Rapid detection of mcyG gene of microcystins producing cyanobacteria in water samples by recombinase polymerase amplification combined with lateral flow strips

In recent years, more and more aquatic ecosystems are suffering from cyanobacteria blooms in frequency, intensity, and duration, which was caused by eutrophication, rising CO₂ levels, and global warming (Huisman et al. 2018). Microcystins (MCs) are a class of toxic metabolites produced by MCs producing cyanobacteria, such as Microcystis, Nostoc, andAnabaena. Once MCs enter the water, they will threaten the health of aquatic animals and humans by accumulating in the food chain and polluting the drinking water sources. MCs can enter the human body through the digestive tract, causing damage to the liver and other organs, and also have potential carcinogenesis (Svircev et al. 2017). The area and frequency of MCs producing cyanobacteria blooms and the concentration of cyanobacteria in freshwater around the world are increasing year by year (He et al. 2016).

At present, the detection methods of MCs producing cyanobacteria are divided into two categories. One is to detect MCs, including high-performance liquid chromatography (HPLC) (Hemmati et al. 2019), liquid chromatography-mass spectrometry (LC-MS) (Palagama et al. 2018), high-performance capillary electrophoresis (Massey et al. 2020), enzyme-linked immunosorbent assay (ELISA) (Lu et al. 2020b), protein phosphatase inhibition assay, and biosensors (Moore et al. 2016). The other is to detect MC synthesis genes (e.g., mcyA, mcyB, mcyE, and mcyG), such as polymerase chain reaction (PCR) (Manali et al. 2019), quantitative PCR (qPCR; Otten et al. 2017), and loop-mediated isothermal amplification (LAMP) (Ramya et al. 2018). These methods can meet the need for highly sensitive detection of MCs producing cyanobacteria and have important reference significance for the detection of the MC synthesis genes. However, most of these methods have very high requirements for operating personnel and advanced detection equipment. Only ELISA and LAMP can be applied for the rapid detection of MCs producing cyanobacteria. However, ELISA is very sensitive, and the probability of false-positive results is high (Guo et al. 2017). The primer design of LAMP technology is complicated, and the reaction needs to be incubated at 65–67 °C for 30–60 min (Notomi et al. 2015). These shortcomings have limited the application of these two methods in the rapid detection of MCs producing cyanobacteria.

Recombinase polymerase amplification (RPA) is an isothermal amplification technology mediated by a recombinase, which is a highly selective and sensitive detection method. The amplification reaction can work at the temperature range of 37–42 °C in less than 20 min (Lobato & O'Sullivan 2018). This technology has no requirement for advanced equipment, and the reaction mixture can be placed in a simple constant temperature device to initialize the reaction. The amplification product analysis of RPA can be performed by gel electrophoresis, and it can also be visualized by combining the lateral flow (LF) strips with the labeled downstream primer and labeled probe (Boluk et al. 2020). RPA has been widely used in the rapid detection of bacteria (Gao et al. 2018), pathogens (Ahn et al. 2018), protozoans (Molina-Gonzalez et al. 2020), and viruses (Davi et al. 2019). The current study reported the development of a rapid method for the detection of MC synthesis gene mcyG, which was based on the RPA method combined with LF technology. This method can be potentially applied to detect MCs producing cyanobacterial blooms in the field samples.

In this study, DNA was extracted by the freeze-thaw cycles, and the specific process of this method was as follows. 1 mL of pure algae culture liquid was transferred to a 1.5 mL centrifuge tube, and cyanobacteria cells were collected by high-speed centrifugation (15,000 rpm/min, 5–10 min). 50 μL of Tris-EDTA (TE) buffer was added into the precipitate to resuspend the sediment. The resuspended algal cells were placed in liquid nitrogen or −80 °C freezer for 3–5 min and then incubated at 40 °C for 1–2 min. The above operations were repeated three to five times. The sample was centrifuged at 15,000 rpm/min for 10 min, and the supernatant was used as the DNA template for subsequent experiments.

McyG is responsible for the first step of Adda (3-amino-9-methoxy- 2,6,8-trimethyl-10-phenyl-4,6-decadienoic acid) synthesis, which is directly related to the toxicity of MCs (Hicks et al. 2006). The sequence of the mcyG gene is ubiquitous in the genome of MCs producing cyanobacteria, which is widely used in PCR or qPCR detection of MCs producing cyanobacteria (Lu et al. 2020a). The RPA primers (Supplementary Material, Table S1) were designed using the primer design software primer 3.0, based on the conserved sequence of mcyG (GenBank: AB444808.1, AY910575.1, and JQ290097.1). These primers were combined into 20 primer pairs (Supplementary Material, Table S2). The primer pair with the best amplification efficiency was selected from these 20 primer pairs throughout the RPA experiment.

The RPA experiment was performed using the TwistAmp Basic kit (TwistDX, UK). The RPA reaction mixture was prepared in a total volume of 50 μL, including 2.4 μL of 10 μM forward primer, 2.4 μL of 10 μM reverse primer, 29.5 μL of rehydration buffer containing the freeze-dried enzyme, 1 μL of DNA template, 12.2 μL of sterile distilled water, and 2.5 μL of 280 mM magnesium acetate. After centrifugation, the RPA reaction mixture was incubated at 37 °C for 30 min. The reaction product was purified using the SanPrep Column PCR Product Purification Kit (Sangon Biotech, China). After purification, 5 μL of the purified product was mixed with 1 μL of 6× loading buffer containing DNA dye and was separated in 2% agarose gel electrophoresis. The results were visualized using a gel imaging system (VILBER QUANTUM ST4, France).

A suitable LF-RPA probe was designed to match the selected primers used in the LF-RPA assay (Table 1). This assay was performed using the TwistAmp nfo kit (TwistDX, UK). The reaction mixture contained 2 μL of 10 μM forward primer, 2 μL of 10 μM reverse primer, 0.2–0.6 μL of 10 μM probe, 29.5 μL of rehydration buffer with the freeze-dried enzyme, 1 μL of DNA template, 2.5 μL of 280 mM magnesium acetate, and sterile distilled water to make up the volume of 50 μL. The reaction mixture was placed at 4 °C before starting the reaction. After centrifugation and mixing, the reaction mixture was incubated at 37 °C for 15–35 min. For analysis by LF strips, the reaction product was diluted by 1/50–1/200 with the PBST (Phosphate Buffered Saline with Tween-20) running buffer, and 5–10 μL was loaded onto the sample pad of LF strips, Milenia GenLine HybriDetect (Milenia Biotec, Germany). The strip was placed in 100 μL of PBST running buffer for 10 min. These results were positive when the control band and the test band had the color at the same time, and the results were negative when only the control band had the color. To optimize the reaction conditions, the probe dosage, reaction temperature, amplification time, dilution ratio, and loading volume were confirmed through the single dependent variable experiment.

PCR reaction mixture was composed of the following ingredients: 0.5 μL of 10 μM forward primer, 0.5 μL of 10 μM reverse primer, 12.5 μL of 2× Easy Taq PCR SuperMix (Transgene, China), 0.5 μL of DNA template, and 11 μL of sterile distilled water to make up the total volume of 25 μL. The reactions were performed with master cycler gradient PCR (Eppendorf, Germany). The PCR reaction program consisted of pre-denaturation at 94 °C for 4 min, 35 cycles of denaturation at 94 °C for 30 s, annealing at 59 °C for 15 s, extension at 72 °C for 30 s, and a final extension at 72 °C for 5 min. After being mixed with a 6× loading buffer containing DNA dye, the PCR products were separated in a 2% agarose gel electrophoresis. The electrophoresis conditions were set at 100 V for 30 min. The analysis was performed with a gel imaging system (VILBER QUANTUM ST4, France). The primers used in the PCR experiment are shown in Table 1.

The qPCR reaction mixture was set up in a total volume of 20 μL, containing 0.4 μL of 10 μM forward primer, 0.4 μL of 10 μM reverse primer, 10 μL of 2× SYBR Green Pro Taq HS Premix (AG, China), 1 μL of DNA template, and 8.2 μL of sterile distilled water. The reaction procedures included pre-denaturation at 94 °C for 30 s, 35 cycles of denaturation at 94 °C for 5 s, annealing, and extension at 60 °C for 34 s with fluorescence acquisition. The primers used in the qPCR experiment are shown in Table 1. Microcystis aeruginosa FACHB-905 was used as the standard MCs producing cyanobacteria in this study. The cell number of FACHB-905 was counted using a flow sight cytometer (Amnis, Merck Millipore, Darmstadt, Germany) and was diluted to ∼10⁶ cells/L. 100 mL of the diluent were filtered through a 0.45 μM membrane to enrich algal cells, and the DNA of the membrane was extracted using FastDNA SPIN Kit for Soil (MP bio, USA). The water samples also went through the same filtration and DNA extraction process. The standard curve is the relationship between the Ct value of standard DNA and the log₁₀ of cell density. The algal density of the water samples is calculated by comparing their Ct value with the standard curve.

By consulting the following reference (Chen et al. 2012; Zhang et al. 2015), 15 algae strains (Table 2) were selected to verify the specificity of the selected primers, including eight strains of MCs producing cyanobacteria, six strains of non-MCs producing cyanobacteria, and one nontoxic diatom. All algae strains were purchased from Freshwater Algae Culture Collection at the Institute of Hydrobiology (FACHB), China. The PCR assay and the qPCR assay were performed as the control methods of the LF-RPA assay. The list of all these algae species is shown in Table 2.

The detection limit of LF-RPA was evaluated from two aspects: the copy number of the target gene mcyG and the cell number of MCs producing cyanobacteria. These two kinds of DNA samples were prepared as follows: the product of the target gene mcyG was obtained by PCR amplification with the selected primers. After purification, DNA concentration was determined based on the absorbance value at 260 nm, from which the copy number of mcyG was deduced and calculated. The purified target gene mcyG was diluted to prepare standard samples of 10⁶, 10⁵, 10⁴, 10³, 10², 10, 1, 0.1, and 0.01 copies/mL. The cell number of M. aeruginosa FACHB-905 was counted using the flow sight cytometer (Amnis, Merck Millipore, Darmstadt, Germany). FACHB-905 was processed by the freeze cracking method, which was diluted into 10⁷, 10⁶, 10⁵, 10⁴, 10³, 10², 10, 1, and 0.1 cells/mL. After centrifugation at 15,000 rpm/min for 5–10 min, the supernatants were used as the DNA template. These two kinds of DNA samples were detected using the LF-RPA method, the PCR method, and also the qPCR method to verify the detection limits, in which the latter two were performed as the control experiment.

The water samples were collected once a month from March 2019 to January 2020 from Dongzhang Reservoir located in Xiamen (Fujian, China), and the sampling site was selected at the water surface near its dam. 500 mL of each water sample was filtered through a 0.45 μm membrane, and 1 mL of TE buffer solution was added into the cut membrane to resuspend the algal cells. The algae DNA of the natural water samples was quickly extracted using the freeze-thaw cycles. The LF-RPA method was applied to detect MCs producing cyanobacteria in these water samples. By comparison, the PCR and qPCR analyses were performed as control experiments.

The LF-RPA reaction was performed using the DNA extracted from 10⁶ cells/mL M. aeruginosa FACHB-905 with different reaction temperatures. The strips of both 37 and 45 °C showed positive results, and the other test strips showed negative results, suggesting that the LF-RPA assay worked well at a temperature range of 37–45 °C (Figure 1). The recommended reaction temperature of the TwistAmp nfo kit is 37–39 °C, and thus 37 °C was chosen as the reaction temperature in this study.

When the amplification time was more than 15 min, the test band presented a colored line (Figure 2). As the reaction time increased, the colored line became darker, indicating the gradually increased product. These results suggested that 15 min of amplification time might be sufficient for the LF-RPA assay to detect the MCs producing cyanobacteria. Properly prolonging the amplification time can improve the detection sensitivity of samples with lower concentrations.

In this study, the effect of different dosages of the 10 μM probe on the detection results was investigated. Figure 3 shows that 0.2–0.6 μL of 10 μM probe in 50 μL of reaction mixture had no significant difference in the color of test band in detecting 10⁶ copies/mL mcyG. However, the difference became very obvious in detecting the blank sample. The test band presented false-positive when the volume of the 10 μM probe was more than 0.4 μL in 50 μL of the reaction mixture. The strips showed the correct results with no false-positive when the volume of 10 μM probe was used as 0.2–0.3 μL. These findings indicated that an appropriate reduction of probe dosage could reduce the false-positive and improve the data accuracy. The results of 0.2 μL probe dosage were more significant and accurate than those of 0.3 μL in the blank control and positive groups. Therefore, the probe dosage in 50 μL of the reaction mixture was determined to be 0.2 μL.

To present the amplification results of the RPA reaction more accurately, the dilution ratio and loading volume were optimized in this study. Figure 4 shows that the false-positive rates of 1/100 dilution and 1/200 dilution were lower than those of 1/50 dilution. However, the larger the dilution times were, the weaker the color of the test band was. Considering the color of the test band in the positive group would be weakened by increasing the dilution times, 1/100 was considered as the best dilution ratio. The difference between the positive group and the blank control group of 10 μL loading volume was more prominent than that of the 5 μL loading volume. Therefore, 10 μL was confirmed as the better loading volume. The results suggested that both dilution ratio and loading volume in LF analysis had an important influence on the data accuracy.

The specificity of the LF-RPA assay was verified by detecting eight MCs producing cyanobacteria (including seven strains of toxic Microcystis and one strain of toxic Nostoc sp.) and seven strains of nontoxic algae (Table 2). It was observed that the strips of eight MCs producing cyanobacteria presented positive results, and the others presented negative results (Figure 5). The results of the PCR assay were consistent with those of the LF-RPA assay. The above results indicated that the LF-RPA assay combined with the selected primers had a high specificity for amplification of mcyG of Microcystis and Nostoc. All algal species used in this study were purchased from the FACHB. There are not many algal strains confirmed to produce MCs in the FACHB, so the species of cyanobacteria that produce MCs tested in this study were limited. In this study, only two genera of MCs producing cyanobacteria, Microcystis and Nostoc, were used to test the specificity of the selected primers. Whether these selected primers can be used to detect MCs producing cyanobacteria in other genera remains to be verified.

HPLC-MS analysis (Chen et al. 2012) showed that MCs were produced by Microcystis flos-aquae FACHB-1028, but the MCs concentration in each cell was the lowest, which was only 1–10% of other M. flos-aquae. However, mcyG was not detected in FACHB-1028 by PCR in this study, which is consistent with the results of published research (Zhang et al. 2015). It is speculated that the analysis detection result of HPLC-MS of FACHB-1028 may be false-positive caused by contamination. Another possible reason is that the sequence of MC synthesis gene of M. flos-aquae FACHB-1028 may be very different from other Microcystis, but this inference needs to be verified by full sequencing of MC synthesis genes of FACHB-1028. Therefore, whether FACHB-1028 secretes MCs remains to be further confirmed.

The detection limit of the LF-RPA assay was assessed from two aspects, i.e., one was from the copy number of mcyG, and the other was from the cell number of MCs producing cyanobacteria. The detection limits for mcyG and M. aeruginosa FACHB-905 were 10³ copies/mL and 10⁴ cells/mL, respectively. By comparison, the detection limits of the PCR assay were 10³ copies/mL mcyG and 10⁵ cells/mL M. aeruginosa FACHB-905, and the detection limits of the qPCR assay were 10² copies/mL mcyG and 10³ cells/mL M. aeruginosa FACHB-905. The detection sensitivity of the LF-RPA assay was similar to that of the PCR assay, while the detection limit of the qPCR assay was slightly higher by one to two orders of magnitude than that of the LF-RPA assay.

Theoretically, excluding those algal cells in the period of proliferation and division, a cell of MCs producing cyanobacteria usually contains a copy of mcyG gene. The detection limit of LF-RPA is directly related to the copy number of mcyG gene, which is not directly related to the species of algae. The cell density of MCs producing cyanobacteria determines the copy number of mcyG gene. FACHB-905 is an MC producing cyanobacteria known to produce MC-LR, which is often used in the research of MCs producing cyanobacteria (Wan et al. 2020). Under the experimental culture conditions, M. aeruginosa FACHB-905 is in a single cell state. Therefore, FACHB-905 is selected as the standard algae species to verify the detection limit of LF-RPA.

The reason why the detection limit of this method was very high is that we reduced the amount of the probe in the reaction system. When the amount of probe is added less, the detection limit becomes higher. However, the amount of probe is increased, and it is easy to be false-positive. We sacrificed the detection limit for accuracy because we thought accuracy was more important than the detection limit for this method. An algal cell producing MCs contains at least one copy of the mcyG gene, and the detection limit of LF-RPA for standard DNA mcyG and DNA samples extracted from M. aeruginosa FACHB-905 should be the same in theory. However, Figure 6(a) and 6(b) shows that the detection limit of LF-RPA for standard DNA was slightly higher than that of DNA samples extracted from M. aeruginosa FACHB-905 by one order of magnitude. The DNA of FACHB-905 was rapidly extracted by freeze-thaw cycles. The cell lysis efficiency of freeze-thaw cycles was about 7.18–35.80% (see Supplementary Material, Table S4), by comparing the cell number of FACHB-905 before and after freeze-thaw cycles, which explained the reason why the detection limits of LF-RPA for two types of samples were different.

The LF-RPA assay was used to detect the MCs producing cyanobacteria in Dongzhang Reservoir to assess its feasibility. The results showed that MCs producing cyanobacteria were detected in June, July, and August of 2019, while the cell numbers of MCs producing cyanobacteria in July were more than those in June and August (Figure 7(a)). The PCR assay showed the same results as the LF-RPA assay (Figure 7(a)). The results of qPCR showed that the cell number of MCs producing cyanobacteria in June, July, and August 2019 was 10⁵–10⁶ cells/L (10²–10³ cells/mL). The detection limit of LF-RPA for M. aeruginosa FACHB-905 is 10⁴ cells/mL (Figure 6(b)). After the algal density in the water sample was concentrated 500 times by filtration, the detection limit was increased to 10² cells/mL (10⁵ cells/L). The LF-RPA assay was established to quickly identify whether there are MCs producing cyanobacteria in algal bloom samples. Generally, when the algal density exceeds 10⁷ cells/L (Ying et al. 2014), it is considered that bloom has occurred. The detection limit of 10⁵ cells/L is sufficient to meet the detection of algal bloom samples. For the water samples with the density of MCs producing cyanobacteria less than 10⁵ cells/L, the application of this method can be realized by increasing the multiple of concentration and extracting DNA by DNA extraction kit. The above results confirmed that the LF-RPA assay established in this study could be applied for the rapid detection of MCs producing cyanobacteria blooms in natural waters.

In this study, a rapid detection method based on LF-RPA was established for detecting MCs producing cyanobacteria. Due to its high sensitivity, the detection results were often prone to false positives. In this study, we found that reducing probe dosage and increasing the dilution ratio could effectively reduce the false-positive. The specificity and detection limit of this method were similar to those from the PCR assay, indicating that the LF-RPA assay can be used as an accurate and reliable detection method. The shortest reaction time of the LF-RPA assay for the detection of MCs producing cyanobacteria was only 30 min, including the preparation time of the reaction mixture, the amplification time, and the time of LF strips analysis. Compared with the PCR assay, the LF-RPA assay had the advantages of much simpler operation, shorter reaction time, and no need for a PCR instrument. Overall, the LF-RPA assay can be applied for the rapid detection of MCs producing cyanobacteria blooms in the field samples, which can help to improve the efforts for controlling and treating MCs producing cyanobacterial blooms.

This work was supported by the STS program supporting projects of Fujian Province (No. 2018T3003) and the National Natural Science Foundation of China (Nos 41703074, 51678551, and U2005206).

The authors declare no conflict of interest.

All relevant data are included in the paper or its Supplementary Information.