INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The microcystins (MCYSTs) are a group of cyclic heptapeptide toxins produced by several cyanobacterial species. Of the more than 60 MCYSTs characterized to date (19, 27, 29), most are potent inhibitors of protein phosphatases 1 and 2A from both plants and animals (17). One of the most common MCYST-producing cyanobacteria is the bloom-forming Microcystis aeruginosa (Kützing) Lemmermann. Due to the widespread distribution and potential toxicity of this species (toxic strains have been found worldwide), M. aeruginosa has been implicated in a number of animal-poisoning incidents (e.g., reference 7) and more recently in human fatalities (11, 23).

M. aeruginosa is a unicellular, colonial freshwater cyanobacterium which often forms blooms during warmer months in eutrophic lakes and reservoirs (37). For this reason, much research has been concerned with the environmental factors which lead to bloom formation and toxin production in this species. A wide range of batch culture studies have shown that the variables influencing MCYST content include trace metal supply (15), nitrogen (N) and phosphorus (P) (31), light and temperature (38), and pH (34). Comparative studies on MCYST production by M. aeruginosa in continuous culture, however, have been limited to examination of the effects of photon irradiance (35), N, P, and Fe³⁺ limitation (16, 36), and more recently P limitation (20). Despite this considerable pool of data concerning MCYST production, few studies (with the exception of the work carried out by Rapala and coworkers [25, 26]) have been able to quantitatively relate MCYST content to any growth determinant.

In a previous batch culture study, we presented data on the effect of N supply on the cellular production of MCYSTs (21). This work showed that the net specific rate of MCYST production was equal to the cell specific growth rate. The application of these findings to previously published batch culture studies suggested that the relationship held under a variety of culture conditions and that MCYST production was indirectly affected by environmental factors through their effects on cell division. A consequence of this linear relationship was that the cell quota of MCYST (Q_MCYST) should remain constant over a range of growth rates. However, Q_MCYST varied significantly, though inconsistently, throughout the growth cycle in separate batch culture experiments (21), suggesting a more complex relationship than predicted. Hence, in an attempt to specifically determine the relationships between growth rate, net rates of MCYST production, and Q_MCYST, we here report the results of a N-limited chemostat study using the same strain of M. aeruginosa MASH 01-A19 under growth conditions similar to those used in our previous batch culture study (21). MCYST data are expressed both as cell quotas and as a ratio to a number of biomass indicators (viz., protein, dry weight, and chlorophyll a [Chl a]) to emphasize the importance of cell quotas in determining cellular physiology of MCYST production.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Organism and growth conditions. M. aeruginosa MASH 01-A19 (3, 4) was provided by the CSIRO Marine Laboratories culture collection. Cells were grown in triplicate 500-ml continuous-culture vessels under constant illumination (40 ± 5 µmol of photons [PAR] m² s¹) using cool white fluorescent lights at 26 ± 1°C. Cultures were supplied with a continuous flow of sterile modified MLA medium (4) containing 0.2 mM NaNO₃ (1/10 original concentration), 0.02 mM K₂HPO₄, and 3.0 mM 2-(N-cyclohexylamino)ethanesulfonic acid (pH 8.0) via a Gilson Multiplus 2 peristaltic pump. Previous batch culture studies had revealed that the concentration of NO₃ used ensured N limitation (21). A single air pump provided constant airflow (through sterile, 0.45-µm-pore-size filters) to all cultures throughout the experiment. Cultures were grown at dilution rates ranging from 0.1 to 1.08 day¹ (as determined by flow rate). At steady state, dilution rate is equivalent to growth rate (24).

Estimation of µ_max. Triplicate batch cultures were used to estimate the maximum specific growth rate (µ_max) of M. aeruginosa MASH 01-A19 grown under the same medium, temperature, and light conditions as chemostats. The specific cell division rate and specific rate of dry weight accumulation were determined from a simple first-order rate law after frequent sampling of cultures postinoculation.

Sampling and analysis. To ensure steady-state conditions at each growth rate, the stability of culture populations was determined by cell counting and dry weight analysis. Once populations had stabilized at each growth rate (not more than 3% variation in cell concentrations or dry weight between four successive samplings), cultures were allowed to grow at steady state for at least five doubling times before sampling. Approximately 75 ml of each culture was removed for analysis of cell number, cell dry weight, MCYST content, cell volume, total cellular protein, and Chl a. Cell counting was done using a hemocytometer (Neubauer) after disruption of colonies in approximately 1.0 ml of culture by heating (80°C, 20 min), using the method of Humphries and Widjaja (10). The effect of heat treatment on cell volume was found to be minimal and consistent with minor volume changes noted by Porter and Jost (22) after collapsing gas vacuoles. For the determination of cell volume, the cells were concentrated by centrifugation (13,000 × g, 5 min), and photomicrographs of the cells and a standard scale (10 µm) were taken using an Olympus compound microscope fitted with a camera. Once developed, photomicrographs were digitally scanned and cross-sectional cell area was determined using NIH Image software (National Institutes of Health, Bethesda, Md.). The cross-sectional area of between 40 and 100 cells from each culture at each growth rate was determined and used to determine mean cell volume, assuming spherical cells. Dividing and nondividing cells were treated equally.

1

1

Data analysis. Where appropriate, data were statistically analyzed by regression analysis and analysis of variance using SPSS for Windows 10.0.5 (SPSS Inc.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Steady-state growth. As determined by consistent dry weight and cell concentrations, cultures were at steady state at each growth rate prior to sampling. Steady-state cell concentration significantly increased with increasing growth rate (P < 0.001), but cell dry weight decreased from 43.4 to 17.7 pg cell¹ (Table 1). Hence, steady-state biomass concentrations increased only slightly with increasing growth rates (P < 0.03) ranging from 50.1 to 61.9 mg liter¹ between the lowest and highest growth rates (0.10 to 1.08 day¹). The reduction in cell weight with increasing growth rate was associated with a decrease in cell volume of approximately fivefold, from 111 to 19.2 µm³ (Table 1).

MCYST analysis. As described in our previous study, two MCYST peaks were determined in M. aeruginosa MASH 01-A19 by HPLC and liquid chromatography-mass spectrometry analysis (LC-MS) (21). The first of these was MCYST-LR; although it probably contains desmethyl isomers of MCYST-LR (21), it is expressed in MCYST-LR molar equivalents (assuming similar molar absorption coefficients). The second MCYST, giving the characteristic absorbance maximum at 238 nm but an uncharacteristic LC-MS spectrum with high fragmentation, has not been identified and was not included in the measurement of total MCYST. At all growth rates, this compound constituted less than 15% of total MCYST-LR equivalents.

MCYST content and production. Q_MCYST ranged from 0.052 to 0.116 fmol cell¹ and showed a positive linear correlation with growth rate (r = 0.952) (Fig. 1). Extrapolation of the fitted regression suggests that Q_MCYST reaches a minimum value (Q_MCYSTmin) at µ = 0 and a maximum value (Q_MCYSTmax) at µ_max (Fig. 1). The linear relationship between Q_MCYST and µ can be described in terms of the predicted Q_MCYSTmin and Q_MCYSTmax, and µ_max as follows:

(1)

Since µ_max cannot be achieved in chemostat cultures, this parameter was determined from analysis of separate batch culture data and estimated to be 1.2 day¹. Using this value, Q_MCYSTmin and Q_MCYSTmax were subsequently calculated from linear regression analysis to be 0.050 ± 0.004 (standard error [SE]) and 0.129 ± 0.006 (SE) fmol cell¹, respectively (Table 2).

View larger version (22K): [in this window] [in a new window]   FIG. 1.   Cellular QMCYST of M. aeruginosa grown in N-limited chemostats. Error bars represent standard errors of the means of triplicate chemostat cultures. The solid line shows cell quotas predicted from experimental growth rates (r = 0.952) using equation 1 (see text), and the dashed line represents the extrapolation of this relationship to µmax (where QMCYST = QMCYSTmax) and µ = 0 (where QMCYST = QMCYSTmin). d, day.

1

View larger version (19K): [in this window] [in a new window]   FIG. 2.   MCYST expressed as a ratio to dry weight (A), to protein (B) to Chl a (C), and on an intracellular concentration basis (D) in M. aeruginosa grown in N-limited chemostats. Error bars represent the standard errors of the means of triplicate chemostats. d, day.

View larger version (17K): [in this window] [in a new window]   FIG. 3.   Intracellular MCYST concentration in cells of M. aeruginosa, grown in N-limited chemostats, as a function of cell volume.

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

(2)

1

1

1

1

View larger version (20K): [in this window] [in a new window]   FIG. 4.   RMCYST as a function of growth rate in M. aeruginosa grown in N-limited chemostats. Error bars represent standard errors of the means of triplicate cultures. The solid line shows net rates of MCYST production calculated using equation 2 (see the text) and the values in Table 2.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

This study shows, for the first time, that the cellular MCYST content of N-limited M. aeruginosa can be predicted from growth rate, with faster-growing cells containing higher intracellular concentrations of MCYST. We believe that these results were achievable by ensuring that cultures were maintained in steady-state, nitrogen-limited growth conditions at all times. Our data also highlight the importance of determining cellular MCYST quotas in experiments examining MCYST content and production. Clearly there are difficulties in interpretation that may arise when MCYST content is measured as a ratio to another cell component that itself may be varying independently in response to a change in growth rate or experimental treatment. This is evidenced by our observations that with increasing growth rate, MCYST increased linearly as a ratio to cell dry weight, generally increased but reached a maximum at intermediate growth rate as a ratio to protein content, and yet was invariant as a ratio to Chl a.

The model put forward in equation 1 proposes that Q_MCYST at any growth rate is dependent on the constants Q_MCYSTmax, Q_MCYSTmin, and µ_max in N-limited cultures (Table 2; Fig. 1). The parameters Q_MCYSTmax and Q_MCYSTmin determine a fixed range of cellular MCYST quotas. This implies that toxigenic strains will always contain, at least, a minimum Q_MCYST and that they will not exceed a maximum Q_MCYST determined by the nutrient saturated µ_max (for the given temperature and light growth conditions). Although a growth rate of zero cannot be achieved in a chemostat, our predicted Q_MCYSTmin is very similar to that observed for batch cultures of this strain at stationary phase, where Q_MCYST remained stable for at least 2 weeks (21). In the same study, Q_MCYSTmax ranged from 0.13 to 0.16 fmol cell¹, again similar to the value predicted from this chemostat study (Table 2). Collectively, the data are consistent with the generalization that MCYST production is constitutive and that toxigenic strains remain so under a variety of growth conditions (33). In support of this conclusion, data from other researchers suggest that minimum and maximum cell quotas of MCYSTs exist in other strains, and even in different Microcystis spp. (34, 40). We also note that the maximum MCYST/dry weight ratio reported in this study (7.6 mg g¹ [dry weight]) is very similar to that found in late log-phase of the original strain MASH 01 (the parent culture) by Bolch et al. (5) (i.e., 7.24 mg g¹ [dry weight]), indicating a conserved process of toxin production in this strain for several years.

Böttcher et al. (6) recently found Q_MCYST to remain constant while µ increased with increasing irradiance in turbidostat experiments. These results may at first appear to contradict ours. However, light-limited turbidostats differ from chemostats in that µ at any irradiance is always nutrient saturated µ_max, and therefore Q_MCYST always equals nutrient saturated Q_MCYSTmax. Thus, their findings suggest a constant Q_MCYSTmax while nutrient saturated µ_max increases with increasing irradiance. In contrast, recent batch culture studies under nutrient-replete conditions over a range of temperatures revealed that Q_MCYSTmin decreased in response to increasing temperature (B. M. Long, unpublished data). Further work is needed to confirm the exact details of how Q_MCYSTmin or Q_MCYSTmax may vary in response to physical conditions limiting growth (temperature, irradiance, etc.).

In addition to describing the cell quota of MCYST, the constants Q_MCYSTmin, Q_MCYSTmax, and µ_max also determine the net rate of MCYST production (equation 2). R_MCYST is the product of Q_MCYST and µ, and as a consequence, equation 2 predicts no net MCYST production at µ = 0 (or stationary phase in batch culture). Also, R_MCYST is constrained by the maximum cell division rate, as Q_MCYST will not exceed that which is achieved at µ_max (i.e., R_MCYSTmax = µ_max × Q_MCYSTmax). This is inconsistent with the regression model advanced by Oh et al. (20), however, which predicted a net production of MCYST at µ = 0 (0.082 mg g¹ [dry weight] day¹). This implies that when cell division stops, MCYST production continues, resulting in increasingly toxic nondividing cells. We can find no other published data to support this proposition. In addition, Oh et al. (20) found that the MCYST/dry weight ratio correlated negatively with µ in P-limited chemostats. When MCYST data are expressed as a ratio to another cell constituent (e.g., protein) or group of constituents (e.g., dry weight) which may be under independent and varying cellular regulation in response to the limiting nutrient, it is very difficult to understand the cellular regulation of MCYST content and production. Cyanobacterial dry weight is affected differentially by N and P limitation (1), demonstrating that the physiological regulation of dry weight production is quite different under different nutrient limitations. We suggest that the observed differences between our findings and those of Oh et al. (20) may result from differential dry weight changes under N and P limitation. Hence, simple comparisons of MCYST/dry weight ratio data cannot be made between cultures grown under different nutrient limitations. The near absence of MCYST cell quotas from the existing literature makes comparison of our data with other studies almost impossible.

Shi et al. (30) found that MCYST was associated with the thylakoid membranes of M. aeruginosa PCC 7820, suggesting a close physical association between MCYSTs and the photosynthetic machinery of the cell. The constant ratio of MCYST to Chl a (1:2 [mol:mol] [Fig. 2C]) found in this study supports this contention and suggests that MCYST synthesis and or function could be linked to photosynthetic processes. The absence of reports of major perturbations in the photosynthetic activity of M. aeruginosa PCC 7806 after knocking out MCYST production (8), however, would suggest that MCYSTs are not essential in photosynthesis. Nevertheless, the MCYST synthetase knockout mutant of strain PCC 7806 was found to have slightly altered thylakoid structure and also to exhibit irregularities in the structure of gas vesicles (E. Dittmann and T. Börner, personal communication).

The decrease in cell size with increasing growth rate is consistent with cell volume variations reported previously for M. aeruginosa by Krüger and Eloff (13). The same authors suggest that cell size is a likely indicator of the physiological state of a cell, with stressed cells being larger. This supports the generally held view that MCYST production is greatest when conditions are favorable for growth (32, 33), as larger cells occur at lowest growth rates (Table 1).

Our finding that smaller cells contain more MCYST than larger ones (Fig. 3) may have implications for toxicity toward grazing zooplankton. Since some daphnids, though notably not all (18), are sensitive to toxic strains of M. aeruginosa (28), it is conceivable that zooplankton could ingest a greater number of smaller Microcystis spp. cells, thus receiving a considerably larger dose of toxin. This is speculation, however, and further work is required to determine the importance of cell size and toxin ingestion rates.

Given the observed relationship between µ and Q_MCYST, our model predicts that Q_MCYST can be determined for any value of µ if µ_max, Q_MCYSTmax, and Q_MCYSTmin are known. Since obtaining these constants in a chemostat study is time-consuming, we suggest that a more practical approach can be made with less complicated apparatus. As Q_MCYSTmin represents Q_MCYST at nitrogen-limited stationary phase in a batch culture, and Q_MCYSTmax represents Q_MCYST during nitrogen-saturated logarithmic growth, these parameters can be determined from a single batch culture experiment (Fig. 5). The only caveat is that the initial nitrogen concentration in the batch culture medium must be sufficient to ensure nitrogen-saturated growth during logarithmic phase but not so high as to allow high biomass development to the point of self-shading (light limitation) or CO₂ limitation; i.e., stationary phase must arise only due to nitrogen depletion.

View larger version (17K): [in this window] [in a new window]   FIG. 5.   Theory for the determination of the constants QMCYSTmax and QMCYSTmin from batch cultures. QMCYSTmax occurs shortly after inoculation when cells are in log phase (µ = µmax). QMCYSTmin will occur in batch cultures at stationary phase (µ = 0). The determination of cell quotas of MCYST at these times should permit the relationship between µ and QMCYST to be calculated according to equation 1 (see text).

Our findings quantitatively demonstrate that under N-limited growth, Q_MCYST in M. aeruginosa is a function of µ. As µ is controlled by cellular N quota (Q_N) under N-limited growth (9), this is consistent with Q_MCYST also being regulated by Q_N. Whether this is the case, or whether there is a more general relationship between Q_MCYST and growth limitation by any environmental factor, remain to be elucidated.