Development of a Screening Method for Isolation of Microalgae Strains That Accumulate Lipids Under Nitrate-replete Conditions

Background: Microalgae biofuels have attracted global attention as an alternative to fossil fuels as an energy resource. Microalgae generally accumulate lipids under nitrogen-depleted conditions, but cell growth is depressed under these conditions which causes decrease in lipid productivity. To realize one-step cultivation for biofuel production, microalgae that highly accumulate lipids even under nitrogen-replete conditions are needed. This study aimed to develop a screening method for microalgae mutants with high lipid content even in the presence of a nitrogen source. Results: Mutant cells were generated by irradiating the oleaginous green microalga Chlamydomonas sp. KOR1 with carbon ion beams, cultured under nitrate-replete conditions, and then subjected to FACS-based screening for lipid-rich cells. By repeatedly performing the sequential procedures of cultivation and selection, strains KAC1710 and KAC1801, which highly accumulate lipids under nitrate-replete conditions, were successfully obtained. These mutants formed signicant lipid droplets in the cells even in the presence of abundant nitrate and achieved 1.5- and 2.1-fold greater lipid content compared to KOR1, respectively. Conclusion: This study developed a novel nitrogen-conditioned screening method for microalgae mutants that accumulate lipids in the presence of a nitrogen source. This method should contribute to microalgae biofuel production via one-step cultivation under nitrogen-replete conditions. lipids. Further studies are required to elucidate the lipid accumulation mechanism in KAC1710 and KAC1801, This study developed a screening method to identify valuable microalgae mutants that can accumulate lipids under nitrate-replete conditions. The mutants KAC1710 and KAC1801 were obtained by nitrogen-conditioned screening with FACS. Under nitrate-replete conditions, KAC1710 and KAC1801 formed signicant lipid droplets in the cells and accumulated 1.5- and 2.1-fold more lipid than the parent strain, respectively. For microalgae biofuel production, the results of this study should contribute to the establishment of a method for one-step cultivation under nutrient-replete conditions.

Previous studies have identi ed target genes for metabolic engineering aimed at improving lipid accumulation in the presence of nutrients. In Nannochloropsis gaditana, a transcription factor downregulated under nitrogen de ciency, Zn (II) 2 Cys 6 , was identi ed, and down-regulation of this transcription factor improved lipid production under growth condition [20]. In Cyanidioschyzon merolae, overexpression of glycerol-3-phosphate acyltransferase (GPAT) induced a 56.1-fold increase in TAG productivity in the growth phase [21,22]. However, the regulatory mechanism of lipid accumulation under nitrogen-replete conditions remains poorly understood, even though it is important for identifying target genes for metabolic engineering. Information regarding this regulatory mechanism could also be obtained from comparative analyses of lipid-rich microalgae mutants.
Irradiation with heavy-ion beams functions as a useful mutagen that induces drastic changes in the genome, and this approach has been applied to microalgae such as Nannochloropsis [28], Euglena [29], and Chlamydomonas [30]. Fluorescence-activated cell sorting (FACS) is a powerful tool for screening lipid-rich cells from a large number of random mutants. Boron-dipyrromethene (BODIPY) and Nile red, uorescent dyes that stain lipid droplets in living cells, are commonly used in combination with FACS [29,[31][32]. Although high lipid-producing strains have been obtained using these tools, the screenings were only conducted under conditions suitable for lipid production, for example, nitrogen-de cient conditions. Thus, a screening method for microalgae that accumulate lipids under nitrogen-replete conditions has not been established.
This study aimed to develop a screening method to obtain microalgae mutants that accumulate lipids under nitrogen-replete conditions. As the parent for mutational breeding, the oleaginous microalga Chlamydomonas sp. KOR1 was exposed to carbon ion beam for random mutagenesis, and the resultant mutant cells were cultured in the presence of abundant nitrate until just before FACS-based screening for lipid-rich cells. This sequential screening approach led to the isolation of strain KAC1710, which highly accumulated lipids under nitrate-replete conditions. By repeating the nitrogen-conditioned screening procedure using KAC1710 as the parent, KAC1801, in which lipid accumulation under the nitrate-replete conditions was further improved compared to KAC1710, was obtained. In the presence of nitrate, the lipid productivity of KAC1710 and KAC1801 was 1.3-and 1.8-fold higher than that of KOR1, respectively. Thus, this study developed a novel nitrogen-conditioned screening method that is useful for obtaining lipidaccumulating mutants under nitrogen-replete conditions.

Results
First-round breeding using Chlamydomonas sp. KOR1 as the parent For one-step lipid production, microalgae that can accumulate lipids under nitrogen-replete conditions are needed. It was assumed that screening of lipid-rich cells from randomly mutated cells cultured in the presence of an abundant source of nitrogen would be useful to obtain such mutants. To examine this hypothesis, mutational breeding was performed using Chlamydomonas sp. KOR1 as the parent. KOR1 was irradiated with carbon ion ( 12 C 5+ ) beams to induce random mutations. The mutant cells were cultured under the nitrate-replete condition in MB 12N medium containing 2% (w/v) sea salt for 3 days. MB12N medium contains abundant NaNO 3 (1588.4 mg/L) than previously used medium [11]. Nitrate concentration in the medium after 3 days cultivation was approximately 800 mg/L which shows nitrate in the medium was replete. Then, mutants were subjected to FACS-based screening of lipid-rich cells highly stained by BODIPY 505/515.
In the rst sorting experiment, the chlorophyll uorescence and BODIPY uorescence intensity of most cells were 1.0 × 10 5 ~ 1.0 × 10 6 (a.u.) and 1.0 ×10 4 ~ 2.0 × 10 5 (a.u.), respectively (Fig. 1a). From these, 1,000 cells exhibiting high BODIPY uorescence of 1.0 × 10 5 ~ 1.0 × 10 6 (a.u.) were sorted (Additional le1: Fig. S1 and Table S1). The sorted cells were subjected to repeated cultivation under nitrate-replete conditions and FACS-based sorting, for a total of ve times. In the fth sorting experiment, a cell population exhibiting higher BODIPY uorescence than the original cells was observed (Fig. 1b), suggesting that mutant cells that accumulate lipids in the presence of nitrate were successfully screened. Then, the mutant colonies were isolated by inoculating the cells obtained in the fth-sorting on Trisacetate-phosphate (TAP) agar plates. For the secondary screening, 18 mutant strains were randomly selected (designated as KAC17s), cultured under nitrate-replete conditions for 3 days, at which time the lipid content was measured using gas chromatography-mass spectrometry (GC-MS). Among these mutants, KAC1710 exhibited the highest lipid content (18.3%), compared with lipid content of only 7.2% for KOR1 (Fig. 1c).
Second-round breeding using Chlamydomonas sp. KAC1710 as the parent To further increase the lipid content under nitrate-replete conditions, second-round breeding was conducted using KAC1710 as the parent. Similar to the rst-round breeding, KAC1710 was irradiated with a heavy-ion beam, and the nitrogen-conditioned screening of lipid-accumulating cells was performed using FACS. In the rst sorting experiment, chlorophyll uorescence and BODIPY uorescence of most cells were 1.0 × 10 5 ~ 10 6 (a.u.) and 1.0 × 10 5 ~ 1.0 × 10 6 (a.u.), respectively (Fig. 2a). From these, 1,000 cells exhibiting higher BODIPY uorescence of approximately 3.0 × 10 6 (a.u.) were sorted (Additional le1: Fig. S2 and Table S2). The sorted cells were repeatedly subjected to cultivation under nitrate-replete conditions and sorting of high BODIPY uorescence cells, for a total of four times. In the fourth sorting experiment, most cells exhibited BODIPY uorescence higher than 1.0 × 10 6 (a.u.) (Fig. 2b), suggesting that further improvement in lipid accumulation under nitrate-replete conditions was achieved. The cells obtained in the fourth sorting were then spread on TAP agar plates to isolate mutant strains. For the secondary screening, 15 mutant strains were randomly selected (designated as KAC18s), cultured under nitrate-replete conditions, and then their lipid content was analyzed by GC-MS. Among these mutants, KAC1801 exhibited the highest lipid content (23.1%), compared to 15.2% for the parental strain KAC1710 (Fig. 2c).

Increased formation of lipid droplets in KAC1710 and KAC1801
Transmission electron microscopy (TEM) analysis revealed that KAC1710 and KAC1801 cells accumulated lipid droplets even in the presence of nitrate, and KAC1801 accumulated more lipid droplets than KAC1710 (Fig. 3). In addition, KAC1710 and KAC1801 cells were larger in size than KOR1 cells, which might have been caused by the increased lipid droplet formation in these mutants. These results suggested that formation of lipid droplets under nitrate-replete conditions was increased by the two random mutagenesis and nitrate-conditioned screening steps.
Evaluation of lipid productivity of KAC1710 and KAC1801 To evaluate lipid productivity under nitrate-replete conditions, KOR1, KAC1710, and KAC1801 were cultured in MB 12N medium containing 2% (w/v) sea salt, and cell density, biomass, nitrate concentration in the medium, and lipid content were measured. Maximum cell density of KOR1, KAC1710, and KAC1801 during 6 days of cultivation was 8.0 × 10 6 cells/mL, 5.7 × 10 6 cells/mL, and 5.3 × 10 6 cells/mL, respectively (Fig. 4a). Maximum biomass of KOR1, KAC1710, and KAC1801 was 4.1 g-DCW/L, 3.6 g-DCW/L, and 3.3 g-DCW/L, respectively (Fig. 4b). These results suggest that delayed cell division in KAC1710 and KAC1801 led to the lower biomass production. In addition, nitrate consumption was decreased in KAC1710 and KAC1801 (Fig. 4c). KAC1710 and KAC1801 consumed 79.8% and 62.9% of the nitrate in the MB 12N medium, respectively, whereas KOR1 completely consumed the nitrate in the medium over 6 days. The lipid content of KAC1710 and KAC1801 at day 5 was approximately 19.1% and 26.6%, respectively, which was 1.5-and 2.1-fold higher than that of KOR1 (12.5%) (Fig. 4d). Although biomass production was reduced, lipid production and productivity of KAC1710 and KAC1801 were improved based on the increase in lipid content. Lipid production of KOR1, KAC1710, and KAC1801 at day 5 was 457.8 mg/L, 605.4 mg/L, and 810.4 mg/L, respectively (Fig. 4e). Lipid productivity of KAC1710 and KAC1801 at day 5 was 1.3-fold (121.1 mg/L/day) and 1.8-fold (162.1 mg/L/day) higher, respectively, compared to KOR1 (91.6 mg/L/day) (Fig. 4f). Thus, by increasing lipid accumulation, lipid productivity under nitrate-replete conditions was improved.
Lipid productivity of KAC1710 and KAC1801 under nitrate-de cient conditions was also investigated using MB 6N medium containing 2% (w/v) sea salt, which contains half the amount of nitrate compared to MB 12N medium (Additional le1: Fig. S3). Biomass production of KAC1710 and KAC1801 after 10 days of cultivation was 5.1 g-DCW/L and 4.4 g-DCW/L, respectively, whereas that of KOR1 was 5.7 g-DCW/L (Additional le1: Fig. S3a). Nitrate in the medium was completely consumed in 3 days by KOR1, whereas 4 days were required by KAC1710 and KAC1801 (Additional le1: Fig. S3b). The lipid content of KOR1, KAC1710, and KAC1801 at day 10 was 43.5%, 50.7%, and 37.1%, respectively (Additional le1: Fig.  S3c); thus, under nitrate-depleted conditions, improvement was observed only in KAC1710. Also, lipid production of KOR1 and KAC1710 was 249.7 mg/L/day and 259.8 mg/L/day, respectively, whereas in KAC1801, it decreased to 163.5 mg/L/day (Additional le1: Fig. S3d).

Discussion
Microalgae such as Chlorella, Nannochloropsis, Scenedesmus, and Chlamydomonas are promising biofuel producers, but lipids are generally accumulated under adverse environmental conditions such as nitrogen de ciency [19]. To examine lipid production under nitrogen-replete conditions, this study developed a novel nitrogen-conditioned screening method to obtain lipid-accumulating mutants. Previous research also succeeded in screening for lipid-rich mutant microalgae [28,29,31,32]; however, those screenings were performed only under conditions suitable for lipid accumulation. This study employed nitrogen-replete conditions in which microalgae generally do not accumulate lipids and conducted nitrogen-conditioned screening in reference to previous studies involving random mutagenesis and FACSbased screening. Strains KAC1710 and KAC1801 obtained in the present screening formed many lipid droplets in the presence of nitrate (Fig. 3) and accumulated 1.5-and 2.1-fold more lipids than KOR1, respectively (Fig. 4d). This is the rst study that successfully screened lipid-accumulating strains under nitrogen-replete conditions from randomly mutated microalgae.
The lipid productivity of KAC1801 under nitrate-replete conditions was 162 mg/L/day ( Fig. 4f and Table   1), higher than that of Nannochloropsis sp. (55 mg/L/day) and Scenedesmus obliquus (78 mg/L/day) [33,34] and comparable to that of Isochrysis zhangjiangensis (141 mg/L/day) and Nannochloropsis oculata (142 mg/L/day) under nitrogen-depleted conditions [35,36]. Although lipid productivity was still lower than that of Dunaliella salina (240 mg/L/day) and Chlamydomonas sp. (306 mg/L/day) cultured in the absence of a nitrogen source [37,11], these results suggest that KAC1801 is a promising strain that can produce a comparable amount of lipids to previously reported microalgae under nitrate-replete conditions.
Under nitrate-depleted conditions, on the other hand, the lipid productivity of KAC1710 and KAC1801 did not change or even decreased compared to that of KOR1 (Additional le1: Fig. S3). Thus, it was suggested that the response to nitrate was modi ed in these mutants, whereas the lipid synthesis pathway itself was not generally improved. Strains KAC1710 and KAC1801 exhibited decreased nitrate consumption compared to KOR1 (Fig. 4c). This suggested that nitrate assimilation was reduced in these mutants. Target of rapamycin (TOR) signaling reportedly plays an important role in nutrient sensing in Chlamydomonas reinhardtii. Previous research reported that TOR inactivation mimics nitrogen-de cient conditions and induces lipid accumulation despite the presence of a nitrogen source in the medium. It was also reported that TOR-inactivation condition increases the expression of lipid synthesis related genes such as diacylglycerol acyltransferase [21]. It is hypothesized that partial dysfunction of TOR in KAC1710 and KAC1801 enabled the cells to accumulate lipid droplets under nitrate-replete conditions. In Nannochloropsisgaditana, the transcriptional factor Zn (II) 2 Cys 6 was identi ed as a negative regulator of lipid accumulation. Zn (II) 2 Cys 6 down-regulated mutant showed biosynthesis switching from proteins to lipids. In Zn (II) 2 Cys 6 down-regulated mutant con rmed to decrease the expression levels of nitrogen assimilation genes such as nitrate transporter (NRT2), nitrate reductase (Nir), glutamine synthetases (GS1 and GS2), and ammonium transporter (AMT1) than wild-type strain [20]. KAC strains also decreased nitrogen assimilation and increased lipid accumulation (Fig. 4 c-d). This result also suggested that suppression of nitrogen assimilation may restrict carbon ux to protein, resulting in allocating carbon to lipids. Further studies are required to elucidate the lipid accumulation mechanism in KAC1710 and KAC1801, however.

Conclusions
This study developed a screening method to identify valuable microalgae mutants that can accumulate lipids under nitrate-replete conditions. The mutants KAC1710 and KAC1801 were obtained by nitrogenconditioned screening with FACS. Under nitrate-replete conditions, KAC1710 and KAC1801 formed signi cant lipid droplets in the cells and accumulated 1.5-and 2.1-fold more lipid than the parent strain, respectively. For microalgae biofuel production, the results of this study should contribute to the establishment of a method for one-step cultivation under nutrient-replete conditions.

Strains
Chlamydomonas sp. KOR1, a lipid-rich mutant derived from Chlamydomonas sp. JSC4 [38], and its mutants were used in this study. Microalgae were maintained on BG-11 plates containing 1.5% agar under continuous illumination at 50 µmol photons m −2 s −1 and 25°C.

Mutagenesis
Microalgae were cultured for 2 days in TAP medium [39] using double-deck photobioreactors constructed using two asks [40]; the upper stage contained 70 mL of cell culture, and the lower stage contained K 2 CO 3 /KHCO 3 solution adjusted to supply 2% CO 2 gas to the upper stage under continuous illumination at 100 µmol photons m −2 s −1 at 30°C. The optical density at 750 nm (OD 750 ) was measured using a UV mini-1240 UV-VIS spectrophotometer (Shimadzu, Kyoto, Japan). The cell culture was diluted with TAP medium to make cell suspension of OD 750 = 0.5, and 100 µL of the diluent was seeded onto TAP agar plates. The plates were covered with polyimide lm Kapton 30EN (Du Pont-Toray CO. Ltd., Aichi, Japan) and irradiated with 50 Gy of heavy-ion beams ( 12 C 5+ , accelerated energy; 220 MeV, surface LET; 107 keV/ μm) accelerated by an azimuthal varying eld cyclotron at Takasaki Ion Accelerators for Advanced Research Application of the National Institutes for Quantum and Radiological Science and Technology [30]. and metals described in a previous report [41]) containing 2% (w/v) sea salt (Sigma-Aldrich Co., St. Louis, MO, USA) and 2% CO 2 . A total of 5.0 × 10 6 cells were collected by centrifugation at 8,000 × g for 1 min at 25°C and resuspended in 1 mL of PBS. Then, 50 μM BODIPY 505/515 (4,4-di uoro-1,3,5,7-tetramethyl-4bora-3a,4a-diaza-s-indacene, Thermo Fisher Scienti c, MA, USA) was added to the cell suspension. After incubation for 5 min at room temperature in the dark, cells with the highest BODIPY uorescence were sorted using a uorescence-activated cell sorter SH-800 (SONY, Tokyo, Japan). To perform the screening procedures repeatedly, sorted cells were subjected to the next cultivation and FACS-based sorting described above. After the nal sorting, sorted cells were seeded on TAP agar plates and cultured for 1 week until colony formation.

Nitrogen-conditioned screening
For secondary screening, cells from isolated colonies were cultured for 3 days under nitrate-replete conditions using 12-well plates under continuous illumination of 100 µmol photons m −2 s −1 supplying 2% CO 2 at 30°C with rotary shaking at 100 rpm. Lipid content of the cells was measured as described below.

Measurement of nitrate concentration
Nitrate concentration was measured as previously reported [14,42]. The culture broth was centrifuged at 8,000 × g for 1 min. The supernatant was diluted with distilled water, and the optical density at 220 nm was measured using a UV mini-1240 UV-VIS spectrophotometer (Shimadzu). Nitrate concentration was calculated using a calibration curve.

Lipid analysis
Lipid content was measured as described in a previous study [11]. Cells in culture broth were collected by centrifugation at 8,000 × g for 1 min at 25°C and washed once with distilled water. Cell pellets were stored at −30°C until subjected to freeze-drying. A total of 2-3 mg of dried cells was fractured with 300 µL of 0.5mm glass beads YGB05 using a multi-beads shocker MB1001C (S) (Yasui Kikai, Osaka, Japan) at 2,700 rpm for 1 min: On, 1 min; Off × 30 cycles, 4°C. Released lipids were esteri ed using a fatty acid methylation kit (Nacalai Tesque, Kyoto, Japan) and analyzed on a GCMS-QP2010 plus (Shimadzu) instrument equipped with a DB-23 capillary column (60 m, 0.25 mm internal diameter, 0.15 µm lm thickness; Agilent Technologies, CA, USA) for identifying and quantifying fatty acids. Heptadecanoic acid (Sigma-Aldrich Co.) was used as an internal standard. The intracellular lipid content was shown by calculating lipid weight per dry cell weight.

TEM analysis
Cells were xed overnight with 2% paraformaldehyde, 2% glutaraldehyde, and 50 mM cacodylic acid. After dehydration in graded ethanol solutions (50-100%), samples were in ltrated with propylene oxide and transferred to 100% resin and polymerized at 60°C for 2 days. The specimens were sectioned and stained with 2% uranyl acetate at room temperature for 15 min. Observation was performed using a JEM-1400 plus electron microscope with a CCD camera EM-14830RUBY2 (JEOL Ltd., Tokyo, Japan).
Evaluation of the lipid productivity of KOR1 and its mutants Batch culture for evaluating lipid productivity was performed using the KOR1, KAC1710, and KAC1801 strains. Pre-culture and primary culture were performed using MB 12N medium containing 2% (w/v) sea salt for 3 days and 6 days under continuous illumination at 250 µmol photons m −2 s −1 at 30°C. During the primary culture period, the cell density, biomass density, nitrogen source concentration in the medium, and intracellular lipid content were analyzed every day. The cell density was analyzed using a TC20 TM Automated Cell counter (Bio-Rad, USA), and the other parameters were analyzed by the above method.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
The data set obtained in this study can be con rmed from the main article and supplementary information.

Competing interest
The authors declare no competing interests.

Funding
This study was conducted with the support of the Impulsing Paradigm Change through Disruptive Technologies (ImPACT) program, Cabinet O ce, Government of Japan.