Abstract

Algicidal bacteria play an important role in mitigating harmful algal blooms (HABs). In the study, five bacterial strains were isolated from the East China Sea. One strain of algicidal bacterium, named DH-e, was found to selectively inhibit the motor ability of Prorocentrum donghaiense, Alexandrium tamarense (ATDH-47) and Karenia mikimotoi Hansen. Both 16S rDNA sequence analysis and morphological characteristics revealed that the algicidal DH-e bacterium belonged to Halomonas. Furthermore, results showed that the metabolites in the DH-e cell-free filtrate could kill algae directly, and the minimum inhibitory concentrations (MICs) of the bacterial metabolites on the cells of the three dinoflagellate species ranged from 35.0–70.0 μg/mL. Following short-term inhibitory tests, the dinoflagellates in mixed crude extract solution (0.7 mg/mL) ceased movement after 5 min. The algicidal mechanism of the metabolites was investigated through enzyme activities, including that of catalase (CAT), alkaline phosphatase (AKP), acetone peroxide (T-ATP) synthetase and nitrite reductase (NR). Results indicated that metabolites did not disrupt the energy or nutrient routes of the algae (P > 0.05), but did initiate an increase in free radicals in the algal cells, which might explain the subsequent death of sensitive algae. Thus, the metabolites of the DH-e bacterium showed promising potential for controlling HABs.

INTRODUCTION

Harmful algal blooms (HABs) are naturally occurring events that can destroy the ecological balance in a marine environment and kill aquatic organisms, thus causing serious economic losses (Liu et al. 2015; Hoagland & Scatasta 2016) and even threatening human life. Phototrophic dinoflagellates are important red-tide organisms in marine ecosystems due to their adverse impact on other living things (Guo et al. 2016; Lee et al. 2016). To date, chemical, physical and biological control measures have been applied to manage or relieve the occurrence of HABs (Gkelis et al. 2014). However, many of these methods are limited at the laboratory level due to the huge outlays and ongoing costs as well as resulting secondary pollution. In recent years, increasing attention has been paid to biological control methods, such as the use of algicidal bacteria (Luo et al. 2013).

With their discovery and development, the use of algicidal bacteria species as a tool to mitigate HABs has gained interest. In many cases, there are direct and indirect interactions between algicidal bacteria and algae. Direct attack by algicidal bacteria can kill algal cells via cell-to-cell contact mechanisms, while indirect interaction is thought to induce chemically mediated inhibition by the action of algicides from bacteria-secreted metabolites (Yang et al. 2014; Tan et al. 2016), which include proteins, polypeptides, and glycolipids (Dabas et al. 2014). Although many bacteria have been screened, few algicidal compounds have been isolated and purified due to the apparent variation in characteristics across different species of algicidal bacteria (Lei et al. 2014). Therefore, the ability to enrich bacterial algicides directly in media without isolation or purification is of importance.

Three species Prorocentrum donghaiense, Alexandrium tamarense and Karenia mikimotoi are key HAB dinoflagellate and are associated with several toxins, including paralytic shellfish poisoning (PSP) and hemolytic toxin, which can damage the fishery and aquaculture industry in regards to the breeding and farming culture of aquatic species. In this study, a strain of bacterium which was algicidal to three toxic dinoflagellate species, Halomonas sp. DH-e, was isolated from the red-tide area in the East China Sea, China. Identification, characterization of the bacterium and of the mode of algicidal action were undertaken in this study.

MATERIALS AND METHODS

Algal cultures

The three dinoflagellate species (P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen) were kindly supplied by Professor Wei-Dong Yang (Jinan University, China); Chlorella sp., Phaeodactylum tricornutum, Nitzschia closterium forma minutissima, Skeletonema costatum, and Chaetoceros gracilis were kindly supplied by Professor Da-Zhi Wang (Xiamen University, China). Algae cultures were maintained in f/2-Si medium at 19 ± 1 °C under an illumination of 72 μmol·m−2·s−1 in a 12 L/12D cycle.

Screening of algicidal bacteria

Seawater samples from the red-tide areas of the East China Sea were serially diluted (10-fold) using sterilized seawater, with 100 μL aliquots of each dilution then added to LB media for 72 h incubation at 28 ± 2 °C. Individual colonies of distinct morphology were further purified and stored at −80 °C in LB medium supplemented with 30% (v/v) glycerol. Five bacterial strains (Thalassospira sp., Staphylococcus sp., Alteromonas sp., Alcanivorax sp., and Halomonas sp. were symbolized as DH-a, DH-b, DH-c, DH-d, DH-e, respectively) were then isolated and cultured in LB media (28 ± 2 °C, 120 rpm) for 10 d to secrete metabolites and for 24 h to reach the exponential phase. The biomasses of the culture broths were centrifuged for 10 min at 5,000 g to collect supernatants and pellets. The pellets were suspended in aseptic seawater (6.0 × 106 CFU mL−1). An aliquot (10.0 mL) of each bacterial solution was inoculated in triplicate in 90 mL axenic cultures of A. tamarense (ATDH-47) (2,500 cell/mL). Aseptic seawater (10.0 mL) was added to each algal culture as the control 1. The supernatants were filtered to remove bacteria using 0.45 μm Millipore filters. An aliquot (10 mL) of each bacterial supernatant was inoculated in 90 mL axenic culture of A. tamarense (ATDH-47) (2,500 cell/mL). Fresh broth (10.0 mL) was also added to each algal culture to remove the effect of the bacteria medium on the algae as the control 2. The inoculated flasks were then incubated for 2 d. To investigate the algicidal activity of the five bacterial strains, algal cells were counted under a microscope after being fixed with Lugol's iodine reagent. All experiments were repeated in triplicate. The algicidal effect was measured using following Equation (1): 
formula
(1)
where D-Control (cells mL−1) and D-Treatment (cells mL−1) are the cell density of A. tamarense (ATDH-47) in the control and treatment, respectively.

Identification of algicidal bacteria

After 24 h of culture, cells from the five bacterial strains were collected by centrifugation (5,000 g for 10 min). The bacterial genomes were extracted using a Bacterial Genomic DNA Extraction Kit (Dongsheng, China), followed by polymerase chain reaction (PCR) amplification of the 16S rRNA gene. The primers used for 16 s rRNA gene amplification and sequencing were 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-TACGGCTACCTTGTTACGACTT-3′). An initial 94 °C denaturation period of 5 min was followed by 30 cycles at 94 °C for 30 s, 55 °C for 30 s, 72 °C for 1 min, and a final extension time at 72 °C for 10 min. The PCR products were checked and purified using 1% agarose gel electrophoresis, then recovered from the agarose gel with a GeneClean Turbo Kit (Qbiogene, Quebec, Canada) and sent to the Shanghai Sangon Company (China) for gene sequencing. The algicidal bacterial sequences were aligned with the same bacterial sequences obtained from GenBank.

Given its algicidal effect on the three toxic dinoflagellate species, a phylogenetic tree for the DH-e strain was constructed via the neighbor-joining method using MAGE 6.0 software (Damanka et al. 2016). The capacity to use different carbon sources of isolates DH-e was evaluated using the API 20E biochemical test kit (Table 2).

Algicidal characteristics of bacterial metabolites

Five algae species and three dinoflagellate species were used to determine the algicidal range of bacterial strain DH-e (the only strain to exhibit such activity). The original densities of the eight algae species that were in logarithmic growth period ranged from 2,000–5,000 cell/mL. For quantitative analysis, 100 mL of the DH-e culture supernatant was dried by vacuum freeze drying. The dried solids were then melted in the same volume of different organic solvents, including methyl alcohol, formic acid, ethanol, acetic acid, petroleum ether, and ethyl acetate, for 30 min. Acetic acid was found to be the optimal solvent, and could desalt and dissolve metabolites well. Thus, 100 mL of acetic acid was vaporized using a vacuum rotary evaporator to collect the crude extract. The extract was weighed with a precision electronic balance and diluted to solution gradually with sterile seawater to OD267 = 1.4–1.6 (7.0 mg/mL), which was equal to the original supernatant's OD267 value. The pH of crude extract solution was ∼9.0.

The crude extract solution at a series of volumes (1 mL, 0.5 mL, 0.25 mL, and 0.1 mL) was inoculated in 50 mL cultures of each alga in triplicate, respectively. Control 1 consisted of 1 mL of sterile seawater in 50 mL of algae culture, and control 2 consisted of 1 mL of LB medium in 50 mL of algae culture. Algal growth was measured through cell density after 8 d of culture.

Minimum inhibitory concentrations (MICs) of crude bacteria extract against the three dinoflagellates

The crude extract solution at a series of volume gradients (1 mL, 0.50 mL, 0.25 mL, and 0.10 mL) was added to 100 mL cultures of P. donghaiense (2,200 cell/mL), A. tamarense (ATDH-47) (2,500 cell/mL) and K. mikimotoi Hansen (2,100 cell/mL) in triplicate. Control 1 consisted of 1 mL of sterile seawater in 100 mL of algae culture, control 2 consisted of 1 mL of LB medium in 100 mL of algae culture. The concentration of algal cells was measured every two days.

Short-term inhibitory effect of bacterial metabolites on P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen

10 mL of the crude extract solution were inoculated in triplicate in 90 mL of algal culture, with 10 mL of aseptic seawater inoculated in algal culture without DH-e used as the control to probe the short-term inhibitory effects of the DH-e bacterial metabolites on P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen motility. The short-term inhibitory effects of the bacterial extract against the three sensitive dinoflagellates were recorded by fluorescent microscopy.

Characterization of the enzyme activities of the three sensitive dinoflagellates

To study the inhibition mechanism of the metabolites, the total-ATP enzyme, alkaline phosphatase, nitrite reductase, catalase assay kits (Nanjing Jiangcheng Bioengineering Institute, China) were used to test the basic indices of the three toxic dinoflagellate species. 10 mL of the crude extract solution and 10 mL of aseptic seawater was pooled into 100 mL of algae culture in triplicate. The enzyme activities were tested after 1 h and 24 h.

Data analysis

All statistical analyses were carried out using the SPSS statistical package (SPSS20.0 for Windows). We tested for normal distribution of all samples using the Shapiro–Wilk test. Mean and standard deviation values were calculated for three independent replicate experiments and data are presented as mean values ± standard error of the means. To identify significant differences between the different treatment-groups, the Student's t-test or one-way analysis of variance (ANOVA) was applied for normal-distributed data. A p-value was used to determine the statistical significance of differences the samples. ‘p < 0.05’ used to identify significant differences between samples and ‘p < 0.01’ used to further identify strongly significant differences between samples.

RESULTS

Algicidal effect of the five bacterial strains

After filtration and bacterial treatment, the inhibiting effects of the five bacteria (Thalassospira sp. DH-a, Staphylococcus sp. DH-b, Alteromonas sp. DH-c, Alcanivorax sp. DH-d, and Halomonas sp. DH-e) isolated from the East China Sea were determined. Cell-free filtrate of strain DH-e showed the strongest alga-lysing activity (92 ± 8.3%) against A. tamarense (ATDH-47) (2,500 cell/mL) at 48 h (Table 1). For the bacterial concentration treatment, no obvious algae inhibition was observed among the different bacterial strains. Overall, these results indicated that DH-e is a promising bacterium for algal inhibition and showed algicidal activity through indirect attack.

Table 1

Algicidal activity following different bacterial treatments

Bacteria species Algicidal effect by different bacterial treatments
 
Bacterial cells Cell-free filtrate 
Thalassospira sp. DH-a – – 
Staphylococcus sp. DH-b – – 
Alteromonas sp. DH-c – – 
Alcanivorax sp. DH-d – – 
Halomonas sp. DH-e 24.6 ± 9.6 92.0 ± 8.3 
Bacteria species Algicidal effect by different bacterial treatments
 
Bacterial cells Cell-free filtrate 
Thalassospira sp. DH-a – – 
Staphylococcus sp. DH-b – – 
Alteromonas sp. DH-c – – 
Alcanivorax sp. DH-d – – 
Halomonas sp. DH-e 24.6 ± 9.6 92.0 ± 8.3 

–: No inhibition. Numbers refer to the rate of algae dissolution. Data are expressed as Mean ± SD (n = 3) (p > 0.05, Shapiro–Wilk test on SPSS).

Identification of the strain DH-e

Based on the 16S rDNA sequence, DH-e (GenBank accession number KP144872) was found to belong to Halomonas (Figure 1), and the bacterial cells were gram stain-negative. It was found that DH-e could take up saccharose, lactose and glucose, but not starch or inose. Furthermore, the results of the urea decomposition (ammonia production), indole, and citrate tests were negative, whereas those of the Voges-Proskauer (VP) and methyl (MR) tests were positive (Table 2). The DH-e strain was deposited in the China Centre for Type Culture Collection (CCTCC M 2014541).

Figure 1

Phylogenetic tree based on multiple alignments (ClustalX) of 16 rDNA sequences from the DH-e strain. The tree was generated using the neighbor-joining method. Algicidal bacterium DH-e is marked by the black dot. Codes before names are GenBank accession numbers.

Figure 1

Phylogenetic tree based on multiple alignments (ClustalX) of 16 rDNA sequences from the DH-e strain. The tree was generated using the neighbor-joining method. Algicidal bacterium DH-e is marked by the black dot. Codes before names are GenBank accession numbers.

Table 2

Physiological and biochemical tests of bacterium DH-e

Experiment project  Experiment project  
Starch Ammonia production – 
Lactose VP 
Glucose MR – 
Saccharose Indole – 
Inose – Citrate – 
Experiment project  Experiment project  
Starch Ammonia production – 
Lactose VP 
Glucose MR – 
Saccharose Indole – 
Inose – Citrate – 

+: positive; −: negative.

Algicidal characteristics of the bacterial metabolites

After 8 d of cultivation, the effect of the crude extract solution on five species of algae was different (Figure 2(a)). The concentration values of Chlorella sp., P. tricornutum and Nitzschia closterium f. minutissima, were 1.10, 1.09, 1.27 times of the control value, respectively. The concentration of C. gracilis was decreased, but the effect of algae is basically the same in different concentrations of the crude extract solution. The concentration value of S. constatum was 0.99 times the control value. However, the crude extract solution exhibited strong inhibition on the three dinoflagellate species (Figure 2(b)), P. donghaiense (F = 5,047.706 > F0.01(5,12) = 5.06, df = 17, p < 0.01), A. tamarense (ATDH-47) (F = 800.857 > F0.01(5,12) = 5.06, df = 17, p < 0.01) and K. mikimotoi Hansen (p < 0.01) (F = 588.873 > F0.01(5,12) = 5.06, df = 17, p < 0.01). It was found that the crude extract solution of DH-e specifically inhibited the dinoflagellate species but did not impact the growth of the other algal species.

Figure 2

Algicidal activity of the DH-e strain against different algal species after 8 d. (a) Five algae species were treated by DH-e strain, (b) three dinoflagellate species were treated by DH-e strain (p > 0.05, Shapiro–Wilk test on SPSS; P < 0.01, one-way ANOVA on SPSS). All data were mean ± SD (n = 3).

Figure 2

Algicidal activity of the DH-e strain against different algal species after 8 d. (a) Five algae species were treated by DH-e strain, (b) three dinoflagellate species were treated by DH-e strain (p > 0.05, Shapiro–Wilk test on SPSS; P < 0.01, one-way ANOVA on SPSS). All data were mean ± SD (n = 3).

MICs of crude bacteria extract against the three dinoflagellates

To determine MICs of crude bacteria extract, the crude extract solutions were added during the logarithmic-phase of P. donghaiense, A. tamarense (ATDH-47), K. mikimotoi Hansen growth at serial concentrations ranging from 0.7 to 7.0 mg/100 mL in duplicate flasks (Figure 3). Inhibition concentration is defined as the concentration of the inoculated extract in the case that the algae species concentration does not increase when compared with the initial inoculation concentration. A greater concentration increased the algicidal activity of bacteria against the three dinoflagellates. All gradient concentrations of the crude bacteria extracts, compared with the control 1, showed significant algicidal activity after 8 d (F = 2,083.128 > F0.01(5,12) = 5.06, p < 0.01) (Figure 3(a)), where all of the P. donghaiense were killed within 6 d at concentrations of 70.0 μg/mL and 35.0 μg/mL. For P. donghaiense, the minimum inhibitory concentration of the crude extract solution was approximately 35.0 μg/mL. For A. tamarense (ATDH-47), the crude extract solutions with concentration of 35.0 μg/mL and 70.0 μg/mL exhibited significant algicidal activities after 8 d (F = 839.270 > F0.01(5,12) = 5.06, p < 0.01) (Figure 3(b)). The MIC of the crude extract solution also was 35.0 μg/mL. The growth rates of K. mikimotoi Hansen were much lower than that of the control 1 at the eighth day, although the numbers of K. mikimotoi Hansen increased in all experimental groups compared to initial values (F = 728.691 > F0.01(5,12) = 5.06, p < 0.01) (Figure 3(c)). The MIC of the crude extract solution was 70.0 μg/mL and also exhibited the strongest inhibition on the number increase of K. mikimotoi Hansen cells. The MICs of the crude extract solutions on the cells of the three dinoflagellate species range from 35.0–70.0 μg/mL.

Figure 3

Minimum inhibitory concentrations (MICs) of bacterial metabolites on the cells of (a) P. donghaiense, (b) A. tamarense (ATDH-47) and (c) K. mikimotoi Hansen. All data were mean ± SD (n = 3) (p > 0.05, Shapiro–Wilk test on SPSS; P < 0.01, one-way ANOVA on SPSS).

Figure 3

Minimum inhibitory concentrations (MICs) of bacterial metabolites on the cells of (a) P. donghaiense, (b) A. tamarense (ATDH-47) and (c) K. mikimotoi Hansen. All data were mean ± SD (n = 3) (p > 0.05, Shapiro–Wilk test on SPSS; P < 0.01, one-way ANOVA on SPSS).

Short-term inhibitory effect of bacterial metabolites on P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen

The short-term (5 min) inhibitory experiments was conducted to evaluate the possibility of practical application of algicidal bacteria for the mitigation of HABs (Figure 4). In the control group, the algae cells showed complete and normal state (Figure 4(a), 4(e) and 4(k)). In the experimental group, the swimming movements of the P. donghaiense cells were inhibited when the metabolites started to work. As time increased, some cell walls detached from the main cellular bodies or were disrupted, and cellular substances were released and decomposed (Figure 4(b)–4(e)). Similar phenomena were found in A. tamarense (ATDH-47) (Figure 4(g)–4(j)) and K. mikimotoi Hansen (Figure 4(l)–4(o)). Compared with the control groups, there was almost no mobile algae in the experimental groups after 5 min.

Figure 4

Different lysis shapes of P. donghaiense ((a)–(e), (a) is normal cell, others are pathological cells), A. tamarense (ATDH-47) ((f)–(j), (f) is normal cell, others are pathological cells), and K. mikimotoi Hansen ((k)–(o), (k) is normal cell, others are pathological cells) photographed at 40× magnification.

Figure 4

Different lysis shapes of P. donghaiense ((a)–(e), (a) is normal cell, others are pathological cells), A. tamarense (ATDH-47) ((f)–(j), (f) is normal cell, others are pathological cells), and K. mikimotoi Hansen ((k)–(o), (k) is normal cell, others are pathological cells) photographed at 40× magnification.

Characterization of the enzyme activities of the three sentive dinoflagellates

After 1 h, the catalase (CAT) activity values of P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen were 1.18, 1.07, 1.20 times of the control value, respectively (Figure 5(a)). The CAT content of the three dinoflagellate species after 24 h were distinctly lower than that of the control (t = −5.176, df = 2, p = 0.029 < 0.05 for P. donghaiense; t = −33.837, df = 2, p = 0.001 < 0.01 for A. tamarense (ATDH-47); t = −16.408, df = 2, p = 0.003 < 0.01 for K. mikimotoi Hansen) (Figure 5(b)). The CAT activity values of P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen were 0.78, 0.73, 0.71 times of the control value, respectively. Alkaline phosphatase (AKP) activities of the different treatment groups varied based on filtrates (Figure 5(c) and 5(d)). The AKP content of the three dinoflagellate species were descending after 1 h, but the effect was not significant (t = −3.809, df = 2, p = 0.063 > 0.05 for P. donghaiense; t = −2.496, df = 2, p = 0.130 > 0.05 for A. tamarense (ATDH-47); t = −2.350, df = 2, p = 0.143 > 0.05 for K. mikimotoi Hansen). After 24 h, all three treatment groups showed a significant decrease of AKP activities incubation with DH-e filtrate (t = −63.548, df = 2, p < 0.01 for P. donghaiense; t = −58.467, df = 2, p < 0.01 for A. tamarense (ATDH-47); t = −17.536, df = 2, p = 0.003 < 0.05 for K. mikimotoi Hansen). The AKP content values of P. donghaiense, A. tamarense (ATDH-47) and K.=mikimotoi Hansen were 0.68, 0.48, 0.49 times of the control value, respectively. T-ATP activities of algal cells changed slower than CAT activities (Figure 5(e) and 5(f)). The acetone peroxide (T-ATP) content of the three dinoflagellate species were basically unchanged after 1 h, and T-ATP activities were significantly decreased after 24 h (t = 59.295, df = 2, p < 0.01 for P. donghaiense; t = 72.181, df = 2, p < 0.01 for A. tamarense (ATDH-47); t = −14.698, df = 2, p = 0.005 < 0.01 for K. mikimotoi Hansen). The T-ATP content values of the three dinoflagellate species were 0.65, 0.67, 0.70 times of the control value, respectively. Nitrite reductase (NR) activities shared a similar pattern with the T-ATP activities (Figure 5(g) and 5(h)). After 1 h and 24 h of exposure, NR activity values declined from 0.91 to 0.48 times, 1.01 to 0.51 times and 0.99 to 0.55 times, respectively.

Figure 5

Activities of basic enzymes, including CAT, AKP, T-ATP synthetase and NR, in P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen after 1 h ((a), (c), (e), (g)) and after 24 h ((b), (d), (f), (h)) at a DH-e metabolite concentration of 0.70 mg/mL. Control group represents 100 mL algal culture treated 10 mL of aseptic seawater. Experiment group represents 100 mL algal culture treated 10 mL of crude extract solution. All data were mean ± SD (n = 3) (p > 0.05, Shapiro–Wilk test on SPSS). *Represents a statistically significant difference of p < 0.05 (on based Student's t-test) when compared to the control group; **represents a statistically significant difference of p < 0.01 (on based Student's t-test); ns represents no statistical difference.

Figure 5

Activities of basic enzymes, including CAT, AKP, T-ATP synthetase and NR, in P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen after 1 h ((a), (c), (e), (g)) and after 24 h ((b), (d), (f), (h)) at a DH-e metabolite concentration of 0.70 mg/mL. Control group represents 100 mL algal culture treated 10 mL of aseptic seawater. Experiment group represents 100 mL algal culture treated 10 mL of crude extract solution. All data were mean ± SD (n = 3) (p > 0.05, Shapiro–Wilk test on SPSS). *Represents a statistically significant difference of p < 0.05 (on based Student's t-test) when compared to the control group; **represents a statistically significant difference of p < 0.01 (on based Student's t-test); ns represents no statistical difference.

DISCUSSION

In coastal seawater, where red tides occur frequently, marine bacteria may control phytoplankton blooms (Hu et al. 2015). Bacteria are closely associated with and thought to influence phytoplankton population dynamics and toxicity (Gutierrez et al. 2014; Bunse et al. 2016). Most marine algicidal bacteria have been characterized as either Bacteroidetes or Gammaproteobacteria (Gong et al. 2016). In this study, according to the phylogenetic analysis of 16S rDNA sequence, the strain DH-e was identified as the Halomonas sp. belonging to Gammaproteobacteria (Figure 1). Shi et al. (2013) found that three bacterial strains P5, N5 and N21 lysed algae through direct attack. Strain P5 most likely killed the algal cells by attaching and decomposing the cells directly, whereas strains N5 and N21 killed the algal cells by attachment or penetration, causing the algal cells to inflate and ultimately lyse. An indirect attack refers to a process in which an extracellular product is responsible for cell lysis (Yang et al. 2014). To understand which mechanism is used by DH-a, DH-b, DH-c, DH-d and DH-e, bacterial cells and cell-free filtrate tested separately (Table 1). Cell-free filtrate of strain DH-e showed the algicidal activity was up to approximately 92.0%. The results indicated that DH-e might produce extracellular bioactive molecules and release dissolved A. tamarense (ATDH-47) by indirect attacking. In general, algicidal bacteria can act on a range of algal species, from one to several, also belonging to different taxonomic groups (Guan et al. 2014). Our results showed there was no distinctly negative impact on growth compared with the control groups (Figure 2(a)). The above indicated that the metabolites of DH-e only had specific algicidal effects on the three toxic dinoflagellate species (Figure 2(b)). These results may provide useful options to control the red tide caused by these three dinoflagellates and have little effect on other species of algae.

The isolation, purification, and characterization of algicidal compounds are difficult, due to their various characteristics across different species of algicidal bacteria. Some research demonstrated that algicidal compounds include proteases, peptides, biosurfactants, antibiotic-like substances, pigments and so on (Li et al. 2015). In the present study, culture supernatant filtrate was dissolved with some reagents after freezing and drying. The supernatant of culture dissolved in acetic acid showed the strongest algicidal activity (data not shown). This algicidal activity suggests that crude extract solution produced by DH-e are aciduric. However, what exactly the crude extract solution is needs to be further studied.

In the present study, the algicidal effects of crude extract solution were concentration-dependent. The initial concentration of crude extract solution produced by DH-e added to P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen was an important factor that influenced the algicidal effectiveness (Figure 3). A greater concentration increased the algicidal activity against three dinoflagellate species. The concentrations of crude extract solution treated with acetic acid that make the algal concentration of P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen no longer increase are 35.0 μg/mL, 35.0 μg/mL and 70.0 μg/mL, respectively. The results indicated that the trace crude extract solutions also have the strong inhibition algal effect. What is more, the results of experiments showed that DH-e metabolites could kill or irreversibly damage algae within 5 min (Figure 4), which greatly increases metabolites applicability in practice as algicides must be rapidly effective following dilution with seawater. We expect that the algicidal bacterium, Halomonas sp. DH-e, can be used as the engineering bacteria to control red tide.

The reactive oxygen species (ROS) are the byproducts of photosynthesis when the photosynthetic system is not functioning normally (Liu et al. 2017). As an important antioxidant enzyme, CAT level of the algae was increased to neutralize the residual ROS (Zhang et al. 2013). Consequently, we measured the content of CAT to demonstrate the response of the antioxidant system. In the present study, exposure to crude extract solution from the strain DH-e increased the levels of CAT in P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen cells with1 h (Figure 5(a)), which indicated that crude extract solution from the strain DH-e induced membrane lipid peroxidation and caused oxidative damage on cell membranes (Figure 4). However, the levels of CAT in P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen cells decreased after 24 h (Figure 5(b)), and it indicated that the number of algal cells was decreasing. The activity of algal cells is very closely related to the nutrient metabolism of phosphorus in organism, and AKP can promote the utilization of the inorganic phosphorus in the environment by the cell body (Feng et al. 2015). In our results, the levels of AKP in P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen cells decreased after 1 h and 24 h (Figure 5(c) and 5(d)), which indicated that the algal cells' damage might be serious, which obstructed the phosphatase synthesis and the activity decreased. T-ATP synthetase and NR are major enzymes in the growth of algae, whose content is affected by external conditions and the growth status of algae. In this study, T-ATP synthetase, and NR showed no obvious decrease or increase compared with the control groups after 1 h (Figure 5(e) and 5(g)). It indicated that the direct target of crude extract solution from the strain DH-e was not energy metabolize and nutrient uptake. To summarize, on the basis of the above experiments, it can be inferred that the metabolites might have damaged the algae by breaking the algae cell wall structure and impacting its photosynthetic system.

CONCLUSION

Bacterium DH-e, which is a Halomonas species, strongly inhibited the red-tide forming marine dinoflagellates P. donghaiense, A. tamarense (ATDH-47) and K. mikimotoi Hansen, especially in the short term. Bacterium DH-e killed the dinoflagellate cells in the absence of physical contact by releasing algicidal compounds that cause oxidative damage, showing potential as a biological algal suppressant. Further studies are needed to pinpoint the mechanism(s) of dinoflagellate inhibition by DH-e, and to refine its potential as a biocontrol agent for red-tide dinoflagellates.

ACKNOWLEDGEMENTS

This work was supported by the National Natural Science Foundation of China under contract No. 41230961; the Open Fund Project of Marine Biotechnology Key Laboratory of Guangdong Province in China under contract GPKLMB2. 01201; and Shantou University Campus Research Fund NFC1400. We thank Prof. Yang Weidong from Jinan University.

REFERENCES

REFERENCES
Dabas
P.
,
Tripathi
S.
,
Chauhan
S.
,
Malik
I.
,
Kumar
N.
,
Kumari
S.
,
Bhatnagar
S.
&
Deepali
,
, .
2014
Cyanobacteria – an unexplored source for biomolecules: a review
.
Advanced Science
6
(
6
),
921
930
.
Damanka
S.
,
Dennis
F. E.
,
Agbemabiese
C.
,
Lartey
B.
,
Adiku
T.
&
Nyarko
K.
2016
Identification of OP354-like human rotavirus strains with subtype P[8]b in Ghanaian children with diarrhoea
.
Virology Journal
13
(
1
),
1
6
.
Feng
T. Y.
,
Yang
Z. K.
,
Zheng
J. W.
,
Xie
Y.
,
Li
D. W.
&
Murugan
S. B.
2015
Examination of metabolic responses to phosphorus limitation via proteomic analyses in the marine diatom Phaeodactylum tricornutum
.
Scientific Reports
5
(
10373
),
10373
.
Hoagland
P.
&
Scatasta
S.
2016
The economic effects of harmful algal blooms
.
Ecology of Harmful Algae
619
(
2
),
391
402
.
Hu
X.
,
Yin
P.
,
Zhao
L.
&
Yu
Q.
2015
Characterization of cell viability in Phaeocystis globosa cultures exposed to marine algicidal bacteria
.
Biotechnology & Bioprocess Engineering
20
(
1
),
58
66
.
Lee
K. H.
,
Jeong
H. J.
,
Kwon
J. E.
,
Kang
H. C.
,
Kim
J. H.
&
Jang
S. H.
2016
Mixotrophic ability of the phototrophic dinoflagellates Alexandrium andersonii, A. affine, and A. fraterculus
.
Harmful Algae
59
,
67
81
.
Lei
X.
,
Li
Y.
,
Chen
Z.
,
Zheng
W.
,
Lai
Q.
&
Zhang
H.
2014
Altererythrobacter xiamenensis sp. nov., an algicidal bacterium isolated from red tide seawater
.
International Journal of Systematic and Evolutionary Microbiology
64
(
Pt 2
),
631
637
.
Li
Z.
,
Geng
M.
&
Yang
H.
2015
Algicidal activity of Bacillus sp. Lzh-5 and its algicidal compounds against Microcystis aeruginosa
.
Applied Microbiology & Biotechnology
99
(
2
),
1
10
.
Liu
T. T.
,
Peng
Y. F.
,
Liu
L.
&
Xia
X. S.
2015
Isolation and characterization of the algicidal bacterium DCJ-2
.
Advanced Materials Research
1065–1069
,
3077
3082
.
Yang
X.
,
Li
X.
,
Zhou
Y.
,
Zheng
W.
,
Yu
C.
&
Zheng
T.
2014
Novel insights into the algicidal bacterium DH77-1 killing the toxic dinoflagellate Alexandrium tamarense
.
Science of the Total Environment
482–483(1)
,
116
.