Molecular adaptations to phosphorus deprivation and comparison with nitrogen deprivation responses in the diatom Phaeodactylum tricornutum

Abstract

Phosphorus, an essential element for all living organisms, is a limiting nutrient in many regions of the ocean due to its fast recycling. Changes in phosphate (Pi) availability in aquatic systems affect diatom growth and productivity. We investigated the early adaptive mechanisms in the marine diatom Phaeodactylum tricornutum to P deprivation using a combination of transcriptomics, metabolomics, physiological and biochemical experiments. Our analysis revealed strong induction of gene expression for proteins involved in phosphate acquisition and scavenging, and down-regulation of processes such as photosynthesis, nitrogen assimilation and nucleic acid and ribosome biosynthesis. P deprivation resulted in alterations of carbon allocation through the induction of the pentose phosphate pathway and cytosolic gluconeogenesis, along with repression of the Calvin cycle. Reorganization of cellular lipids was indicated by coordinated induced expression of phospholipases, sulfolipid biosynthesis enzymes and a putative betaine lipid biosynthesis enzyme. A comparative analysis of nitrogen- and phosphorus-deprived P. tricornutum revealed both common and distinct regulation patterns in response to phosphate and nitrate stress. Regulation of central carbon metabolism and amino acid metabolism was similar, whereas unique responses were found in nitrogen assimilation and phosphorus scavenging in nitrogen-deprived and phosphorus-deprived cells, respectively.

Fig 1. Growth curve and efficiency of PSII in +P and -P cells.

(A) Growth of P. tricornutum in +P (36 μM phosphate) and -P (phosphate-free) cultures. (B) Efficiency of PSII (F_V/F_m) in +P and -P cultures from day zero (start of the experiment) to day 3. Values are mean ± 1 SD of four replicates.

Fig 2. Accumulation of neutral lipids during P deprivation.

(A) Fluorescence intensity in P. tricornutum cells from +P and -P cultures stained with BODIPY 505/515 at 48 h and 72 h. Fluorescence was measured in 20–30 randomly selected cells using confocal microscopy. Significant differences (**, P < 0.01) between +P and -P cultures are indicated. au, arbitrary units. (B) Z-stack projections of P. tricornutum in +P and -P cultures at 48 h and 72 h. Bar, 5 μm.

Fig 3. GO analysis of significantly regulated genes after 72 h of P deprivation.

The dataset was divided into (A) up- and (B) down-regulated genes and analysed for process GO terms. The 16 most frequent GO terms were listed, and the rest were combined into “others”. The number in the “others” section indicates the percentage of hits within this category. The total number of GO term hits is listed below the diagram.

Fig 4. Venn diagram of up- and down-regulated genes from Yang et al. [34] and two microarray datasets (48 h and 72 h) in P-deprived P. tricornutum.

(A) Up-regulated genes. (B) Down-regulated genes. Numbers indicate genes that have shared or unique regulation among the three datasets. Genes similarly regulated in this study and Yang et al. [34] are indicated in bold.

Fig 5. Changes in carbon metabolism in -P cells.

Coloured squares indicate the regulation pattern of genes encoding putative enzymes involved in central carbon metabolism after 48 h (left square) and 72 h (right square) of P deprivation, compared with P-replete cultures. Squares with a diagonal line inside indicate no significant difference in expression (P > 0.01). The scale on the right represents gene expression ratio values, log2 transformed. Numbers indicate Phatr2 gene IDs. Metabolites detected are underlined. Red, blue and black text indicates up-, down-, and no regulation of pathways or metabolites, respectively. Red arrow indicates allosteric activation. 2-OG, 2-oxoglutarate; 2-PGA, 2-phosphoglycerate; CISY, citrate synthase; DHAP, dihydroxyacetone phosphate; FBA, fructose-1,6-bisphosphate aldolase; FBP, fructose-1,5-bisphosphatase; FUM, fumarate hydratase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GLK, glukokinase; GPI, glucose phosphate isomerase; ICL, isocitrate lyase; IDH, isocitrate dehydrogenase; MDH, malate dehydrogenase; ME, malic enzyme; MS, malate synthase; OAA, oxaloacetic acid; OGD, 2-oxoglutarate dehydrogenase; P2FK/F2BP, 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase; PDHC, pyruvate dehydrogenase complex; PEP, phosphoenolpyruvate; PEPCK, PEP carboxykinase; PGK, phosphoglycerate kinase; PK, pyruvate kinase; PPDK, pyruvate orthophosphate dikinase; PYC, pyruvate carboxylase; SBP, sedoheptulose bisphosphatase; SCS, succinyl-CoA synthase; SDH, succinate dehydrogenase; TPI, triosephosphate isomerase; UGP/PGM, UDP-glucose pyrophosphorylase/phosphoglucomutase. Annotation for glycolysis/gluconeogenesis enzymes is taken from [52], other annotation is taken from [51].

Fig 6. Lipid remodelling in -P cells.

Putative enzymes functioning in phospholipid degradation is shown to the right, the triacylglycerol biosynthetic pathway is shown in the middle and non-phosphorus-containing lipid biosynthesis to the left. Coloured squares indicate the regulation pattern of genes after 48 h (left square) and 72 h (right square) of P deprivation, compared with P-replete cultures. Squares with a diagonal line inside indicate no significant difference in expression (P > 0.01). The scale on the right represents gene expression ratio values, log2 transformed. Numbers indicate Phatr2 gene IDs. Metabolites detected are underlined. Red, blue and black text indicates up-, down-, and no regulation of metabolites. The dashed arrow indicates an unidentified enzyme. GPDE, glycerophosphoryldiester phosphodiesterase;; LAH, lipid acyl hydrolase; PLA, phospholipase A; PLC, phospholipase C; PLD, phospholipase D;; UGP/PGM, UDP-glucose pyrophosphorylase/phosphoglucomutase.

Fig 7. Comparison of the effect of nitrate and phosphate deprivation on fatty acids, glycerides and organic acids.

Hierarchical cluster analysis based on mean ratio values, log2 transformed (deprived vs. replete condition at 48 and 72 h; n = 4) from the current P deprivation experiment and a previously published N deprivation experiment [25]. Treatments are depicted in columns while metabolites are represented by rows. Heat map visualization displays differences in metabolite profiles; bluish colours indicate decreased metabolite concentration in deprived compared to the corresponding replete condition, and reddish colours increased metabolite levels (see colour scale).