Recent Advances in the Research on the Anticyanobacterial Effects and Biodegradation Mechanisms of Microcystis aeruginosa with Microorganisms

Harmful algal blooms (HABs) have attracted great attention around the world due to the numerous negative effects such as algal organic matters and cyanobacterial toxins in drinking water treatments. As an economic and environmentally friendly technology, microorganisms have been widely used for pollution control and remediation, especially in the inhibition/biodegradation of the toxic cyanobacterium Microcystis aeruginosa in eutrophic water; moreover, some certain anticyanobacterial microorganisms can degrade microcystins at the same time. Therefore, this review aims to provide information regarding the current status of M. aeruginosa inhibition/biodegradation microorganisms and the acute toxicities of anticyanobacterial substances secreted by microorganisms. Based on the available literature, the anticyanobacterial modes and mechanisms, as well as the in situ application of anticyanobacterial microorganisms are elucidated in this review. This review aims to enhance understanding the anticyanobacterial microorganisms and provides a rational approach towards the future applications.


Introduction
Harmful cyanobacterial blooms (HCBs) caused by cyanobacteria (including Microcystis, Anabaena, Nodularia, Oscillatoria, and so on) have become a common occurrence in freshwater worldwide [1,2]. Among the blooming cyanobacteria, Microcystis aeruginosa is one of the most common and widespread species [3]; specifically, it is known to be a representative species due to the dominant production of microcystins [4,5]. The rapid and excessive growth of M. aeruginosa is harmful to drinking water treatments and aquatic ecosystems due to the release of algal organic matters and cyanobacterial toxins [6,7]. As a result, the control of HCBs in water sources is a matter of great urgency.
Many approaches have been adopted for M. aeruginosa removal over the past few decades [8]. Physical methods including mechanical salvage, physical aeration, and ultrasonic treatment are usually high cost and take a long time; chemical methods such as chemical oxidants are highly efficient and low-cost methods for controlling HCBs within a short time [9]. However, chemicals may lead to a secondary contamination that may lead to potential threats to the aquatic ecosystem [10,11]. Compared with the physical and chemical methods, biological approaches such as plant allelopathy, aquatic animals and anticyanobacterial microorganisms are considered to be an economic and environmentally friendly way for cyanobacteria inhibition/biodegradation [2,10,12]. Among these methods, anticyanobacterial microorganisms are used as efficient biological agents M. aeruginosa [13]; furthermore, the microcystins can be biodegraded by certain anticyanobacterial microorganisms at the same time [6,14,15].
Up to now, several review articles have been published to introduce the anticyanobacterial microorganisms including bacteria, viruses, and fungi [2,10,13,16,17]. However, the previous reviews have concentrated mainly on both the freshwater and marine cyanobacterial/algal species or diatoms. While few studies have focused on elimination and degradation of the toxic cyanobacterium M. aeruginosa by bacteria and fungi. Moreover, the important role of anticyanobacterial microorganisms on the key genes expression and the anticyanobacterial activities regulated by quorum sensing (QS) system hasn't been mentioned. In order to clarify the current situation of anticyanobacterial microorganisms for M. aeruginosa control, the available literature on the bacteria and fungi (studies that focused on bacteriophages against Microcystis spp. are not included in this review) are adapted to review the current progress. In this review, anticyanobacterial substances and their acute toxicities (the half maximal effective concentration, EC 50 ), anticyanobacterial modes and mechanisms, as well as in situ application of anticyanobacterial microorganisms are elucidated. This review will enhance understanding the anticyanobacterial microorganisms and provide a rational attitude towards future application

Anticyanobacterial Effects for M. aeruginosa 2.1. Anticyanobacterial Microorganisms
Over the past few decades, the isolation and identification of microorganisms with anticyanobacterial effects have attracted extensive concern. Based on the literature, a variety of anticyanobacterial microorganisms have been isolated from the natural environment, and most of them belong to the anticyanobacteria and anticyanobacterial fungi.
It is shown in Table 1 that the largest number of anticyanobacterial Firmicutes are the Bacillus group, accounting for 77.3% of the total number of Firmicutes, while the remaining strains are from the genera Exiguobacterium [44,45] and Staphylococcus [35]. Li et al., (2015) revealed that Bacillus sp. Lzh-5 releases anticyanobacterial substances to attack M. aeruginosa, M. viridis, Chroococcus sp., and Oscillatoria sp. [46]; B. licheniformis Sp34 can also effectively destroy the cell membrane of M. aeruginosa and inhibit the synthesis of microcystins [47]; moreover, the simultaneous application of Bacillus sp. T4 and toxindegrading bacteria could eliminate both Microcystis sp. and microcystins [48]. These results demonstrate that Bacillus not only inhibits the growth of M. aeruginosa [49,50], but also inhibits the expression of microcystins synthesis gene mcyB [47,51] and degrades the cyanobacterial toxins [48]. Obviously, Bacillus has a potential application for HCBs control.
There is only one strain of Deinococcus metallilatus MA1002 attached to Thermus that has been reported to inhibit M. aeruginosa [52]. The bacterium Deinococcus sp. also shows an anticyanobacterial effect on the toxic dinoflagellate Alexandrium tamarense [53]. Except for the genera mentioned above, other genera connected with anticyanobacterial or flocculation activities also exist, including Citrobacter sp. [28,54] and Sphingopyxis sp. [55]. The above anticyanobacteria can destroy the M. aeruginosa cells by causing cell membrane damage, and oxidative stress and by inhibiting the gene expression from a wide range of temperatures (−20 to 121 • C) and pH (3 to 11) [5,32,33]. Not only that, the photosynthesis system of M. aeruginosa is also reduced [56]. To summarize, the anticyanobacteria can effectively inhibit the growth of M. aeruginosa, and cause an inhibition effect at a low concentration.  NA means the date is not available, not mentioned, or unclear. An asterisk (*) stands for the Chl a concentration, µg L −1 ; Two asterisks (**) represent the cell concentrations of anti-cyanobacterial microorganisms, cfu mL −1 ; Three asterisks (***) represent the dry cell weight concentrations of the anti-cyanobacterial microorganisms, mg L −1 .

Anticyanobacterial Fungi
Compared with the studies of anticyanobacteria, the research and application of fungi for eliminating or inhibiting M. aeruginosa cells has not received much attention until 2010 [105,106]. Only Ascomycetes and Basidiomycetes have been found to have the anticyanobacterial effects against M. aeruginosa. It has been reported that the fungus Trichaptum abietinum 1302BG can eliminate four cyanobacteria directly including M. aeruginosa FACH-918 and M. aeruginosa PCC 7806 in 48 h [106]. Some other fungi such as Trichoderma citrinoviride [6], Penicillium chrysogenum [97], Aureobasidium pullulans KKUY070 [98], Lopharia spadicea [99], Phanerochaete chrysosporium [100,101], Irpex lacteus T2b [102], Trametes versicolor F21a [107] and Bjerkandera adusta T1 [103] also show good inhibitory activities against M. aeruginosa. It has been stated that T. citrinoviride and A. pullulans have highly specific anticyanobacterial effects towards Microcystis spp. while they have an insignificant effects on the green algae or diatoms [6,98]; furthermore, the biodegradation of M. aeruginosa cells may be due to the excretion of the lytic enzyme (N-β-acetylglucosaminidas) [98], which can degrade the peptidoglycan from the cyanobacterial cell wall. The extracellular enzymes of cellulase, β-glucosidase, protease, and laccase from T. versicolor F21a have also been proven to be responsible for the degradation of Microcystis spp. [107,108].
On the contrary, the M. aeruginosa cells are damaged in a short time under the treatment of T. abietinum 1302BG, I. lacteus T2b or T. hirsuta T24, and the anticyanobacterial process occurs "cell to cell" through the following steps: (1) the fungus comes into physical contact with the surface of the cyanobacterial cells; (2) cyanobacterial cells are encompassed with mycelia, which destroy the cyanobacterial cell wall and membrane; and (3) the nucleic acids and other substances of cyanobacteria cells are released [17]. Fungi have the natural ability to destroy Microcystis cells by secreting anticyanobacterial substances or through "cell to cell" contact. Apart from the growth inhibition and cell lysis of M. aeruginosa, some fungi are able to remove microcystins [6,98,106], and the removal mechanism is related to the adsorption/biodegradation of fungus or the inhibition expression of microcystins synthesis gene [15].
Previous studies have indicated that amino acids have powerful anticyanobacterial effects against Microcystis spp. at concentrations between 0.6 and 5.0 mg L −1 [11,110,111], and the inhibition effect of L-lysine against Microcystis sp. is remarkable [110]. Moreover, the eutrophic lake with the dominant species of cyanobacterium M. aeruginosa is selectively controlled by lysine [111]. The amino acids and proteins have commonly been identified and reported as the anticyanobacterial substances for M. aeruginosa. Two amino acids (L-lysine and L-phenylalanine) are purified from B. amyloliquefaciens T1 that have an inhibition effect against M. aeruginosa FACHB-905 [94]; the L-valine, which shows a better anticyanobacterial activity than L-lysine, is also isolated from S. jiujiangensis JXJ 0074 [37]. It is interesting that the anticyanobacterial efficiency of tryptamine and tryptoline on M. aeruginosa FACHB-905 is 80 ± 1% and 100 ± 2%, respectively, but the growth of M. aeruginosa is recovered as tryptamine (tryptoline) and is completely used or degraded by microorganisms [38]. Therefore, the persistence of amino acids should be further considered when they are used for eutrophication control [112].   [117] NA means the date is not available, not mentioned or unclear; An asterisk (*) stands for the Chl a concentration, µg L −1 .

Anticyanobacterial Modes
In general, the anticyanobacterial modes by microorganisms are divided into direct attack (bacterial and cyanobacterial cell contact) and indirect attack (the release of anticyanobacterial substances) (Figure 1) [10,32,72,118]. To date, although anticyanobacteria can directly kill several different kinds of cyanobacteria, only few has been reported. A wide range of cyanobacteria including M. aeruginosa, M. wesenbergii, M. viridis, Anabaena flos-aquae, Oscillatoria tenuis, Nostoc punctiforme and Spirulina maxima are lysed by B. cereus DC22 with the direct attack mode, as well as chlorophyceae (Chlorella ellipsoidea and Selenastrum capricornutum) [89]. In addition to B. cereus, other anticyanobacteria that destroy M. aeruginosa with direct attack have also been reported. For example, the anticyanobacterial modes of Aeromonas bestiarum HYD0802-MK36 [20], Chryseobacterium sp. [40], Streptomyces globisporus G9 [83], Alcaligenes denitrificans [59], and Shigella sp. H3 [60] on M. aeruginosa are regarded as direct attack, and a number of cyst-like cells are formed in cyanobacteria during the direct attack [10]. It is speculated that the cyanobacterial cell walls are partially destroyed at the contact point with the anticyanobacteria, and the formation of cyst-like cells is a potential defense system against anticyanobacteria [2,10].

Anticyanobacterial Modes
In general, the anticyanobacterial modes by microorganisms are divided into direct attack (bacterial and cyanobacterial cell contact) and indirect attack (the release of anticyanobacterial substances) (Figure 1) [10,32,72,118]. To date, although anticyanobacteria can directly kill several different kinds of cyanobacteria, only few has been reported. A wide range of cyanobacteria including M. aeruginosa, M. wesenbergii, M. viridis, Anabaena flosaquae, Oscillatoria tenuis, Nostoc punctiforme and Spirulina maxima are lysed by B. cereus DC22 with the direct attack mode, as well as chlorophyceae (Chlorella ellipsoidea and Selenastrum capricornutum) [89]. In addition to B. cereus, other anticyanobacteria that destroy M. aeruginosa with direct attack have also been reported. For example, the anticyanobacterial modes of Aeromonas bestiarum HYD0802-MK36 [20], Chryseobacterium sp. [40], Streptomyces globisporus G9 [83], Alcaligenes denitrificans [59], and Shigella sp. H3 [60] on M. aeruginosa are regarded as direct attack, and a number of cyst-like cells are formed in cyanobacteria during the direct attack [10]. It is speculated that the cyanobacterial cell walls are partially destroyed at the contact point with the anticyanobacteria, and the formation of cyst-like cells is a potential defense system against anticyanobacteria [2,10]. The indirect attack mode has been observed in the numerous metabolites from most of the reported anticyanobacterial microorganisms, and the anticyanobacterial characteristics of these bacteria seem to be unique to M. aeruginosa. Up to now, the genus Acinetobacter [22,72,119] and Exiguobacterium [44,45,96], which firstly attach to M. aeruginosa and then cause serious damage to the cyanobacterial cell structure and morphology, are recognized as degrading M. aeruginosa by producing anticyanobacterial substances. Nevertheless, some anticyanobacteria can inhibit or kill green alga and cyanobacteria with an indirect attack simultaneously. For instance, B. amyloliquefaciens FZB42 can efficiently eliminate M. aeruginosa, Anabaena sp., A. flos-aquae and Nostoc sp. by secreting bacilysin [91]. In line with this genus, B. amyloliquefaciens T1 produces amino acids to inhibit the growth of four Microcystis spp., but not of Anabaena flos-aquae or Chlorella pyrenoidosa [49,94]; S. amritsarensis HG-16 kills A. flos-aquae, Phormidium sp. and five Microcystis spp. by secreting active substances, but has a small inhibitory effect on C. vulgaris and a promoting effect on Oscillatoria sp. [5]. Along with this, the anticyanobacterial modes of Aquimarina salinaria on green algae and cyanobacterium, which is a direct attack on C. vulgaris 211-31 and an indirect attack on M. aeruginosa MTY01, is quite different [39]. Furthermore, a recent study firstly demonstrated that Paucibacter aquatile DH15 inhibits M. aeruginosa by both direct and indirect attacks [61], which would be interesting and could shed further light on the anticyanobacterial modes by microorganisms. The indirect attack mode has been observed in the numerous metabolites from most of the reported anticyanobacterial microorganisms, and the anticyanobacterial characteristics of these bacteria seem to be unique to M. aeruginosa. Up to now, the genus Acinetobacter [22,72,119] and Exiguobacterium [44,45,96], which firstly attach to M. aeruginosa and then cause serious damage to the cyanobacterial cell structure and morphology, are recognized as degrading M. aeruginosa by producing anticyanobacterial substances. Nevertheless, some anticyanobacteria can inhibit or kill green alga and cyanobacteria with an indirect attack simultaneously. For instance, B. amyloliquefaciens FZB42 can efficiently eliminate M. aeruginosa, Anabaena sp., A. flos-aquae and Nostoc sp. by secreting bacilysin [91]. In line with this genus, B. amyloliquefaciens T1 produces amino acids to inhibit the growth of four Microcystis spp., but not of Anabaena flos-aquae or Chlorella pyrenoidosa [49,94]; S. amritsarensis HG-16 kills A. flos-aquae, Phormidium sp. and five Microcystis spp. by secreting active substances, but has a small inhibitory effect on C. vulgaris and a promoting effect on Oscillatoria sp. [5]. Along with this, the anticyanobacterial modes of Aquimarina salinaria on green algae and cyanobacterium, which is a direct attack on C. vulgaris 211-31 and an indirect attack on M. aeruginosa MTY01, is quite different [39]. Furthermore, a recent study firstly demonstrated that Paucibacter aquatile DH15 inhibits M. aeruginosa by both direct and indirect attacks [61], which would be interesting and could shed further light on the anticyanobacterial modes by microorganisms.

Anticyanobacterial Mechanisms
Currently, the anticyanobacterial mechanisms of microorganisms against M. aeruginosa are mainly dependeent on the attack modes, and these mechanisms are revealed with the changes in the photosynthesis system, antioxidant enzymes system, gene expression and QS system (Figure 2).

Anticyanobacterial Mechanisms
Currently, the anticyanobacterial mechanisms of microorganisms against M. aeruginosa are mainly dependeent on the attack modes, and these mechanisms are revealed with the changes in the photosynthesis system, antioxidant enzymes system, gene expression and QS system (Figure 2).

Effects of Anticyanobacterial Microorganisms on Photosynthesis
Photosynthesis, which converts solar energy into chemical energy through the photosynthesis system (PS) II and PS I, is the principal mode of energy metabolism in cyanobacteria [120]. Anticyanobacterial microorganisms can significantly affect the photosynthesis of M. aeruginosa cells in several ways, including decreasing the chlorophyll a (Chl a) contents and photosynthetic pigments [56], and the disruption of the electron transport pathway in PS [23,93]. Chl a is one of the important components of cyanobacterial pigments. It is markedly decreased in M. aeruginosa under the exposure of anticyanobacteria such as P. aeruginosa [18,63], Streptomyces sp. [33,36], Exiguobacterium sp. [44,45], and so on. For the photosynthetic pigments, phycocyanobilin (PC), allophycocyanin (APC) and phycoerythrin (PE) are major indicators of cyanobacterial photosynthetic efficiency and are essential apparatus for light harvesting [61], and the addition of anticyanobacterium results in a significant decrease in the PC, APC and PE by disrupting the synthesis of an photosynthetic pigments [56]. In addition, the expressions of pcA and apcA genes for PC and APC synthesis in M. aeruginosa are down-regulated by Paucibacter aquatile DH15, which shows an inhibition effect on active chlorophyll [61]. It has been noted that the Chl a decrease is closely related to the reduction in photosynthetic pigments, and the cyanobacterial membrane is sensitive and easily damaged by anticyanobacterium [56].
The variations of cyanobacterial energy kinetics have also been evaluated by Chl fluorescence parameters, such as the maximum photochemical quantum yield of PS II (Fv/Fm), the effective quantum yield (Φe), and the maximum electron transport rate (ETRmax) [41,95]. With the addition of fermentation filtrate (5%, v/v) of Paenibacillus sp. SJ-73, the Fv/Fm values of M. aeruginosa PCC7806 and M. aeruginosa TH1701 dramatically decline from 0.52 and 0.29 to 0 [95]; similarly, it is only 0.08 (14.3% of the initial value) for M. aeruginosa FACHB-905 after being treated for 24 h by the fermentation filtrate (5%, v/v) of Raoultella sp. S1 [23]. Besides, the Φe and ETRmax of M. aeruginosa 9110 following the treatment of Chryseobacterium sp. GLY-1106 decrease gradually with time [41]; the

Effects of Anticyanobacterial Microorganisms on Photosynthesis
Photosynthesis, which converts solar energy into chemical energy through the photosynthesis system (PS) II and PS I, is the principal mode of energy metabolism in cyanobacteria [120]. Anticyanobacterial microorganisms can significantly affect the photosynthesis of M. aeruginosa cells in several ways, including decreasing the chlorophyll a (Chl a) contents and photosynthetic pigments [56], and the disruption of the electron transport pathway in PS [23,93]. Chl a is one of the important components of cyanobacterial pigments. It is markedly decreased in M. aeruginosa under the exposure of anticyanobacteria such as P. aeruginosa [18,63], Streptomyces sp. [33,36], Exiguobacterium sp. [44,45], and so on. For the photosynthetic pigments, phycocyanobilin (PC), allophycocyanin (APC) and phycoerythrin (PE) are major indicators of cyanobacterial photosynthetic efficiency and are essential apparatus for light harvesting [61], and the addition of anticyanobacterium results in a significant decrease in the PC, APC and PE by disrupting the synthesis of an photosynthetic pigments [56]. In addition, the expressions of pcA and apcA genes for PC and APC synthesis in M. aeruginosa are down-regulated by Paucibacter aquatile DH15, which shows an inhibition effect on active chlorophyll [61]. It has been noted that the Chl a decrease is closely related to the reduction in photosynthetic pigments, and the cyanobacterial membrane is sensitive and easily damaged by anticyanobacterium [56].
The variations of cyanobacterial energy kinetics have also been evaluated by Chl fluorescence parameters, such as the maximum photochemical quantum yield of PS II (Fv/Fm), the effective quantum yield (Φe), and the maximum electron transport rate (ETRmax) [41,95]. With the addition of fermentation filtrate (5%, v/v) of Paenibacillus sp. SJ-73, the Fv/Fm values of M. aeruginosa PCC7806 and M. aeruginosa TH1701 dramatically decline from 0.52 and 0.29 to 0 [95]; similarly, it is only 0.08 (14.3% of the initial value) for M. aeruginosa FACHB-905 after being treated for 24 h by the fermentation filtrate (5%, v/v) of Raoultella sp. S1 [23]. Besides, the Φe and ETRmax of M. aeruginosa 9110 following the treatment of Chryseobacterium sp. GLY-1106 decrease gradually with time [41]; the ETRmax values of M. aeruginosa are also depressed significantly under the stress of Raoultella sp. S1 [23] and Bacillus sp. B50 [93]. The decreases in Fv/Fm, Φe and ETRmax demonstrate that the photosynthetic system is seriously damaged and the electron transport chain is blocked, resulting in the inhibition of cyanobacterial cell photosynthesis [55]. In consequence, the possible mechanism underlying the photosynthetic reduction could be due to the reduction in Fv/Fm, Φe and ETRmax in M. aeruginosa.

Effects of Anticyanobacterial Microorganisms on Antioxidant Enzymes System
The oxidative damage of the cyanobacterial cells can occur under different environmental stress conditions, and it will results in an increase in reactive oxygen species (ROS), which includes the superoxide anion radical, hydrogen peroxide and hydroxyl radicals [51,61]; while excess ROS often leads to oxidative stress, lipid peroxidation, and DNA damage [56,121]. The enzymatic antioxidants (such as catalase (CAT), superoxide dismutase (SOD), peroxidase (POD), and so on) and non-enzymatic antioxidants (such as ascorbic acid (AsA) and glutathione (GSH)) are responsible for removing the overproduction of ROS [2,31,41]. For instance, Streptomyces eurocidicus JXJ-0089 inhibits the growth of cyanobacterial cells in various ways, including promoting ROS production (e.g., O 2 • − ), inhibiting the antioxidant synthesis, removing chlorophyll and destroying cell walls [38].
The ROS of cyanobacteria increases excessively by either the direct attack or indirect attack of anticyanobacterial microorganisms. The O 2 • − content in M. aeruginosa cells is induced largely by 4 µg mL −1 3, 4-dihydroxybenzalacetone (DBL) secreted from Phellinus noxius HN-1 and increased from 0.360 ± 0.001 to 0.400 ± 0.001 µg g −3 [104]. The ROS level of M. aeruginosa NIES 843 treated with Bacillus sp. AF-1 (cell-free filtrate) was lower than that of the control at the first 48 h but much higher at 72 h, indicating that some evasive mechanisms were taken to prevent the ROS accumulation in cyanobacterial cells at the initial stage [51]. Similar variations of ROS have been observed in M. aeruginosa KW after being treated with Paucibacter aquatile DH15, and the malondialdehyde (MDA) content and SOD activity related to remove ROS also increased at first and then decreased [61]; The MDA content, CAT and POD activity of M. aeruginosa FACHB-905 also increased quickly when fermentation liquid (5%, v/v) of P. aeruginosa [18] and P. chrysosporium was added quickly [101]; moreover, the responses of M. aeruginosa FACHB-905 cells to Streptomyces sp. KY-34 and Streptomyces sp. HJC-D1 following a similar pattern with the increases of CAT, SOD and POD, and the MDA further increased during the incubation time [56,121]. Although the antioxidants increased immediately to relieve the damage caused by anticyanobacteria, the cyanobacterial cell membrane may have decompose due to the accumulation of MDA [18,67,121].
For the non-enzymatic antioxidants, the variation of GSH is opposite to that of the antioxidase activity. The Bacillus licheniformis Sp34 induces more GSH production in M. aeruginosa at first to clear ROS, but the GSH content is much lower at 20 h (compared with the control) [47]. Such a phenomenon is also obtained in the anticyanobacterial process of Raoultella sp. S1 [23]. The prodigiosin from Hahella sp. KA22 also leads to the variation of GSH content, while the GSH content decreases slightly after exposure for 36 h [31]. These results demonstrate that the ROS levels and MDA contents decrease under prolonged exposure to anticyanobacteria [31,33,65]; in addition, the non-enzymatic antioxidants also play a critical role in protecting the cyanobacterial cells from oxidative damage under anticyanobacterial stress [23].

Regulating the Anticyanobacterial Activity by QS System
QS system is the regulator control system for microorganisms that sense the cell density of their own species and make themselves to coordinate gene expression and physiological accommodation on a community scale [122,123]. It is a cell-to-cell communication that relies on the signal molecules [124], and the accumulated QS signals can bind to the cognate receptors and regulate biological activities and cellular functions [69,125]. Previous studies have shown that microbial behaviors such as the secondary metabolites, cell motility and antibiotic resistance are all influenced by QS [122,123]; in addition, QS signals that contribute to the interactions between planktonic microalgae and bacteria are summarized as the N-acyl-homoserine lactones (AHLs) [69], the 2-alkyl-4-quinolones (AQs) [123], long-chain fatty acids and fatty acid methyl esters (autoinducer-2, AI-2) and dihydroxypentanedione furanone derivates [12]. It is agreed that most of the anticyanobacterial activities by Gram-negative bacteria (such as Pseudomonas sp., Acinetobacter sp., etc.) are the consequence of bacterial-cyanobacterial QS rather than bacterium-cyanobacteria interactions [12,124]. Some species of Serratia sp. [109] and Hahella sp. [31] can produce prodigiosin to inhibit M. aeruginosa, and the prodigiosin production is regulated by LuxI and LuxR, which are the crucial genes of AHLs [126]. The QS signal molecule (C4-HSL), which belongs to the classic AHL-based LuxIR-type QS system of Gram-negative bacteria, is responsible for the synthetic process of the anticyanobacterial compound (3-methylindole) from Aeromonas sp. GLY-2107 [69]. During the anticyanobacterial process, the QS systems of Gram-negative bacteria produce AHLs signaling molecules, which are synthesized by the basic regulatory protein of LuxI [69,88,126].
In contrast, a wide range of the Gram-positive anticyanobacteria (such as Streptomyces sp., Bacillus sp., etc.) generally use AI-2 as the signal molecules in QS systems [125]. The anticyanobacterium S. xiamenensis Lzh-2 exhibits QS behavior, and the LuxS gene is crucial for the AI-2 type QS system; obviously, the anticyanobacterial activity of S. xiamenensis Lzh-2 is regulated through the LuxS/AI-2 QS system by inducing the production of anticyanobacterial compounds 2, 3-indolinedione and cyclo(Gly-Pro) [126]. The AI-2 type QS behavior is present in Bacillus sp. [127]. Genomic analysis of B. subtilis JA has indicated the existence of the LuxS gene that regulates the pheromone biosynthesis, and the high-molecular-weight anticyanobacterial compounds (>3 kDa) produced by Bacillus sp. S51107 have been proven to be primarily regulated by the NprR-NprX-type (AI-2) QS system [88]. As a consequence, the AI-2 QS system has been considered as a possible strategy to regulate the behavior of the anticyanobacterial effects of Gram-positive bacteria. Although QS behavior has been reported in recent years, there is still an improved understanding of the interaction between cyanobacteria and anticyanobacterial microorganisms.

Application of Anticyanobacterial Microorganisms
In consideration of the drawbacks of physical and chemical methods, the biological control of HCBs is of great importance for the aquatic ecological environment. In particular, the application of anticyanobacterial microorganisms (bacteria and fungi) or their anticyanobacterial substances is regarded as the most suitable approach due to the economical and environment-friendly performance. It is well known that it is difficult for microorganisms to exist persistently in the aquatic environment [128]. To overcome this limitation, microbial immobilized technology using different porous matrices for enhancing the cyanobacterial removal efficiency has been attempted. For example, a biological treatment system equipped with coconut packing carriers has been established to enrich anticyanobacteria. The results indicate that the average anticyanobacterial efficiency of 87.69 ± 2.44% is obtained and 13 genera anticyanobacteria, which account for 10.17% of the total bacteria, are responsible for the removal of HCBs [129]. As the Brevundimonas sp. AA06 is immobilized using polyvinyl alcohol-sodium alginate beads and B. methylotrophicus ZJU is immobilized with Fe 3 O4 nanoparticles, the inhibition effects are much better than freely suspended cells [24,50]; meanwhile, the extracellular polymeric substances produced by P. aeruginosa ZJU1 are made as bioflocculants, and the removal efficiency of M. aeruginosa reached 100 ± 0.07% in 5 min at the dosage of 2.75 g/L bioflocculant [130]. These strategies demonstrating the "indirect attack" of microorganisms could be immobilized by multi-functional systems and their anticyanobacterial products could be further enriched. Taking full account of the uncertainties of using anticyanobacterial microorganisms to control/eliminate HCBs in natural waters, the "direct attack" microorganisms may be as ineffective as "indirect attack" microorganisms in actual applications.
In situ eutrophication controls have also been carried out in other researche. It was found that the Chl a removal efficiency reached 99.2% when the anticyanobacterium B. cereus N-1 was immobilized with a floating carrier for natural eutrophication water [48]; the wild cyanobacteria from a shallow eutrophic pond were significantly controlled by adding solid B. amyloliquefaciens T1 agent at the concentration of 0.5 mg L −1 (or above) [49]. Taking the recycling utilization of the industrial waste product into account, approximately 80.0% of the M. aeruginosa and 48.1% of the microcystin-LR were removed by the biosorbent, which originated from the Escherichia coli biomass [131]. Apart from the persistent existence of microorganisms, anticyanobacterial effects are concerned with environmental conditions and nutrient concentrations [132]. As the previous study indicates, the yeast Candida utilis F87, which converts the nitrogen and phosphorus into microbial protein, can inhibit the growth of M. aeruginosa by nutrient competition [133]. Therefore, the issue of nutrient competition in cyanobacterial control using microorganisms is a crucial consideration. Based on the current collection of literature, the anticyanobacterial microorganisms have a potential application for HCBs control in the natural environment.

Summary and Prospective
Interactions between cyanobacteria and microorganisms are considered to be an integral part of the geochemical cycle. However, with the spatial and temporal heterogeneity, these interactions can be modulated in various ways, and highly efficient anticyanobacterial strategies in the eutrophic environment can be obtained from microorganisms. Plentiful studies have reported on ecological interactions between anticyanobacteria and cyanobacterium M. aeruginosa, which are focused on the anticyanobacterial microorganisms, substances, modes and mechanisms. Although the anticyanobacterial approach by microorganisms seems to be safe and effective, it is still appreciated that there are limitations and challenges in field applications. A drawback of this approach is that anticyanobacterial microorganisms must be chosen carefully to secrete specific anticyanobacterial compounds and the dosage of the microorganism inoculum or microbial agent is of great importance. On the other hand, the abiotic and biotic factors of the natural environment may have a remarkable influence on the distribution of cyanobacteria and the cyanobacterial response to anticyanobacterial substances.
Besides the target specificity, the complicating factors in realistic eutrophic environment research are the complexity of consortia with multiple species and the unsustainability of anticyanobacteria. It is delightful to see that the studies for HCBs control in situ have contributed to a better understanding of the role of anticyanobacterial microorganisms, especially the multiple regulations for microcystins. Further investigations should be focused on the simultaneous removal of nitrogen, phosphorus and microcystins by mixed microbial community, and the understanding of the cell-to-cell communication and the defense mechanisms of QS systems. Besides, more insights are needed for the specific genes encoding photosystem synthesis, peptidoglycan synthesis, membrane proteins, cyanotoxin microcystins, DNA repair and so on.