Association of Phosphatidylinositol-Specific Phospholipase C with Calcium-Induced Biomineralization in the Coccolithophore Emiliania huxleyi

Biomineralization by calcifying microalgae is a precisely controlled intracellular calcification process that produces delicate calcite scales (or coccoliths) in the coccolithophore Emiliania huxleyi (Haptophycea). Despite its importance in biogeochemical cycles and the marine environment globally, the underlying molecular mechanism of intracellular coccolith formation, which requires calcium, bicarbonate, and coccolith-polysaccharides, remains unclear. In E. huxleyi CCMP 371, we demonstrated that reducing the calcium concentration from 10 (ambient seawater) to 0.1 mM strongly restricted coccolith production, which was then recovered by adding 10 mM calcium, irrespective of inorganic phosphate conditions, indicating that coccolith production could be finely controlled by the calcium supply. Using this strain, we investigated the expression of differentially expressed genes (DEGs) to observe the cellular events induced by changes in calcium concentrations. Intriguingly, DEG analysis revealed that the phosphatidylinositol-specific phospholipase C (PI-PLC) gene was upregulated and coccolith production by cells was blocked by the PI-PLC inhibitor U73122 under conditions closely associated with calcium-induced calcification. These findings imply that PI-PLC plays an important role in the biomineralization process of the coccolithophore E. huxleyi.


Introduction
Coccolithophores play an essential role in the global carbon cycle through considerable carbon dioxide fixation and carbon sequestration as they are major producers of marine biogenic calcites and produce large blooms in the open ocean. Emiliania huxleyi (E. huxleyi), a calcifying haptophyte, is widespread throughout the oceans worldwide [1,2], including polar [3] and coastal regions [4]. Due to their ability to form dense blooms at mid and high latitudes, they are intensively studied to understand the environmental implications of calcification [5,6]. Furthermore, they produce ornate calcium carbonate scales, known as coccoliths, as cell covering. The calcite plate originates from a separate intracellular compartment known as the coccolith vesicle (CV) [7][8][9][10]. Moreover, since the complex shape of calcite crystals, which is similar to coccoliths, cannot be produced artificially at a nanometer scale, coccoliths are an attractive biogenic product for bioengineers and material scientists and are desirable for various applications in nanotechnology [11]. Despite the pivotal roles of this calcifying alga, the underlying molecular mechanism remains unknown. Understanding the critical

Algal Strain and Culture Conditions
E. huxleyi strain CCMP 371 was purchased from the Provasoli-Guillard National Center for Marine Algae and Microbiota (East Boothbay, ME, USA); this strain has unique characteristics for calcification as described in the introduction. The cells were cultured in sterile artificial seawater [27] enriched with nitrate, phosphate, trace metals, and vitamins at f/2 concentrations [28] as well as 0.01 µM selenium [29]. The cells were maintained at 20 • C under constant irradiation at 100 µmol m −2 s −1 and agitated twice daily by manual shaking to homogenize the culture. To inhibit PI-PLC, the cells were treated with U73122 (Sigma-Aldrich Co., St. Louis, MO, USA) at final concentrations of 0.01, 0.1, and 0.2 µM. U73122 is a PI-PLC and A 2 inhibitor that inhibits the hydrolysis of phosphatidylinositol to inositol triphosphate, leading to a decrease in free cytosolic Ca 2+ ; it also acts to inhibit the coupling of G-protein phospholipase C activation while being unaffected by cAMP production (U6756, Product Information Sheet, Sigma). As a negative control, U73343 (Sigma-Aldrich Co., St. Louis, MO, USA), a less active structural analog of U73122, has been used for the nonspecific effects of U73122 on PI-PLC. As U73122 and 73343 were dissolved in dimethyl sulfoxide (DMSO), control cells were treated with DMSO only. Twenty-four hours after inoculating the cells, U73122, U73343 and DMSO were applied, and [Ca 2+ ] was concurrently shifted from 0.1 to 10 mM. Cells were harvested after 72 h of treatment for further experiments using flow cytometry in triplicate.

Inorganic Phosphate Measurement
The inorganic phosphate (Pi) in the medium was analyzed by the molybdenum blue method [30]. For the analysis, 20 µL of acid ascorbate solution (2.5% ammonium molybdate, 0.1% potassium antimonyl tartrate sesquihydrate, and 3.15 M H 2 SO 4 ), and 20 µL of acid molybdate solution (10% ascorbic acid and 2.25 M H 2 SO 4 ) was added to 1 mL of the supernatant of the cell suspension.

AP Activity Measurement
According to the method described by Reichardt et al. [31], alkaline phosphatase (AP) activity was estimated using p-nitrophenyl phosphate (p-NPP) as a substrate to measure its degradation to p-nitrophenol. First, p-NPP (100 µL, 36 mM) and 700 µL of 3-(cyclohexylamino)-1-propane sulfonic acid buffer (200 mM, pH 10.0) was added to 200 µL of the cell suspension with intact cells, followed by incubation at 40 • C for 15 min. The reaction was terminated by adding 100 µL of 4 M NaOH, and the absorbance at 397 nm was measured using a UV-VIS spectrophotometer (UV mini 1240; Shimadzu, Kyoto, Japan).

Determination of TIC and TOC
According to Pelusi et al. [32], 50 mL of each cell culture sample was harvested during the exponential growth phase. Half of the culture sample (25 mL) was centrifuged (1720× g, 5 min) to remove cells. Then, the supernatant was filtered using a Millex HA filter (pore size 0.45 µm, Merck Millipore, Darmstadt, Germany) to acquire culture medium without any particles, including cell debris (culture medium fraction). Total organic carbon (TOC) and total inorganic carbon (TIC) values of both the cell suspension and culture medium fraction (25 mL each) were measured using a TOC analyzer (Shimadzu Corporation, Kyoto, Japan). TIC was calculated by subtracting TOC from the total carbon. The cellular organic and inorganic carbon quantities (Cell C and Cell IC, respectively) were calculated using the following equations: Cell IC = (TIC in cell suspension) − (TIC in culture medium fraction), where the TOC and TIC values are expressed as cells contained in a 1-L culture (mg/L).

Cell Growth and Microscopic Observation
The cell number was determined by counting the cells using a hemocytometer (Marienfeld, Bad Mergentheim, Germany) or a cell counter (Sysmex CDA-1000, Kobe, Japan). Cells surrounded by coccolith were observed under polarized light, and micrographs were taken using a microscope (BX53; Olympus Ltd., Tokyo, Japan) equipped with a fluorescent microscope digital camera (Keyence, Osaka, Japan).

Sample Preparation
Triplicate samples from each time point (−3, 0, 1, and 3 h before and after loading to 10 mM [Ca 2+ ]) were harvested in a conical tube by centrifugation (1720× g, 5 min, 4 • C) and immediately frozen in liquid nitrogen. The total RNA was isolated for RNA-seq using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), the RNeasy Mini Kit (Qiagen, Hilden, Germany), and DNase (Qiagen, Hilden, Germany) treatment according to the manufacturer's instructions. RNA quality and integrity were estimated using an Agilent Technologies 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).

RNA-seq and Transcriptome Analysis
Library construction and RNA-seq using an Illumina HiSeq 3000 (Illumina, Inc., San Diego, CA, USA) were performed by Macrogen (Seoul, Korea). The sequencing libraries were prepared according to the manufacturer's instructions using the Illumina TruSeq 3000 SBS Kit v3 and sequenced in pair-end reads (Table S1). Read quality was assessed using FastQC v0.11.5 [33]. The raw reads were trimmed with Trimmomatic v0.36 [34]. Trimmed reads for each sample were aligned using Bowtie v1.1.2 (http://bowtie-bio.sourceforge.net/index.shtml) with the previously assembled reference transcripts (DDBJ/EMBL/GenBank accession GHJP00000000) [24]. Sequenced raw reads are deposited in the Sequence Read Archive of the National Center for Biotechnology Information under Bioproject number PRJNA623073.
The RSEM algorithm (v1.2.29) was used to count the aligned reads, and contigs with 0 counts in all 12 samples were removed from the analysis [35]. Thus, we conducted statistical analysis of 16,061 contigs after excluding 32,589 out of a total of 48,650 contigs. Unigenes were distinguished as DEGs if they showed a fold change (| log 2 FC |) > 1 between the 0.1 mM calcium group (−3 and 0 h (time of the shift), Group 1) and 10 mM calcium group (1 and 3 h, Group 2) using a cut-off probability score to ensure a false discovery rate (FDR) less than 0.05. The domain enrichment (pfam and Interpro) and protein-protein interaction networks of DEGs were analyzed using STRING (v11.0) [36]. The interaction score parameters were set at the medium confidence level for minimum requirements.

qPCR
Total RNA was extracted from cell cultures using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and the RNeasy Mini Kit (Qiagen, Hilden, Germany); cDNA was synthesized using a Superscript III Kit (Invitrogen, Carlsbad, CA, USA) primed with oligo dT primers, following the manufacturer's protocol. The cDNA was used as a template for qPCR. SYBR green was applied for amplicon detection. SYBR premix (Takara, Tokyo, Japan) and the Thermal Cycler Dice Real-Time System TP8200 (Takara, Tokyo, Japan) were used for cDNA amplification. Actin was used as the reference gene for qPCR [22]. Target sequences of primers are listed in Table S2.

Calcein Staining and Flow Cytometry Analysis
For each sample, 7 mL of cells were used for calcium carbonate staining by calcein (Sigma-Aldrich Co., St. Louis, MO, USA). Cells were calcein labelled for 2 h according to Fox et al. [37] and washed three times by gentle centrifugation (840× g, 1 min, 20 • C). Then, cells were resuspended in fresh f/2-si media before flow cytometry analysis. The BD FACS Canto II flow cytometer (BD Biosciences, San Jose, CA, USA) was used to analyze the calcein-stained E. huxleyi samples. A 488-nm excitation laser with an FITC emission filter was used to detect calcein fluorescence and with a PerCP-Cy5.5 filter to detect chlorophyll autofluorescence. The populations of cells and coccoliths were distinguished by FSC and side scatter (SSC). In total, of 10,000 or 20,000 cells were gated, and chlorophyll-positive populations were counted for each sample. Data were analyzed using FlowJo software version 10.0.7 (Tree Star, Ashland, OR, USA). Statistically significant differences were determined by Student's t-test (* p < 0.05, ** p < 0.01).  Figure 1a). In the other vessel, the cell suspension was diluted with fresh medium on Days 3 and 6 to avoid Pi-depleted conditions (medium supply (MS) in Figure 1a). Both cultures remained in the exponential growth phase during the experiment (Figure 1b), even under different Pi conditions (Figure 1c). The change in extracellular alkaline phosphatase (AP) activity, induced by Pi limitation in the cells, revealed that Pi limitation started on Day 5 in the control but was not observed in MS conditions (Figure 1c). AP activity expressed on the cell surface was determined as a parameter of Pi limitation since the activity is known to be induced quantitatively depending on the extent of Pi limitation in E. huxleyi [13]. Cells grown in 0.1 mM [Ca 2+ ] and harvested on Day 5 (0 h in Figure 1d) had no coccolith on the cell surface in both control and MS conditions. Coccolith production started already at 24 (Day 6) and 48 h (Day 7) after the shift to ambient [Ca 2+ ] (10 mM), and the amount of coccoliths covering the cells became more obvious in MS than in the control (Figure 1d,e). The results indicate that coccolith production in E. huxleyi CCMP 371 is mainly regulated by [Ca 2+ ] conditions, but not Pi concentration. PerCP-Cy5.5 filter to detect chlorophyll autofluorescence. The populations of cells and coccoliths were distinguished by FSC and side scatter (SSC). In total, of 10,000 or 20,000 cells were gated, and chlorophyll-positive populations were counted for each sample. Data were analyzed using FlowJo software version 10.0.7 (Tree Star, Ashland, OR, USA). Statistically significant differences were determined by Student's t-test (* p < 0.05, ** p < 0.01).

Coccolith Formation in E. huxleyi CCMP 371 under Phosphate-Replete Conditions
Firstly, cells cultivated in ambient seawater culture medium containing 10 mM calcium chloride and 36 μM inorganic phosphate (Pi) as the stock culture were transferred to culture medium containing 0.  Figure 1a). In the other vessel, the cell suspension was diluted with fresh medium on Days 3 and 6 to avoid Pi-depleted conditions (medium supply (MS) in Figure 1a). Both cultures remained in the exponential growth phase during the experiment (Figure 1b), even under different Pi conditions ( Figure 1c). The change in extracellular alkaline phosphatase (AP) activity, induced by Pi limitation in the cells, revealed that Pi limitation started on Day 5 in the control but was not observed in MS conditions ( Figure 1c). AP activity expressed on the cell surface was determined as a parameter of Pi limitation since the activity is known to be induced quantitatively depending on the extent of Pi limitation in E. huxleyi [13]. Cells grown in 0.1 mM [Ca 2+ ] and harvested on Day 5 (0 h in Figure 1d) had no coccolith on the cell surface in both control and MS conditions. Coccolith production started already at 24 (Day 6) and 48 h (Day 7) after the shift to ambient [Ca 2+ ] (10 mM), and the amount of coccoliths covering the cells became more obvious in MS than in the control (Figure 1d,e). The results indicate that coccolith production in E. huxleyi CCMP 371 is mainly regulated by [Ca 2+ ] conditions, but not Pi concentration.

Morphological and Quantitative Monitoring of Calcium-Induced Coccolith Production
According to the observations under light and polarized microscopes, cells pre-cultivated at 0.1 mM [Ca 2+ ] for 5 days (0 h) were confirmed to have no coccolith (Figure 2a). Subsequently, after shifting [Ca 2+ ] from 0.1 to 10 mM in MS conditions with sufficient [Pi], coccolith production began to increase the amount of calcite scales surrounding the cells (coccoliths). The increase in the TIC concentration in the cells was directly proportional to the coccolith formation observed by light and polarized microscopes ( Figure 2). These data show that coccolith formation was induced by the increase in calcium concentration. Meanwhile, the quantity of coccoliths produced as CaCO 3 (which was estimated as TIC by a TOC analyzer that quantified both TIC and TOC as described in 'Materials and Methods') increased during calcification for 24 h and then decreased slightly toward 48 h in the cell fraction ( Figure 2b) and cells (Figure 2c). Both TIC (TIC/mL) and TOC (TOC/mL) in the cell fraction increased linearly as observed with coccoliths in the whole cells, whereas cellular TOC was maintained at nearly constant levels owing to the increase in cell number during the experiment ( Figure S1). The decrease in TIC in the medium was quantitatively accompanied by an increase in TIC in the whole cells ( Figure S1c), indicating that most carbon absorbed from the medium was transferred into the coccoliths.

Morphological and Quantitative Monitoring of Calcium-Induced Coccolith Production
According to the observations under light and polarized microscopes, cells pre-cultivated at 0.1 mM [Ca 2+ ] for 5 days (0 h) were confirmed to have no coccolith (Figure 2a). Subsequently, after shifting [Ca 2+ ] from 0.1 to 10 mM in MS conditions with sufficient [Pi], coccolith production began to increase the amount of calcite scales surrounding the cells (coccoliths). The increase in the TIC concentration in the cells was directly proportional to the coccolith formation observed by light and polarized microscopes ( Figure 2). These data show that coccolith formation was induced by the increase in calcium concentration. Meanwhile, the quantity of coccoliths produced as CaCO3 (which was estimated as TIC by a TOC analyzer that quantified both TIC and TOC as described in 'Materials and Methods') increased during calcification for 24 h and then decreased slightly toward 48 h in the cell fraction ( Figure 2b) and cells (Figure 2c). Both TIC (TIC/mL) and TOC (TOC/mL) in the cell fraction increased linearly as observed with coccoliths in the whole cells, whereas cellular TOC was maintained at nearly constant levels owing to the increase in cell number during the experiment ( Figure S1). The decrease in TIC in the medium was quantitatively accompanied by an increase in TIC in the whole cells ( Figure S1c), indicating that most carbon absorbed from the medium was transferred into the coccoliths.

Calcium-Induced Biomineralization-Associated Genes
The calcium-induced calcifying cells were already observed at 3 h after shifting [Ca 2+ ] from 0.1 to 10 mM ( Figure 2). Therefore, the cells at 1 and 3 h after the shift were used to analyze genes influenced by calcium at the transcriptional level ( Figure 3a); additionally, samples harvested at 0 and 3 h before the shift (−3 and 0 h, respectively) were examined. These samples were used for RNA sequencing (RNA-seq) (Figure 3a) by generating an average of 68 million reads per sample (Table S1). Mapping of these sequenced reads against the previously assembled transcripts [24] clearly showed that samples were separated by shift time into groups of DEGs: before (Group 1: non-calcified cells harvested at −3 and 0 h) and after (Group 2: calcified cells harvested at 1 and 3 h) shifting [Ca 2+ ] from 0.1 to 10 mM (Figure 3b). Expression levels verified by quantitative polymerase chain reaction (qPCR) with randomly selected DEGs confirmed the differential gene expression between the two groups ( Figure 3c, Table S2). Comparison of DEGs between groups revealed that the number of upregulated and downregulated DEGs was 116 and 47, respectively, in cells calcified by [Ca 2+ ] enrichment (Group 2) (Figure 3d, Table S3).

Calcium-Induced Biomineralization-Associated Genes
The calcium-induced calcifying cells were already observed at 3 h after shifting [Ca 2+ ] from 0.1 to 10 mM ( Figure 2). Therefore, the cells at 1 and 3 h after the shift were used to analyze genes influenced by calcium at the transcriptional level ( Figure 3a); additionally, samples harvested at 0 and 3 h before the shift (−3 and 0 h, respectively) were examined. These samples were used for RNA sequencing (RNA-seq) (Figure 3a) by generating an average of 68 million reads per sample (Table  S1). Mapping of these sequenced reads against the previously assembled transcripts [24] clearly showed that samples were separated by shift time into groups of DEGs: before (Group 1: noncalcified cells harvested at −3 and 0 h) and after (Group 2: calcified cells harvested at 1 and 3 h) shifting [Ca 2+ ] from 0.1 to 10 mM (Figure 3b). Expression levels verified by quantitative polymerase chain reaction (qPCR) with randomly selected DEGs confirmed the differential gene expression between the two groups ( Figure 3c, Table S2). Comparison of DEGs between groups revealed that the number of upregulated and downregulated DEGs was 116 and 47, respectively, in cells calcified by [Ca 2+ ] enrichment (Group 2) (Figure 3d, Table S3).  The enriched domain search using public databases identified transcripts encoding phosphoinositide (PI) system and polyketide synthase (PKS) as functioning DEGs that were upregulated with [Ca 2+ ]-stimulated coccolith production (Tables 1 and 2). However, we did not find any domains enriched from the downregulated DEGs. Notably, PI-PLC domains were identified among the PI system in the PFAM database (PF00388 in Table 1) and the Interpro database (IPR000909, IPR001711, and IPR001192 in Table 2). Accordingly, GO enrichment analysis indicated PI-PLC as the gene that was upregulated in cells grown in 10 mM [Ca 2+ ] in our previous report [24]. Additionally, according to the protein-protein interaction network of DEGs, which are essential for understanding the cellular processes induced/activated under calcium-replete conditions, these upregulated DEGs showed an important role of PI-PLC in signal transduction ( Figure S3).

Inhibition of Coccolith Production by PI-PLC inhibitor U73122
Based on the identification of PI-PLC as a calcium-upregulated gene, the effect of the PI-PLC inhibitor U73122 on coccolith production was tested in E. huxleyi CCMP 371. We first examined the effect of U73122 at 0.01, 0.1, and 0.2 µM on cell growth; consequently, no inhibition of cell growth was observed for at least for 3 days when 10 mM [Ca 2+ ] and U73122 were simultaneously added to cells grown in 0.1 mM [Ca 2+ ] on Day 1 ( Figure S2). Based on these results, the effect of U73122 on coccolith production during the culture (as shown in Figure S2) was monitored by flow cytometry using a previously established calcein staining method, in which calcein binds to calcium ions in biominerals such as coccoliths [37]. This method allowed calcium detection in both coccoliths that were attached and those that were detached from the cells as a larger or smaller forward scatter (FSC, cell size), respectively (Figure 4a). Cells were additionally gated as chlorophyll-positive populations. The calcein-positive cell population was analyzed under each condition to estimate the quantity of coccoliths that were attached to intact cells or detached in the medium. The FSC value of the organic parts of the cells (recognized with chlorophyll fluorescence and shown as 'Cells (Chl+)' in Figure 4) was suppressed by 0.2 µM U73122 to 17% of the control, whereas that of the coccoliths was greatly diminished to 51% of the control at 10 mM [Ca 2+ ] (Figure 4a,b). The average geometric mean fluorescence intensity of the coccolith-gated calcein-positive population showed a significant decrease in all three U73122-treated conditions (23%, 20%, and 42% less at 0.01, 0.1, and 0.2 µM, respectively) ( Figure 4c). However, the coccolith production examined by U73343 treatment, which is known as an inactive analog of U73122, did not show a significant decrease in U73343-treated conditions compared to the control ( Figure S4). These results suggest the involvement of PI-PLC in the calcium-induced biomineralization by E. huxleyi CCMP 371.

Physiological Regulation of Cell Culture for Providing Calcium-Induced Calcified Cells
In this study, we first established a physiologically well-organized cell cultivation system to investigate the calcium-specific induction of intracellular calcification using the coccolithophore E. huxleyi CCMP 371 strain. To identify genes closely associated with coccolith production as well as intracellular calcification, we performed transcriptomic analysis in E. huxleyi cells by transferring non-calcifying cells grown in low [Ca 2+ ] to ambient seawater medium containing 10 mM [Ca 2+ ], where coccolith production was greatly stimulated (Figures 1 and 2). To promote coccolith production by cells, we selected 10 mM [Ca 2+ ] (final concentration) based on the ambient seawater calcium concentration (10 mM) that is reported to be the optimal condition for calcification in the literature [22][23][24][38][39][40][41].
This study provided experimental evidence supporting the fact that the calcification of E. huxleyi CCMP 371 is induced by the calcium supply and is independent of Pi conditions, since coccolith production was observed in both Pi-limited and -replete conditions (Figure 1). Therefore, the E. huxleyi CCMP 371 strain is very useful for investigating calcium-induced biomineralization without the effects of Pi. As phosphate limitation is generally known to increase calcification in other strains of E. huxleyi [13,[18][19][20], the regulatory mechanism appears to differ among strains, even within E. huxleyi. This strain-specific variation may be supported by evidence showing a wide variation of Emiliania genomes, including pan-genomes among strains isolated from various environments [42].
The quantitative estimation of coccoliths and cellular organic materials as TIC and TOC, respectively, by a TOC analyzer demonstrated that the cellular IC content (cell IC) represents coccoliths produced from IC absorbed from the carbon in the medium (Figure 2 and Figure S1). Although the total TOC in the cell fraction increased along with the cell number, the cellular TOC content (Cell C) remained nearly constant during culture at ambient [Ca 2+ ] ( Figure S1), indicating that the increase in [Ca 2+ ] resulted in the promoted uptake of both calcium and bicarbonate, followed by CaCO 3 crystallization for coccolith production.
We observed visible coccoliths on the cell surface at 3 h after the [Ca 2+ ] shift ( Figure 2a). The data are comparable with previous data by Paasche (1962) [43], who reported that the production of one coccolith takes less than 3 h under illumination, although the time taken varies among strains. Therefore, the experimental design used for the transcriptomic analysis of calcium-induced E. huxleyi CCMP 371 cells in this study was adequate for understanding the molecular mechanisms of calcium-induced coccolith production as an important biomineralization mechanism.

Identification of Calcium-Induced Biomineralization-Associated Genes
Using the established experimental design, transcriptomic analysis was performed to identify calcium-induced biomineralization-associated genes in E. huxleyi CCMP 371 (Figure 3). For RNA-seq in this study, over 50 million reads per sample (in four conditions with three biological replicates) were generated and processed for mapping the reads against previously assembled transcripts by Nam, Park, Lee and Jin [24]. The multidimensional scaling represents the degree of similarity between samples and shows that groups 1 and 2 were distinguished by the [Ca 2+ ] shift. The transcriptional response to calcium-induced biomineralization resulted in the identification of 116 upregulated and 47 downregulated DEGs. The enriched domains in the upregulated DEGs were associated with polyketide synthesis, signal transduction, and organelle transport (Tables 1 and 2). Inversely, no domains were enriched in the downregulated DEGs, probably owing to the small number of gene lists.

Association of PKS with Calcium-Induced Biomineralization
Several findings have linked PKS with calcification in the literature: (1) the biosynthesis of polyketides shares large similarities with fatty acid biosynthesis in bacteria, fungi, and plants [44]. (2) The link between PKS and calcium carbonate biomineralization was reported in teleost fish medaka Oryzias latipes [45]. In a spontaneous medaka mutant, PKS facilitated nucleation in otolith (calcareous ear stone) mineralization. (3) During the developmental stage of sea urchins, PKS expression is limited to skeletogenic cells and their precursors, which are responsible for calcium carbonate biomineralization [46]. The following data were found for E. huxleyi in the literature: (1) type I PKS genes in E. huxleyi were previously reported by John et al. [47]. (2) Microarray analysis revealed high PKS levels in a calcifying diploid strain of E. huxleyi RCC 1216 [48]. (3) Proteomic data obtained in E. huxleyi confirmed the expression of the modular type I PKS protein [49]. Despite these reports, the association of PKS with biomineralization has rarely been reported in E. huxleyi.
In the current study, PKS transcription was upregulated after the shift from 0.1 to 10 mM [Ca 2+ ], which might be associated with biomineralization. Actually, many functional domains associated with type I PKSs, such as ketoacyl synthase (PF02801, PF00109, IPR14031, IPR14030, IPR018201), ketoacyl reductase (PF08659), acyl transferase (IPR016035), enoyl reductase (IPR020843), acyl carrier protein (IPR036736), dehydratase (IPR020807), and thioesterase (IPR001031), were identified to be upregulated by the increase in [Ca 2+ ] used to induce and stimulate coccolith production (Tables 1 and 2). These data demonstrate that PKS has a possible role in calcium-induced calcification in E. huxleyi. Therefore, unveiling the relationship between PKS and biomineralization may advance the present knowledge on this bloom-forming cosmopolitan coccolithophorid alga E. huxleyi.

Association of PI-PLC with Calcium-Induced Biomineralization
Among the enriched domain functions related to signal transduction and organelle transport in DEGs upregulated by the elevation of [Ca 2+ ], phosphatidylinositol-specific phospholipase C (PI-PLC) and the phosphoinositide PI-PLC family were clearly identified as possible candidates associated with calcium-induced biomineralization (PF00388, IPR000909, IPR001711, and IPR001192 in Tables 1  and 2). These results, obtained in the present study, strongly confirm our previous study results, which suggested the possibility that PI-PLC participates in calcification in calcium-enriched E. huxleyi cells [24]. Further, the protein-protein interaction network analysis performed in the present study confirmed that the PI system is also working in the coccolith production mechanism in E. huxleyi strain CCMP 371 ( Figure S3).
Concerning the function of PI-PLC, PI-specific PI-PLC catalyzes the degradation of PIs, known as regulatory membrane components, in eukaryotes [50]; additionally several PI-PI-PLCs are known to be activated via interaction with their pleckstrin homology domains and the βγ subunits of G-proteins [51][52][53]. In mammalian cells, PI-PLC activity is necessary to regulate the formation of transport carrier fission in the trans-Golgi network [54].
As coccoliths are produced in CV, which should be derived from the Golgi body in E. huxleyi cells, the intracellular calcification process must have a complex interaction with the endomembrane systems [7,8,10,[55][56][57]. Based on this evidence, through endomembrane systems such as the endoplasmic reticulum (ER), the trans-Golgi trafficking can be regulated by IP 3 /Ca 2+ signaling system via PI-PLC in actively calcifying E. huxleyi cells.
The inhibition of coccolith production by the PI-PLC inhibitor U73122 (shown in Figure 4) demonstrates the association of PI-PLC with calcium-induced biomineralization in E. huxleyi. The coccolithophore E. huxleyi cells release the coccoliths from the cell surface into the medium during culture as revealed by a previous report, which estimated that 10-20 coccoliths per cell were usually detached from exponentially growing cells per day [8]. Our data evaluating the population of coccoliths covering cells and those detached/released into the medium showed that the number of both cell-attached and detached coccoliths, namely the entirety of coccoliths produced, were increased by [Ca 2+ ] enrichment, and all coccolith production was hindered by the PI-PLC inhibitor U73122 ( Figure 4). These data indicate that PI-PLC must be involved in the intracellular calcification process resulting in coccolith production.
Recently, successful genetic transformation methods have been established for other haptophytes, especially in the coccolithophore Pleurochrysis carterae [58,59]. Although P. carterae has a calcification mechanism in the subcellular system for intracellular calcification that differs morphologically and physiologically, the basic molecular system of coccolith production, such as calcium accumulation and bicarbonate transport, may be roughly similar. In E. huxleyi, the calcium/phosphate-rich phase in a different compartment from the coccolith-producing vesicles functions as a calcium reservoir [60]. Additionally, it has been shown lately that P. carterae also contains a calcium storage compartment [61]. Based on this evidence, our finding in this study carefully suggests that PI-PLC functions to accumulate calcium in this reservoir and supports the stimulation of intracellular calcification that is promoted under calcium-rich conditions in E. huxleyi cells. Further analysis using such genetic manipulation should be introduced to understand the regulatory mechanism of individual genes related to calcium-induced biomineralization in vivo, although the investigated genetic transformation techniques need to be established in E. huxleyi.

Conclusions
Here, we demonstrate that the biomineralization of E. huxleyi CCMP 371 is not entirely dependent on phosphate limitation, which is known to be a primary trigger for coccolith production in other strains of E. huxleyi, and the regulatory mechanism appears to differ among strains as they were isolated from various environmental conditions and habitats. Actually, this study clearly showed that the elevation of calcium concentration from 0.1 to 10 mM triggers and promotes the mineralization process of the E. huxleyi CCMP 371 strain even under phosphate-sufficient conditions in which coccolith production is usually suppressed in other strains. Based on the RNA-seq and PI-PLC inhibitor experiments, the present study suggests the involvement of PI-PLC in the calcium-induced coccolith formation mechanism in E. huxleyi CCMP 371.  Table S1: Raw data statistics of RNA sequencing, Table S2: Primers used for qPCR, Table S3: Differentially expressed genes in calcium-induced biomineralization.

Conflicts of Interest:
We declare that we have no conflict of interest.