We investigated the applicability of the green fluorescent protein of Aequorea victoria as a reporter for gene expression in the strictly fermentative Gram-negative ethanologenic bacterium Zymomonas mobilis and in the moderately halophilic bacterium Halomonas elongata. We have succeeded to express a mutated gene of green fluorescent protein under the control of different promoters in Z. mobilis and H. elongata grown under various glucose or salt concentrations, respectively. Our results demonstrate that gfp can serve as an easily assayable reporter gene in both organisms. Maximum fluorescence was obtained in Z. mobilis grown aerobically and in H. elongata grown under elevated salt concentration in solid medium. For both bacteria the fluorescence obtained was higher when the gfp gene was placed under the control of a native promoter.
Zymomonas mobilis is a Gram-negative anaerobic bacterium with industrial importance that produces ethanol from hexoses at high rates and yields . It has an unusual tolerance for high concentrations of ethanol and glucose [1,2]. Plasmid vectors have been developed for Z. mobilis, mutants can be isolated by various mutagenic treatments, and some Z. mobilis genes have been cloned in heterologous hosts . However, several other routine methods and tools for molecular genetic investigations, including the use of reporter genes for analysis of gene transcription and promoter function, have not been studied widely in Z. mobilis, with the exception of reports on the use of lacZ [4,5,6] and inaZ .
Halomonas elongata, a moderately halophilic bacterium, has recently attracted scientific and technological attention because it exhibits a wide salt tolerance, it is an excellent tool to study the molecular biology of osmoregulation processes and it is a good source for halophilic enzymes and protecting agents (compatible solutes) for both enzymes and whole cells . However, knowledge concerning the genetics and the use of reporter systems in H. elongata is very limited: only recently an inaZ reporter system was developed in H. elongata and other moderately halophilic bacteria [9,10].
The green fluorescent protein (GFP), a 27-kDa protein from the marine bioluminescent jellyfish Aequorea victoria , is a unique marker that can be identified by non-invasive methods. Neither substrates, other enzymes nor co-factors are required for detection of GFP. In the present study we have investigated the use of gfp as a reporter gene in Z. mobilis and H. elongata. We have used a mutated version of this gene , which allows the establishment of GFP as a convenient expression marker for bacteria, because it can fold correctly and remain soluble in the cells, resulting in 100-times increased fluorescence as compared to the wild-type GFP when expressed in Escherichia coli. We here show that gfp can be expressed and serve as an easily assayable reporter gene in both bacteria.
Materials and methods
Strains and growth conditions
Z. mobilis strain CP4RifR, a spontaneous mutation of strain CP4 , was grown semianaerobically at 30°C in complete medium as described before . H. elongata strain ATCC 33174 was cultured as described by Vargas et al. . E. coli DH5α (BRL, Gaithersburg, MD, USA), which was used as a host for subcloning and maintaining recombinant plasmids, was grown in LB medium as routinely described . Solid media were obtained by adding 2% (w/v) agar. Antibiotics, when needed for genetic selections or plasmid maintenance, were added at the following concentrations: tetracycline 20 μg ml−1 for E. coli; chloramphenicol (Cm) 20 μg ml−1 for E. coli, 100 μg ml−1 for Z. mobilis; kanamycin 50 μg ml−1; rifampicin (Rif) 20 μg ml−1 for Z. mobilis, 25 μg ml−1 for H. elongata; streptomycin (Sm) 40 μg ml−1 for H. elongata.
Plasmid preparations from E. coli, restriction enzyme digestions, ligations, E. coli transformations, DNA electrophoresis and Southern blot analysis were performed according to standard protocols . Plasmid DNA from Z. mobilis was isolated as described before . Plasmid DNA from H. elongata was isolated according to the protocol of Wizard SV Plus mini prep isolation kit (Promega). DNA bands from agarose gels were purified according to protocol of GENECLEAN II BIO 101 (La Jolla, CA, USA). DNA labeling and hybridization were performed with the DIG non-radioactive labeling method (Boehringer Mannheim cat. No. 1093657).
Conjugal transfer of recombinant plasmids in Z. mobilis CP4RifR or H. elongata ATCC 33174 was performed by triparental filter-matings as previously described [17,14] by using pRK2013 as the helper plasmid . Z. mobilis transconjugants were selected in Cm and H. elongata in Sm. In both organisms, Rif was used for counter-selection against donor cells. All transconjugants were tested for their plasmid content with plasmid isolation, back transformation in E. coli when needed, restriction analysis and Southern blot hybridization by standard methodology .
The fluorescence of GFP-expressing E. coli cells was assessed by eye using an UV lamp (Ultra Violet Products Inc. 366 nm). The amount of fluorescence emitted by Z. mobilis and H. elongata cells in liquid culture was assayed as described before  using a fluorimeter (Perkin-Elmer LS-3) set to excite the cells at 488 nm and detect emission at 511 nm. Cells from 30-ml liquid culture were harvested by centrifugation (6000×g, 10 min), washed with buffer (10 mM Tris–HCl, pH 8.0, 600 mM NaCl), resuspended in the same buffer and the fluorescence measured immediately. The amount of fluorescence in cells grown on agar plates was assayed as described by Lissemore et al. , with minor modifications. A number of 50–70 colonies were scraped from the plates and resuspended in the same buffer as above, after storage at 4°C for 28 h. Protein concentration was determined by the Lowry method.
Reverse transcription-polymerase chain reaction (RT-PCR) analysis
Total RNA was isolated from 5 ml of Z. mobilis CP4 and H. elongata ATCC 33104 cells grown up to late exponential phase (OD600 0.6 and 1.0, respectively), using the High Pure RNA Isolation Kit (Boehringer cat. No. 1828665) according to the manufacturer's instructions. RNA was quantitated by photometric measurement . An amount of 1 μg of total RNA from each sample was used in an RT-PCR reaction employing the RobusT RT-PCR Kit (Finnzymes) according to the manufacturer's instructions. RT-PCR primers were selected (MWG-Biotech AG) on the basis of the gfp DNA sequence contained in the pGreenTIR cloning vector , yielding a product of 600-bp length. The 5′→3′ sequences of the GFP primers used were the following: (A) forward GTGGAGAGGGTGAAGGTGATGC, (B) reverse CATCCATGCCATGTGTAATCCC.
Determination of plasmid copy number
Plasmid DNA was extracted from Z. mobilis transconjugants grown with or without agitation, as well as from H. elongata transconjugants grown in 2, 5 or 10% NaCl. Total plasmid DNA in different quantities was electrophoresed on 0.8% agarose gel at 5 V cm−1. The DNA was transferred to nylon paper by Southern blotting and hybridized with a gfp gene probe to distinguish plasmids pAEG1, pDS3154G, pHSG1 and pHSG2 from the native plasmids of the hosts. The intensities of the hybridizing plasmid bands were compared to each other, after correcting with the number of cells, using the Kodak Digital Science™ 1D Image Analysis Software and plasmid copy number was estimated as previously described .
Results and discussion
Expression of gfp in Z. mobilis
For the investigation of the gfp expression in Z. mobilis, two plasmids were constructed, in which the gfp gene was placed under the control of a native and a heterologous promoter, respectively.
In the first case, the recombinant plasmid vector pAEG1 was constructed by a two-step subcloning, as shown in Fig. 1. Initially, the intermediate plasmid pAE1 was created by inserting the 1.8-kb BamHI–SalI fragment of plasmid pBZIP1 in pBR325 . This fragment contains the replication region of the Z. mobilis ATCC 10988 plasmid pZMO3 , which allows its replication and stable maintenance in different Z. mobilis strains . It also carries the strong promoter of the Z. mobilis CP4 pyruvate decarboxylase gene, Ppdc. Plasmid pAE1 is a suitable vector for transfer of heterologous genes in Z. mobilis because it additionally contains the oriT of pBR325, which allows its conjugal mobilization from E. coli donors, as well as the chloramphenicol resistance gene, a good selection marker for Z. mobilis. Subsequently, the gfp gene from pGreenTIR  was subcloned as a 0.8-kb BamHI fragment in pAE1 (Fig. 1) under the control of Ppdc. The gfp gene from pGreenTIR is a mutated gene appropriate for prokaryotic expression, which encodes a GFP with two substitutions (F64L and S65T). The first mutation confers increased GFP solubility , whereas the second one increases the fluorescence of this GFP variant by a factor of 35 compared to the wild-type . In the second case, the gfp gene was placed under the control of the heterologous promoter Pbla, subcloned as a 0.8-kb PstI fragment in pDS3154, a shuttle vector for Z. mobilis , producing plasmid pDS3154G. Transformation of E. coli DH5α with plasmids pAEG1 and pDS3154G gave fluorescent transformed colonies visible by an UV lamp.
Plasmids pAE1, pAEG1, pDS3154 and pDS3154G were transferred in Z. mobilis CP4 by helped conjugation. In all cases transconjugants were isolated and their amount of emitted fluorescence was assayed as described in Materials and methods. RT-PCR analysis demonstrated gfp expression in both pAEG1 and pDS3154G transconjugants (Fig. 3). Z. mobilis CP4/pAEG1 isolates exhibited significant amount of fluorescence as compared with the negative control CP4/pAE1, whereas the strain CP4/pDS3154G emitted a lower amount of fluorescence in comparison to its corresponding negative control CP4/pDS3154 (Table 1). Fluorescence reached a maximum value at late exponentially phase and the concentration of glucose in the medium varying between 2% and 10% had no effect (data not shown). Both gfp carrying plasmids were stable in the Z. mobilis hosts as could be maintained structurally intact for at least 100 cell cycles under non-selective conditions. Additionally, a correlation between biomass and expression of GFP was analyzed during cell growth under anaerobic and aerobic conditions. It is well known that the full GFP activity is achieved after an oxidation of the protein . The aeration does not affect the growth of Z. mobilis cells . However, the results showed that the fluorescence increased seven-fold (Table 1) in the case of aerobic cell growth, thus confirming earlier reports . Increase of fluorescence under aerobic growth is not due to higher gene dosage because it was found that the copy number of pAEG1 and pDS3154G was not affected by the different growth conditions (data not shown). Furthermore, it is shown that gfp is more efficiently expressed under the control of Ppdc. The gfp reporter system is more convenient than other reporter systems in Z. mobilis because it does not require specific substrates (lacZ) or low growth temperatures (inaZ). The intensity of fluorescence depends on the culture aeration. Additionally, gfp is not as sensitive as inaZ and hence expression from distal promoters can be minimal .
|Fluorescence mg−1 protein*|
|Fluorescence mg−1 protein*|
The fluorescence was measured as described in Section 2. Values were estimated from four independent measurements (mean±standard deviation) at late exponential phase. *: Fluorescence due to GFP was detected by subtraction of values of fluorescence of CP4/pAE1 and CP4/pDS3154 from those obtained with CP4/pAEG1 and CP4/pDS3154G, respectively. **n.d.: Not detected; <0.05.
Expression of the gfp gene in H. elongata
For the investigation of the gfp expression in H. elongata a shuttle vector for Halomonas strains, pHS15, was used , in which the gfp gene was also placed under a native and an heterologous promoter, respectively.
In the first case, the gfp gene was excised as a 0.8-kb PstI fragment from pGreenTIR and subcloned in the PstI site of pHS15 producing plasmid pHS15G1 (Fig. 2). In the case of the heterologous promoter, gfp gene was subcloned under the control of Ppdc as a 1-kb partial EcoRI fragment excised from pAEG1 and subcloned in the EcoRI site of pHS15. The outcome was the recombinant plasmid pHS15G2 (Fig. 2). Transformation of E. coli DH5α with plasmids pHS15G1 and pHS15G2 gave fluorescent colonies visible under an UV lamp.
The plasmids pHS15G1 and pHS15G2 were transferred to H. elongata ATCC 33174 by assisted bacterial conjugation. The amount of fluorescence emitted by ATCC 33174/pHS15G1 and ATCC 33174/pHS15G2 cells was assayed as described in Section 2. Transconjugants were grown in various concentrations of NaCl (2%, 5% and 10%). The gfp gene was expressed in all transconjugants, as shown by RT-PCR analysis (Fig. 3). Strains ATCC 33174/pHS15G1 and ATCC 33174/pHS15G2 showed significantly higher fluorescence as compared with the strain ATCC 33174/pHS15, which was used as a negative control (Table 2). The expression of gfp in the ATCC 33174/pHS15G1 strain controlled by the native promoter of plasmid pHS15  appeared to be significantly higher as compared with the strain ATCC 33174/pHS15G2. The expression of gfp in ATCC 33174/pHS15G1 was elevated as the concentration of NaCl in the growth medium increased from 2 to 10%, whereas the opposite effect was observed with the ATCC 33174/pHS15G2 (Table 2). Additionally, the latter exhibited a delayed growth in high NaCl concentrations (10%, w/v) unlike strain ATCC 33174/pHS15G1, which grew normally (data not shown).
|Strain||fluorescence mg−1 protein*|
|liquid cultures||solid cultures|
|NaCl (% w/v)|
|Strain||fluorescence mg−1 protein*|
|liquid cultures||solid cultures|
|NaCl (% w/v)|
The fluorescence was measured as described in Section 2. Values were estimated from four independent measurements (mean±standard deviation). *: Fluorescence due to GFP was detected by subtraction of values of fluorescence of ATCC 33174/pHS15 from those obtained with ATCC 33174/pHS15G1 or ATCC 33174/pHS15G2. **: No growth.
Since others have reported a similar poor growth of bacterial strains that express GFP in liquid cultures but not on agar plates , we checked the growth of ATCC 33174/pHS15G1 and ATCC 33174/pHS15G2 strains on solid media with various NaCl concentrations. No growth was observed for the ATCC 33174/pHS15G2 on agar plates with high NaCl concentrations (10%, w/v). On the contrary, under these conditions the ATCC 33174/pHS15G1 strain showed good growth and appeared to have eight-times more fluorescence as compared with the corresponding liquid cultures (Table 2). This does not seem to be due to the toxicity of GFP, as higher GFP amounts were obviously tolerated by ATCC 33174/pHS15G1. This unexpected finding is further under investigation at present. Furthermore, as described for the Z. mobilis case, the gfp plasmids were reasonably stable in the H. elongata transconjugants for at least 100 cell cycles grown under non-selective conditions and retained a similar copy number regardless of the salt concentration (data not shown). These results demonstrate that gfp can also be used as a reporter system in H. elongata in liquid, and much better in solid cultures. The intensity of fluorescence depends on the salinity and solid nature of the culture medium.
This is the first report demonstrating that gfp can be a useful reporter gene in Z. mobilis and H. elongata strains. It was found that gfp is more efficiently expressed under the control of native promoters in each bacterium. Maximum fluorescence was obtained in Z. mobilis at late exponential phase grown aerobically, whereas the glucose concentration in the medium had no effect. In H. elongata, GFP was expressed more efficiently on solid cultures with high NaCl concentration.
The authors wish to thank Dr. William Miller (University of California, Berkeley, CA, USA) for providing the pGreenTIR and Prof. Antonio Ventosa (University of Seville, Seville, Spain) for critical reading of this manuscript. This work was supported financially by the Greek General Secretariat of Research and Technology (Program PENED 1999, Contract No. 99ED67).