Introduction

Microcystis spp. is a common genus of bloom-forming cyanobacteria, which generates Microcystis blooms worldwide1. Microcystis blooms is one of the serious harmful algae blooms because many Microcystis species produce microcystins having high toxicity2. These blooms also cause fish mortality due to depletion of oxygen3 and loss of biodiversity and affect the cycles of biogenic elements in freshwater ecosystems1,4. Thus, an insight into the distribution, succession and diversity of Microcystis species is important to understand the life-cycle of Microcystis as well as ecology of Microcystis blooms.

During the past decades, many studies have been carried out to investigate the processes of Microcystis bloom formation5,6. Multiple Microcystis species have been recorded according to their morphological characteristics, especially their colonial morphology7. The life cycle8, spatial distribution9, seasonal succession10 and physiology of Microcystis11 has been well studied based on this morphological taxonomy. In addition, the competition between Microcystis spp. and other algae and also the competition among different Microcystis species have been investigated to reveal the ecology of Microcystis bloom formation12,13.

Recently, Microcystis has been well documented having high phenotypic plasticity14,15. Otsuka et al.16 demonstrated that the colonial morphology of Microcystis in culture could change from time to time. Sun et al.17 indicated that colonies with colonial morphology of M. aeruginosa under culture conditions could change their morphology to that of a typical M. novacekii. Li et al.18 illustrated that solubilization of mucilage could induce changes in colonial morphology and the authors suggested that seasonal succession of Microcystis species was due to morphological changes. Therefore, the taxonomy of this genus should be re-evaluated via molecular genetic analyses.

The phenotype-genotype correlation of Microcystis is helpful in filling the gap between classification of molecular biology and the morphological taxonomy of Microcystis. The phylogenetic analysis based on 16S rDNA was considered as one of the most reliable criteria for determining relationships among organisms with close relation19. However, the similarity of colonies in different morphology was high as measured by 16S rDNA sequencing20,21, and thus the unification of five species of Microcystis has been proposed22. In addition, the events of horizontal gene transfer would cause flexibility of several informative genes including 16S rDNA of Microcystis23. A more reliable gene sequence should be explored to analyze the phenotype-genotype correlation of Microcystis. Otten and Paerl24 indicated that M. wesenbergii could be identified from four different Microcystis morphospecies using 16S-23S rDNA-ITS sequences, but the other four morphospecies could not. Tan et al.25 indicated cpcBA-IGS could be used as an effective tool to identify M. wesenbergii. Several polysaccharide biosynthesis-related genes were also found to identify morphospecies of Microcystis26. Thus, these genes were hypothesized to be significantly related to Microcystis colonial morphology, and this hypothesis has been preliminarily verified by Xu et al.27.

In addition, microcystin-producing genes were also postulated to divide Microcystis into toxic species and non-toxic species28. The morphospecies was considered to relate to the toxicity of Microcystis. Generally, M. ichthyoblabe was considered as non-toxic species29, while M. aeruginosa and M. wesenbergii as toxic species30,31,32. The microcystin synthetase (mcy) gene cluster in different Microcystis morphospecies was thus analyzed to reveal the phenotype-genotype correlation of Microcystis colonies33. However, it was still poorly understood whether there was a relationship between the phenotype and microcystin-producing genes.

The current study aimed to gain insight into the phenotype-genotype correlation of Microcystis. The 16S rDNA, 16S-23S rDNA-ITS, cpcBA-IGS, mcy gene (mcyB)34 and several polysaccharide biosynthesis-related genes were analyzed along with the identification of the morphology of Microcystis colonies collected in the field. This study also attempted to resolve that polysaccharide biosynthesis-related genes might distinguish the Microcystis morphospecies as EPS played great roles in colony formation and morphological changes of Microcystis18,35.

Materials and Methods

Experimental design

This study has two parts. (I) Seeking novel functional gene which may distinguish the Microcystis morphospecies. Individual Microcystis colonies were isolated from natural samples and then axenically cultured for PCR amplification and sequencing. Afterwards, phenotype-genotype correlation of Microcystis colonies was investigated and the function gene was identified. (II) Confirming the functional gene. Microcystis “scum” at different seasons were collected and divided into varying classes consisting of various Microcystis morphospecies according to colony size. The functional genes of the subsamples were then analyzed to confirm that this gene succeed in distinguishing the Microcystis morphospecies.

Sample collections

Algal samples for colony isolation and culture in part I were collected during a Microcystis bloom in Meiliang Bay in northern Lake Taihu (China) on 15 August and 1 November 2014. Lake Taihu was selected in the current study because Microcystis spp. is the dominant species at most of the time and heavy Microcystis blooms occurs frequently10. In addition, the colony morphology and phylogenetic inference of Microcystis species has been well investigated in this lake8,24,36, which could be referred to. The water samples containing abundant Microcystis colonies were collected directly from the lake surface (30 cm depth) and were transferred into plastic bottles with a capacity of 5 L. The samples were then stored in a cold closet and transported to the laboratory as soon as possible for culture. Algal samples for confirming the functional gene in part II were collected on 4 June, 16 July, 17 August, 29 September, 15 October and 15 November, 2015, respectively.

Microcystis colony separation

Water samples for part I were diluted with BG-11 culture medium until a single Microcystis colony could be separated by a pipette. The separated colony was examined under a microscope (×100), and the colonial morphology was recorded. M. aeruginosa and M. wesenbergii were found in the sample collected on 15 August. M. ichthyoblabe was found in the sample collected on 1 November. Five colonies of each morphology were separated for culture. M. ichthyoblabe colonies were named M. ichthyoblabe colonies TH11, TH12, TH13, TH14 and TH15. M. aeruginosa colonies were named M. aeruginosa colonies TH21, TH22, TH23, TH24 and TH25. M. wesenbergii colonies were named M. wesenbergii colonies TH31, TH32, TH33, TH34 and TH35.

Single colony culture

Each colony was washed with BG-11 medium three times. Then, the colonies were cultured in 10 mL of BG-11 medium in glass tubes at 25 °C under a 12 h:12 h light-dark cycle with a light density of approximately 45 μmol m−2 s−1. After one month of culture, the M. ichthyoblabe colonies TH11, TH12, TH13, TH14, TH15, the M. aeruginosa colonies TH21 and TH22 and the M. wesenbergii colonies TH31 and TH32 grew well but the others died. The DNA of the growing Microcystis was extracted.

DNA extraction

The DNA extraction method was referred to Sun et al.17. Microcystis pellets were dispersed into 0.8 mL extraction buffer (1.5 M NaCl, 1% CTAB, 100 mM Tris-HCl, 100 mM Na2EDTA, 100 mM Na3PO3, pH 0.8) and 20 μL of proteinase K (30 mg mL−1). Afterwards, they were incubated at 37 °C for 30 min and then, 0.48 mL of 20% SDS was added to each sample, incubating at 65 °C for 1 h. The samples were extracted using phenol-chloroform-isoamyl (25:24:1) and chloroform-isoamyl (24:1) successively. Centrifuged at 8000 × g for 5 min, the supernatant was transferred to new tubes. Thereafter, 0.6 mL pure isopropyl alcohol was injected to purify the DNA sample. After 20-min centrifugation at 16000 × g, 70% ethanol was used to rinse the DNA sample. Each DNA sample was dried and dissolved in 100 μL of Tris-EDTA (10 mM Tris and l mM EDTA, pH 8.0). Finally, the DNA sample was analyzed using a Nanodrop-2000.

PCR amplification and sequencing

Seven pairs of primers targeting the 16S rRNA, 16S-23S ITS(A)/(S), cpcBA-IGS, mcyB, TagH and epsL genes were used for the amplification and sequencing of all of the samples (see Table 1). A total volume of 50 μL containing 25 μL of 2 × PCR mixture buffer with tag enzyme (Bioteke, Beijing, China), 1.2 μL of each primer (10 μM), 2 μL DNA (10–20 ng μL−1) and 21.8 μL ddH2O was used for the PCR amplifications. The PCR amplification was run with an initial denaturation of the DNA at 94 °C for 5 min, followed by 34 cycles of 50 s at 94 °C, 50 s at 42 °C (mcyB) or 30 s at 50 °C (16S, 16S-23S) or 30 s at 52 °C (cpcBA-IGS) or 30 s at 55 °C (TagH, epsL), and 1 min at 72 °C. The reaction was completed after 10 min at 72 °C. The detection and the size of the amplicons were determined by agarose (1.0%) gel electrophoresis compared with a DL2000 DNA Marker (Tiangen, Beijing, China). The amplicons with the correct length were used for sequencing by the Tianyihuiyuan biotechnology company (except mcyB gene).

Table 1 List of primer pairs for the amplification and sequencing of Microcystis.
Full size table

Treatment of samples for part II

The sample for part II was poured gently through sieves (divided into four classes: >500 μm, 300–500 μm, 150–300 μm and 75–150 μm). Each class was re-suspended in BG-11 medium. For each subsample from sieving, the photomicrographs were taken using an Olympus C-5050 digital camera coupled with an optical microscope (Olympus CX31). The length and width of Microcystis colonies was analyzed using the UTHSCSA ImageTool (v3.00, University of Texas Health Science Center, San Antonio, TX, USA). The biovolume of Microcystis colony was calculated as volume = π/6 (length × width)3/2 as it is hard to measure the thickness of colonies. A total of 300 colonies were analyzed in each sample. Afterwards, the percentage of different Microcystis morphospecies in the total Microcystis biovolume of each subsample was calculated. Microcystis morphospecies was identified according to Yu et al.7. In the current study, M. ichthyoblabe, M. aeruginosa and M. wesenbergii was identified as in Fig. 1 and other Microcystis colonies were defined as unidentified Microcystis.

Figure 1
figure 1

Micrographs of Microcystis species collected in Lake Taihu.

(A) M. aeruginosa; (B) M. wesenbergii; (C) M. ichthyoblabe.

Full size image

For each subsample, DNA for PCR templates was extracted. Only epsL and TagH were used for amplification and sequencing according to the results of part I. All the procedure and method was as same as those described for part I.

Data analysis

Alignment for all of the sequences was determined by Muscle and edited by software Bioedit37. Some related sequences in the NCBI database were also used for alignment. MEGA5 was used to construct neighbor-joining tree of phylogeny analysis38, with bootstrap for 1000 replications, Maximum Composite Likelihood, and d: Transitions + Transversions.

Results and Discussion

Relationship between species and toxicity

Figure 2 shows an electropherogram of the PCR products with the primer of mcyB. Our results showed that one M. aeruginosa colony contained mcyB but the other did not. Two out of five M. ichthyoblabe colonies contained mcyB in this study. Mazur-Marzec et al.39 showed similar results in the Vistula Lagoon (southern Baltic Sea). However, M. aeruginosa colonies are generally considered as toxic species30,40. M. ichthyoblabe has never been reported to produce microcystins29,41,42. M. wesenbergii was classified as a non-toxic species31, but our results showed that both two M. wesenbergii colonies contained mcyB. Nevertheless, some investigations32,42 also illustrated that M. wesenbergii is toxic. All of the conflicting conclusions above indicated that there is not an exact relationship between the phenotype and microcystin-producing genes.

Figure 2
figure 2

Electropherogram of the PCR products with the primer of mcyB.

Full size image

Yoshida et al.32 divided 47 strains of Microcystis into three clusters based on the sequences of 16S-23S rDNA-ITS. Their results showed that the first cluster contained both non-toxic and toxic strains, the second only had toxic ones, and the last only had non-toxic strains. This result implied that the 16S-23S gene may distinguish the toxic and non-toxic Microcystis species, which was also reported by Janse et al.43. On the contrary, our results demonstrated that the 16S-23S gene sequences failed to distinguish nine strains with different phenotypes, four of which possessed the mcyB gene. This result suggested that 16S-23S rDNA-ITS gene failed to distinguish toxic and non-toxic strains. Yoshida et al.44 suggested that 16S rDNA could used to identify toxic and non-toxic Microcystis species in some bloom stages. However, our results did not reach a similar conclusion. Therefore, the Microcystis species identified by morphology or molecular tools (16S rDNA, 16S-23S ITS and cpcBA-IGS) could not be used to distinguish toxic and non-toxic species.

Phylogenetic trees based on 16S, 16S-23S and cpcBA-IGS

The phylogenetic trees referring to 16S, 16S-23S and cpcBA-IGS are illustrated in Figs 3, 4 and 5, respectively. The 16S sequences divided all of the samples into two clusters. All of the M. ichthyoblabe colonies were in the same clade, but this clade also included M. wesenbergii colony (TH22). Both of the M. aeruginosa colonies and M. wesenbergii colonies were found in clade 1. However, these colonies had high homozygosity in 16S with M. ichthyoblabe 0BB39S02 (AJ635433), Microcystis novacekii TAC20 (AB012336) and Microcystis viridis TAC17 (AB012328). 16S rDNA sequences could not be used to distinguish different phenotypes of Microcystis20. Lepère et al.21 also reported that the 16S rDNA sequences of six Microcystis strains assigned to four different morphospecies based on colonial morphology were similar.

Figure 3
figure 3

Phylogenetic tree based on the analysis of the 16S gene sequences.

Full size image
Figure 4
figure 4

Phylogenetic tree based on the analysis of the 16S-23S gene sequences.

Full size image
Figure 5
figure 5

Phylogenetic tree based on the analysis of the cpcBA-IGS gene sequences.

Full size image

Sanchis et al.45 used both the 16–23S rDNA ITS and the cpcBA-IGS sequences to identify Microcystis. Their results suggested that M. novacekii could be distinguished from M. wesenbergii, but there was a close relationship between M. novacekii and M. aeruginosa. Otten and Paerl24 also indicated that M. wesenbergii could be identified within four different Microcystis morphospecies based on the 16S-23S rDNA-ITS sequences. Similarly, Yoshida et al.32 found that M. aeruginosa could be distinguished from M. wesenbergii and M. viridis by the 16S-23S rDNA-ITS sequences. Do Carmo Bittencourt-Oliveira et al.46 successfully distinguished the M. aeruginosa morphospecies from the morphospecies of M. wesenbergii and M. viridis based on the DNA sequences of cpcBA-ITS.

All the above studies considered that M. wesenbergii could be distinguished using the 16–23S rDNA ITS and the cpcBA-IGS sequences. Conversely, in the current study, the sequences displayed high homozygosity for each 16S-23S and cpcBA-IGS in all of the samples except for the M. aeruginosa colony, TH32 (Figs 4 and 5). Similarly, the phylogenic tree for the 63 Microcystis strains in China based on the cpcBA-IGS gene sequences showed that this gene did not always succeed in identifying different morphospecies47. These occasional failures may be resulted from genetic variations among the strains of Microcystis48. One Microcystis genotype was reported to have more than one phenotype29,49. In East Africa, 24 isolated strains of M. aeruginosa could be separated into 10 genotypes based on the DNA sequences of the PC-IGS and ITS1 rDNA regions50. Thus, there was no obvious relationship between these phenotypes and the phenotypes of Microcystis based on 16S, 16S-23S and cpcBA-IGS because of the significant genetic variations among the strains of Microcystis.

Polysaccharide biosynthesis-related genes

Figure 6 shows a phylogenetic tree based on the analysis of the sequences of the polysaccharide biosynthesis-related genes (espL and TagH). The results demonstrate that the M. wesenbergii colonies could be divided directly from other colonies using espL. Xu et al.27 suggested that the polysaccharide biosynthesis-related gene TagH may explain the diversity of the Microcystis morphospecies. In the current study, TagH divided all of the samples into two clusters but failed to distinguish the different phenotypes.

Figure 6
figure 6

Phylogenetic tree based on the analysis of the polysaccharide biosynthesis-related gene sequences (espL and TagH).

Full size image

Since very small amount of colonies were tested and cultured, there would be a risk that the final Microcystis morphotype would change compared with the initially identified Microcystis due to intraspecific competition. Therefore, part II was carried out to confirm as the polysaccharide biosynthesis-related genes could distinguish the Microcystis phenotypes. The phylogenetic tree based on the analysis of the sequences of the polysaccharide biosynthesis-related genes (espL and TagH) extracted from Microcystis “scum” collected from June and November 2015, was shown in Figs 7 and 8, respectively. The gene espL divided all of the samples into two clusters and the first cluster was divided into three subclades (Fig. 7). The samples in clade 2 was dominated by M. wesenbergii (60–100%). The samples in subclade 1 of clade 1 was dominated by M. ichthyoblabe (60–100%). As shown in Fig. 8, the gene TagH divided all of the samples into two clusters. All the samples collected in June and November were brought into subclade 1 in clade 1 and samples in August were brought into subclade 2 in clade 1. The former samples was dominated by M. ichthyoblabe (60–100%) and the latter samples was dominated by M. wesenbergii (60–100%). In consequence, it was confirmed that M. wesenbergii and M. ichthyoblabe could be distinguished by the polysaccharide biosynthesis-related genes espL and TagH. However, the two polysaccharide biosynthesis-related genes (epsL and TagH) may not be qualified for identifying all the species of Microcystis. These two genes combined with some other functional genes may succeed in identifying all the Microcystis species based on further researches.

Figure 7
figure 7

Phylogenetic tree based on the analysis of the sequences of espL genes extracted from Microcystis scums collected in different months with different morphospecies composition.

Full size image
Figure 8
figure 8

Phylogenetic tree based on the analysis of the sequences of TagH genes extracted from Microcystis scums collected in different months with different morphospecies composition.

Full size image

Extracellular polysaccharide (EPS) was considered to be the material basis of Microcystis colony formation. A positive relationship between colony size and EPS content has been reported during recent years51,35. Li et al.18 illustrated that solubilization of mucilage, which consists of EPS, induced changes in Microcystis colonial morphology. Forni et al.52 indicated that the composition of EPS in different Microcystis species varied. The EPS content of various Microcystis morphospecies was also different53. Therefore, the content and composition of EPS has been postulated to be related to Microcystis colony morphology. In conclusion, the polysaccharide biosynthesis-related genes could distinguish the Microcystis phenotypes.

Conclusions

  1. 1

    Microcystis species identified by morphology or molecular tools (16S rDNA, 16S-23S ITS and cpcBA-IGS) could not be distinguished as toxic and non-toxic species.

  2. 2

    There was no obvious relationship between the phenotypes of Microcystis species based on 16S, 16S-23S and cpcBA-IGS because of the significant genetic variations among the strains of Microcystis.

  3. 3

    It was confirmed that polysaccharide biosynthesis-related genes could distinguish the Microcystis phenotypes.

Additional Information

How to cite this article: Xu, S. et al. Polysaccharide biosynthesis-related genes explain phenotype-genotype correlation of Microcystis colonies in Meiliang Bay of Lake Taihu, China. Sci. Rep. 6, 35551; doi: 10.1038/srep35551 (2016).