Fig 1.
Overview of sampling times in the short-term and the long-term experiment.
Overview of the sampling time points conducted during the long-term (LT) and the short-term (ST) experiment. The different treatments are indicated on the left, with the sampling times in hours (h) to the right. Orange arrows indicate the addition of iron with only the iron-recovery (rec), the actD-Fe and the dmso-Fe samples receiving 10 μM final concentration FeCl3 immediately after the initial measurement T0.
Fig 2.
Changes in cellular chlorophyll content.
Box plots (A) and short- and long-term time-courses (B) of chlorophyll. High-iron, low-iron, and iron-recovery samples are shown in black, blue, and orange, respectively. A two-tailed Student’s t-test was used for the statistical analysis of box plots (A). Box plots represent duplicate measurements of three experiments (n = 6) for each time point (0 h, 1 h, 6 h). Line graphs (B) show trends throughout the long-term and short-term experiments. Error bars represent standard errors (S.E.). Statistically significant P values are indicated as * <0.05, **, 0.01, *** < 0.001, **** < 0.0001. A line without stars indicates a test that is not statistically significant.
Fig 3.
Changes in photo-physiological parameters throughout 6 h during the short-term (ST) experiment.
High-iron, low-iron, and iron-recovery samples are shown in black, blue, and orange, respectively. The 0 h, 1 h, and 6 h sampling times are indicated as triangles, squares, and circles, respectively. (A) Electron transfer rate (ETR) in μmol e m-2s-1. (B) Bar graphs of ETRmax with Student’s t-test indicate statistically significant differences. (C) Photochemical quantum yield (PS-quantum yield Y(II)), non-photochemical quenching (Y(NPQ)), and non-regulated energy dissipation (Y(NO)). (D) Bar graphs of Y(II), Y(NPQ), and Y(NO) with Student’s t-test. (E) Light-induction curves and the corresponding Fv/Fm values with ‘Student’s t-test showing statistical differences. Statistically significant P values are indicated as * <0.05, ** < 0.01, *** < 0.001, **** < 0.0001. A line without stars indicates a test that is not statistically significant.
Fig 4.
Comparison of physiological parameters in low-iron, high-iron, and iron-recovery samples.
Characterization of cells grown under low-iron, high-iron, and iron-recovery conditions. High-iron samples are black, low-iron samples are blue, and iron-recovery samples are shown in orange. Results from ST- and the LT-experiment are combined, with every dot representing one sample. We used a two-tailed Student’s T-test for statistical analysis between high-iron and low-iron cultures. Statistically significant P values are indicated as * <0.05, ** < 0.01, *** < 0.001, **** < 0.0001. A line without stars indicates a test that is not statistically significant. (A) Growth rates μ [day-1]. (B) Chlorophyll content per cell (fg/cell). (C) Fv/Fm measurements of all timepoints measured during LT- and ST-experiment. (D) Nitrogen/cell in fmol/cell. (E) Carbon/cell in fmol/cell. (F) C:N molar ratio.
Fig 5.
Time course of change in targeted transcript counts in high-iron, low-iron, and iron-recovery samples over the 6 h and the 22 h periods.
Normalized transcript counts are on the y-axes vs time in hours on the x-axes, with 0 h representing 7 AM. Grey boxes indicate the dark phase. The response within the first hour is shown on the left, and the addition of iron is indicated with a red dashed line. Samples from the long-term experiment are shown on the right. High-iron, low-iron, and recovery samples are shown in black, blue, and orange, as in previous figures, with error bars indicating standard errors (S.E.). Transcripts are grouped: (A) iron starvation-induced proteins (ISIP), (B) proteins involved in photosynthesis, (C) Fructose-bisphosphate aldolases (FBA), (D) proteins involved in the glycosylation pathway, (E) transcription factors (HSF, Sigma70TF), iron reductase (FRE1) and a protein of unknown function (CREGx2).
Fig 6.
Overview of actinomycin D treatment on transcript counts compared to iron-induced reduction of mRNA transcripts.
(A) Changes in transcript counts over time (6 h) in iron-recovery (rec; orange), actD-treated cells (actD–red, actD-Fe–grey). All samples shown here were low-iron samples. ActD cultures received actinomycin D and actD-Fe cultures received actinomycin D and iron. Rec are low-iron samples that received only iron. (B) Table of half-lives for transcripts analyzed in the ST-experiment (in min). Values with an asterisk are excluded from the average. (C) Bar graphs of half-lives of the 7 rapidly declining iron-responsive genes. Transcripts with an asterisk revealed significant differences between actD and iron-recovery samples. (D) Bar graphs of half-lives for non-iron-responsive genes.
Fig 7.
Putative transcription factors in T. oceanica and their transcript levels.
(A) Comparison of previously published data on transcription factors in P. tricornutum and T. pseudonana [34] to T. oceanica. (B) Expression patterns of the transcription factors from high-iron (black), low-iron (blue), and iron-recovery (orange) treatments. Transcript counts are on the y-axes, and time (h) is on the x-axis. The grey box indicates the dark phase. (C) Transcript counts of high- (black) and low-iron (blue) treatments. A two-tailed Student’s t-test was used for statistical analysis. Transcript counts are on the y-axis, and the gene names are indicated on the x-axis.
Fig 8.
Normalized transcript counts for 21 targeted genes (A) and heatmap of Spearman correlation between transcript levels (B).
(A) Normalized transcript counts (counts/100ng of total RNA). Low-iron (blue), high-iron (black), and iron-recovery (rec) (orange) samples are plotted for each gene. The x-axis is plotted in log-scale, with the genes plotted in order of decreasing transcript abundance. (B) Heatmap with clustering based on Spearman correlation. Transcript counts are plotted as log2 values, with yellow and dark blue indicating low and high counts, respectively. The four coloured bars represent treatments, sampling times, iron levels, and experiment type. For the treatment, low-iron samples are shown in blue, dimethyl sulfoxide (DMSO)-treated samples are purple, iron-recovery samples are shown in orange, and high-iron samples are black. Timepoints are separated into Initial, 5–15 min, 30–45 min, 1 h, and >1 h in increasingly darker shades of green, with darker shades indicating the time progression from earlier to later timepoints. Iron levels are indicated in grey for samples without iron and black for samples with iron. Samples from the long-term experiment are shown in dark brown, and samples from the short-term experiment are shown in light brown. The dendrogram above the heatmap indicates the following clusters: a) downregulated genes following iron addition and b) upregulated genes. The dendrogram on the left indicates clusters 1 and 3 with samples from 12 h and 14 h after iron addition. The split region shown as cluster 2 consists of iron-recovery samples from time points with rapid acclimation of the iron-responsive transcripts. The following abbreviations were used: flavodoxin (FLDA1), fructose-bisphosphate aldolase (FBA), plastocyanin (PETE), ferredoxin (PETF), iron starvation-induced protein (ISIP), cellular repressor of EA1-stimulated genes (CREG), heat shock factor (HSF), manganese-superoxide dismutase (Mn-SOD), transcription factor (TF), oligosaccharyltransferase (OST), nuclear import-export receptor house-keeping (nucl_hk), N-acetylglucosaminyltransferase (GnT1), UDP-glucose glucosyltransferase (UGGT), iron reductase (FRE1).
Fig 9.
Principle component analysis (PCA) of all samples. Low-iron samples are shown in blue, high-iron in black and recovery samples are shown in orange. Sampling times are divided into groups and are shown as different shapes. The initial samples are represented by a plus sign, samples from 5–15 min are shown as a square, 30–45 min as a triangle, samples from one hour are shown as a dot, and all samples after one hour are shown as a crossed square.
Fig 10.
Structure and evolutionary distance to other CREG proteins of CREGx2_To.
(A) CREGx2_To protein structure. Transmembrane domain (TM) and N-glycosylation sites are indicated as barrels and squares, respectively. (B) Phylogenetic tree of CREG proteins. The T. oceanica sequence is shown in bold (THAOC08512). The maximum likelihood tree was generated in MEGA [44] using the top 100 blastp hits. The sequence alignment is based on a muscle alignment, and the tree was generated with the WAG model and gamma distribution. The black circle represents a bootstrap of > 0.5. Three conserved domains are indicated. Complex I intermediate-associated protein 30 (CIA-30) is in red, Pyrodoxamin-5’phosphate oxidase (Pyr-ox) in blue and histidine-rich region (His-rich) is indicated in green. The signal peptide is abbreviated with SP.
Fig 11.
Cellular overview of the targeted transcripts and their localization in the cell.
The scheme of a T. oceanica cell is shown with chloroplast (light-green), mitochondrion (Mito/blue), endoplasmic reticulum (ER/beige), Golgi apparatus (Golgi/orange), and the nucleus (Nucl/brown) indicated. The position of the gene name indicates the localization with respect to published data (ISIP1, ISIP2, ISIP3, FBA1, and FBA3) or putative location based on a prediction by TargetP [58] and ChloroP [59]. The surrounding colour shows the state and the temporal change of regulation upon iron resupply with upregulated genes shown in orange, dark orange, light red, and dark red indicating slow to fast acclimation, respectively. The downregulated genes are shown in two shades of green with the darker green indicating the fastest acclimation, and neutral genes are shown in yellow.