Analysis of catabolic genes for naphthalene and phenanthrene degradation in contaminated New Zealand soils


*Corresponding author. Tel.: +64 (7) 858-3700; Fax: +64 (7) 858-4164;


Culture-dependent and culture-independent methods were used to investigate the diversity of three polycyclic aromatic hydrocarbon (PAH) catabolic genes in contaminated soils. PAH-degrading bacteria were isolated based on growth at the expense of naphthalene (44 isolates) or phenanthrene (35 isolates). Of these 79 PAH-degraders, 53% (42 isolates) failed to hybridise with three gene probes specific for PAH degradation. The gene for the naphthalene dioxygenase iron–sulfur protein (nahAc) from Pseudomonas putida G7 hybridised to 45% (20/44) of the culturable naphthalene-degrading bacteria of the ‘classical’nah-type, whilst analogues of the bacterial glutathione S-transferase (GST) encoding gene of Sphingomonas paucimobilis EPA505 were associated with culturable phenanthrene-degrading isolates and hybridised to 29% (10/35) of these isolates. Apart from the host strain Burkholderia RP007, we were not able to detect the phnAc gene amongst cultured isolates by hybridisation or PCR, though could directly amplify this gene from contaminated soils.


Enumeration of the PAH-degrading bacterial population in contaminated soils using traditional microbiological methods can take an inordinate length of time, and often underestimates numbers as a result of our inability to cultivate the majority of soil organisms. Molecular biological approaches using hybridisation or polymerase chain reaction (PCR) methods are valid alternatives. An impediment to these approaches is the available database of appropriate gene sequences, which for PAH-degrading bacteria is currently limited. Another obstacle arises from extrapolating sequence data originally obtained from readily cultivated bacterial groups which can bias our perception of the importance of readily isolated genotypes in soil microbial ecology, and does not provide for an adequate description of natural populations.

A prime example of a readily isolated group of soil bacteria are Pseudomonas and Pseudomonas-like strains isolated for their ability to degrade naphthalene and its close analogues. Numerous examples of catabolic genes from such strains have been described which show a surprising (>90%) degree of conservation [1–6], and broad distribution [7–11]. For the purposes of this study we will describe these ‘classical’nah genes as nah-like sequences. It is accepted that this highly conserved group of genes cannot reflect the true diversity of microbial genes involved in the catabolism of PAHs, such as naphthalene and phenanthrene [12–14].

We have targeted three different PAH catabolic genes, using both DNA hybridisation and PCR amplification, to analyse the genotype of the culturable phenanthrene- and naphthalene-degrading population isolated from soils contaminated with aromatic hydrocarbons. In addition, analysis of DNA directly isolated from the same contaminated soils has allowed us to determine the presence or absence of these same PAH genes amongst the total microbial population in these soils. The probes and PCR primers used in this study were designed to target catabolic genes involved in the degradation of PAHs by soil bacteria. These are based on genes originally isolated from Pseudomonas, Sphingomonas and Burkholderia, genera commonly isolated for their ability to degrade the PAHs, such as naphthalene and phenanthrene [12, 15].

Two of the primer pairs target genes that encode the iron–sulfur protein (α-subunits) of two divergent initial dioxygenases which are integral components of the first enzyme of bacterial oxidative attack on PAHs [16, 17]. One primer pair specifically targets the nah-like genes [1–6], the other the disparate phn-type phenanthrene-catabolic genes [18]. The phn genes were isolated from Burkholderia sp. RP007 and show only 24–73% amino acid sequence homology with the corresponding nah-like sequences [18].

The third primer set targets a glutathione S-transferase (GST) encoding gene that has previously been described as a probe for PAH-degrading bacteria [19]. GST is associated with catabolic genes involved in aromatic hydrocarbon degradation in a number of other Sphingomonas strains including Sphingomonas sp. WP01, Sphingomonas sp. RP003, S. yanoikuyae Q1, and B1, and S. aromaticivorans F199 [19], whilst a GST analogue has also been found adjacent to PAH catabolic genes in Pseudomonas U2 [6], and Comamonas testosteroni GZ42 [14]. Given the ever-increasing evidence linking bacterial GSTs to the detoxification of harmful or persistent chemicals [20, 21], bacterial GST genes may yet fill a valuable role as an environmental indicator, mirroring that of eukaryotic GSTs.

2Materials and methods

2.1Soil collection

Soil samples, contaminated with PAH and BTEX compounds, were collected throughout the Waikato region of New Zealand's North Island in April and December of 1997 and stored at 4°C. Samples 1–7 were contaminated with automotive fuels and lubricants, samples 8–10 were obtained from the site of a former town gas generating plant in Hamilton, New Zealand, whilst samples 11 and 12 represent pristine soils obtained from a maize pasture and a native New Zealand forest. Further descriptions of soil samples are provided in Table 1.

Table 1.  Description of soil samples
  1. aSamples 1–7 are industrial and roadside samples consisting of disturbed soil/road runoff/scoria contaminated with oil and fuels.

  2. bBTEX benzene/toluene/ethylbenzene/xylenes.

  3. c∑16 EPA priority PAHs.

1Contaminated soil below leaking oil barrelsOil
2Diesel-contaminated soil from gravel carparkDiesel
3Fuel oil-contaminated soilFuel oil
4Fuel oil-contaminated soilFuel oil
5Oil-contaminated soil adjacent to railwayOil/BTEXb
6Soil near heavy goods vehicle drivewayOil
7Roadside oily sludge sedimentOil/BTEX
8PAH-contaminated soil from Hamilton town gas site (8.5% organic carbon)<15 ppm BTEX, 41–340 ppm PAHsc
9PAH-contaminated soil from Hamilton town gas site (9.2% organic carbon)<260 ppm BTEX, <7580 ppm PAHs
10PAH-contaminated soil from Hamilton town gas site (6.3% organic carbon)<224 ppm BTEX, <2937 ppm PAHs
11Maize pasture (6% organic carbon)Pristine
12Native forest soil (30–40% organic carbon)Pristine

2.2Isolation and analysis of naphthalene and phenanthrene degrading bacteria

For the purposes of this study we have defined the culturable PAH-degrading population as colony-forming isolates able to utilise either naphthalene or phenanthrene in pure culture as sole carbon and energy sources. The inherent biases of such a definition, and that of our isolation approach, precludes isolation of any auxotrophic PAH-degrading bacteria, PAH-degrading consortia, or simply PAH-degrading bacteria that were unable to survive the transition to laboratory culture. Our approach was based upon the direct plating of soil bacteria onto agar plates, followed by incubation in the presence of PAHs as growth substrates. Enrichment cultures were not used for isolation to avoid preselection of the most rapidly growing PAH-degrading bacteria at the expense of a perhaps more diverse population of isolates. To reduce the chance of duplication, we elected to use ten contaminated soil samples collected from geographically disparate areas from which to isolate PAH-degrading bacteria. Five grams of each soil, added to 45 ml of sterile 0.1% (w/v) tetra-sodium pyrophosphate containing 15 g of 2 mm glass beads in a 100-ml Schott bottle, were shaken at 200 rpm for 1 h at 28°C. Serial dilutions made in 0.1% (w/v) tetra-sodium pyrophosphate were spread directly onto minimal agar plates (4 g l−1 Na2HPO4, 2 g l−1 KH2PO4, 1 g l−1 (NH4)2SO4, 2 ml l−1 salts solution [22] solidified with 1.5% purified agar (Difco)). Both naphthalene (naphthalene isolates) and phenanthrene (phenanthrene isolates) were used as sole carbon and energy sources to assess the influence of carbon source on the diversity of culturable PAH-degrading bacteria that were isolated. Naphthalene (E. Merck, Germany) was supplied as a vapor by incubating unsealed Petri dishes in a desiccator containing naphthalene crystals placed in its base, and phenanthrene (Sigma, St. Louis, MO, USA) vapor was provided by placing phenanthrene crystals within the lids of sealed Petri dishes. Following up to 6 weeks of incubation at 28°C, morphologically distinct colonies from each plate were selected. Isolates were identified by soil sample number/carbon source used for isolation/isolate number, thus 1N7 was isolated from soil sample 1 using naphthalene as carbon source. In addition, phenanthrene-degrading isolates obtained from the site of a former coal carbonisation plant at Rotowaro, New Zealand were included, these are identified as RP002, RP004 and RP005. Soils sampled from this site were not included in this study.

Presumptive growth of naphthalene- and phenanthrene-degrading isolates was confirmed by comparing growth on substrate-free minimal agar plates in the presence and absence of naphthalene (nah) and phenanthrene (phe). Nutrient-rich (2.5 g l−1 Bacto tryptone; 1.25 g l−1 Bacto yeast extract; 0.5 g l−1 glucose; 15 g l−1 Bacto agar (Difco)) grown colonies were also screened for ability to oxidise indole or catechol. The ability to oxidise indole (ind), provided in the vapor phase from crystals of indole placed in the lid of a sealed Petri dish, leads to the formation of the blue product indigo, which is indicative of the presence of aromatic oxygenase genes [23]. Colonies able to express extradiol dioxygenase activity were detected by their ability to form the yellow ring-fission product 2-hydroxymuconic semialdehyde on spraying colonies with 0.1 M catechol (cat) [24].

2.3Isolation of soil and genomic DNA

DNA was extracted from 0.5-g soil samples using a bead beating method followed by purification through PVPP (polyvinylpolypyrrolidone) spin columns [25]. Genomic DNA was isolated from naphthalene- and phenanthrene-degrading bacteria using a modified CTAB (hexadecyl trimethylamonium bromide) method [26].

2.4Dot blots and Southern hybridisation

To reveal the proportion of colonies hybridising to each probe, dot blots were prepared with 3 μg of genomic DNA from each isolate applied to a positively charged nylon membrane (GeneScreen Plus (NEN Research Products)). Southern blots consisting of 10 μl of PCR products (see below) amplified from DNA extracted directly from each soil sample, and resolved on an agarose gel, were also blotted to a positively charged nylon membrane for analysis by hybridisation with the corresponding 32P-labelled probes. The probes, amplified using the PCR primers described below, consisted of; a 992-bp region encompassing nt 63–1055 of nahAc from Pseudomonas putida G7; a 993-bp region encompassing nt 82–1075 of phnAc from Burkholderia sp. strain RP007; and 487-bp encompassing nt 1–487 of the GST gene from Sphingomonas sp. strain WP01. Strains WP01 and RP007 have been deposited in the culture collection at Landcare Research (ICMP, c/o Landcare Research, Private Bag 92170, Auckland, New Zealand) as strain numbers 13533 and 13529.

Membranes were hybridised overnight at 65°C in 5×SSC, followed by stringency washes with 2×SSC, 0.1% SDS for 15 min at 65°C, and 0.2×SSC, 0.1% SDS for 15 min at 65°C. Burkholderia LB400 [27], Sphingomonas CB3 [28], P. putida F1 [29] and P. putida mt-2 [30] were used as negative control strains for dot-blot hybridisation. Under these conditions, the phnAc and nahAc probes do not hybridise to characterised iron–sulfur protein (α-subunits) of initial dioxygenases present in control strains (bphA, carAa, todC1, and xylX share only 36–42% nucleotide sequence homology with probe sequences). Homology between characterised phnAc and nahAc-like sequences of 55% was also too low to allow hybridisation. Similarly, the GST gene probe amplified from WP01 shows <55% nucleotide sequence homology to GST from Burkholderia LB400 [19] which is insufficient to hybridise under these conditions. Sequence data from nahAc amplicons characterised during this study suggest that the mismatch stringency for the G7 nahAc probe under these hybridisation conditions is approximately 20%. The similar GC content of the phnAc and GST probes suggests they would behave similarly.

2.5PCR analysis

A comparison of the nucleotide sequences for the primers targeting α-subunits of phn and nah initial dioxygenases relative to annealing sites on characterised gene sequences is presented in Fig. 1. The sequence of the nahAc gene from P. putida G7 (GenBank accession number M83949) was used as the basis for primer design to generate primers (nahAcfor and nahAcrev) specific for the conserved nahAc sequences of nah-like genotypes. We have confirmed that these primers amplify the pahAc gene of P. putida OUS82 despite three and two mismatches at the primer annealing sites (data not shown). The P8073/P9047 primer pair will amplify phnAc similar to that present in Burkholderia sp. RP007 (GenBank accession number AF061751), but due to sequence mismatches, does not allow amplification of nah-like genes including pahAc from P. putida OUS82. The third primer pair, p3-24F (5′-ATGAAACTGTTCATCAGCC) and p3-24R (5′-CCAGCATCACGTACAGATAG) is based on the bacterial glutathione S-transferase (GST) encoding gene of Sphingomonas paucimobilis EPA505 (GenBank accession number AF001779) [19]. The genes for 16S rDNA were targeted in DNA extracted from each soil sample using the primers 27F (5′-AGAGTTTGATCCTGGCTCAG) and 1492R (5′-TACGGGTACCTTGTTACGACTT) [31].

Figure 1.

Alignment of partial nucleotide sequences encoding for the α-subunits of the initial PAH dioxygenase enzyme from naphthalene- and phenanthrene-degrading bacteria. Sequences correspond to nt 63–97 and 1036–1072 according to nahAc from P. putida G7, and nt 66–100 and 1040–1075 for phnAc from Burkholderia sp. strain RP007. Sequence gaps are shown as dots, and PCR primers are highlighted. The nahAcfor/nahAcrev primers shown above the alignment are designed to amplify a 992-bp region encompassing nt 63–1055 of nahAc from Pseudomonas putida G7, and the P8073/P9047 primers shown below the alignment are designed to amplifiy a 993-bp region encompassing nt 82–1075 of phnAc from Burkholderia sp. strain RP007. Nucleotide sequence mismatches to the nahAcfor/nahAcrev primers for aligned sequences are underlined, and for the P8073/P9047 primers are shaded.

PCR amplification was carried out in 20 mM Tris-HCl (pH 8.4); 50 mM KCl; 1.25 mM MgCl2; dNTPs at 200 μM each, 2.5 units PLATINUM Taq DNA polymerase (Gibco-BRL); 0.2 μM forward and reverse primers; and template DNA at 0.1 μg per 50 μl reaction. The cycling conditions (Techne Cyclogene Thermal Cycler) were 5 min at 94°C, followed by 25 cycles (40 cycles when amplifying from DNA isolated from soil) of 94°C for 2 min, 50°C (16S rDNA and GST primers) or 55°C (nahAc and phnAc primers) for 1 min, 72°C for 1 min, maximal ramp rates throughout, with the final 72°C segment of the cycle extended to 10 min before cooling to 4°C. PCR products amplified from genomic DNA of PAH-degrading isolates were visualised by agarose gel electrophoresis. Restriction fragment length polymorphism (RFLP) analysis was used to discern polymorphisms. A PstI digest was used to identify nah-like α-subunits (P. putida 9816-4 nahAc[1], P. putida 9816 ndoB[2], Pseudomonas C18 doxB[5], P. putida G7 nahAc[1], P. putida PAK1 pahAc, P. putida OUS82 pahAc[4]) as 270-bp and 722-bp fragments.

To reveal the diversity of the nahAc sequences amplified using the nahAcfor and nahAcrev primers, partial nucleotide sequences of selected amplification products were determined. PCR amplicons were sequenced directly using the nahAcfor primer and PRISM Ready Reaction DNA Terminator Cycle Sequencing Kit (Perkin-Elmer). Reactions were resolved using an ABI model 377 sequencer by the Waikato DNA Sequencing facility.


3.1Genetic markers in culturable PAH-degrading bacteria

The genotype of the majority of PAH-degrading isolates could not be defined by the probes used in this study. Only 47% (37/79) of the 79 PAH-degrading bacteria (44 isolated on naphthalene, and 35 on phenanthrene) hybridised to any of the three PAH catabolic probes. The three probes, nahAc from P. putida G7, phnAc from Burkholderia RP007, and the GST encoding gene from Sphingomonas WP01, hybridised to the corresponding positive controls. Negative controls failed to hybridise due to low nucleotide homology.

The nah-like genotype was found to predominate among naphthalene isolates. nahAc was amplified from 45% (20/44) of bacterial strains isolated using naphthalene as sole carbon source. Isolates within this group were all Gram-negative, only one of which was also able to degrade phenanthrene. Positive indole (ind+) and catechol (cat+) reactions also correlate well with this genotype, and were present in 85% (17/20) of these isolates. The high incidence of the nah-like genotype amongst bacteria isolated using naphthalene as carbon source is commonly observed, and often compounded by classical enrichment. Compared with classical enrichment, where 73% (19/26) [32] of isolates possessed nah genotypes, our direct plating approach has resulted in the isolation of a more diverse population, since only 45% (20/44) of these isolates contained nah-like genotypes.

The nah-like nahAc sequence was only amplified from 17% (6/35) of phenanthrene isolates, three of which also conformed to the nah+phe+ind+cat+ phenotype. A study of phenanthrene-degrading marine sediment bacteria produced similar data, with <12% of phenanthrene isolates hybridising to a nah-like probe [12].

Additional genetic markers different to the nah-like sequences are therefore required before PCR amplification, or probe hybridisation, can adequately provide a reliable indication of the potential catabolic capacity of phenanthrene-degrading bacteria in nature. An example is the gene encoding GST. Of the three probes used in this study, GST was most common amongst the phenanthrene isolates and was amplified from, and hybridised to, 29% (10/35) of these isolates. Only one naphthalene isolate, which also degraded phenanthrene, hybridised to the GST gene probe. None of the isolates which hybridised to the GST gene also hybridised to either nahAc or phnAc, which is perhaps surprising given that GST encoding genes are frequently associated with PAH catabolic genes [6, 19, 20].

The third PAH marker gene used in this study is the phnAc gene from the phn operon of Burkholderia RP007, originally isolated for its ability to degrade phenanthrene [18]. Surprisingly, the phnAc probe did not hybridise to, and was not amplified from, any culturable PAH-degrading isolate obtained during this study. This is particularly interesting when we consider that Burkholderia RP007 was isolated in a similar manner, and from a similar environment, to those sampled in this study [18]. Two reasonable hypotheses can be formulated to account for the absence of phn-type PAH catabolic genes amongst the culturable PAH-degrading bacteria we have analysed. Either this genotype is rare in the environment we have sampled, or alternatively, the customary hosts for phn genes are genera which we are not able to cultivate under laboratory conditions.

3.2Diversity of nahAc obtained from pure cultures

nahAc amplicons were compared to determine the level of sequence conservation within this gene amongst our PAH-degrading bacteria which had a nah-like genotype. On the basis of RFLP analysis, the nahAc amplicons were subdivided into two classes. The majority of nahAc PCR products (21/26) have an identical RFLP to the ‘classical’nahAc, whilst the other five lack this PstI site. Partial sequencing and phylogenetic analysis of 539 bp from ten nahAc PCR products comprising representatives of each group shows that the four nahAc amplicons (9P4, 1N7, 2N1, 8N3) are >90% homologous to the nah-like group, and sequences (RP005, RP002, 4N2C, 4NA, 4N2B, RP004) that do not contain an internal PstI site fall into a recently diverged and highly conserved (>90% homologous) branch of the iron–sulfur proteins (α-subunits) which include nagAc, ntdAc, and dntAc[6, 33, 34](Fig. 2).

Figure 2.

Dendrogram based on a comparison of 539 bp of nucleotide sequence (corresponding to nt 111–650 of nahAc from Pseudomonas putida G7) from genes encoding the α-subunits of the initial dioxygenase enzymes. The analysis is based on corresponding regions from published ‘classical’nah-like, the dnt/ntd, and phn-type sequences [18]. Data for nahAc amplicons from naphthalene- and phenanthrene-degrading isolates are included to show their position relative to published sequences. The unrooted tree is rerooted using phnAc as an outgroup. Bootstrap values of the major branches are shown. The tree was constructed using the PHYLIP software package (version 3.57c) [36] using the SEQBOOT, DNADIST, NEIGHBOR and CONSENSE programs. Dendrograms were redrawn to scale using the ClustalW [37] homology data.

The nah-like and dnt/ntd-groups contain both naphthalene- and phenanthrene-degrading phenotypes. Thus the nahAcfor/nahAcrev primers are able to target not only the nah-like, but also representatives of the dnt/ntd-group nahAc sequences with naphthalene- and phenanthrene-degrading phenotypes. While the divergent phnAc gene was included in the phylogenetic analysis as an outgroup, sequences closely related to phnAc were not amplified by the nahAc specific primers.

3.3Genetic markers for PAH-degrading bacteria amplified from soil

In the second part of our study, we sought to determine the presence or absence of each genetic marker in contaminated soil samples. We were particularly interested in evaluating whether the divergent phn genes were detectable in DNA directly extracted from each soil since none of the PAH-degrading bacteria isolated from these soils contained phn-type genes. Furthermore, we were curious as to whether the predominance of nah-like sequences over GST gene and phn-type sequences observed amongst cultivated soil bacteria was also reflected in the total microbial population of the same soils.

Twelve soil samples were selected and each yielded 1–5 μg of DNA per gram (wet weight) of sample. Fragments corresponding to approximately 1400-bp of 16S rDNA were successfully amplified from each soil DNA extract (data not shown), confirming that the DNA extracts isolated from soil were not inhibitory to PCR analysis. We then applied a PCR amplification strategy based on the three different PAH genetic markers to analyse the PAH-degrading population present in aromatic hydrocarbon contaminated soils. Two pristine soils, from a maize pasture and a native New Zealand forest, were included to provide an indication of the distribution of these genes in soils not affected by anthropogenic aromatic hydrocarbon contamination.

From the 12 soil samples analysed, nahAc was amplified from seven soils and GST from nine soils. Perhaps more significantly, the phn genotype was relatively common and amplified from eight soils screened during this study. Our failure to detect the phn genes in PAH-degrading bacteria isolated from these same soils implies that the usual host range of the phn genes are bacterial species not readily cultivated under laboratory conditions. Fig. 3 shows that fragments amplified from contaminated soils were identical in size and hybridised to the corresponding PCR products amplified from the nahAc, phnAc and GST genes.

Figure 3.

Amplification of (A) phnAc, (B) nahAc, and (C) GST amplicons from contaminated and pristine soils. The PCR products amplified from each soil sample were resolved by agarose gel electrophoresis. The corresponding hybridisation patterns are shown to confirm the identity of the PCR products. Lanes 1–12 correspond to the contaminated soils sampled for this study. Negative and positive controls appear in lanes 13 and 14.

Genotype distribution was not uniform amongst the soils, four soils contained all three genotypes, four other soils contained only two genotypes, whilst a single genotype was present in the four remaining soils. These results clearly demonstrate that focusing on a single gene subgroup, for example, the nah-like group (present in only seven soils), can make the reliable prediction of biodegradation potential difficult [11, 12, 32]. Together, the α-subunits of two divergent initial dioxygenases, nahAc and phnAc, were enriched in all soils contaminated with BTEX and PAHs. The GST gene was detected in pristine as well as contaminated soils which suggests that this gene is a broadly distributed and ubiquitous bacterial enzyme.

To confirm that the nahAc and phnAc probes were indeed specific for different populations of PAH-degrading bacteria, we cross-hybridised the phnAc soil PCR products with the G7 nahAc probe, and the nahAc soil PCR products with the RP007 phnAc probe. These two probes show 55% nucleotide homology to each other which is not sufficient to cross-hybridise under the hybridisation conditions used in this study. Cross-hybridisation was not detected to either set of soil PCR products, confirming that consensus sequences, intermediate to both nahAc and phnAc, were not amplified from contaminated soils by either primer set. This confirms that the nahAc and phnAc genotypes targeted by the PCR primers used in this study represent divergent populations of PAH-degrading bacteria, and supports the phylogenetic evidence that the phn genes have evolved as a deeply branching divergent lineage from nah-like sequences [18].


Our results reveal some of the potential difficulties of analysing any ecologically significant genotype. Although subgroups of the potential PAH-degrading population can be detected in most contaminated soils using either microbiological or molecular biological approaches, it is also evident that both approaches can lead to different conclusions. From a microbiological perspective, we are restricted to what we can cultivate, which we have demonstrated includes isolates which host divergent catabolic genes not represented in the sequence databases. Molecular approaches are also restricted due to the limited number of characterised sequences available for most phenotypes. In this study, we have shown that the phn sequences which are very difficult to isolate in pure culture, may be as widely distributed in contaminated soils as those that are readily isolated such as nah, or even those encoding GST.

Our microbiological analysis has shown us that using closely related PAHs, such as naphthalene and phenanthrene, as growth substrates leads to the isolation of different genotypic groups. Naphthalene results in the isolation of a greater proportion of Pseudomonas-like organisms with a nah-like genotype, compared with phenanthrene. nah-like isolates were shown to be readily cultivated, with 45% of naphthalene degrading organisms isolated during this study confirmed as having a nah-like genotype. Primers or probes based on nah-like sequences therefore seem best suited to detecting ‘classical’ naphthalene-degrading bacteria, and it is apparent that the paradigm of using nah-like sequences as probes for analysis of genes that encode phenanthrene or PAH degradation is flawed, since the nah genes hybridise to only a small fraction of phenanthrene-degrading isolates. By comparison, the predominant genes which encode phenanthrene degradation are less well characterised. Although the GST probe was the best probe of those used in this study, the majority (54%) of phenanthrene-degrading isolates remained uncharacterised. The phn genes we had hoped would fill this gap failed to hybridise to any of the PAH-degrading isolates, yet were detected in contaminated soils. This emphasises dramatically that the cultured fraction is unrepresentative of the total population, and despite the apparent enrichment of the phn genes in contaminated soils, our inability to detect these genes amongst the culturable population also suggests that this genotype would rarely feature in microbiological studies of PAH degrading bacteria.

We acknowledge that the probes used in this study cannot be expected to provide a complete picture of the overall genetic diversity of PAH catabolic genes, any more than genes originally isolated from Pseudomonas, Sphingomonas and Burkholderia strains can be considered to be the sole catalysts of PAH degradation. Thus it is not surprising that the largest genotypic class amongst the cultivated bacteria was not delineated by the probes and PCR primers used in this study. This class must contain either divergent homologues of the targeted genes with low sequence homology, or alternatively, contains unrelated and uncharacterised genes for PAH catabolism. That over 50% of the PAH-degrading isolates remain unclassified does suggest that we have a poor grasp of the available diversity of PAH catabolic genes. This diversity can only be described by applying alternative approaches to the isolation of PAH-degrading bacteria and their catabolic genes. For example, the isolation of aromatic catabolic genes on the basis of positive colorimetric reactions to either indole or catechol, despite undoubted value in the cloning of aromatic monooxygenase and dioxygenase genes, represents a poor approach for isolating new PAH degradation genes since many genotypically unclassified isolates gave negative reactions to indole and catechol screens. More subtle culturing approaches allied to different methods for isolating PAH genes, such as using analogues of indole [35], or hybridisation using GST genes [19], may prove useful for targeting and describing novel PAH catabolic genes.

Applying molecular biological techniques to screen for a particular genotype has revealed a single probe does not adequately describe a broad phenotype in all contaminated soils. Within the limits of our screening approach, all genotypes were not represented in all soil samples; however, we were able to compare the presence of different genotypes in a range of contaminated soil samples. The nah and phn genotypes both represent a potential capacity for PAH degradation that is enriched in the majority of contaminated soils sampled during this study. The GST gene was also present in contaminated samples, though since it appears to be more ubiquitous, some care must be taken when interpreting the GST gene as a marker for PAH catabolic genes, since this gene is also present in other catabolic pathways [19, 20]. Although we are only able to sample a small fraction of the available PAH-degrading population, using either microbiological or molecular biological techniques, we must determine the significance of these populations in contaminated soils before valid ecological or biotechnological conclusions can be made. Currently studies are underway to determine the relative contributions of the nah and phn genotypes to PAH-degrading populations in different soils.


Support provided by the New Zealand Foundation for Research Science and Technology is acknowledged.