Isolation and characterization of a bio-agent antagonistic to diatom, Stephanodiscus hantzschii

Authors


M.-S. Han, Department of Life Science, Hanyang University, Seoul 133-791, Korea (e-mail: hanms@hanyang.ac.kr).

Abstract

Aims:  Identification of bacterium HYK0203-SK02 and its lysis of Stephanodiscus hantzschii.

Methods and Results:  In an effort to identify a bio-agent capable of controlling S. hantzschii blooms, we used the algal lawn method to identify 76 bacteria in relevant water samples. Of these, the seven isolate showed algicidal activity against S. hantzschii; isolate HYK0203-SK02 exhibited the strongest algicidal activity, and was used for further analysis. 16S rDNA sequencing of this isolate allowed us to identify HYK0203-SK02 as a strain of Pseudomonas putida (99·2%). Growth of S. hantzschii was strongly suppressed by bacteria in all growth phases, with the strongest algicidal activity noted against diatoms in the exponential stage (5–18 days). Host range assays revealed that isolate HYK0203-SK02 also strongly inhibited the growth of Microcystis aeruginosa, but stimulated growth of the diatom Cyclotella sp., which has a similar structure to that of S. hantzschii. Biochemical assays revealed that the algicidal substance seemed to be localized in the cytoplasmic membrane of this newly identified algicidal bacterium.

Conclusion:  The algicidal bacteria P. putida HYK0203-SK02 caused cell lysis and death of not only diatom S. hantzschii but also cyanobacteria M. aeruginosa, dramatically. Algicidal substance might be located at the compartment of cytoplasmic membrane.

Significance and Impact of the Study:  Taken together, our results indicate that P. putida HYK0203-SK02 may be a potential bio-agent for future use in controlling freshwater diatomic blooms.

Introduction

The small centric diatom Stephanodiscus hantzschii, which is generally considered an indicator of eutrophication, flourishes worldwide in most rivers, reservoirs and lakes during periods of water shortage and low temperatures (Kilham et al. 1986; Leitao 1995; Ha et al. 2002). Blooms are seen annually in the Pal'tang Reservoir and Naktong River, South Korea, mainly from late autumn to the following spring (Cho et al. 1998; Ha et al. 2002; Han et al. 2002; Hong et al. 2002a). These blooms cause significant decreases in dissolved oxygen, water transparency and recreational amenity (Kolmakov et al. 2002), negatively impacting the drinking water supply (Oksiyuk 1965; Sakevich 1970; Lee et al. 2001), and may cause expensive problems at water treatment plants such as filter clogging and reduced efficiency of coagulation and sedimentation (Lim et al. 2000). Because of this, researchers are currently investigating various methods for controlling algal blooms through physical, chemical or biological means (Sigee et al. 1999; Jeong et al. 2000). Of these, biological control agents (bio-agents) such as viruses, fungi, bacteria, actinomycetes and protozoa (Sigee et al. 1999) are of particular interest. Recent work has focused on the identification of bacteria capable of inhibiting or degrading algal blooms in marine and freshwater environments (Lovejoy et al. 1998; Wu et al. 1998; Doucette et al. 1999; Imai et al. 2001; Manage et al. 2001). However, although a variety of bacteria have been tested against freshwater algal blooms (Manage et al. 2000; Hong et al. 2002b; Jang et al. 2003; Kim and Han 2003), to date, no bacterium has been shown to be effective against diatomal blooms in freshwater.

In an effort to identify and characterize a bio-agent, we sought (i) to isolate and identify diatom-lysing bacteria effective against S. hantzschii, (ii) to evaluate the algicidal activity of the isolated bacteria, and (iii) to examine the cellular localization of the relevant algicidal substances.

Materials and methods

Diatom culture

Stephanodiscus hantzschii UTCC 267 was purchased from the University of Toronto Culture Collection of Algae and Cyanobacteria (UTCC) in Canada. The diatom was maintained as unialgal axenic culture at 20°C, pH 7, under 50 μ mol photons m−2 s−1 with a 12 : 12 (light : dark) cycle The diatom cells were incubated in diatom medium (DM) adjusted to pH 6·9 containing 2·00 × 10−2 g Ca(NO3)2· 4H2O, 1·24 × 10−2 g KH2PO4, 2·5 × 10−2 g MgSO4·7H2O, 1·09 × 10−2 g NaHCO3, 2·25 × 10−3 g EDTA-Fe-Na, 2·25 × 10−3 g EDTA-Na2, 2·48 × 10−3 g H2BO3, 1·39 × 10−3 g MnCl2·4H2O, 1·00 × 10−3 g (NH4)6Mo7O24·4H2O, 4·00 × 10−5 g cyanocobalamin, 4·00 × 10−5 g thiamine HCl, 4·00 × 10−5 g biotin and 5·7 × 10−2 g NaSiO3·9H2O in 1 l DW (Beakes et al. 1988) and transferred to fresh DM media once 2 week.

Isolation and identification of algicidal bacteria

Pseudomonas putida HYK0203-SK02 with algicidal effects against S. hantzschii was isolated by a modified soft-agar over-layer technique (Sakata et al. 1991). Surface water samples were collected from the Pal'tang Reservoir, Korea (from February to November 2002), and filtered through 0·8 μm-nucleopore membrane filters. For generation of the diatomic lawn, S. hantzschii was cultured for 20 days, harvested by centrifugation at 18 000 g for 20 min, spread on liquid DM soft agar (1·0% agar) and equilibrated to 50°C. A fixed quantity (20 ml) of the diatom mixture was poured onto DM bottom agar (1·5% agar) and solidified for 2 days under the algal growth conditions described above. The surface water filtrates (200 μl) were spread on the S. hantzschii algal lawn, and the plates were incubated for 10 days under algal growth conditions. Micro-organisms forming a clear zone around the colonies were picked, incubated at 30°C in the dark for 2 days and purified as described previously (Yamamoto and Suzuki 1990). In this manner, we successfully isolated the algicidal bacterium HYK0203-SK02, which showed significant effects against S. hantzschii. Once purified, the bacterium was axenically maintained in the dark on nutrient agar (NA) plates containing 1·5% agar, at 30°C and pH 7, or was cryopreserved at −76°C in nutrient broth (NB) medium containing 20% glycerol.

For identification of isolate HYK0203-SK02, bacterial chromosomal DNA was isolated as previously described (Hong et al. 2002b).The 16S rDNA was PCR amplified with primers 27F, 5′-AGAGTTTGATCATGGCTCAG-3′ and 1492R, 5′-GGTTACCTTGTTACGACTT-3′ in 50 μl reactions containing 20 ng template DNA, 1X PCR buffer, 5 mm MgCl2, 10 mm dNTP, 10 pm of each primer and 2·5 units of Taq DNA polymerase. The PCR was run with 35 thermal cycles of denaturation for 1 min for 94°C, annealing for 2·5 min at 55°C, extension for 2·5 min for 72°C, and with a final elongation step of 7 min at 72°C in a DNA thermal cycler, (Genetic analyser 377; Perkin-Elmer, Boston, MA, USA), employing the thermal profile. The PCR products were purified with the QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) and sequenced by automated DNA sequencer (Bionex, Seoul, Korea) using the SequiTherm EXCEL II Labelled Primer Sequencing Kit (LI-COR Inc., Lincoln, NE, USA) with the T3F and M13R primers. Identification of the HYK0203-SK02 isolate was determined as compared the full length sequencing of this bacterium with a collection of 16S rDNA obtained from the DDMJ/EMBL/Genbank database. Our sequencing and comparison of the 16S rDNA gene of this isolate indicated that it was most closely related to P. putida strain ATCC 17485 (99·2% homology).

16S rDNA sequence of strain HYK0202-SK02 has been deposited in the GenBank database under accession numbers AY738626, and this bacterium culture on NA agar plate is deposited in the Korean Culture Center of Microorganism (KCCM), South Korea, under accession number KCCM-10464.

Algicidal range of the HYK0203-SK02

To examine the host range of isolate number, we evaluated changes in the biomass of several algae and cyanobacteria in the presence of HYK0203-SK02. The bacterium was incubated at 29°C under 120 rev min−1 in NB medium for 24 h, harvested by centrifugation at 18 000 g for 20 min, and adjusted to 1·8 at A660 nm with sterilized DM medium (yielding an initial density of ca 1 × 108 CFU ml−1). Bacterial suspensions (1 ml) were inoculated into test tubes containing 19 ml of exponential-phase cultures of the following test-cultures of Chlorella sp., Microcystis aeruginosa NIES 298 and NIES 44, Anabaena cylindrica NIES 19, A. macrospora ATCC 22664, Oscillatoria sancta NIES 10027, Synechococcus bacillaris CCAP 1479/7, Coelastrum sp., S. hantzschii UTCC 267 and CCAP 1709/4, Cyclotella sp. HYK0210-A1 and, Aulacoseira granulata CCAP 1002/1. Green algae and cyanobacteria were cultured in BG11 medium adjusted to pH 9 (1·5 g NaNO3, 3·0 × 10−2 g K2HPO4, 7·5 × 10−2 g MgSO4·7H2O, 3·6 × 10−2 g CaCl2·2H2O, 6·0 × 10−3 g EDTA, 1·0 × 10−3 g Na2CO4, 2·0 ×10−2 g citric acid, 2·86 × 10−3 g H3BO3, 1·81 × 10−3 g MnCl2·4H2O, 2·22 × 10−4 g ZnSO4·7H2O, 3·90 × 10−4 g NaMoO4, 7·9 × 10−5 g CuSO4·5H2O and 4·94 × 10−5 g Co(NO3)2·6H2O in 1 l DW, Rippka et al. 1979) at 30°C, pH 9 and with 120 rev min−1 shaking, under 50 μmol photons m−2 s−1 on a 12 : 12 (light : dark) cycle. The two diatoms were cultured under the same conditions as S. hantzschii. All cultures of algae and cyanobacteria were unialgal axenic culture except for the Coelastrum sp., Chlorella sp. and S. hantzschii CCAP 1709/4. Algae were counted daily up to a period of 2 days as described below.

Bacterial density and algicidal effect

To elucidate the algicidal activity of the different densities of the bacterium HYK0203-SK02 against S. hantzschii, the bacterial culture was prepared as above and then serially diluted with DM (5 ml) to initial densities of 103, 104, 105, 106, 107 and 108 CFU ml−1 in 100 ml of S. hantzschii culture (1·1 × 106 cells ml−1) at mid-exponential phase (13 days after incubation). Algae were counted daily up to a period of 25 days, as described below.

The effects of diatom and bacterial growth phases on the observed algicidal ability

The algicidal activities of HYK0203-SK02 against different growth phases of S. hantzschii were studied. The bacterial culture was prepared as above, adjusted to a density of 1·8 at A660 nm, and 2·5 ml of the bacterial suspension was inoculated into 100-ml flasks containing 50 ml of lag, exponential or stationary phase of S. hantzschii (2·3, 20·7 and 22·7 × 105 cells ml−1 respectively).

In addition, we studied the algicidal activity of HYK0203-SK02 in different growth phases (lag, exponential and stationary phase) against exponential phase S. hantzschii. Aliquots of HYK0203-SK02 were incubated for 3 (lag phase), 18 (exponential phase) and 36 (stationary phase) hours in 100-ml flasks containing 50 ml of NB, harvested by centrifugation at 18 000 g for 20 min, and adjusted up to 1·8 at A660 nm. These samples (2·5 ml; initial density ca 1 × 108 CFU ml−1) were inoculated into 100-ml flasks containing 50 ml of mid-exponential phase S. hantzschii (13·5 × 105 cells ml−1; cultured for 13 days). Algae were counted daily, as described below.

Analysis of algicidal activity

Algae and cyanobacteria were quantified by haemocytometer under a light microscope and bacteria were counted by CFU method. The algicidal activity of isolate HYK0203-SK02 was calculated by the following equation: Algicidal effect (%) = (1−Tt/Ct) × 100, where T (treatment) and C (control) are the S. hantzschii cell densities with and without HYK0203-SK02, respectively, and t is the inoculation time. After the isolate HYK0203-SK02 harvested by centrifugation was suspended into fresh DM, a suitable volume of this bacterium was inoculated into S. hantzschii culture in the treatment. In the control, was the equal volume the fresh DM used in treatment inoculated into S. hantzschii culture. All experiments were repeated in triplicate and results are given by mean and standard deviation of raw data.

Scanning electron microscopy (SEM)

For SEM analysis, a 0·2 ml sample from the growing cultures of S. hantzschii was mixed with the isolate HYK0203-SK02 was subsampled, fixed with 2% of glutaraldehyde, and prepared by critical-point drying overnight. The samples were coated with gold for 5 min and examined in a XL series scanning electron microscope (Philips, Eindhoven, the Netherlands) with an accelerating voltage of the electron beam up to 120 keV.

Preparation of cell-free extracts and culture supernatants

Cell-free extracts were separated according to the method of Niviere et al. (1986). For preparation of cell-free extract, cells were incubated in NB medium for 24 h at 30°C and pH 7·0, harvested by centrifugation at 750 000 g for 30 min, and suspended in 10 mm phosphate-citrate buffer (pH 7·0). The bacteria were homogenized with a mortar and pestle for 40 min at 4°C. The cellular debris was removed by centrifugation at 1 400 000 g for 2 h at 4°C and bacteria were completely removed through 0·2-μm polycarbonate filters to prepare cell-free extracts and supernatants. Finally, acetone was added to the clear supernatant (crude extract) at 10% (v/v) and the mixture was immediately assayed for degradation activity.

Preparation of cell fractions

Periplasm fraction was prepared by the method of van der Western et al. (1978), and the other cell fractions were separated according to the method of Niviere et al. (1986). A bacterial pellet (7 g; prepared as above) was suspended in phosphate buffer (pH 9) containing 50 mm EDTA and incubated at 40°C for 30 min. The periplasm fraction was obtained by centrifugation at 18 000 g for 30 min at 4°C. The pelleted spheroplast fraction was homogenized with a mortal and pestle for 90 min in 10 mm phosphate buffer (pH 7) at 5°C. The homogenate was centrifuged at 30 000 g and the supernatant further centrifuged at 1 500 000 g for 5 h at 4°C. The supernatant was used as the cytoplasmic fraction. The pellet was resuspended in 10 mm phosphate buffer (pH 7), and represented the cytoplasmic membrane fraction.

Preparation of crude cell extracts of S. hantzschii

Mid-exponential phase S. hantzschii cells were harvested by centrifugation at 8000 g for 30 min at 4°C, washed twice with sterilized water and dried at 4°C. Dried cells (2 g) were suspended in 200 ml of 4% SDS, boiled at 100°C for 30 min, harvested by centrifugation at 120 000 g for 30 min, and cooled to room temperature. This process was repeated five times, and then the S. hantzschii cell extracts were suspended in 400 ml of boiled distilled water, harvested and cryopreserved at −20°C until they were used for measurement of specific activity.

Assay of protein content and degradation activity

The protein concentration of protein was determined by the Bradford method, using bovine serum albumin as the standard (Bradford 1976). For the measurement of the ability of the bacteria to degrade the algae, the HYK0203-SK02 cell fractions and cell free-extract were adjusted to 1·2 at A660 in 25 mm Tris–HCl buffer (pH 8·0), inoculated into crude cell extracts of S. hantzschii, and incubated at 40°C for 30 min. The assay of protein content, degradation activity and specific activity were determined by modified method of Kim et al. (2002) as below. The degradation activity was calculated by comparison of absorption at 660 nm before and after the reaction. One unit of degradation activity indicates the ability to decrease absorbance 0·001 in 1 min of reaction time. The specific activity of HYK0203-SK02 was calculated by the following equation: specific activity (unit mg−1) = the degradation activity (unit)/the protein concentration (mg).

Results

Seventy-six bacterial isolates identified by different colony morphologies on agar plates (Leboffe and Pierce 1999) were screened in surface water samples from the Pal'tang Reservoir, Korea, taken from February to November in 2002. Only seven bacterial isolates of bacteria exhibited the inhibitory activity towards the diatom S. hantzschii UTCC 267. Of these, isolate P. putida HYK0203-SK02 appeared to have the strongest algicidal activity when 5 ml (1 × 108 CFU ml−1) of those isolates were inoculated into 100 ml of S. hantzschii culture (1 × 106 cells ml−1) a second time for 10 days. This bacterium was a Gram-negative, rod type, nonpigmented and showed slow gliding in NB medium with 1·5% agar. The optimal temperature and pH for the growth ranged between 10 and 30°C, and 5 and 9 respectively (data not shown).

The isolate exhibited algicidal activity against S. hantzschii (average reduction in abundance: 80·4%), two strains of Microcystis aeruginosa (70·0 and to 71·4%), Aulacoseira granulata (31·2%), Anabaena cylindrica (41·6%) and Chlorella sp., (30·3%). Interestingly, the diatom Cyclotella sp. HYK0210-A1, which is structurally similar to Stephanodiscus, was stimulated by the isolate, not inhibited (Table 1).

Table 1.  Algicidal effect of the isolate HYK0203-SK02 against several algal strains
StrainAlgicidal effect (%)
  1. *Coelastrum sp., Cyclotella sp. HYK0210-A1 and Chlorella sp. were isolated from the Gyeongan stream in Korea. The algicidal activity of isolate HYK0203-SK02 was calculated by the following equation: Algicidal effect (%) = (1−Tt/Ct) × 100, where T (treatment) and C (control) are the S. hantzschii cell densities with and without HYK0203-SK02, respectively, and t is the inoculation time (t = 2 days).

  2. NIES, National Institute of Environmental Studies, Japan; ATCC, American Type Culture Collection, Manassas, VA, USA; CCAP, Culture Collection of Algae and Protozoa, USA; UTCC, University of Toronto Culture Collection of Algae and Cyanobacteria, Canada.

  3. Data are the mean ± s.d. from at least three independent assays.

Chlorella sp.*30·3 ± 9·5
Coelastrum sp.*0 ± 2·0
Microcystis aeruginosa NIES 29871·4 ± 12·1
Microcystis aeruginosa NIES 4470·0 ± 12·0
Anabaena cylindrica NIES 1941·6 ± 6·7
Anabaena macrospora ATCC 226642·0 ± 2·0
Oscillatoria sancta NIES 100274·4 ± 3·2
Synechococcus bacilaris CCAP 1479/79·8 ± 0·4
Stephanodiscus hantzschii CCAP 1709/480·4 ± 2·5
Stephanodiscus hantzschii UTCC 26780·4 ± 2·0
Cyclotella sp. HYK0210-A1*−23·5 ± 0·9
Aulacoseira granulata CCAP 1002/131·2 ± 5·1

To determine the effective algicidal threshold density of the isolate HYK0203-SK02 isolate against S. hantzschii, S. hantzschii was incubated with six different initial densities of the bacterium (103, 104, 105, 106, 107, and 108 CFU ml−1) (Fig. 1). The bacterial densities of <106 CFU ml−1 showed little or no algicidal activity against S. hantzschii; initial bacterial cell densities of 106, 107 and 108 CFU ml−1, showed algicidal activities of 33·4, 61·5 and 85·5% (percentage decrease in cell count) respectively. However, Isolate HYK0203-SK02 effectively inhibited growth of S. hantzschii during all three diatom growth phases, showing biomass decreases of 67·7% (lag phase), 86·9% (exponential phase) and 75·1% (stationary phase) for S. hantzschii (Fig. 2), which they were significantly different at 95% level. HYK0203-SK02 of different growth phases consistently inhibited mid-exponential S. hantzschii, yielding inhibitions of 85·4% (lag), 86·5% (exponential) and 81·6% (stationary) (Fig. 3).

Figure 1.

Density of Stephanodiscus hantzschii in cultures incubated with various concentrations of algicidal bacterium HYK0203-SK02 (∘: 108, bsl00072: 107, bsl00000: 106, bsl00001: 105, bsl00000: 104, bsl00000: 103 CFU ml−1 and •: without the bacteria). An arrow indicates the time of bacterium inoculation. Data are the mean ± s.d. from at least three independent assays

Figure 2.

Density of Stephanodiscus hantzschii in cultures incubated with (∘) algicidal bacterium HYK0203-SK02 at three different growth phases of S. hantzschii (a, lag phase; b, exponential phase; c, stationary phase) and without (•) algicidal bacteria. The arrow indicates the time of bacterium inoculation. Data are the mean ± s.d. from at least three independent assays

Figure 3.

Density of Stephanodiscus hantzschii in cultures incubated with algicidal bacterium HYK0203-SK02 at three different growth phases of algicidal bacteria (∘: lag phase, bsl00072: exponential phase and bsl00000: stationary phase) and without the algicidal bacteria (•). The arrow indicates the time of bacterium inoculation. Data are the mean ± s.d. from at least three independent assays

No immediate morphological change was observed in the diatom when HYK0203-SK02 was inoculated into S. hantzschii culture (Fig. 4a). Over time, the bacteria gradually surrounded the S. hantzschii cells until the entire shell was covered with bacteria. The swarming bacteria breached the cell wall of the diatom within 4 h (Fig. 4b), and subsequently the diatom chloroplasts and cytoplasm (Fig. 4c). By 12 h after inoculation, the diatom cell structure was unrecognizable (Fig. 4d). Scanning electron microscopy revealed that the diatom cells were heavily deconstructed in the light of silicate frustle (Fig. 5). We next examined the localization of the algicidal substance capable of degrading S. hantzschii, by testing cell culture and or cell-free extracts of HYK0203-SK02 against crude cell extracts of S. hantzschii. We observed degradation in the presence of the cell-free extract (52·1 units), but not in the presence of the cell filtrates (Table 2). To fine-tune the localization, we then examined degradation of S. hantzschii by the various cell fractions of P. putida HYK0203-SK02 (Table 2). Although there were very slight differences in protein concentrations among the three cell factions, we observed significant differences in the degradative activities of the periplasm (0 unit), cytoplasm (32·4 units) and cytoplasmic membrane (361·5 units) fractions. In terms of specific activity, the cytoplasmic membrane showed higher specific activity (62·5 units) than did the periplasm (0 units) and cytoplasm (7·7 units) fractions.

Figure 4.

Light microscopic observation of diatom Stephanodiscus hantzschii in the absence (a) and presence (b–d) of isolate algicidal bacterium HYK0203-SK02 (b, 4 h; c, 6 h; d, 12 h). Scale bar = 10 μm

Figure 5.

Scanning electron micrographs of Stephanodiscus hantzschii in the absence (a) and presence (b) of algicidal bacterium HYK0203-SK02. Scale bar = 5 μm

Table 2.  Degradation of Stephanodiscus hantzschii by cell culture filtrates, cell-free extracts of isolate HYK0203-SK02 and various fractions of cell-free extracts of isolate HYK0203-SK02
Cell compartmentsCell fractionsDegradation activity (unit)Protein concentration (mg ml−1)Specific activity (unit mg−1)
  1. Data are the mean from at least three independent assays.

  2. The HYK0203-SK02 cell compartments and cell fractions were adjusted to 1·2 at A660 nm in 25 mm Tris–HCl buffer (pH 8·0), inoculated into crude cell extracts of S. hantzschii, and incubated at 40°C for 30 min. The degradation activity was calculated by comparison of absorption at 660 nm before and after the reaction. One unit of degradation activity indicates the ability to decrease absorbance 0·001 in 1 min of reaction time. The specific activity of HYK0203-SK02 was calculated by the following equation: specific activity (unit mg−1) = the degradation activity (unit)/the protein concentration (mg).

Cell filtrates 0·000·120·00
Cell-free extracts 52·1115·103·45
 Periplasm0·04·50·0
 Cytoplasm32·34·07·7
 Cytoplasmic membrane361·55·862·5

Discussion

Blooms of the diatom S. hantzschii degrade water quality and have adverse effects on ecosystems (Cho et al. 1998; Ha et al. 2002; Kolmakov et al. 2002; Kravchuk et al. 2002). Although this is a serious issue, no previous work has identified possible methods for controlling S. hantzschii blooms in the freshwater ecosystems. Here, we report for an algicidal bacterium that appears to act against S. hantzschii.

The genus Pseudomonas is known to effectively degrade cyanobacteria in eutrophic lake environments (Yamamoto et al. 1993; Kodani et al. 2002), and more particularly P. putida is a metabolically versatile bacteria that is known for its diverse metabolism, and potential for development of bioconversion to break down aromatic hydrocarbons such as polycyclic aromatic hydrocarbons, phenols and polychlorinated biphenyl (Billingsley et al. 1999; Kim et al. 2000; Manohar et al. 2001; Fernandes et al. 2002).

Among 12 algae and cyanobacteria selected in our initial screen, P. putida HYK0203-SK02 had the highest algicidal activity and was chosen for further study. This isolate proved to have a selective host range, suppressing growth of S. hantzschii and M. aeruginosa, while facilitating the growth of the diatom Cyclotella sp. HYK0210-A1, which is similar in structure to S. hantzschii. This is consistent with the results of Fukami et al. (1992), who also reported that bacteria may have selectively beneficial and/or harmful effects on specific algae. Natural bacteria can influence significantly the development and decline of algal bloom (Fukami et al. 1992, 1997). Laboratory studies presented that the different algicidal activities of HYK0203-SK02 may seem to have species- and/or strain-specific activity without regard to general taxonomical relationships or morphological characteristics. Host range test of algicidal bacteria plays an important role in application for biological control to algal freshwater blooms in field trials. Further research will demand on elucidation of attack type and interaction between algicidal substance and component of cell wall.

Doucette et al. (1999) suggested that the initial algicidal bacteria concentration has an influence on manifestation of algicidal activity, and Fraleigh and Burnham (1988) postulated that above threshold density of algicidal bacteria will occur, with little dependence on inorganic nutrient concentrations or host density. Our data indicate that the threshold density at which isolate HYK0203-SK02 had algicidal effects on S. hantzschii was 1 × 106 CFU ml−1, i.e. more algicidal activity at densities over this threshold was observed (Fig. 1). Algicidal activity depends on different stages of S. hantzschii (exponential > stationary > lag phase) (Fig. 2). In contrast, the algicidal activity of the bacteria differed little based on the growth stage of the bacteria itself (Fig. 3). Manage et al. (2001) reported that the physiological status of the host algae might play a crucially important role in the success of bacterial attack. Toncheva-Panova and Ivanova (2000) found that lysis effect of bacteria was accompanied by change of algal physiology. Thus, our results suggest that the physiological status of the algae, not the bacterium, is a crucial factor in growth inhibition of S. hantzschii. Why was bacterial attack successful in exponential phase? One possible explanation could be that cell wall of algae at which the exponential phase of algae repeats the most actively cell division than other phases, was easily destroyed by bacterial attack, similar to the results of Cole (1982).

Our microscopic observations (Fig. 4) revealed that the algicidal activity of the bacterium HYK0203-SK02 relies on cell-to-cell contact, with the bacteria rapidly besieging around the algal cell and lysing most of the S. hantzschii within 12 h. The algicidal mechanisms of bacteria generally proceed in one of three ways (Doucette et al. 1999; Sigee et al. 1999): direct attack (Imai et al. 1993), release of extracellular compounds (Lovejoy et al. 1998) or bacterial entrapment, which may involve lysis through release of lysozyme-like enzymes on the bacterial surface (Burnham et al. 1981). Here, we found that diatom S. hantzschii was rapidly lysed directly by bacterial entrapment.

As an early effort in identifying a lytic agent that may be produced by HYK0203-SK02, we investigated the cellular location of the algicidal substance capable of degrading S. hantzschii. Because of the entrapment strategy of this bacterium, we hypothesized that the algicidal substance was likely localized near the inner bacterial surface. No degradation activity was detected in the cell culture fraction, while the cell-free extract showed strong activity (Table 2). These results suggest that P. putida HYK03-SK02 likely possesses algicidal substances on the bacterial surface or the inside of the membrane. Indeed, we found strong lytic activity within the cytoplasmic membrane fraction (Table 2), confirming that the algicidal substance acting against S. hantzschii localizes to the cytoplasmic membrane.

We report the first identification of an algicidal bacterium that acts against a freshwater diatom. Moreover, this study confirms that the genus P. putida has a significant algicidal activity not only on cyanobacteria like M. aeruginosa but also on the diatom S. hantzschii. We identified the presence of an algicidal substance in the cytoplasmic membrane fraction of isolate HYK0203-SK02, which most likely promotes lysis by degrading the cell wall of S. hantzschii. Thus, this work enhances our understanding of the relationship between algal cell structures and algicidal substances during degradation of the diatom S. hantzschii, and may provide a new strategy for controlling harmful blooms of this common freshwater diatom.

Acknowledgements

Funds for this article were provided by the National Research Laboratory Program (2000-N-NL-01-C-290) of the Korean Ministry of Science and Technology. Finally, my special thanks go to my colleague Sung-Su Hong and Wook-Sei Lee in our laboratory for selflessly encouraging my research and giving me technical assistance.

Ancillary