Nannochloropsis oceanica was obtained from Leadingtec biotechnology company, Shanghai, China. The F/2 seawater culture medium was used in the experiment as described elsewhere (Cheng et al., 2016), with the specific contents of NaNO₃ and NaH₂PO₄ H₂O adjusted to 2.64 and 0.20 mmol L⁻¹, respectively.

The algal species were cultivated at 27°C in a 300 mL column glass bioreactor with 160 mm height, 56 mm inner diameter (ID) and 62 mm outer diameter (OD) in an artificial greenhouse, with the light intensity controlled at 7,000 Lux. A 1% CO₂ mixture in nitrogen was bubbled into the bioreactor at a flow rate of 30 mL min⁻¹. The bioreactor parameters and biomass measurements are described in our previous work (Cheng et al., 2016). All growth and oxygen evolution experiments were done in duplicate.

Solutions of N. oceanica with initial OD₇₅₀ of 1.5 were placed in fifteen 50 mL triangle bottles (three bottles for each radiation dosage) and irradiated with ¹³⁷Cs-γ rays at dosages of 200, 300, 500, 700, and 900 Gy and a dose rate of 1 Gy min⁻¹. Experiments were performed at the Academy of Agricultural Sciences, Zhejiang Province, China. The nuclear radiation generator was the same as described in the previous work (Cheng et al., 2016). After radiation, microalgal cells were placed in an incubator for 1 month for self-healing at 20°C and under an illumination of 1,000 Lux. The survival recovery percentage at each irradiation dosage was calculated as follows:

The survival recovery percentage was defined as 100% for the wild type, because no cells were died when there was no radiation.

Afterwards, 0.2 μL of microalgae from each triangle bottle treatment was injected into agar plates made of F/2 seawater medium described above with 1.5% agar, which was irradiated at different dosages, for purification of a single strain. After purifying thousands of mutants on agar plates, single colonies were picked into 96-well plate with F/2 seawater medium and cultivating in 96-well plate for 7 days, three strains exhibiting the highest biomass concentration at each radiation dosage [reflected by the optical density value from microplate reader (Tecan Infinite 200, Switzerland), OD405 was used as instructed] were screened and incubated for 8 days in 20 mL test tubes containing 5 mL of culture medium. During the large-scale cultivation, cell concentrations were measured by the spectrophotometer with a wavelength at 750 nm (Griffiths et al., 2011) and the strains with the fastest growth rate were further screened.

The relationship between dried weight of microalgal biomass and optical density of 750 nm is as follows:

Oxygen evolution was measured in duplicate by Oxygraph (Hansatech, UK). For each treatment, a 1 mL culture sample was taken into the center column for the oxygen evolution test. The center column was illuminated at 350 μmol (m²·s)⁻¹ using a red-light source. The oxygen evolution rate was calculated using the accumulated oxygen concentration and the time of the steady increase phase as following:

where C_t1 is the accumulated O₂ concentration at time t₁ and C_t0 is the accumulated O₂ concentration at time t₀, t₁ = t₀ + 60 s. The oxygen concentration was measured every second, while the data points shown in the figures represent every 5 s.

Main components (protein, carbohydrates, and lipids) in N. oceanica and the mutant were measured on the 4th day of 1% CO₂ bubbling cultivation. The measurement methods of total carbohydrates content in N. oceanica biomass were determined by modified Anthrone method (Carrieri et al., 2010). The measurement methods of proteins and lipids contents in N. oceanica biomass were same as described in a previous work (Su et al., 2010).

On the 4th day of cultivation, two duplicate samples of wild type and mutant ZJU700 were filtered, quick-frozen in liquid nitrogen, collected and stored at −80 °C, which were sent to Shanghai OE Biotech Co., Ltd for further analysis. High-throughput transcriptome sequencing was conducted on both the wild type and mutant ZJU700 samples. The manufacturer's protocols were followed when using the mirVana miRNA kit (Thermofisher) for RNA extraction; the Truseq stranded mRNA LT sample prep kit (Illumina) for sequencing library preparation; and the Bioanalyzer's High Sensitivity DNA kit (Agilent) for quality control of the libraries. Gene expression levels were compared between the wild type and mutant ZJU700.

The P-value analysis was conducted by the software “IBM SPSS Statistics 20,” which could give significance test result (P-value) when there were two replicates or more. This test could tell if the difference between two samples was significant.

Figure 1A shows the survival recovery percentage of N. oceanica after self-healed for 30 days from the irradiation with various dosages of γ rays. The survival recovery percentage decreased with the increasing γ radiation dosages, indicating that microalgae had a higher mortality rate under higher radiation intensity. This was because γ rays from nuclear radiation generate oxidative stress on thylakoid structures in microalgal cells, thereby impacting the PSII and Calvin cycle, or even causing death (Fuma et al., 2009).

Meanwhile, photosynthesis of the remaining cells was more likely to be affected under higher radiation intensity, thus generating more phenotypes. After 8 days of large-scale cultivation in test tubes, the average cell concentration of the three best screened strains (reflected by OD₇₅₀ value) were shown in Figure 1A. These three strains showed various growth phenotypes, which might have led to a large error bar. Little effect was observed on the screened mutant under the lowest irradiation (200 GY) as the cell concentration was nearly identical to the wild type. The biomass concentration of screened mutants increased with increasing radiation dosages, peaking at 700 GY, and then decreased at the irradiation of 900 GY. This might because genes in Nannochloropsis mutant cells were destructured and recombined to a greater extent under a higher radiation intensity, with the survived cells undergoing greater changes via acute evolutionary processes. After 700 GY radiation dosage, the average biomass concentration of the screened mutants increased by 84.2% compared with the wild type; however, the biomass concentration (OD₇₅₀ value) decreased to 0.80 when the radiation dosage further increased to 900 GY, which might due to the extensive damage to the mutant cells caused by the excessive radiation.

Wild type Nannochloropsis and mutant strains with the highest concentrations at radiation dosages of 300, 500, and 700 GY were then cultivated in the column photobioreactors. Growth curves were shown in Figure 1B. The dried biomass weight of all mutants were higher than the wild type. The mutant strain irradiated at 700 GY showed the highest biomass productivity, which was 26.7% higher than that of the wild type. This mutant strain showed the best growth phenotype after several rounds of screening; thus, it was designated as Nannochloropsis ZJU700 and used for the subsequent experiments. The mutant strain ZJU700 had been cultivated for tens of generations in the lab and showed similar growth phenotype among each generation, which proved the stability of this mutant. On the 4th day of the 1% CO₂ bubbling cultivation, the carbohydrates content of the mutant (30%) was higher than the wild type N. oceanica (23%), whereas proteins and lipids content of the mutant (27 and 32%, respectively) were a little lower than the wild type (30 and 35%, respectively). The following two pathways were among the KEGG enrichment top 20 of wild type N. oceanica and the mutant: carbon fixation in photosynthetic organisms and citrate cycle, which were the most important pathways for carbon and biomass accumulation (Figure 2). This result indicated that the carbon fixation process changed greatly in the mutant ZJU700 cells, and finally led to its higher biomass productivity.

Interestingly, N. oceanica reached its highest biomass productivity at a radiation dosage of 700 GY, which was the same dosage at which Chlorella sp. peaked in our previous study (Cheng et al., 2016). Conversely, Spirulina (a prokaryotic organism) grew fastest after irradiation at 9,000 GY (Cheng et al., 2017). Therefore, the “best intensity” of Spirulina was 10 times higher than that of Nannochloropsis and Chlorella (eukaryotic microalgal strains). These results indicated that eukaryotic strains were more sensitive to γ rays from nuclear radiation and were more likely to attain the best following irradiation at around 700 GY, whereas prokaryotic strains, such as Spirulina would achieve the best phenotype at a much higher radiation intensity.

To further explore the reason for the improved growth phenotype of Nannochloropsis mutant ZJU700, oxygen evolution rates of mutant ZJU700 and the wild type were measured (Figure 3). On the 4th day, the oxygen evolution rate of mutant ZJU700 was 44.63 nmol/mg/min, which was 2.7% higher than the wild type (Figure 3B). When taking the biomass concentration into account, the oxygen evolution rate of mutant ZJU700 was 41.95 nmol/mL/min, which was 30.4% higher than the wild type (Figure 3b). This result shows that mutant ZJU700 produces oxygen much faster than the wild type.

Given the steady increase of accumulated oxygen (over 3 min) during the oxygen evolution measurement, constant oxygen evolution rates were obtained in both the wild type and mutant ZJU700, indicating steady downstream carbon assimilation during the Calvin cycle to pull the photosynthetic oxygen evolution by consuming its products. A higher carbon fixation rate could be inferred from the higher oxygen evolution rate, thereby contributing to the higher growth rate in the mutant. On the 6th day, the oxygen evolution rates of mutant ZJU700 and the wild type both decreased when compared with those on the 4th day, indicating that the carbon fixation rates decreased and the growth of the microalgae shifted from the logarithmic to the stationary phase.

The O₂ evolution was significantly affected by photosynthetic and respiration processes. Gene expression levels between the wild type Nannochloropsis and mutant ZJU700 were analyzed, expression levels of photosynthesis related genes were much higher in the mutant than the wild type, while the respiration related genes were not found differently expressed.

A photoreaction splits water into oxygen, synthesizes NADPH and ATP, and provides energy and materials for carbon fixation in cells. After nuclear radiation mutagenesis, mutant ZJU700 showed not only a faster oxygen evolution rate at the macro level, but also a higher expression of photoreaction-related genes at the genetic level (Figure 4).

The Calvin cycle, an energy-consuming process that converts CO₂ into organic carbon, is an open system connected to light photosynthetic reactions and plays a central role in algal metabolism (Jablonsky et al., 2011; Yang et al., 2017).

Results of transcriptome sequencing showed that most Calvin cycle-related genes were up-regulated in mutant ZJU700 (Figure 5). The rbcL and rbcS genes, which catalyse the primary carboxylation of ribulose 1,5-bisphosphate to yield two 3-phosphoglycerate molecules (Hügler and Sievert, 2011), were expressed 9.38 and 5.18 times higher, respectively, than the wild type (Table 1). The expression of gapdh, which catalyses the conversion of 1,3-bisphosphoglycerate into glyceraldehyde-3-phosphate (Michelet et al., 2013), increased by 2,024%. The prk gene catalyses phosphorylation of ribulose-5-phosphate to ribulose-1,5-bisphophate using ATP generated by thylakoid ATP synthase (Michelet et al., 2013). In mutant ZJU700, prk gene was expressed 2.30 times higher than in the wild type.

γ radiation led to the up-regulation of Calvin cycle-related genes and was likely to accelerate biomass accumulation while consuming more ATP generated from upstream photoreactions. This was echoed by the enhanced upstream processes, such as the higher oxygen evolution rate.