Variations of Growth and Toxin Yield in Cylindrospermopsis raciborskii under Different Phosphorus Concentrations

The bloom-forming cyanobacteria, Cylindrospermopsis raciborskii, is a producer of the cytotoxic cylindrospermopsin (CYN). In this study, the growth, toxin yield, and expression of CYN biosynthesis genes of C. raciborskii were examined under varying phosphorus (P) concentrations. The results show the cell number at 0.00 and 0.01 mg·L−1 P was significantly lower than that at higher P concentrations (≥0.5 mg·L−1). The chlorophyll a content, filament length, heterocyst, and akinete numbers at P ≤ 0.05 mg·L−1 were also significantly reduced. The intracellular and extracellular CYN concentrations and the extracellular proportions increased during the culture period, and larger values were observed at higher P concentrations. Total CYN content reached 45.34–63.83 fg·cell−1 and extracellular CYN proportion reached 11.49%–20.44% at the stationary growth phase. A significantly positive correlation was observed between CYN production and cell growth rate. Three cyr genes were expressed constantly even at P-deficient conditions. The transcription of cyr genes at P-replete conditions or after P supplementation increased from 1.18-fold to 8.33-fold. In conclusion, C. raciborskii may rapidly reorganize metabolic processes as an adaptive response to environmental P fluctuations. CYN production and cyr gene expression were constitutive metabolic processes in toxic C. raciborskii.


Introduction
Increased incidence of harmful cyanobacterial blooms were amongst the most severe environmental problems over the past two decades. Cyanobacterial blooms have attracted considerable scientific attention because of their potential risks related to human sickness and animal mortality and the disruption of aquatic ecosystems [1,2]. Toxic metabolites of cyanobacteria, i.e., cyanotoxins, can be classified into four major categories based on their mode of action. Among these compounds are the hepatotoxic microcystins [3] and nodularins [4]; the cytotoxic cylindrospermopsins (CYNs) [5]; the neurotoxic saxitoxins [6], anatoxin-a [7], and beta-methylamino-L-alanine [8]; and the dermatotoxic Lyngbya toxins [9] and aplysiatoxins [10]. Specifically, the alkaloid CYN, which is composed of a tricyclic guanidine moiety coupled with a hydroxymethyluracil and a sulfonic acid group, is of particular exponential growth in seven cultured strains of C. raciborskii. In addition, Bormans et al. [51] observed no relationship between CYN concentration and growth rate in benthic species Oscillatoria sp. PCC 6506 and the percentage of extracellular CYN varied between 56% and 96%. Moreover, several studies have inferred that the release of CYN increases as cells age, based on observations that the highest dissolved CYN content was found in older blooms [46] or during stationary phase of batch cultures [49,52]. Consistent with these reports, Davis et al. [53] regarded CYN release into the water column as a result of cell lysis during the stationary phase or environmental stress.
Given the critical importance of P availability in growth and blooms development of cyanobacteria, we investigated the growth and toxin yield of a toxic C. raciborskii strain under a range of inorganic P concentrations. This study measured the changes in morphology, toxin yield and the expression of cyr genes in C. raciborskii. The key genes chosen were involved in CYN biosynthesis and transport, such as cyrA (amidinotransferase, the first step in CYN biosynthesis), cyrJ (sulfotransferase), and cyrK (putative transport).

CYN Measurement
CYN contents in the C. raciborskii culture were measured and normalized to the total cell numbers ( Figure 2). The cell quotas of intracellular, extracellular, and total CYN gradually increased under all treatments over the experimental period. The total CYN cell quota was calculated as the sum of intracellular and extracellular CYN cell quota. Moreover, the groups with higher P concentrations had larger CYN amounts than the groups with lower P concentrations at most time points. Intracellular CYN accounted for the majority of total CYN with 38.72 ± 1.12, 39.21 ± 1.61, 44.71 ± 4.08, 45.06 ± 5.78, and 50.79 ± 2.13 fg·cell −1 for P concentrations of 0.00, 0.01, 0.05, and 0.50 mg·L −1 and the control group, respectively ( Figure 2). A minor portion of CYN was extracellular with 6.75 ± 0.32, 6.13 ± 0.66, 5.80 ± 1.00 and 7.61 ± 1.16 fg·cell −1 for cells grown in P concentrations of 0.00, 0.01, 0.05, and 0.50 mg·L −1 , which were significantly lower than 13.04 ± 0.14 fg·cell −1 in the control group (p < 0.01; Figure 2). The final total CYN contents reached 45.47 ± 1.43, 45.34 ± 2.23, 50.51 ± 4.43, 52.67 ± 6.59, and 63.83 ± 2.09 fg·cell −1 for P concentrations of 0.00, 0.01, 0.05, and 0.50 mg·L −1 and the control group, respectively ( Figure 2). With regard to intracellular CYN and total CYN, no significant differences were found among cells grown with P ≤ 0.05 mg·L −1 (p > 0.05), while the values of cells grown with P > 0.05 mg·L −1 were significantly higher than that of those under P depletion condition (p < 0.01).  The percentage of extracellular CYN also increased during the sampling period. At the last day of the experiment, a significantly higher CYN percentage was observed to be extracellular in the control group (20.44% ± 0.75%), compared with extracellular percentages of 14.83% ± 0.23%, 13.50% ± 0.81%, 11.49% ± 1.86%, and 14.45% ± 1.60% at P concentrations of 0.00, 0.01, 0.05, and 0.50 mg·L −1 , respectively (p < 0.01; Figure 2). Furthermore, both µ CYN-total and µ CYN-in showed a significantly positive correlation with µ c (p < 0.05; Figure 3). The slope were 0.96 and 0.95, respectively, both of which were not significantly different from 1.00 (ANCOVA: p = 0.83 for µ CYN-total and p = 0.80 for µ CYN-in ).   Correlations between specific toxin production rate (µ CYN ) and the corresponding specific growth rate (µ c ) during culture period. (A) µ CYN-total represents specific total toxin production rate and (B) µ CYN-in represents specific intracellular toxin production rate. The linear regression equations are as follows: (A) y = 0.96x + 0.02, R 2 = 0.54, p = 0.00; and (B) y = 0.95x + 0.01, R 2 = 0.48, p = 0.00.

cyr Gene Expression
The expression levels of three cyr genes, namely, cyrA, cyrJ, and cyrK, were examined, and their transcript abundances were expressed as the fold changes with respect to time = 0. In the P starvation stage, upregulated expression levels of cyrA and cyrK genes were observed for the control and experimental groups compared with the initial expression levels ( Figure 4A,C). The expression of cyrJ gene was also increased in the early stage of P starvation and was reduced to initial levels subsequently ( Figure 4B). Furthermore, the expression levels of cyrA gene in the P-depleted group were 12%, 26%, and 72% of those in the control group after 0.5, 1, and 7 days, respectively, with significant difference (p < 0.05). The expression levels of cyrJ gene in the P-depleted group were 25%, 36%, and 77% of those in the control group after 0.5, 1, and 2 days, respectively, with significant difference (p < 0.05). The expression levels of cyrK gene in the P-depleted group were 20%, 57%, and 85% of those in the control group after 0.5, 1, and 5 days, respectively, with significant difference (p < 0.05). However, a significant 1.74-fold increase in cyrK gene expression was observed in the P-depleted group after seven days compared with the control group. Moreover, no significant difference was observed for the expression levels of cyr genes between P-depleted and control groups at the last day of P starvation (p > 0.05).  After P supplementation on the eighth day, the expression levels of cyrA and cyrK genes were significantly increased to 2.81-fold and 1.77-fold, respectively, at 0.50 mg·L −1 P compared with those in the P-depleted group after 6 h (p < 0.05; Figure 4D,F). Similarly, the expression levels of cyrJ gene at 0.05 and 0.50 mg·L −1 P were 1.97-fold and 2.23-fold higher, respectively, than those in the P-depleted group after 6 h, with significant difference (p < 0.05; Figure 4E). Afterward, the cyr gene expression levels at 0.05 mg·L −1 P were reduced and were 55%, 60%, and 33% of those in the P-depleted groups for the cyrA, cyrJ, and cyrK genes, respectively, after 24 h, with significant difference (p < 0.05). The cyr gene expression levels at 0.50 mg·L −1 P were also reduced and were 30%-79%, 41%-65%, and 32%-61% of those in the P-depleted group for the cyrA, cyrJ, and cyrK genes, respectively, with significant difference, except after 12 h. The cyr gene expression levels at 0.01 mg·L −1 P were not significantly different from that in the P-depleted group at most time points, except for significantly higher expression levels of the cyrA and cyrJ genes at 0.01 mg·L −1 P after 48 h.

Correlation between the Biomass of Cylindrospermopsis and TP Concentration in Freshwater Bodies
Among a series of phytoplankton species and water quality parameters, we centered on Cylindrospermopsis and TP concentration. The percentage of Cylindrospermopsis in total cyanobacterial biomass showed the tendency of negative correlation with TP concentration (y = −1.68x + 6.44, R 2 = 0.50, p = 0.00), compared with the positive correlation between the percentage of Microcystis and TP concentration (y = 0.60x + 0.58, R 2 = 0.08, p = 0.03) ( Figure 5).

Discussion
In this study, we explored the P-dependent growth, toxin yield, and the expression levels of the cyr genes cyrA, cyrJ, and cyrK, in C. raciborskii CHAB3438. This strain could grow at very low P concentrations during the experimental period, and this corresponds with the high P affinity and large P storage that has been previously reported for C. raciborskii [54][55][56]. However, the chlorophyll a content per cell was lower at the lower P concentrations compared to the higher P concentrations. Additionally, filament length and formation of heterocysts and akinetes were also reduced at low P concentrations. The explanation for this physiological response is that C. raciborskii is able to economize resources by adjusting metabolic pathways and reducing high energy-consuming physiological processes under limited nutrient supply, a response also found in Microcystis aeruginosa [57]. The phenotypic plasticity of multiple traits may allow C. raciborskii to optimize growth in P fluctuations, which may explain its dominance in the phytoplankton of many aquatic ecosystems, particularly those with low P concentrations [58]. Furthermore, field surveys of this study also showed a trend where the relative abundance of Cylindrospermopsis increased with the decline of TP concentration in lakes. This result is consistent with the investigation of Bonilla et al. [59] that observed a high biomass of C. raciborskii under low TP concentration. Therefore, Cylindrospermopsis is highly competitive and adaptive under low P concentrations compared with other bloom-forming cyanobacteria, which significantly promotes its invasive spread.
Previous investigations revealed that P deprivation inhibited CYN production in Chr. ovalisporum [60]. On the other hand, replete P supply was found to boost CYN yield in C. raciborskii and cyanobacterial blooms samples [61,62]. However, only intracellular CYN was measured in these studies, and the extracellular portion of CYN was ignored. On the contrary, an increase in the total CYN content was reported under P-depleted conditions, coupled with upregulated expression levels of the CYN biosynthesis genes [45,63]. Therefore, it is necessary to explore the extracellular CYN concentration under different P concentrations to clarify the contradiction of these results. In this study, the concentrations of intracellular, extracellular and total CYN, as well as the extracellular proportion of CYN, all increased during the culture time and at all P concentrations. The CYN production rate exhibited a positive correlation with cell growth rate, therefore, CYN production is constitutive in C. raciborskii CHAB3438. The effects of P concentrations on these processes were produced through affecting cell growth. Therefore, CYN yield would be reduced when cell maximum biomass is reduced under P-limited conditions. These results were supported by previous findings that the CYN pool size of C. raciborskii was constant under different nitrogen, phosphorus, light, and CO 2 conditions [64,65].
At the end of the experiment, the proportion of extracellular CYN reached 11.49% to 20.44% of the total CYN, indicating that CYN is mainly stored in cyanobacterial cells, whereas only a minor proportion of CYN is released into the extracellular environment, consistent with the results of Orr et al. [66] with field cyanobacterial samples. Cell lysis may contribute to extracellular CYN accumulation during the stationary phase as suggested in previous research [53]. In addition, active CYN release from cyanobacterial cells has been emphasized in other studies [63], but it remains to be proved. The hypothetical CYN transporter CyrK has been deduced from the biosynthesis cluster [20] but its function needs to be confirmed by examining a mutant strain lacking the cyrK gene.
The transcription of three cyr genes that are involved in CYN biosynthesis and transport was examined to better understand the potential response of cyr gene cluster in C. raciborskii to various P concentrations. The results showed significantly high expression levels of cyr genes under P-replete conditions in the early stage of P starvation and in the initial 6 h of P supplementation thereafter. Afterwards, the expression levels of cyr genes at various P concentrations maintained similar levels during the experimental period; however, significantly lower expression levels were observed for 0.05 and 0.50 mg·L −1 P in comparison to other P concentrations at several time points. Although the cyr gene expression did not completely coincide with the production and release of CYN, which was constitutive as aforementioned, the constant expression of cyr genes further confirmed CYN as a constitutive metabolite of the toxic C. raciborskii. In a study with C. raciborskii CS-505, Stucken et al. [67] investigated the relationship between expression levels of four cyr genes and CYN production in response to nitrogen source. They also found that these genes showed a constitutive transcription and concluded that CYN biosynthesis was not regulated at the transcriptomic level. Furthermore, Pierangelini et al. [64] reported differential expression of cyr genes in contrast to constant CYN cell quotas in different growth stages of C. raciborskii, and suggested CYN biosynthesis was post-transcriptionally regulated. Our results support the post-transcriptional regulation of CYN for a constant cell quota.

Conclusions
In summary, the morphology of C. raciborskii and the expression of the genes cyrA, cyrJ, and cyrK, altered in response to variations in environmental P, however, CYN production and release remain constitutive processes. The majority of CYN occurred intracellularly with only a small proportion (4.38%-14.83%) occurring extracellularly. Finally, the toxin yield of C. raciborskii blooms would be reduced with a reduction in available P, as P limits the maximum total cell number.

Strain and Culture Conditions
CYN-producing strain C. raciborskii CHAB3438 was previously isolated from Lake Xianghu and maintained in liquid MA medium [68]. Quintuplicate 1 L pure cultures in sterile 2 L Erlenmeyer flasks were incubated in 25 • C with a light intensity of 35 µmol m −2 ·s −1 (12/12 h light/dark) provided by cool white fluorescent tubes. The cultures were manually shaken thrice daily during incubation until the exponential phase.

Phosphorus Starvation and Supplementation
The cultures were filtered using Millipore membranes (3.0 µm pore size, Merck Millipore, Darmstadt, Germany), washed thrice with sterile P-free MA medium, resuspended, and subcultured into five groups. Each group included triplicate cultures containing 50 mL concentrated cultures and 1150 mL P-free MA medium (four experimental groups) or full medium (one control group with 10 mg·L −1 P). The starting cell concentration was 2.36 × 10 6 ± 2.10 × 10 5 cells·mL −1 . The cultures were incubated for eight days to deplete intracellular polyphosphate stores. Afterward, the four experimental groups were supplemented with K 2 HPO 4 solution to obtain the final P concentrations of 0.00, 0.01, 0.05, and 0.50 mg·L −1 . Fresh media with corresponding P concentrations were supplemented after each sampling to maintain constant culture volumes. Samples were collected on alternate days to determine cyanobacterial growth, CYN content, and gene expression level. In the P starvation stage, three of the 12 experimental cultures were randomly selected to measure each parameter. Therefore, the data of growth and CYN are shared by four experimental groups during the first eight days of P starvation.

Measurement of Growth
Growth was measured spectrophotometrically at 750 nm. Subsamples (1 mL) were preserved in Lugol's iodine solution for microscopic measurement (Olympus BX51, 400× magnification, Olympus, Tokyo, Japan) of morphological features, including cell concentration, heterocyst and akinete numbers, and filament length. Cell concentration was measured as described in Pierangelini et al. [69]. Filament length measurements were performed on at least 100 trichomes. For chlorophyll a measurement, 20 mL of cell culture was filtered through a glass fiber filter (47 mm diameter, 1.2 µm pore size; Whatman GF/C) and extracted using 90% acetone solution in the dark for 24 h at 4 • C. The extracts were centrifuged at 8000 rpm for 30 min, and the absorbance of the supernatant was measured at 750, 663, 645, and 630 nm. The concentration of chlorophyll a was calculated using the formula applied by the China Environmental Protection Administration [70].

Extraction and Quantitation of CYN
A cell culture volume of 25-50 mL was filtered by Millipore filters (0.22 µm pore size). Intracellular and extracellular CYN were extracted from filters and filtrates, respectively. The collected cells on the filters were resuspended in 1 mL distilled water, fractured by a SPEX 6870 Freezer Mill for 10 min, shaken for 1 h at room temperature, and centrifuged at 12,000 rpm for 30 min. The supernatants were collected and combined after the extraction step was repeated. The supernatants and filtrates were subjected to solid-phase extraction to enrich CYN, as described previously [71]. Afterward, the elutes were further concentrated by rotary evaporation. The precipitate was redissolved in 0.5 mL of Millipore water and filtered through an ultra-centrifugal filter (10 kDa). CYN was detected by high-performance liquid chromatography, and the column and elution conditions were set as descried by Jiang et al. [48]. Standard CYN (Enzo Life Sciences, New York, NY, USA) was used for qualitative and quantitative analyses. CYN concentrations in the samples were calculated based on UV absorbance at 262 nm. The specific growth rate (µ c ) and specific toxin production rate (µ CYN ) were calculated using the equation: where C 1 and C 2 are the cell number (cells·mL −1 ) and CYN concentration (µg·CYN·L −1 ) at times t 1 and t 2 , respectively [64]. Data of days 3-21 were used to calculate µ c and µ CYN . The slope of regression curves were statistically analyzed, as described previously [67].

RNA Isolation and cDNA Synthesis
Cells were collected via centrifugation of 9.0-40 mL of culture at 8000 rpm, 4 • C for 2 min, and total RNA was extracted using the RNAiso Plus Kit (Takara, Otsu, Japan) according to the manufacturer's instructions. DNA digestion was performed after RNA isolation with RNase-free DNase (Promega, Madison, WI, USA) and RNase Inhibitor (Thermo, Waltham, MA, USA). RNA quality and purity was assessed using a Nanodrop spectrophotometer (Thermo, Waltham, MA, USA) and a total of 900 ng of RNA was taken for reverse transcription (RT) reaction using the PrimeScript™ RT reagent Kit with gDNA Eraser (Takara, Otsu, Japan) according to the manufacturer's instructions. Genomic DNA was further removed with gDNA Eraser during this process. The absence of genomic DNA in the RNA samples was then verified by a RT-PCR control containing no reverse transcriptase.

Quantitative PCR (qPCR) Assay
A three-step qPCR programme was performed in an iCycler iQ5 (Bio-Rad, Hercules, CA, USA). The procedure consisted of 95 • C for 1 min, followed by 40 cycles of 95 • C for 15 s, 55 • C for 15 s, 72 • C for 45 s and then a melt-curve cycle of 0.5 • C increases at 15 s intervals to verify the quality of the product. All reactions were run in triplicate, which comprised 10 µL 2× THUNDERBIRD SYBR qPCR Mix (Toyobo, Osaka, Japan), 0.2 µM forward and reverse primers, 2 µL cDNA and sterile Millipore water to a final reaction volume of 20 µL. Sanger sequencing was performed on the amplicons to confirm the presence of cyr gene fragments. The qPCR primers were designed according to cyr gene sequences of C. raciborskii CHAB3438 (GenBank accession number: KJ139743). Standard curve of each primer pair was established by 10-fold dilutions of a PCR template. The qPCR primer sequences, amplicon sizes and amplification efficiencies are shown in Table 1. The time = 0 point sample was used as a calibrator and 16S rDNA was chosen as a reference gene for calculating fold changes of cyrA, cyrJ and cyrK genes using the 2 −∆∆Ct method, where ∆∆C t = (C t, target gene − C t, 16S rrn ) t − (C t, target gene − C t, 16S rrn ) t = 0 . (2)

Collection of Water Samples
A total of 55 sampling sites belonging to 23 shallow lakes were covered in this study. These water bodies were located in the east of China and water samples were collected in the summer of 2013 (Table S1). Analysis of water quality parameters and phytoplankton species composition were conducted as previously described [72].

Statistical Analysis
Data are presented as the mean ± standard deviation. Statistical difference was tested by one-way ANOVA and Tukey's post hoc comparison test implemented in SPSS v19.0 software (IBM, Chicago, IL, USA) for Windows. Differences with p values less than 0.05 were considered significant.