Smith and Osborn QPCR review

M IN I R E V I EW Advantages and limitationsofquantitativePCR (Q-PCR)-based approaches inmicrobial ecology Cindy J. Smith & A. Mark Osborn Department of Animal and Plant Sciences, University of Sheffield, Western Bank, Sheffield, UK Correspondence: A. Mark Osborn, Department of Animal and Plant Sciences, University of Sheffield, Western Bank, Sheffield S10 2TN, UK. Tel.: 144 114 222 4626; fax: 144 114 222 0002; e-mail: a.m.osborn@sheffield.ac.uk Received 20 June 2008; revised 10 October 2008; accepted 24 October 2008. First published online December 2008. DOI:10.1111/j.1574-6941.2008.00629.x Editor: Michael Wagner Keywords Q-PCR; RT-Q-PCR; 16S rRNA gene; mRNA. Abstract Quantitative PCR (Q-PCR or real-time PCR) approaches are now widely applied in microbial ecology to quantify the abundance and expression of taxonomic and functional gene markers within the environment. Q-PCR-based analyses combine ‘traditional’ end-point detection PCR with fluorescent detection technologies to record the accumulation of amplicons in ‘real time’ during each cycle of the PCR amplification. By detection of amplicons during the early exponential phase of the PCR, this enables the quantification of gene (or transcript) numbers when these are proportional to the starting template concentration. When Q-PCR is coupled with a preceding reverse transcription reaction, it can be used to quantify gene expression (RT-Q-PCR). This review firstly addresses the theoretical and practical implementation of Q-PCR and RT-Q-PCR protocols in microbial ecology, high- lighting key experimental considerations. Secondly, we review the applications of (RT)-Q-PCR analyses in environmental microbiology and evaluate the contribu- tion and advances gained from such approaches. Finally, we conclude by offering future perspectives on the application of (RT)-Q-PCR in furthering understanding in microbial ecology, in particular, when coupled with other molecular approaches and more traditional investigations of environmental systems. Introduction The application of PCR in combination with the extraction of nucleic acids (DNA and RNA) from environmental matrices has been central to the development of culture- independent approaches in microbial ecology. These meth- ods, which have been applied since the early 1990s (e.g. Giovannoni et al., 1990), enabling the analysis of the total microbial communities present within environmental sys- tems, have revolutionized our understanding of microbial community structure and diversity within the environment. Coupling environmental nucleic acid isolation to subse- quent PCR amplification of both taxonomic (i.e. rRNA) and functional gene markers and in combination with DNA fingerprinting- and sequencing-based analyses has enabled description of the hitherto uncharacterized majority of environmental microorganisms (Head et al., 1998) driving the discovery of new microbial lineages and enabling the description of genetic diversity in a wealth of functional gene markers (Larkin et al., 2005). Although recently devel- oped ultra-high-throughput sequencing technologies such as pyrosequencing (Margulies et al., 2005; Edwards et al., 2006) now dwarf PCR-based sequence studies in terms of sequence coverage, the ability of the PCR to specifically target particular taxonomic or functional markers from domain – down to strain – or phylotype levels means that PCR will remain an invaluable method in the molecular microbial ecologist’s toolbox. Nevertheless, PCR has inher- ent limitations (von Wintzingerode et al., 1997), particularly those that result in biases in the template to product ratios of target sequences amplified during PCR from environmental DNA (Suzuki & Giovannoni, 1996; Polz & Cavanaugh, 1998), with such amplification biases found to increase with increasing numbers of PCR cycles. These limitations pre- sented a significant challenge to microbial ecologists who were interested in determining the abundance of individual genes present in environmental samples. To circumvent such challenges, an adaptation of the PCR method developed by Holland et al. (1991) utilizing the so-called ‘50 nuclease assay’ was applied to quantify target 16S rRNA genes amplified from environmental DNA by PCR (Becker et al., 2000; Suzuki et al., 2000; Takai & Horikoshi, 2000). This FEMS Microbiol Ecol 67 (2009) 6–20c� 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved development had been facilitated by the earlier combination of the 50 nuclease assay developed by Holland et al. (1991) with fluorescence detection following cleavage of an internal (TaqManTM) DNA probe (Livak et al., 1995), enabling the accumulation of amplicons to be monitored after each cycle (in real-time) and hence facilitating quantitative determina- tion of the initial template gene (or transcript) numbers. Quantitative-PCR or Q-PCR (often referred to as real- time PCR) is now widely used in microbial ecology to determine gene and/or transcript numbers present within environmental samples. The target specificity of any Q-PCR assay is determined by the design of the primers (and in some cases an internal probe), allowing quantification of taxonomic or functional gene markers present within a mixed community from the domain level down to the quantification of individual species or phylotypes. Q-PCR has been shown to be a robust, highly reproducible and sensitive method to quantitatively track phylogenetic and functional gene changes across temporal and spatial scales under varying environmental or experimental conditions. Moreover, the quantitative data generated can be used to relate variation in gene abundances and/or levels of gene expression (in terms of transcript numbers) in comparison with variation in abiotic or biotic factors and/or biological activities and process rates. The provision of Q-PCR data sets that describe the abundance of specific bacteria or genes to complement other quantitative environmental data sets is of increasing importance in microbial ecology as it furthers understanding of the roles and contributions of particular microbial and functional groups within ecosys- tem functioning. Furthermore, reverse transcription (RT) analyses are now increasingly combined with Q-PCR methods in RT-Q-PCR assays, offering a powerful tool for quantifying gene expression (in terms of numbers of rRNA and mRNA transcripts) and relating biological activity to ecological function. In this review, we firstly discuss the mechanistic aspects of Q-PCR and RT-Q-PCR methodologies, hereafter defined collectively as (RT)-Q-PCR, and highlight the key experi- mental considerations in the design and implementation of (RT)-Q-PCR protocols and the analysis of resultant data sets. Secondly, we explore the application of (RT)-Q-PCR approaches in microbial ecology, and finally we discuss how these methods can be applied together with other molecular and also conventional approaches to provide an increased understanding of microorganisms within environmental systems. Advantages of Q-PCR over traditional end- point PCR Q-PCR approaches combine the detection of target template with quantification by recording the amplification of a PCR product via a corresponding increase in the fluorescent signal associated with product formation during each cycle in the PCR. Quantification of gene (or transcript) numbers is determined during the exponential phase of the PCR amplification when the numbers of amplicons detected are directly proportional to the initial numbers of target sequences present within the environment (discussed in more detail in Target quantification). Quantification of the target gene during exponential amplification avoids pro- blems that are associated with so-called ‘end-point’ PCR (in which amplicons are only analysed after completion of the final PCR cycle). In end-point PCR, the proportions of numerically dominant amplicons do not necessarily reflect the actual abundances of sequences present within the environment due to the inherent biases of PCR that are associated with amplification of targets from mixed template community DNA (Reysenbach et al., 1992; Suzuki & Giovannoni, 1996; Polz & Cavanaugh, 1998). Moreover, Q-PCR that uses fluorescence-based detection offers greater sensitivity and enables discrimination of gene numbers across a wider dynamic range than is found with end-point PCR; for example twofold changes in target concentration can be discriminated using Q-PCR. Before the development of fluorescence-based Q-PCR-based methods, two alterna- tive PCR-based methods for gene number quantification had been developed, namely competitive PCR (Diviacco et al., 1992) and limiting dilutions or most probable number (MPN)-PCR (Skyes et al., 1992). However, these methods are time- and resource-consuming, requiring post-PCR analysis, and have now largely been replaced by fluores- cence-based Q-PCR methods. Fluorescence detection chemistries used to detect template amplification during Q-PCR Quantitative real-time PCR works in essentially the same manner as end-point PCR, i.e. multiple amplification cycles in which template DNA is initially denatured, followed by annealing of oligonucleotide primers targeting specific sequences, followed by subsequent extension of a comple- mentary strand from each annealed primer by a thermo- stable DNA polymerase, resulting in an exponential increase in amplicon numbers during the PCR. However, in contrast to end-point PCR, the increase in amplicon numbers is recorded in ‘real-time’ during the PCR via detection of a fluorescent reporter that indicates amplicon accumulation during every cycle. Two reporter systems are commonly used, namely, the intercalating SYBR green assay (Wittwer et al., 1997) and the TaqMan probe system (Holland et al., 1991; Livak et al., 1995). SYBR green binds to all double-stranded DNA via inter- calation between adjacent base pairs. When bound to DNA, FEMS Microbiol Ecol 67 (2009) 6–20 c� 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved 7Application of Q-PCR in microbial ecology a fluorescent signal is emitted following light excitation (Fig. 1a). As amplicon numbers accumulate after each PCR cycle, there is a corresponding increase in fluorescence. Because SYBR green binds to all double-stranded DNA, it is essential to use primer pairs that are highly specific to their target sequence to avoid generation of nonspecific products that would contribute to the fluorescent signal, resulting in an overestimation of the target. Extensive optimization of primer concentrations used in SYBR green Q-PCR assays may be required to ensure that only the targeted product is formed. Primer pairs that exhibit self-complementarity should also be avoided to prevent primer–dimer formation. A post-PCR dissociation (melting) curve analysis should be carried out to confirm that the fluorescence signal is generated only from target templates and not from the formation of nonspecific PCR products. During a dissocia- tion curve, the double-stranded template is heated over a temperature gradient and fluorescence levels are measured at each discrete temperature point. As the double-stranded template is heated, it denatures, resulting in a corresponding decline in fluorescence due to SYBR green dissociation from the double-stranded product (Giglio et al., 2003; Gonzalez- Escalona et al., 2006). The temperature at which 50% of the double-stranded template is denatured can be used to confirm that the template being targeted is present, along with the presence of other nonspecific template and primer dimers in much the same way as agarose gel electrophoresis of an end-point PCR product is used. The TaqMan probe method utilizes a fluorescently labelled probe that hybridizes to an additional conserved region that lies within the target amplicon sequence. The TaqMan probe is fluorescently labelled at the 50 end and contains a quencher molecule at the 30 end (Livak et al., 1995). The close proximity on the probe of the quencher molecule to the fluorophore prevents it from fluorescing due to fluorescent resonance energy transfer. During the annealing step of each cycle of the PCR, primers and the intact probe bind to their target sequences. During subse- quent template extension, the 50 exonuclease activity of the Taq polymerase enzyme cleaves the fluorophore from the TaqMan probe and a fluorescent signal is detected as the fluorophore is no longer in close proximity to the quencher (Fig. 1b). Amplification of the template is thus measured by the release and accumulation of the fluoro- phore during the extension stage of each PCR cycle. The additional specificity afforded by the presence of the TaqMan probe ensures that the fluorescent signal generated during Q-PCR is derived only from amplification of the target sequence. Multiple TaqMan probes and primer sets can be used in different Q-PCR assays to differentiate between closely related sequences (Smith et al., 2007), or alternatively, probes can be labelled with different fluoro- phores, facilitating the development of multiplex Q-PCR protocols whereby different targets can be coamplified and quantified within a single reaction (Neretin et al., 2003; Baldwin et al., 2003, 2008). For example, Baldwin et al. (2003) developed a multiplex Q-PCR assay targeting a number of different aromatic oxygenase genes using bacter- ial strains and then subsequently applied the assay to simultaneously quantify aromatic oxygenase genes in con- taminated groundwater (Baldwin et al., 2008). TaqMan probes are, however, a more expensive option than using SYBR green chemistry and the former requires the presence of an additional conserved site within the short amplicon sequence to be present. Identification of three conserved regions within a short region (typically c.o 100 bp) may not Fig. 1. Real-time PCR chemistries: (a) SYBR green detection. SYBR green binds to all double-stranded DNA and emits a fluorescent signal. In its unbound state, SYBR green does not fluoresce. Template amplification is therefore measured in each cycle by the corresponding increase in fluorescence. (b) TaqMan (50 nuclease) assay using TaqMans probes. During annealing, the TaqMan probe and primers bind to the template. When the TaqMan probe is intact, energy is transferred between the quencher and the reporter; as a result, no fluorescent signal is detected. As the new strand is synthesized by Taq polymerase, the 50 exonuclease activity of the enzyme cleaves the labelled 50 nucleotide of the probe, releasing the reporter from the probe. Once it is no longer in close proximity, the fluorescent signal from the probe is detected and template amplification is recorded by the corresponding increase in fluorescence. FEMS Microbiol Ecol 67 (2009) 6–20c� 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved 8 C.J. Smith & A.M. Osborn always be possible, especially when primer/probe combina- tions are being designed to target divergent gene sequences. More recent advances in TaqMan probe technology have involved the introduction of the minor groove binder (MGB) probe (Kutyavin et al., 2000). The MGB molecule is attached to the 30 end of the probe and essentially folds back onto the probe. This not only increases the stability of the probe, but allows the design of shorter probes (13–20 bp) than are required for traditional TaqMan probes (20–40 bp), while at the same time, maintaining the required hybridiza- tion annealing temperature. Target quantification using the cycle threshold (C t) method Irrespective of the fluorescence chemistry used, quantifica- tion of the target template DNA is carried out in essentially the same manner. There are a number of different commer- cially available instruments to carry out Q-PCR, each with its own associated software. Currently, there is considerable debate as to which algorithms are the best used to analyse Q-PCR data (reviewed in Rebrikov & Trofimov, 2006). All the Q-PCR platforms collect fluorescent data from every amplification cycle and the increase in fluorescence is plotted against the cycle number, resulting in the typical amplification curve shown in Fig. 2. The Q-PCR amplifica- tion curve can be subdivided into four stages, namely background noise, where the background fluorescence still exceeds that derived from initial exponential template accumulation, exponential amplification, linear amplifica- tion and a plateau stage. During the exponential phase of the amplification, the amount of target amplified is propor- tional to the starting template and it is during these cycles that gene numbers are quantified using the Ct method. The Ct is reached when the accumulation of fluorescence (tem- plate) is significantly greater than the background level Fig. 2. Q-PCR amplification from known concentrations of template DNA to construct standard curves for quantification of unknown environmental samples. (a) Log plot of the increase in fluorescence vs. cycle number of DNA standards ranging from 1� 104 to 1�108 16S rRNA gene amplicons mL�1. (b) Linear plot indicating the three phases of a PCR amplification, the corresponding Ct values for each of the amplified standards and for the NTC. (c) Simple linear regression of the Ct values (from b) vs. log of the initial rRNA gene number. Q-PCR descriptors are shown (boxed). FEMS Microbiol Ecol 67 (2009) 6–20 c� 2008 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved 9Application of Q-PCR in microbial ecology (Heid et al., 1996). During the initial cycles, the fluorescence signal due to background noise is greater than that derived from the amplification of the target template. Once the Ct value is exceeded, the exponential accumulation of product can be measured. When the initial concentration of the target template is higher, the Ct will be reached at an earlier amplification cycle. Quantification of the initial target sequences of an unknown concentration is determined from the Ct values and can be described either in relative or in absolute terms. In relative quantification, changes in the unknown target are expressed relative to a coamplified steady state (typically a housekeeping) gene. Any variation in the presence (or expression) of the housekeeping gene can potentially mask real changes or indicate artificial changes in the abundance of the gene of interest. While this approach is commonly applied for studying eukaryotic gene expression (reviewed in Bustin, 2002), it is more difficult to apply this method for studying prokaryotic genes where the identification of a valid steady-state reference gene is problematic. Burgmann et al. (2007) nevertheless successfully utilized such an approach when confirming microarray-based determination of the transcriptional responses of Silicibacter pomeroyi to dimethylsulphoniopropionate additions. From microarray experiments, they identified a gene whose expression was not altered by experimental conditions and used the expres- sion of this gene to normalize levels of expression of the target genes of interest in RT-Q-PCR assays. In a number of other studies, gene and transcript numbers of the target gene of interest have been normalized to the numbers of 16S rRNA gene or transcripts (Neretin et al., 2003; Treusch et al., 2005; Kandeler et al., 2006). For example, Treusch et al. (2005) normalized the number of amoA transcripts to numbers of 16S rRNA gene transcripts