbial communities from the Maritime Antarctic


  • Cecilia G. Flocco,

    Corresponding author
    1. Julius Kühn-Institute, Federal Research Centre for Cultivated Plants (JKI), Messeweg 11/12, D-38104 Braunschweig, Germany.
    2. Consejo Nacional de Investigaciones Científicas y Tecnológicas (CONICET), Buenos Aires, Argentina.
    Search for more papers by this author
  • Newton C. M. Gomes,

    1. CESAM, Departamento de Biologia, Universidade de Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal.
    Search for more papers by this author
  • Walter Mac Cormack,

    1. Instituto Antártico Argentino, Cerrito 1428 (C1010AAZ) Buenos Aires, Argentina.
    Search for more papers by this author
  • Kornelia Smalla

    1. Julius Kühn-Institute, Federal Research Centre for Cultivated Plants (JKI), Messeweg 11/12, D-38104 Braunschweig, Germany.
    Search for more papers by this author

*E-mail flocco@ffyb.uba.ar; Tel. (+49) 5312993814; Fax (+49) 5312993013.


The diversity of naphthalene dioxygenase genes (ndo) in soil environments from the Maritime Antarctic was assessed, dissecting as well the influence of the two vascular plants that grow in the Antarctic: Deschampsia antarctica and Colobanthus quitensis. Total community DNA was extracted from bulk and rhizosphere soil samples from Jubany station and Potter Peninsula, South Shetland Islands. ndo genes were amplified by a nested PCR and analysed by denaturant gradient gel electrophoresis approach (PCR-DGGE) and cloning and sequencing. The ndo-DGGE fingerprints of oil-contaminated soil samples showed even and reproducible patterns, composed of four dominant bands. The presence of vascular plants did not change the relative abundance of ndo genotypes compared with bulk soil. For non-contaminated sites, amplicons were not obtained for all replicates and the variability among the fingerprints was comparatively higher, likely reflecting a lower abundance of ndo genes. The phylogenetic analyses showed that all sequences were affiliated to the nahAc genes closely related to those described for Pseudomonas species and related mobile genetic elements. This study revealed that a microdiversity of nahAc-like genes exists in microbial communities of Antarctic soils and quantitative PCR indicated that their relative abundance was increased in response to anthropogenic sources of pollution.


Polycyclic aromatic hydrocarbons (PAH) are a class of ubiquitous hydrophobic organic pollutants (Skupinska et al., 2004). They arise from natural sources (such as forest fires and volcanic activity) as well as from anthropogenic activities, being the combustion of fossil fuels and some industrial processes the main input into the environment (Shuttleworth and Cerniglia, 1995). Pollution events caused by PAH have been reported all around the world (Gioia et al., 2006; Gomes et al., 2007; Iqbal et al., 2007; Zhang et al., 2008). Antarctica, although largely considered as a pristine region, is not free of this type of pollution. The contamination with PAH (derived mostly from petroleum hydrocarbons) is caused by human activities in the Antarctic and several chronically contaminated sites have been reported (Aislabie et al., 1999; Mazzera et al., 1999; Martins et al., 2004). In Jubany research station (62°14′S, 58°40′W South Shetland Islands; Fig. 1) in particular, a relatively low but chronic contamination with PAH of surface and subsurface soils and sediments was recently reported (Curtosi et al., 2007), attributed not only to fuel spills but also to the low temperature combustion of organic materials.

Figure 1.

Location of experimental sites in Jubany station and neighbouring areas from Potter Peninsula, King George Island (Isla 25 de Mayo) South Shetland Islands, Antarctica. Also the location of other stations mentioned in the text is shown.

In situ remediation of contaminated sites exploiting the catabolic capacities of microorganisms (bioremediation) is proposed as a convenient tool for the clean-up of soils and coastal waters in Antarctic environments, since it can be implemented with a minimal disruption of treated sites, a fundamental prerequisite for nature-protected areas (Aislabie et al., 2004; 2006). Also plant-associated microbes can contribute to the establishment of vegetation in extreme environments or degraded landscapes, such as contaminated soil (Siciliano et al., 2003). Only two vascular plants, named Colobanthus quitensis and Deschampsia antarctica, can thrive under the drastic Antarctic environmental conditions. Interestingly, we observed that these plants can also grow in the chronically petroleum hydrocarbon-contaminated areas in Jubany station. This might indicate that the plants and their associated microflora growing in these sites can either tolerate or catabolize the contaminants (Karthikeyan and Kulakow, 2003). It is known that the plant exudates released in the rhizhosphere (defined as the portion of soil influenced by plant) can be used as carbon sources by microorganisms and that in some cases they can induce microbe-derived degrading enzymes (Chaudhry et al., 2005).

Several bacterial and fungal strains were reported as active PAH degraders and different catabolic pathways have been described (Cerniglia, 1984; Cerniglia et al., 1985; Bouchez et al., 1995; Zocca et al., 2004; Vinas et al., 2005; Vandermeer and Daugulis, 2007). However, it is assumed that only microbial communities adapted to the harsh environment that characterize the polar areas can exhibit a significant catabolic activity under such extreme conditions (Stallwood et al., 2005; Aislabie et al., 2006; Ruberto et al., 2008). Considering also that the protocol to the Antarctic Treaty prohibits the introduction of foreign organisms into the Antarctic area, studying the catabolic potential in microbial communities from Antarctic soils is of crucial importance.

The dominant aerobic catabolic pathway for low-molecular-weight PAH is initiated by a multicomponent enzyme system called naphthalene dioxygenase (NDO; E.C. (Kauppi et al., 1998). Hence, molecular biology strategies targeting this gene have been used for tracing PAH-degrading genes of great environmental relevance (Dionisi et al., 2004; Zhou et al., 2006; Gomes et al., 2007). In this study, the occurrence and diversity of ndo genes in Antarctic soils from PAH-contaminated and non-contaminated sites from the Maritime Antarctic (Jubany station and adjacent areas in Potter Peninsula) were assessed. In addition, we directed efforts to elucidate whether ndo diversity is influenced by the two vascular plants that grow in Antarctica: D. antarctica and C. quitensis. Ndo genes were amplified from total community DNA (TC-DNA) by polymerase chain reaction (PCR), with primers designed to detect a broad range of ndo genes, and analysed by denaturant gradient gel electrophoresis (DGGE) and cloning and sequencing, as previously described by Gomes and colleagues (2007). We aimed to expand the knowledge on the genetic potential for PAH biodegradation of local microorganisms by targeting a wide range of ndo genotypes of environmental importance, rather than targeting specific ndo genotypes, as been done so far. To the best of our knowledge, this is the first application of the above described strategy to study the occurrence of ndo genes in Antarctic soils, which considers as well the possible influence of vegetation at the plant species level.


Detection and profiling of ndo genes by PCR-DGGE

The nested ndo PCR approach employed produced amplicons of the expected size for all TC-DNA samples originating from bulk and rhizosphere soil from site B (as detected in 1% agarose gel with ethidium bromide staining; data not shown). For bulk and rhizosphere soil samples originating from sites PV and PI, typically two to three replicates out of four did yield PCR products. The highest amounts of amplicons were obtained from samples originating from the site related to oil hydrocarbons contamination (site B; data not shown).

A representative ndo-DGGE fingerprint displaying the ndo genotype diversity of bulk and rhizosphere samples is shown in Fig. 2. Typically two to four strong dominant bands per site and soil type were found, together with some less intense bands. Noticeably, a stable and reproducible fingerprinting pattern was observed for those samples taken in the vicinity of oil storage tanks and station facilities (site B), characteristic that did not apply to fingerprints corresponding to samples of the rest of the sites. In fact, those samples originating from the sites not affected by PAH contamination (sites PV, PI) had variable and sometimes not reproducible patterns between replicates. It is noteworthy that the band pattern corresponding to the oil contaminated area (site B) was not influenced either by the presence of plants or by the plant species. Among the dominant genotypes (bands) revealed by the fingerprint analysis, those that co-migrated with genotypes with mobility types II and III represented in the ndo marker (both corresponding to nahAc ndo genotypes) were found for all the sites and soil types. Another type of dominant genotype (mobility IVa), present only in bulk soils from site PV, had an electrophoretic mobility similar to that of nagAc ndo marker (mobility IV). The rest of the dominant bands did not co-migrate with any of the ndo gene markers used. Those bands were labelled as well according to their DGGE mobility as follows: bands with mobility types VI and VII associated mainly to site B, band Ib for site PI and bands IVa and VIII detected for site PV.

Figure 2.

DGGE fingerprints of ndo gene fragments amplified from bulk and rhizosphere soil DNA templates. Band positions indicated in the gel correspond to the electrophoretic mobility of the ndo fragments used as marker (M) and to dominant bands originating from the Antarctic soil samples (which in some cases co-migrated with the marker ndo gene fragments). ndo gene fragments represented in the marker are: I (5ndo-S3 environmental clone EF455676.1, Gomes et al., 2007); II (nahAc Pseudomonas sp. ARS 10, I. Kosheleva, unpubl. data); III (nahAc Pseudomonas putida KT2442 pNF142, Gomes et al., 2005); IV (nagAc environmental clone, Gomes, NCM, unpubl. data); V (3ndo-S1, environmental clone EF455675.1, Gomes et al. 2007); and IX (phnAc environmental clone AY540615, Gomes et al., 2005).

Identification of the main ndo types by means of cloning and sequencing

Clone libraries of ndo amplicons of each sample type were constructed in order to assign the identities to the dominant DGGE bands and to detect possible new ndo genotypes that can potentially be underestimated by DGGE fingerprinting analysis.

We obtained clones with inserts that co-migrated with dominant bands of all sites and sample types as well as some clones with intermediate mobilities. After discarding sequences that were of poor quality, a total of 109 clones matched with more than 95% similarity to ndo gene sequences contained in the GenBank, and two clones matched with 94% similarity. Based on their nucleotide sequences, the ndo gene fragments (384 nucleotides) were classified into five tentative groups, according to the corresponding phylogenetic neighbour with highest similarity (Table 1). All clones were linked to closely related nahAc genotypes described for Pseudomonas species and/or plasmids associated to this genus. The main characteristics of the closest phylogenetic neighbours are summarized in Table 2. The geographic origin of the closest related ndo genotypes and the habitats from which they were obtained were diverse and comprised soil, sediments and marine environments related to oil contamination as well as non-contaminated sites from both mesophilic and Antarctic environments.

Table 1.  Distribution of ndo gene fragment clones originating from bulk and rhizospheric soils from Jubany station and neighbouring sites of Potter Peninsula (South Shetland, Antarctica).
Phylogenetic neighbour (accession No.)
Clone number for each combination of site-sample type (clone library one-letter code)
PV-BS (D)PV-Da (E)PI-BS (F)PI-Cq (G)B-BS (A)B-Cq (B)B-Da (C)
  • *

    The bold numbers in columns two to eight indicate the number of clones in the particular group. The percentages values within square brackets represent the percentage similarity of the clones with the nearest phylogenetic neighbour; when less than 95% indicated with an asterisk ().

nahAc pNAH7
Pseudomonas putida
D14, D37, D38,E2, E17, E18, E21, E28, E38, E40F8, F19G25A6, A8B19C38
G.22  12376
Pseudomonas sp.
5N1-1 mRNA
D13, D22  G1, G5, G9, G16, G19, G22, G24, G26, G33, G35, G39, G40A17, A27, A32B2, B6, B7, B21, B23, B27, B31C5, C6, C8, C11, C20, C31
[97–98%]  [97–100%][99–100%][97–100%][96–99%]
nahAc pDTG1
NCIB 9816-4
D3, D30, D39E9, E14, E26, E31, E33, E37F20, F26,F27G6, G7, G8, G10, G11, G13, G14, G17, G20, G30, G36, G38A3, A21, A25*, A30, A40*B4, B11, B13, B14, B28C1, C2, C3, C9, C10, C13, C15, C16, C17, C21, C23, C24*, C26, C27, C33, C35, C37, C39, C40
[96–99%][98–99%][98–99%][96–99%][98–99%], *[94%][95–99%][97–99%], *[94%]
balearica LS402
D6, D40      
G.5 18    
Pseudomonas sp.
2N1-1 mRNA
 E5F4, F5, F12, F23, F28, F32, F37, F38    
Table 2.  Description of the closest relative retrieved from GenBank to environmental ndo clones originating from Antarctic soils from Jubany station and Potter Peninsula (South Shetland, Antarctica).
GroupClosest relativeReferences
NumberMain DGGE mobilities associatedaAccession (GenBank)Gene nameBacterial strainLocation or molecule typeRelevant characteristics
  • a. 

    Corresponding to the environmental ndo gene fragment clones that clustered into the defined groups.

putida, strain G7
Plasmid pNAH7Archetypal PAH degradation pathwayYen and Gunsalus (1982); Sota et al. (2006)
sp. 5N1-1
mRNANaphthalene-degrading isolate from marine sediments, AntarcticaBosch, R., Lalucat, J. and Rosselló-Mora, R. (unpublished)
3IIIAF491307.2nahAcPseudomonas putida
NCBI 9816-4
Plasmid pDTG1Archetypal PAH degradation pathwayDennis and Zylstra (2004)
Genomic DNAMarine isolate, Western Mediterranean region, Barcelona SpainRosselló (1992); Ferrero et al. (2002)
5IIIaAJ496391nahAcPseudomonas sp. 2N1-1mRNANaphthalene-degrading isolate from marine sediments, AntarcticaBosch, R., Lalucat, J. and Rosselló-Mora, R. (unpublished)

The main phylogenetic groups are represented in the tree constructed with the corresponding amino acid sequences (128 amino acid residues) of the environmental clones (Fig. 3). Noticeably, the phylogenetic relationships and distribution of ndo genotypes were well predicted by the DGGE fingerprinting analysis (Fig. 2). For example, clones with DGGE mobilities II and III clustered into two groups (Group 1 and Group 3 respectively) being both closely related to archetypal nahAc gene sequences carried on Pseudomonas-associated plasmids. As shown in Fig. 3, the Group 1 contains ndo gene fragments closely related to the nahAc gene described for Pseudomonas putida G7 pNAH7 (sequence identity within the group 99.2%) and the Group 3 contains ndo gene fragments closely related to the nahAc gene described for P. putida NCIB 9816-4 pDTG1 (sequence identity within the group 98.2%). The sequence identity between Group 1 and Group 3 was 95.7%. These two groups are numerically the most important ones (see Table 2) and they contain representative clones from all the sites and sample types analysed (distribution revealed by the DGGE profile) as well as some less frequent clones with DGGE mobilities revealed by the analysis of the clone libraries, which corresponded to very similar variants of the main genotypes. Another important group is Group 2, which includes clones with DGGE mobility types VI and VII retrieved almost exclusively from the contaminated site B (also in this group cluster some clones with intermediate mobilities indicated with V and a subindex). Members within this group showed 99.1% sequence identity and are closely related to a nahAc gene carried by a naphthalene-degrading Pseudomonas strain isolated from Antarctic marine sediments (see Table 2). The sequences of this group are related to Groups 1 and 3, having with them 95.1% and 93.5% sequence identity respectively. The clone C20 (with mobility type VI) is apparently located between the Group 2 and the branch of Groups 1 and 3. The last two groups (Groups 4 and 5) are more distantly related to the former ones and are less frequent. Their particularity is that they contain clones obtained only from non-contaminated sites. The Group 4 had an amino acid identity with the Groups 1, 2 and 3 of 83.5%, 88.2% and 85.3% respectively. Only two clones corresponding to the dominant bands originating from bulk soil of site PV (Fig. 3) were recovered from the library and sequenced: clone D6 (mobility VIII) and clone D40 mobility (IVa) (Table 2; only clone D40 was included in the tree). The closest relative for both clones was a nahAc genotype described for Pseudomonas balearica (94.8% amino acid identity for clone D40; 97% nucleotide similarity for clone D6). Although the electrophoretic mobility of clone D40 (IVa) was similar to that of the nagAc gene fragment used as marker for the fingerprint (mobility IV; see Fig. 2) it was not related to that genotype. The other closely related group that contained ndo genotypes found also in non-contaminated sites is Group 5 (sites PV-Da and PI-Bs, only for the latter site a clone is shown in the tree). Members within this group had 99.6% sequence identity and were closely related to ndo genotypes described for Pseudomonas sp. 2N1-1, which is a naphthalene-degrading strain isolated from Antarctic marine sediments. The main DGGE mobility of ndo clones in this group was denominated IIIa. The Group 5 had 94.5% amino acid identity with Group 4 and 86.1%, 90.7% and 87.9% amino acid identity with Groups 1, 2 and 3, respectively.

Figure 3.

Phylogenetic relationships of the amino acid sequences of partial large alpha subunit corresponding to ndo fragments of the seven clone libraries generated for bulk and rhizosphere soil for sites B, PV, PI. The tree (128 amino acid residues) was constructed using the neighbour-joining method and the bootstrapping analysis (1000 repetitions); the phnAc gene from Burkholderia sp. RP007 was used as outgroup. The number on the branches indicates the percentages of bootstrap values when greater than 50%. The scale bar represents the amino acid divergence. The prefix in the clone number indicates the origin of the corresponding library as follows: A (B-BS); B (B-Cq); C (B-Da); D (PV-BS); E (PV-Da); F (PI-BS); and G (PI-Cq). Roman numbers in parentheses indicate the corresponding DGGE mobility and Arabic numbers indicate the number of clones with identical sequence. Ndo gene fragments corresponding to PAH-degrading isolates recovered from Antarctic stations reported by Ma and colleagues (2006) were included and are indicated with a circle (CR: Frei Station and LCY: Great Wall Station, in King George Island; P/L ZT: Zhongshan station, East Antarctic).

16S rRNA and nahAc gene copy number determined by real-time PCR

An estimation of the bacterial abundance in the Antarctic soils studied was performed through the quantification of the 16S rRNA gene copy number (Fig. 4). Results were interpreted in a comparative fashion, considering that multiple rRNA gene copies per cell can occur. The analyses revealed that the average values for sites with low impact of anthropogenic activities (sites PV, PI) ranged between 1.94 × 108 and 2.83 × 108 16S small-subunit (SSU) gene copies per gram of soil on dry weight (DW) basis; no significant effect of the vegetation or the location was evidenced (P > 0.05). The average values for sites impacted by oil hydrocarbon contamination (site B) ranged from 2.48 × 108 to 2.17 × 109 16S SSU gene copies per gram of soil DW. The top value of this range corresponds to site B-Cq, for which the trend indicates a higher 16 SSU gene copy number, in comparison with other sample types. However, no statistically significant differences with other sample types from site B or with samples from non-polluted sites (PV and PI) were found (P > 0.05).

Figure 4.

Quantification of 16S rRNA gene and nahAc gene copy numbers contained in bulk and rhizosphere soil of Jubany station neighbouring areas. Different upper-case letters indicate statistically significant differences in 16S rRNA gene copy numbers; different lower-case letters indicate statistically significant differences in nahAc gene copy numbers (< 0.05).

After having analysed the type of ndo genes present in the Antarctic soils studied, it was decided to quantify the amount of nahAc gene copies using a specific probe. The analysis of the copy number of nahAc gene (Fig. 4) revealed that the quantities detected in the bulk soil and both rhizospheric soils from the PAH-affected site (site B; range: 6.35 × 104 to 1.61 × 106nahAc gene copies per gram of soil DW) were significantly higher (< 0.05) compared with the corresponding counterparts of non-polluted sites (sites PV and PI; range: 1.78 × 103 to 5.57 × 104nahAc gene copies per gram of soil DW). For both contaminated and non-contaminated sites no significant effect of the vegetation was evidenced (P > 0.05).

The integral analysis of the described results confirm that differences of nahAc copy numbers among contaminated and non-contaminated sites were not due to differences in microbial biomass, since there was no significant difference in the 16S SSU gene copy numbers between the sites or samples types.


Diversity and occurrence of ndo genes in Antarctic soils environments

Soils, sediments and coastal waters in the surrounding of Antarctic stations are increasingly polluted with PAH and other toxic compounds due to the operation of research stations, fishing activities and an increasing influence of the tourism recorded in the last decade (Frenot et al., 2005). Bioremediation is proposed as the best tool for cleaning and recovering polluted sites in Antarctica (Ruberto et al., 2003; Aislabie et al., 2006). Consequently, the characterization of the biodegradation potential of indigenous microbes represents a crucial point. In this study the occurrence and diversity of ndo genes in PAH-contaminated sites from Jubany station and non-contaminated sites from the Potter Peninsula was investigated by means of a ndo PCR-DGGE fingerprinting method recently developed by Gomes and colleagues (2007). The ndo PCR primers used were designed to amplify a wide range of dioxygenase genes, belonging to the main clade of the Group 3, according to the classification system for large alpha subunits of terminal dioxygenases proposed by Nam and colleagues (2001). This clade includes several ndo types of environmental importance which are found in strains of Pseudomonas (nahAc, nahA3, pahAc, doxB), Alcaligenes (phnAc), Ralstonia (phnAc, nagAc), Burkholderia (phnAc, dntAc), Comamonas (pahAc) and Herbaspirillum (phnAc). A cultivation-independent analysis of ndo genes in Antarctic microbial communities was achieved by amplifying ndo genes from total community DNA and subsequent analysis of the PCR amplicons by means of DGGE or cloning and sequencing. This approach was first applied to study sediments from urban mangrove forests affected by different degree of PAH contamination (Gomes et al., 2007). In the TC-DNA from mangrove sediment and rhizosphere samples several ndo genes such as nagAc- and phnAc-like genes but also novel ndo types were detected by sequencing dominant bands of the ndo patterns. Interestingly, Gomes and colleagues (2007) observed a higher number of ndo genotypes in the sediments of the less polluted sites and an increased relative abundance of certain ndo types in the sediment from the heavily polluted site. In our study, bulk and rhizosphere soil samples from a very contrasting environment, the Antarctic, were studied. While ndo PCR products were obtained from all replicates originating from a site chronically contaminated with PAH from Jubany station, ndo amplification failed for some replicates originating from non-contaminated sites from Potter Peninsula. The lack of amplification is most likely due to a low abundance of ndo genes in the non-contaminated environments rather than a suboptimal amount of DNA template or PCR-inhibiting substances, given that the abundance of bacteria in the different sites was comparable (as estimated by the quantitative analyses of the 16S rRNA gene copy number) and bacterial 16S rRNA gene DGGE fingerprints were obtained without problem (data not shown). Also, the high variability among the fingerprints obtained for the samples from the non-contaminated sites points to a relative low abundance of the target sequences. As expected, the lower abundance of ndo genes in non-contaminated sites corresponded well with the results obtained for the quantification of the copy numbers of the dominant ndo gene detected in the soils studied, the nahAc gene. In this study, we unequivocally showed that ndo genes are present in Antarctic soils and that hydrocarbon pollution seems to increase the abundance of bacteria carrying these genes. A possible explanation for the presence of ndo genes in the non-contaminated areas could be related to the presence of naturally occurring PAH compounds, as those sites are not directly impacted by activities in the station and are separated by relevant terrain irregularities. In contrast to the rather high variability among replicates for samples originating from non-contaminated sites, the fingerprints of the samples originating from the PAH-affected site (site B) revealed stable and reproducible patterns. These results suggest that the presence of contamination exerted selection pressure, resulting in a higher numerical abundance of organisms carrying PAH-degrading genes. This selection pressure effect was also observed by Dionisi and colleagues (2004) for PAH-contaminated freshwater sediments and was described as well for other types of contamination (van der Meer et al., 1998; Newby et al., 2000; Shaw and Burns, 2004).

Concerning the identity of the ndo genes, the phylogenetic analysis of clone libraries revealed the occurrence of exclusively nahAc-like genes in both contaminated and non-contaminated sites from Antarctic soils from Jubany station and Potter Peninsula respectively. This result is surprising, taking into account that the DGGE fingerprints showed dominant ndo types of very different electrophoretic mobilities, which could, in principle, suggest the occurrence of diverse ndo genes, as observed in the study by Gomes and colleagues (2007). In our study, the finding of dominant bands with very different electrophoretic mobility reflected a microdiversity of closely related nahAc genes detected by the PCR-DGGE method used. The closest relatives to these nahAc genotypes are genes encoding proteins described in Pseudomonas species originating from both mesophilic and Antarctic environments (Table 2). As mentioned before, the ndo primers used in this study were designed to detect a wide range of ndo genes of environmental importance, including the divergent phnAc genes (Laurie and Lloyd-Jones, 1999). In this regard, it was not possible to target the less related ndo alleles, such as narA-like gene present in Rhodococcus strains or the nidA/pdoA1-like genes from Mycobacterium, among others (Habe and Omori, 2003).

PAH contamination and ndo gene types

Previous studies on ndo genes in Antarctic regions reported the occurrence of nahAc genotypes in isolates originating from marine sediments (Bosch, R., Lalucat, J. and Rosselló–Mora, R., unpublished) affected by oil contamination. Ma and colleagues (2006) described the occurrence of nahAc-like genotypes carried by PAH-degrading bacteria recovered from PAH-contaminated soil after enrichment in agitated flasks with minimal medium amended with naphthalene or phenanthrene as the sole carbon source, followed by selective plating. Luz and colleagues (2004) searched specifically for ndoB genotypes by PCR hybridization of TC-DNA directly extracted from soil samples and detected those ndo genotype in several contaminated and uncontaminated Antarctic soils. An earlier study carried out with pristine soil samples taken at Ross Island (77°S, 166°E) (Laurie and Lloyd-Jones, 2000) showed the occurrence of phnAc genotypes by means of competitive PCR quantification (QC-PCR), performed with TC-DNA extracted from the soil after enrichment in microcosms amended with naphthalene or phenanthrene, supplied as vapour. In the same work and using the same approach, also phnAc genotypes were detected by QC-PCR in pristine soils originating from very distant places such as Siberia (61°N, 89°E) and New Zealand (38°S, 175°E). Naphthalene and phenanthrene are also the main PAH pollutants found in the chronically contaminated areas from Jubany station studied in this work (Curtosi et al., 2007), and are likely to contribute to a natural enrichment of PAH-degrading bacteria in these soils under Antarctic conditions. Surprisingly, we did not detect other genotypes than nahAc, using the same molecular approach (PCR-DGGE, from TC-DNA directly extracted from soil) that has already been proven to be useful for detection of phnAc, nagAc as well as other novel ndo genotypes in PAH-contaminated mangrove forests (Gomes et al., 2007). Additionally, we analysed our samples by PCR amplification with phnAc-specific primers, as described by Laurie and Lloyd-Jones (2000); none of the samples retrieved phnAc amplicons (data not shown). The mentioned study, as well as some other studies corresponding to temperate environments, indicated a possible correlation of the type of PAH contamination with the abundance of ndo alleles. For example, a noteworthy result emerging from a culture-dependent study (in which naphthalene-degrading bacteria were isolated from tar-contaminated sites from mesophilic environments without prior enrichment) was that the nahAc gene was present only in isolates from the heavily contaminated sites and the phnAc gene was distributed among divergent hosts isolated from marginally contaminated sites (Wilson et al., 2003). Another study of a tar-contaminated site, based on the quantification of nagAc-like gene sequences by real-time PCR assays applied to DNA extracted from the freshwater sediments, showed a positive correlation of the concentration of naphthalene with the abundance of nagAc-like alleles (Dionisi et al., 2004). Ni Chadhain and colleagues (2006) studied the occurrence of ndo genes in PAH impacted soils (by means of PCR amplification with degenerate primers) and described the dominance of a nahAc-like allele detected after enrichment with naphthalene and that this ndo type was rarely detected when phenanthrene was used for enrichment. This study described as well a succession of different ndo populations responding to different PAH-contaminated environments. However, results from enrichment experiments can introduce a severe bias, as the array of organisms growing under the enrichment conditions defined do not necessarily reflect what occurs in natural environments (Dunbar et al., 1997).

The above mentioned studies represent only a few examples described in the literature that have shown a correlation between ndo alleles and contamination level and type, at least for each particular site studied. However, extrapolations of such correlations remain mostly as speculations. On one side comparisons are difficult since different approaches and molecular tools were used for assessing ndo gene diversity. On the other side, it is hard to disentangle the complexity of the interactions of biotic and abiotic factors that can influence the occurrence and distribution of these genes. Studies that consider habitat characteristics, besides the type presence of contamination, are necessary to understand the ecological and genetic aspects of the biodegradation of PAH.

Influence of vegetation and PAH contamination on ndo gene diversity

One important habitat characteristic is the presence of vegetation, since it has been demonstrated that plant species can introduce profound changes in the composition of microbial communities (Smalla et al., 2001; Kowalchuk et al., 2002; Costa et al., 2006) and potentially of the microorganisms carrying degradative genes. Although most of the studies on the rhizosphere effect were performed on mesophilic environments, a recent broad-scale study showed that the bacterial diversity in Antarctic soils (as determined by cultivation and molecular methods) decreased along a latitudinal gradient (51°S to 72°S), being the vegetated sites the remarkable exception to this trend (Yergeau et al., 2007a). Similar results were found when analysing the bacterial abundance (Yergeau et al., 2007b). Although in that comprehensive study the vegetation was evaluated as a cover with no dissection of the type (the regions explored are mostly vegetated by cryptogams with a few vascular plants), it was suggested that the vegetation could exert an important effect on microbial communities also under harsh Antarctic conditions. Noteworthy, the presence of vegetation was also suggested as a differentiating factor for the occurrence of ndo genes in tar-contaminated sites with high and low contamination level, for which, respectively, nahAc and phnAc genotypes were distinctively detected (Wilson et al., 2003). Although the authors only speculate about the influence of vegetation on the occurrence of genes which are carried by bacteria related to PAH degradation (since it is mentioned that the sampling scheme was not specifically designed for assessing it), it is an important observation arising from the analyses of the experimental data, which indicates the importance of considering the vegetation as a variable when studying the distribution of ndo genes.

In our study, we designed the sampling strategy to differentially analyse the occurrence of ndo genotypes in the bulk soil and the rhizosphere from the only two vascular plants growing in Antarctic regions (D. antarctica and C. quitensis). Furthermore, the sampling was designed to differentiate the effect of vascular plants at the species level. Due to the variability of the fingerprints of non-contaminated sites (likely linked to the relative low abundance of ndo genes in these sites) it was difficult to detect a rhizosphere effect on ndo genotype distribution. Noticeably, the DGGE fingerprints for contaminated sites revealed clearly no effect of plant species or globally of vegetation on the relative abundance of ndo genes, as almost identical fingerprinting patterns were found for the contaminated soil, irrespectively of the presence of two different vascular plant species. It is important to point out that when using the same TC-DNA to study the bacterial diversity by PCR-DGGE, reproducible 16S rRNA gene DGGE profiles were obtained for all sample types and a very clear rhizosphere effect at the species level was detected (data not shown). From the ndo genetic profile obtained it can be deduced that the preponderant factor on the selection of ndo genotypes was the contamination. A possible explanation would be that the drastic Antarctic environmental conditions as a sum (which comprise not only low temperatures but also repeated thaw–freezing cycles, draught and high UV radiation) together with the contamination might constitute a stronger factor than a possible rhizosphere effect on the bacterial communities carrying ndo genotypes. It is likely that these soils are enriched in organisms adapted to the mentioned drastic conditions and if it occurs that these organisms are carrying or are able to acquire genes conferring the capacity to grow on contaminated sites, they might numerically expand and predominate. Furthermore, the ndo-DGGE fingerprints showed that some of the dominant ndo genotypes were present in both contaminated and non-contaminated soils and that the contamination seemed to increase the relative abundance of these genotypes. These results were confirmed by sequencing ndo clone libraries generated for all combinations of sites and soil types (seven clone libraries). Sequences from all sample types showed the highest similarity with ndo genotypes described for members of the Pseudomonas genus and associated plasmids such as IncP-9, which might indicate that horizontal gene transfer (HGT) is an important factor which can drive the occurrence and spread of ndo genes in this environment. Our results are in accordance with the findings arising from cultivation-dependent studies carried out by Ma and colleagues (2006). These researchers studied Antarctic soils from contaminated sites from stations located in King George Island, in the vicinity of Jubany station (Great Wall station: 62°12′S, 58°57′W; Frei station: 62°19′S, 58°55′W) and on East Antarctic (Zhongshan station, China: 69°22′S, 76°22′W) (Fig. 1). PAH-degrading bacteria isolated only from PAH-contaminated sites after enrichment with naphthalene or phenanthrene most frequently belonged to the genus Pseudomonas. The plasmid-located ndo genotypes detected for the PAH-degrading isolates exhibited high similarity with nahAc genotypes contained in the Groups 1 and 3 (associated to DGGE mobilities II and III respectively) described in our cultivation-independent study (Figs 2 and 3).

Concluding remarks

Our culture-independent study described for the first time the type of ndo genes present in soils of Jubany station and neighbouring areas in the Maritime Antarctic. A microdiversity of nahAc-like genes, closely related to Pseudomonas species and associated mobile elements, was revealed. The relative abundance of these genes was shaped in response to anthropogenic sources of pollution and apparently not affected by the presence of vascular plants. Investigations based on cultivation-independent methods targeting specifically the Pseudomonas group and related plasmids in the Antarctic soils are in progress.

Experimental procedures

Characteristics of the Antarctic location

The place under study is the Argentinean scientific station Jubany and surrounding sites from Potter Peninsula (62°14′S, 58°40′W), King George Island (Isla 25 de Mayo), South Shetland Islands (Fig. 1). The Potter Peninsula presents diverse vegetated microenvironments, which range from non-human-disturbed areas to places heavily influenced by anthropogenic activities. Fossil fuel (mainly diesel) is used for heating and machinery operation. Accidental spills occurring during fuel manipulation, combustion of diesel and a history of open air garbage burning are the main sources of PAH contamination in the area of the station. A monitoring study of the PAH levels in Jubany station was performed during the austral summer Antarctic expeditions 2003/4 (the same expedition in which the samples for this study were collected) and 2004/5 (Curtosi et al., 2007). The authors reported that the levels of total PAH in the chronically polluted surface and subsurface soils from Jubany station ranged between 10 and 1182 ng g−1 DW, being two to three ring PAH (mostly represented by methylnaphthalene, phenanthrene and anthracene) detected as the main compounds.

Experimental sites and sampling procedure

Samples were collected during the Summer Antarctic Campaign 2003/4. A field exploration and a census of vegetation covering the station field and surrounding protected areas of Potter Peninsula were performed prior to the collection of soil and plant material, aiming to characterize and define the experimental sites. The required permits for working in this Antarctic Specially Protected Area (ASPA 132) were obtained from the corresponding authorities of the Dirección Nacional del Antártico (Argentina).

Three different areas were finally chosen for this study, aiming to select those sites with contrasting characteristics, in terms of anthropogenic impact (Fig. 1). The selected experimental sites were named with acronyms as follow: site B (located near the oil storage tanks) comprised areas chronically contaminated with hydrocarbons. This location corresponds to the site B from Curtosi and colleagues (2007), who determined that the total PAH concentration of this site was 15–21 ng g−1 soil DW, up to a soil depth of 30 cm (which corresponds approximately to the soil depth that the roots of the vascular plants studied can reach). The sites PI and PV which corresponded to unpolluted sectors were located far from the station facilities (2.5–4.5 km), and separated from that by relevant terrain irregularities and ice masses (although the places are not considered absolutely pristine because crew members of the station very occasionally arrive to them). The levels of PAH compounds from sites PI and PV were below the detection limit (1 ng g−1; US EPA method 8100). In all these sites, D. antarctica (Da) or C. quitensis (Cq) or both were present. The origin of each sample (sites PV, PI, B), the presence of plant species (Da, Cq) and soil properties are summarized in Table 3. From each of these sites, bulk soil and plant specimens with their rhizosphere soil were collected. Quadrants of 5 × 5 m were delimited, numbered and sorted with random numbers for sampling. Bulk soil samples as well as plants with roots were collected with a metal sterile shovel, trying to reach as deep as the roots of plants were expanding (varied from 15 to 35 cm depth, depending on the soil properties and plant species). The roots were vigorously shaken to separate soil not tightly adhering to them. Soil and plant samples were placed in sterilized plastic bags, kept at approximately 0°C while sampling and transported to the station laboratory for further processing within the next 24 h. At each defined sampling site, two composite samples were collected for rhizosphere and bulk soil. Each composite rhizosphere sample consisted of the soil adhering to the roots of five randomly chosen plants. Composite samples of bulk soil were obtained by mixing five shovels of root-free soil taken till approximately 20 cm depth. A total number of 14 composite soil samples (bulk and rhizosphere) were obtained. The resulting sampling scheme was a compromise between the number of samples required for analysis and the legal restrictions concerning the removal of plant specimens and soil from Antarctic protected areas.

Table 3.  Soil physical and chemical properties of selected sampling sites from Jubany station and Potter Peninsula (South Shetland, Antarctica).
LocationSamples collected (acronym)Chemical propertiesTexture
pHConductivity (dS m−1)P (p.p.m.)C (%)Total N %Clay %Silt %Sand %Class
  1. For soil chemical properties, average values from two composite samples are given. For texture analysis, two homogenously pooled composite samples were used.

Peñón VBulk soil (PV-BS)6.750.1035.960.760.107.52.590.0Sandy
Rhizosphere D. antarctica (PV-Da)4.870.2797.460.710.09102.587.5Loamy sand
Peñón IBulk soil (PI-BS)6.500.1210.050.690.087.52.590.0Sandy
Rhizosphere C. quitensis (PI-Cq)5.380.3528.710.600.0812.52.585.0Loamy sand
Base (Jubany Station)Bulk soil (B-BS)6.910.0949.020.990.1010.07.582.5Loamy sand
Rhizosphere C. quitensis (B-Cq)6.190.424.560.530.0711.31.387.5Loamy sand
Rhizosphere D. antarctica (B-Da)6.280.456.990.570.0910.02.587.5Loamy sand

Initial steps of sample processing such as pooling and homogenization were performed immediately in the station laboratory facilities. Then the samples were shipped at 0°C to the University of Buenos Aires at 0°C. Soil physicochemical properties were determined by standard methods and are presented in Table 3. A complete subset of the samples was stored at −20°C for molecular analyses.

Total community DNA isolation

For molecular analyses each composite sample was further divided into two subsamples, in order to obtain a total of four DNA extractions – PCR amplifications per sample type. TC-DNA was extracted from bulk and rhizosphere soil samples (0.8 g) with the Fast DNA Spin kit for soils (MP Biomedicals, Solon, OH, USA), used according to the supplier's instructions. TC-DNA samples were re-purified with the GENECLEAN Spin kit (Q Biogene, Carlsbad, CA, USA) following manufacturer's recommendations. Genomic DNA extracts were checked under UV light after 0.8% agarose gel electrophoresis and ethidium bromide staining.

Real-time PCR

Real-time PCR assays were performed using the Applied Biosystems 7000 Real-Time PCR System. Quantification of bacterial ribosomal genes in soil was carried using primers and probe previously described (Takai and Horikoshi, 2000). Polymerase chain reaction amplifications were carried out in a 50 μl reaction volume containing 1.25 U of AmpliTaq Gold and buffer II (Applied Biosystems), 0.2 mM of each deoxynucleoside triphosphate, 2.5 mM MgCl2, 0.1 mg ml−1 bovine serum albumin (BSA) and 0.25 mM of primers and probe. Thermocycles were 10 min at 94°C, and 40 cycles consisting of 15 s at 95°C, 15 s at 50°C and 60 s at 60°C. Templates to generate standard curves were serial dilutions of gel-purified PCR products from Escherichia coli 16S rRNA genes (1506 bp, 0.533 OD260). Quantification of ndo genes (nahAc) was performed using primers and probe previously described (Debruyn et al., 2007); the TaqMan probe was slightly modified and its sequence was as follows: 5′-FAM-TGGGR TTGAAAGAAGTCGCTCG-BHQ1. The real-time PCR assay was performed as above described, using serial dilutions of a purified PCR product from a cloned nahAc fragment as standard (940 bp, 0.488 OD260). Results were analysed by means of one-way anova test and post hoc comparisons were performed with Tukey's honestly significant difference (Tukey's HSD, confidence level of 95%).

PCR amplification of ndo gene fragments for DGGE analyses

The TC-DNA extracted from bulk and rhizosphere soil was used as template for PCR amplification targeting ndo genes, using the method developed by Gomes and colleagues (2007). Briefly, amplicons of approximately 896 bp were obtained using the primer pair NAPH1F and NAPH1R and subsequently 1 μl of the PCR product was used as template for a second PCR with primers NAPH2F and NAPH2R-GC, the last one possessing a GG clamp attached to the 5′ end, as described by Heuer and colleagues (1997). The second PCR produced GC-clamped ndo gene fragments of approximately 740 bp suitable for DGGE analysis.

Denaturant gradient gel electrophoresis

DGGE analyses were performed with the Ingeny porU system (Ingeny, Goes, the Netherlands). Gels were prepared with a double gradient of 30–65% denaturants (100% denaturants defined as 7 M urea and 40% formamide) and 6–9% acrylamide. Aliquots of PCR products were loaded on the gel as well as a mixture of PCR products corresponding to fragments of different ndo gene types (Gomes et al., 2007), used as marker to check the electrophoresis run and to compare fragment migration between gels. Electrophoresis was carried out with 1× Tris acetate-EDTA buffer at 58°C and constant voltage of 240 V for 20 h. Gels were silver-stained according to Smalla and colleagues (2001) and air-dried. The different bands in the marker were designated according to their electrophoretic mobilities (with Roman numbers), as well as dominant bands originating from the environmental samples that did not co-migrate with bands corresponding to the ndo marker. Dried gels were scanned transmissively with high resolutions settings (720 dpi) with an Epson 1680 Pro equipment (Seiko-Epson Japan).

Cloning and sequencing

The ndo PCR fragments obtained after the second PCR round for each sample type were used for cloning. The amplicons obtained from each of the four DNA extractions from composite samples were pooled together and cleaned using the GENECLEAN Spin kit (QBiogene, Carlsbad, CA, USA) following supplier's recommendations. Products were ligated into pGEM-T vectors (Promega) and transformed into competent E. coli cells (E. coli JM109; Promega) following manufacturer's instructions.

The clone libraries, corresponding to each sample type (seven in total), were denominated with a one-letter code (A to G) which was used as prefix for numbering the corresponding clones (see Table 1). Typically, 40 positive clones per library were selected and subjected to PCR-DGGE to check the corresponding electrophoretic mobilities. Clones containing inserts that showed electrophoretic mobilities comparable to that of dominants bands in DGGE profiles were selected for sequencing analyses, as well as those that exhibited DGGE mobilities in intermediate positions, aiming to detect possible new ndo gene types not revealed by PCR-DGGE profiling of the environmental samples. After this selection, approximately one-half of the clones originally picked for each library was sent for sequencing.

Phylogenetic analyses

Sequence chromatograms were manually cured and sequences below 200 bp were discarded from the analyses. Identities searches were performed using blast-n and blast-x (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Amino acid sequences were deduced and aligned and a homogeneous block of 128 amino acid residues was selected for the phylogeny analysis. Subsequently, sequences corresponding to ndo genotypes recovered from neighbouring Antarctic stations described by Ma and colleagues (2006) were included in the analysis, for comparison purposes. The evolutionary history was inferred using the neighbour-joining method (Saitou and Nei, 1987), with a bootstrap test (1000 replicates) (Felsenstein, 1985). The evolutionary distances were computed using the JTT matrix-based method (Jones et al., 1992). The mentioned phylogenetic and molecular evolutionary analyses were conducted using MEGA version 4 (Tamura et al., 2007).

Nucleotide sequence accession numbers

The ndo gene sequences determined in this study were deposited in the GenBank database under the numbers EU660591 to EU660669.


This work was supported by the Alexander von Humboldt Foundation. Field work was supported by Instituto Antártico Argentino and initial sample processing was performed at the University of Buenos Aires. Many thanks to Jubany station crew for logistical support and to E. Krögerrecklenfort and H. Heuer JKI for assistance in real-time PCR analyses.