OUP user menu

Effect of inoculum pretreatment on survival, activity and catabolic gene expression of Sphingobium yanoikuyae B1 in an aged polycyclic aromatic hydrocarbon-contaminated soil

Michael Cunliffe , Akitomo Kawasaki , Emma Fellows , Michael A. Kertesz
DOI: http://dx.doi.org/10.1111/j.1574-6941.2006.00167.x 364-372 First published online: 1 December 2006


The survival and effectiveness of a bioaugmentation strain in its target environment depend not only on physicochemical parameters in the soil but also on the physiological state of the inoculated organism. This study examined the effect of variations in inoculum pretreatment on the survival, metabolic activity (measured as rRNA content) and polycyclic aromatic hydrocarbon (PAH)-catabolic gene expression of Sphingobium yanoikuyae B1 in an aged PAH-contaminated soil. RNA denaturing gradient gel electrophoresis analysis showed stable colonization of PAH-contaminated soil by S. yanoikuyae B1 after four pretreatments (growth in complex or minimal medium, starvation, or acclimation to phenanthrene). By contrast, extractable CFUs decreased with time for all four treatments, and significantly faster for Luria Bertani-grown inocula, suggesting that these cells adhered strongly to soil particles while remaining metabolically active. Pretreatment of the inoculum had a dramatic effect on the expression of genes specific to the PAH-degradation pathway. The highest levels of bphC and xylE expression were seen for inocula that had been precultivated on complex medium, and degradation of PAHs was significantly enhanced in soils treated with these inocula. The results suggest that using complex media instead of minimal media for cultivating bioaugmentation inocula may improve the subsequent efficiency of contaminant biodegradation in the soil.

  • Bioaugmentation
  • Sphingomonas
  • Sphingobium
  • polycyclic aromatic hydrocarbon
  • inoculum pretreatment
  • gene expression


Sphingomonads are prevalent in polycyclic aromatic hydrocarbon (PAH)-contaminated soil communities (Leys et al., 2004) and are routinely isolated from PAH-contaminated soils (Van Broekhoven et al.et al., 2004) and rhizosphere (Daane et al.et al., 2001). Sphingomonads have ecological properties that may facilitate their survival in the PAH-contaminated soil environment (Johnsen & Karlson et al., 2004; Johnsen et al., 2005), including biofilm formation (Johnsen & Karlson et al., 2004) and motility. They are therefore prime candidates for bioremediation of PAH-contaminated soils through bioaugmentation (Kästner et al.et al., 1998).

A range of biological and physical factors can contribute to the successful colonization of soils by inoculated microorganisms. These include many factors characteristic of the target environment, such as soil pH, texture, and water content, and the presence of protozoan populations (for a review see van Veen et al.et al., 1997). The physiological state of the microbial inoculum can also influence the efficiency of soil colonization and subsequent survival. Prestarvation of the inoculum, for example, might be expected to enhance utilization of alternative carbon sources by the introduced bacteria and promote both biodegradation and survival. In one example this was indeed shown to be the case (enhanced survival of Ralstonia eutropha E2 in activated sludge; Watanabe et al.et al., 2000), but other studies have shown no effect of prestarvation on inoculum survival (van Overbeek et al., 1995).

In situ monitoring of catabolic gene expression is a desired component of bioremediation (Sanseverino et al., 1993) and can be used to measure the success of bioaugmentation. Quantitative PCR (qPCR) has been used to monitor the presence of specific organisms in an environment (e.g. phenanthrene-degrading bacteria in contaminated soil; Schwartz et al., 2000) and to monitor defined functional activity (e.g. catechol 2,3-dioxygenase gene expression during benzene, toluene and xylene bioremediation; Mesarch et al., 2000). Degradation of PAHs by sphingomonads follows a pathway similar to that seen for other microorganisms (Habe & Omori et al., 2003), but requires a unique set of genes that are only distantly related to those in pseudomonads and other genera (Pinyakong et al., 2003). These genes are often not well clustered in sphingomonads, which makes them difficult to study. However, PAH degradation by Sphingobium yanoikuyae B1 has been shown to proceed via the pathway shown in Fig. 1 (Kim & Zylstra et al., 1999), and most of the genes required by the strain have been sequenced, facilitating the study of their in situ expression by quantitative reverse transcriptase (qRT)-PCR.


The catabolic pathway for the degradation of naphthalene by Sphingobium yanoikuyae B1, showing the role played by the 2,3-dihydroxybiphenyl 1,2-dioxygenase (bphC) and catechol 2,3-dioxygenase (xylE) encoding genes (boxed) used in this study. The pathways for the degradation of phenanthrene, biphenyl and m-xylene converge at xylE (pathways not shown). Modified from Zylstra & Kim (1997).

We have previously shown that S. yanoikuyae B1, a PAH-degrader, is able to invade the soil community in an aged PAH-contaminated soil (Cunliffe & Kertesz et al., 2006a). In this study we evaluate the effect of a variety of inoculum treatment conditions on colonization by S. yanoikuyae B1, and on the expression of PAH-catabolism genes by this species. The metabolic activity of the strain after inoculation was unaffected by starvation or acclimation procedures prior to inoculation. However, inoculum pretreatment had a major effect on PAH catabolic gene expression and subsequent PAH degradation.

Materials and methods

Bacteria and growth conditions

Sphingobium yanoikuyae B1 (DSMZ 6900) was obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), Braunschweig, Germany. The strain was maintained aerobically at 25°C in complex medium [Luria Bertani (LB) medium (Sambrook et al., 1989), Kings B medium (King et al., 1954) or Nutrient Broth No. 2 (Difco)], R2A medium (Difco), or Brunner Minimal Medium (MM; Cunliffe & Kertesz et al., 2006a). In MM medium, carbon was provided either as succinate (20 mM), or as phenanthrene-saturated medium. [This was prepared by adding 1 mL of a phenanthrene solution in acetone (10% w/v) to a sterile 250 mL Erlenmeyer flask, allowing the acetone to volatilize, and redissolving in 100 mL fresh MM overnight on an orbital shaker. The phenanthrene-saturated MM was then filter-sterilized (0.22 μm).] For solid MM medium, naphthalene crystals were provided in the Petri dish lid. Streptomycin (500 μg mL−1) was added to complex media as required.

Soil, microcosm and inoculation conditions

An aged contaminated soil that had been previously characterized (Cunliffe & Kertesz et al., 2006a) was collected from a decommissioned industrial site in the north of England. The soil was heavily contaminated with PAHs (total EPA 16PAHs 5.98 mg g−1), had a sandy loam texture, pH 7.67, and a relatively low phosphorus content (8 μg g−1). Microcosms were set up in triplicate with 30 g of soil in 50 mL screw-cap polypropylene tubes. Sphingobium yanoikuyae B1 was grown in appropriate medium to exponential phase (A600=0.8), and then starved in carbon-free MM or exposed to phenanthrene in phenanthrene-saturated MM for 4 h, as required. After pretreatment, the cells were washed twice with sodium pyrophosphate buffer (6.3 mM, pH 7.0) and resuspended in dH2O (c. 3 × 1010 CFU mL−1) before being added directly to the soil (1 mL inoculum 30 g−1 microcosm) and mixed with a spatula for 2 min. To ensure that cell density and viability were not affected by the pretreatment used, serial dilutions of each treatment were plated onto LB. The final addition of water with the inoculum brought the water content of the microcosm to 30% of the total water-holding capacity for that soil. Microcosms were maintained in the dark at 25°C, and water content adjusted as necessary.

Microcosm sampling and RNA extraction

Microcosms were sampled at days 0, 2, 5, 10, 20 and 40. Survival of S. yanoikuyae B1 was measured by extraction of the bacterial cells from soil (0.5 g) into 10 mL phosphate-buffered saline (10 mM phosphate, 2.7 mM KCl, 137 mM NaCl, pH 7.4), by incubation on a rotating shaker (30 min, 4°C), and enumeration of S. yanoikuyae colonies after growth on streptomycin-containing LB medium. RNA was extracted from soil samples (0.5 g) using a soil DNA extraction kit (QBiogene), following the manufacturer's instructions [E lysis matrix tube, Fast Prep FP120 (BIO101) bead beater at speed 5.5 for 30 s]. After centrifugation (30 s, 16 000 g), the supernatant was removed and mixed with 250 μL of protein precipitation solution (QBiogene). This mixture was inverted 10 times, centrifuged for 5 min, and the supernatant was then extracted twice with 1 mL chloroform : isoamyl-alcohol (24 : 1). RNA was precipitated by the addition of two volumes of diethyl pyrocarbonate-treated 30% (w/v) polyethylene glycol/1.6 M NaCl and incubation at room temperature for 2 h, and collected by centrifugation (10 min, 13 000 g, 4°C). After two washes in ice-cold 70% ethanol, RNA was re-dissolved in 100 μL sterile H2O. The RNeasy kit (Qiagen) was then used to clean the RNA sample, including a DNase digestion step to remove DNA contamination, following the manufacturer's instructions. RNA quality was assessed using gel electrophoresis and quantified using spectrophotometry.

RT-PCR and denaturing gradient gel electrophoresis

Reverse transcription was performed with a random hexamer primer and 100 ng total RNA per reaction using the Reverse-iT first-strand synthesis kit (ABgene, UK), following the manufacturer's instructions. cDNA generated from random hexamers was used for PCR of the V3 region of the 16S rRNA gene using primers 341F and 518R (Muyzer et al., 1993) and subsequent denaturing gradient gel electrophoresis (DGGE), as previously described (Cunliffe & Kertesz et al., 2006a). Ten percent (w/v) acrylamide/bisacrylamide DGGE gels were cast with a 40–60% denaturing gradient and run for a total of 1008 V H before being stained with SYBR Gold stain (Molecular Probes) for visualization. All RNA-DGGE experiments were carried out in triplicate, using samples from three separate microcosms for each treatment.

Quantitative PCR of catabolic genes

PCR primers were designed for the genes bphC and xylE, which encode 2,3-dihydroxybiphenyl 1,2-dioxygenase and catechol 2,3-dioxygenase, respectively, in the PAH catabolic pathway of S. yanoikuyae B1 (Kim & Zylstra et al., 1999; Fig. 1). The primer sequences used were bphCfor GGTAAGCCCG ACTACAACACAA; bphCrev GGAGGCATGCAACACGATGC; xylEfor GGCACTGACCGGTGTACTTCG; and xylErev CGA CCTTGAAGGCCATCC. PCR was performed in a total volume of 50 μL containing 20 μM dNTPs, 10 pmol of each primer, and 0.75 μL (3.75 U) Taq DNA polymerase (Roche) with the buffer provided by the manufacturer. The PCR programme for bphC consisted of (xylE conditions in parentheses) initial denaturation at 95°C for 5 min, followed by 30 cycles of 95°C for 30 s, 60°C (57°C) for 30 s, 72°C for 30 s, and a final elongation at 72°C for 10 min, in a Whatman Biometra T1 thermocycler (Biometra). The external standard for each qPCR contained 101−108 amplicon molecules μL−1, and was constructed using PCR products that had been purified using a spin-column PCR purification kit (Qiagen) and quantified spectrophotometrically.

qPCR was performed in a Roche Lightcycler (Roche), using 20-μL glass capillaries (Roche) containing 5 μL DyNAmo capillary SYBR Green qPCR master mix (Finnzymes, Finland), 0.3 pmol of each primer, and cDNA synthesized from 100 ng RNA, in a total volume of 10 μL. Running conditions for bphC analysis (xylE analysis in parentheses) in the Lightcycler were 95°C for 10 min followed by 40 cycles of 95°C for 10 s, 60°C (57°C) for 10 s and 72°C for 20 s. A melting curve was generated by heating samples from 60 to 95°C at a ramp rate of 20°C s−1.

Analysis of variance (anova) was used to highlight significant differences in qPCR results (n=3; P<0.05). Where significant differences were seen, a Tukeys test was used to compare data within a defined set. Both anova and Tukey's test were performed using spss software.

Measurement of PAH degradation

Fluorene and phenanthrene degradation was monitored as described previously (Cunliffe & Kertesz et al., 2006a). Briefly, the total organic component of the soil was extracted into acetone/dichloromethane (1 : 1 v/v) and PAHs were analysed by gas chromatography using a capillary column (phase, EC-1; length, 30 m; internal diameter, 0.25 mm; film thickness, 0.25 μm; Alltech) at a gas flow rate of 1 mL min−1, with flame ionization detection.


Survival of S. yanoikuyae B1 in PAH-contaminated soil after different inoculum pretreatments

The effect of inoculum pretreatment in bioaugmentation experiments was examined using four pretreatments to prepare S. yanoikuyae for inoculation into a PAH-contaminated soil. These were: (a) growth to exponential phase in complex medium (LB), (b) growth to exponential phase in MM-succinate medium (MM), (c) growth to exponential phase in MM-succinate medium followed by incubation in carbon-free MM medium for 4 h (MM-S), and (d) growth to exponential phase in MM-succinate medium followed by acclimation to phenanthrene in phenanthrene-saturated MM medium for 4 h (MM-E). Cell counts carried out with the treated cells immediately prior to inoculation confirmed that S. yanoikuyae B1 population density and cell viability were not significantly affected by the latter two pretreatments (data not shown).

Identical numbers of S. yanoikuyae cells were inoculated into the soil after each pretreatment (109 cells per gram soil). Survival of S. yanoikuyae in the inoculated microcosms was monitored for 40 days as CFUs, and also by examination of the active S. yanoikuyae population by RT-PCR DGGE (RNA DGGE) (see below). The population of S. yanoikuyae that could be extracted from soil immediately after inoculation was lower than that determined in the inoculum itself (Fig. 2a), suggesting that many cells either bound very strongly to the soil immediately after inoculation, or did not survive the transfer. The proportion of cells that could be recovered from the soil immediately after inoculation (day 0) varied with pretreatment, with about 20% recovery for the inoculated cells from the MM-grown pretreatments (including carbon-starved and phenanthrene-treated cells) and a significantly lower proportion for the LB-grown inoculum (Fig. 2a). The extractable population of S. yanoikuyae decreased logarithmically from day 5 to day 20 for all four treatments, before stabilizing at 105–106 CFU g−1 soil. This decrease of extractable CFUs to a stable holding capacity is normal for an inoculated strain (see e.g. Molina et al., 2000), but it contrasted unexpectedly with the population dynamics observed by RNA DGGE, discussed below.


(a) Survival of Sphingobium yanoikuyae B1 cells in soil according to pretreatment. Viable S. yanoikuyae B1 cells were extracted from soil microcosms and enumerated as CFUs on streptomycin-containing plates. (b) Total RNA extracted from microcosms inoculated with S. yanoikuyae B1. ○, inoculum precultivated in LB medium; ◻, inoculum precultivated in 20 mM succinate-MM; ▲, inoculum carbon-starved in MM; ▪, inoculum exposed to phenanthrene-saturated MM; ♦ noninoculated control. Mean±standard error (n=3).

RNA content of S. yanoikuyae B1 in soil

To confirm and extend the survival data, the total RNA in each microcosm was quantified at defined time intervals through the experiment (Fig. 2b). The total RNA content in the inoculum varied with pretreatment. Thus, S. yanoikuyae B1 LB-grown cells contained 13.2 fg RNA per cell, significantly more than for cells grown with MM-succinate (4.5 fg RNA per cell), or for the two MM-succinate-grown populations that had been further treated (phenanthrene-treated cells, 1.8 fg RNA per cell; carbon-starved cells, 0.7 fg RNA per cell). This last value represents a decrease of 85% in the 4 h of starvation treatment, confirming that the starvation response of this strain involved significant changes in RNA content. The phenanthrene-treated cells also contained lower quantities of RNA, indicating that these cells had undergone a partial starvation response.

Because LB-grown inocula contained significantly more RNA than cells from the other pretreatments, the soil inoculated with LB-grown cells initially contained more extractable RNA than the soil in the other treatments (Fig. 2b). This difference was evident in the total RNA extracted from the soils immediately after inoculation, but by day 10 the elevation in RNA content attributable to the LB pregrown cells had decreased, and total RNA levels were not significantly different among the four treatments. For MM, MM-E and MM-S treatments, the total extractable RNA did not change significantly over the study period. Importantly, the amounts of RNA obtained from the control microcosms were significantly lower than those obtained from inoculated microcosms (Fig. 2b), suggesting that the difference was attributable to the presence of the inoculum. Because the total RNA values in the microcosms were maintained at a relatively stable level throughout the experiment, these results contrast with the logarithmic decrease in the survival data measured as extractable CFUs (Fig. 2a) of S. yanoikuyae RNA in the microcosms.

To explore this further, the population of active S. yanoikuyae B1 in the soil community was evaluated by RNA DGGE over the course of the experiment. The community profiles observed for the inoculated microcosms were the same for all four pretreatments, as were their developments over time. Community profiles from uninoculated microcosms, microcosms inoculated with phenanthrene-acclimated cells, and microcosms inoculated with LB-grown cells are presented here (Fig. 3). The intensity of the band corresponding to strain B1 was approximately constant throughout the experiment, and was absent in the uninoculated microcosm, indicating that no large changes in the population size or metabolic activity of strain B1 occurred over the time period studied. This suggests that inoculum pretreatment had no effect on the ability of S. yanoikuyae B1 to remain active in the soil community. More importantly, it also suggests that the survival data obtained as CFUs significantly underestimated the soil populations of strain B1. We conclude that this may be as a result of the extractability of strain B1 from the soil decreasing significantly with time. This is consistent with the known ability of S. yanoikuyae B1 to form biofilms (Cunliffe & Kertesz et al., 2006b), where the cells are metabolically active, but not actively dividing.


RT-PCR DGGE of the V3 region of the 16S rRNA gene, showing the presence of an active Sphingobium yanoikuyae B1 population in two treatment microcosms. Control, uninoculated microcosm; MM-E, preinoculation cells exposed to phenanthrene-saturated MM; LB, preinoculation cells grown in LB medium. The band corresponding to S. yanoikuyae B1 is indicated, as are three further bands that increased in intensity after inoculation with S. yanoikuyae B1 (bands S1−S3).

Although the native bacterial community structure remained fairly stable over time within each treatment (Fig. 3), three bands in the community profile were only observed after inoculation with S. yanoikuyae B1 (Fig. 3; bands S1, S2, S3). Band S1 in particular was quite intense in the presence of S. yanoikuyae B1, indicating that this native population benefits from the presence of the inoculated strain.

PAH catabolic gene expression of S. yanoikuyae B1, and PAH degradation in inoculated microcosms

The expression of specific genes required for PAH metabolism by S. yanoikuyae B1 was monitored by quantitative RT-PCR. Primers were designed to amplify fragments of genes encoding an early and a late enzyme in the PAH-degradation pathway, namely 2,3-dihydroxybiphenyl 1,2-dioxygenase (bphC), and the ring-cleavage enzyme catechol 2,3-dioxygenase (xylE) (Fig. 1). Expression of bphC in the inoculum was strongly affected by pretreatment, and LB-grown cells had significantly higher levels of bphC expression compared with all other treatments (Fig. 4a). This suggested that the LB medium used may have contained a specific inducer of the PAH-degradation pathway, but expression levels of bphC were very similar when the strain was cultivated in other media containing different complex components (Nutrient Broth, King's B medium, R2A medium). xylE expression showed a different pattern from that observed for bphC, with carbon-starved cells showing significantly reduced xylE expression compared with the other treatments (Fig. 5a). Expression of bphC in LB-grown cells was 48-fold higher than xylE expression in the same cells.


qPCR of 2,3-dihydroxybiphenyl 1,2-dioxygenase (bphC) expression in Sphingobium yanoikuyae B1. (a) bphC expression in inocula after growth in LB medium, MM medium, phenanthrene exposure (MM-E) or starvation (MM-S). (b) bphC expression after inoculation into an aged contaminated soil (inset image − increased scale). ○, inoculum precultivated in LB medium; ◻, inoculum precultivated in 20mM succinate-MM; ▲, inoculum carbon-starved in MM; ▪, inoculum exposed to phenanthrene-saturated MM; ♦ noninoculated control. Mean±standard error (n=3).


qPCR of catechol 2,3-dioxygenase (xylE) expression in Sphingobium yanoikuyae B1. (a) xylE expression in inocula after growth in LB medium, MM medium, phenanthrene exposure (MM-E) or starvation (MM-S). (b) xylE expression after inoculation into an aged contaminated soil (insert image − increased scale). ○, inoculum precultivated in LB medium; ◻, inoculum precultivated in 20 mM succinate-MM; ▲, inoculum carbon-starved in MM; ▪, inoculum exposed to phenanthrene-saturated MM; ♦ noninoculated control. Mean±standard error (n=3).

After inoculation into the soil, expression of the two genes followed a similar trend, with bphC and xylE expression both significantly higher in the microcosms inoculated with LB-grown cells than in those for the other treatments (Figs 4b and 5b). Even though gene expression levels decreased with time, significantly higher bphC expression was still seen in microcosms inoculated with LB-grown cells up to day 20 (P≤0.005). xylE expression declined rapidly compared with bphC expression, and microcosms containing LB-grown cells showed expression levels similar to those of other treatments at day 5. Significant fluorene and phenanthrene degradation was observed only in microcosms that had been inoculated with LB-grown cells (Table 1), and this correlated well with the elevated bphC and xylE gene expression observed in microcosms inoculated with LB-grown cells.

View this table:

Fluorene and phenanthrene concentrations (μg g−1 soil wet weight) measured by gas chromatography in contaminated soil microcosms inoculated with Sphingobium yanoikuyae B1


In this study the effect of inoculum pretreatment on the survival and PAH-catabolic gene expression of S. yanoikuyae B1 in an aged PAH-contaminated soil was examined. Neither carbon starvation nor acclimation to phenanthrene for 4 h led to significant enhancement of bphC or xylE expression. Pregrowth of the strain with complex medium (LB) gave much higher levels of PAH-degradation gene expression than pregrowth with minimal medium, and PAH degradation in the soil over the length of the study (40 days) was correspondingly increased with the LB-grown inoculum.

When adding a bacterial inoculum to a contaminated soil system, it is important that the inoculated strain is able to adapt quickly to the new environment and utilize the contaminant carbon sources present. In response to nutrient starvation, many bacteria undergo changes in expression of specific genes of carbon metabolism, but they also experience changes in macromolecular composition, including a severe reduction in cellular RNA content. In Pseudomonas putida, for example, ribosome content falls by 50% within 2 h, and continues to fall over several weeks (Givskov et al., 1994), while Rhizobium leguminosarum showed a 57-fold reduction in RNA synthesis on extended carbon starvation (Thorne & Williams et al., 1997). We observed that after 4 h of carbon starvation the RNA content of carbon-starved S. yanoikuyae B1 cells fell to 15% of that for exponential-phase cells, a value similar to that obtained with Sphingomonas sp. RB2256 (Fegatella et al., 1998), a marine ultramicrobacterium adapted to oligotrophic conditions that showed a reduction in ribosome content of 90% during starvation.

Because of the oligotrophic nature of the soil environment, we anticipated that prestarved cells would be better adapted for survival in contaminated soils. However, the CFU survival curves were very similar for cells grown in minimal medium and those subsequently starved or phenanthrene-treated (Fig. 2a), although the starved cells gave a slightly higher final population. This is consistent with results obtained in a previous study on Pseudomonas fluorescens (van Overbeek et al., 1995), although the earlier report did not compare survival after pregrowth in complex medium, where we found survival to be significantly impaired (Fig. 2a). However, RNA DGGE analysis (Fig. 3) showed that the active population of strain B1 in the microcosms did not decline in parallel with the extractable CFUs, but remained approximately constant throughout the experiment, and this was reflected in the total RNA that could be extracted from the microcosms at each time point (Fig. 2b). Because of the inherent instability of RNA released from dead cells in soil, this suggests that the CFU decrease did not arise from cell death, but rather from the cells binding to soil. A similar effect has been seen with P. fluorescens (van Overbeek et al., 1995), where cell counts were measured by immunofluorescence, and with Ralstonia eutropha (Watanabe et al., 2000; cells binding to activated sludge flocs), using qPCR instead of colony counting. Interestingly, the latter study showed a significant increase in inoculum survival in activated sludge when cells had been starved for 2 days, but this effect was lost after longer starvation (7 days). A decrease in colony counts while maintaining total cell population is consistent with the inoculum entering a ‘viable but not culturable state’, but this does not seem to be happening here because total rRNA levels remain approximately constant, suggesting that the cells are metabolically active. Since strain B1 is an effective biofilm-former (Cunliffe & Kertesz et al., 2006b), the decrease in extractable population may be because the inoculum adheres strongly to soil particles and remains metabolically active in a biofilm state while becoming more difficult to extract from the soil.

Unexpectedly, the expression of genes specific to the PAH-degradation pathway was not strongest in cells that had been carbon-starved before inoculation, nor in cells that had been subjected to acclimation with phenanthrene, but rather in cells that had been pregrown in complex medium (Figs 4 and 5). Degradation of phenanthrene and fluorene over the period of the experiment was correlated with bphC and xylE gene expression measured by qPCR, and was only significant for soils inoculated with LB-grown cells. bphC and xylE expression were much higher in LB-grown cells even prior to the addition to soil – this effect was seen after growth in several complex media, and probably reflects higher overall levels of gene expression in the cells, owing to higher overall metabolic activity. Earlier studies on naphthalene degradation in P. putida and Alcaligenes sp. showed that there is considerable constitutive naphthalene mineralization even after growth in Nutrient Broth, although gene expression levels were not reported (Guerin & Boyd et al., 1995). Once induced, naphthalene mineralization in the two strains was stable in the cells for many months, even in the absence of inducer, and although this may have been a result of extreme stability of the relevant enzymes (Guerin & Boyd et al., 1995), low but constitutive expression of the corresponding genes is more likely. In P. fluorescens, by contrast, expression of the bphC gene was reported to be very low in contaminated soil (Brazil et al., 1995), demonstrating that the regulation of PAH-catabolic genes is very strain-dependent.

The elevated levels of bphC and xylE expression observed after growth in LB medium are matched by a slower drop in the expression of these genes over time compared with other pretreatments (Figs 4 and 5). This suggests that a major factor involved in maintaining specific gene expression is the increased initial fitness of LB-grown cells compared with other treatments, particularly with regard to intracellular energy and nutrient reserves, and this is borne out by the fact that similar levels of PAH-catabolic gene expression were observed after growth in other complex media. In addition, enhanced nutrient reserves may be important for motility in soil (P. putida G7, for example, exhibits chemotaxis to naphthalene, although it only remains motile for 1 h in buffer without an exogenous carbon source; Grimm & Harwood et al., 1997). Enhanced chemotaxis would increase exposure to PAHs in the soil and stimulate expression of PAH-specific genes. However, inoculum cells subjected to all four tested pretreatments displayed similar levels of chemotaxis towards phenanthrene-saturated medium (data not shown), using a standard capillary assay (Cunliffe & Kertesz et al., 2006b), so if this property is impaired in starved cells inoculated into the soil, it is not apparent in the inoculum.

Sphingobium yanoikuyae B1 is not limited to the use of PAHs as carbon source, but can metabolize a broad range of ecologically relevant carbon sources, including amino acids, organic acids and phenolic compounds (Cunliffe & Kertesz et al., 2006a). All the S. yanoikuyae B1 populations in this study, except that derived from LB-grown cells, showed a rapid reduction in PAH-catabolic gene expression after inoculation into the soil, but they still survived well within the soil community (Fig. 3). Given the broad range of compounds within the metabolic scope of this organism it seems probable that populations expressing PAH-catabolic genes at a lower level are accessing other carbon sources. Indeed, multiple-substrate utilization is a potential strategy to allow access to poorly available substrates such as PAHs (Wick et al., 2003), either by constitutive expression of PAH-catabolic genes, or via their induction by other substrates, such as plant terpenes (Jung et al., 2002). The same is probably true in the aged PAH-contaminated soil used in this study, with low PAH availability limiting induction of the PAH-degradation genes.

Inoculation of S. yanoikuyae B1 into the contaminated soil used here had very little effect on the community already present in the soil. However, the presence of strain B1 did enhance the active population corresponding to one native RNA DGGE band (Fig. 3, Band S1). This effect has also been seen at the level of DNA DGGE, and the obtained sequence was the same as that found here, closely related to known sphingomonad sequences (Cunliffe & Kertesz et al., 2006a). Potential reasons for this stimulation include the utilization of PAH metabolites released by strain B1, or strain S1 profiting from other contaminants mobilized by strain B1. Interestingly, however, S. yanoikuyae B1 promoted activity of the band-S1 population for all four pretreatments, and hence independently of enhanced bph gene expression. This indicates that the community interactions revealed here involve other processes that have yet to be explored.


We are grateful to Rajasekaran Palanisamy for help with soil RNA extraction, and to David Cooke, Andrew Surdevan and Carole Webb for advice on PAH analysis. This work was supported by the Natural Environment Research Council.


  • Editor: Max Häggblom


View Abstract