|Home | About | Journals | Submit | Contact Us | Français|
The chip and quantitative real-time PCR (qPCR) assays were optimized to study the expression of microcystin biosynthesis genes (mcy) with RNA samples extracted from cyanobacterial strains and environmental water samples. Both microcystin-producing Anabaena and Microcystis were identified in Lake Tuusulanjärvi samples. Microcystis transcribed the mcyE genes throughout the summer of 2006, while expression by Anabaena became evident later in August and September. Active mcyE gene expression was also detectable when microcystin concentrations were very low. Detection of Anabaena mcyE transcripts by qPCR, as well as certain cyanobacterial 16S rRNAs with the chip assay, showed slightly reduced sensitivity compared with the DNA analyses. In contrast, even groups undetectable or present in low quantities as determined by microscopy could be identified with the chip assay from DNA samples. The methods introduced add to the previously scarce repertoire of applications for mcy expression profiling in environmental samples and enable in situ studies of regulation of microcystin synthesis in response to environmental factors.
Toxic mass occurrences (blooms) of cyanobacteria are common worldwide. They pose a serious health risk to humans and have caused a number of animal poisonings (34, 38). In freshwaters, hepatotoxic blooms occur more often than neurotoxic blooms (39). Hepatotoxicity is caused by microcystins, cyclic peptides consisting of seven amino acids. Microcystis spp., Anabaena spp., and Planktothrix spp. are the main microcystin producers in freshwater environments (39). In cyanobacterial blooms, microcystin producers of different genera often cooccur (32, 45). In addition, blooms can contain both microcystin-producing and -nonproducing strains of the same genus (39, 45).
Microcystin-producing strains cannot be discriminated from nontoxic strains by traditional microscopy, since they are similar in appearance (39). The difference between toxic and nontoxic strains is the presence of microcystin synthetase genes (mcy) in the genome of toxic strains. These biosynthetic genes encode nonribosomal peptide synthetase, polyketide synthase, and tailoring enzymes that in turn construct the toxin molecules (4, 7, 26, 27, 35, 41). The mcy genes are currently used widely to detect, differentiate, and identify the potential microcystin-producing cyanobacteria in environmental samples with various molecular methods (29, 37). The most commonly used method is conventional PCR, which is a robust method for mcy gene detection either with genus-specific (e.g., see references 20, 32, 42, and 44) or general primers (10, 14, 16, 25, 31). However, when general primers are used, a post-PCR method is required for identification of the producers. The post-PCR methods used include restriction enzyme digestion of PCR amplicons (14), sequencing (16), denaturing gradient gel electrophoresis (10), and a DNA chip (33). The DNA chip can be used to simultaneously detect hepatotoxin producers of the genera Anabaena, Microcystis, Planktothrix, Nodularia, and Nostoc, as well as 19 cyanobacterial groups with probes for the mcyE (33) and 16S rRNA (3) genes. The DNA chip thus gives a view of the cyanobacterial community and identifies the potential microcystin producers present at the same time. In comparison to conventional PCR, which can be used to detect the presence of mcy genes, quantitative real-time PCR (qPCR) makes it possible to quantify the microcystin producers and also to reveal the dominant toxin producer, if several genera cooccur, by determining the mcy copy numbers (37).
Most molecular detection methods use DNA as the target (29, 37). However, DNA-based methods can reveal only the presence of genes involved in microcystin production. In environmental samples, DNA can originate from dead as well as living cells. Additionally, mcy genes can be mutated, preventing or diminishing their transcription and subsequent toxin production (5, 6, 12, 17, 20, 23, 42). Use of RNA, instead, allows detection of potential microcystin-producing cyanobacteria that are alive and actively transcribing the mcy genes. Consequently, observation of actual toxin producers is more reliable. However, few studies are available in which RNA was used for identification of microcystin producers in natural environments by PCR (8, 11). Expression of biosynthesis genes of nodularin, a toxin resembling microcystin but produced solely by Nodularia spumigena, was detected by qPCR in Baltic Sea samples (15).
Our aim was to develop molecular detection methods for metabolically active cyanobacteria, with an emphasis on identifying the active microcystin producers. We chose the mcyE gene, since this region can reliably detect various microcystin producers, as shown by previous studies (31-33, 44). A previously designed DNA chip assay (3, 33) was optimized for use with RNA extracted from cyanobacterial strains and environmental water samples (Lake Tuusulanjärvi). In addition, Anabaena- and Microcystis-mcyE-targeted qPCR assays were designed to verify the chip results.
DNA and RNA were extracted from axenic hepatotoxin-producing Anabaena sp. 90 and Microcystis aeruginosa PCC 7806 cultures. The strains were grown in Z8 medium (19) for 21 days at 20 to 25°C in continuous light, 5 to 6 μmol m−2 s−1, and the cells were collected on 5-μm-pore-size polycarbonate filters (Poretics; GE Osmonics Labstore). In addition, for qPCR optimization, DNA from 15 Microcystis, 13 Anabaena, two Planktothrix, two Nostoc, and two Nodularia strains was used (see Table S1 in the supplemental material).
A total of five water samples were collected from Lake Tuusulanjärvi from June to September 2006. Composite water samples from depths of 0 to 2 m were taken with a Ruttner sampler (Limnos Ltd., Turku, Finland), and 100 to 250 ml water was filtered through 5-μm- and 1-μm-pore-size polycarbonate filters (Poretics, GE Osmonics Labstore). Samples were collected for the DNA, RNA, and microcystin extractions. The water samples for RNA extraction were filtered at the lakeshore immediately after sampling and frozen in liquid nitrogen on site, while the water samples for DNA and microcystin extraction were transported to the laboratory before filtration. The DNA and microcystin samples were stored at −20°C and RNA samples at −70°C until extraction.
For microscopic analysis, a 200-ml composite water sample was preserved with acid Lugol's solution. The Utermöhl technique (28, 43) was used for counting phytoplankton cells and cyanobacterial colonies (subsection I, Chroococcales) or filaments (subsection IV, Nostocales; subsection III, Oscillatoriales). The biomasses (fresh weight) were estimated, using species-specific conversion values.
The DNAs from the strains and environmental samples were extracted by bead beating and the cetyltrimethylammonium bromide (CTAB) method, as described earlier by Kolmonen et al. (18), and purified further with a Gene Clean Turbo kit according to the manufacturer's instructions (Q-Biogene). The quantity and quality of the extracted DNA were measured with a UV spectrophotometer (Biophotometer; Eppendorf).
For RNA extraction, the filters were transferred to FastPrep Lysing matrix E tubes (Q-Biogene) and the cells were disrupted mechanically in 1 ml PMR1 solution (MO BIO Laboratories Inc.) with a FastPrep FP120 bead beater (Thermo Electron Corporation) for 40 s at a speed of 5 m s−1. RNA isolation was then performed, using an Ultraclean Plant RNA kit, as instructed by the manufacturer (Mo Bio Laboratories). For the environmental RNA samples, two filters were combined to increase the total RNA amount. The extracted RNA was treated twice with DNase I according to the manufacturer's instructions (Promega). The DNase-treated RNA was phenol-chloroform extracted to inactivate and remove the enzyme, ethanol precipitated, and diluted to a 50-μl final volume with water. The quantity and quality of the RNA were measured with a NanoDrop-1000 spectrophotometer (Nanodrop Technologies). The mcyE qPCR was used to verify that the RNA samples were DNA free. The RNA was reverse transcribed to cDNA, using 5 to 15 μl of RNA as a template, with an iScript cDNA synthesis kit (Bio-Rad) according to the manufacturer's instructions.
Four different glass types and printing solutions were tested to determine the best combination (see note S1 and Fig. S2 in the supplemental material). To print the slides for the experiments, Nexterion slide A (Schott Nexterion) and Micro Spotting Solution Plus (MSP) buffer (TeleChem International) were selected. The ZipCodes were printed on the arrays according to the scheme used in the previous version of the chip (3, 33). All the slides were printed in the Biomedicum Biochip Center, University of Helsinki, Helsinki, Finland.
Prior to the ligation detection reaction (LDR), the target sequences of 16S rRNA and mcyE were amplified from the DNA and cDNA samples. The general primers (mcyE-F2/R4) that detect all the main microcystin-producing genera (31, 44) were used to amplify the mcyE sequences, as described by Rantala et al. (33). All the reactions were done in a 20-μl final reaction volume of 1× Super Taq plus buffer (HT Biotechnology) with 1 U of Super Taq plus polymerase (HT Biotechnology Ltd.), 0.5 μM primers (Sigma-Genosys), 250 μM deoxynucleoside triphosphates (dNTPs) (Finnzymes), and 1 μl of DNA or cDNA. Some of the cDNA samples were reamplified, using 1 to 3 μl of the first-round PCR product as a template. For DNA amplifications, 1.25 μg μl−1 bovine serum albumin (Promega) was added to the reaction mixture. The PCR conditions were as follows: 95°C for 3 min, 35 × (94°C for 30 s, 56°C for 30 s, and 68°C for 1 min), and 68°C for 10 min.
The 16S rRNA and internal transcribed spacer (ITS) region were amplified from the DNA samples as previously detailed by Castiglioni et al. (3). The PCR was done with 0.5 μM primers (Sigma-Genosys) pA (9) and 23S30R (21) in 1× Super Taq plus buffer (HT Biotechnology) with 1 U of Super Taq plus polymerase (HT Biotechnology), 200 μM dNTPs (Finnzymes), and 1 μl of DNA sample in a 50-μl reaction volume. The PCR program was as follows: 94°C for 5 min; 10 × (94°C for 45 s, 57°C for 45 s, and 68°C for 2 min), 25 × (92°C for 45 s, 54°C for 45 s, and 68°C for 2 min), and 68°C for 20 min. A different reverse primer, pH• (9), was used for amplification of the cDNA samples, because the ITS region is not present in 16S rRNA transcripts. The PCRs were otherwise the same as with the DNA samples, except that 220 μM concentrations of the dNTPs were used. The PCR conditions were as follows: 94°C for 5 min, 30 × (94°C for 1 min, 58°C for 1 min, and 68°C for 1 min), and 68°C for 10 min. All PCR products were purified with a GFX PCR purification kit (GE Healthcare) according to the manufacturer's instructions. The size and quantity of the amplicons were evaluated with 1.5% agarose gel electrophoresis and a UV spectrophotometer (Biophotometer, Eppendorf), respectively.
The LDR step of the chip assay (see Fig. S1 in the supplemental material) was carried out as described by Rantala et al. (33). The oligonucleotide mix contained 250 fmol μl−1 of each common and discriminating probe for the cyanobacterial 16S rRNA (3) and mcyE sequences (33). The LDR was performed in a final volume of 20 μl containing 1× Pfu DNA ligase buffer (Stratagene), 1 μl of oligonucleotide mix, and 25 fmol of both 16S rRNA and mcyE PCR products. The reaction mixture was preheated for 2 min at 94°C and centrifuged for 1 min prior to addition of 4 U Pfu ligase (Stratagene). The LDR included 30 cycles of 94°C for 30 s and 63°C for 4 min in the iCycler (Bio-Rad) thermal cycler.
Hybridization of the slides was carried out as detailed by Rantala et al. (33). The slides were first prehybridized in 5× saline sodium citrate buffer (SSC) (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-1% bovine serum albumin (BSA) solution for 1 h at 42°C and rinsed with water five times for 30 s (each). The hybridization mixture contained the entire LDR mixture, 5× SSC, 1 mg ml−1 of salmon testes DNA (Sigma-Aldrich), and 10 fmol of hybridization control oligonucleotide in a total volume of 65 μl. The entire hybridization mixture was placed on a glass slide, divided into eight hybridization chambers with a Press-To-Seal silicone isolator (1.0 by 9 mm; Schleicher and Schuell BioScience). The slides were hybridized for 2 h in a dark 65°C water bath and washed with preheated 1× SSC-0.1% sodium dodecyl sulfate (SDS) for 15 min at 65°C, with 0.1× SSC for 5 min at room temperature, and finally with water three times for 5 min (each) at room temperature.
The fluorescent signals were acquired at 5-μm resolution with a GenePix 4200AL laser scanner (Molecular Devices) at a wavelength of 532 nm. The signal intensities of the fluorescent spots and background were analyzed with GenePix Pro 5.1 microarray image analysis software (Molecular Devices). Detection of the signals and spots and thus the presence of 16S rRNA or mcyE sequences in the samples was evaluated according to criteria described previously (24), with minor modifications. For positive detection, the following conditions should be fulfilled: (i) signal-to-noise ratio (SNR) of the spot should be ≥3, (ii) ≥70% of the spot's signal intensity should be two standard deviations higher than the background signal intensity, and (III) the first and second conditions should be met by at least two of the four replicated spots. The signal intensity was calculated by subtracting the mean background signal intensity from the mean signal intensity of the spot. The average signal intensities of the group-specific spots (four replicates) were presented as percentages of the average signal intensity of the hybridization control spots (eight replicates) to normalize variation introduced by different chip assays.
Primer pairs and hydrolysis probes specific for the Anabaena and Microcystis mcyE genes were designed, and the PCR conditions were optimized (Table (Table1;1; see also note S2, Fig. S4, and Table S1 in the supplemental material). The qPCR was used to detect the mcyE sequence copy numbers in Lake Tuusulanjärvi DNA (5-μm filters) and RNA samples (5-μm and 1-μm filters). In addition, it was used to detect potential DNA contamination of the RNA samples prior to reverse transcription to cDNA. Quantification of the copy numbers was performed with external standards. Genomic DNA from Anabaena sp. 90, Anabaena sp. 202A1, or Microcystis strain PCC 7806 was used to prepare a dilution series containing 101, 102, 103, 104, 105, and 106 copies of the mcyE genes, as described by Vaitomaa et al. (44). As templates, 5 μl of 1:100-diluted DNA or DNase-treated RNA samples or 1.5 μl of cDNA samples were used. The reactions included 1× PCR mix, a 300 nM concentration of both primers, and a 200 nM concentration of the hydrolysis probe (5′ labeled with 6-carboxyfluorescein (FAM) and 3′ labeled with 6-carboxytetramethylrhodamine (TAMRA). The PCRs were carried out with an ABI7000 or ABI7300 instrument (Applied Biosystems). The PCR protocol consisted of preincubation steps at 50°C for 2 min and 95°C for 10 min, followed by 40 amplification cycles of denaturation at 95°C for 30 s and annealing at 62°C for 60 s. The amount of fluorescence was measured at each cycle after elongation. All reactions were carried out in three replicates. The samples were under the detection limit if their cycle threshold (CT) values were beneath the CT value of the most diluted standard (101 copies).
Analysis of the microcystins was carried out, using high-performance liquid chromatography (HPLC) combined with a diode array detector and mass spectrophotometer (MS) to determine the microcystin concentration in Lake Tuusulanjärvi samples. The extraction and analysis of microcystins from filtered cells were performed as described in detail by Halinen et al. (13), with the following exceptions. The microcystins were extracted directly from the frozen cells in 70% methanol by first mixing the extraction tubes and then heating the closed tubes for 30 min in an 80°C air incubator. Mixing and heating were repeated once. The samples were then centrifuged for 5 min at 10,000 × g, and the extracts were analyzed with liquid chromatography/mass spectrometry (LC/MS). The retention times of the microcystin standards and product ion spectra from positive ionization were used to identify the microcystins. Negative [M-H]− ion signals were used to quantify microcystins. Of the standards, microcystin-LR was obtained from the Faculty of Chemistry, University of Gdansk, Gdansk, Poland, and commercial microcystin-RR was obtained from Alexis Corp., Switzerland.
The performance of the chip assay with RNA samples was first assessed, using DNA and RNA (as cDNA) extracted from the microcystin-producing strains Anabaena sp. 90 and Microcystis aeruginosa PCC 7806 and their mixture as targets in the assay. In general, the same signals were detected with the 16S rRNA and mcyE probes with either nucleic acid type (Fig. (Fig.1).1). In addition, the signal intensities were clearly above the detection limit, due to the low background signal levels (see Fig. S3 in the supplemental material), and allowed unambiguous recognition of real spots. No signal was found from the negative-control spots. Faint signals were also identified with the 16S rRNA probes for Gloeothece when Microcystis DNA was included in the target (Fig. (Fig.1).1). These signals most probably arose from nonspecific hybridization, since this genus was not among the strains studied.
Next, to evaluate the performance of the chip assay with environmental samples, DNA and RNA (as cDNA) extracted from Lake Tuusulanjärvi water samples were targeted. Probes for Microcystis, Anabaena/Aphanizomenon, Woronichinia, and Gloeothece 16S rRNA genes gave signals with all the DNA samples (Fig. (Fig.22 A). In addition, Synechococcus/Prochlorococcus 16S rRNA genes were identified from water samples collected in June, late July, and August. In the RNA samples, Microcystis 16S rRNA was recognized at every sampling date and Anabaena/Aphanizomenon 16S rRNA was recognized in early July, August, and September (Fig. (Fig.2A).2A). Signals from Woronichinia, Gloeothece, and Synechococcus/Prochlorococcus 16S rRNA spots were absent from the RNA samples. Signals with probes targeted to the 16S rRNA of all cyanobacteria (universal probes) were detected with all the DNA and RNA samples.
Microcystis mcyE signals were detected in all the water samples except the early July RNA sample (Fig. (Fig.2B).2B). Signals from the Anabaena mcyE spots were present in both DNA and RNA of the August and September samples (Fig. (Fig.2B).2B). In all, the results showed that the chip assay is well suited for studying RNA samples coming from cultured strains, as well as complex environmental samples, and thus has the potential for more-reliable identification of metabolically active cyanobacterial groups and microcystin producers than is achieved with the DNA samples.
The microcystin concentrations were measured from the Lake Tuusulanjärvi water samples by LC/MS. The microcystins were detected in all samples, and the total concentration varied from 18 ng liter−1 to 532 ng liter−1 (Fig. (Fig.3).3). Demethylated microcystin-RR was the most common variant and was identified in every sample. In September, it constituted more than 50% of the total microcystin amount. The other main variants detected were microcystin-RR, microcystin-LR, and demethylated microcystin-LR (Fig. (Fig.3).3). In addition, microcystin-YR and -WR were present in small amounts (Fig. (Fig.33).
Anabaena and Microcystis mcyE-targeted qPCR was used to assess the presence (DNA) and expression (RNA) of the mcyE genes in the Lake Tuusulanjärvi samples. Microcystis mcyE was detected in every sample except the early July RNA sample (Fig. (Fig.4).4). Anabaena mcyE was found in all DNA samples except the one collected in June. However, expression of Anabaena mcyE was detected only later in the August and September RNA samples (Fig. (Fig.4).4). The mcyE sequences were detected only in the DNA and RNA samples extracted from 5-μm filters, whereas no mcyE sequences were amplified in samples extracted from 1-μm filters. The qPCR results showed that the Microcystis mcyE gene was present and was expressed earlier in the summer than the Anabaena mcyE gene. The highest copy numbers of the mcyE genes coincided with the highest microcystin concentrations in August and September. Similar trends were seen in the mcyE expression levels, although the quantitation was based on the DNA standard curve. In contrast, mcyE gene expression was not observed at all in early July, when the microcystin concentration was lowest (Fig. (Fig.44).
The qPCR results were also used for verifying the mcyE probe signals detected by the chip assay. In general, the two molecular methods performed well and were highly consistent (Fig. (Fig.2B2B and and4).4). However, Anabaena mcyE was not detected by the chip assay in the two July DNA samples, while it was detected by qPCR (Fig. (Fig.2B2B and and4).4). This was most probably due to the more sensitive detection of Anabaena mcyE by the genus-specific qPCR primers than by the general mcyE primer pair used in the chip assay. Specific and sensitive performance of the qPCR assays was also shown by the nearly unaltered amplification of the target genes when large amounts of competing DNA were present (see Fig. S4 in the supplemental material). No statistically significant variation was found (P = 0.109 to 1.000) in the mean CT levels between the unspiked and spiked reactions.
The signals detected from the 16S rRNA spots of the chip assay were compared with microscopic analyses of the water samples to assess which of the cyanobacterial groups could be detected by the chip assay. Microscopy showed that all the cyanobacteria (24 to 36 morphotypes) belonged to subsections I (order Chroococcales in the botanical system), III (order Oscillatoriales), and IV (order Nostocales) (Table (Table2)2) (2). Subsection I was the major group in almost every sampling. However, the numbers of the genera Microcystis and Woronichinia, for which probes were available in the chip assay, were very low (Table (Table2).2). The other cyanobacteria of subsection I could not be identified to the genus level or represented genera for which there are no probes designed (data not shown). Subsection III was represented almost exclusively by the genus Pseudanabaena (data not shown), which has no corresponding probes in the chip assay. In contrast, the Anabaena/Aphanizomenon probe pair represented the whole of subsection IV, since only these two genera were observed by microscopy.
Microcystis was detected in every DNA and RNA sample by the chip assay (Fig. (Fig.2A)2A) and by microscopy in all but the June sample (Table (Table2).2). Since colonies are counted in microscopic analysis, Microcystis may still have been present in June, but as single cells. Similarly, Woronichinia colonies were present in the June, late July, and September samples as determined by microscopy (Table (Table2)2) but were detected in all DNA samples by the chip assay (Fig. (Fig.2A).2A). However, Woronichinia signals were not detected in any of the RNA samples. Signals from Synechococcus/Prochlorococcus and Gloeothece probe pairs were also detected in the DNA samples by the chip assay (Fig. (Fig.2A),2A), but they were not observed by microscopy. Identification of Synechococcus cells is difficult, due to their small size, and Gloeothece signals could have been caused by nonspecific hybridization, as suggested by the assays performed with axenic cyanobacterial strains (Fig. (Fig.1).1). Genera belonging to subsection IV were observed in each sample by microscopy (Table (Table2),2), and signals with the corresponding Anabaena/Aphanizomenon probes were detected in all five DNA samples and three of the RNA samples by the chip assay (Fig. (Fig.2A).2A). The missing signals of the two RNA samples (June and late July) could have been caused by the lower cell numbers for Anabaena and Aphanizomenon in these samples (Table (Table22).
We applied the DNA chip and qPCR methods to assess the activity of potential microcystin-producing cyanobacteria in lake samples. The results showed that all the expected signals were found, using RNA and DNA extracted from microcystin-producing cyanobacterial strains by the new chip assay. In environmental samples, detection was highly consistent with both nucleic acid types, as well. The only discrepancy was that Anabaena mcyE was not detected in the two July DNA samples by the chip assay. This was most likely caused by the preferential amplification of Microcystis mcyE by the general primer pairs used in the chip assay, while it was avoided by the genus-specific primers and probes used in qPCR. The same phenomenon was observed in our previous study; only Microcystis mcyE was detected by the DNA chip assay, although the Anabaena and Planktothrix mcyE genes were also present in the sample in low quantities according to SYBR green-based qPCR (33). The qPCR assays designed in this study were specific and resistant to competing DNA of other microcystin producers (see Fig. S4 in the supplemental material). Thus, the amplification of the Anabaena mcyE gene in qPCR very likely reflected the real presence of potential microcystin-producing Anabaena in the lake already in July. Both Microcystis and Anabaena commonly occur in Lake Tuusulanjärvi (22, 30). Previously both genera, as determined by detection of the mcyE genes, were also suspected of including potential microcystin-producing strains (32, 33, 44). Our present results confirmed that both Microcystis and Anabaena actively transcribed the microcystin biosynthesis genes in the lake during the summer of 2006. Microcystis expressed the mcyE genes earlier, while Anabaena became more abundant and active later. This is in accordance with a 2-year survey of cyanobacterial community composition in Lake Tuusulanjärvi, where Anabaena and Aphanizomenon became more common in late summer and fall. The proliferation was related to long preceding periods of low inorganic nitrogen concentrations in the water, giving the advantage to these nitrogen-fixing cyanobacteria (30).
In general, both the chip and qPCR assays successfully detected the presence and expression of mcyE genes. Thus, they provide new tools for the scarce repertoire of methods for assessing the transcription of hepatotoxin biosynthesis genes in complex environmental samples. However, to accurately quantify transcripts, the qPCR method should be amended by a more appropriate standard system (1), using a relative quantitation method with an internal armored RNA standard (36) or with a suitable reference gene (1). Previously Anabaena- and Microcystis-specific reverse transcription (RT)-PCR assays were used to detect and identify the mcyE genes in ballast waters (8) and in bloom samples in Lake Agawam, New York (11). Both the present study and that by Gobler et al. (11) showed that Microcystis mcyE genes were expressed during the summer months. However, in Lake Agawam, toxic Anabaena was not detected. Previously the qPCR assay for the ndaF gene of Nodularia spumigena was used to study gene expression changes in response to nutrient levels in Baltic Sea samples (15). The chip assay, although not quantitative, can be used to simultaneously identify all the main microcystin and nodularin producers present, expediting the analysis of large sample sets. Although RNA is more demanding than DNA to sample, handle, and store under field conditions, it provides valuable additional information. While DNA samples target both living and dead cells, as well as free DNA, RNA samples focus the study on living cells that actively express the genes for microcystin production. Using RNA also prevents detection of cyanobacteria with inactive microcystin synthesis genotypes, i.e., those strains that have the biosynthesis genes but produce no microcystins. Inactive genotypes are relatively common among Planktothrix strains (5, 20) and are also encountered among Microcystis (23, 42, 46) and Baltic Sea Anabaena (12) strains. Expression of the mcy genes may still not be a definitive sign of microcystin production, since regulation at the translational level was suspected to be the reason for a loss of microcystin production in one Microcystis strain (23). However, it could prove to be beneficial in setting more accurate time frames for transcriptional changes and subsequently in making correlations with the environmental factors and microcystin levels detected.
The microcystin concentrations measured from the Lake Tuusulanjärvi water samples in the summer of 2006 were rather low, 18 to 532 ng liter−1, compared with results in previous studies including samples taken in the summers of 1999 and 2002, in which concentrations measured with enzyme-linked immunosorbent assay (ELISA) were above 1 μg liter−1 (32) and 8 μg liter−1 in the July samples (44). Accordingly, the amount of mcyE genes was also clearly lower than that in July 1999 (44). And yet, microcystin producers could be detected and identified with the molecular methods used, even in the June and early July samples, in which the microcystin concentrations were only slightly above the detection limit. However, in these samples, detection was more reliable with the DNA samples. The reduced sensitivity was also noted in identification of the cyanobacterial groups in the lake samples by the chip assay, using probes targeted to 16S rRNA, although rRNA is present in large quantities and is more stable than mRNA due to binding of ribosomal proteins. The extraction, purification, and DNase I treatment steps could also have resulted in an additional loss of sample RNA. A low RNA template concentration may reduce the efficiency of reverse transcription, leading subsequently to lower cDNA concentrations (40). The primers used for amplification of cDNA samples may have been more biased toward the abundant cDNA types present, leaving rarer cDNA molecules, e.g., those of Woronichinia and Synechococcus/Prochlorococcus, underrepresented.
The chip assay detected three to five cyanobacterial groups in the lake DNA samples. This result was in accordance with results in a previous study, in which other molecular methods, cloning and denaturing gradient gel electrophoresis, detected Anabaena/Aphanizomenon, Microcystis, and Synechococcus, the main cyanobacterial genera present in Lake Tuusulanjärvi during a 2-year period in 2000 to 2002 (30). The biomasses of cyanobacteria belonging to subsections I (Microcystis) and IV (Anabaena/Aphanizomenon) were also predominant in the present study. These genera were consistently identified with the chip assay. In addition, Woronichinia and Synechococcus were detected by the chip assay, although they could not be identified in all samples by microscopy. Accordingly, Anabaena/Aphanizomenon, Microcystis, and Woronichinia were recognized in a Lake Tuusulanjärvi sample by the previous version of the chip (3). In general, the chip assay detected those cyanobacterial groups that were in its scope, i.e., those for which probes have been designed. However, more probes should be added to the chip for it to serve as a real future alternative to microscopy.
In the present study, we showed that the chip and Anabaena- and Microcystis-specific qPCR assays were suitable for detecting mcyE transcripts in RNA extracted from axenic strains and from lake water samples. With both methods, Microcystis was identified as the main microcystin producer in Lake Tuusulanjärvi, while transcription of Anabaena mcyE genes was observed only in autumn. The methods developed here enrich the limited selection of methods previously available for studying expression of microcystin synthesis genes. In the future, quantitative analysis of mcyE expression could be used to relate transcriptional changes to changes in environmental factors possibly affecting microcystin synthesis. The chip assay instead offers a high-throughput platform for simultaneous assessment of the living cyanobacterial community and active microcystin and nodularin producers present.
This work was supported by grants from the Academy of Finland to A.R.-Y. (128480), to I.O. (213382), and to K.S. (Research Center of Excellence grants 53305 and 118637 and grant for Academy professors 214457).
We thank the Regional Environment Centre of Uusimaa (Leena Villa) for the water samples and Reija Jokipii and Maija Niemelä from the Finnish Environment Institute for the microscopic analysis. We are grateful to Matti Wahlsten for helping with the microcystin analysis and to Lyudmila Saari for culturing the cyanobacterial strains.
Published ahead of print on 16 April 2010.
†Supplemental material for this article may be found at http://aem.asm.org/.