Comparison of Quantitative PCR and Droplet Digital PCR Multiplex Assays for Two Genera of Bloom-Forming Cyanobacteria, Cylindrospermopsis and Microcystis
ABSTRACT
The increasing occurrence of harmful cyanobacterial blooms, often linked to deteriorated water quality and adverse public health effects, has become a worldwide concern in recent decades. The use of molecular techniques such as real-time quantitative PCR (qPCR) has become increasingly popular in the detection and monitoring of harmful cyanobacterial species. Multiplex qPCR assays that quantify several toxigenic cyanobacterial species have been established previously; however, there is no molecular assay that detects several bloom-forming species simultaneously. Microcystis and Cylindrospermopsis are the two most commonly found genera and are known to be able to produce microcystin and cylindrospermopsin hepatotoxins. In this study, we designed primers and probes which enable quantification of these genera based on the RNA polymerase C1 gene for Cylindrospermopsis species and the c-phycocyanin beta subunit-like gene for Microcystis species. Duplex assays were developed for two molecular techniques—qPCR and droplet digital PCR (ddPCR). After optimization, both qPCR and ddPCR assays have high linearity and quantitative correlations for standards. Comparisons of the two techniques showed that qPCR has higher sensitivity, a wider linear dynamic range, and shorter analysis time and that it was more cost-effective, making it a suitable method for initial screening. However, the ddPCR approach has lower variability and was able to handle the PCR inhibition and competitive effects found in duplex assays, thus providing more precise and accurate analysis for bloom samples.
INTRODUCTION
Cyanobacteria, also known as blue-green algae, constitute the most notorious phylum of phytoplankton capable of forming harmful blooms in freshwater aquatic ecosystems (1). Their presence in freshwater and brackish and coastal marine waters is of particular interest because of their massive accumulation and proliferation into nuisance blooms in nutrient-enriched water. These cyanobacterial blooms are the cause for a multitude of water quality concerns, due to their potential to produce secondary metabolites, some of which are toxins and compounds that compromise taste and odor. Cyanotoxins have been linked to human and animal illness and death and pose serious health hazards to communities which use surface waters and reservoirs as potable waters (2). Hence, the detection of cyanobacteria is crucial for reliable and prudent water management. The use of efficient detection methods in routine monitoring of waters is necessary to safeguard precious water resources.
Microscopic counting combined with chemical detection of cyanotoxins in water samples is the conventional method to evaluate harmful cyanobacterial blooms (3). However, morphological identification is prone to limitations such as being time-consuming and unable to distinguish between toxic and nontoxic species and prone to misinterpretation when limited morphological differences are available (4, 5). In light of these discrepancies, DNA-based detection methods are becoming increasingly popular because of increasing bioinformatics knowledge and the availability of genetic databases that allow the design of specific, sensitive, and speedy molecular assays (6). These assays were developed to amplify unique nucleotide sequences of a target gene of interest, for example, functional genes responsible for the biosynthesis of cyanotoxins (mcy, cyr, and nda), or housekeeping gene (16S rRNA and RNA polymerase gene) sequences exclusively for certain cyanobacterial species (7, 8). The high sensitivity of molecular methods allows the detection of harmful cyanobacterial species even at low abundance, thus providing rapid evaluation before the occurrence of a cyanobacterial bloom, which cannot be achieved by conventional counting methodology.
Real-time quantitative PCR (qPCR) has been successfully used to monitor harmful cyanobacterial populations in several studies. Most of these assays were designed for a single target, while only a few multiplex assays were developed to detect multiple analytes simultaneously (9–14). Despite the advantages of multiplexing qPCR in identification of taxa (such as offering quicker screening and being less labor-intensive), there is currently no method which can detect and quantify several genera of bloom-forming cyanobacteria in one assay.
Digital PCR (dPCR) is an alternative method of quantitative PCR that derives the target's abundance from the ratio of the number of positive partitions to a large number (hundreds to millions) of total reaction partitions generated from a known volume of PCR mix. The absolute number of copies of the target is calculated from the proportion of positive partitions and statistically corrected with a Poisson distribution (15, 16). Droplet digital PCR (ddPCR) is one of the commercially available dPCR approaches and generates microfluid droplets in oil, forming tens of thousands of tiny PCRs in an oil-water emulsion. The intrinsic nature of ddPCR allows detection of rare nucleic acids in the presence of a high abundance of standard sequences (15). The applications of this technology have mainly been focused on clinical research and diagnosis, genetic modification, food product screening, and viral surveillance (17–23), while the potential of using ddPCR in environmental studies has yet to be explored.
In this study, we developed duplex quantification assays targeting two ubiquitous cyanotoxin-producing cyanobacterial genera found in most freshwater systems—Microcystis assay (MIC) and Cylindrospermopsis assay (CYL)—using both qPCR and ddPCR techniques. The assay performances of the two methods are also compared. The developed assays can be used to detect and quantify the two most widely reported cyanobacterial genera in both laboratory and environmental samples.
MATERIALS AND METHODS
Primer and probe design.
Primers and probes were designed based on specific sequences of the RNA polymerase C1 gene (rpoC1) for Cylindrospermopsis species and the c-phycocyanin beta subunit-like gene (cpcB) for Microcystis species. Nucleotide sequences (43 rpoC1 sequences and 73 cpcB sequences) from various Cylindrospermopsis and Microcystis strains were extracted from the National Center for Biotechnology Information (NCBI) database and aligned using Mega 4.0. The aligned sequences were examined manually to determine conserved regions suitable as targets. Potential primer and probe sequences were first generated using IDT PrimerQuest software and then adjusted manually following the instructions of the TaqMan Multiplex PCR Optimization User Guide (Life Technologies) for optimum assay efficiency, with the following specifications.
1.
The primer melting temperatures (Tm) should be similar for all primers and for the two probes.
2.
The Tm of the probe should be ∼10°C higher than the Tm of the primers.
3.
The sizes of amplicons should be between 50 and 150 bp.
4.
The primers and probes should not be self-complementary or complementary to each other.
5.
GC content should be between 40% and 60%.
6.
The probe length should be 13 to 30 bp, the 5′ end should not be a G residue, and repeating identical nucleotides should be avoided.
Possible primer dimers and hairpins were examined using MFEprimer-2.0 (http://biocompute.bmi.ac.cn/CZlab/MFEprimer-2.0/index.cgi/check_dimer), with a ΔG value of not less than −5 kcal/mole as the recommended value. Specificities of the target regions were initially determined with BLASTN (NCBI).
In addition, qPCR was performed on DNA from laboratory cultures of species of various taxa of cyanobacteria, including Microcystis, Cylindrospermopsis, Limnothrix, Anabaena, Pseudanabaena, Planktothrix, Hapalosiphon, and Synechococcus, as well as noncyanobacterial species, including Actinastrum and Chlorella species, to verify the specificity of the assays. qPCR product sizes from the DNAs of the cultures were checked by gel electrophoresis performed with a 1% agarose gel stained with GelRed (Biotium).
Cloning and plasmid standard synthesis.
Plasmid DNAs were used to establish the standard curves. Given the known vector size and target qPCR regions (quantified in base pairs), an estimate of gene copy numbers (GCNs) based on the plasmid DNA is more accurate than estimates based on the genomic DNA. To avoid the possibility of mispaired nucleotides at the ends of the PCR product, two sets of extended primers (Table 1) were used to amplify longer sequences flanking the target qPCR regions. PCR amplifications of the DNA of M. aeruginosa PCC7806 and Cylindrospermopsis strain CS505 were carried out in a Mastercycler Pro PCR system (Eppendorf) subjected to the following steps: polymerase activation at 95°C for 2 min; 35 amplification cycles of 95°C for 20 s, 58°C for 30 s, and 72°C for 30 s; and a final extension at 72°C for 5 min. Each reaction mixture contained 10 μl of 2× GoTaq Hot Start Green master mix (Promega), 0.5 μM (each) forward and reverse primers, and 2 μl of the DNA template. PCR products were cleaned with a Wizard SV gel and PCR clean-up system (Promega) and cloned into pGEM-T Easy vector (Promega) following the manufacturer's instructions. Successful Escherichia coli clones were inoculated in 10 ml of LB broth medium and grown overnight using 37°C incubation. Plasmid DNA was extracted using a QIAprep Spin Miniprep kit (Qiagen), followed by restriction enzyme digestion (SalI; Promega). The quality and concentration of linearized plasmids were determined with a NanoDrop 1000 Spectrophotometer (BioFrontier). Plasmid DNAs generated were also sequenced in a ABI 3730xl DNA analyzer using an ABI BigDye Terminator v3.1 cycle sequencing kit (Life Technologies) to verify correct target sequences (see the supplemental material).
TABLE 1
| Primer or probe | Nucleotide sequence (5′–3′)a | Tm (°C) | Amplicon size (bp) |
|---|---|---|---|
| CYL rpo F | GCATTCCTAGTTATATTGCCAT | 59.0 | 154 |
| CYL rpo R | TCTAACCATTGGTCTTCTGTTAA | 59.7 | |
| CYL probe | FAM-CTAGACATGCCCCTACGAGAYGTTGAGC-BHQ-1 | 73.5 | |
| MIC cpc F | GYTATGTYACCTACGCTACCTTC | 60.5 | 107 |
| MIC cpc R | TCCWGGTACTCCTAAAGCTACA | 59.8 | |
| MIC probe | Yakima yellow/HEXb-ACCATTTAAGCAACGATCATCGAGAACAC-BHQ-1 | 71.7 | |
| Cloning | |||
| CYL rpo F (extended) | TGTAGCCCATGTTTGGTATCTCAAAG | 67.8 | 206 |
| CYL rpo R (extended) | TGAGAATCTTCACTATAAATCCTATCC | 61.0 | |
| MIC cpc F (extended) | TTACGCGACATGGAAATCAT | 63.3 | 152 |
| MIC cpc R (extended) | TTGCTTACGCCAGCGGCTA | 69.2 |
a
FAM, 6-carboxyfluorescein; BHQ-1, black hole quencher 1.
b
Yakima Yellow reporter dye was used in the qPCR assay; HEX reporter dye was used in the ddPCR assay.
qPCR optimization and standard curves.
A duplex qPCR assay was developed to enumerate gene copy numbers of CYL rpoC1 and MIC cpcB genes simultaneously. Amplifications were carried out using a 0.1-ml MicroAmp Fast Reaction 8-tube strip (Life Technologies) in a 20-μl reaction volume using a StepOnePlus real-time PCR system (Life Technologies). Plasmid standards cloned with targets were diluted 10-fold and amplified using primers and dual-label probes designed in this study. To optimize the duplex assay, amplifications were run in monoplex and duplex fashions in parallel. The assay conditions were changed until the monoplex and duplex assays generated comparable quantification cycle (Cq) values for the same samples (Cq difference < 1) with an amplification efficiency of 100% ± 10%. The duplex qPCR assay was optimized in two ways via optimization of (i) amplification conditions and (ii) reagent composition. For the amplification conditions, assay performances for 2-step and 3-step amplifications at various annealing temperatures (from 50 to 60°C) were compared. For the reagent composition, different master mixes, primers, probes, and MgCl2 concentrations were tested. The optimum assay consisted of reagents as follows: 10 μl of 2× QuantiFast multiplex PCR master mix (Qiagen), 0.5 μM (each) MIC cpc F/R primer, 0.75 μM (each) CYL rpo F/R primer, 0.2 μM (each) probe, 1 μM MgCl2, and 2 μl of DNA template in a total reaction volume of 20 μl. The amplification cycles included an enzyme activation step at 95°C for 5 min and 40 cycles of 3-step amplification of 15 s at 95°C, 25 s at 57°C, and 25 s at 72°C.
Standard curves of monoplex and duplex assays were established in a similar way. Serially diluted plasmid samples with 100 to 106 copies of target were prepared in duplicate or triplicate, and the standard curve was generated as a linear regression between Cq values and the logarithmic target or gene copy number (GCN). The GCN in 1 μl of sample was calculated from the concentration of plasmid used.
ddPCR optimization.
The duplex ddPCR assay optimization was similar to the optimization for qPCR, where various thermal cycling conditions (2-step versus 3-step; temperature gradient for annealing; cycle number) and primer-probe concentrations were evaluated. Satisfactory separation of droplets positive for the target from those negative for the target, assay sensitivity comparable to qPCR sensitivity, and linearity were the criteria used in optimization. The same primers and probes were used in ddPCR, except that for the reporter dye for the MIC assay the probe was changed to HEX in ddPCR to allow optimum florescence detection. The optimized PCR mixture contained 10 μl of 2× ddPCR Supermix for probes (Bio-Rad) (no dUTP), 0.9 μM (each) primer, 0.25 μM (each) probe, and 2 μl of the DNA template. Quantification of targets was carried out in a QX200 Droplet Digital PCR system (Bio-Rad Laboratories, Inc.). First, 20 μl of each well-mixed PCR mixture was transferred to a droplet generator cartridge (Bio-Rad). After 70 μl of droplet generation oil (Bio-Rad) was added into the oil wells, the cartridge was covered with a rubber gasket and loaded onto a QX200 droplet generator (Bio-Rad). The emulsions of droplets generated were transferred to a 96-well PCR plate (Eppendorf), sealed with PX1 PCR plate sealer (Bio-Rad), and subjected to amplification in a C1000 Touch thermal cycler (Bio-Rad). The amplification was carried out at a uniform ramp rate of 2.5°C/s at 95°C for 10 min; 45 cycles of 95°C for 15 s followed by 58°C for 1 min; and a final enzyme deactivation at 98°C for 10 min. Fluorescent signals from amplified droplets were captured individually in the QX200 droplet reader (Bio-Rad) and analyzed with QuantaSoft 1.6.6 software. The distinction between positive and negative droplets (with and without target) was based on the threshold values assigned for all samples (9,000 for CYL and 5,400 for MIC) and on the threshold values automatically obtained from different sample types (i.e., plasmid DNA and DNA from cyanobacterial cultures and environmental sample extracts). The target concentrations were reported as the numbers of copies per microliter of the PCR mixture after correction with the Poisson distribution.
Sensitivity, precision, and competitive effect.
Plasmid samples containing 10, 102, 103, 104, or 105 target copies per reaction, estimated from previous ddPCR results, were prepared and evaluated with qPCR and ddPCR duplex reactions. To evaluate assay sensitivity, reaction mixtures containing less than 10 target copies were prepared and run for each technique. The limit of detection (LOD) for each assay, defined as the smallest amount of analyte in a sample that could be confidently detected (95th percentile), was calculated following guidelines approved by the NCCLS (24). To examine possible competitive effects in CYL and MIC assays in duplex reactions, samples containing low copy numbers of one target were mixed with a 10-fold-increased level of another target and were then analyzed with both qPCR and ddPCR.
Assay evaluation using laboratory cultures and environmental samples.
Once the conditions were optimized, the assay was verified using laboratory cultures and environmental water samples. Six strains of cyanobacteria, three Microcystis strains (PCC7806, NIES843, and local isolate RKC), and three Cylindrospermopsis strains (CS505, CS509, and local isolate CYL2) grown in MLA medium (temperature, 23 to 25°C; light intensity, 25 μmol quanta/m2 · s) (45) were harvested during the early stationary phase. Concentrations of Microcystis strains ranged from 1.9 × 106 to 9.4 × 106 cell/ml, while concentration of Cylindrospermopsis strains ranged from 1.4 × 107 to 2.5 × 107 cell/ml. The number of cells was determined by a microscopic count method where 10 μl of sample fixed with Lugol's solution (25) was loaded onto a disposable hemocytometer (C-Chip; iNCYTO) and the cells were counted under an inverted microscope (Leica DM IL light-emitting-diode [LED] fluorescence microscope). Duplicate measurements were made for each culture. For cyanobacterial genomic DNA extraction, 15 ml of each laboratory culture was filtered onto a 0.45-μm-pore-size cellulose nitrate membrane, followed by DNA extraction using a PowerWater DNA isolation kit (MoBio). These samples were also analyzed for total cyanobacterial 16S rRNA gene abundance, using previously reported primers and probe (26) that had been optimized on qPCR and ddPCR platforms.
Environmental water samples were also harvested and extracted with the same method and analyzed with qPCR and ddPCR techniques using the CYL and MIC assays.
Statistical analyses.
Significant differences between GCNs and Cq values of different GCNs were determined using the independent-sample t test and paired t test (Predictive Analytics SoftWare [PASW] version 18).
RESULTS
Assay specificity.
The specificities of the primers and probes were evaluated using BLASTN. The combination of CYL primer and probe sequences matched 100% of 35 sequences from Cylindrospermopsis species and 2 sequences from Anabaena sphaerica var. tenuis, while the combination of MIC cpcB primers and probe was specific only to Microcystis species (100% identity). Monoplex qPCR conducted on laboratory algal cultures also indicated that the primers and probes were specific to the desired targets (Table 2).
TABLE 2
| Cyanobacterial/algal culture | Assay result | |
|---|---|---|
| MIC qPCR monoplex | CYL qPCR monoplex | |
| Microcystis sp. strain RK18b | + | − |
| Microcystis sp. strain RKCb | + | − |
| Microcystis sp. strain RK1b | + | − |
| Microcystis sp. strain RK7b | + | − |
| Microcystis sp. strain RP1Db | + | − |
| Microcystis aeruginosa PCC7806 | + | − |
| Microcystis aeruginosa CS-573 | + | − |
| Microcystis aeruginosa NIES843 | + | − |
| Cylindrospermopsis sp. strain LPRbb | BLOQ | + |
| Cylindrospermopsis sp. strain CYL1b | − | + |
| Cylindrospermopsis sp. strain CYL2b | − | + |
| Cylindrospermopsis sp. strain CYL8b | − | + |
| Cylindrospermopsis raciborskii CS-505 | − | + |
| Cylindrospermopsis raciborskii CS-509 | − | + |
| Cylindrospermopsis raciborskii CS-511 | − | + |
| Anabaena circinalis CS-337-01 | − | − |
| Anabaena circinalis CS-541-06 | − | − |
| Nodularia spumigena CS-336-05 | − | − |
| Nodularia spumigena CS-588-02 | − | − |
| Limnothrix sp. strain RU5Ab | BLOQ | − |
| Pseudanabaena sp. strain RM13Ab | − | − |
| Synechococcus sp.b | − | − |
| Chlorella sp. strain RM1b | BLOQ | BLOQ |
| Actinastrum sp. strain PR1b | − | − |
a
+, Cq within quantification range; BLOQ, Cq below limit of quantification (LOQ); −, amplification undetectable. The concentrations of algal culture DNA were 1.25 to 23.90 ng/μl.
b
Strains isolated from local reservoirs.
qPCR assay performance.
The linear dynamic range of quantification for duplex qPCR was between 100 and 106 gene copies/reaction (i.e., 7.5 × 100 to 7.5 × 106 gene copies/reaction for Cylindrospermopsis and 8.5 × 100 to 8.5 × 106 gene copies/reaction for Microcystis). The monoplex and duplex reaction efficiencies and regressions were almost identical (Fig. 1A and B), with differences of less than 1 between the Cq values of the monoplex and duplex assays for each concentration (Table 3). The LOD values for the CYL and MIC assays were 6.88 and 7.36 gene copies/reaction, respectively.
FIG 1
TABLE 3
| No. of copies of plasmid DNA | Cq for MIC cpcBa | Cq for CYL rpoC1a | ||
|---|---|---|---|---|
| Monoplex | Duplex | Monoplex | Duplex | |
| 100 | 36.03 (0.11) | 35.39 (0.30) | 35.26 (0.78) | 35.75 (0.75) |
| 101 | 32.18 (0.27) | 32.13 (0.17) | 31.43 (0.33) | 32.34 (0.21) |
| 102 | 28.70 (0.09) | 28.56 (0.31) | 28.38 (0.23) | 29.00 (0.21) |
| 103 | 25.21 (0.10) | 25.06 (0.17) | 25.07 (0.47) | 25.56 (0.22) |
| 104 | 21.74 (0.08) | 21.56 (0.17) | 21.76 (0.26) | 22.14 (0.17) |
| 105 | 18.20 (0.18) | 18.18 (0.16) | 17.91 (0.63) | 18.68 (0.23) |
| 106 | 14.89 (0.15) | 14.75 (0.29) | 14.57 (1.18) | 15.41 (0.21) |
a
Values shown represent averages of 5 readings from duplicate and triplicate samples analyzed separately. Standard deviations are in parentheses.
ddPCR assay performance.
Using the same dilution series of plasmid DNA for qPCR standard curves, each concentration of plasmid standard was run in triplicate on two separate runs. The dynamic range for ddPCR reactions was between 100 and 104 gene copies/reaction (Fig. 2), while the LOD values were 4.26 gene copies/reaction for the CYL assay and 3.52 gene copies/reaction for the MIC assay. The linearity decreased for lower target concentrations at 100 copies, and reaction saturation was reached at a concentration of 105 gene copies/reaction. The GCNs obtained for the monoplex and duplex reactions were indistinguishable (t test, P > 0.05) for the range from 100 to 103 gene copies/reaction. Slightly lower amplification efficiency was observed for the duplex assay when both targets reached a concentration of 104 gene copies/reaction (t test, P < 0.05), but the discrepancy between monoplex and duplex assays was below 5%.
FIG 2
Comparison between qPCR and ddPCR.
The variability of the GCNs measured with the qPCR and ddPCR techniques is shown in the quantitative correlations of qPCR and ddPCR measurements (Fig. 3), where the slopes are 1.1238 for Cylindrospermopsis and 1.1436 for Microcystis with an R2 value of ≥0.98. The numbers of target copies measured by the two platforms were well correlated, with qPCR generally providing higher GCNs for both targets. GCNs estimated with qPCR were approximately 1.3-fold to 6.8-fold (average, 2.8-fold) higher than those measured by ddPCR.
FIG 3
A sensitivity test using samples containing small amounts of targets, where 94.4% of the qPCR samples gave positive amplification compared to 80.5% of the ddPCR samples, indicated that qPCR has greater sensitivity than ddPCR. However, ddPCR showed higher precision, with lower values for standard deviation (SD) and percent coefficient of variation (%CV) (Table 4).
TABLE 4
| Parametera | Valueb | |||||||
|---|---|---|---|---|---|---|---|---|
| CYL rpoC1 | MIC cpcB | |||||||
| qPCR | ddPCR | qPCR | ddPCR | |||||
| 1 | 2 | 1 | 2 | 1 | 2 | 1 | 2 | |
| Target concn (GCN/reaction) | 8.40 | 2.10 | 8.40 | 2.10 | 5.20 | 1.30 | 5.20 | 1.30 |
| Measured mean concn (GCN/reaction) | 7.24 | 3.34 | 7.20 | 2.33 | 7.54 | 6.80 | 7.27 | 1.80 |
| SD | 4.05 | 4.17 | 3.51 | 0.82 | 4.12 | 7.11 | 2.92 | 0.59 |
| % CV | 55.91 | 125.09 | 48.75 | 35.10 | 54.69 | 104.65 | 40.17 | 32.96 |
| Ratio of no. of positive samples to total no. of samplesc | 9/9 | 8/9 | 8/9 | 6/9 | 9/9 | 8/9 | 9/9 | 6/9 |
a
SD, standard deviation; CV, coefficient of variation.
b
Columns labeled 1 and 2 represent two different sets of results, based on target concentration.
c
Positive sample, sample with detectable amplification signal.
Competition between CYL and MIC assays was observed in the duplex qPCR. The effect was considered acceptable (Cq ≤ 1) when the concentration of one target was <10 times higher than the concentration of the other target. Beyond this range, however, the competitive effect increased, with greater differences in the two target concentrations (Fig. 4A and B) (t test, P < 0.05). However, measured target copy numbers were still of the same magnitude when the concentration for the second target was 1,000 (CYL assay) or 100 (MIC assay) times higher. Notably, this competitive effect was not detected in the ddPCR reaction when low copy numbers of one target were amplified in the presence of large amounts of the other target. The monoplex and duplex reactions gave identical results as shown in Fig. 4C and D (t test, P > 0.05), although the latter contained at least 104 copies of the other target.
FIG 4
The copy numbers of the CYL rpoC1, MIC cpcB, and 16S rRNA genes for 6 laboratory cultures were quantified using qPCR and ddPCR. The ratios of the 16S rRNA gene copy numbers to the CYL rpoC1 gene copy numbers for Cylindrospermopsis spp. were 2.33 for qPCR and 3.00 for ddPCR, whereas the ratios of the 16S rRNA gene copy numbers to the MIC cpcB gene copy numbers for Microcystis spp. were 1.50 for qPCR and 1.84 for ddPCR (Table 5). This result is consistent with the complete genome sequences in NCBI database, showing that C. raciborskii CS 505 contains 3 copies of the 16S rRNA gene and 1 copy of the rpoC1 gene per genome whereas M. aeruginosa NIES 843 contains 2 copies of the 16S rRNA gene and 1 copy of the cpcB gene per genome (NCBI accession number AP009552; GCA_000175835.1). Our study showed that ddPCR was more accurate in the prediction of the copy numbers of target genes in a genome.
TABLE 5
| Cylindrospermopsis strain | 16S rRNA copy no./rpoC1 copy no. ratio (SD)a | Microcystis strain | 16S rRNA copy no./cpcB copy no. ratio (SD)a | ||
|---|---|---|---|---|---|
| qPCR | ddPCR | qPCR | ddPCR | ||
| CYL2 | 2.64 (0.02) | 3.03 (0.05) | RKC | 1.32 (0.01) | 1.66 (0.01) |
| CS505 | 2.47 (0.01) | 2.94 (0.06) | PCC7806 | 1.82 (0.05) | 1.88 (0.02) |
| CS509 | 1.88 (0.03) | 3.01 (0.07) | NIES843 | 1.36 (0.01) | 1.98 (0.02) |
| Overall | 2.33 (0.36) | 3.00 (0.06) | 1.50 (0.25) | 1.84 (0.15) | |
a
Data represent the averages of the results of triplicate determinations, with standard deviations indicated in parentheses.
Evaluation of environmental samples.
The GCNs measured by qPCR and ddPCR for 17 environmental samples are presented in Fig. 5. Overall, the data generated by the two techniques fell within the same order of magnitude. Several samples showed significant differences in GCNs using the two techniques (Cq > 1). For the CYL assay, the mean difference between qPCR and ddPCR was 0.344 log10 (GCN)/ml (SD, 0.210 log10 [GCN]/ml). The MIC assay showed less variability for the two techniques, with a mean difference of 0.264 log10 (GCN)/ml (SD, 0.113 log10 [GCN]/ml). The MIC cpcB GCN measured by qPCR was generally higher than that measured by ddPCR (for 15 of 17 samples) (paired t test, P < 0.05). This is in agreement with the quantitative correlation of qPCR and ddPCR obtained from the plasmid DNA assays (Fig. 3). However, this result was not found in the CYL assay (paired t test, P > 0.05), where the qPCR assay produced lower GCNs than the ddPCR assay for 11 of 17 samples tested. Of these 11 samples, 9 had a 10-fold-higher MIC cpcB GCN than CYL rpoC1 GCN, suggesting that a competitive effect was present in these samples in analyses performed with duplex qPCR.
FIG 5
DISCUSSION
Freshwater cyanobacterial blooms are of major concern globally due to the natural toxin and off-flavor compounds produced by several genera, causing esthetical and public health issues. Microcystis and Cylindrospermopsis are two of the most widespread and successful bloom-forming genera across continents (27, 28). Many previous investigations have focused on a single dominant genus during bloom events. However, cooccurrence of these genera has also been reported, with changes in dominance affected by nutrients, light, season, and grazing effects (28–30). The impacts of Microcystis and Cylindrospermopsis blooms are not restricted only to microcystins and cylindrospermopsins but have also recently extended to include compounds such as microsin, microviridin, and β-N-methylamino-l-alanine (BMAA), known to cause adverse ecological and health effects (31–33). Thus, monitoring the total abundance of these genera is highly relevant from the perspective of water quality risk assessment and bloom management.
The 16S rRNA gene is by far the most frequently used molecular marker in microbial ecology studies (34). Assays based on the 16S rRNA gene have been designed to determine the total cyanobacterial community members or the members of a particular genus or species (8). However, issues such as inconsistent copy numbers of gene and heterogonous sequences in cyanobacteria make translation from 16S rRNA GCN to cell number difficult. To avoid quantification biases, we selected rpoC1 and cpcB genes as targets for which only a single gene copy is present in a genome. The primers and probes are specific to only the target genera, except for two sequences from Anabaena sphaerica var. tenuis that match CYL rpoC1 primers and probes. In fact, a phylogenetic study of four genes (the 16S rRNA gene and hetR, nifH, and rpoC1 genes) of Anabaena morphospecies led to the conclusion that A. sphaerica is phylogenetically closer to Cylindrospermopsis and Raphidiopsis and, thus, should be considered a sister group with Cylindrospermopsis instead of Anabaena (35), thus supporting our assay design.
The qPCR technique has become a popular method in the monitoring of algal blooms over the past decade. However, most of the published assays identify only single targets whereas only a few use multiplex techniques. The multiplex assays available in the literature focus on toxin-producing species. These assays were designed in such a way that they estimate the percentage of toxigenic species within a genus (4, 9, 36, 37) or quantify multiple toxin synthesis genes covering several toxicological groups (hepatotoxins and neurotoxins) (38). The assay developed in this study is the first duplex assay that enumerates abundances of two genera simultaneously, enabling fast and effective assessment of changes in subpopulations, interpopulation competition, and dynamics for an aquatic system.
This was the first study comparing qPCR and ddPCR as methods to quantify harmful cyanobacterial species in laboratory and freshwater systems. The designed primers and probes are suitable for duplex detection on both PCR platforms, with the assays achieving satisfactory specificity, sensitivity, and efficiency. Although both techniques required substantial optimization efforts to achieve the desired assay performance and sensitivity, once optimization was achieved, multiplexing reduced the number of reactions required to generate data sets as well as sample requirements and costs of reagents and other laboratory consumables. The optimized qPCR assay developed in this study has a larger quantification dynamic range and is more sensitive in detecting targets of low concentrations. While the ddPCR assay was lower in sensitivity, other criteria required for operation are comparable to or even surpass those of the qPCR assay.
ddPCR is known to provide several advantages over qPCR: it is less susceptible to PCR inhibition and high background DNA levels; it gives absolute quantification without relying on an external standard reference; and it is more precise than qPCR (20–22). Our findings are in agreement with those previous observations showing that the ddPCR method was indeed higher in precision for either plasmid standard or environmental sample analysis.
The accurate prediction of ratios of rpoC1 and cpcB gene copy numbers to 16S rRNA gene copy numbers in Cylindrospermopsis and Microcystis strains indicated that ddPCR is excellent in predicting gene copy number variations in organisms. Multiple gene copy numbers found in human eukaryotic organisms and bacteria have been shown to be associated with human diseases, a faster cellular response to resource availability, and a higher cell growth rate (39). Variation in 16S rRNA gene copy numbers for cyanobacterial genera has been described previously (40, 41). It is believed that gene copy number variation at a specific gene is a result of positive selection, contributing to phenotypic differences in adaptive traits, and it could be phylogenetically informative (39, 42). Therefore, ddPCR is useful in shedding light on natural selection by providing valuable information on copy number variation.
Optimization is required in multiplex assays due to competition between individual assays as well as an increased probability of unwanted cross-oligonucleotide interactions. Inhibition and an effect of competition between targets are major limitations for multiplex qPCR, especially for environmental samples, where the targets may be present in low quantities with a large amount of nontarget background DNA, leading to under- or overestimation of (43, 44) of target gene copy numbers. Quantification bias due to the competitive effect, which was evident in this study, needs to be addressed carefully, as many natural bloom samples are dominated by a single cyanobacterial species. Our results showed that the ddPCR technique was able to overcome this problem, providing accurate quantification for at least two targets with a 1,000-fold difference in concentration.
Nevertheless, the qPCR assay has several advantages in environmental sample analysis. Its higher sensitivity and larger quantification dynamic range allow more flexibility in evaluating environmental samples, where the target could be of a very low concentration (nonbloom/nondominant species) or a very high concentration (bloom occurrence). In addition, qPCR was generally the cheaper of the two techniques. Based on duplex assays developed in this study, the cost to analyze a full 96-well plate was $156.00 for qPCR (∼$1.60 per reaction) and $470.00 for ddPCR (∼$4.90 per reaction). The qPCR assay was also less laborious than the ddPCR assay. For 96 samples, the preparation time needed for qPCR was 45 min, the amplification time was 80 min, and the data processing time was 15 min; for ddPCR, the preparation time was 150 min, the amplification time was 100 min, and the plate reading time was 120 min. Therefore, for routine water resource monitoring and screening involving large numbers of samples and a quick operational response, the qPCR technique is more cost-effective and able to generate results more rapidly. ddPCR, however, is a better technique when accuracy and precision are of major importance or when PCR inhibition and competitive effects are likely, such as when analyzing bloom samples with a duplex qPCR assay.
Knowledge about molecular techniques and genetic understanding of cyanobacteria have increased tremendously over the past decade, and yet the application of molecular methods in the detection of cyanobacteria for water management and risk assessment is still in its infancy. The potential of molecular quantification assays to monitor cyanobacteria can be fully realized if these assays are robust and reliable and possess high efficiency. In this study, development of duplex assays based on the rpoC1 and cpcB genes shows promise to detect Cylindrospermopsis and Microcystis on two molecular quantification platforms, qPCR and ddPCR. To the best of our knowledge, this is the first duplex ddPCR assay developed to quantify cyanobacterial species abundance in natural environments.
ACKNOWLEDGMENTS
This research is supported by a grant from the Singapore National Research Foundation under its Environmental & Water Technologies Strategic Research Programme and administered by the Environment & Water Industry Programme Office (EWI) of the PUB (grant 1102-IRIS-14-02).
We thank Masaaki Kitajima at the Center for Environmental Sensing and Modeling (CENSAM) of the Singapore-MIT Alliance for Research and Technology (SMART) for his assistance in sample analysis and instrument access and Nur Hazimah Binte Mohamed Nor at NUS Environmental Research Institute (NERI) for conducting cell counting.
Supplemental Material
File (zam999116444so1.pdf)
- Download
- 40.84 KB
REFERENCES
1.
Paerl HW, Fulton RS, Moisander PH, Dyble J. 2001. Harmful freshwater algal blooms, with an emphasis on cyanobacteria. ScientificWorldJournal 1:76–113.
2.
Graham JL, Loftin KA, Ziegler AC, Meyer MT. 2008. Guidelines for design and sampling for cyanobacterial toxin and taste-and-odor studies in lakes and reservoirs. US Department of the Interior, US Geological Survey, Reston, VA.
3.
Ouellette AJ, Wilhelm SW. 2003. Toxic cyanobacteria: the evolving molecular toolbox. Front Ecol Environ 1:359–366.
4.
Rasmussen J, Giglio S, Monis P, Campbell R, Saint C. 2008. Development and field testing of a real-time PCR assay for cylindrospermopsin-producing cyanobacteria. J Appl Microbiol 104:1503–1515.
5.
Baker JA, Entsch B, Neilan BA, McKay DB. 2002. Monitoring changing toxigenicity of a cyanobacterial bloom by molecular methods. Appl Environ Microbiol 68:6070–6076.
6.
Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL. 2009. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem 55:611–622.
7.
Kurmayer R, Christiansen G. 2009. The genetic basis of toxin production in cyanobacteria. Freshw Rev 2:31–50.
8.
Martins A, Vasconcelos V. 2011. Use of qPCR for the study of hepatotoxic cyanobacteria population dynamics. Arch Microbiol 193:615–627.
9.
Rinta-Kanto J, Ouellette A, Boyer G, Twiss M, Bridgeman T, Wilhelm S. 2005. Quantification of toxic Microcystis spp. during the 2003 and 2004 blooms in western Lake Erie using quantitative real-time PCR. Environ Sci Technol 39:4198–4205.
10.
Kurmayer R, Kutzenberger T. 2003. Application of real-time PCR for quantification of microcystin genotypes in a population of the toxic cyanobacterium Microcystis sp. Appl Environ Microbiol 69:6723–6730.
11.
Vaitomaa J, Rantala A, Halinen K, Rouhiainen L, Tallberg P, Mokelke L, Sivonen K. 2003. Quantitative real-time PCR for determination of microcystin synthetase E copy numbers for Microcystis and Anabaena in lakes. Appl Environ Microbiol 69:7289–7297.
12.
Marbun YR, Yen H-K, Lin T-F, Lin H-L, Michinaka A. 2012. Rapid on-site monitoring of cylindrospermopsin-producers in reservoirs using quantitative PCR. Sustain Environ Res 22:143–151.
13.
Rasmussen J, Barbez P, Burgoyne L, Saint C. 2008. Rapid preparation of cyanobacterial DNA for real-time PCR analysis. Lett Appl Microbiol 46:14–19.
14.
Koskenniemi K, Lyra C, Rajaniemi-Wacklin P, Jokela J, Sivonen K. 2007. Quantitative real-time PCR detection of toxic Nodularia cyanobacteria in the Baltic Sea. Appl Environ Microbiol 73:2173–2179.
15.
Huggett JF, Foy CA, Benes V, Emslie K, Garson JA, Haynes R, Hellemans J, Kubista M, Mueller RD, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wittwer CT, Bustin SA. 2013. The digital MIQE guidelines: minimum information for publication of quantitative digital PCR experiments. Clin Chem 59:892–902.
16.
Hindson CM, Chevillet JR, Briggs HA, Gallichotte EN, Ruf IK, Hindson BJ, Vessella RL, Tewari M. 2013. Absolute quantification by droplet digital PCR versus analog real-time PCR. Nat Methods 10:1003–1005.
17.
Laird PW. 2010. Principles and challenges of genome-wide DNA methylation analysis. Nat Rev Genet 11:191–203.
18.
Day E, Dear PH, McCaughan F. 2013. Digital PCR strategies in the development and analysis of molecular biomarkers for personalized medicine. Methods 59:101–107.
19.
Cai Y, Li X, Lv R, Yang J, Li J, He Y, Pan L. 2014. Quantitative analysis of pork and chicken products by droplet digital PCR. Biomed Res Int 2014:810209.
20.
Hayden RT, Gu Z, Ingersoll J, Abdul-Ali D, Shi L, Pounds S, Caliendo AM. 2013. Comparison of droplet digital PCR to real-time PCR for quantitative detection of cytomegalovirus. J Clin Microbiol 51:540–546.
21.
Morisset D, Štebih D, Milavec M, Gruden K, Žel J. 2013. Quantitative analysis of food and feed samples with droplet digital PCR. PLoS One 8:e62583.
22.
Strain MC, Lada SM, Luong T, Rought SE, Gianella S, Terry VH, Spina CA, Woelk CH, Richman DD. 2013. Highly precise measurement of HIV DNA by droplet digital PCR. PLoS One 8:e55943.
23.
Pohl G, Shih I-M. 2004. Principle and applications of digital PCR. Expert Rev Mol Diagn 4:41–47.
24.
NCCLS. 2004. Protocols for determination of limits of detection and limits of quantitation; approved guideline. National Committee for Clinical Laboratory Standards, Wayne, PA.
25.
Throndsen J. 1978. Preservation and storage, p 69–74. In Sournia A (ed), Monographs on oceanographic methodology, no. 6. Phytoplankton manual. UNESCO, Paris, France.
26.
Rinta-Kanto JM, Konopko EA, DeBruyn JM, Bourbonniere RA, Boyer GL, Wilhelm SW. 2009. Lake Erie Microcystis: relationship between microcystin production, dynamics of genotypes and environmental parameters in a large lake. Harmful Algae 8:665–673.
27.
Marinho M, Souza M, Lürling M. 2013. Light and phosphate competition between Cylindrospermopsis raciborskii and Microcystis aeruginosa is strain dependent. Microb Ecol 66:479–488.
28.
Moustaka-Gouni M, Vardaka E, Tryfon E. 2007. Phytoplankton species succession in a shallow Mediterranean lake (L. Kastoria, Greece): steady-state dominance of Limnothrix redekei, Microcystis aeruginosa and Cylindrospermopsis raciborskii. Hydrobiologia 575:129–140.
29.
Costa IAS, Azevedo SMFO, Senna PAC, Bernardo RR, Costa SM, Chellappa NT. 2006. Occurrence of toxin-producing cyanobacteria blooms in a Brazilian semiarid reservoir. Braz J Biol 66:211–219.
30.
Soares M, Rocha M, Marinho M, Azevedo S, Branco C, Huszar V. 2009. Changes in species composition during annual cyanobacterial dominance in a tropical reservoir: physical factors, nutrients and grazing effects. Aquat Microb Ecol 57:137–149.
31.
Rohrlack T, Christoffersen K, Hansen P, Zhang W, Czarnecki O, Henning M, Fastner J, Erhard M, Neilan B, Kaebernick M. 2003. Isolation, characterization, and quantitative analysis of microviridin J, a new microcystis metabolite toxic to Daphnia. J Chem Ecol 29:1757–1770.
32.
Wiegand C, Peuthert A, Pflugmacher S, Carmeli S. 2002. Effects of microcin SF608 and microcystin-LR, two cyanotobacterial compounds produced by Microcystis sp., on aquatic organisms. Environ Toxicol 17:400–106.
33.
Cox PA, Banack SA, Murch SJ, Rasmussen U, Tien G, Bidigare RR, Metcalf JS, Morrison LF, Codd GA, Bergman B. 2005. Diverse taxa of cyanobacteria produce β-N-methylamino-l-alanine, a neurotoxic amino acid. Proc Natl Acad Sci U S A 102:5074–5078.
34.
Case RJ, Boucher Y, Dahllöf I, Holmström C, Doolittle WF, Kjelleberg S. 2007. Use of 16S rRNA and rpoB genes as molecular markers for microbial ecology studies. Appl Environ Microbiol 73:278–288.
35.
Thomazeau S, Houdan-Fourmont A, Couté A, Duval C, Couloux A, Rousseau F, Bernard C. 2010. The contribution of sub-Saharan African strains to the phylogeny of cyanobacteria: focusing on the Nostocaceae (Nostocales, cyanobacteria). J Phycol 46:564–579.
36.
Briand E, Escoffier N, Straub C, Sabart M, Quiblier C, Humbert J-F. 2009. Spatiotemporal changes in the genetic diversity of a bloom-forming Microcystis aeruginosa (cyanobacteria) population. ISME J 3:419–429.
37.
Briand E, Gugger M, Francois J-C, Bernard C, Humbert J-F, Quiblier C. 2008. Temporal variations in the dynamics of potentially microcystin-producing strains in a bloom-forming Planktothrix agardhii (cyanobacterium) population. Appl Environ Microbiol 74:3839–3848.
38.
Al-Tebrineh J, Pearson LA, Yasar SA, Neilan BA. 2012. A multiplex qPCR targeting hepato- and neurotoxigenic cyanobacteria of global significance. Harmful Algae 15:19–25.
39.
Schirrmeister B, Dalquen D, Anisimova M, Bagheri H. 2012. Gene copy number variation and its significance in cyanobacterial phylogeny. BMC Microbiol 12:177.
40.
Engene N, Cameron Coates R, Gerwick WH. 2010. 16S rRNA gene henerogeneity in the filamentous marine cyanobacterial genus Lyngbya. J Phycol 46:591–601.
41.
Engene N, Gerwick WH. 2011. Intra-genomic 16S rRNA gene heterogeneity in cyanobacterial genomes. Fottea (Praha) 11:17–24.
42.
Schrider DR, Hahn MW. 2010. Gene copy-number polymorphism in nature. Proc Biol Sci 277:3213–3221.
43.
Cook RF, Cook SJ, Li F, Montelaro RC, Issel CJ. 2002. Development of a multiplex real-time reverse transcriptase-polymerase chain reaction for equine infectious anemia virus (EIAV). J Virol Methods 105:171–179.
44.
Stals A, Baert L, Botteldoorn N, Werbrouck H, Herman L, Uyttendaele M, Van Coillie E. 2009. Multiplex real-time RT-PCR for simultaneous detection of GI/GII noroviruses and murine norovirus 1. J Virol Methods 161:247–253.
45.
Bolch C, Blackburn S. 1996. Isolation and purification of Australian isolates of the toxic cyanobacterium Microcystis aeruginosa Kütz. J Appl Phycol 8:5–13.
Information & Contributors
Information
Published In
Applied and Environmental Microbiology
Volume 81 • Number 15 • 1 August 2015
Pages: 5203 - 5211
Editor: D. W. Schaffner
PubMed: 26025892
Copyright
© 2015 American Society for Microbiology.
History
Received: 22 March 2015
Accepted: 18 May 2015
Published online: 7 July 2015
Contributors
Editor
D. W. Schaffner
Editor
Metrics & Citations
Metrics
Note: There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
Citation counts come from the Crossref Cited by service.
Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.