TEXT

Top Abstract Text References

Due to the widespread use of petroleum products and the current regulations requiring underground storage tanks to be upgraded, replaced, or closed by December 1998 (4), the number of petroleum-contaminated sites has abounded. Of particular concern for drinking water quality are the more water-soluble components, benzene, toluene, ethylbenzene, and xylenes (BTEX). Natural attenuation, which relies on in situ biodegradation of pollutants, has received a large amount of attention, especially for petroleum contaminants (15). While microorganisms capable of biodegradation of BTEX compounds are usually present at these sites, there is a need to know whether or not conditions are favorable for biodegradation to occur. A recent approach to determine whether compounds are bioavailable and what conditions are favorable for degradation is the use of whole-cell bioluminescent reporters (9).

Bioluminescent reporters have been widely used for the real time nondestructive monitoring of gene expression. Heitzer et al. (8) developed a quantitative assay for naphthalene bioavailability and biodegradation by using a nah-lux reporter strain constructed by King et al. (13) and expanded its use as an on-line optical biosensor for application in groundwater monitoring (10). Other lux fusions have been constructed for monitoring the expression of catabolic genes, including those for degradation of isopropylbenzene (21) and toluene (1, 5). lux fusions have also been constructed for monitoring heat shock gene expression (24, 25), oxidative stress, (3), the presence of Hg(II) (20), and alginate production (26). In all of these cases, the lux fusions were plasmid based and were constructed by placing the promoter of interest in front of the promoterless lux genes from Vibrio fischeri contained in pUCD615 (18).

In this study, a strategy was pursued to introduce a single copy of the lux fusion into the bacterial chromosome via a transposon delivery system. A mini-Tn5 delivery vector constructed by Herrero et al. (11) provided the basic model for this work. By this approach, a tod-lux fusion was constructed and introduced into Pseudomonas putida F1 to examine the induction of the tod operon when exposed to BTEX compounds and aqueous solutions of JP-4 jet fuel constituents. Since this system contains the complete lux cassette (luxCDABE), bacterial bioluminescence was measured on-line in whole cells without addition of an aldehyde substrate. The resultant strain was also evaluated for its stability and fitness compared to those of the wild-type strain, F1.

Organisms and culture conditions. The strains used in these experiments are shown in Table 1. All cultures were grown at 28°C with appropriate antibiotic selection, except for Escherichia coli strains, which were grown at 37°C.

DNA isolation and manipulation. Large-scale plasmid DNA isolation was done by a modified alkaline lysis protocol (16). Chromosomal DNA was prepared by the protocol outlined by Ausubel et al. (2). All DNA preparations were further purified by CsCl-ethidium bromide ultracentrifugation (19). DNA modifications and restriction endonuclease digestions were performed as outlined by Sambrook et al. (19).

Transposon and plasmid construction. The transposon mini-Tn5KmNX was constructed with two 58-base oligonucleotides 5' and 3' with respect to the kanamycin resistance gene (Km^r) in pCR II (Invitrogen, San Diego, Calif.) (I end, 5'GGGCGCTAGCGAAATGTTGACTGTCTCTTGATCAGATC TTTCAATTCAGAAGAACTCG3'; O end, 5'CGAATTCTGACTCTTATACACAAGTTCTAGATTGCGGCCGCTTGG TTAAAAAATGAGC3'). Oligonucleotides were synthesized with a Beckman Oligo 1000 DNA synthesizer (Palo Alto, Calif.). Base substitutions were made to generate both I and O insertion sequences as well as unique NotI and XbaI sites inside the transposon for cloning. An extra adenine was mistakenly added between the NotI and XbaI sites in the O primer, but it did not affect the construction. Primers were used to amplify the kanamycin resistance gene from pCR II by using touchdown PCR (7). The manufacturer's protocol was used with the following thermocycler conditions. Initial denaturation at 94°C for 5 min, followed by five cycles of denaturation at 94°C for 1 min, 72°C annealing for 1 min, and 72°C extension for 2 min. The annealing temperature was then lowered 5°C every five cycles until 42°C, at which point, eight cycles were run, followed by a final extension of 15 min at 72°C. The resultant PCR fragment was cloned into the transposon delivery vector pUT, generating pUTK210. The cloning vector pLJS was constructed from pBluescript II (KS) (Stratagene, LaJolla, Calif.) by cleavage with BssHII removing the multicloning site (MCS). The resultant plasmid was named "pBSMCS()." Two oligonucleotides (a 47-mer and a 44-mer) (KpnI end, 5'CCAAGCGCGCAACTAGTCTAGACTAAAGCTAGCCTAGGCTGGGATCC3'; SacI end, 5'GTGAGCGCGCGTAATACGAGCTAGCCTAGGGCGAATTGGAGCA C3') were synthesized to regenerate the MCS and add the restriction sites XbaI, NheI, SpeI, and AvrII. The orientation of the added sites can be seen in Fig. 1. The new MCS was amplified from pBluescript II (KS) by using the manufacturer's protocol with the following thermocycler conditions. Initial denaturation was at 94°C for 5 min, followed by 38 cycles of denaturation at 94°C for 30 s, annealing at 42°C for 1 min, extension at 72°C for 30 s, and final extension at 72°C for 15 min. The amplified fragment was cleaved with BssHII, ligated into pBSMCS(), and transformed into DH5. A portion of pLJS was sequenced, confirming the base substitutions and integrity of the MCS, with an Applied Biosystems model 373A sequencer (Foster City, Calif.). Plasmid pLJST2 was generated by directional cloning of the 0.8-kb HindIII-HincII fragment containing the 5S ribosomal rrnB T₁T₂ transcription terminator from pKK223-3 (Pharmacia, Piscataway, N.J.) into pLJS cleaved with HindIII and SmaI. The NotI-AvrII terminator fragment from pLJST2 was subsequently cloned into the NotI-XbaI site of pUTK210, yielding pUTK211 containing mini-Tn5KmT2. This allowed for the subsequent destruction of the XbaI site by heterologous ligation and the regeneration of the NotI and XbaI unique sites downstream of the terminator. Mini-Tn5Kmtod-lux (pUTK214) was generated by directional cloning of the 10.2-kb NotI-XbaI tod-lux fragment from pUC18Not-todlux (Table 1) into the NotI-XbaI site of pUTK211. Both the insert and vector DNA were purified by agarose gel electrophoresis and electroeluted prior to cloning. Electrocompetent E. coli S17-1(pir) cells were prepared and ligations were electroporated as outlined by the manufacturer (BTX, San Diego, Calif.). All other plasmids and relevant constructs are described in Table 1.

View larger version (20K): [in this window] [in a new window]   FIG. 1.   (A) Construction of mini-Tn5Kmtod-lux. A/X and Nh/X represent the AvrII-XbaI and NheI-XbaI heterologous cloning sites, respectively. N, NotI; Sa, SalI; X, XbaI. (B) Cloning plasmid pLJS with unique restriction sites. A, AvrII; Ac, AccI; Ap, ApaI; B, BamHI; Bs, BstXI; C, ClaI; D, DraII; E, EcoRI; Ea, EagI; EV, EcoRV; H, HindIII; Hc, HincII; K, KpnI; Nh, NheI; P, PstI; S, SpeI; Sc, SacI; ScII, SacII; Sm, SmaI; Xh, XhoI.

Strain construction. Plasmid pUTK214 was conjugated into P. putida F1 from E. coli S17-1(pir) as previously described (6). Strains carrying transposon insertions were selected on Pseudomonas isolation agar (Difco, Detroit, Mich.) supplemented with 50 µg of kanamycin/ml. Colonies which produced light upon exposure to toluene were grown in mineral salts media (MSM) (23) with toluene vapor to ascertain that the transposon had not inserted into a gene required for cell growth and also to evaluate their performance as bioreporters in liquid, growing-cell assays (8). A strain designated TVA8 was selected for further study and subjected to DNA-DNA hybridization to verify transposition, as opposed to recombination, by using a ³²P-labeled probe specific for the Tn5 transposase (tnp) contained on pUT. Equal target amounts of luxA, todC, and tnp DNA were loaded onto a Biotrans nylon membrane (ICN, Irvine, Calif.) by using a Bioslot blot apparatus (Bio-Rad, Hercules, Calif.) according to the manufacturer's protocol. The blot consisted of chromosomal DNA from F1, TVA8, and the aforementioned controls. The DNA was loaded in triplicate, the blot was subdivided, and each separate blot was hybridized with either luxA, todC, or tnp PCR-generated P-labeled DNA probes. Blots, hybridized and washed as previously described (1), verified that TVA8 contained luxA and todC but not tnp (data not shown). The negative transposase result confirmed that transposition had occurred.

Stability assays. Batch stability assays were performed by transfer of 1 ml of a 100-ml overnight culture grown in Luria-Bertani (LB) broth with 50 µg of kanamycin/ml (LBKm₅₀) to a 250-ml Erlenmeyer flask with toluene used as the sole carbon source (supplied as vapor). One milliliter of culture was transferred every day for 5 days to flasks containing 100 ml of MSM supplied with toluene vapor (without Km₅₀). Assays were performed in triplicate. Before each transfer, cells were plated on selective (LBKm₅₀) and nonselective (LB) media to ascertain loss of kanamycin resistance resulting from deletion or excision of the transposon. Colonies were subjected to colony hybridization with a 295-bp luxA DNA probe (12). Stability was also assayed in continuous culture with a New Brunswick Bio Flow fermentor (Edison, N.J.) with a 370-ml vessel operated at 28°C at 180 rpm. The feed consisted of MSM supplemented with toluene at approximately 100 mg/liter at a flow rate of 1.0 ml/min. Toluene was fed by simultaneously adding toluene-saturated MSM at a flow rate of 0.2 ml/min and MSM at a flow rate of 0.8 ml/min by using FMI metering pumps (Oyster Bay, N.Y.). The fermentor was operated for 14 days (100 generations) with daily bioluminescence and optical density (OD) measurements. Plate counts (at 7 and 14 days) from selective (LBKm₅₀) and nonselective (LB) media were compared to determine if the kanamycin marker was being lost, and luxA colony blot hybridization was performed to confirm that all colonies contained the lux transposon insert. In batch and chemostat stability studies, TVA8 did not demonstrate instability when subjected to the same evaluation. From batch assays, the selective/nonselective plate count ratio was 1.12 ± 0.13 after five daily transfers, and all colonies hybridized with the luxA probe. After a 14-day continuous cultivation, the selective/nonselective plate count ratio was 1.05 ± 0.13, and all colonies from selective and nonselective plates were lux positive.

Comparison of growth of TVA8 and F1 on toluene. To examine the effect of bioluminescence on the fitness of TVA8, growth curves of TVA8 and F1 were obtained by growing cells in 100 ml of MSM in 250-ml Erlenmeyer flasks with toluene vapor supplied as the sole carbon source. Flasks were inoculated from a fresh overnight culture, grown to an OD at 546 nm (OD₅₄₆) of 1.0 in 100 ml of LB, washed twice in 100 ml of MSM, and resuspended in 100 ml of MSM. A 1-ml aliquot of this suspension was added to the toluene flasks. The cultures were shaken at 200 rpm at 28°C and sampled approximately every hour. The OD₅₄₆ was measured for each culture, and rates of increase in OD were determined from the linear portion of the curves. Growth curves for TVA8 and the parent strain F1 on toluene vapor are shown in Fig. 2. The curves show similar shapes with different lag times for TVA8 and F1 that can be attributed to slightly different inoculum concentrations. Rates were computed from the slopes of the linear portion of the growth curve for both strains. The average rates of increase in OD for F1 and TVA8 were (2.1 ± 0.3) and (2.2 ± 0.3) min¹ × 10³, respectively and were not statistically different ( = 0.05). These results demonstrate that the bioluminescence reactions do not appear to affect cell growth.

View larger version (16K): [in this window] [in a new window]   FIG. 2.   Bioluminescence and growth of TVA8 and growth of F1 on toluene vapor under batch conditions. , , and  represent individual replicates of bioluminescence readings over time. The solid squares (TVA8) and circles (F1) represent the average OD546 of three replicates.

View larger version (11K): [in this window] [in a new window]   FIG. 3.   Specific bioluminescence of TVA8 grown on toluene vapor. The regression line equation is y = 116x  418 (r2 = 0.982). (The first four time points were not included in the linear regression because the organisms were in lag phase.)

Bioluminescence response of TVA8 to toluene, BTEX compounds, and JP-4 jet fuel. Bioluminescence assays were conducted as described by Heitzer et al. (8). Aqueous solutions of toluene, benzene, ethylbenzene, phenol, isomers of xylene, and JP-4 jet fuel constituents were prepared by adding pure hydrocarbon or JP-4 to MSM in a 1:10 (vol/vol) ratio. The solutions were placed on a rotary shaker for 24 h. After phase separation, aqueous-phase aliquots were added to test vials. Final concentrations of dissolved toluene in growing-cell assays ranged from 0 to 50 mg/liter (based on water solubility). The final concentration of the other hydrocarbons was 50 mg/liter (based on their water solubility), and the percentage of JP-4-saturated MSM in the test samples was 2%. Vials containing test solutions and cells were shaken at 150 rpm on a rotary shaker, and bioluminescence was measured every 30 min. Sample vials were placed in a light-tight box, and light output was measured with a liquid light pipe and an Oriel photomultiplier and digital display (models 77340 and 7070; Stratford, Conn.) (8). The light detection methods for continuous culture and growth curves were similar, except that the light-tight box was modified to hold a cuvette, allowing for light measurements and OD readings to be taken consecutively.

View larger version (13K): [in this window] [in a new window]   FIG. 4.   Bioluminescence response of TVA8 to increasing concentrations of toluene after a 2-h exposure. Values are averages of three replicates and have been normalized to the cell density (OD546).

Conclusions. The majority of bioluminescent reporter systems currently being used are the result of cloning of a promoter in front of either a promoterless luxAB or luxCDABE gene cassette and transfer of the plasmid construct into the strain that contained the particular promoter. Plasmid-based systems have obvious drawbacks, such as the need for constant selective pressure to ensure plasmid maintenance (17). Another important consideration is plasmid copy number. In a positively regulated system, copy number can negatively effect gene expression. Multiple copies of the promoter binding region for the regulatory protein on the plasmid compete with the binding site on the chromosome, causing less expression of the operon being studied (27). This negative effect is important when lux fusions are used for on-line monitoring of bacterial processes.