Fluorescence Resonance Energy Transfer (FRET) based molecular detection of a genetically modified PCB degrader in soil


*Corresponding author. Tel.: +353-59-9716220; fax: +353-59-9170517, E-mail address: david.dowling@itcarlow.ie


Genetic analysis of the location of a mini-Tn5 promoted insertion of the LB400 bph operon in the rhizosphere coloniser Pseudomonas fluorescens F113rifPCB, allowed the development of a specific PCR detection system based on the unique DNA sequence at this insertion site. Real time PCR using both SYBR green chemistry and Fluorescence Resonance Energy Transfer probes allowed the precise identification of the recombinant strain and its quantitative detection in soil microcosms over a (bacteria/g) range of five orders of magnitude. This new assay can detect the genetically modified microorganism from soil in less than 90 min and at levels below the detection limits of standard PCR or cultivable counts on selective media.


Polychlorinated biphenyls (PCBs) are chemically inert compounds which have long been used in hydraulic fluids, electrical components, plasticisers, paints, adhesives and casting agents [1]. Their use was banned when they were identified as suppressors of the immune and reproductive systems. However, due to their chemical and physical properties they persisted in the environment, where they accumulated in food chains and in the fat tissue of animals and humans [2–5]. The major source of PCBs in the environment is redistribution due to volatisation from soil and water after disposal of wastes and leakage from electric transformers, resulting in PCB detection from undisturbed regions such as the arctic. PCBs are listed on both the Dutch and American priority lists and research is ongoing worldwide on methods to clean up this dangerous group of environmental contaminants. Bioremediation offers a solution to the treatment of such dilute pollutants. One of the most promising bioremediation strategies is rhizoremediation which involves the use of plants and rhizosphere microorganisms (for recent review see [6]). Bacterial colonisation of plant roots is a complex process and a possible approach to improve rhizoremediation is to introduce degradation genes into good root colonisers, for example, TCE degradation [7] and PCB degradation [8]. The genetically modified microorganism (GMM), Pseudomonas fluorescens strain F113rifPCB, was constructed by introducing the bphLB400 operon (for biphenyl and PCB co-metabolism) into the chromosome of the root colonising strain F113. This strain has the ability to break-down PCBs similar to the donor of the bphLB400 operon, Burholderia sp. strain LB400, and can be used in the plant rhizosphere to aid the rhizoremediation of these compounds. At the single cell level, F113rifPCB cells containing a gfp reporter construct can degrade and sense PCB-2 in the alfalfa rhizosphere [9].

In any analysis of a bioremediation process it is important that the introduced or inoculant strain can be accurately identified and enumerated. This is of particular importance if the inoculant is a GMM, where molecular methods in addition to plate counts should be developed [10]. The location and DNA sequence of the bphLB400 insertion site in F113rifPCB can provide the basis of a unique strain detection system. This report describes the sequence analysis of the insertion site and the development of a Real-time PCR assay, which made use of SYBR green chemistry and amplicon specific, i.e. Fluorescence Resonance Energy Transfer (FRET) probes. FRET probes give a high level of specificity due to the use of two adjacent hybridisation probes. In this work they were used to identify and quantify the PCB degrader, F113rifPCB, from the soil environment at levels that are below the detection limit of conventional plate counts.

2Materials and methods

2.1Strains, media and reagents

Bacterial strains are described in Table 1. Bacterial strains were routinely grown on Luria–Bertani (LB) or Pseudomonas minimal media (MM) plates supplemented with selective agents as required [8,11]. On MM plates, biphenyl (BP) or 4-chlorobiphenyl (4-CBP) crystals were provided as a carbon and energy source by placing crystals on the lid of the petri dish and sealing with parafilm before incubation [12]. The detection of a functional bphC gene by spraying colonies on LB plates with 2,3-dihydroxybiphenyl (2,3-OHBP) indicating the presence of the bph operon, was also routinely used, as described previously [13].

Table 1.  Bacterial strains used in this study
StrainDescriptionRelevant characteristicsSource or reference
  1. Flu = production of fluorescent siderophore, phl = production of antifungal compound 2,4-diacetylphloroglucinol, HCN/Protease = production of hydrogen cyanide and extracellular protease.

P. fluorescens F113 WTSugarbeet rhizosphere isolateFlu+; phl+; HCN+; protease+[29]
P. fluorescens F113 rifSpontaneous Rifampicin resistant mutant of F113Rifr[8]
P. fluorescens F113rifPCBbphLB400 cassette inserted in F113rifRifr; bph+[8]
P. fluorescens F113rifPCBlac3.1bphA promoter-lacZ insertion in F113rifpcbRifr; bph+; lac+; SpcR/SmR[8]
P. fluorescens F113lacZYF113 with mini-Tn5 lacZY insertionLac+, growth on lactose+[30]
E.coli CC118 λpir pDDPCBTransposable bph operon (mini-Tn5) constructAmpr; MobRP4; Pttr; Tnp[13]
Pseudomonas species B13 FRIPCB3-chlorobenzoate degrader with Bph cassette randomly inserted in chromosome3CB+; bph+[13]

2.2DNA manipulations

Genomic DNA from pure cultures was obtained using Wizard® Genomic DNA purification kit as per manufacturers instructions. The DNA was diluted to give standards of 100–106 gene copy numbers. Gene copy numbers were calculated from concentrations of the positive control assuming 6.48 × 1012 bp/μg of DNA, approximately 6.5 Mbp per genome, and one gene copy per genome [14]. Additional DNA techniques were by published protocols [11]. Sequencing of cloned DNA was carried out either using an ABI Prism® 310 Genetic Analyser (Applied Biosystems) or commercially by MWG-Biotech. Sequence analysis was carried out using the nucleic acid and amino acid basic local alignment sequence tool (BLAST) analysis programs [15], CLUSTALW [16] and Omiga software (Accelrys). Standard PCR reactions were carried out on a Hybaid PCR Sprint Thermal Cycler or OMNI-E Thermal Cycler using standard protocols [17]. PCR primer pairs used in this study were designed using Primer 3 [18] and were synthesised by Sigma–Genosys.

2.3Soil microcosms

A medium to heavy textured free draining grey podzolic soil derived from limestone boulder clay (40–45% clay and 35% silt) was inoculated with P. fluorescens F113rifPCB at 1.24 × 108 to 1.24 × 102 colony forming units per g (CFU/g), to establish the efficacy of the detection method and to quantify the strain. After inoculation, 10 g soil microcosms were maintained overnight at 4 °C before analysis. Soil DNA was isolated with a MoBio Soil DNA extraction kit as per manufacturers instructions

2.4Southern blotting and hybridisation analysis

To determine the number of random insertions of the bph operon into the P. fluorescens F113rifPCB chromosome, total genomic DNA was restricted using HindIII, electrophoresed and Southern blotted [19]. A 1.1 kbp DNA fragment encoding the bphC gene (and part of the bphB gene) was isolated as an EcoR1 fragment from the plasmid clone p12C (Sherlock, O and Dowling, D.N. unpublished), purified using the QIAEX II gel extraction kit, labelled using the DIG DNA Labelling Kit (Boehringer Mannheim) and used to probe the Southern blot. The resulting membrane was exposed to Kodak X-OMAT LS film for 20 min in a hypercassette (Amersham lifesciences; 18 × 24 cm). The film was developed using Kodak GBX developer and fixer according to the manufacturers instructions.

2.5Isolation of sequences of the bph insert::F113 chromosome junction region

To identify the junction region between the terminal end of the bph operon and the P. fluorescens F113rif chromosome, the Vectorette cloning kit (Sigma–Genosys) was used according to manufacturers instructions. This kit allowed a region of DNA to be amplified by PCR despite the sequence of only one end of the target DNA being available. The P. fluorescens F113rifPCB chromosome was restricted with EcoR1. A ligation was carried out between the restricted chromosomal DNA and EcoR1 ended “vectorette units”. A PCR was then carried out using a primer (P1 ∼GGGATGTGACTGCCAGACTTGG∼) designed from the internal sequence of the bph operon (bphD gene) [20] and a second primer designed to amplify from the inserted EcoR1 unit. A 750 base pair (approx) fragment downstream of the bphD gene was obtained. This was cloned into a pGEM T-easy vector and sequenced. As the transposon inverted repeat sequence ACTTGTGTATAAGAGTCAG [21] was not evident in this fragment, a second round of EcoR1 based vectorette reactions was carried out using a forward primer (P4 ∼GGCTACAGCATCGGCTTCT∼) and a nested primer (P5 ∼GGCTCGGTACGGCACTAAC∼) based on this bphD downstream sequence. The entire junction region, starting from the 3′ end of bphD has been sequenced (Accession Number AJ493164).

2.6Design of realtime PCR primers and amplification of the junction region

The oligonucleotide primers used in this study were designed using – OMIGA version 1.1.3 (Accelrys) and primer 3 software [18]. The primers; OD fwd 5′GCTTCTCGGTTCAGAAGCTG3′ and OD rev 5′CTGCTGATCCTGCTGTTCAA3′ were manufactured and provided by Sigma Genosys, UK. This OD primer pair amplified a 146bp DNA section that contains the terminal end of the putative bphO gene of strain LB400 and a fragment of the P. fluorescens F113 chromosome (Fig. 3). Real-time PCR was carried out using SYBR Green chemistry and the Roche Lightcycler(R). Quantification of PCR product was carried out by comparing the fluorescence of a PCR product of unknown concentration with the fluorescence of several dilutions of a specific external standard. By defining the background fluorescence of a set of standards, the fluorescence data could be used for quantification. The parameter, Ct (threshold cycle), was the fractional cycle number at which the fluorescence emission crossed an arbitrarily defined threshold within the logarithmic increase phase. The higher the amount of initial template DNA, the smaller was the Ct. The Ct values obtained for each sample were compared with a standard curve to determine copy number of the target region.

Figure 3.

Sequence of F113rif and the bph operon junction site. An amplicon size of 146bp is expected with the OD primer pair fwd and rev. The amplicon encompasses 81bp of the putative bphO, the 19bp Tn5 IR region (underlined and boxed) and 46bp of the F113 chromosome. An amplicon of 197bp is expected with the FRET primers 12 prb fwd and 12 prb rev.

2.7FRET primer and probe design

FRET PCR utilises fundamentally different chemical and thermal cycling conditions than standard PCR. The inclusion of internal FRET probes constrains the design of the PCR primers [22], so these primers were designed manually. The primers 12-prb-fwd (5′-GTTTTCAGCACCAAAGATGG-3′) and 12-prb-rev (5′-CTGATCCTGCTGTTCAACC-3′) (MWG-Biotech, Germany) targeted a 297 base pair product, (encompassing the region targeted by SYBR green PCR) extending from the bphO gene into the P. fluorescens F113rifPCB chromosome (Fig. 3). The probes EJRHGM-1-12 (5′-TTGCGGCCGCACTTGTGTAT-3′FLU) and EJRHGM-2-12 (5′-LC Red640-AGAGTCAGGCACCGTGGCAT-3′-P) (TIB Molbiol, Germany) were also designed manually to overlie the bphO– Tn5 IR end – F113 chromosome junctions, (Fig. 4), while allowing for their close proximity requirements and a 5–10 °C melting temperature in excess of the primers. The detection probe EJRHGM-1-12, was labelled with fluorescein, and the anchor probe EJRHGM-2-12 was labelled with Lightcycler Red 640 and was also modified at the 3′ end by phosphorylation to block extension during the PCR reaction.

Figure 4.

Schematic representation of the bphLB400 operon and its insertion into the P. fluorescens F113rif chromosome. The figure also highlights the location of two FRET probes (EJRHGM-1-12 and ERJHGM-2-12) in relation to the unique bphO– Tn5 IR-F113rif chromosome junction.

2.8FRET PCR optimisation

PCR was performed on the Lightcycler™ Instrument (Roche Diagnostics, Mannheim, Germany) in lightcycler glass capillaries (Roche) as 20 μl reactions and the final optimised reaction conditions were as follows: the final concentration for each reaction was, 1 μl of DNA template (less than 50 ng of total genomic or soil DNA), 10 pmol of each primer, 3 μM of each probe, 4 mM of MgCl2 and 2 μl of lightcycler DNA master hybridisation probe mix (Roche). The conditions for thermal cycling were as follows: initial denaturation at 94 °C for 30 s, followed by 50 amplification cycles at 94 °C for 3 s, 56 °C for 8 s and 72 °C for 12 s. Fluorescence was measured at 640 nm (F2 channel) at the end of each annealing phase [23]. The amplification was followed by a melting program which started at 55 °C for 30 s and then increased to 95 °C at 0.1 °C/s, with the fluorescence signal being continuously monitored. The entire PCR process took approximately 36 min.


3.1P. fluorescens F113rifPCB contains a single copy of the bphLB400 operon

It was necessary to confirm that only one copy of the bphLB400 operon was present in F113rifPCB. Southern blot and hybridisation analysis were used to determine the number of insertions of the mini-Tn5 containing the bphLB400 operon that took place in the construction of P. fluorescens F113rifPCB. Total genomic DNA from P. fluorescens F113rifPCB and P. fluorescens F113 wild type was prepared for use in the Southern blot procedure. The restriction enzyme HindIII was used to restrict the genomic DNA from these strains, as there was a single HindIII restriction site upstream of the bphLB400 operon, ensuring that the operon would remain intact when the genomic DNA was restricted. The expected fragment size was >12 kbp and was dependant on the location of the mini-Tn5 insertion in the chromosome. One bphLB400 insertion site was detected in the strain P. fluorescens F113rifPCB and no operon insertion was detected in parent strain P. fluorescens F113 WT (Fig. 1). The presence of a single copy of the bphLB400 operon allowed sequencing and further analysis to be carried out to determine the chromosomal location of the operon in P. fluorescens F113rifPCB.

Figure 1.

Southern blot showing a single mini-Tn5 insertion as HindIII DNA fragments hybridising to a bphC DNA probe. P. fluorescens strains shown on the gel as follows: Lane 1: size standards, Lane 2: P. fluorescens F113 WT and Lane 3: P. fluorescens F113rifPCB.

3.2The bph operon in P. fluorescens F113rifPCB contains an additional predicted gene, bphO

As a result of operon fragment amplification, cloning and sequencing, a 1522 bp sequence downstream of the bphD stop codon and including the bph operon/P. fluorescens F113rif junction region was obtained (Accession Number AJ493164). Sequencing of the bphD 3′ terminal end of the bph operon identified a putative open reading frame. This ORF, referred to as bphO, encoded 185 amino acids and has not been previously described from LB400. The deduced amino acid sequence showed significant homology to a number of proteins in the databanks. Most significant was homology to an open reading frame (ORF1) in Pseudomonas sp. KKS102. ORF 1 is also located in the bphD 3′ terminal of the bph operon. The function of the bphO gene and ORF1 from strain KKS102 has not been elucidated. Other genes with some similarity to bphO, include, doxH of the dox operon (Pseudomonas species), the naqQ gene (Ralstonia) and the pahQ gene (P. aeruginosa, P. putida), [24–26].

3.3The insertion of the bphLB400 operon in P. fluorescens F113rifPCB interrupts an F113 ORF encoding a putative ABC transporter protein

The mini-Tn5 inverted repeat (IR) [21] was identified and located in the junction region (Fig. 3) indicating the precise location of the transposon promoted insertion site of the bph operon in the P. fluorescens F113rifPCB chromosome. BlastN analysis of the 703 base pair region downstream of the bph P. fluorescens F113rif chromosome junction revealed an 88% identity with a region of P. aeruginosa strain PAO1 complete genome sequence [27]. BlastX analysis revealed that this ORF corresponded to a probable ATP-binding/permease fusion ABC transporter PA0860 (72% positives) (Fig. 2). This is a 596 amino acid protein of which 196 amino acids are evident in this sequence, indicating that the insertion of bph interrupted a potential P. fluorescens F113rif homologue of this gene. The insertion has not had a detectable effect on the fitness of the strain. It has been previously shown that neither growth nor rhizosphere colonisation are affected [8].

Figure 2.

Bioinformatic analysis of the bphLB400 operon insertion site of P. fluorescens F113rifPCB. Information from P. fluorescens genome sequence adapted from the MIPS database of unfinished genome sequences, PEDANT (http://pedant.gsf.de/).

3.4PCR amplification of the bphO-F113 junction sequence is the basis of a unique molecular detection system for F113rifPCB derivatives

Real time PCR amplification of the terminal end of the bphLB400 operon bphO and the F113 chromosome insertion site (using the OD primer set) enabled the unique detection of F113rifPCB strains. No amplicon is produced from the F113 wildtype strain or the PCR negative control. The size of the amplicon was confirmed by agarose gel electrophoresis and its identity confirmed by DNA sequencing (data not shown).

The specificity of these primers was tested against several GMM bacteria in a Real-time PCR experiment. The melting curve data produced using this primer set shows that the fragment of DNA amplified was unique to the target strain (Fig. 5).

Figure 5.

Melting curve analysis of bphO-F113rif chromosome junction region showing specificity of amplification following real-time PCR (SYBR green) with the OD primer set. Sample 1: F113rifPCB, sample 2: F113WT, sample 4: F113rifPCBlac3.1; sample 5: Pseudomonas species B13FR1PCB, sample 6: P. fluorescens F113lacZY, sample 7: plasmid pDDPCB and samples 10 and 11: water negative controls.

The only strains that produced an amplicon with these primers were F113rifPCBlac3.1 and its precursor, F113rifPCB (Fig. 5). Another genetically modified F113 derivative, containing a miniTn5-lacZY insertion, F113lacZY yielded no amplicon. Likewise, no amplicon was produced from the cloning vector harboring the bphLB400 operon, pDDPCB (Fig. 5). Pseudomonas sp. B13FRIPCB [13], is a genetically modified strain containing a random insertion of the bphLB400 operon using pDDPCB, did not produce a product. Similarly, other F113 derivatives with random insertions of bphLB400 did not give positive results, confirming the specificity of assay for F113rifPCB and F113rifPCBlac3.1.

3.5FRET PCR based detection of P. fluorescens F113rifPCB in soil

The FRET system has a much higher degree of sensitivity and confidence than other detection/enumeration methods, such as colony forming units and standard PCR. Using it for F113rifPCB required optimisation of MgCl2 and primer concentrations using SYBR green chemistry. The calculated concentrations were then transferred to the FRET hybridisation reactions and modified to give the best amplification. To apply the FRET assay to quantify the strain, a standard curve was constructed. This standard curve, Fig. 6(b), was generated from the crossing point parameter (Ct threshold) using genomic DNA of F113rifPCB diluted to estimated standards [14] of 100–106 junction target copy numbers, Fig. 6(a). The melting curve profile (Fig. 6(c)) shows the amplicon peaks from the soil microcosm experiment (at the temperature 67.4 °C) and confirmed the specificity of the PCR reaction [14].

Figure 6.

(a) Amplification plot for FRET hybridisation reaction using F113rifPCB standards, crossing points (Ct values) are calculated from such curves. (b) FRET standard curve for F113rifPCB (•) standards and the determination of copy number in positive samples (X). (c) FRET melting peak, a confirmatory test for the PCR detection of F113rifPCB from soil.

Soil microcosms were inoculated with a 10-fold dilution series of F113rifPCB (Table 2) and incubated overnight. Total DNA was extracted and analysed by FRET, standard PCR and selective plate counts. The number of target amplicons (corresponding to F113rif-bph junction copies) was calculated from the standard curve and the data is summarised in Table 2. In this experiment the limits of cultivable detection using selective plate counts was 2.3 × 103 (corresponding to sample number 5) which also was the sample that showed the limit of detection by standard PCR. In contrast, FRET PCR allowed detection and enumeration to the end of the dilution series (input of 102 CFU/g), which was two orders of magnitude more sensitive than standard PCR or plate counts. As expected, microcosms that were not inoculated did not produce a signal with any detection method. The GMM could be also detected with a similar level of sensitivity in a naturally PCB contaminated soil microcosm (results not shown).

Table 2.  Enumeration of the GMM PCB degrader (F113rifPCB) in soil by FRET PCR
Microcosm numberEstimated inoculated CFUs/mlEstimated junction copies/g soilDetected by standard PCR
  1. *=junction copy numbers were too high to read off standard curve.

  2. +=junction copy detected using standard PCR.

  3. ND=not detected.

  4. The limit of detection of cultivable cells (CFU/g) recovered from the microcosms for this experiment was 103 corresponding to sample 5.

31062.5 × 107+
41053.17 × 106+
51045.63 × 105+
61031.5 × 105ND
71024 × 103ND


Vectorette PCR of the F113rifPCB chromosome enabled us to map the location of the bphLB400 operon in this GMM. A total of 1522bp of DNA encompassing the terminal end of the bphD gene, uncharacterised Burkholderia sp. LB400 DNA sequence and P. fluorescens F113rif chromosome was cloned sequenced and analysed.

In the 1522bp DNA sequence, the Tn5 IR [21] was identified indicating the transposon insertion site of the bphLB400 operon in the P. fluorescens F113rif chromosome. BlastN and Clustal W analysis of the 703 base pair region downstream of the bphLB400/P. fluorescens F113rifPCBlac3.1 chromosome junction revealed an 88% identity with a probable ATP-binding/permease fusion ABC transporter PA0860 (72% positives). There are multiple copies of such transporters in the P. fluorescens genome, which explains why such an insertion does not have a significant detrimental effect on the cell [8,12]. This region of the F113 chromosome may provide a permissive site for the integration of recombinant genes into P. fluorescens.

DNA sequencing and subsequent analysis of the bphD 3′ terminal end of the bph operon identified a putative open reading frame. This ORF, referred to as bphO, encodes 185 amino acids. A gene, doxH, with some similarity to bphO, was identified from Pseudomonas sp. C18 as one gene in a group of genes encoding proteins capable of metabolising dibenzothiophene. However, a function for DoxH in this pathway was not determined. It was suggested that doxH and/or doxJ may be involved in conversion of 2-hydroxy-4-2-(2′-oxo-3,5-cyclohexadienyl)-buta-2,4-dinoate to cis-o-hydroxybenzylidnepyruvate via the Kadoma pathway [24]. It is possible that bphO may be involved in the processing of a similar intermediate of the lower biochemical pathway for the degradation of PCBs.

A rapid and unique detection method for P. fluorescens F113rifPCB was developed based on this new sequence information. The OD primer pairs used to detect this strain were designed to span a DNA sequence including the terminal end of bphLB400 and the insertion site of this operon in the F113 chromosome (bphO/F113 junction region). This DNA amplicon appears to be unique to F113rifPCB and F113rifPCBlac3.1. The inherent difficulties with the SYBR green system, e.g., primer dimer formation, was resolved by using a FRET based Real-time PCR. By using a similar primer set and FRET hybridisation probes for the junction region, a precise detection and quantification assay was developed. In a soil microcosm experiment with authochtonous microbiota, FRET detection and enumeration was possible over an inoculum density range of five orders of magnitude. High inoculum levels (>107/g soil) were detected, but were difficult to quantify accurately. These results are similar to a Real-time PCR TaqMan assay that was developed to detect and quantify a recombinant Rhodococcus sp. RHA1 strain in soil [28]. We found that the FRET assay was 100-fold more sensitive than plate counts or standard PCR in enumerating F113rifPCB from soil. This is in contrast to a TaqMan real-time quantification of a engineered P. putida strain, where the PCR based quantification was 10-fold lower than plate counts under conditions of substrate depletion [26].

In conclusion, DNA sequencing of the integration junction of recombinant::host DNA can be used to develop a real-time FRET PCR assay for the unambiguous detection and quantification of a GMM PCB degrader in soil. Fast and sensitive tools for the analysis of population dynamics of augmented bacteria will be essential for the effective bioremediation and management of contaminated sites.


This work was supported in part by grants from the Higher Education Authority of Ireland (HEA) PRTLI programmes and European Union (EU) contracts BIO4-CT-2227, QLK3-CT2000-00164 & QLK3-CT2001-00101.