The Bloom-Forming Dinoflagellate Karenia mikimotoi Adopts Different Growth Modes When Exposed to Short or Long Period of Seawater Acidification

Impacts of ocean acidification (OA) on noncalcifying organisms and the possibly responsible mechanism have aroused great research interests with the intensification of global warming. The present study focused on a noxious, noncalcifying, bloom-forming dinoflagellate, Karenia mikimotoi (K. mikimotoi), and its variation of growth patterns exposed to different periods of seawater acidification with stressing gradients was discussed. The dinoflagellates under short-time acidifying stress (2d) with different levels of CO2 presented significant growth inhibition (p < 0.05). The cell cycle was obviously inhibited at S phase, and the photosynthetic carbon fixation was also greatly suppressed (p < 0.05). Apoptosis was observed and the apoptotic rate increased with the increment of pCO2. Similar tendencies were observed in the key components of mitochondrial apoptotic pathway (the mitochondrial membrane potential (MMP), Caspase-3 and -9, and Bax/Bcl-2 ratio). However, under prolonged stressing time (8 d and 15 d), the growth of dinoflagellates was recovered or even stimulated, the photosynthetic carbon fixation was significantly increased (p < 0.05), the cell cycle of division presented little difference with those in the control, and no apoptosis was observed (p > 0.05). Besides, acidification adjusted by HCl addition and CO2 enrichment resulted in different growth performances, while the latter had a more negative impact. The results of present study indicated that (1) the short-time exposure to acidified seawater led to reduced growth performance via inducing apoptosis, blocking of cell cycle, and the alteration in photosynthetic carbon fixation. (2) K. mikimotoi had undergone adaptive changes under long-term exposure to CO2 induced seawater acidification. This further demonstrated that K. mikimotoi has strong adaptability in the face of seawater acidification, and this may be one of the reasons for the frequent outbreak of red tide. (3) Ions that dissociated by the dissolved CO2, instead of H+ itself, were more important for the impacts induced by the acidification. This work thus provides a new perspective and a possible explanation for the dominance of K. mikimotoi during the occurrence of HABs.


Results
Different growth modes of microalgae exposed to short or long period of seawater acidification were explored from three aspects, namely photosynthetic carbon fixation, that was extremely sensitive to changes in CO 2 concentration, as well as apoptosis and cell cycle, which were closely related to cell proliferation.

Changes of Population Dynamics to Different pCO 2 Levels
The growth of K. mikimotoi in different pCO 2 treatments is shown in Figure 1. The specific growth rate (µ) was inhibited in both treatment groups during the first four days after inoculation, while a relatively higher one was found thereafter compared with that of the control. A logistic equation was applied to simulate the growth performance, and the results showed the carrying capacity (K) as well as the maximum instantaneous growth rate (r) in either treatment groups to be higher than those in the control. However, the smaller Tp (the inflection point of the population growth curve) of the treatment groups may be related to the significantly increase in growth rate which made the population enter the stationary phase faster (Table 1). It seems that short-term acidification stress inhibited the population increase within a certain range, but relatively long-term (4~14 days in the present study) acidification stress stimulated growth. This observation differs with the documented results on microalgae exposed to environmental stress and carbon-related processes because the specificity of CO 2 treatment was speculated to provide more evidence. study. Therefore, the growth performances of K. mikimotoi exposed to different acidification levels lasting for different durations were analyzed, and the possible adaptive mechanism involved in this process was discussed. The present study would provide a new perspective and a possible explanation for the dominance of K. mikimotoi during the occurrence of HABs regarding physiological adaption to seawater acidification.

Results
Different growth modes of microalgae exposed to short or long period of seawater acidification were explored from three aspects, namely photosynthetic carbon fixation, that was extremely sensitive to changes in CO2 concentration, as well as apoptosis and cell cycle, which were closely related to cell proliferation.

Changes of Population Dynamics to Different pCO2 Levels
The growth of K. mikimotoi in different pCO2 treatments is shown in Figure 1. The specific growth rate (μ) was inhibited in both treatment groups during the first four days after inoculation, while a relatively higher one was found thereafter compared with that of the control. A logistic equation was applied to simulate the growth performance, and the results showed the carrying capacity (K) as well as the maximum instantaneous growth rate (r) in either treatment groups to be higher than those in the control. However, the smaller Tp (the inflection point of the population growth curve) of the treatment groups may be related to the significantly increase in growth rate which made the population enter the stationary phase faster (Table 1). It seems that short-term acidification stress inhibited the population increase within a certain range, but relatively long-term (4~14 days in the present study) acidification stress stimulated growth. This observation differs with the documented results on microalgae exposed to environmental stress and carbon-related processes because the specificity of CO2 treatment was speculated to provide more evidence. Figure 1. The specific growth rate of K. mikimotoi exposed to different pCO 2 levels. * indicated the significant difference between the treatment groups with the control group at the p < 0.05 level, and data were expressed as means ± SD (n = 3). The increase in pCO 2 decreased the activity of CA during the first eight days after exposure, and the most obvious inhibition was found in the 2000 ppmv treatment group compared with that of the control (p < 0.01). However, the opposite occurred after 15 days of exposure. The CA activity of the 2000 ppmv group was significantly increased compared to the control (p < 0.05) (Figure 2A). Different tendencies were observed in the changes of Rubisco activity. In fact, a clear time-dependent trend was observed in the 2000 ppmv treatment group, and the elevated pCO 2 significantly decreased the activity of Rubisco in K. mikimotoi at 24 h after exposure. Rubisco activity increased steadily over time from the day 8 and reached its peak at the end of the experiment. Only a slight difference was observed in the 1000 ppmv group compared with the control group during the whole treatment ( Figure 2B). Short-term exposure to seawater acidification inhibited the photosynthetic carbon fixation ability of K. mikimotoi, but with long-term exposure, the photosynthetic carbon fixation ability could recover and even improve. As a result, there was a consistency between the population dynamics and photosynthetic carbon fixation, and both changed with acidification time. The alteration in photosynthetic carbon fixation may affect the energy distribution of K. mikimotoi and then the population dynamics. This result provided a possible explanation for the changes in the population dynamics.

Alteration in the Photosynthetic Carbon Fixation of K. mikimotoi Exposed to Different pCO2 Levels
The increase in pCO2 decreased the activity of CA during the first eight days after exposure, and the most obvious inhibition was found in the 2000 ppmv treatment group compared with that of the control (p < 0.01). However, the opposite occurred after 15 days of exposure. The CA activity of the 2000 ppmv group was significantly increased compared to the control (p < 0.05) (Figure 2A). Different tendencies were observed in the changes of Rubisco activity. In fact, a clear time-dependent trend was observed in the 2000 ppmv treatment group, and the elevated pCO2 significantly decreased the activity of Rubisco in K. mikimotoi at 24 h after exposure. Rubisco activity increased steadily over time from the day 8 and reached its peak at the end of the experiment. Only a slight difference was observed in the 1000 ppmv group compared with the control group during the whole treatment ( Figure 2B). Short-term exposure to seawater acidification inhibited the photosynthetic carbon fixation ability of K. mikimotoi, but with long-term exposure, the photosynthetic carbon fixation ability could recover and even improve. As a result, there was a consistency between the population dynamics and photosynthetic carbon fixation, and both changed with acidification time. The alteration in photosynthetic carbon fixation may affect the energy distribution of K. mikimotoi and then the population dynamics. This result provided a possible explanation for the changes in the population dynamics. Carbonic anhydrase activity of K. mikimotoi exposed to different pCO2 levels; (B) Rubisco activity of K. mikimotoi exposed to different pCO2 levels. * indicated the significant difference between the treatment groups with the control group at p < 0.05 level and ** indicated the most significant at p < 0.01 level, the data are shown as the mean ± SE (n = 3).

Alteration of Cellular Apoptosis of K. mikimotoi When Exposed to Different pCO2 Levels
Elevated pCO2 induced cellular apoptosis of K. mikimotoi at 24 h after exposure, and the apoptotic rate greatly increased in a concentration-dependent manner from approximately 2.8% in the control to 10.41% and 13.03% in the 1000 ppmv and 2000 ppmv groups, respectively ( Figure 3A). Statistical significance was observed between the treatment groups and the control (p < 0.05). However, the apoptotic rate in the 1000 ppmv and 2000 ppmv treatment groups almost decreased to zero on the eighth day after treatment, which was only slightly different from that in the control. When the population grew to approximately the plateau phase, the cells began to undergo apoptosis again. At this time, the apoptosis rate of the 390 ppmv group (i.e., the control group) was higher than 1000 ppmv and 2000 ppmv groups. B Figure 2. (A) Carbonic anhydrase activity of K. mikimotoi exposed to different pCO 2 levels; (B) Rubisco activity of K. mikimotoi exposed to different pCO 2 levels. * indicated the significant difference between the treatment groups with the control group at p < 0.05 level and ** indicated the most significant at p < 0.01 level, the data are shown as the mean ± SE (n = 3).

Alteration of Cellular Apoptosis of K. mikimotoi When Exposed to Different pCO 2 Levels
Elevated pCO 2 induced cellular apoptosis of K. mikimotoi at 24 h after exposure, and the apoptotic rate greatly increased in a concentration-dependent manner from approximately 2.8% in the control to 10.41% and 13.03% in the 1000 ppmv and 2000 ppmv groups, respectively ( Figure 3A). Statistical significance was observed between the treatment groups and the control (p < 0.05). However, the apoptotic rate in the 1000 ppmv and 2000 ppmv treatment groups almost decreased to zero on the eighth day after treatment, which was only slightly different from that in the control. When the population grew to approximately the plateau phase, the cells began to undergo apoptosis again. At this time, the apoptosis rate of the 390 ppmv group (i.e., the control group) was higher than 1000 ppmv and 2000 ppmv groups.
To further support the abovementioned observation, a series of indices closely related to the apoptotic pathway was analyzed. The trend of the change in the percentage of JC-1 monomer was somewhat similar to that of the apoptotic rate ( Figure 3B). The percentage of cells in the JC-1 monomer that indicated mitochondrial membrane potential (MMP) damage was greatly increased, by approximately 5.96% and 9.17% in the 1000 ppmv and 2000 ppmv groups, respectively, after 24 h of exposure to elevated pCO 2 . Similarly, the JC-1 monomer in the 1000 ppmv and 2000 ppmv groups was almost undetectable on the eighth day after treatment, which was not significantly different from the results of the control group. However, when the population grew to approximately the plateau phase, the JC-1 monomer in the experimental group increased significantly, and the percentage of JC-1 monomer at 1000 ppmv and 2000 ppmv was approximately 2.67% and 5.03% lower than the percentage of JC-1 monomer at 390 ppmv, respectively.
In addition, the expression levels of Bcl-2 and Bax were analyzed by ELISA, and the ratio of Bax/Bcl-2 was estimated ( Figure 3C). The results showed that the 1000 ppmv and 2000 ppmv groups, which had ratios approximately 1.33-and 1.42-fold higher than that of the 390 ppmv group in the first two days after exposure, were highly significant. However, as the acidification time increased, seawater acidification caused by 1000 ppmv and 2000 ppmv had little effect on the Bax/Bcl-2 ratio. There was almost no difference in Toxins 2021, 13, 629 6 of 21 the ratio of Bax/Bcl-2 between the treatment groups and the control group, and the ratio of each group was close to 1 after exposure for eight days.   To further support the abovementioned observation, a series of indices closely related to the apoptotic pathway was analyzed. The trend of the change in the percentage of JC-1 monomer was somewhat similar to that of the apoptotic rate ( Figure 3B). The percentage of cells in the JC-1 monomer that indicated mitochondrial membrane potential (MMP) damage was greatly increased, by approximately 5.96% and 9.17% in the 1000 E D Figure 3. Alteration of cellular apoptosis of K. mikimotoi when exposed to different pCO 2 levels. (A) Representative graphs of cell apoptosis obtained from flow cytometry analysis and the percentage of apoptotic cells were shown by histogram. (B) changes of mitochondirol membrane potential (MMP): percentage of JC-1 with monomer and J-aggregate were shown by histogram; (C) changes of Bax/Bcl-2 ratio; (D) changes of Caspase-9 activity; (E) changes of Caspase-3 activity. * indicated the significant difference between the treatment groups with the control group at p < 0.05 level and ** indicated the most significant at p < 0.01 level, and data were expressed as means ± SD (n = 3). Note: the damaging level indicated from the percentage of JC-1 monomer with green fluorescence.
The key components of the apoptosis pathway, i.e., Caspase-9 and Caspase-3, were examined. This examination consistently showed that both CO 2 concentration and acidification time were associated with the Caspase-3 and -9 activities of K. mikimotoi. On the second day, the activity of Caspase-9 in the treatment groups steadily increased as the Toxins 2021, 13, 629 9 of 21 CO 2 concentration increased and reached a peak at 2000 ppmv, which was approximately 1.15-fold higher than that in the control group ( Figure 3D). However, as the acidification time increased, the Caspase-9 activity of the treatment groups significantly decreased, and after eight days of exposure there was almost no difference in Caspase-9 activity between the treatment and the control groups. A similar trend was observed for Caspase-3, which increased by approximately 1.29-fold (p < 0.05) ( Figure 3E) in the 2000 ppmv group compared with that in the control group on the second day. As the acidification time increased, the Caspase-3 activity in the 1000 ppmv and 2000 ppmv treatment groups significantly decreased, which presented little difference with that in the control.
The above results indicated that short-term (i.e., 24 h) elevated pCO 2 induced the apoptosis of K. mikimotoi. However, with prolonged exposure, the effect of elevated pCO 2 on the apoptosis of K. mikimotoi gradually weakened or even disappeared.

Alteration of the Cell Cycle of K. mikimotoi Due to Elevated pCO 2
The alteration of the cell cycle was analyzed by FCM ( Figure 4). The percentage of cells in the G 2 phase decreased from approximately 52.15% to 25.7% while those in S phase significantly increased from 29.08% to 58.67% (p < 0.05) with the increase of pCO 2 in the first two days after exposure (p < 0.05). However, recovery was appeared in the alteration of the cell cycle after e days of exposure. Thereafter, there was no significant difference in cell cycle changes between the final treatment groups and the control group. It seemed that cell cycle arrest in S phase occurred at the initial period of acidifying stress, but prolonging the exposure time would alleviate or even eliminate this negative impact.

Discrimination of H + and Carbon System Alterations in Apoptosis and the Cell Cycle
Based on acute toxicity experiment, a further study was performed to identify whether H + or ions in the carbon system played a decisive role on the change of apoptosis and cell cycle. Cellular apoptosis of K. mikimotoi was induced by seawater acidification after 24 h of exposure, and the apoptotic rate greatly increased from 2.5% in the t=2 t=8 t=15 Figure 4. Cell cycle distribution of K. mikimotoi exposed to different pCO 2 levels. "**": p < 0.01. Note: The cells was divided into three groups by flow cytometry: Phase G 0 /G 1 , S and G 2 /M cells, The DNA content of G 1 /G 0 phase cells: diploid cells (2n); the G 2 /M phase: tetraploid cells (4n); S phase: between diploid and tetraploid (2n-4n).

Discrimination of H + and Carbon System Alterations in Apoptosis and the Cell Cycle
Based on acute toxicity experiment, a further study was performed to identify whether H + or ions in the carbon system played a decisive role on the change of apoptosis and cell cycle. Cellular apoptosis of K. mikimotoi was induced by seawater acidification after 24 h of exposure, and the apoptotic rate greatly increased from 2.5% in the control to 28.65% and 46.6%, respectively, in the HCl addition and CO 2 enrichment groups when the pH was 6.5 (p < 0.05) ( Figure 5A). CO 2 enrichment had a more noxious impact on the apoptosis of K. mikimotoi than that of HCl treatment at the same pH value. The percentage of JC-1 monomer cells, which indicated mitochondrial membrane potential (MMP) damage, was found to greatly increase (by approximately 26.7% and 41.3%) in the treatment groups with a pH of 6.5 compared with that of the control ( Figure  5B). Furthermore, a more obvious increase was found in the CO2 enrichment group, indicating that the impairment was more serious than that in the HCl treatment group.
The expression of Bax in the treatment groups with a pH level of 6.5 was significantly higher than that of Bcl-2, and Bax/Bcl-2 increased greatly, by approximately 1.5and 1.8-fold, respectively, compared with that of the control (p < 0.05) ( Figure 5C). The results showed that short-term (i.e., 24 h) seawater acidification activated the apoptotic pathway and further induced the apoptosis of K. mikimotoi, and the impact induced by CO2 enrichment was more serious than that induced by HCl addition.
The Caspase-3 and -9 activity in the treatment groups (HCl and 18000 ppmv) was significantly higher than that in the control group (390 ppmv). The activity of Caspase-9 increased greatly, by approximately 1.14-and 1.46-fold higher in the treatment groups than that of the control (p < 0.05) ( Figure 5D). A similar trend was observed for Caspase-3, which increased by approximately 1.28-and 1.48-fold (p < 0.05) ( Figure 5E) higher in the treatment groups than that of the control. The results suggested that Caspase-9 and Caspase-3 were activated throughout the experiment by seawater acidification and the CO2 enrichment group was more obvious.
The cell cycle distribution of K. mikimotoi is shown in Figure 6. After 24 h of treatment with seawater acidification, the percentage of cells in the G2 phase significantly decreased, from 59.65% to 45.57% and 38.51% in the HCl addition and CO2 enrichment groups, respectively (p < 0.05). In contrast, the number of cells in the S phase increased significantly (from 7.02% to 43.02% and 45.45% in the HCl addition and CO2 enrichment groups, respectively) compared with that of the control group (p < 0.05) ( Figure 6). This result indicated that the CO2 enrichment and HCl addition partially arrested cells in the S (E) changes of Caspase-3 activity. * indicated the significant difference between the treatment groups with the control group at p < 0.05 level and ** indicated the most significant at p < 0.01 level, and data were expressed as means ± SD (n = 3). Note: the damaging level indicated from the percentage of JC-1 monomer with green fluorescence.
The percentage of JC-1 monomer cells, which indicated mitochondrial membrane potential (MMP) damage, was found to greatly increase (by approximately 26.7% and 41.3%) in the treatment groups with a pH of 6.5 compared with that of the control ( Figure 5B). Furthermore, a more obvious increase was found in the CO 2 enrichment group, indicating that the impairment was more serious than that in the HCl treatment group.
The expression of Bax in the treatment groups with a pH level of 6.5 was significantly higher than that of Bcl-2, and Bax/Bcl-2 increased greatly, by approximately 1.5and 1.8-fold, respectively, compared with that of the control (p < 0.05) ( Figure 5C). The results showed that short-term (i.e., 24 h) seawater acidification activated the apoptotic pathway and further induced the apoptosis of K. mikimotoi, and the impact induced by CO 2 enrichment was more serious than that induced by HCl addition.
The Caspase-3 and -9 activity in the treatment groups (HCl and 18,000 ppmv) was significantly higher than that in the control group (390 ppmv). The activity of Caspase-9 increased greatly, by approximately 1.14-and 1.46-fold higher in the treatment groups than that of the control (p < 0.05) ( Figure 5D). A similar trend was observed for Caspase-3, which increased by approximately 1.28-and 1.48-fold (p < 0.05) ( Figure 5E) higher in the treatment groups than that of the control. The results suggested that Caspase-9 and Caspase-3 were activated throughout the experiment by seawater acidification and the CO 2 enrichment group was more obvious.
The cell cycle distribution of K. mikimotoi is shown in Figure 6. After 24 h of treatment with seawater acidification, the percentage of cells in the G 2 phase significantly decreased, from 59.65% to 45.57% and 38.51% in the HCl addition and CO 2 enrichment groups, respectively (p < 0.05). In contrast, the number of cells in the S phase increased significantly (from 7.02% to 43.02% and 45.45% in the HCl addition and CO 2 enrichment groups, respectively) compared with that of the control group (p < 0.05) ( Figure 6). This result indicated that the CO 2 enrichment and HCl addition partially arrested cells in the S phase, and the effect of the CO 2 enrichment was more obvious than that of the HCl addition. phase, and the effect of the CO2 enrichment was more obvious than that of the HCl addition. The results of the acute toxicity experiments showed that treatment with CO2 enrichment and HCl addition induced apoptosis in K. mikimotoi and partially arrested cells in the S phase, but the effect of CO2 enrichment was more harmful than HCl addition to K. mikimotoi in seawater of the same pH. This result was presumed to indicate that H + was not the only factor that resulted in the impact of seawater acidification and that carbon system alteration may play an essential role in the impacts of seawater acidification.

What's the Possible Explanation for The Different Growth Modes of K. mikimotoi Exposed to Short or Long Period of Seawater Acidification
Elevated CO2 levels have been confirmed to stimulate the growth of many phytoplankton species, including the diatom Pseudo nitzschia sp [41], the raphidophyte Heterosigma akashiwo [42,43], and Alexandrium fundyense [44]. The dinoflagellate Karlodinium veneficum [45] also displayed abundant growth rates under increased pCO2 concentrations. We observed a slightly different phenomenon in K. mikimotoi with different pCO2 exposures. The specific growth rate was inhibited during the first four days after exposure compared with that of the control, while prolonging the duration time alleviated the inhibition. In other words, short-term seawater acidification inhibited the growth of K. mikimotoi in a clear concentration-dependent manner, while adaptation occurred with increasing time, and growth stimulation occurred within a certain time range in the exponential phase. It has been shown that short-term acidification has a negative effect on algae. This might be because sharply elevated pCO2 in seawater, together with other chemical changes, altered the periplasmic redox activity or the permeability of cellular membranes [46] and perturbed cell membrane ion channels, therefore acting as a stressor and affecting red tide algae [47]. Gazeau et al. [10] also showed that reducing carbonate ion concentrations through acidification reduced the iron uptake of phytoplankton and Figure 6. Cell cycle distribution of K. mikimotoi exposed to seawater acidification caused by CO 2 enrichment and HCl addition. "**": p < 0.01. Note: The cells was divided into three groups by flow cytometry: Phase G 0 /G 1 , S and G 2 /M cells, The DNA content of G 1 /G 0 phase cells: diploid cells (2n); the G 2 /M phase: tetraploid cells (4n); S phase: between diploid and tetraploid (2n-4n).
The results of the acute toxicity experiments showed that treatment with CO 2 enrichment and HCl addition induced apoptosis in K. mikimotoi and partially arrested cells in the S phase, but the effect of CO 2 enrichment was more harmful than HCl addition to K. mikimotoi in seawater of the same pH. This result was presumed to indicate that H + was not the only factor that resulted in the impact of seawater acidification and that carbon system alteration may play an essential role in the impacts of seawater acidification.

What's the Possible Explanation for the Different Growth Modes of K. mikimotoi Exposed to Short or Long Period of Seawater Acidification
Elevated CO 2 levels have been confirmed to stimulate the growth of many phytoplankton species, including the diatom Pseudo nitzschia sp [41], the raphidophyte Heterosigma akashiwo [42,43], and Alexandrium fundyense [44]. The dinoflagellate Karlodinium veneficum [45] also displayed abundant growth rates under increased pCO 2 concentrations. We observed a slightly different phenomenon in K. mikimotoi with different pCO 2 exposures. The specific growth rate was inhibited during the first four days after exposure compared with that of the control, while prolonging the duration time alleviated the inhibition. In other words, short-term seawater acidification inhibited the growth of K. mikimotoi in a clear concentration-dependent manner, while adaptation occurred with increasing time, and growth stimulation occurred within a certain time range in the exponential phase. It has been shown that short-term acidification has a negative effect on algae. This might be because sharply elevated pCO 2 in seawater, together with other chemical changes, altered the periplasmic redox activity or the permeability of cellular membranes [46] and perturbed cell membrane ion channels, therefore acting as a stressor and affecting red tide algae [47]. Gazeau et al. [10] also showed that reducing carbonate ion concentrations through acidification reduced the iron uptake of phytoplankton and hypothesized that this resulted from a decreased ability of phytotransferrin, an iron binding protein, to bind with complex inorganic iron at the cell surface. However, growth differentiation was found with increased acidification time until the cells reached the exponential phase. The specific growth rate in the treatment groups increased obviously compared with that in the control, indicating the restoration of growth inhibition. The simulation results of the logistic equation provided further evidence of the observation. The carrying capacity (K) was almost equal to the maximum value of the cell density in the exponential phase. We also found that the peak appeared in the 2000 ppmv group. In fact, similar results have been reported. For instance, Alexandrium cell densities were significantly and consistently enhanced when natural populations were incubated at 150 Pa pCO 2 compared with those in populations incubated at 39 Pa. During natural Alexandrium blooms in Northport Bay, pCO 2 concentrations increased over the course of a bloom to more than 170 Pa and were highest in regions with the greatest Alexandrium abundances, suggesting Alexandrium may further exacerbate acidification and/or be especially adapted to these acidified conditions [44]. Cell proliferation was achieved through the cell cycle. In this study, we found that short-term acidification arrested microalgal cells in the S phase and prolonging the duration of exposure to acidification alleviated cellular arrest. This means that short-term acidification inhibited cell division, which was consistent with the results of many previous studies [48], while an adaptation occurred with increasing time. This was one possible explanation for the observed growth performance at the population level.
In addition, dinoflagellates have been suggested to be at a disadvantage with regard to photosynthetic carbon fixation due to the presence of low affinity forms of CO 2fixing enzymes, including ribulose-1, 5-bis-phosphate carboxylase-oxygenase (type II Rubisco) [49][50][51], under the present ocean conditions of low CO 2 and high O 2 . The carbon fixation ability of dinoflagellates is said to be well below saturation at present CO 2 levels [50,52], and thus, extremely sensitive to changes in CO 2 concentrations. We found that Rubisco activity and the carbonic anhydrase (CA) activity of K. mikimotoi in the treatments were significantly decreased during the first 24 h, indicating that the reduction in photosynthesis and CCMs and the changes in seawater chemistry caused by pCO 2 might be responsible for the result. However, a difference was observed on the eighth day, when Rubisco activity increased while CA decreased significantly (p < 0.05) in both treatments compared with that of the control, and adaptation was assumed to occur. Chen et al. [47] reported that short-term exposure to reduced pHnbs (7.70) decreased the photosynthesis and light use efficiency in algae. However, acclimation to a reduced pH level for 1~19 generations led to recovered photosynthetic activity. The results of this study were consistent with those of the abovementioned reports. Mackey et al. [53] reported that an increase in the CO 2 concentration could potentially decrease the need for CCMs (e.g., by acquiring CO 2 via diffusive uptake rather than active pumping of HCO 3 − or reducing the requirement to express CAs), and this downregulation could potentially allow for energy and resources to be allocated to other physiological processes, such as the assimilation of other nutrients, leading to stimulated growth and photosynthesis [16,54,55]. In fact, these characteristics of adaptation to seawater acidification are one of the causative reasons for dinoflagellate bloom formation. Some documents have suggested that HABs easily develop in acidified environments. Vicicitus globosus, for instance, had a selective advantage under ocean acidification, increasing its abundance in natural plankton communities at CO 2 levels higher than 600 µatm and developing blooms above 800 µatm CO 2 [25]. HABs have increased over the past half-century [39,56], and the levels of atmospheric and surface water CO 2 concentrations have concurrently increased by more than 25% [57]. Changing levels of dissolved inorganic carbon in surface waters may impact phytoplankton inorganic carbon fixation [16,58]. Thus, rising CO 2 concentrations in surface waters may potentially contribute to the global expansion of HABs (i.e., the fertilization effect) [59,60]. On the other hand, the increase of CO 2 or decrease of pH have been proven to alleviate this carbon limitation and reduce the need for energetically taxing CCMs, and the resultant shift in competitive balance may alter the phytoplankton community composition [50,61]. In other words, some dinoflagellates might benefit from decreasing pH because of their wide diversity of CCM efficiencies [62]. Previous studies have also found carbon fixation rates generally mirrored the trends of toxicity. One plausible explanation is that enhanced toxin synthesis could be due to an overall enhanced metabolic activity, and elevated pCO 2 could be the catalyst for increased amounts of energy and carbon needed for phytotoxin production [63]. Likewise, Karlodinium veneficum, an ichthyotoxic dinoflagellate, has been shown to increase growth rates and produce more carbon-based karlotoxin under elevated levels of pCO 2 [45]. In a similar study, increased growth of A. tamarense was seen, as the photosynthetic properties also showed a positive growth response when exposed to higher pCO 2 . Increased pCO 2 also induced an increase in cellular toxin production in the dinoflagellate A. tamarense [64]. It can be inferred that K. mikimotoi was not only significantly increased in growth and photosynthetic carbon fixation, but also increased its toxicity when exposed to high pCO 2 , which may further enhance the allelopathy of K. mikimotoi to other species, resulting in the competitive potential of toxic HAB species as ocean acidification intensifies. K. mikimotoi is one of the most causative bloom-forming species in coastal China and is usually a dominant species. The adaptive mechanism of photosynthetic carbon fixation is thus assumed to explain this phenomenon.
We observed microalgal apoptosis within 24 h. The elevated pCO 2 increased the Bax/Bcl-2 ratio and magnified apoptosis signaling by transferring extracellular apoptotic signaling to the apoptosis pathway. The simultaneous impairment of the mitochondrial membrane activated Caspase-9 and its downstream effector Caspase-3 trigger apoptosis. An interaction was observed between cellular apoptosis and the change in cell density during the first 24 h, and apoptosis was thought to be the adaptative response to the elevated pCO 2 . A hypothesis called the "altruistic adaptation theory" was proposed, the main content of which was that damaged cells could be removed from a population of single-celled organisms by programmed cell death, reducing the burden on the surviving cells to the benefit of the whole group [65]. The inflection point from negative to positive effects of elevated CO 2 was affected by the degree of acclimation to acidification. Again, on the eighth day after exposure, apoptosis in the treatment groups was no longer observed. The results further demonstrated the speculation that K. mikimotoi had undergone adaptive changes under long-term exposure to CO 2 -induced seawater acidification. Interestingly, the rates of apoptosis in the 2000 ppmv and 1000 ppmv groups were significantly reduced on the 15th day after treatment, and the abovementioned indicators showed a downward trend compared with those of the control group. Some studies have suggested that high CO 2 treatments allow cells to potentially downregulate their carbon concentrating mechanisms (CCMs), which saves up to 20% of CCM-related energy expenditures [66,67]. This would mean that high CO 2 would raise the energy use efficiency and excessive energy dissipation strategy of K. mikimotoi. We might assume that K. mikimotoi cells reallocate intercellular energy to adapt to the new environment and that more energy is used to extend cell life. Previous studies have proven that, while K. mikimotoi has a wide range of suitable pH values for growth, values of 7.2~9.2 can maintain faster growth. This was consistent with our study, which found that K. mikimotoi has the ability to acclimate to the expected rise in atmospheric CO 2 , up to 2000 ppmv. The red tide alga K. mikimotoi can increase its tolerance to decreased pH values and its competitiveness in phytoplankton communities after it has acclimated to CO 2 -induced seawater acidification. This is the main reason for the frequent bloom of harmful red tides, such as those caused by K. mikimotoi, in environments in which seawater acidification is intensified, and the strong adaptability of harmful algae is pivotal.

What's the Possible Explanation
Leading to the Different Physiological Responses of K. mikimotoi When Exposed to HCl or CO 2 -Induced Acidifying Conditions?
There were ions in the acidified seawater that seemed even more noxious than H + to K. mikimotoi. The reduced pH level resulting from the different treatments (HCl addition and high CO 2 concentration enrichment) led to large differences in the microalgal apoptotic rates and cell cycle, and the impacts induced by the CO 2 treatment were more severe than those induced by the HCl treatment in seawater of the same pH. We speculated that the effects of seawater acidification caused by a high CO 2 concentration on K. mikimotoi were not dominated by an increase in H + concentration and that the effects of other factors related to CO 2 were even more pronounced. Previous studies have suggested that the toxicity of CO 2 might be largely underestimated compared with that of H + . Elevated seawater CO 2 concentrations can affect the functioning of molecules, cells, tissues, and whole organisms through factors such as CO 2 itself as well as CO 2 -related changes in pH and bicarbonate levels [68,69]. Moreover, Xu et al. [70] reported that acidification interferes with the processes of ingestion and digestion in Mytilus edulis. However, the negative impacts induced in the CO 2 group were more severe than those induced in the HCl group. This was consistent with the results of our study. Altogether, these results provide further evidence that the impact of seawater acidification on K. mikimotoi is based on factors related to CO 2 , not just H + .

Conclusions
Short-term seawater acidification inhibited the growth of K. mikimotoi, and the combination of inhibition of photosynthetic carbon fixation, cell cycle arrest, and apoptosis were the possible causes. However, prolonged exposure would induce the adaptive responses, presented by the recovery or even stimulation of the algal growth. Ions dissociated by CO 2 dissolution other than H + itself were more important to the observation. Results might elucidate the dominant role of K. mikimotoi during the occurrence and throughout the duration of blooms in the East China Sea.

Microalgal Cultivation
K. mikimotoi was obtained from the Algal Culture of Fisheries College Collection at the Ocean University of China. A batch cultivation was performed in the present study, and cells were cultured in f/2 medium [71] at 20 ± 1 • C with a photon flux density of 75 µmol photon m −2 s −1 . The light-dark cycle was set at 12:12 h. Cells in the exponential phase were used for the experiments.

Acidifying System Set-Up and Experimental Design
Two level of acidification were applied, namely 1000 ppmv (pH 7.8; the prediction of atmospheric CO 2 level in 2100) and 2000 ppmv (pH 7.5; the prediction of atmospheric CO 2 level in 2300) based on our previous studies [38,72]. pCO 2 of 390 ppmv (pH 8.1; ambient seawater pH) was used as the control. The pCO 2 alteration was achieved in the experimental system with gentle bubbling of 0.22 µm-filtered ambient air and air/CO 2 mixtures, which were generated by CO 2 light incubator (HP400G-D, Ruihua Instrument & Equipment Ltd., Wuhan, China), with a variation of less than 5%. The pH values used during the experiment were determined using a pH meter (Seven Compact™ S210k, Mettler Toledo, Switzerland), which was calibrated daily with a standard buffer system from the National Bureau of Standards (NBS). The salinity was measured daily using a handheld salinometer (WY028Y, Huarui, China). All other carbonate system variations were calculated using CO 2 SYS software according to the method of Lewis et al. [73].
The sterile seawater was divided into 3 groups, and 3 different CO 2 treatments (390, 1000 and 2000 ppmv) were performed 24 h before microalgal inoculation to allow the carbonate system to stabilize. Then, cells in the exponential phase were inoculated into 500 mL flasks with 300 mL acidified seawater enriched with f/2 medium, and the initial cell density was adjusted to 1 × 10 4 cells mL −1 . To maintain consistent CO 2 in the cultures during the experiments, the cultures were continuously aerated by air and two different CO 2 concentrations (1000 and 2000 ppmv) controlled by the incubator through the experiment. The cultivation was conducted for 15 days thereafter, and measurements were conducted on 2, 8, and 15 days after exposure.

Population Dynamics
Breifly, samples of 1 mL of the microalgal solution from each treatment were fixed every two days with acidic Lugol's and the cell density was determined according to the method of Guillard [74].
The specific growth rate (µ) was calculated using the following equation: where N j and N i represent the average cell numbers at times t j and t i , respectively.
A logistic equation was applied to quantitatively characterize the growth performance in the different pCO 2 treatments.

Determination of Photosynthetic Carbon Fixation
Rubisco and carbonic anhydrase (CA) are two key enzymes that determine the rate of carbon assimilation in the dark reaction of photosynthesis. Their enzymatic activities were spectrophotometrically determined using ELISA kits (Shanghai Jining Shiye, Shanghai, China) according to the manufacturer's protocol, and the results were expressed as U mg −1 protein. The soluble protein content was measured using the Bradford method [75].

Estimation of the Key Index in the Apoptotic Pathway
(1) Cellular apoptotic rate (Annexin V-FITC/PI) The cell apoptotic rate was detected using the Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) double staining method by flow cytometry (FCM). Cells in each treatment were collected on the same day and washed twice with cold phosphatebuffered saline (PBS, pH 7.2-7.4). The staining was performed using the Annexin V-FITC double staining kit (BD Pharmingen, San Diego, CA, USA) at 26 • C in the dark for 15 min, and the PI was added thereafter. The abovementioned prepared solutions were analyzed with the FL-1 channel for FITC fluorescence and the FL-3 channel for PI fluorescence using an FC 500 MPL flow cytometer (Beckman Coulter, CA, USA). The apoptotic rate was indicated by the percentage of early apoptotic cells that were stained by Annexin V-FITC.
(2) Mitochondrial membrane potential (MMP) changes The mitochondrial membrane potential (MMP) was detected by FCM, which indicated the percentage of monomers and J-aggregates (JC-1, Beyotime Bio. Inc., Nantong, China). JC-1 probes, at a concentration of 6 µM, were added to the culture medium and incubated for 20 min at 26 • C, as described by the manufacturer. MMP changes were estimated using FCM equipped with FL-1 and FL-3 channels for monomers and J-aggregates (Beckman Coulter, Brea, CA, USA), and the MMP was indicated by the percentage of JC-1 with monomers and J-aggregates.
(3) Changes in the Bax/Bcl-2 ratio Bax is a pro-apoptotic protein and Bcl-2 is an anti-apoptotic protein. The elevation of its ratio indicated the increase of cell apoptosis. The content of Bax and Bcl-2 was measured in the microplate using ELISA kits (Shanghai Jining Shiye, Shanghai, China) according to the manufacturer's instructions. The absorbance (OD) of Bax and Bcl-2 was measured at 450 nm using a Perkin-Elmer microplate reader, and the content of Bax and Bcl-2 in the samples was calculated based on the standard curve.
(4) Activities of Caspase-9 and Caspase-3 The activity of Caspase-9 and Caspase-3 was measured using colorimetric activity assay kits (Beyotime Bio. Inc., Nantong, China) according to the manufacturer's instructions and the descriptions found in our previous studies [48,76], and the activity level was recorded as U mg −1 protein.

Determination of the Cell Cycle
Propidium iodide (PI), a stain that can be embedded in double-stranded DNA, produces red fluorescence, and the fluorescence intensity is proportional to the DNA content. Microalgal cells used for the analysis of the cell cycle were prepared according to the method of propidium staining (Beyotime Bio. Inc., Nantong, China). The treated cells were incubated for 30 min at room temperature and thereafter were analyzed using the FL-3 channel for PI fluorescence. The DNA content was determined simultaneously by FCM. The percentage of cells in the different phases of the cell cycle was determined using MultiCycle software (Beckman Coulter).

Discrimination of H + and Ions in Carbon System Alteration on Apoptosis and Cell Cycle
Based on Acute Toxicity Experiment Geological CO 2 sequestration was a practical tool in reducing emissions [77], however, it is accepted that sub-seabed storage leaks are possible over time [78]. These leaks could cause severe reductions in pH (even achieved pH 5.6). Therefore, pH 6.5 was used as an acute toxicity test to identify the impacts of H + and ions in the carbon system on the observed results by using HCl addition and CO 2 enrichment. A pH value at extreme acidifying condition of pH 6.5 (about 18,000 ppmv) was simulated by 1 M HCl or CO 2 mixed gas. Cellular growth and apoptosis were analyzed on the second day after exposure according to the description in 2.6 and 2.7 without other description.

Statistical Analysis
Analysis of variance was carried out using SPSS 16.0. All experiments were performed at least three times. The results are expressed as the mean ± standard deviation (SD), and one-way analysis of variance was used for comparisons between groups. Values of p < 0.05 were considered statistically significant.