Physiological and transcriptome analysis elucidates the metabolic mechanism of versatile Porphyridium purpureum under nitrogen deprivation for exopolysaccharides accumulation

Porphyridium purpureum is a mesophilic, unicellular red alga rich in phycoerythrin, sulfate polysaccharides, and polyunsaturated fatty acids. Nitrogen deficiency inhibited the growth of P. purpureum and resulted in yellowing of the cells and thickening of the extracellular viscousness sheath. Under nitrogen stress, the contents of total lipids and exopolysaccharides in P. purpureum were increased by 65.2% and 188.0%, respectively. We demonstrate that the immediate response of P. purpureum to nitrogen deficiency is mediated by carbon flow to polysaccharide synthesis, while the synthesis of lipids is enhanced as a permanent energy storage substance at the later stage. Based on transcriptome annotation information, we elucidate the synthesis pathway of polysaccharides from P. purpureum from the perspective of glycosyl-donor interconversion, and demonstrate that the n-6 pathway is the main synthesis pathway of polyunsaturated fatty acids. This study not only provides a production strategy for polysaccharides and fatty acids by single-celled marine red algae P. purpureum, but also provides targets for further genetic modification. Supplementary Information The online version contains supplementary material available at 10.1186/s40643-021-00426-x.


Introduction
Ancestral lineages of red algae have acted as mediators between prokaryotes and photosynthetic eukaryotes and formed a monophyletic lineage that consists of several unicellular mesophilic and extremophilic species as well as economically important seaweeds, thereby significantly increasing the diversity of photosynthetic organisms. Porphyridium purpureum is a unicellular autotrophic marine red alga that is rich in phycoerythrin, energy transfer (Ma et al. 2020). P. purpureum exopolysaccharides are composed of xylose, galactose, glucose, and other monosaccharides and are rich in glucuronic acid and sulfate groups, which confers stronger antioxidant capacity (Sun et al. 2009). Arachidonic acid (ARA), a long-chain ω-6 PUFA, is important for brain and cardiovascular system development, especially for infant cognitive development, but the traditional deep-sea fish oil-derived ARA may be contaminated by heavy metals (Timilsena et al. 2017). Thus, a healthier and more sustainable source is needed. Actually, P. purpureum is one of the few microalgae that are rich in ARA, and the culture medium is free of heavy metals, making it a potential host for producing ARA (Khozin-Goldberg et al. 2011).
Abiotic stress is one of the strategies to improve microalgae productivity. Studies have shown that high light intensities promote lipid accumulation and increase the proportion of triacylglycerol (TAG) in microalgae (Kurpan Nogueira et al. 2015). Appropriate increases in temperature can promote lipid accumulation and improve neutral lipid content, whereas low temperature reduces PUFA content and membrane fluidity (Renaud et al. 2002). For nutrient stress, nitrogen and phosphorus stress has been widely used for metabolism regulation in microalgae (Jaeger et al. 2017;Liu et al. 2019;Miller et al. 2010;Siaut 2011). Transcriptome analysis has shown that Chlamydomonas reinhardtii reinforces the de novo synthesis of fatty acids and membrane lipid turnover to promote carbon flow to lipid accumulation under nitrogen deficiency (Miller et al. 2010). Studies combining transcriptome with lipidome analyses have shown that several metabolic pathways are stimulated to cope with nitrogen depletion in Nannochloropsis oceanica IMET1, which include upregulation of Kennedy pathway-related genes, utilization of both pyruvate dehydrogenase complex (PDHC) and PDHC bypass pathway, and enhancement of genes involved in tricarboxylic acid (TCA) cycle (Li et al. 2014). Multi-omics analysis of Chromochloris zofingiensis under nitrogen deprivation not only clarified the oleaginous mechanism but also revealed some key pathway-related genes, which provided targets for further genetic modification .
Although P. purpureum is a non-model organism, its genetic background is relatively clear, and its genome information has been published (Bhattacharya et al. 2013). Studies have shown that bacteria-derived plasmids could stably and freely exist in P. purpureum, and the expression level of foreign genes was as high as 5% of the total soluble protein (TSP) due to the high copy property of the plasmids, which not only provided a new method for genetic engineering research of P. purpureum, but also explored its potential as a bioreactor for the production of high bioactive substances (Li and Bock 2018). The stable nuclear transformation system of P. purpureum has also been established, which further enriched its genetic engineering research (Prasad et al. 2019).
However, due to the lack of suitable gene targets, culture optimization remains the major foci of current studies on P. purpureum, and its genetic engineering has not been reported until recently by a study showing that knock out of the chlorophyll synthase gene (CHS1) resulted in phycoerythrin production (Jeon et al. 2021). In the early stage, our laboratory conducted culture optimization and carbon-to-nitrogen ratio regulation on P. purpureum, and assessed differences in the physiological and biochemical components of P. purpureum under nitrogen-deficient conditions but did not elucidate its metabolic mechanism at the molecular level (Li et al. 2020). Unlike the well-studied mechanisms of nitrogen deficiency in green algae, P. purpureum acts as a mediator of horizontal gene transfer between prokaryotes and photosynthetic eukaryotes (Bhattacharya et al. 2013), which makes it significant to study the response mechanism of P. purpureum to nitrogen deprivation. In this study, we performed transcriptome analysis of P. purpureum under nitrogen deprivation for the first time. Synthesis pathways of polysaccharides and PUFAs and the metabolic mechanism underlying nitrogen deficiency were elucidated, and several key pathway-related genes were identified, providing targets for further rational genetic engineering. This study further lays a foundation for the accumulation of high-value bioactive substances from P. purpureum as a "green factory".

Strains and growth conditions
P. purpureum FACHB-806 was purchased from the Freshwater Algae Culture Collection at the Institute of Hydrobiology. The ASW medium previously optimized by our laboratory was used for culture (Li et al. 2020), and the culture temperature, rotation speed, and illumination conditions were 25 °C, 120 rpm, and 50 μmol/ m 2 /s, respectively.

Nitrogen deficiency experimental design
The KNO3 concentration in nitrogen-deficient ASW medium was 0.075 g/L (1.000 g/L in normal ASW medium), and other components were the same as those in normal ASW medium. Seed culture was conducted using a 1-L flask with 500 mL normal ASW medium. The algal cells in logarithmic phase were collected by centrifugation at 4000 rpm for 10 min, washed with sterile deionized water, and then resuspended in nitrogen-deficient and normal ASW medium. Approximately 500 mL of mixed algal cells was transferred to a vertical columnar photoreactor (6 × 60 cm) with an initial inoculation density of 2.0 × 10 6 cells/mL, and the culture temperature, aeration rate (sterile air), and illumination conditions were 25 °C, 1 vvm, and 50 μmol/m 2 /s, respectively.
Samples were taken every 2.5 days and 5 days for growth characteristics and biochemical components measurement, respectively. Algal cells were collected at 5 and 15 days for transcriptome analysis. Cells from day 12 were collected for scanning electron microscopy (SEM) and transmission electron microscopy (TEM).

Physiological and biochemical analysis
For growth characteristics, cell density was measured by a hemocytometer, and dry cell weight was calculated as previous reported (Li et al. 2020). For chlorophyll a concentration determination, 2 mL algal cells (1.0 × 10 7 cells/ mL) were centrifuged, washed with sterile deionized water, and finally soaked in 90% acetone at 4 °C for 24 h in the dark. Then, the supernatant was collected (4 °C, 5000 rpm for 10 min), and the absorbance was determined with a spectrophotometer at a wavelength of 665 nm. Chlorophyll a concentration was further calculated as described elsewhere (Jia 2006). The OJIP test was measured with a portable fluorometer AquaPen-C 100 (Photon System Instrument, Czech) using the FluoPen 1.0 software to access the data . Two milliliters of algal cells (1.0 × 10 7 cells/mL) were collected and adapted to the dark for 20 min before measurement. For SEM and TEM (HITACHI SU9010, Japan), cell pretreatment and section preparation were performed according to standard procedures (Arsiya et al. 2017). Total soluble protein, carbohydrate, total lipid, polyunsaturated fatty acid, phycoerythrin, and exopolysaccharide content were determined as previously described (Li et al. 2020).

RNA sequencing and data analysis
RNA samples from four growth phases (8 samples) were submitted to Beijing Biomarker Technologies Corporation for Illumina sequencing. The mRNA-seq paired-end sequencing data are available at the NCBI's Sequence Read Archive database with an accession number PRJNA682303. Clean data (clean reads) were obtained by removing reads containing adapter, reads containing poly-Ns, and low-quality reads from raw data. At the same time, Q20, Q30, and GC content of the clean data were calculated. Clean reads were then mapped to the P. purpureum genome (GenBank Assembly Accession: GCA_008690995.1) using HISAT2 tools software (http:// ccb. jhu. edu/ softw are/ hisat2/ index. shtml) . The mapped reads were spliced using StringTie software (https:// ccb. jhu. edu/ softw are/ strin gtie/ index. shtml) and compared with the reference genome annotation to discover new genes (Pertea et al. 2015). DEseq (http:// www. bioco nduct or. org/ packa ges/ relea se/ bioc/ html/ DESeq. html) was used for differential expression analysis (Wang et al. 2010), and the false discovery rate (FDR) < 0.05 and |Log 2 FC| ≥ 1 were set as the threshold for significant differential expression. Gene function was annotated based on the Nr, Nt, Pfam, KOG/COG, Swiss-Prot, KO, and the GO databases using BLAST (http:// blast. ncbi. nlm. nih. gov/ Blast. cgi). Gene Ontology (GO) enrichment analysis of DEGs was implemented by the GOseq R packages based Wallenius non-central hyper-geometric distribution (Young et al. 2010). KOBAS software was used to test the statistical enrichment of differential expression genes in KEGG pathways (Mao et al. 2005). Carbohydrate-active enzyme prediction was conducted using CAZY (http:// www. cazy. org/).

Statistical analysis
Each experiment and each assay had three biological replicates and two technical replicates, respectively. The data were expressed as the mean and standard deviation (SD). For mRNA-sequencing, two separate biological replicates were analyzed in parallel. Student's t test was used as implemented in the IBM SPSS Statistic 24 software. For all the data analysis, differences with a P value < 0.05 were considered statistically significant.

Results and discussion
Under nitrogen stress, the color of P. purpureum turned yellow, photosynthesis decreased, the structure of phycobilisome was damaged, but the thickness of the extracellular viscousness sheath increased Owing to fluorescent phycoerythrin, P. purpureum exhibits a vivid red color under normal conditions, but when cultured under nitrogen deprivation conditions, the algae turned yellow by the fifth day ( Fig. 1a), which is consistent with chlorosis in green algae under nitrogen deficiency (Arsiya et al. 2017;Li et al. 2020). Images from an optical microscope further confirmed the color changes of the single cell. P. purpureum under normal conditions rapidly grew, and cell density peaked at 3.46 ± 0.31 × 10 7 cells/mL on day 20, but cells grew slowly under nitrogen deprivation conditions and stopped growing on day 5, reaching a cell density of 1.07 ± 0.10 × 10 7 cells/mL (Fig. 1a).
Photosynthetic system (PS) plays an irreplaceable role in the development of microalgae. Fv/Fm, the maximum and effective quantum yields of PSII electron transport, showed a relatively flat trend under normal conditions, but significantly decreased under nitrogen deprivation conditions in P. purpureum (Additional file 1: Figure S1), as has been reported in other microalgae (Schmollinger et al. 2014). To further investigate the photosynthetic response of nitrogen deprivation, we ran the OJIP test on the cells under two different conditions on day 20. The results showed that the total energy absorption, trapping, and transfer (Fv/Fm, ψ0, and φE0, respectively) were much lower in ND cells (cells under nitrogen deprivation), leading to a drop in PI (Fig. 1b), which is an overall evaluation of photosynthetic efficiency (Aksmann and Tukaj 2008). However, the specific energy fluxes for absorption, trapping, and electron transport per reaction center (RC), namely, TR0/RC, ET0/RC, and ABS/ RC, respectively, were higher in ND cells (Fig. 1b), indicating bigger size and stronger photosynthetic efficiency of individual RC (Aksmann and Tukaj 2008). This may be explained by the lower number of RCs in the cells under ND, which coincides with a decrease in the number of phycobilisome units observed by TEM (Fig. 1c).
Because exopolysaccharides (EPS) and phycobilisome are barely visible under the optical microscope, SEM and TEM were subsequently adopted (Fig. 1c). Both SEM and TEM showed that cells under ND tended to secrete more exopolysaccharides outside the cell, thereby resulting in the accumulation of cell wall polysaccharides (viscousness sheath), which facilitated intercellular contact. There was an intact star-shaped chromophore (chloroplast)  Fig. 1 Growth characteristics, photosystem parameters and morphological changes of P. purpureum under normal and nitrogen-deficient conditions. a Growth curves and color changes of P. purpureum under normal and nitrogen-deficient conditions, and individual cell status under optical microscope at day 20. b Spider-plot presentation of selected parameters quantifying the behavior of PS II of P. purpureum cells exposed for 20 days to nitrogen stress. c Scanning electron microscope and transmission electron microscope images of normal and nitrogen-deficient cells on day 12. ND means nitrogen deprivation. Data are expressed as mean ± SD (n = 3) within the CK cells (cells under normal conditions) of P. purpureum, but the chromophore under ND was much smaller, just about a third of the size of the CK cells. However, cell size did not significantly change in either conditions. Intracellular polysaccharides aggregated in the form of 'starch granules' , which significantly accumulated in ND cells. We found the suspected phycobilisome structure inside the chromophore of CK cells, consistent to an earlier report (Ma et al. 2020), but was almost invisible in ND cells. Therefore, the lack of nitrogen not only caused shrinkage of the chromophores, but also reduced the content and stability of phycobilisomes in P. purpureum.
Taken together, nitrogen deficiency impaired the structure of chromophores and phycobilisomes, leading to a decrease in photosynthetic efficiency and further inhibition of P. purpureum growth. However, thickening of the extracellular viscousness sheath may be a protective mechanism under stress.

Under nitrogen deficiency, the total soluble protein content decreased significantly, while the polysaccharide and lipid content increased
In general, nitrogen-containing substances are recycled in the absence of nitrogen to produce essential proteins such as key enzymes (nitrogen transporters or nitrate reductase) or to provide molecular building blocks for cell survival, with an overall reduction in protein content (Corteggiani Carpinelli et al. 2014;Levitan et al. 2015). Under normal conditions, the intracellular TSP content of P. purpureum remained stable, but during the 5 days of nitrogen deficiency, the intracellular TSP content sharply dropped and finally reached equilibrium. On the 20th day, the intracellular TSP of the control group and the nitrogen deficiency group was 0.24 ± 0.04 mg/10 7 cells and 0.02 ± 0.01 mg/10 7 cells, respectively (Fig. 2a). Not surprisingly, nitrogen deficiency led to a decrease in the nitrogen-containing B-phycoerythrin and chlorophyll a content, which was most pronounced in the early stages of nitrogen deficiency ( Fig. 2b; Additional file 1: Figure  S2). Chlorophyll a is the key pigment in photosynthesis of green algae and higher plants, while phycobilisome plays a similar role in cyanobacteria and red algae (Adir et al. 2020;Ma et al., 2020), and phycoerythrin content in P. purpureum is much higher than chlorophyll a ( Fig. 2b; Additional file 1: Figure S2), which may play a leading role in the capture and transfer of light energy. Reductions in B-phycoerythrin and chlorophyll a in P. purpureum caused by nitrogen deficiency may be the reason for the decrease of photosynthetic efficiency.
Despite the observed growth inhibition, large amounts of cell wall polysaccharides were synthesized by nitrogendeficient cells of P. purpureum, which was considered a protective mechanism under stress (Arad et al. 1988). In this study, we observed that the carbohydrate content in both the control group and the nitrogen deficiency group decreased in the early stage ( Fig. 2c), possibly because the cells were forced to secrete polysaccharides to adapt to the new environment (Fig. 2d). After the adaptation period (day 5), the carbohydrate content in the control group increased with growth, while that in the nitrogen-deficient group remained stable (Fig. 2c). Exopolysaccharide content increased under both conditions, but the accumulation rate was faster in the nitrogen-deficient group, reaching 0.14 ± 0.01 mg/10 7 cells on day 20, which was 2.88 times that of the control group (Fig. 2d). Interestingly, the high yield of polysaccharides under normal conditions appears to be a combination of increased cell density and increased polysaccharide content in a single cell (Figs. 1a, 2d), which differs from B-phycoerythrin accumulation. Exopolysaccharide yield in the two groups was similar (166.90 ± 5.40 mg/L in the control group; 137.25 ± 7.19 mg/L in the nitrogen deficiency group), but the nitrogen deficiency group had fewer cells (Figs. 1a, 2d), which was more conducive to downstream separation and purification. In terms of total polysaccharides, P. purpureum tended to secrete more polysaccharides under nitrogen-deficient conditions.
Nitrogen stress can promote lipid accumulation, particularly TAG, which has been reported in many microalgae (Li et al. 2014;Liu et al. 2019). Compared with the control group, the total lipid (TL) content of P. purpureum in the nitrogen-deficient group increased by 65.2%, reaching as high as 0.32 ± 0.04 mg/10 7 cells on the 20th day (Fig. 2e). The fatty acids with the highest content mainly included palmitic acid (PA), linolenic acid (LA), and ARA (Fig. 2f ). Under nitrogen stress, the contents of total fatty acids (TFA) and individual fatty acids increased, except for eicosadienoic acid (C20:2), which decreased by 53% (Fig. 2f ). Among these, the contents of PA and eicosatrienoic acid (C20:3) most significantly changed, which increased by 236% and 232%, respectively (Fig. 2f ). We hypothesized that the increase in C20:3 might be related to the decrease in C20:2, as the substrate C18:2 shared by the two seems to flow more to C20:3 through the n-6 pathway. Although ARA content increased by 55%, EPA content increased by 123%, resulting in a decrease in the ARA to EPA ratio (ARA/EPA) from 4.97 to 3.50 ( Fig. 2f; Table 1). Normally, PUFAs account for the highest proportion of total fatty acids in P. purpureum, while SFA and MUFAs accounted for the highest proportion in the nitrogen deficiency group (Table 1), indicating that although overall fatty acid content increased, unsaturation of total fatty acids decreased, which was concordant to the findings of previous reports (Msanne et al. 2012).

De novo transcriptome assembly and annotation and global gene expression analysis
To further elucidate the molecular mechanism by which P. purpureum responds to nitrogen deficiency, we performed time-resolved comparative transcriptome analysis, in which cells from normal and nitrogen deprivation conditions on days 5 and 15 were collected for RNA-seq analysis. Eight samples (two biological replicates per case) were used for transcriptome assembly, and a total of 80.06 Gb of clean data were obtained. The clean data of each sample reached 7.90 Gb, and the percentage of Q30 was at least 93.67%. The clean reads of each sample were mapped to the reference genome, with alignment efficiency ranging from 87 to 94% (Additional file 1: Table S1).
The annotation of gene function in each database was determined (Additional file 1: Figure S3; Additional file 2: Data S1). In terms of COG function classification, the most significant ones were "basic function prediction", "ribosome regeneration", and "signal transduction" (Additional file 1: Figure S4). In the "molecular function" category, the top four GO terms were "catalytic activity", "binding", "structural molecule activity", and "transporter activity" (Additional file 1: Figure  S5). For KEGG enrichment analysis, the number of genes related to "protein processing in the endoplasmic reticulum" ranked first (Additional file 1: Figure S6). In addition, we annotated 21 transcription factor (TF) families (183 genes), 18 transcription regulator (TR) families (94 genes), and 48 protein kinase (PK) families (202 genes). Among the predicted transcription factors, the number of C2H2 and MYB-related families was the highest, containing 66 and 32 genes, respectively. The GANT (23 genes) and SNF2 (17 genes) families were the predominant transcription regulatory factors. In terms of protein kinases, the CAMK-CDPK family (60 gens) was by far the most abundant (Additional file 3: Data S2). Studies have shown that the TF NIT2, which regulates nitrate metabolism, is upregulated sixfold under nitrogen-deficient conditions (Miller et al. 2010).
Differentially expressed genes (DEGs) are the main foci of transcriptome analysis. In this study, we adopted a comparative transcriptome approach to define the DEGs by comparing the gene expression between the nitrogen-deficient group and the control group at a specific time point (days 5 and 15). We further analyzed the enrichment of DEGs in the COG, GO, and KEGG pathways (Additional file 1: Figures S7-11), and the results showed that in the COG function classification, DEGs on day 5 (R5_vs_Y5) mainly enriched the categories of "general function prediction only", "translation, ribosomal structure and biogenesis", "posttranslational modification, protein turnover, chaperones", and "signal transduction mechanisms" (Fig. 3a), which suggested the complex response mechanism of P. purpureum under the conditions of nitrogen deficiency. In addition, we analyzed gene expression levels of the annotated TRs, TFs, and PKs that have regulatory or signal transduction functions (Fig. 3b), and the highest number of DEGs was observed on the 15th day, indicating that P. purpureum had a more complex and urgent metabolic regulatory mechanism under nitrogen stress. Compared with normal conditions, most of the DEGs encoding TFs, TRs, and PKs were upregulated under nitrogen-deficient conditions (R5_vs_Y5, R15_vs_Y15), which was different from the expression of these genes during growth (R5_vs_R15, Y5_vs_Y15).

Active chemical energy promotes metabolic migration
Consistent with previous observations in various microalgae (Miller et al. 2010;Sun et al. 2013), most photosynthesis-related proteins were significantly downregulated at prophase (day 5) under nitrogen deficiency. A total of 5 ATP synthase-encoding genes were annotated, all of which were downregulated on the 5th day, while two genes were upregulated on the 15th day, indicating the important role of these two genes in photosynthetic ATP production under nitrogen deficiency (Additional file 1: Table S2).
A total of 70 genes were annotated to oxidative phosphorylation, among which 30 genes were differentially expressed, including 25 upregulated genes and 5 downregulated genes (Additional file 1: Table S3). In addition, most of the ATP synthase-encoding genes were upregulated (Additional file 1: Table S3). The overall upregulation of oxidative phosphorylation-and ATP synthase-encoding genes indicates that the production of ATP under nitrogen-deficient conditions is mainly due to oxidative phosphorylation. It has been reported that oxidative phosphorylation is the main source of ATP under nitrogen stress (Schmollinger et al. 2014).

Table 1 Fatty acids composition of P. purpureum under normal and nitrogen deprivation conditions
Data are expressed as mean ± SD (n = 3) SFA saturated fatty acid; MUFA monounsaturated fatty acid; PUFA polyunsaturated fatty acid; UFA unsaturated fatty acid; TFA total fatty acid; ARA arachidonic acid; EPA eicosapentaenoic acid a Significant difference (t test, p < 0.05) b Significant difference (t test, p < 0.01) c Significant difference (t test, p < 0.001)

SFA (mg/10 9 cells)
MUFA (mg/10 9 cells) PUFA (mg/10 9 cells) UFA (mg/10 9 cells) TFA (mg/10 9 cells) ARA/EPA In the pentose phosphate pathway, GPD (glucose-6-phosphate 1-dehydrogenase) and PGL (6-phosphogluconolactonase) catalyze the conversion of G-6-P to gluconate-6P, which generates ribulose-5P under the catalysis of PGD (6-phosphogluconate dehydrogenase). These two steps are accompanied by the production of reducing power NADPH. The upregulation of the pentose phosphate pathway under nitrogen deficiency to provide reducing power for free fatty acid synthesis has been reported in microalgae (Jaeger et al. 2017;Liu et al. 2019). Two GDP genes and two PGD genes were annotated in P. purpureum that showed varying degrees of upregulation This image shows central carbon metabolic pathways, including the Calvin cycle, gluconeogenesis, glycolysis, pentose phosphate pathway, and the tricarboxylic acid cycle. The red plus sign indicates upregulation, the blue minus sign indicates downregulation, and the black sign indicates no significant difference. The front and back positions indicate the differential expression at day 5 and day 15, respectively. The regions in red and blue indicate the upregulation and downregulation of the overall pathway, respectively. Abbreviations for metabolites were shown in black. PEP, phosphoenolpyruvate; OAA, oxaloacetate; MAL, malate; Pyr, pyruvate; 3PG, 3-phosphoglycerate; 1,3-2PG, 3-phospho-glycerol phosphate; G-3-P, glyceraldehyde 3-phosphate; F-1,6-2P, fructose 1,6-bisphosphate; F-6-P, fructose 6-phosphate; G-6-P, glucose 6-phosphate; G-1-P, glucose under nitrogen deprivation ( Fig. 4; Additional file 1: Table S4). The expression of one GPD gene increased by nearly 89-fold (Log 2 FC = 6.49), and another PGD gene was more than 500-fold upregulated (Log 2 FC = 8.98), indicating the production of large amounts of NADPH in the pentose phosphate pathway ( Fig. 4; Additional file 1: Table S4). In addition, not only the abovementioned ME (nearly four times upregulated) plays a role in carbon sequestration, but its catalytic process requires NADP + as the substrate to generate NADPH. The upregulation of ME provides reducing power for cell metabolism, which has long been reported in other microalgae (Jaeger et al. 2017).

Carbon flow leads to the synthesis of polysaccharides and lipids in the prophase and anaphase, respectively
Among the 31 genes noted in the KEGG pathway for carbon fixation in photosynthetic organisms, 14 DEGs and 10 downregulated genes were found ( Fig. 4; Additional file 1: Table S5). However, among the 14 DEGs, 4 genes were upregulated under nitrogen deficiency, including a glutamate glyoxylate aminotransferase-encoding gene GGAT , a phosphoenolpyruvate carboxylase-encoding gene PEPC, and two malic enzyme-encoding genes (ME) ( Fig. 4; Additional file 1: Table S5). These are key genes in the C4 pathway, which are significantly upregulated in the process of nitrogen deficiency, suggesting that there may be a C4 pathway in P. purpureum, which plays a role in CO 2 supplementation under nitrogen deprivation. Zhao et al. reported that four key genes in the C4 pathway were significantly upregulated (> 10-fold) under nitrogen-deficient conditions, suggesting that both C3 and C4 pathways exist in Haematococcus pluvialis . Phosphoenolpyruvate carboxykinase-encoding gene PPCK was significantly upregulated under nitrogendeficient conditions on days 5 and 15, but the fructose-1,6-bisphosphatase-encoding gene FBP, another key enzyme in the gluconeogenic pathway, was significantly downregulated on day 15 (Log 2 FC < − 1.78), suggesting that gluconeogenesis is mainly upregulated at prophase ( Fig. 4; Additional file 1: Table S6). In addition, during the conversion of hexose to disaccharide, we found that UGP (UTP-glucose-1-phosphate uridylyltransferase encoding gene) and TPS/TP (trehalose 6-phosphate synthase/phosphatase encoding genes) were upregulated on day 5 but downregulated on day 15 ( Fig. 4; Additional file 1: Table S6). In response, the expression of the α,αtrehalase (TL) gene did not significantly change on day 15 from downregulation on day 5 ( Fig. 4; Additional file 1: Table S6). The upregulation of gluconeogenesis and the downregulation of polysaccharide decomposition on day 5 indicate that polysaccharide synthesis is enhanced at the early stage, which was concordant to the observed increase in polysaccharide content.
There are three rate-limiting enzymes in glycolysis, namely, hexokinase (HK), 6-phospho-fructose-kinase (PFK), and pyruvate kinase (PK). Differential expression of two HK and PFK genes revealed opposite directions on the 15th day, but the upregulated extent was higher than the downregulated one ( Fig. 4; Additional file 1: Table S6). Therefore, we believe that both HK and PFK were upregulated on the 15th day. However, one of the three genes encoding PK showed significant downregulation at day 15 ( Fig. 4; Additional file 1: Table S6). In addition, the downregulation of PGK (phosphoglycerate kinase) on day 15, together with the upregulation of PGM (phosphoglycerate mutase) ( Fig. 4; Additional file 1: Table S6), indicating that Calvin cycle-derived 3PG (3-phosphoglycerate) is directed towards glycolysis than gluconeogenesis, which has also been reported in Monoraphidium neglectum (Jaeger et al. 2017). Enhancement of polysaccharides catabolism and upregulation of glycolysis pathway genes indicate that glycolysis is accelerated at the late stage, resulting in the enhancement of pyruvate to acetyl-CoA synthesis.
The upregulation of the pyruvate dehydrogenase complex (PDHC) and PDHC bypass pathway under nitrogen-depleted conditions has been previously reported in the oleaginous microalga Nannochloropsis oceanica IMET1 (Li et al. 2014). Transcriptome analysis showed that although aldehyde dehydrogenase-encoding gene ALDH (Log2FC = − 1.23) was downregulated under nitrogen deprivation, pyruvate decarboxylase-encoding gene PDC (Log 2 FC = 2.78) was significantly upregulated; hence, we believe that the PDHC bypass pathway is generally upregulated (Fig. 4; Additional file 1: Table S6). Furthermore, PDHC was downregulated on the 5th day and upregulated on the 15th day ( Fig. 4; Additional file 1: Table S6). Therefore, we hypothesized that in the early stage of nitrogen deficiency, the generation of acetyl-CoA occurs mainly through the PDHC bypass pathway, and in the later stage, the PDHC and PDHC bypass pathways work together to accelerate the generation of acetyl-CoA and provide precursors for fatty acid synthesis.
There were no significant changes in most TCA cyclerelated genes, except for two fumarate hydratase-encoding genes (FUM), suggesting that pyruvate-derived acetyl-CoA is directed towards free fatty acid synthesis than into the TCA cycle for complete oxidation ( Fig. 4; Additional file 1: Table S7). Nevertheless, the carbon skeleton produced by protein degradation may enter the TCA cycle, increasing the TCA cycle from the substrate level and promoting oxidative phosphorylation (Additional file 4: Data S3). Most of the genes related to ubiquitin-mediated proteolysis were upregulated under nitrogen deprivation conditions, which indicated enhanced proteolysis (Additional file 1: Table S8). In addition, the genes related to aminoacyl-tRNA synthesis were downregulated, while the ribosomal protein-coding genes were upregulated (Additional file 1: Table S9). The downregulation of genes related to aminoacyl-tRNA synthesis indicates a decrease in aminoacyl-tRNAs; thus, more amino acids from protein degradation may enter the TCA cycle to provide a carbon skeleton for cell metabolism. We believe that ribosomal protein-encoding genes are upregulated because cells need to synthesize new proteins that play a key role in nitrogen stress such as nitrate reductase and nitrate transporter, which help cells absorb nitrogen sources from outside the cell (Additional file 1: Table S10).
Although the genes of de novo fatty acid synthesis pathway were downregulated, the increased content of acetyl-CoA, a precursor of lipid synthesis, derived from glycolysis pathway, and the downregulation of fatty acid β oxidation pathway could explain the lipid accumulation in P. purpureum during the late period of nitrogen deficiency. Besides, the downregulation of de novo fatty acid synthesis has also occurred in other microalgae, and the relatively plausible explanation is that gene expression in the de novo synthesis of fatty acids is normally excessive (Jaeger et al. 2017;Li et al. 2014;Liang et al. 2012Liang et al. ,2019Msanne et al. 2012).

Proposed pathways for polysaccharide biosynthesis in P. purpureum
Unlike brown algae polysaccharides, which are composed of fucose or mannose (Ficko-Blean et al. 2015;Skriptsova 2015), the structure of P. purpureum polysaccharides is much more complex. Gloaguen et al. demonstrated by nuclear magnetic resonance (NMR) spectroscopy that the polysaccharides of P. purpureum are composed of three neutral monosaccharides (Xyl, Glc, and Gal) and one uronic acid (GlcA) (Gloaguen et al. 2004).
To date, only a few studies on the synthesis pathway of polysaccharides from microalgae have been conducted. Here, from the perspective of amino and nucleotide sugars and carbohydrate-active enzymes (CAZymes), we mainly focused on the interconversion between glycosyldonors and the differential expression of glycosyltransferase-encoding genes to elaborate the synthesis pathway of polysaccharides in P. purpureum. We believe that the synthesis of polysaccharides in P. purpureum begins with the synthesis of disaccharides (trehalose), and then the glycosyl donors are constantly connected to the repeated units with the catalysis of glycosyltransferase to form polysaccharides. According to the results of transcriptome annotation, there are nearly 10 different glycosyl donors in P. purpureum (Fig. 5). In addition to the previously reported monomer structure, there are also Man and GlcNAc (Fig. 5). We hypothesize that there may be other monosaccharides in P. purpureum polysaccharides, but the content is relatively low and undetectable. Among the 21 genes annotated by the transcriptome that related to glycosyl-donor interconversion, 12 genes were significantly differentially expressed under nitrogen deprivation conditions, and almost all these DEGs are closely related to the synthesis of GDP-Man and UDP-GlcNAc, suggesting that P. purpureum may regulate the composition and structure of polysaccharides to cope with nitrogen stress (Fig. 5).
As for the CAZymes, a total of 119 genes were annotated in the transcriptome, among which the number of genes encoding glycoside transferases (GTs) and glycoside hydrolases (GHs) was the largest (Table 2). This is consistent with the 83 GTs annotated in the genome information, which also indirectly reflects the complex polysaccharide structure in P. purpureum (Bhattacharya et al. 2013). For GHs, the number of significantly downregulated genes was higher than the number of upregulated genes, indicating that the decomposition of polysaccharides in P. purpureum was inhibited, which corresponds to the increase of its content (Table 2; Fig. 2d). As for the quantitative differences of carbohydrate-binding modules (CBMs) in the genome and transcriptome, we believe that some genes with inaccurate predictions at that time have been endowed with other functions, resulting in a decrease in the number of CBMrelated genes (Table 2).
Based on the transcriptome annotation information and GC-MS analysis of the structure of free fatty acids, we are more inclined to believe that the synthesis of ARA and EPA in P. purpureum involves the n-6 synthesis pathway ( Fig. 6; Table 3; Additional file 1: Figures S12-14).