Cyanobacterial Blue Color Formation during Lysis under Natural Conditions
ABSTRACT
Cyanobacteria produce numerous volatile organic compounds (VOCs), such as β-cyclocitral, geosmin, and 2-methylisoborneol, which show lytic activity against cyanobacteria. Among these compounds, only β-cyclocitral causes a characteristic color change from green to blue (blue color formation) in the culture broth during the lysis process. In August 2008 and September 2010, the lysis of cyanobacteria involving blue color formation was observed at Lake Tsukui in northern Kanagawa Prefecture, Japan. We collected lake water containing the cyanobacteria and investigated the VOCs, such as β-cyclocitral, β-ionone, 1-propanol, 3-methyl-1-butanol, and 2-phenylethanol, as well as the number of cyanobacterial cells and their damage and pH changes. As a result, the following results were confirmed: the detection of several VOCs, including β-cyclocitral and its oxidation product, 2,2,6-trimethylcyclohexene-1-carboxylic acid; the identification of phycocyanin based on its visible spectrum; the lower pH (6.7 and 5.4) of the lysed samples; and characteristic morphological change in the damaged cyanobacterial cells. We also encountered the same phenomenon on 6 September 2013 in Lake Sagami in northern Kanagawa Prefecture and obtained almost the same results, such as blue color formation, decreasing pH, damaged cells, and detection of VOCs, including the oxidation products of β-cyclocitral. β-Cyclocitral derived from Microcystis has lytic activity against Microcystis itself but has stronger inhibitory activity against other cyanobacteria and algae, suggesting that the VOCs play an important role in the ecology of aquatic environments.
INTRODUCTION
Blooms of freshwater cyanobacteria, particularly the genera Microcystis and Anabaena, have caused increasing problems in recent years. These blooms have frequently resulted in water quality deterioration with adverse effects on lake ecology, livestock, human water supplies, and recreational activities (1). The most direct method of control involves the application of algicides (2–4), but this is potentially damaging to the environment, may cause secondary pollution, and cannot be applied to public water areas in Japan. Although alternative methods, such as nutrient removal and suppression of growth by aeration devices, have been developed for the elimination of cyanobacteria (5, 6), these methods are expensive and have no immediate obvious effect. There have been very few successful examples of aeration devices in Japan. Therefore, no effective methods have yet been developed. In order to control cyanobacteria, lytic organisms have been investigated, and many lytic protozoa, bacteria, and bacteriophages have been isolated (1, 7–9). Although lytic organisms may play a primary role in causing declines of cyanobacteria, no definite conclusions have been established.
In a natural environment, the lytic phenomena associated with the occasional blue color are observed (Fig. 1A). Fallon and Brook reported that lysis was indicated by the presence of a blue opalescent sheen on the surface waters, resulting from the release of phycocyanin pigments and gas vesicles from the lysing algae (7). For the lysis with blue color formation, they considered that the lytic organisms (bacteria and protozoa) and photooxidation may potentially be harmful to the cyanobacterial populations. However, they could neither identify the cause nor clarify the mechanism of the blue color formation. We also observed the lysis of cyanobacteria involving blue color formation in a natural environment on 5 August 2008, 14 September 2010, and 6 September 2013 and detected phycocyanin from the filtrate of the lysed cyanobacterial scum.
FIG 1
Cyanobacteria produce numerous volatile organic compounds (VOCs), such as β-cyclocitral, geosmin, and 2-methylisoborneol (2-MIB) (10–13). These volatile compounds have been shown to possess lytic activity against cyanobacteria (14). Among these compounds, only β-cyclocitral caused the characteristic color change in the culture broth from green to blue during the lysis process (14, 15). β-Cyclocitral was more easily oxidized than similar aldehyde compounds; therefore, the pH of the solution quickly decreased. As a result, chlorophyll a and β-carotene disappeared due to the acid stress, but phycocyanin remained. The inherent blue color from the tolerant water-soluble pigments thus become visible in the cultured broth. An oxidation product of β-cyclocitral in an aqueous solution was isolated and identified as 2,2,6-trimethylcyclohexene-1-carboxylic acid (β-cyclocitric acid) (15).
Lakes Sagami and Tsukui in northern Kanagawa Prefecture, Japan, are typical eutrophic reservoirs (Fig. 2). We have continued to monitor the presence of cyanobacteria and algae in these lakes every week since 1972 because the lakes are important sources of drinking water for Kanagawa Prefecture. Several large-scale growths of cyanobacteria have occurred with blue color formation, as shown in Fig. 1A. In August 2008 in Lake Tsukui, we encountered a color change similar to that observed in the laboratory experiments shown in Fig. 1. In the present study, we collected lysed cyanobacteria on 5 and 12 August 2008 and investigated the number of cyanobacterial cells and their damage, pH, and VOCs. We also encountered the same phenomenon on 14 September 2010 in Lake Tsukui and on 6 September 2013 in Lake Sagami and collected scum samples and investigated the VOCs (Fig. 2). The aim of the present study is to report blue color formation under natural conditions, which includes measurement of the cyanobacterial cells, observation of the surface of the damaged Microcystis cell membrane, measurement of pH, and detection of phycocyanin and VOCs, such as β-cyclocitral.
FIG 2
MATERIALS AND METHODS
Study sites in Lakes Tsukui and Sagami.
Lake Tsukui (35°35′32″N, 139°16′21″E) is a reservoir that has a storage volume of 62,300,000 m3. The intake is located 85 m upstream of the dam and 55 m offshore from the left bank (Fig. 2) and at 5 to 10 m from the lake bottom. Residence time is approximately 3 weeks. Nine aeration diffuser tubes have been established at a 17-m water depth in the lake to control the growth of cyanobacteria. As shown in Fig. S1 in the supplemental material, a thermocline is formed at 3 depths (4-m and 17-m water depths and 5 m from the lake bottom). The influent water flows below the thermocline at a 17-m water depth. Therefore, the cyanobacterial biomass remains in the surface layer of the lake during a bloom. The cyanobacteria can thus continue to accumulate and move only via the wind-driven currents. The dam site and survey area are shown in Fig. 2. During the daytime, although the wind blows from downstream toward upstream because of the seasonal breeze and sea breeze, the survey area is not exposed to the winds due to the barrier formed by the dam. At night, the wind blows from upstream to downstream because of the land breeze, and the water blooms drift to the dam site. Thus, dense blooms appear often in the survey area, and the lysis occurs at this location.
Lake Sagami (35°36′50.3″N, 139°10′56.1″E) is a reservoir that has a storage volume of 63,200,000 m3 (Fig. 2). The maximum depth of the current is 25 m, and the intake is located on the right bank of the dam from a 0-m to a 23-m water depth (see Fig. S1 in the supplemental material). That is, it has become an all-layers intake. The residence time is approximately 2 weeks. Eight intermittent aerohydraulic guns have been established at a 12-m water depth in the lake to control the growth of cyanobacteria. As shown in Fig. S1 in the supplemental material, a thermocline is formed at 2 depths (3-m and 8-m water depths). The influent water flows below the thermocline at an 8-m water depth. Algae are distributed to a depth of 10 m from the surface. The nutrient concentration of Lake Sagami is higher than that of Lake Tsukui, but it is more difficult to accumulate the cyanobacterial bloom there than in Lake Tsukui. However, in case of a drought, such as in 2013, the water was a headrace from Lake Miyagase to Lake Tsukui, the discharge amount of Lake Sagami was reduced, the residence time was longer, and the cyanobacterial bloom was formed in 2013.
Collection of water containing cyanobacteria.
We conducted a fixed-point observation at 80 m from the left bank of the dam of Lake Tsukui near the water intake opening (station [St.] 1) (Fig. 2) and at the center of the Sagamiko Oohashi Bridge (St. 5) (Fig. 2) once a week. A 1-liter sample of the surface water was collected using a stainless steel bucket and then placed in a polyethylene bottle. Samples from Lake Tsukui were brought to the laboratory within 10 min and analyzed. Samples from Lake Sagami were analyzed within 30 min. Cyanobacterial blue color formation was observed at 30 m from St. 1 at Lake Tsukui on 5 August 2008 (St. 2) (Fig. 2), and it continued to 12 August. The surface scum samples at the lysing point were collected in the same way at the fixed point on 5 and 12 August 2008. We also encountered the same phenomenon at Lake Tsukui on 14 September 2010 and also collected a surface scum sample at the lysing point (St. 3) (Fig. 2) and a nonlysing scum sample (St. 4) (Fig. 2). We also encountered the same phenomenon at Lake Sagami on 6 September 2013 and also collected a surface scum sample at the lysing point (St. 8) (Fig. 2) and a water sample (St. 6) (Fig. 2) and a nonlysing scum sample (St. 7) (Fig. 2). The samples from Lake Sagami were brought to the laboratory within 30 min and analyzed.
Absorption spectrum.
The scum sample was filtered using GF/A glass fiber filter paper (Whatman, Maidstone, United Kingdom), and the absorption spectrum of the filtrate was recorded at 800 to 190 nm using a Jasco V-560 UV-visible (Vis) spectrophotometer (Jasco Co., Tokyo, Japan). The standard phycocyanin was extracted from a cyanobacterium (Microcystis aeruginosa NIES-298) and used for identification (16).
Measurement of pH.
The pHs of the samples were measured with an HM-60V pH meter (DKK TOA Co., Tokyo, Japan).
Microscopic examination.
The samples were fixed with equal quantities of 2% glutaraldehyde in 30 mM HEPES buffer (Dojindo Laboratory, Kumamoto, Japan). A microscopic examination was performed to identify the cyanobacteria using a positive phase-contrast microscope (BX50; Olympus Co., Tokyo, Japan). The number of trichomes of Aphanizomenon and the number of cells of Anabaena were determined in a 0.1-ml sample at a magnification of ×150. The number of Microcystis cells was examined using a hemacytometer at a magnification of × 600 after they were separated into single cells by chlorination (5 mg liter−1; 1 h) and sonication (approximately 30 W; 1 min) using an ultrasonic homogenizer (US50; Nissei Co., Tokyo, Japan) (17).
Observation of cell surfaces using a SEM.
To observe the surfaces of the algal cells, a scanning electron microscope (SEM) (JSM-6380LA; JEOL, Tokyo, Japan) was used. The samples were fixed with 2% glutaraldehyde, as for the microscopic examination. The specimens were transferred to distilled water and rinsed and then dropped onto a nanopercolator (SEM Pore) made of polycarbonate (JROL, Tokyo, Japan), which were slowly sucked and rapidly frozen using liquid nitrogen. The filter was then attached to a specimen holder for use in a low-vacuum SEM, freeze-dried in the specimen chamber, and then sputter coated with platinum (JFC-1600; JEOL). Photographs were taken under high-vacuum conditions.
Chemicals.
For the VOCs, β-cyclocitral and 2,2,6-trimethylcyclohexanone were purchased from Alfa Aesar (Ward Hill, MA, USA), and β-ionone was purchased from Sigma-Aldrich (St. Louis, MO, USA). 2,2,6-Trimethylcyclohexene-1-carboxylic acid was prepared as follows. Silver oxide (Ag2O) was added to water, and then sodium hydroxide was added to the Ag2O solution. β-Cyclocitral was added to this solution with shaking at room temperature for 20 h. The reaction solution was extracted with diethyl ether to remove the starting material, and 0.5% HCl was added to the lower layer for acidification. The resulting acidic solution was extracted with diethyl ether, and the extract was evaporated. The residue was recrystallized with benzene and cyclohexane to give colorless prism crystals with melting points of 92 to 94°C. The preparation of 2,2,6-trimethylcyclohex-1-en-1-yl formate was carried out as follows. An aqueous solution of β-cyclocitral was stirred for 40 h at room temperature. The solution was extracted with diethyl ether. The extract was evaporated to dryness to provide an oily substance, which was subjected to silica gel column chromatography with benzene–n-hexane (1:1) as the mobile phase. The enolester was obtained as an oily compound.
GC analysis of VOCs by SPME.
The water samples collected in 2008 were subjected to a headspace solid-phase microextraction (HS-SPME) coupled with gas chromatography-mass spectrometry (GC-MS) for quantitative determination of the volatile compounds β-cyclocitral, β-ionone, geosmin, 2-MIB, 2-methyl-1-butanol, and 3-methyl-1-butanol. In 2010, 1-propanol, isobutanol, and 2-phenylethanol were also examined, in addition to the above-mentioned VOCs, by the same method. Twenty milliliters of raw lake water was filtered through a GF/A glass filter; then, 10 ml of this raw water and the resulting filtrate were used for the subsequent analyses. The samples were placed in a headspace septum vial containing 4 g NaCl (12). After an internal-standard solution (10 μl) of geosmin-D3 was added, the vial was sealed with a Teflon-lined cap and then gently agitated to dissolve the salt. After refrigerated storage for up to 24 h, the vial was placed in a stirrer equipped with a heater block preheated to 60°C (Combi Pal system autosampler; CTC Analytics, Zwingen, Switzerland). The outer needle of the SPME fiber (50/30-μm divinylbenzene-carboxen-polydimethylsiloxane [DVB-CAR-PDMS]; Supelco, Bellefonte, PA, USA) assembly was passed through the septum, and the fiber was extended into the headspace. After 20 min, the fiber was retracted and placed in the injector of a gas chromatograph (Agilent 6890; Agilent Technologies, Palo Alto, CA, USA) connected to a J&W Scientific (Folsom, CA, USA) column (DB-624; 30 m by 0.25-mm inside diameter [i.d.] by 1.4-μm film) and to a mass selective detector (Agilent 5973). Fiber desorption was performed in the splitless mode at 200°C for 10 min. The injector temperature was 200°C, and the column temperature program was 40°C (5 min), from 40°C to 210°C at 10°C min−1, and a 5-min hold at 210°C. Helium was used as the carrier gas (1 ml min−1). The detector temperature was 230°C. Electron ionization (EI) was used for the ionization and for the selected ion monitoring (SIM) mode: m/z 152 for β-cyclocitral, m/z 177 for β-ionone, m/z 57 for 2-methyl-1-butanol, m/z 55 for 3-methyl-1-butanol, m/z 91 for 2-phenylethanol, m/z 31 for 1-propanol, m/z 43 for 2-methyl-1-propanol, m/z 56 for 1-butanol, m/z 112 for geosmin, m/z 95 for 2-MIB, and m/z 115 for geosmin-D3 were monitored.
GC-MS analysis after extraction with diethyl ether.
Five milliliters of the sample was filtered through a GF/A glass filter. Two milliliters of the filtrate was extracted with 2 ml of diethyl ether, with 20 μl of geosmin-D3 as the internal standard. The mixture was centrifuged at 2,000 rpm for 5 min, and the upper layer was placed in an amber vial sealed with a Teflon-lined cap. GC-MS was performed using a gas chromatograph (HP 6890; Agilent Technologies, Palo Alto, CA, USA) connected to an Agilent Technologies HP-5ms column (20 m by 0.25-mm i.d. by 0.25-μm film) and to a mass selective detector (HP 5973). The injector temperature was 250°C, and the column temperature program started at 40°C (3 min), then from 40°C to 100°C at 15°C min−1, and then from 100°C to 250°C at 25°C min−1 with a 5-min hold at 250°C. Two microliters of the sample was injected into the GC injection port in the splitless mode using an injector (HP 6890 series). The injection port was equipped with a 2.0-mm-i.d. single-taper deactivated liner (Restek, Bellefonte, PA, USA). Helium was used as the carrier gas (1 ml min−1). The transfer line temperature was 280°C, and the detector temperature was 230°C. EI was used for ionization at an electron energy of 70 eV and for the SIM mode: m/z 137 for β-cyclocitral, m/z 107 for β-cyclocititric acid, m/z 177 for β-ionone, m/z 91 for 2-phenylethanol, m/z 82 for 2,6,6-trimethylcyclohexanone, m/z 125 for 2,2,6-trimethylcyclohex-1-en-1-yl formate, and m/z 115 for geosmin-D3 were monitored.
RESULTS
Occurrence of lysed cyanobacteria involving blue color formation.
In 2008, cyanobacteria appeared at the beginning of June, and the water bloom was formed in the middle of July at Lake Tsukui. The scum was then formed at the dam site at the end of July. We observed cyanobacterial blue color formation during the lysis process under natural conditions at St. 2 (Fig. 2) on 5 August (Fig. 1B), and it extended to 12 August at 30 m from the fixed point. Although this phenomenon continued for two more weeks, a heavy rain removed it on 24 August. Following this rainfall, the cyanobacteria sharply decreased, and the phenomenon was not observed again in that year. The surface scum samples at the lysed point (St. 2) (Fig. 2) were collected in the same way at the fixed point on 5 and 12 August 2008. The phenomenon was observed again in 2010 and in Lake Sagami in 2013. In 2010, cyanobacteria appeared at the beginning of June, but the scum formed at the dam site at the beginning of September at Lake Tsukui. We also encountered the same phenomenon at St. 3 on 7 September 2010 (Fig. 1C and 2). At the time, although the phenomenon occurred, we were unable to take samples because the area was away from the dam. One week later, we were able to collect the surface scum samples at the lysed point because it was possible to approach it from the dam. In 2013, cyanobacteria appeared at the beginning of mid-June, and the water bloom was formed in mid-July at Lake Sagami. We observed blue color formation at St. 8 (Fig. 1D and 2).
Observation of blue color and pH change in the lysed scum samples.
As shown in Fig. 3, the filtrate of the lysed scum samples in August 2008 showed a blue color. The visible spectrum of the filtrate had an absorption maximum at 620 nm, which was identical to that of the extracted phycocyanin from a cyanobacterium (M. aeruginosa NIES-298) (16). The lake water in the lysed sample in September 2010 became visibly blue after agitation followed by rest (see Fig. S2A in the supplemental material). The living cells floated, while the dead cells sank, as shown in Fig. S2B and C in the supplemental material, respectively. As shown in Table 1 and Fig. S2 in the supplemental material, the pH of water on the lake surface in the nonlysing area was approximately 8.5, whereas the pH of the lysed sample taken on 5 August was 6.7 and that taken on August 12 was 5.4. In 2010, the pH was 5.9 at the lysed point (St. 3) (Fig. 2), while it was approximately pH 9 at St. 1 and pH 9.9 in a nonlysing scum sample (St. 4) (Fig. 2). As shown in Table 2, the pH of the lysed sample taken on 6 September 2013 at Lake Sagami was 6.2, whereas that taken at St. 6 was 9.3.
FIG 3
TABLE 1
| Parameter | Value for indicated sampling site and date | |||||||
|---|---|---|---|---|---|---|---|---|
| 2008 (St. 2 [with BCFa]) | 2010 (14 September) | |||||||
| 5 August | 12 August | St. 3 (with BCF) | St. 4 (without BCF) | |||||
| Raw | Filtrate | Raw | Filtrate | Raw | Filtrate | Raw | Filtrate | |
| pH | 6.7 | 5.4 | 5.9 | 9.9 | ||||
| VOC concn (μg liter−1)b | ||||||||
| β-Ionone | 9.6 (3.2) | 7.6 | 98 (2.7) | 15 | 300 (0.7) | 1.6 | 26 (<0.1) | 0.5 |
| β-Cyclocitral | 50 (79) | 0.9 | 100 (1.5) | 55 | 1400 (3.3) | 3.2 | 52 (<0.1) | 1.2 |
| 2-Methyl-1-butanol | 2.1 (1.9) | 0.9 | 130 (1.7) | 76 | 140 (0.3) | 17 | 6.6 (<0.1) | 2.6 |
| 3-Methyl-1-butanol | 9.2 (7.9) | 4.3 | 470 (5.5) | 300 | 430 (0.8) | 92 | 37 (<0.1) | 17 |
| 1-Butanol | ND | ND | ND | ND | 350 (0.7) | 65 | 3.8 (<0.1) | 2.4 |
| 1-Propanol | ND | ND | ND | ND | 1100 (1.5) | 450 | 83 (<0.1) | 60 |
| Isobutanol | ND | ND | ND | ND | 180 (0.4) | Not detected | 12 (<0.1) | Not detected |
| 2-Phenylethanol | ND | ND | ND | ND | 440 (1.0) | 5.6 | 45 (<0.1) | Not detected |
| Geosmin | 0.6 | 0.3 | 6.5 | 6.1 | 95 | 13 | 7.7 | 4.9 |
| No. of Microcystis cells | 6.2 × 104 | 3.1 × 106 | 4.2 × 107 | 7.6 × 107 | ||||
a
BCF, blue color formation.
b
VOC concentrations are expressed in μg liter−1, and in parentheses are the analytical results for the original intracellular concentrations (fg cell−1) of the VOCs in the raw samples. ND, not determined.
TABLE 2
| Parameter | Value at indicated sampling site | ||
|---|---|---|---|
| St. 6 | St. 7 | St. 8 (with BCFa) | |
| pH | 9.3 | 8.7 | 6.2 |
| VOC concnb (μg liter−1) | |||
| β-Cyclocitral | Not detected | Not detected | Not detected |
| 2,2,6-Trimethylcyclohex-1-en-1-yl formate | Not determined | Not detected | 25 (0.15) |
| β-Cyclocitric acid | Not detected | 52 (0.31) | 120 (0.71) |
| 2,2,6-Trimethylcyclohexanone | Not detected | 11 (0.08) | 170 (1.2) |
| β-Ionone | 10 (0.05) | 120 (0.61) | 150 (0.80) |
| No. of Microcystis cells | 4.5 × 104 | 2.0 × 105 | 8.5 × 105 |
a
BCF, blue color formation.
b
VOC concentrations are expressed in μg liter−1, and in parentheses are molar concentrations.
Observation of lysed cyanobacteria by microscopy.
Figure 4 shows light micrographs of the lysed scum samples collected from Lakes Tsukui and Sagami. Numerous dead cells of cyanobacteria collected on 12 August 2008 lost their cell walls, and the resulting cytoplasm was aggregated in the sample on 12 August (Fig. 4B). While one trichome of Anabaena affinis was composed of 100 or more cells on 5 August (Fig. 4A), the number of cells in one trichome was reduced to less than 30 after 1 week (Fig. 4B). In the sample collected from St. 8 of Lake Sagami on 6 September 2013, most of the vegetative cells of Microcystis and Anabaena (Fig. 4C) were lysed and the heterocysts of Anabaena (Fig. 4D) and Pseudoanabaena mucicola (Fig. 4E) remained. Anabaena crassa looked whitish because there was no gas vesicle, as a result of damage (Fig. 4C). Observation of the cell surface using a SEM showed that the surface of the damaged Microcystis cell membrane sampled on 14 September 2010 was wrinkled (see Fig. S3B in the supplemental material), whereas that of the healthy cell was smooth (see Fig. S3A in the supplemental material).
FIG 4
Figure 5 shows the species composition of the cyanobacteria collected from Lake Tsukui in 2008 and 2010 and from Lake Sagami in 2013. The species composition of the living cells at the biovolume rate (18, 19) in the sample taken on 5 August, when the phenomenon had just occurred, was as follows: Aphanizomenon flos-aquae, 43.7%; M. aeruginosa, 36.4%; and A. affinis, 20.0%. For the sample taken on 12 August, Microcystis was almost 95.9%, A. affinis was 4.1%, and Aphanizomenon had disappeared. In 2010, such a change was not found at the same point, because we were able to collect the samples only one time. However, the percentage of Microcystis within the species composition at the lysing point was greater than 99% at St. 3, whereas it was 89% at the fixed point (St. 1) (Fig. 2). In 2013, Anabaena mucosa was the dominant species in Lake Sagami. The percentage of A. mucosa within the species composition was 83.8% at St. 5 on 3 September and 75.6% at St. 6 on 6 September. However, the percentages were 36.6% at St. 7 and 5.7% at St. 8 on September 6. On the other hand, the percentage of Microcystis within the species composition was 14.4% at St. 5 on 3 September and 22.4% at St. 6 on 6 September. However, it increased to 62.8% at St. 7 and 72.0% at St. 8 on September 6. A. crassa also increased from 1.8% to 22.2% at St. 8 (Fig. 5).
FIG 5
Detection of VOCs and related compounds using GC-MS.
Table 1 shows the analysis results for the VOCs and pH values (the concentration of the raw material is expressed as the absolute amount [μg liter−1], and the intracellular concentration is expressed as the amount [raw filtrate] [μg liter−1]/number of cells [fg/cell]). From the samples collected on 5 August 2008, β-cyclocitral was detected in significant amounts in the raw sample, whereas the concentrations of the other VOCs were approximately 5-fold lower than that of β-cyclocitral. After 1 week, the concentration of β-cyclocitral doubled compared to that in the raw sample. The concentrations of β-ionone, 2-methyl-1-butanol, and 3-methyl-1-butanol significantly increased. Among them, a particularly high concentration of 3-methyl-1-butanol was detected in the filtrate. As we previously reported, an oxidation product of β-cyclocitral, β-cyclocitric acid, was also detected, although no quantitative analysis was conducted (20) in 2008 and 2010. In the raw water sample collected in 2010, β-cyclocitral was detected at a concentration higher than that in the 2008 samples, and β-ionone and the alcohols were also detected. Specifically, the concentration of 1-propanol was very high. However, the concentrations were quite low at the fixed point. When the intracellular VOC concentrations were measured on 5 August 2008 (Table 1), that of β-cyclocitral was the highest among the VOCs. However, the concentration of β-cyclocitral drastically decreased 1 week later, probably due to oxidation to the corresponding acid, because the pH of the raw sample was reduced to 5.4 (Table 1). In September 2010, the cell density of Microcystis at St. 3 (Fig. 2) with blue color formation was higher than that in 2008, and the VOC concentration was higher than that in 2008. However, the concentration per cell was rather low, and some of the VOC concentrations at St. 4 (Fig. 2) without any formation of blue color were roughly equivalent, but the concentrations per cell were very low.
Detection of VOCs using GC-MS after extraction.
Table 2 shows the analytical results (μg liter−1) for the volatile organic compounds and the pH values collected from the three stations at Lake Sagami on 6 September 2013 with or without blue color formation. In the sample from St. 6 (Fig. 2), 84% of the cyanobacteria were Anabaena, and only β-ionone was detected. In the samples from St. 7 and 8 (Fig. 2), β-cyclocitral was not detected, but β-cyclocititric acid, 2,2,6-trimethylcyclohex-1-en-1-yl formate, and 2,6,6-trimethyl-cyclohexanone, which are the oxidation products of β-cyclocitral, were detected, in addition to β-ionone. The structures of β-cyclocitral and its oxidation products, together with β-ionone, which are probably derived from β-carotene, are shown in Fig. 5.
DISCUSSION
Cyanobacteria produce numerous volatile organic compounds, such as β-cyclocitral, geosmin, and 2-MIB (13, 21). The occurrence of β-cyclocitral is correlated with the appearance of Microcystis (11, 21). In the present study, we detected the following volatile organic compounds from the raw water and the filtrate of the lysed cyanobacterial scum of Lake Tsukui: β-cyclocitral, β-ionone, 2-methyl-1-butanol, 3-methyl-1-butanol, 1-propanol, and 2-phenylethanol (Table 1). Some of these volatile compounds from the cyanobacteria showed lytic activity against the cyanobacteria (14). In particular, β-cyclocitral caused an interesting color change in the culture broth from green to blue during the lysis process (14, 15). When β-cyclocitral was added to the laboratory strains of any genera and to bloom samples, including many species of cyanobacteria, it caused the characteristic result, so that the absorption maxima of chlorophyll a and β-carotene disappeared but that of phycocyanin remained for 12 h. This indicated that the oxidation of β-cyclocitral leads to acidification, which then preferentially decomposes chlorophyll a and β-carotene rather than phycocyanin, so that the inherent color from the tolerant water-soluble pigments became observable. Simis and Kauko bleached chlorophyll and carotenoid by the addition of β-cyclocitral to many cell cultures of cyanobacteria and obtained the mass-specific absorption spectra of phycocyanin and phycoerythrin (22). However, the addition of β-ionone gradually reduced the green color and produced a colorless solution with white precipitates after 10 h (15). This result was probably due to the structurally characteristic feature of β-cyclocitral.
The concentrations measured at the lysing point were lower than those from the laboratory experiments performed by Ozaki et al. (14) and Harada et al. (15). However, β-cyclocitral, whose pH-dependent distribution coefficient (log D) is 3.3, is hydrophobic; consequently, only 1/2,000 of the added β-cyclocitral dissolves in water and shows lytic activity (23). The concentration at which the lysis involving blue color formation took place in the laboratory experiment was 6.5 mM, which equates to 1,000 mg liter−1, and its concentration taking into consideration the 1/2,000 dissolution factor was 500 μg liter−1. This concentration roughly agreed with the data for the Lake Tsukui sample taken on 14 September 2010 (Table 1). The cell quota of β-cyclocitral was estimated to be approximately 10 fg/cell (10). According to the analytical results for the VOCs from the scum samples with blue color formation on 5 August, the intracellular β-cyclocitral concentration was 75 fg/cell, which was much higher than 10 fg/cell (Table 1). One week later, the intracellular β-cyclocitral concentration became lower than 10 fg, and the concentration per cell was very low in the St. 4 sample, which showed no formation of blue color. Although β-cyclocitral was formed during the analysis using SPME (20), we observed that the potential for the formation of β-cyclocitral was higher in cells in which blue color formation occurred than in healthy cells.
β-Cyclocitral is derived from β-carotene by an oxidation reaction (11, 21). Jüttner et al. reported that the production of β-cyclocitral from β-carotene was activated by the disintegration of the Microcystis cells undergoing freezing and thawing (11). These results indicated that β-cyclocitral is not originally present in the cells and can be derived from β-carotene when the Microcystis cell is disintegrated by stimulations, such as some pretreatments and biological attacks by viruses and bacteria, and the subsequent activation of the carotenoid cleavage dioxygenases (CCD) (24, 25). β-Cyclocitral was more easily oxidized than similar aldehyde compounds, and the pH of the solution, therefore, quickly decreased to 4.5. An oxidation product of β-cyclocitral in a water solution was isolated and identified as β-cyclocititric acid (13). Huang et al. reported that cyanobacterial (specifically, Synechocystis sp. strain PCC 6308) blue color formation may be due to acid stress (26). They used hydrochloric acid for acidification, and the blue color was formed below pH 3.6. As shown in Table 1, we measured the pHs of the lysed scum samples from Lake Tsukui, and the pH decreased to 5.4 from the typical pH (approximately 8.5), suggesting the presence of an acidic compound. The acid formed from β-cyclocitral probably contributed to the pH-lowering effect, thus allowing the characteristic blue color to form, although the pH was higher than 3.6 (15). This may have been due to damage to the cell wall, and these details of blue color formation will be reported elsewhere.
On 6 September 2013 at Lake Sagami, β-cyclocitral was not detected, but β-cyclocitric acid, 2,2,6-trimethylcyclohex-1-en-1-yl formate, and 2,6,6-trimethyl cyclohexanone were detected; the concentration of 2,6,6-trimethyl cyclohexanone was the highest among these VOCs (Table 2 and Fig. 6). β-Cyclocitric acid is an oxidation product of β-cyclocitral, and 2,2,6-trimethylcyclohex-1-en-1-yl formate is an enolester, also produced from β-cyclocitral, probably due to Baeyer-Villiger oxidation. The enolester was easily transformed into 2,6,6-trimethyl cyclohexanone by alkaline hydrolysis. As shown in Table 2, the molar sums of these three VOCs at St. 7 and St. 8 were 0.39 μM and 2.07 μM, respectively, suggesting that the parental β-cyclocitral was present at a concentration of 61 μg liter−1 at St. 7 and 330 μg liter−1 at St. 8. These concentrations and pH values are similar to the 2008 data from Lake Tsukui. The detailed chemical and analytical aspects of the VOCs will be discussed elsewhere.
FIG 6
Ozaki et al. studied the morphological change in the lysed Microcystis cells after the spiking of β-cyclocitral using electron microscopic methods in the laboratory (27). They found that the cells initially shrank and then gradually became wrinkled when using the SEM, and damage to the cell membrane or cell wall on the surface was observed. However, by using a transmission electron microscope, it was shown that the thylakoid membrane with the phycobilisome still remained in the center of the cell. In the present study, we examined the wrinkled cells using an SEM (see Fig. S3B in the supplemental material) and observed that the cell walls were broken; the resulting cytoplasm was identified using an optical microscope in the same manner as previously reported (27) (Fig. 4B). This result suggested that the cyanobacteria collected from Lake Tsukui were lysed with β-cyclocitral. These dead cells became aggregated and sank on a large scale; therefore, the lake clearly showed a blue color (Fig. 1A).
As shown in Fig. 5, the percentage of the living Microcystis cells within the species composition increased from 36.4% to 95.9% 7 days after blue color formation occurred, but that of A. affinis decreased from 20% to 4.1% in 2008. That of A. flos-aquae was 43.7%, but it disappeared 1 week later. The lytic activity of β-cyclocitral for A. affinis and Aphanizomenon appeared stronger than that for Microcystis. When β-cyclocitral was added to laboratory strains, blue color formation was observed after 6 h; however, it was observed 4 days after β-cyclocitral was added to natural bloom samples (15). Microcystis under natural conditions has a sheath on the outside of the cells (28, 29), but such a sheath was not found in A. affinis and Aphanizomenon. Therefore, the Microcystis composition of the living cells appeared to increase 7 days after blue color formation. In 2013 at Lake Sagami, the percentage of the living Microcystis cells within the species composition increased from 22.4% to 72.0% and that of A. mucosa cells drastically decreased from 83.8% to 5.7% (Fig. 5). Chang et al. exposed β-cyclocitral to two cyanobacteria (M. aeruginosa PCC 7005 and 7820) and one diatom (Nitzschia palea) (30). They determined the effect of β-cyclocitral on cell integrity using an SEM and a flow cytometer. As a result, a higher concentration of β-cyclocitral was needed for the two Microcystis strains than for N. palea to cause the cells to rupture. In the present study, we observed that A. affinis, Aphanizomenon, and A. mucosa decreased more quickly than Microcystis. Microcystis has higher resistance to β-cyclocitral than the other cyanobacteria and algae. β-Cyclocitral derived from Microcystis has lytic activity toward Microcystis itself but would have stronger inhibitory activity against other cyanobacteria and algae. Jüttner et al. also reported that the role of β-cyclocitral was as a grazer repellent against Daphnia (31).
Several alcohols, such as 1-propanol, 3-methyl-1-butanol, and 2-phenylethanol, were detected together with β-cyclocitral, as shown in Table 1. In a previous laboratory experiment, these alcohols were also detected together with β-cyclocitral from the cyanobacterium M. aeruginosa NIES-843. The alcohols were optimally produced after 35 days of culture, in which the nitrate nitrogen in the cultured broth was exhausted. Additionally, they were definitely produced using the 2-keto-acid decarboxylase (MaKDC) in the Microcystis strains (32). It is known that 3-methyl-1-butanol induced the formation of the hypha-like extensions and pseudohyphae in yeasts (33). These results suggested that, because these compounds are not produced by the other genera of cyanobacteria, the alcohols from Microcystis were significant for its life (34). Watson mentioned that although many algal VOCs may represent waste by-products from the metabolic or cell degradation process, they could serve as important chemical signals of changes in the algal growth or metabolism, in community structures, and in ecosystem functions and health (13). The production of β-cyclocitral and 3-methyl-1-butanol showed a unique pattern, so that they were not produced during the cyanobacterial growth stage but appeared during an early stage of aging (14). Therefore, it is considered that these VOCs play an important role in the life cycle of the cyanobacteria.
While blue color formation is often observed during midsummer when the cyanobacterial blooms produce scum, it is rare in the autumn, when the cyanobacteria decrease. This may be due to the stress related to the density of the cyanobacteria, which leads to the disintegration of the Microcystis cells, followed by the production of β-cyclocitral. In a separate laboratory experiment, it was found that the accumulation of Microcystis cells caused blue color formation (data not shown). Additionally, the characteristic feature of Lake Tsukui, which is that there is no leakage of water blooms of cyanobacteria from the surface layer and they continue to accumulate, facilitates the occurrence of this phenomenon, as shown in Materials and Methods. As seen in Fig. 1A, blue color formation appeared over a large portion of the surveyed area. Lysis with blue color formation may be one of the survival strategies employed by Microcystis spp. during cyanobacterial blooms, when nutrients become scarce.
In conclusion, a color change similar to that seen in laboratory experiments was observed in the natural environment. We tried to clarify a mechanism for blue color formation. The mechanism may be as follows: first, β-cyclocitral is released from Microcystis cells by a lytic microorganism or high cell density as a trigger; second, β-cyclocitral lyses other Microcystis cells; third, the resulting β-cyclocitral is rapidly oxidized to the corresponding carboxylic acid; then, the pH quickly decreases, and chlorophyll a and β-carotene are decomposed by acid stress. As a consequence of these consecutive events, the blue color due to phycocyanin is formed. In this study, β-ionone, 1-propanol, 3-methyl-1-butanol, and 2-phenylethanol were detected, together with β-cyclocitral. These VOCs may be associated with the life cycle of Microcystis, and more detailed studies will contribute to establishing a biological control system for Microcystis. Blue color formation is often observed during the midsummer, when the cyanobacterial blooms produce scum. Therefore, it might be related to the stress of the density. β-Cyclocitral derived from Microcystis has lytic activity toward Microcystis itself but would have stronger inhibitory activity against other cyanobacteria and algae, suggesting that these VOCs play an important role in the ecology of aquatic environments.
ACKNOWLEDGMENTS
We thank Daiki Fujise, Kawasaki Waterworks Bureau, Japan, for the SEM experiments and Mark Rebuck, Meijo University, for helpful discussions.
Supplemental Material
File (zam999116167so1.pdf)
- Download
- 1.09 MB
REFERENCES
1.
Sigee DC, Glenn R, Andrews MJM, Bellinger EG, Butler RD, Epton HA, Hendry RD. 1999. Biological control of cyanobacteria: principles and possibilities. Hydrobiologia 395-396:161–172.
2.
Wu ZX, Gan NQ, Huang Q, Song LR. 2007. Response of Microcystis to copper stress: do phenotypes of Microcystis make a difference in stress tolerance? Environ Pollut 147:324–330.
3.
Jancula D, Marsalek B. 2011. Critical review of actually available chemical compounds for prevention and management of cyanobacterial blooms. Chemosphere 85:1415–1422.
4.
Hrudey S, Burch M, Drikas M, Gregory R. 1999. Remedial measured, p 278–283. In Chorus I, Bartram J (ed), Toxic cyanobacteria in water: a guide to their public health consequences, monitoring, and management. E & FN Spon, London, United Kingdom.
5.
Pastorok RA, Lorenzen MW, Ginn TC. 1982. Environmental aspects of artificial aeration and oxygenation of reservoirs: a review of theory, techniques, and experiences. NTIS technical report E-82-3. U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
6.
Piehler MF. 2008. Watershed management strategies to prevent and control cyanobacterial harmful algal blooms. Adv Exp Med Biol 619:259–273.
7.
Fallon RD, Brook TD. 1979. Lytic organisms and photooxidative effects: influence on blue-green algae cyanobacteria in Lake Mendota, Wisconsin. Appl Environ Microbiol 38:499–505.
8.
Tsuji K, Asakawa M, Anzai Y, Sumino T, Harada K-I. 2006. Degradation of microcystins using immobilized microorganisms isolated in a eutrophic lake. Chemosphere 65:117–124.
9.
Tucker A, Pollard P. 2005. Identification of cyanophage Ma-LBP and infection of the cyanobacterium Microcystis aeruginosa from an Australian subtropical lake by the virus. Appl Environ Microbiol 71:629–635.
10.
Jones GJ, Korth W. 1991. In situ production of volatile odor compounds by river and reservoir phytoplanctons in Australia. Water Sci Technol 31:145–151.
11.
Jüttner F, Höflacher B, Wurster K. 1986. Seasonal analysis of volatile organic biogenic substances (VOBS) in freshwater phytoplankton populations dominated by Dinobryon, Microcystis and Aphanizomenon. J Phycol 22:169–175.
12.
Satchwill T, Watson SB, Dixon E. 2007. Odorous algal-derived alkenes: differences in stability and treatment responses in drinking water. Water Sci Technol 55:95–102.
13.
Watson SB. 2003. Cyanobacterial and eukaryotic algal odor compounds: signals or by-products? A review of their biological activity. Phycologia 42:332–350.
14.
Ozaki K, Ohta A, Iwata C, Horikawa A, Tsuji K, Ito E, Ikai Y, Harada K-I. 2008. Lysis of cyanobacteria with volatile organic compounds. Chemosphere 71:1531–1538.
15.
Harada K-I, Ozaki K, Tsuzuki S, Kato H, Hasegawa M, Kuroda EK, Arii S, Tsuji K. 2009. Blue color formation of cyanobacteria with β-cyclocitral. J Chem Ecol 35:1295–1301.
16.
Sarada R, Pillai MG, Ravishankar GA. 1999. Phycocyanin from Spirulina sp: influence of processing of biomass on phycocyanin yield, analysis of efficacy of extraction methods and stability studies on phycocyanin. Process Biochem 34:795–801.
17.
Japan Water Works Association. 2011. Standard methods for the examination of water. Japan Water Works Association, Tokyo, Japan. (In Japanese.)
18.
Sigee DC. 2005. Freshwater microbiology: biodiversity and dynamic interactions of microorganisms in aquatic environment. John Wiley and Sons, Chichester, United Kingdom.
19.
Wood SA, Paul WJ, Hamilton DP. 2008. Cyanobacterial biovolumes for the Rotorua Lakes. Cawthron report no. 1504, p 1–19. Cawthron Institute, Nelson, New Zealand. http://www.boprc.govt.nz/media/32233/Cawthron-090803-CyanobacterialbiovolumesforRotorualakes.pdf.
20.
Fujise D, Tsuji K, Fukusima N, Kawai K, Harada K-I. 2010. Analytical aspects of cyanobacterial volatile organic compounds for investigation of their production behavior. J Chromatogr A 1217:6122–6125.
21.
Jüttner F. 1984. Dynamics of the volatile organic substances associated with cyanobacteria and algae in a eutrophic shallow lake. Appl Environ Microbiol 47:814–820.
22.
Simis SGH, Kauko HM. 2012. In vivo mass-specific absorption spectra of phycobillipigments through selective bleaching. Limnol Oceanogr Methods 10:214–226.
23.
Nozawa Y, Kawashima A, Hashimoto EH, Kato H, Harada K-I. 2009. Application of log D for the prediction of hydrophobicity in advanced Marfey's method. J Chromatogr A 1216:3807–3811.
24.
Marasco EK, Kimleng V, Schmidt-Dannert C. 2006. Identification of carotenoid cleavage dioxygenases from Nostoc sp. PCC 7120 with different cleavage activities. J Biol Chem 281:31583–31593.
25.
Scherzinger D, Al-Babili S. 2008. In vitro characterization of a carotenoid cleavage dioxygenase from Nostoc sp. PCC 7120 reveals a novel cleavage pattern, cytosolic localization and induction by highlight. Mol Microbiol 69:231–244.
26.
Huang JJ, Kolodny NH, Redfearn JT, Allen MM. 2002. The acid stress response of the cyanobacterium Synecocystis sp. strain PCC 6308. Arch Microbiol 177:486–493.
27.
Ozaki K, Ito E, Tanabe S, Natsume K, Tsuji K, Harada K-I. 2009. Electron microscopic study on lysis of a cyanobacterium Microcystis. J Health Sci 55:578–585.
28.
Amemiya Y, Kato K, Nakagawa O. 1988. Extracellular products of Microcystis species: formation of slime layer and DOC pool in surrounding waters. Verh Int Verein Limnol 23:1886–1892.
29.
Amemiya Y, Nakagawa O. 1989. Separation and sugar composition of gelatinous sheath carbohydrates produced by Micocystis. Jpn J Phycol 37:253–262.
30.
Chang DW, Hsieh ML, Chen YM, Lin TF, Chang JS. 2011. Kinetics of cell lysis for Microcystis aeruginosa and Nitzschia palea in the exposure to β-cyclocitral. J Hazard Mat 185:1214–1220.
31.
Jüttner F, Watson SB, Elert EV, Köster O. 2010. β-Cyclocitral, a grazer defense to the cyanobacterium Microcystis. J Chem Ecol 36:1387–1397.
32.
Hasegawa M, Nishizawa A, Tsuji K, Kimura S, Harada K-I. 2012. Volatile organic compounds derived from 2-keto-acid decarboxylase in Microcystis aeruginosa. Microbes Environ 27:525–528.
33.
Dickinson JR. 1996. ‘Fusel’ alcohols induce hyphal-like extensions and pseudohyphal formation in yeasts. Microbiology 142:1391–1397.
34.
Harada K-I, Hasegawa M, Fukushima N, Nishizawa A, Fujise D, Tsuji K. 2010. Degradation behavior of β-carotene during cultivation of cyanobacteria. J Res Inst Meijo Univ 9:83–91.
Information & Contributors
Information
Published In
Applied and Environmental Microbiology
Volume 81 • Number 8 • 15 April 2015
Pages: 2667 - 2675
Editor: H. L. Drake
PubMed: 25662969
Copyright
© 2015 American Society for Microbiology.
History
Received: 18 November 2014
Accepted: 22 January 2015
Published online: 26 March 2015
Contributors
Editor
H. L. Drake
Editor
Metrics & Citations
Metrics
Note:
- For recently published articles, the TOTAL download count will appear as zero until a new month starts.
- There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
- Citation counts come from the Crossref Cited by service.
Citations
Citation
2015.
Cyanobacterial Blue Color Formation during Lysis under Natural Conditions. Appl Environ Microbiol
81:.
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.