Correspondence: Hong-Ying Hu, Environmental Simulation and Pollution Control State Key Joint Laboratory, Department of Environmental Science and Engineering, Tsinghua University, Beijing 100084, China. Tel.: +86 10 6279 4005; fax: +86 10 6279 7265; e-mail: email@example.com
Algal blooms have become a worldwide issue recently, especially those comprised of toxic cyanobacteria. Grazers' predation of bloom-forming algae plays an important role in water clearing. In this study, a species of golden alga (strain ZX1), capable of feeding on the toxic cyanobacteria Microcystis aeruginosa, was isolated and identified as Poterioochromonas sp. (GenBank accession: EU586184) on the basis of morphological characteristics and 18s rRNA gene sequencing. Feeding experiments showed that ZX1 could clear high densities of M. aeruginosa (7.3 × 105–4.3 × 106 cells mL−1) in a short time (40 h), with inhibition ratios higher than 99.9%. ZX1 grew during the feeding processes and achieved a maximum density of 10–20% of the initial M. aeruginosa density. Furthermore, this study is the first to report that ZX1 was able to degrade microcystin-LR (MC-LR) in cells of M. aeruginosa while digesting the whole cells, and that the degradation process was determined to be carried out inside the ZX1 cell. For a total MC-LR (intra- and extracellular) concentration of up to 114 μg L−1, 82.7% was removed in 40 h. This study sheds light on the importance of golden alga in aquatic microbial ecosystems and in the natural transportation/transformation of MC-LR.
Recently, blue algal blooms have broken out frequently in lakes and reservoirs around the world, causing great damage to the biodiversity and to the equilibrium of aquatic ecosystems. Several species of blue algae (e.g. Microcystis and Anabaena) are more harmful than others as they release toxins into water bodies, causing illness or the death of wildlife and humans. Millions of dollars will be spent to deal with the problems (e.g. off-flavor, drinking water, and health care) caused by harmful algal bloom, and >2000 cases of human poisoning are caused globally by cyanotoxins each year (Men et al., 2007). Microcystis aeruginosa is the most common toxic cyanobacterium found worldwide, and produces potent cyclic peptide hepatotoxins called microcystins. So far, almost 60 variants of microcystins have been isolated; microcystin-LR (MC-LR) is regarded as the most common and most toxic variant of microcystins (Vasconcelos et al., 1996; Park et al., 2001). A provisional guideline value of 1 μg L−1 has been issued for MC-LR in drinking water by the World Health Organization (WHO, 1998). Therefore, M. aeruginosa and MC-LR are the standard species of bloom-forming algae and the standard microcystin in studies, respectively.
Algal blooms en masse may be dispersed via physical factors or removed from the water column due to sinking, while the mortality of individual cells within blooms may be affected by autolysis, viruses, predatory bacteria, or grazing zooplankton (Rosetta & McManus, 2003). A number of relevant microorganisms have already been isolated either from bloom-forming water or during laboratory studies. Cyanophage (a virus of cyanobacteria e.g. LPP-1), bacteria (e.g. Bacillus sp., Flexibacter sp., Myxococcus sp.), fungi (e.g. Rhizophidium planktonicum, Acremonium, Emericellopsis, and Verticillium), and some actinomycetes (e.g. Streptomyces) are all potential biocontrol agents. These microorganisms may act in one of three major ways to control the overgrowth of bloom-forming algae: production of extracellular products, contact lysis, or entrapment lysis (Sigee et al., 1999). Also, some kinds of protozoa [e.g. Tintinnids (Adrniraal & Venekamp, 1986), Strombidium lingulum (Montagnes & Humphrey, 1998), Oxyrrhis marirta (Jeong et al., 2003), Gyrodinium dominans (Nakamura et al., 1995), and Naegleria (Liu et al., 2006)] and mixotrophs [e.g. Ochromonas danica (Cole & Wynne, 1974)] may act as alternative biocontrol agents that can graze on certain types of bloom-forming algae; these species can thereby play an important part in balancing the flow of matter in the case of phytoplankton blooms (Tillmann, 2004).
Besides bacteria, viruses and protozoa, other kinds of microorganisms may also influence bloom-forming algae, including other species of algae. These species, especially mixotrophic algae, which can graze on certain types of algae, play an important part in forming the community structure of the aquatic ecosystems (Boraas et al., 1988). Several species of golden algae, such as Ochromonas, Poterioochromonas, and Chrysamoeba, are mixotrophic and can feed on organic particles, bacteria, and certain species of green algae, cyanobacteria, and diatoms (Daley et al., 1973; Zhang et al., 1996; Holen, 1999; Kristiansen, 2005; Ou et al., 2005). Ochromonas danica and Poterioochromonas malhamensis can feed on the toxic bloom-forming cyanobacterium M. aeruginosa (Daley et al., 1973; Zhang et al., 1996), and a species of Poterioochromonas can degrade MC-LR (Ou et al., 2005). However, only a few species of golden alga capable of feeding on M. aeruginosa have been reported, and there is a great dearth of data related to the feeding characteristics and degradation processes of microcystins.
In this study, a species of golden alga (strain ZX1) that feeds on M. aeruginosa was isolated and identified. The feeding characteristics of the golden alga predation on M. aeruginosa were studied along with the degradation of MC-LR during the feeding period. Based on these results, the significance of golden algae in aquatic microbial ecosystems was discussed.
Materials and methods
Microcystis aeruginosa PCC7806 was purchased from the Freshwater Algae Culture collection of the Institute of Hydrobiology (FACHB) (Wuhan, Hubei Province, China) and cultured in sterilized BG11 medium (Rippka et al., 1979). The algae were cultured under the standard condition (800–1100 lx white light; light : dark=14 h : 10 h, 25 °C) for about 14 days to reach the log phase before being used as inoculants. The flasks were shaken at least once a day.
ZX1 was fed with M. aeruginosa cells in BG11 medium and cultured under the standard condition. The culture with only ZX1 was used as the grazer inoculants, after M. aeruginosa cells were cleared out.
Golden alga identification
ZX1 was identified by both morphological characteristics and 18s rRNA gene sequencing. The morphological characteristics were observed via light microscopy (Leica DM6000B) with an electronic flash device for photography and transmission electron microscopy (TEM). The processes of DNA analysis included PCR amplification of the 18s rRNA genes, blast analysis, and comparison with sequences in the GenBank nucleotide database (http://www.ncbi.nlm.nih.gov/BLAST/). The polygenetic tree was used to describe the relationship between the golden alga and other strains. The primers and PCR amplification processes were according to Ou (2005).
Feeding experiments of ZX1 on M. aeruginosa were carried out in 500-mL flasks, with different initial volumes of the two algal inoculants and sterilized BG11 medium, reaching 200 mL in total. Three different initial densities of M. aeruginosa were prepared: 7.3 × 105, 2.4 × 106, and 4.3 × 106 cells mL−1. In the test groups, 10 mL liquid inoculants containing 8 × 106 ZX1 cells were added to each flask, while no ZX1 was added to the controls. Groups were tested in triplicate and cultured under the standard condition. Subcultures were taken periodically over a 40-h incubation period, and cell densities of the two algae were determined using a hemocytometer under light microscopy (magnification: 10 × 40). The feeding experiments were repeated three times.
Analysis method for MC-LR
The concentrations of both extra- and intracellular MC-LR were determined by the SEP-HPLC method, according to Men & Hu (2007). The concentration of MC-LR was determined by developing a calibration curve from commercially available MC-LR. All chemicals used were of HPLC or of analytical grade.
The inhibition ratio of M. aeruginosa (IR, %) was calculated as IR=(1−Ntt/Nct) × 100, where Ntt and Nct are the densities (cells mL−1) of M. aeruginosa in the test group and in the control at time t, respectively.
The growth-curve parameters during the log phase of M. aeruginosa were calculated following the formulas below. Because the grazing period was short, the growth of M. aeruginosa during this period was ignored. A linear model was used to describe the reduction of M. aeruginosa according to the grazing of the golden alga. The clearance rate of each ZX1 (RC, nL per cell of golden alga h−1) was calculated as
where N0 and Nt are the densities (cells mL−1) of M. aeruginosa at the beginning and the ending points of the log phase, respectively; M0 and Mt are the densities (cells mL−1) of ZX1 at the beginning and the ending points of the log phase of M. aeruginosa, respectively; and Δt is time of the log phase (h).
The specific growth rate of ZX1 (μ, h−1) was calculated as μ=ln(Mt/M0)/Δt.
The specific attenuation rate of M. aeruginosa (ν, h−1) was calculated as ν=ln(Nt/N0)/Δt.
The removal ratio of total MC-LR (η, %) was calculated as η=(1−Ctt/Cct) × 100, where Ctt and Cct are the concentrations (μg L−1) of total MC-LR in the test group and in the control at time t, respectively.
Morphology and taxonomy of ZX1
ZX1 was yellow brown, transparent, free swimming, had no cell walls, and presented in spherical, ovum, or amoeba form with one or two flagella (Fig. 1). The cell lengths and movement speeds were 5–15 μm and 0–5 μm s−1, respectively. The food storage products were oils, and the mode of reproduction was cell division. The species proved to be mixotrophic (autotrophic, osmotrophic, and phagotrophic) and grew faster phagotrophically than autotrophically. ZX1 was identified as a species of golden alga on the basis of its morphological characteristics, referring to the descriptions of golden alga by Kristiansen (2005).
After PCR amplification and 18s rRNA gene sequencing, one sequence of 1477 bp was obtained. Alignments and phylogenetic analysis (Fig. 2) demonstrated that the gene sequences of ZX1 (GenBank accession: EU586184) had 99% maximum identification with the P. malhamensis strain SAG933.1c (GenBank accession: EF165114). Therefore, ZX1 was identified as a species of the Poterioochromonas genus.
Growth and grazing characteristics of ZX1 on M. aeruginosa
Three different initial densities of M. aeruginosa were fed to ZX1, and the density changes of M. aeruginosa in the control and the test groups are shown in Fig. 3. The densities of M. aeruginosa kept growing slowly in the controls but declined steeply in the test groups. These results showed that ZX1 could feed on the cyanobacterium M. aeruginosa and clear M. aeruginosa of high density (4.3 × 106 cells mL−1) in a short time (40 h). The inhibition ratios of M. aeruginosa were higher and the medium was apparently clearer in the lower initial M. aeruginosa density groups during the first 20 h. Therefore, we recommend that ZX1 be placed in the water body during the early periods of blue algal blooms if it is to be applied in controlling future blooms.
In all three test groups, there were lag phases during the first 20 h, which might have been due to low ZX1 densities. As ZX1 density increased up to around 105 cells mL−1, M. aeruginosa density decreased quickly and the green color of the medium faded noticeably. After 34.5 h of incubation, M. aeruginosa densities in all the test groups were <104 cells mL−1 (the detection limit), with inhibition ratios around 99.9%. The medium plateaued at this stage for several days.
The growth curve of ZX1 followed the ‘S’ pattern, and there were three phases as shown in Fig. 4, including the lag, log, and stationary phases. The maximum densities of ZX1 (among 1–5 × 105 cells mL−1) increased linearly with the initial densities of M. aeruginosa, ranging from 5 × 105 to 5.1 × 106 cells mL−1 (R2=0.80), where these maximums represented 10–20% of the initial M. aeruginosa densities. The decreasing density of the grazer met the pyramid of energy flow in the food chain principle. The stationary phase lasted for 10–20 days, at which point ZX1 declined in density and slowed down in movement, approaching the declining phase (data not shown). Based on these results, we can deduce that ZX1 will not form blooms.
Growth and grazing characteristic parameters of ZX1 feeding on M. aeruginosa were calculated during the log phase of M. aeruginosa (Table 1). The specific growth rates of ZX1 and the specific attenuation rates of M. aeruginosa increased with the initial densities of M. aeruginosa.
Table 1. Growth characteristic parameters of ZX1 and Microcystis aeruginosa
Initial density of M. aeruginosa (cells mL−1)
Specific attenuation rate of M. aeruginosa (h−1)
Clearance rate of each ZX1 (nL per cell of golden alga h−1)
Specific growth rate of ZX1 (h−1)
7.3 × 105
0.233 ± 0.070
1.874 ± 0.198
0.034 ± 0.011
2.4 × 106
0.339 ± 0.011
0.815 ± 0.074
0.076 ± 0.011
4.3 × 106
0.334 ± 0.055
0.612 ± 0.102
0.129 ± 0.020
Degradation of microcystins (MC-LR) by ZX1
As shown in Table 2, MC-LR was degraded by ZX1. The concentrations of MC-LR in controls increased with the initial density of M. aeruginosa, and were mostly contained within cells. However, the concentrations of intracellular and total MC-LR decreased dramatically in the test groups, with removal ratios of total MC-LR >82.7% in 40 h. Intracellular MC-LR in the test groups included MC-LR in cells of both M. aeruginosa and ZX1, and its disappearance indicated both that MC-LR within cells of M. aeruginosa was degraded by ZX1 and that there was no MC-LR accumulation in ZX1 cells. Furthermore, the degradation processes were carried out inside ZX1 cells while ZX1 was digesting M. aeruginosa cells. However, the concentrations of extracellular MC-LR in the test groups were higher than those of the controls, which may have been due to an increased release of microcystin by M. aeruginosa under the grazing pressure of ZX1. More studies should be carried out to verify the assumption and to clarify the degradation pathway of MC-LR by ZX1.
Table 2. Concentrations and the removal ratios of MC-LR in different groups
Intracellular MC-LR included MC-LR in cells of both M. aeruginosa and ZX1 in the test groups.
7.3 × 105
2.4 × 106
4.3 × 106
Golden algae and mixotrophy
Golden algae (Chrysophytes) constitute a group of microscopic algae that are primarily characterized by their golden color (Kristiansen, 2005). The chrysophytes show an unusual and remarkable range of different nutrition strategies, including phototrophy, osmotrophy, or phagotrophy. Some species have a combination of phototrophy with any or all of these methods, called mixotrophy. The golden alga ZX1 isolated in this study is mixotrophic (phototrophic, osmotrophic, and phagotrophic).
The mixotrophs are widespread in aquatic habitats and represent an important step of eukaryote evolution; they possess advantages that facilitate their survival under particular conditions, such as low light and scarce mineral nutrients. In the illuminated surface strata of a lake, mixotrophs steeply reduce prey abundance and escape competition with, or losses to, higher grazers that cannot persist. Furthermore, the mixotrophs structure prey abundance along the vertical light gradient (Martin & Hofle, 2001; Tittel et al., 2003). They can even invade established plankton communities depending on the trophic status of the system, affecting the food web structure, species diversity, nutrient dynamics, and the flux of material through planktonic food webs (Jones, 2000; Domaizon et al., 2003; Katechakis & Stibor, 2006). However, until now, quantifying the role of mixotrophs in plankton communities has proven extremely difficult, especially in a field study.
As a mixotroph, ZX1 success is favored in shaded, slightly humic, mesotrophic ponds with some organic nutrients. Mixotrophic chrysophytes may constitute >50% of the total phytoplankton biomass in some lakes (Kristiansen, 2005). Mass developments, or blooms, of chrysophytes (e.g. Dinobryon, Synura, Chrysococcus, and Uroglena, not including Poterioochromonas sp.) may sometimes occur in ponds or lakes, but in smaller quantities than for inducing cyanobacterial blooms. The study of golden algae feeding on bloom-forming cyanobacteria is valuable in the study of mixotrophs and bloom-forming algae.
Feeding on toxic M. aeruginosa
Several protozoa, both in freshwater and in the sea, capable of grazing on algae, have been reported (Tillmann, 2004), but few feed on the toxic cyanobacteria M. aeruginosa. Daphnia can feed on M. aeruginosa, but accumulate microcystins in their bodies at a level of 1.78 μg toxin per 25 daphnids (Mohamed, 2001). Until now, only a few species of golden alga (Poterioochromonas sp. and O. danica) have been reported to be capable of grazing on M. aeruginosa (Cole & Wynne, 1974; Zhang et al., 1996; Ou et al., 2005). The ingestion and digestion processes of grazing by P. malhamensis and O. danica on M. aeruginosa and other organic particles were observed using light and electron microscopy by Cole & Wynne (1974) and Zhang et al. (1996). The processes were as follows: (1) the membrane that was derived from the plasma membrane and that surrounded the prey disappeared sometime after ingestion, (2) the food vacuole was then formed by successive fusion of numerous homogeneous vesicles accumulated around the prey, and (3) the prey was enclosed in a single membrane-bound food vacuole and then digested. Ou et al. (2005) reported that Poterioochromonas sp. could degrade MC-LR. We thus believed that MC-LR was degraded in ZX1 cells while ZX1 was digesting M. aeruginosa cells.
In conclusion, the golden alga Poterioochromonas sp. is widespread in freshwater environments and is capable of feeding on toxic cyanobacterium M. aeruginosa and biodegrading MC-LR. Based on these results, Poterioochromonas sp. is deduced to play an important part in the flux of material through planktonic food webs and the community structure of the aquatic ecosystems. However, this only heralds the very beginning of studies regarding interactions between golden algae and M. aeruginosa and the degradation of microcystins by golden algae. This leaves a great deal of work to be carried out in determining the roles of golden algae in the aquatic microbial food chains and in natural transformation/transportation of microcystins.
This study was funded by the NSFC-JST joint project (No.50721140017).