Initial characterization of new bacteria degrading high-molecular weight polycyclic aromatic hydrocarbons isolated from a 2-year enrichment in a two-liquid-phase culture system


Réjean Beaudet, INRS-Institut Armand-Frappier, 531 boulevard des Prairies, Laval, Québec, Canada, H7V 1B7 (e-mail:


Aims: To characterize some polycyclic aromatic hydrocarbons (PAH)-degrading microorganisms isolated from an enriched consortium degrading high molecular weight (HMW) PAHs in a two-liquid-phase (TLP) soil slurry bioreactor, and to determine the effect of low molecular weight (LMW) PAH on their growth and HMW PAH-degrading activity.

Methods and results: Several microorganisms were isolated from a HMW-PAH (pyrene, chrysene, benzo[a]pyrene and perylene) degrading consortium enriched in TLP cultures using silicone oil as the organic phase. From 16S rRNA analysis, four isolates were identified as Mycobacterium gilvum B1 (99% identity),Bacillus pumilus B44 (99% identity), Microbacterium esteraromaticum B21 (98% identity), and to the genus Porphyrobacter B51 (96% identity). The two latter isolates have not previously been associated with PAH degradation. Isolate B51 grew strongly in the interfacial fraction in the presence of naphthalene vapours and phenanthrene compared with cultures without LMW PAHs. Benzo[a]pyrene was degraded in cultures containing a HMW PAH mixture but pyrene had no effect on its degradation. The growth of isolates B1 and B21 was improved in the aqueous phase than in the interfacial fraction for cultures with naphthalene vapours. Pyrene was required for benzo[a]pyrene degradation by isolate B1. For isolate B21, pyrene and chrysene were degraded only in cultures without naphthalene vapours.

Conclusion: Consortium enriched in a TLP culture is composed of microorganisms with different abilities to grow at the interface or in the aqueous phase according to the culture conditions and the PAH that are present. Naphthalene vapours increased the growth of the microorganisms in TLP cultures but did not stimulate the HMW PAH degradation.

Significance and Impact of the Study: New HMW PAH-degrading microorganisms and a better understanding of the mechanisms involved in HMW PAH degradation in TLP cultures.


Polycyclic aromatic hydrocarbons (PAHs) are produced during fossil fuel combustion and as by-products of industrial processes. They are found throughout the environment in low concentrations. However, high levels are detected in soils surrounding PAH-releasing industrial plants such as gas works, coking plants and wood-preserving facilities. Several PAHs, most notably those with high molecular weights (HMW), are potentially genotoxic and carcinogenic (Cerniglia 1992; Mersch-Sundermann et al. 1992), and 16 of them are listed as priority pollutants by the US Environmental Protection Agency (USEPA web site: HMW PAHs have high hydrophobicity, low water solubility and a tendency to sorb to the organic fraction of soil and sediments. These properties are largely responsible for their low availability to microorganisms and their persistence in the environment (Hughes et al. 1997). Two-liquid-phase (TLP) bioreactors is a newly developed strategy to increase the bioavailability and the biodegradation of hydrophobic/toxic pollutants (Déziel et al. 1999). In this type of reactor, a water-immiscible liquid, non-biodegradable and non-toxic to microorganisms is used to dissolve the hydrophobic organic substrates and to promote and control their bioavailability to microorganisms. The hydrophobic compounds diffuse from the water-immiscible phase to the aqueous phase and are transformed or degraded by the microorganisms at the interface and/or in the aqueous phase. A variety of hydrophobic solvents have been used to improve the biodegradation of PAHs such as paraffin or silicone oil (Vanneck et al. 1995; Jimenez and Bartha 1996; Villemur et al. 2000).

Previously, we reported the enrichment of a microbial consortium degrading HMW PAHs (pyrene, chrysene, benzo[a]pyrene and perylene) in TLP soil slurry bioreactors (Marcoux et al. 2000; Villemur et al. 2000), and optimization experiments have shown that maximal PAH-degrading activity is obtained with silicone oil as the water-immiscible phase. Furthermore, the addition of low molecular weight (LMW) PAHs such as phenanthrene and naphthalene increased the PAH-degrading activity of the consortium. However, study of the degrading activity of the microorganisms at the interface and in the aqueous phase is difficult because of the complexity and the variability of the microorganisms forming the consortium. The present report aims to characterize some PAH-degrading microorganisms isolated from this consortium, and to determine their localization in TLP cultures and the effects of LMW PAH on their HMW PAH-degrading activity.

Materials and method

Microorganisms and culture conditions

A microbial consortium degrading HMW PAHs was previously enriched in TLP cultures in our laboratory (Marcoux et al. 2000, Villemur et al. 2000). TLP enrichment cultures were carried out in 500 ml Erlenmeyer flasks containing 90 ml of Bushnell–Haas (BH) mineral salt medium (Difco Laboratory, Detroit, MI, USA), 10 g of sterile non-contaminated soil (MS) and 50–100 mg l−1 of pyrene, chrysene, benzo[a]pyrene and perylene dissolved in 20 ml of silicone oil (dimethylpolysiloxane, 20 centistokes, Sigma-Aldrich Chemical, Oakville, Canada). Flasks were incubated at 25°C on a rotary shaker with agitation at 150 rev min−1. The MS soil was taken in a private land in a residential area and was without noticeable pollutant. This soil was composed of 50% sand, 37% silt and 13% clay and contained 3% total organic carbon. The enrichment was maintained by consecutive transfers (10% of the aqueous phase) in fresh medium at 20–30-day intervals.

In a first round of isolation, microorganisms from 27-month-old enrichment cultures (21 days after the last transfer) were isolated after spreading dilutions on R2A plate (Becton Dickinson, Cockeysville, MD, USA). In a second assay, microorganisms from 31-month-old enrichment cultures (12 days after transfer) were plated on 10% trypticase soya agar (TSA), R2A, BH agar containing 0.05% yeast extract (BHY) (Difco Laboratory) and BH agar with naphthalene continuously available as vapour by the presence of crystals in the plate lids. Rich and poor media were used to allow the isolation of a wide range of microorganisms from the consortium. The PAH degrading activities of the colonies were visualized by the production of clearing zones in the PAH film deposited on the surface of the agar as described by Kiyohara et al. (1982). In this assay, plates were sprayed with 1% (w/v) phenanthrene, fluorene or pyrene dissolved in acetone. The solvent was quickly evaporated leaving a thin film of PAH on the agar surface. After few hours/days of incubation, the appearance of clearing zones around the colonies suggested that the PAH was degraded.

Characterization and identification of microorganisms

Dioxygenase activity such as naphthalene dioxygenase was determined by the indole test (Ensley et al. 1983). Indole vapours were produced by adding crystals in the plate lids. Sphingomonas sp. 107-6 was used as a positive control (Dagher et al. 1997). Oxidase tests were performed with 1% (w/v) N,N,N′,N′-tetramethyl-p-phenylenediamine dihydrochloride. Catalase activity was evaluated by the production of bubbles after the addition of one drop of 3% (v/v) hydrogen peroxide on a colony spread on a microscope slide.

16S ribosomal RNA sequencing

Liquid culture (1·5 ml) from bacterial isolates was centrifuged at 16 000 ×g for 30 s. Cells were dispersed in 250 μl of TEN (Tris–HCl 50 mM, EDTA 10 mM, NaCl 150 mM, pH 8·0) to which 250 mg of sterilized glass beads (0·4–0·5 mm) were added, then broken in a glass bead mill for 1–2 min at maximum amplitude. The homogenate was centrifuged at 16 000 ×g for 5 min. The supernatant was then treated with RNAse A (10 μg ml−1) for 10–15 min at 22°C, extracted once with phenol/chloroform/isoamyl alcohol (25 : 24 : 1) and once with chloroform/isoamyl alcohol (24 : 1). Total DNA was precipitated with two volumes of ethanol, centrifuged at 16 000 ×g for 10 min, washed with 70% ethanol and resuspended in water (Milli-Q; Fisher Scientific, Montreal, Canada).

Amplification of 16S ribosomal RNA gene (rDNA) sequences were performed with the 16S rDNA eubacterial primers 5′-AGAGTTTGATCCTGGCTCAG-3′ and 5′-AAGGAGGTGATCCAGCCGCA-3′ corresponding to positions 8-27 and 1541-1522 in the 16S rDNA of Escherichia coli (GenBank accession number J01695). Primers were synthesized by the Gene Assembler Plus (Pharmacia, Baie d'Urfé, Canada). PCR amplifications were accomplished in 50 μl reaction volume with 10 mM Tris–HCl pH 9·0, 1·5 mM MgCl2, 50 mM KCl, 200 μM dNTP, 10 pmol of each primer, and 2·5 U of Taq DNA polymerase (Pharmacia). DNA (1–10 ng) was added to preheated PCR mixture at 80°C (3–5 min) and then amplifications were carriedout at 94°C 5 min, 55°C 5 min, then 30 cycles at 72°C 2 min, 94°C 40 s, 55°C 1 min, and finally an extension time at 72°C for 10 min. The PCR product was directly sequenced with the first PCR primer. Sequence analyses were performed with the BLAST programs at the National Center for Biotechnology Information (

PAH biodegradation

Bacterial isolates were cultured in R2A medium in TLP cultures containing the PAH mixture until exponential growth (O.D. 0·4 at 595 nm), then centrifuged at 8000 ×g for 5 min and washed in sterile BH medium before use. PAH degradation experiments were carried out in 250 ml flask containing 21 ml of BH mineral salt medium, 19 ml of sterile soil extract, 5 ml of silicone oil containing the different PAHs and 1 ml of the respective exponentially growing cells. Abiotic controls containing 4% (w/v) sodium azide were inoculated with the corresponding autoclaved bacterial cultures. Flasks were incubated at 25°C with agitation at 150 rev min−1. The soil extract was the supernatant of a 10% (w/v) MS soil in BH mineral salt medium, autoclaved for 20 min and centrifuged 10 min at 750 ×g. All experiments were performed in duplicate or triplicate. In some experiments, naphthalene vapours were continuously provided by crystals in a small cheese-cloth bag attached to the flask stopper.

Localization of isolates B1, B21 and B51 in the two liquid phase cultures

Five-milli litre samples (including the silicone phase) of TLP cultures with respective bacterial isolates were collected periodically and centrifuged for 10 min at 2200 ×g. Bacteria located in the aqueous phase were collected in the pellet. The remaining bacteria were at the silicone–water interface, and were harvested with a Pasteur pipette. Cells were washed twice in phosphate buffer saline pH 7·0, then dispersed in 0·1 N NaOH and incubated at 70°C for 45 min. The cellular protein concentrations were determined according to Bradford assay with the Bio-Rad Protein Assay Kit (Bio-Rad Laboratories, Mississauga, Ontario, Canada) using bovine serum albumin as the standard.

Chemical analysis

The PAHs in silicone oil were extracted with N,N-dimethylformamide and analysed by HPLC using a method already described (Marcoux et al. 2000). The PAHs and other chemicals were from Aldrich Chemical or Sigma (Oakville, Ontario, Canada). Acetone, dichloromethane and acetonitrile were from Anachemia Science (Lachine, Québec, Canada).


In the first experiment, 39 different isolates were obtained on R2A plates from a 21-day-old TLP culture of the 27 month old TLP enrichment cultures. All the isolates were negative for the dioxygenase test and no clearing zone around the colonies were observed on plates previously sprayed with phenanthrene, fluorene or pyrene. TLP cultures with these bacterial isolates showed no PAH degradation. The second isolation assay was carried out 4 months later with the TLP enrichment culture 12 days after the last transfer. Twelve isolates were obtained on BH plates incubated in presence of naphthalene vapours as sole carbon and energy source. All the isolates were able to degrade PAHs in TLP cultures (Table 1, isolates A1–A12). Phenanthrene was degraded by all the isolates but only three degraded benzo[a]pyrene and none degraded perylene. Additional microorganisms were also isolated on 10% TSA, R2A and BHY plates with respectively 17, 17 and 10 isolates each. Ten of these isolates degraded PAHs in TLP cultures (Table 1, isolates B1–B57). Only two of those isolates degraded benzo[a]pyrene and none degraded perylene. Five isolates produced a brown coloration of the colonies on plates sprayed with pyrene. This coloration is probably because of the formation of quinones, a coloured ring fission products (Shuttleworth and Cerniglia 1996).

Table 1.  PAH degrading activity of bacterial isolates in TLP cultures
 PAH degradation (%) after 25 daysClearing zone
  1. TLP cultures of isolates A1–A12 contained BH medium, soil extract and naphthalene vapours. The PAH concentrations in silicone oil were 102, 113, 63, 115 and 40 mg l−1 of phenanthrene (PHE), pyrene (PYR), chrysene (CHY), benzo[a]pyrene (BAP) and perylene (PER), respectively. PAH biodegradation was determined after 25 days.

  2. TLP cultures of isolates B1–B57 contained BH medium, soil extract and the PAH concentration in silicone oil were 79, 106, 36, 96 and 16 mg l−1 of PHE, PYR, CHY, BAP and PER, respectively. PAH biodegradation was determined after 30 days.

  3. +, clearing zone of PAH film around the colonies on plates; P, production of a brown coloration; Nd, not determined; nt, not tested.

B501002310000++ P
B51100100100220++ P

Four isolates (B1, B21, B44 and B51) presenting differences in their PAH-degrading activity and morphology were further characterized. Part of the 16S rDNA sequence (first 500 nt) of each isolate was amplified by PCR. Sequence analyses suggest that isolate B1 is very closely related to Mycobacterium gilvum (99% identity), isolate B21 is related to Microbacterium esteraromaticum (98% identity), isolate B44 is very closely related to Bacillus pumilus (99% identity) and the isolate B51 belongs to the genus Porphyrobacter (96% identity).

The optimal growth temperature of isolates B1, B44 and B51 is 28°C compared with 37°C for isolate B21. Isolates B1 and B44 can grow with naphthalene vapours on BH plates but not isolates B21 and B51. For isolate B1, B21 and B51, PAH degradation was higher in the following culture medium: BH + soil extract > TSB (trypticase soy broth) (diluted 1/10) > BHY > TSB. However, for isolate B44, the best culture media were: BHY > TSB > TSB (diluted 1/10) > BH + soil extract (data not shown).

The biodegradation of PAH by the four isolates in TLP cultures in presence of soil extract is shown in Fig. 1. None of them showed significant degradation of perylene. Isolates B1 and B51 showed complete degradation of pyrene and chrysene in 7–24 days while approximately 15% of the benzo[a]pyrene was degraded in 39 days. Isolate B21 degraded chrysene completely and approximately 50% of pyrene was degraded in 12 days. However, after 12 days pyrene degradation rate was slowed down. No significant degradation of benzo[a]pyrene was observed. Isolates B44 was the least effective with 10–20% degradation of pyrene and chrysene.

Figure 1.

The PAHs biodegradation in TLP cultures inoculated with the isolates B1, B21, B44 or B51. The cultures contained the mixture of pyrene (♦), chrysene (□), benzo[a]pyrene (▴) and perylene (○). The abiotic loss of each PAH determined in the abiotic controls were subtracted. Data are mean values of triplicates

The effect of LMW PAH on the bacterial growth and PAH degradation performance were tested with isolate B21 in TLP cultures (Fig. 2). Isolate B21 was cultured with four HMW PAHs in presence or absence of naphthalene. Bacterial biomass was then collected in the aqueous phase and at the biphase interface. In presence of naphthalene, isolate B21 grew very well in the aqueous phase with 140 mg cell proteins l−1 observed after 28 days compared with only 2·3 mg cell proteins l−1 at the interface (Fig. 2). However, slow growth was observed in the aqueous phase and at the interface (<6 mg cell proteins l−1) in cultures without naphthalene. Pyrene and chrysene degradations were measured in these cultures. In both cases, isolate B21 cultured in presence of naphthalene showed little degradation (10–15%) in contrast to cultures without naphthalene which showed more than 80% degradation.

Figure 2.

Bacterial growth and pyrene and chrysene degradation in TLP cultures inoculated with isolate B21. The cultures contained the HMW PAH mixture (pyrene, chrysene, benzo[a]pyrene, and perylene) and in presence (a) or absence (b) of naphthalene. Bacterial mass (a, b) (in mg of cell proteins l−1) was determined at the interfacial fraction (□) and in the aqueous phase (♦). Pyrene (c) and chrysene (d) biodegradation in presence (▴) or absence (⋄) of naphthalene. (•) Abiotic controls. Data are mean values of triplicates

Benzo[a]pyrene degradation and bacterial growth were then studied with isolates B1 and B51 in the TLP cultures (Fig. 3 and 4). Isolate B1 cultured in presence of benzo[a]pyrene and naphthalene grew better in the aqueous phase in the first 24 days (maximum 42 mg cell proteins l−1), then the biomass dropped at the same level found at the interface (Fig. 3a). Similar results were observed with cultures in presence of benzo[a]pyrene only (Fig. 3c), and in presence of benzo[a]pyrene and phenanthrene (Fig. 3b) with however lower biomass production (< 20 mg cell proteins l−1 in the aqueous phase). In presence of four HMW PAHs, there were no or little significant differences between the biomass found in the aqueous phase and at the interface (Fig. 3d). At the interface, bacteria growth was linear with slightly better growth rate in presence of LMW PAH (0·38 mg l−1 day−1) than with HMW PAHs (0·26 mg l−1 day−1). Benzo[a]pyrene degradation was better in cultures containing the three other HMW PAHs (chrysene, pyrene and perylene). However, its degradation was observed only in the first 15 days of incubation. No significant degradation was observed in the three types of TLP cultures.

Figure 3.

Bacterial growth and benzo[a]pyrene degradation in TLP cultures inoculated with isolate B1. The cultures contained benzo[a]pyrene with (a) naphthalene vapours or (b) phenanthrene (periodic addition) (c) without other PAH or (d) with the HMW PAH mixture (pyrene, chrysene, benzo[a]pyrene, perylene). Bacterial mass (a–d) (in mg of cell proteins l−1) was determined at the interfacial fraction (□) and in the aqueous phase (♦). (e) Benzo[a]pyrene degradation (initial concentration 90–110 mg l−1) in cultures containing only benzo[a]pyrene (▵), or benzo[a]pyrene with naphthalene (▴) or with phenanthrene (⋄) or in the HMW PAH mixture (○). Abiotic controls (•). Data are mean values of duplicates

Figure 4.

Bacterial growth and benzo[a]pyrene degradation in TLP cultures inoculated with isolate B51. The cultures contained benzo[a]pyrene with (a) naphthalene or (b) phenanthrene (periodic addition), (c) without other PAH or (d) with the HMW PAH mixture (pyrene, chrysene, benzo[a]pyrene, perylene). Bacterial mass (a–d) (in mg of cell proteins l−1) was determined at the interfacial fraction (□) and in the aqueous phase (♦). (e) Benzo[a]pyrene degradation (initial concentration 90–110 mg l−1) in cultures containing only benzo[a]pyrene (▵) or benzo[a]pyrene with naphthalene (▴) or with phenanthrene (⋄) or in the HMW PAH mixture (◯). Abiotic controls (•). Data are mean values of duplicates

In contrast, growth of strain B51 was better at the interface in cultures with benzo[a]pyrene and naphthalene (121 mg cell proteins l−1), and in cultures with benzo[a]pyrene and phenanthrene (48 mg cell proteins l−1) than in the aqueous phase (Fig. 4a,b). Growth of isolate B51 was lower in the aqueous phase and at the interface when cultured without LMW PAHs (<20 mg cell proteins l−1) (Fig. 4c,d). Benzo[a]pyrene was slightly degraded in cultures containing mixture of HMW PAH without naphthalene and phenanthrene (Fig. 4e). No significant degradation was observed in the three types of TLP cultures.

The effects of pyrene and chrysene on the benzo[a]pyrene degradation by isolates B1 and B51 were studied (Fig. 5). TLP cultures with each isolate were carried out with different PAH mixtures: benzo[a]pyrene alone, benzo[a]pyrene and pyrene, benzo[a]pyrene and chrysene and finally benzo[a]pyrene, chrysene and pyrene. Initial doses of benzo[a]pyrene, pyrene and chrysene were of 100, 500 and 70 mg l−1, respectively, and periodic additions of pyrene and chrysene were made when necessary in order to maintain their presence in the cultures. Isolate B1 degraded benzo[a]pyrene only in TLP cultures containing pyrene. The degradation rates of benzo[a]pyrene were 1·2 mg l−1 day−1 in cultures containing pyrene and 1·0 mg l−1 day−1 in cultures containing pyrene and chrysene. There was no significant benzo[a]pyrene degradation in cultures containing only chrysene. There was no significant benzo[a]pyrene degradation with isolate B51 in all the culture conditions tested, except after 40 days in the cultures containing chrysene.

Figure 5.

Benzo[a]pyrene degradation by isolates B1 (a) and B51 (b). The TLP cultures contained benzo[a]pyrene alone (□), or benzo[a]pyrene with chrysene (▵), or with pyrene (♦), or with pyrene and chrysene (◯). Pyrene and chrysene initial concentrations were, respectively, of 500 and 70 mg l−1. Periodically, pyrene and chrysene were added to maintain their concentrations in the cultures. (•) Abiotic controls. Data are mean values of duplicates


Several bacterial isolates were isolated from an enriched consortium degrading HMW PAHs in TLP cultures. None of the isolates of the first round of isolation were capable of degrading PAHs. This first round was carried out with a consortium from a transfer culture of 21 days. The dominant microorganisms in the culture at this time were probably those transforming the degradation products and not the microorganisms initiating the PAH degradation. However, the second experiment carried out from a transfer culture of 12 days was successful in isolating PAH-degrading microorganisms. As the consortium was enriched during 31 months, this period has probably allowed the selection of microorganisms well adapted to degrade HMW PAH in TLP system.

Four isolates were further characterized for their PAH degrading activity. Based on their 16S rDNA sequence, they were related to Myco. gilvum (B1), Mic. esteraromaticum (B21), B. pumilus (B44) and Porphyrobacter (B51). Many authors have already observed that Mycobacterium species can degrade PAH (Dean-Ross and Cerniglia 1996; Churchill et al. 1999; Bastiaens et al. 2000; Ho et al. 2000). Only one B. pumilus strain was recently described by Widada et al. (2002), as being capable of utilizing PAH. It was isolated by enrichment from a soil sample in Japan. This strain grew on naphthalene as a sole carbon source but not on phenanthrene. Isolates B21 and B51 belong to genus or species that were never described before as PAH-degrading bacteria. It is possible that the use of a TLP culture system allows the enrichment of PAH-degrading bacteria not usually selected by other growth conditions. The water-immiscible liquid added to the culture increases the substrate bioavailability. Enrichment of adherent microorganisms to the hydrophobic liquid phase with PAH-degrading activity should be favoured under those conditions (Deziel et al. 1999). Bastiaens et al. (2000) using hydrophobic membranes containing sorbed PAHs, have previously enriched and recovered PAH-degrading bacteria exclusively belonging to Mycobacterium spp from PAH-contaminated soil and sludge samples while, in liquid enrichment cultures, mainly Sphingomonas sp. were isolated.

Isolates B1, B21 and B51 degraded PAH more effectively in TLP cultures when the aqueous phase contained BH and soil extract. TSB was the richest and the least effective medium. Rich media probably inhibit PAH degradation at a genetic level (Guerin and Jones 1988) or because other easily degradable substrates are preferentially used by the microorganisms (Keuth and Rehm 1991). However, isolate B44 was different and PAH degradation was more effective in the richest media (BHY and TSB).

Benzo[a]pyrene was degraded by isolate B1 only when it was present in a mixture of HMW PAH. As seen in Figure 5, the presence of pyrene was responsible for this degradation. As suggested by Juhasz et al. (1997), pyrene or one of its metabolites could induce the enzymes involved in benzo[a]pyrene degradation, especially when one considers the similarities between these two compounds. Once the benzo[a]pyrene degradation started, some of its own metabolites could also act as inducers. Benzo[a]pyrene degradation by co-metabolism with pyrene was also previously observed by Juhasz et al. (2000). Under similar conditions, the isolate B51 (Fig. 5) did not significantly degrade benzo[a]pyrene with the exception of a small decrease observed after 40 days in the cultures containing chrysene or the mixture of pyrene and chrysene. This suggests that chrysene could be responsible for this small degradation.

We have reported that the addition of LMW PAHs such as naphthalene or phenanthrene increased the degradation of HMW PAHs by the enriched consortium (Marcoux et al. 2000). This was not observed with strains B1, B21 and B51, for which the LMW PAHs were inhibitory to HMW PAH degradation. This could be the result of inhibition of the induction of PAH-degrading enzymes, competitive inhibition of the PAH-transport system or preferential degradation of the LMW PAH by the microorganisms. Depending on the bacterial strains and PAH combinations used, addition of LMW PAH was found to be favourable (Juhasz et al. 1996, 1997) or unfavourable (Bouchez et al. 1995) to biodegradation of HMW PAHs. Bouchez et al. (1995) reported that naphthalene was toxic to all strains not isolated in this compound, probably because of its relatively high water solubility. However, strain associations were found efficient in relieving inhibition phenomena including the toxic effect of naphthalene. In the present TLP cultures, naphthalene was supplied in the vapour phase, and toxic concentrations were not achieved and allowed the growth of strains B1, B21 and B51 at the interface or in the aqueous phase.

The formation of a biofilm at the interface of the aqueous and organic phases was observed in the present study and in other reports on TLP cultures (Efroymson and Alexander 1991; Ortega-Calvo and Alexander 1994; Ascon-Cabrera and Lebeault 1995; Jimenez and Bartha 1996; Osswald et al. 1996). The microorganisms at the interface can presumably acquire the PAHs directly from the organic phase, and are probably more hydrophobic than the microorganisms freely dispersed in the aqueous phase. Only isolate B51 displayed a preferential growth at the interface in TLP cultures containing LMW PAH. Inversely, the growth of strains B21 in the aqueous phase was much more important than the one observed at the interface in cultures with naphthalene. As the naphthalene inhibited pyrene and chrysene degradation, the best approach in a TLP treatment with the strain B21 would be to use, in the first step, naphthalene for a few days to produce an important biomass which, in a second step, would rapidly degrade the HMW PAH in the absence of naphthalene. As LMW PAH such as naphthalene and phenanthrene are soluble in both phases although less in aqueous phase, their effects on the microbial growth in a mixed culture should be observable at both the interface (as strain B51) and in the aqueous phases (strain B1 and B21).

Our results show that an HMW PAH-degrading consortium enriched in a TLP culture system is composed of microorganisms with different abilities to grow at the interface or in the aqueous phase according to the culture conditions and the PAH that are present. Cultures in TLP systems as an enrichment technique does not select only HMW PAH-degrading microorganisms growing preferentially at the interface, but also in the aqueous phase. Naphthalene vapours stimulated the growth of the microorganisms in TLP cultures but were generally inhibitory to the HMW PAH degradation presumably by substrate competition. Isolation of bacterial species never previously reported to grow on PAHs demonstrates the usefulness of the TLP system as an enrichment procedure.