Potential role of multiple carbon fixation pathways during lipid accumulation in Phaeodactylum tricornutum

Abstract

Background: Phaeodactylum tricornutum is a unicellular diatom in the class Bacillariophyceae. The full genome has been sequenced (<30 Mb), and approximately 20 to 30% triacylglyceride (TAG) accumulation on a dry cell basis has been reported under different growth conditions. To elucidate P. tricornutum gene expression profiles during nutrient-deprivation and lipid-accumulation, cell cultures were grown with a nitrate to phosphate ratio of 20:1 (N:P) and whole-genome transcripts were monitored over time via RNA-sequence determination.

Results: The specific Nile Red (NR) fluorescence (NR fluorescence per cell) increased over time; however, the increase in NR fluorescence was initiated before external nitrate was completely exhausted. Exogenous phosphate was depleted before nitrate, and these results indicated that the depletion of exogenous phosphate might be an early trigger for lipid accumulation that is magnified upon nitrate depletion. As expected, many of the genes associated with nitrate and phosphate utilization were up-expressed. The diatom-specific cyclins cyc7 and cyc10 were down-expressed during the nutrient-deplete state, and cyclin B1 was up-expressed during lipid-accumulation after growth cessation. While many of the genes associated with the C3 pathway for photosynthetic carbon reduction were not significantly altered, genes involved in a putative C4 pathway for photosynthetic carbon assimilation were up-expressed as the cells depleted nitrate, phosphate, and exogenous dissolved inorganic carbon (DIC) levels. P. tricornutum has multiple, putative carbonic anhydrases, but only two were significantly up-expressed (2-fold and 4-fold) at the last time point when exogenous DIC levels had increased after the cessation of growth. Alternative pathways that could utilize HCO3- were also suggested by the gene expression profiles (e.g., putative propionyl-CoA and methylmalonyl-CoA decarboxylases).

Conclusions: The results indicate that P. tricornutum continued carbon dioxide reduction when population growth was arrested and different carbon-concentrating mechanisms were used dependent upon exogenous DIC levels. Based upon overall low gene expression levels for fatty acid synthesis, the results also suggest that the build-up of precursors to the acetyl-CoA carboxylases may play a more significant role in TAG synthesis rather than the actual enzyme levels of acetyl-CoA carboxylases per se. The presented insights into the types and timing of cellular responses to inorganic carbon will help maximize photoautotrophic carbon flow to lipid accumulation.

Figure 1

Growth characterization ofP. tricornutum. Cell density growth curve of P. tricornutum cells (▲) showing depletion of exogenous nitrate (○) and phosphate (◊). Phosphate concentrations are multiplied by a factor of 10 for visualization (A). Cell density growth curve showing the depletion and rebound of dissolved inorganic carbon (∆) throughout P. tricornutum growth (B). Arrows indicate time points at which cells were harvested for RNA sequencing analysis.

Figure 2

Characterization of lipid accumulation inP. tricornutumduring increase in Nile Red fluorescence intensity (■) with respect to cell number (▲) (A). Nile Red fluorescence intensity indicating the increase in lipids is shown with the depletion of external nitrate (○) and phosphate (◊) (B). Phosphate concentrations are multiplied by a factor of 10 for scaling purposes (e.g., 0.2 mM = 0.02 mM). Arrows indicate time points at which cells were harvested for RNA sequencing.

Figure 3

Nutrient depletion and cell count growth curve ofP. tricornutum. Cell density (▲) during the depletion and rebounding of dissolved inorganic carbon (∆) and increase in Nile Red fluorescence intensity (∎). Arrows indicate time points at which cells were harvested for RNA sequencing.

Figure 4

Proposed cellular metabolic map forP. tricornutumduring nutrient depletion and initial lipid accumulation as compared to nutrient replete conditions (Q2 vs. Q1). Differences in fold change are based on log₂ scale. Color scale represents up-expressed (green) and down-expressed (red) genes. Genes are represented within organelles based on predicted protein localizations (from the literature) including probable membrane bound proteins.

Figure 5

Proposed cellular metabolic map forP. tricornutumduring extended nutrient depletion and lipid accumulation as compared to nutrient replete conditions (Q3 vs. Q1). Differences in fold change are based on log₂ scale. Color scale represents up-expressed (green) and down-expressed (red) genes. Genes are represented within organelles based on predicted protein localizations (from the literature) including probable membrane bound proteins.

Figure 6

Nitrogen metabolism gene expression ofP. tricornutumduring nutrient depletion and lipid accumulation as compared to nutrient replete conditions (Q2 vs. Q1) (A). Nitrogen metabolism gene expression during extended nutrient depletion and lipid accumulation as compared to nutrient replete conditions (Q3 vs. Q1) (B). Genes are localized to organelles based upon reported literature. Differences in fold change are based on log₂ scale and the color scale represents up-expressed (green) and down-expressed (red) genes.

Figure 7

Proposed C4 metabolism ofP. tricornutumbased on gene expression and gene localizations. Carbon-assimilation gene expression during nutrient depletion and initial lipid accumulation as compared to nutrient replete conditions (Q2 vs. Q1) (A). Carbon-assimilation gene expression during extended nutrient depletion and lipid accumulation as compared to nutrient replete conditions (Q3 vs. Q1) (B). Differences in fold change are based on log₂ scale and the color scale represents up-expressed (green) and down-expressed (red) genes. Font size is adjusted to the transcript abundances of C4 metabolism genes relative to each other.

Figure 8

Fatty acid metabolism, tricarboxylic acid cycle, and glyoxylate shunt related gene expression inP. tricornutumduring nutrient depletion and initial lipid accumulation as compared to nutrient replete conditions (Q2 vs. Q1) (A). Significant gene expression during extended nutrient depletion and lipid accumulation as compared to nutrient replete conditions (Q3 vs. Q1) (B). Fatty acid metabolism genes are denoted by presumptive roles in fatty acid biosynthesis, triacylglyceride assembly, β-oxidation, and chain modifications. Differences in fold change are based on log₂ scale and the color scale represents up-expressed (green) and down-expressed (red) genes.

Figure 9

Schematic representation of temporal biomass sampling from replicate bioreactors.