Transcriptome remodeling drives acclimation to iron availability in the marine N2-fixing cyanobacterium Trichodesmium erythraeum IMS101

Abstract

While enhanced phytoplankton growth as a result of iron (Fe) fertilization has been extensively characterized, our understanding of the underlying mechanisms remains incomplete. Here, we show in a laboratory setup mimicking Fe fertilization in the field that transcriptome remodeling is a primary driver of acclimation to Fe availability in the marine diazotrophic cyanobacterium Trichodesmium erythraeum IMS101. Fe supplementation promoted cell growth, photosynthesis and N₂ fixation, and concomitant expression of the photosynthesis and N₂ fixation genes. The expression of genes encoding major Fe-binding metalloproteins is tightly linked to cellular carbon and nitrogen metabolism and appears to be controlled by the ferric uptake regulator FurA, which is involved in regulating Fe uptake and homeostasis. This feedback loop is reinforced by substitutive expression of functionally equivalent or competitive genes depending on Fe availability, as well as co-expression of multiple Fe stress inducible isiA genes, an adaptive strategy evolved to elicit the Fe-responsive cascade. The study provides a genome-wide perspective on the acclimation of a prominent marine diazotroph to Fe availability, reveals an upgraded portfolio of indicator genes that can be used to better assess Fe status in the environment, and predicts scenarios of how marine diazotrophs may be affected in the future ocean.IMPORTANCEThe scarcity of trace metal iron (Fe) in global oceans has a great impact on phytoplankton growth. While enhanced primary productivity as a result of Fe fertilization has been extensively characterized, the underlying molecular mechanisms remain poorly understood. By subjecting the model marine diazotroph Trichodesmium erythraeum IMS101 to increasing concentrations of supplemented Fe, we demonstrate in it a comprehensively remodeled transcriptome that drives the mobilization of cellular Fe for coordinated carbon and nitrogen metabolism and reallocation of energy and resources. Our data provide broad genomic insight into marine diazotrophs acclimation to Fe availability, enabling the versatility and flexibility in choice of indicator genes for monitoring Fe status in the environment and having implications on how marine diazotrophs persist into the future ocean.

Keywords: Trichodesmium; iron limitation; isiA gene cluster; marine diazotroph; transcriptome remodeling.

Fig 1

Physiological response of T. erythraeum IMS101 to increased iron availability. Shown are the chlorophyll-based specific growth rates and N₂ fixation rates (A), photosynthesis-irradiation curve fitted with the rates of oxygen evolution (B), relative mRNA abundances of two iron (Fe) stress indicator genes, idiA and isiA (C), and measured concentrations of dissolved Fe redox species (II + III) (dFe') in the culture media at designated time points during IMS101 growth (D). Measurements of N₂ fixation rates, photosynthetic O₂ evolution, and qPCR assay of Fe stress genes were conducted using samples collected on days 6, 7, and 5, respectively. IMS101 was grown in media supplemented with 0, 10, and 100 nM of FeCl₃ representing the low (LFe), medium (MFe), and high (HFe) concentration iron treatments, respectively. The data are plotted as mean ± standard deviation of biological triplicates (n = 3). Dissimilar lowercase letters in panels A and C represent significant differences across the treatments (P < 0.05). The shaded area in panel D represents the 95% confidence interval.

Fig 2

Transcriptome landscape of T. erythraeum IMS101 in acclimation to increased iron availability. (A) Circular presentation of differentially expressed genes between the HFe and LFe conditions along the IMS101 genome. The rings of the circos plot from the outermost to the innermost are (i) average read count per gene of the HFe treated samples, (ii) significantly upregulated genes under the HFe condition highlighted in dark color, (iii) heatmap of log₂FC between the HFe and LFe conditions, (iv) significantly downregulated genes under the HFe condition highlighted in dark color, and (v) average read count per gene of the LFe treated samples. For visualization purposes, a maximum read count of 50,000 has been set for rings i and v, and counts above this threshold are highlighted with green bars. Log₂FC, log₂-scaled fold change between the HFe and LFe treatments after normalization. (B) Close-up view of the four gene clusters depicted in panel A, the downregulated isiAB operon (1), isiA2-isiA3-isiA4-cpcG2 (2), a cryptic gene cluster possibly involved in anti-microbial secondary metabolite biosynthesis (3), and the upregulated nitrogenase nif gene cluster (4). (C) Venn diagram summarizing the numbers of differentially expressed genes among treatments. Numbers in black represent the number of genes with >2-fold changes. From these, the numbers of upregulated genes are shown in red and those of downregulated genes in blue.

Fig 3

Changes in the expression of genes related to iron acquisition and utilization in T. erythraeum IMS101 upon iron supplementation. Shown are genes for Fe homeostasis regulatory factors (A), Fe absorption and acquisition (B), Fe-S cluster assembly (C), and Fe storage (D). The relative abundance of gene transcripts at each Fe supplementation concentration was log₂ transformed and plotted according to the formula Z = (x − μ)/σ, where x is the mean of log₂CPM in one treatment, μ is the mean of log₂CPM values in all treatments, and σ is the standard deviation of the overall normal distribution curve. This formula centers and scales a variable to mean 0 and standard deviation 1, which ensures that the criterion for finding linear combinations of the predictors is based on how much variation they explain and, therefore, improves the numerical stability.

Fig 4

Photosynthetic and N₂-fixing genes differentially expressed upon iron supplementation. (A) Expression of genes involved in N₂ fixation is shown as fold changes in the level of transcripts between the HFe and LFe treatments. Also shown is the organization of genes in the 22,106 bp long nif gene cluster. (B) Expression of genes involved in photosynthesis is shown as Z-scores corresponding to each iron supplementation treatment. Z-scores were calculated according to the following formula: Z = (x − μ)/σ, where x is the mean of log₂CPM in one treatment, μ is the mean of log₂CPM values in all treatments, and σ is the standard deviation of the overall normal distribution curve. Genes are labeled with gene symbol (if available) followed by locus tag of the open reading frame (e.g., psbA1_4763 denoting the psbA1 gene in the locus tag Tery_4763 in the Trichodesmium genome). For the gene products associated with different forms of Fe (in red text), the numbers and types of the corresponding Fe ions/compounds are given. All Fe-containing molecules are shown in ball-and-stick representation with the Fe, S, Mo atoms depicted as green, yellow, magenta spheres, respectively.

Fig 5

Diversity and conservation of isiA gene family in cyanobacteria. (A) Variants of isiA genes in Trichodesmium erythraeum IMS101, Crocosphaera watsonii WH0003, Anabaena sp. PCC7120 and Leptolyngbya sp. JSC-1 genomes, and their possible roles in the formation of super-complexes around photosystems under iron limitation conditions. Discontinued regions are marked with “//.” (B) Maximum likelihood tree showing the phylogenetic relationship of isiA, psbC, and pcb genes found in diverse cyanobacteria. The phylogenic tree was constructed with IQ-TREE using aligned amino acid sequences of the IsiA homologs and was rooted with PsbC as the outgroup. Numbers at the branch denote the topological robustness of the tree evaluated by Bayesian criterion with 1,000 bootstrap replicates. Tree scale represents amino acid substitutions per site.

Fig 6

Expression of genes involved in the de novo synthesis of chlorophyll a, vitamin B₁₂, and heme upon iron supplementation. Most genes in the biosynthesis pathways of chlorophyll a and heme were significantly upregulated under the HFe treatment, while genes in the vitamin B₁₂ biosynthesis pathway were significantly downregulated. Expression of genes is shown as Z-scores corresponding to each iron supplementation treatment. Z-scores were calculated according to the following formula: z = (x − μ)/σ, where x is the mean of log₂CPM in one treatment, μ is the mean of log₂CPM values in all treatments, and σ is the standard deviation of the overall normal distribution curve. The genes are labeled with a gene symbol (if available) followed by a locus tag of the open reading frame (e.g., chlM_4469 denoting the chlM gene in the locus tag Tery_4469 in the Trichodesmium genome).

Fig 7

Schematic diagram illustrating significantly regulated gene products for coordinated nutrient transport and carbon and nitrogen metabolism in T. erythraeum IMS101 in acclimation to increased Fe availability. Fe supplementation promotes the expression of the genes encoding major Fe-binding metalloproteins involved in photosynthetic electron transport and structural subunits of the nitrogenase complex. The upregulated photosynthesis and N₂ fixation are accompanied by a balance between the central carbon and nitrogen metabolism, which is tightly linked through the GS-GOGAT cycle. The Fe redox is primarily under the control of the ferric uptake regulator FurA that functions as a transcriptional repressor of genes for Fe utilization, specifically those encoding the transporters of extracellular particulate (HTCaBs) and siderophore (FhuD) Fe, the plasma membrane transporters of ferric (FutABC) and ferrous (FeoAB) Fe, and the Fe-S cluster assembly proteins (SufBCDS). FurA also, possibly indirectly, functions as an activator in controlling the expression of genes encoding Fe storage ferritins (Ftn and Bfr) and proteins involved in heme and chlorophyll a biosynthesis. Moreover, the transcription of genes for nitrogen transport and utilization (amt, nrtABCD, urtABCDE, narB, and nirA) is likely under the control of a sophisticated network involving the Fur family. Proteins putatively involved in the extracellular particulate Fe-siderophore adsorption are boxed in dotted lines. Abbreviations are used as follows: TCA, tricarboxylic acid; Glu, glutamate; Gln, glutamine; GS, glutamine synthetase; GOGAT, glutamate synthase; Uro III, uroporphyrinogen III; Phgen IX, photoporphyrinogen IX; Chl a, chlorophyll a; Pchlide, Protochlorophyllide; Pre 2, precorrin 2; Co-pre 2, cobalt-precorrin 2; RuBP, ribulose-1,5-bisphosphate; 3-PGA, 3-phosphoglycerate; 1,3-PGA, 1,3-diphosphoglycerate; F1,6P, fructose-1,6-bisphosphate; F6P, fructose-6-phosphate; Ru5P, ribulose 5-phosphate; 2-OG, 2-oxoglutarate; EM, extracellular membrane; IM, intracellular membrane; TM, thylakoid membrane.