Expression of functional genes in soil can be monitored by detection of their transcription either directly as mRNA or via a reporter gene fused to the target gene. For interpretation it is important to know the mechanism of regulation for expression of the activity of interest, as levels of mRNA may not always be indicative of enzyme activity due to posttranscriptional modification.
To quantify, or detect in situ, a specific cell population or gene activity in soil, a marker gene (expressed from a constitutive promoter resulting in a detectable phenotype) or a reporter gene (expressed from an inducible promoter) can be used (see Chaps. 16-18)
Luminescent marker genes luc and luxAB, derived from eukaryotic and bacterial luciferase, respectively, have been used in microbiological studies; due to limitations such as energetic demands on cell metabolism, their usefulness for studies of soil microbial populations is limited. Consequently, fluorescent marker genes, mainly those encoding fluorescent proteins such as the green fluorescent protein (GFP), have been used, despite limitations such as a requirement for molecular oxygen and a need for specialised detection equipment. In contrast to luminescent markers, fluorescent markers place less demands on cell energy. Chromogenic reporter genes have also been used in many environmental studies, including: xylE encoding catechol 2,3-dioxygenase, lacZ for ^3-galactosidase, gusA encoding E. coli glucuronidase, inaZ encoding ice nucleation activity, and different antibiotic resistance genes such as nptII which produces kanamycin resistance. Most bioreporters use an environmentally or metabolically responsive promoter, which is fused to a suitable reporter gene and introduced into a microbial host. Host cells then react to environmental stimuli by production of an easily detectable reporter protein (Leveau and Lindow 2002; Hinde at al. 2003; Jansson 2003).
Another fluorescent method for the in situ detection of a gene of interest is based on hybridisation of a target gene with a fluorescently labelled probe (Bakermans and Madsen 2002; Ginige et al. 2004). Fluorescent in situ hybridisation (FISH), where 16S rRNA sequences are targeted with specific fluorescently labelled probes to identify active bacteria (having a higher ribosomal count) within a community population, has been widely used in different environments. Studies in the soil environment have been limited due to autofluorescence of soil particles and the small size of indigenous bacteria which show only dim fluorescence (Hahn et al. 1992; Christensen et al. 1999). Metcalfe (2002) used this technique to specifically detect cells of chitinolytic Stenotrophomonas maltophilia in a cell extract from chitin bags buried in soil for a period of a few weeks. Comparing cell extracts from July and November sampling, the July extract showed brighter cells indicative of higher activity. This finding corresponded with higher chitinolytic activity detected in the July sampling.
FISH was recently coupled with autoradiography (MAR-FISH/STARFISH/MICRO-FISH) enabling simultaneous microscopic determination of in situ identities, activities, and specific substrate uptake of bacterial cells within complex bacterial communities (Lee et al. 1999; Ouverney and Fuhrman 1999; Cottrell and Kirchman 2000; Gray et al. 2000; Ito et al. 2002). These methods are based on feeding of a natural community with a radioactive substrate. After incubation cells are fixed on a slide and subjected to FISH treatment and microautoradiography. Using fluorescently labelled oligonucleotide probes with different levels of specificity, identity of individual cells can be determined. As the same slide is also used for the autoradiography, affiliation data from FISH can be connected with activity data obtained by autoradiography.
The original FISH method relies on the presence of many target sequences (16S rRNA) within an individual cell. This was later modified to detect a single copy of a functional gene in a cell. Tani et al. (1996) used this method to identify cells of E. coli. Cells were spotted on a slide, the cell wall permeabilised by lysozyme and proteinase K, and PCR performed directly on the slide using a digoxigenin-labelled primer. The digoxigenin-labelled product was then detected with a labelled anti-digoxigenin antibody and visualised using a suitable technique such as epifluorescent microscopy or confocal laser-scanning microscopy (Tani et al. 1996; Hoshino et al. 2001). The same principle was used for the detection of mRNA in a bacterial cell by Hodson et al. (1996). After permeabilisation of the cell wall, RT-PCR was carried out; cells were washed and used in a second PCR with the labelled primer.
A variety of methods have been used to detect expression of functional genes in the pool of community RNA extracted from a particular environment (ex situ). The first step is the isolation of mRNA from soil. Traditionally, it was believed that mRNA was short-lived and nearly impossible to isolate. This was based on studies of the turnover of prokaryotic mRNA, where, for example, the half-life of specific E. coli mRNAsrangedfromabout 0.5 to 20 min, which was constant with doubling times between 60 min and 10 h (Nierlich and Murakawa 1996). However, it has been shown that slower growing environmental strains have longer mRNA half-lives (Fleming and Sayler 1995), and the situation is rather different with eukaryotic mRNA. In prokaryotes, mRNA is translated nearly simultaneously as it is transcribed and often many genes are transcribed together in a single polycistronic mRNA, whereas transcription and translation in eukaryotes is physically separated by the nucleus membrane resulting in temporal separation. Eu-karyotic mRNA also requires splicing of intervening sequences before it can be translated, which means that it has a longer half-life measured in hours rather than minutes (Sarkar 1997; Rauhut and Klug 1999). Further problems hampering RNA isolation are associated with the isolation process itself. After cell lysis RNA can be quickly degraded by the action of ubiquitous RNases, or bound to soil particles and colloids via functional groups ofthe nitrogenous bases of single-stranded RNA (Ogram et al. 1994, 1995; Men-dum et al. 1998). Despite these technical challenges several methods for the isolation of RNA from natural environments have been published (Selenska and Klingmüller 1992; Ogram et al. 1995; Griffiths et al. 2000; Hurt et al. 2001; Sessitsch et al. 2002; Bürgmann et al. 2003). Once isolated, various methods for RNA analysis are available. One of the traditional methods which has proved to be fast and reliable is a slot or dot blot hybridisation, where RNA is applied to a membrane and subsequently probed with a 32P-labelled in vitro transcribed RNA probe. A comprehensive review of such methods can be found in Sayler et al. (1992).
To detect differences in gene expression in a bacterial strain under different conditions or between different strains, subtractive hybridisation can be used. cDNA prepared from total RNA of a wild type or a strain cultivated under certain conditions is hybridised with RNA from a mutant or strain cultivated under different conditions. Double-stranded cDNA-mRNA hybrids are then selectively removed and the composition of single-stranded cDNA left in the supernatant is analysed (Utt et al. 1995).
Hybridisation is also used in ribonuclease protection assays, where an extracted soil RNA is hybridised to a labelled, in vitro transcribed anti-sense RNA probe complementary to the gene of interest. The non-hybridised single-stranded probe is then removed by ribonuclease action and the protected double-stranded RNA hybrid is quantified by comparison with standards (Fleming et al. 1993).
A differential display method widely used for eukaryotic gene discovery takes advantage of the polyadenylate tail present on most eukaryotic mRNAs. Despite the fact that poly(A) polymerase was isolated from E. coli as early as 1962, and again during the 1970s from other bacteria, most recently in Streptomycetes (Sarkar 1997; Rauhut and Klug 1999; Jones 2002), it was long believed that polyadenylation was a characteristic of eukaryotic mRNA and quite rare in bacteria. Now it is believed that 2-50% of bacterial mRNA have poly(A) tracts of 14-60 nucleotides, in comparison with 100% polyadenylation (80-200 nucleotides long) in eukaryotes. The exact role of polyadenylation is still not fully established but there are indications that the poly(A) tail plays a role in the stability of mRNA.
The 3' primer used in the first step of the differential display method, reverse transcription, and also for the following PCR consists of a poly(T) region with two additional bases that recognise one-twelfth of the total mRNA population. The 5' primer is arbitrary and short (6 to 7 bp) and should ideally anneal within 500 bp of the 3' primer; this means that, in contrast to other methods, there is no need for a priori information on gene sequences for design of specific probes or primers. Resulting PCR products are resolved by a DNA sequencing gel creating a 'fingerprint' of the studied population. A low annealing temperature (42 °C) is used and specificity of DNA amplification can be increased by decreasing the concentration of deoxynucleoside triphosphates (Liang and Pardee 1992). The method can be used to visualise mRNA compositions of cells, cDNAs can be sequenced, or individual bands can be cloned and used as probes for Northern or Southern blotting. Reproducibility is very high (95% of bands were always seen). To study bacterial R/DNA lacking the poly(A)
tail, both primers can be arbitrary. Wong and McClelland (1994) used this method searching for stress-induced RNAs in Salmonella typhimurium, and Fleming et al. (1998) detected differential mRNA transcription both from pure-culture and soil-derived bacterial DNA. After optimisation of the primer concentration, length, annealing temperature, concentration of template, deoxynucleotide triphosphate, and MgCl2 concentration, the detection limit of todC1 was 0.015 ng of total RNA template or approximately 103 transcripts. Another possible variation of this method is the use of "motif" primers using a sequence common for a group of genes of interest where the PCR product is then biased towards this group (Stone and Wharton 1994). An important limitation of the method is that less abundant RNAs will probably be under-represented among the visible products on the gel. A possible solution to this problem might be a nested PCR with identical primers with one or more extra arbitrary bases at the 3' end. The most vexing problem of this method is the occurrence of false positives. Less than half of the putative differentially expressed PCR products excised from gels were truly differentially expressed.
Competitive PCR can be used to determine the relative transcript level of a gene of interest. After the isolation of RNA, RT-PCR using a gene-specific primer is run. For the second PCR a series of dilutions of competitive templates (genomic DNA) is added and resulting products are separated on an agarose gel according to their size, as PCR products of the competitive template are longer due to the presence of introns (Lamar et al. 1995; Bogan et al. 1996a). Recently, Han and Semrau (2004) coupled RT-PCR with capillary electrophoresis, which resulted in higher sensitivity of the detection of RT-PCR products. The method is also very fast, requires only small amounts of sample, and can be completely automated from sample injection to data analysis.
Real-time PCR uses a double-labelled probe to measure the accumulation of fluorescence of a released reporter dye during PCR, which is proportional to the amount of product formed during PCR and can be correlated to the original amount of template DNA or even mRNA. This method is highly suitable for quantification of transcripts of genes of interest (Rodrigues et al. 2002).
Metabolic processes of soil microorganisms are of key significance due to their ecological and economic importance; one important metabolic process is antibiotic production/resistance. A large proportion of antibiotics currently in use are of actinomycete or fungal origin, therefore detection of genes involved in their production and understanding expression is of a great importance. In addition to detection of antibiotic synthesis genes, detection of antibiotic resistance genes is becoming more important in light of the increasing resistance seen in environmental and clinical bacteria.
Another important group of microbial genes are those involved in the N cycle. Nitrogen is the main element limiting plant productivity in nearly all natural systems. The efficient use of N available to plants is essential regardless of whether it is N fixed by symbiotic or free living N-fixing bacteria, or N in organic residues entering soil which has to be released by microbial activity (mineralisation), or N delivered in industrial fertilisers, where the efficiency of its use is important due to the cost of fertiliser production. In addition to the economic importance of efficient utilisation of N fertilisers, ecological impact must be taken into account as high levels of N released in the form of NO- can contaminate ground water and plant crops.
Degradation processes are also of interest, as all organic matter produced must be decomposed to release bound elements back to the element cycles. The soil microbial population conducts a major part of these processes. Diversity of degradative abilities of soil microbes even ensures degra-dation/bioremediation of anthropogenic substances entering the natural environment.
Expression of Antibiotic Activity, Antagonism
A luxAB reporter gene was used to monitor the expression of the antibiotic phenazine-1-carboxylic acid (PCA) by Seveno et al. (2001). The luxAB from Vibrio harveyi was inserted in the phzB gene of the phenazine operon and transcription was monitored by measurement of luminescence in liquid culture, on nutrient agar, on sterile wheat seedlings and in sterile bean rhizosphere. Production of phenazine was confirmed both in liquid culture and on solid media, but it could not be detected on wheat seedlings or in bean plant root rhizosphere despite the fact that transcription of the phzB::luxAB reporter occurred on the bean plant root.
Direct detection of mRNA of an antibiotic production gene using RT-PCR was achieved by Anukool et al. (2004). Streptothricin (ST) production by Streptomyces rochei F20 was monitored in liquid culture, soil and the rhizosphere of spring wheat. After RNA extraction, primers specific for the ST resistance gene (sttR) andSTbiosynthesisgene(sttA) coding for peptide synthetase of S. rochei were used for RT-PCR. A higher level of expression of sstR in comparison with sttA was detected in all experiments. mRNAs of both genes were detected in soil containing approximately 106 cfu g-1 soil. The sttR mRNA was detected both in sterile and non-sterile rhizospheres, but neither sttR nor sttA transcripts were detected in the rhizoplane.
The role of antibiotics in biological control has been reported (Rothrock and Gottleib 1984; Brisbane and Rovira 1988; Thomashowet al. 1990; Howie and Suslow 1991). Detection of antibiotics in soil is difficult due to low level production and adsorption to clays. Thiostrepton production in sterile soil was determined by Wellington etal. (1993) within the range of30-50 ng g-1 soil following extraction and bioassay via a specific thiostrepton-inducible promoter coupled to a resistance gene. A highly sensitive inducible promoter driving gfp expression was used to detect oxytetracycline production in soil where FACS analysis enabled direct counting of fluorescing cells in soil extracts (Hansen et al. 2001). The direct detection of phenazine in the wheat rhizosphere after inoculation and incubation with P.fluorescens 2-79 and P. aureofaciens 30-84 proved the significance of antibiotic production in control of Take-all disease caused by Gaeumannomyces graminis var. tritici (Brisbane and Rovira 1988).
Detection of Expression of N Cycle Genes in the Environment
Gene expression can be correlated to bulk N fixation. Bürgmann et al. (2003) extracted nif H mRNA from Azotobacter vinelandii growing in liquid culture and sterile soil, and using RT-PCR estimated expression. Bhagwat and Keister (1992) used the subtractive RNA hybridisation procedure to characterise genes responsible for increased competitiveness for nodulation of the strain Bradyrhizobium japonicum USDA 438 in comparison with the strain USDA 110. Both genes were isolated and characterised.
Expression of five denitrification genes in two estuarine sediments was studied by Nogales et al. (2002). The presence of all five genes was confirmed by PCR in both sediments although RT-PCR detected transcript of only two genes, the nirS and nosZ genes, the latter only by Southern hybridisation. Consistently stronger RT-PCR products from one sampling site corresponded with a 10-fold higher plate count of denitrifiers from this site indicating the semi-quantitative nature of RT-PCR. The failure to detect all five denitrifying genes on the RNA level could be due to slightly different specificity of primers, inhibitory effects of co-purified impurities (mainly humic substances) in RNA samples, preferential amplification of certain templates, or insufficient quantity of template needed for detection. Cloning and sequencing of nirS RT-PCR products revealed a high diversity of the gene in both environments, where 13 of 16 clusters appeared novel suggesting the presence of unknown denitrifying bacteria. Most sequences obtained were specific to single sampling sites emphasising the high diversity of denitrifiers.
The community structure of ammonia-oxidising bacteria within anoxic marine sediments was studied by Freitag and Prosser (2003). The communities were studied by denaturing gradient gel electrophoresis (DGGE) analysis and sequencing of 16S rRNA genes using group-specific primers by PCR and RT-PCR. The RT-PCR was used in parallel with PCR to achieve greater sensitivity due to the increased number of targets provided by the ribosomes and also indicated which members of the community were most metabolically active. RNA-derived DGGE banding patterns were similar to those derived from DNA, but not all dominant bands present in rDNA were detected, suggesting differences in growth and activity of ammonia-oxidising bacteria. This may reflect the heterogeneity in physicochemical characteristics within the sediments.
The activity of microbial decomposers can be elucidated using various techniques. Transcription of the cellulase/hemicellulase genes of Clostridium cellulovorans were studied using Northern blot, RT-PCR, primer extension and nuclease protective assay (Han et al. 2003). Northern hybridisation illustrated that the cellulosomal cbpA cluster was transcribed as a poly-cistronic mRNA of 8 and 12 kb. Primer extension and nuclease protective assay located transcriptional start sites of four genes in the cluster. Expression of ligninolytic enzymes (lip) of Phanerochaete chrysosporium in wood was studied by Janse et al. (1998). Poly(A) RNA was isolated by magnetic capture and the specific transcriptions of all ten lip genes plus three manganese peroxidase genes (mnp) were quantified by competitive RT-PCR with full-length genomic subclones. Differences in the transcript level ranged up to 10,000-fold and the transcript pattern of different lignin peroxidase genes expressed in wood differed dramatically from those previously obtained with defined media or from soil cultures, indicating specialisation of different enzymes to certain conditions. In addition, Stewart and Cullen (1999) studied transcription of lip genes of Phanerochaete chrysosporium. Using competitive RT-PCR they assessed lip transcript levels in both C-and N-limited media, which showed differential regulation of lip genes and the presence of constitutive and up-regulated transcripts. Iron-responsive genes of P. chrysosporium were studied by Assmann et al. (2003). Using differential display RT-PCR they identified 97 differentially expressed cDNA fragments, the majority of them encoding proteins involved in iron acquisition.
Microbial remediation of pollutants has been measured by quantifying messages from inducible genes involved in degradation. Selvaratnam et al. (1993) used RT-PCR to monitor the 3-chlorobenzoate-degrading catabolic tfdB gene of Pseudomonas putida inoculated into non-sterile activated sludge, and the phenol-degrading dmpN gene of P. putida in a sequencing batch reactor. Expression of dmpN correlated with the degradation of phenol, confirming the suitability of this method for monitoring and controlling operation of a reactor.
Degradation of polycyclic aromatic hydrocarbons (PAHs) in sterile soil microcosms was investigated using competitive RT-PCR for the detection of mRNA of three manganese peroxidase genes (mnp) of Phanerochaete chrysosporium (Bogan et al. 1996a). Levels of mnp mRNA corresponded with the disappearance of high ionisation potential PAHs during soil microcosm experiments. This confirmed that monitoring levels of mRNA could be used as an indicator of physiological state of the fungus under real conditions during remediation processes.
Mechanisms of naphthalene degradation have been studied by quantification of m-RNA transcripts of nahA. A naphthalene-lux reporter system was also used to monitor degradation of naphthalene and determine bioavailability within soils heavily contaminated with PAHs. The naphthalene mineralisation rate correlated positively with the number of nahAgene transcripts present and the naphthalene-lux reporter system confirmed bioavailability of naphthalene illustrating the usefulness of this monitoring system (Sanseverino et al. 1993-94). Bakermans and Madsen (2002) used FISH with tyramide signal amplification (TSA), routinely used to detect mRNA in eukaryotic systems, for the detection of intracellular mRNA for the naphthalene dioxygenase gene (nahAc) in naphthalene-degrading P putida NCIB 9816-4. TSA significantly increased probe fluorescence and only 100-1000 copies of the hybridisation target molecules per cell were needed for detection. The efficiency of the method was tested on naphthalene-contaminated groundwater samples. Positive results represent progress towards the main goal of environmental microbiology, to simultaneously detect identity, activity and biogeochemical impact of microorganisms in situ, including in soil, water, and sediments.
To detect mercury in the environment, Hansen and S0rensen (2000) constructed three cell biosensors fusing the mercury-inducible promoter, Pmer, and its regulatory gene, merR,withreportergenes luxCDABE, lacZYA, or gfp. Transferred into P. putida, this reporter system was able to quantitatively detect mercury in contaminated soil. The construct was cloned into a mini-Tn5 delivery vector which was able to transfer into a variety of Gram-negative bacteria. Jeffrey et al. (1996) extracted mRNA from the concentrated bacterial fraction from a freshwater pond and detected the transcript of the merA gene by probing with a radioactive antisense probe. The quantity of the transcript did not correspond with the concentration of dissolved mercury but rather with activity of the microbial community.
Was this article helpful?