Life-cycle-related gene expression patterns in the brown algae
Figures
Generation-biased gene expression across the brown algae.
(A) Cladogram of the brown algae showing the 10 species for which generation-biased gene expression was compared. Adapted from Cock et al., 2014. (B) Generation-biased gene expression in relation to life cycle dimorphism. Pie diagrams indicating the proportions of sporophyte-specific (dark red), sporophyte-biased (light red), gametophyte-specific (dark blue), gametophyte-biased (light blue), and unbiased (grey) genes in species with different haploid-diploid life cycles ranging from sporophyte-dominant through isomorphic to gametophyte-dominant (indicated by the red and blue wedges). sp., species. (C) Overlaps (Jaccard index) between sets of gametophyte-biased/specific and sporophyte-biased/specific genes across the 10 species.
Schematic life cycles and photographs of the sporophyte and gametophyte generations of brown algal species analysed in this study illustrating the different degrees of generational dimorphism.
Size bars indicate the approximate sizes of each generation of each life cycle, providing an indication of the degree of dimorphism between the two generations. The life cycle and the morphology and size of the sporophyte and gametophyte generations of Saccharina japonica (not shown) are highly similar to those of Saccharina latissima. Photograph credits: M. pyrifera sporophyte, Shannon DeVanney; S. latissima, Samuel Boscq (Roscoff Biological Station); M. clavaeformis, Akira F. Peters (Bezhin Rosco); S. promiscuus sporophyte, Masakazu Hoshino (Kobe University); P. littoralis gametophyte, Inka Bartsch (Alfred Wegener Institut) and Delphine Scornet (Roscoff Biological Station); P. littoralis sporophyte, Alfred Wegener Institut and Delphine Scornet (Roscoff Biological Station); S. polyschides gametophyte, M. pyrifera gametophyte, E. siliculosus, Delphine Scornet (Roscoff Biological Station).
Biological function gene ontology term enrichment in the generation-biased/specific gene sets of the 10 species.
Conserved general functional categories are indicated by coloured boxes. padj, p-value adjusted for multiple testing based on the Benjamini-Hochberg false discovery rate; GA, gametophyte; SP, sporophyte.
Gene expression patterns during the Ectocarpus species 7 life cycle.
(A) Principal component analysis of the Ectocarpus species 7 gene expression across 10 life cycle stages. The numbers in circles refer to the differential gene expression analyses illustrated in C and D. Most of the stages analysed consisted entirely of non-flagellated cells with the exception of the gamete stage and mature sporophyte and gametophyte stages, which bear flagellated spores and gametes, respectively, in sporangia and gametangia, respectively. (B) Percentage changes in gene expression (regression breakpoints) during Ectocarpus species 7 sporophyte development (from left to right, transitions between free-swimming male gamete, sporophyte initial cell (24 hours after gamete release), elongating sporophyte initial cell (48 hours after gamete release), sporophyte 2–5 cell stage, non-fertile adult sporophyte and fertile sporophyte). (C) Proportions of differentially expressed genes in the transitions between (1) free-swimming male gamete and sporophyte initial cell stage, and (2) adult sporophyte and gametophyte stages, determined using DESeq2 (Supplementary file 1). (D) Density distribution of the |log2(FoldChange)| for upregulated and downregulated genes in the differential gene expression analysis described in C. Dotted lines indicate mean values.
Light microscopy images of Ectocarpus species 7 initial cell and elongating initial cell.
(A) Ectocarpus species 7 strain Ec32 initial cell (settled male Ec32 gamete, 24 hr after release). (B) Ectocarpus species 7 strain Ec32 elongating initial cell (settled male gamete, 48 hr after release). Inverted Leica microscope, magnification: x200, scale bar: 100 µm.
Breadth of expression (tau) of genes that were differentially expressed, or not, across the transition from gamete to sporophyte initial cell.
Significantly downregulated and upregulated genes were identified using DESeq2. Dotted lines indicate mean tau values.
Characterisation of gene co-expression modules in Ectocarpus species 7.
(A) Functional categories manually assigned to modules based on Gene Ontology term enrichment analysis of each module. (B) Enrichments in subcellular localisations based on HECTAR predictions: mitochondria (small ellipses), plastid (tube shape with thylakoids) and signal peptide/anchor peptide for secreted proteins. Note that HECTAR does not predict nuclear localisation. (C) Proportion of differentially expressed genes (DESeq2 adjusted p-value<0.05 and |log2FoldChange|>1) in each gene module in the transitions between (1) free-swimming male gametes and sporophyte initial cell stage, and (2) adult sporophyte and gametophyte (Supplementary file 1). Significant enrichment (indicated by a darker shade of green or orange) means that the proportion of differentially expressed genes in the module was significantly greater (ClusterProfiler adjusted p-value<0.05) than the proportion for the entire genome. (D) Average module gene expression profile computed on genes with a WGCNA module MM >0.86 for, from left to right: free-swimming male gamete, sporophyte initial cell (24 hr after gamete release), elongating sporophyte initial cell (48 hr after gamete release), sporophyte 2–5 cell stage, non-fertile adult sporophyte, fertile sporophyte, non-fertile female and male gametophytes, fertile female and male gametophytes developmental stages. (E) Number of genes in the module positively or negatively correlated with the module eigengene. (F) Module eigengene dendrogram showing the relationship between module eigengenes.
Hierarchical cluster trees showing co-expression modules identified using WGCNA.
(A) Ectocarpus species 7. (B) D. dichotoma. Grouped branches correspond to modules, which are indicated by coloured boxes underneath the tree.
Enriched GO terms in Ectocarpus species 7 gene modules.
Statistically significant enriched GO terms in Ectocarpus species 7 gene modules. Enrichment is indicated as log2 of the ratio of the proportion of genes assigned to the GO term in the module divided by the proportion for the whole genome. When it was possible to manually assign a general function to a module, the annotation is indicated in brackets after the module name. padj, p-value adjusted for multiple testing based on the Benjamini-Hochberg false discovery rate.
Transcript abundance heatmap for Ectocarpus species 7 CAZYme genes.
Heatmaps with log2(NormalisedCounts +1) for replicate samples of (from left to right) free-swimming male gamete, sporophyte initial cell (24 hr after gamete release), elongating sporophyte initial cell (48 hr after gamete release), sporophyte 2–5 cell stage, non-fertile adult sporophyte, fertile sporophyte, non-fertile female and male gametophyte, fertile female and male gametophyte Ectocarpus species 7 developmental stages for GT23 and PL41 annotated genes. Left annotation track: differential expression (DEG) analysis results in the transitions between (1) free-swimming male gamete and sporophyte initial cell stage, and (2) adult sporophyte and gametophyte; right annotation track: WGCNA module colours.
Transcript abundance heatmap for Ectocarpus species 7 transcription factor genes that have one-to-one D. dichotoma orthologues.
Hierarchically clustered heatmap of log2(NormalisedCounts +1) for 325 Ectocarpus species 7 TAP genes (Denoeud et al., 2024) and 90 EsV1-7 domain genes (Macaisne et al., 2017) for replicate samples of the following life cycle stages: (from left to right) free-swimming male gamete, sporophyte initial cell (24 hr after gamete release), elongating sporophyte initial cell (48 hr after gamete release), sporophyte 2–5 cell stage, non-fertile adult sporophyte, fertile sporophyte, non-fertile female and male gametophytes, fertile female and male gametophytes. Left annotation track: differential expression (DEG) analysis results in the transitions between (1) free-swimming male gamete and initial cell stage, and (2) sporophyte and gametophyte; right annotation track: WGCNA module colours.
Expression of translation-related genes during Ectocarpus species 7 development.
(A) Mean expression profile computed for genes with a WGCNA module MM >0.86 for (from left to right): free-swimming male gamete, sporophyte initial cell (24 hr after gamete release), elongating sporophyte initial cell (48 hr after gamete release), sporophyte 2–5 cell stage, non-fertile adult sporophyte, fertile sporophyte, non-fertile female and male gametophyte, fertile female and male gametophyte developmental stages in modules ‘purple’, ‘cyan’, ‘greenyellow’, ‘magenta’. (B and C) Heatmaps showing log2(NormalisedCounts +1) values across the same developmental timepoints as in A for manually reannotated ribosomal protein genes corresponding to cytosolic, chloroplast and mitochondrial ribosomal subunits (B) and translation-related genes (C). Left annotation track: differential expression analysis results for the transitions between (1) free-swimming male gamete and sporophyte initial cell stage, and (2) adult sporophyte and adult gametophyte; right annotation track: WGCNA module colours. (D) Schematic representation of translation-related functions enriched in the ‘greenyellow’, ‘purple’, ‘magenta’, ‘cyan’ modules. When it was possible to manually assign a general function to a module, the annotation is indicated in brackets after the module name. (E) Gene ontology terms significantly enriched in the sets of genes clustered within the ‘purple’, ‘greenyellow’, ‘cyan’, and ‘magenta’ modules in Ectocarpus species 7. Enrichment is indicated as log2 of the ratio of the proportion of genes assigned to the GO term in the module divided by the proportion for the whole genome. CC, cellular component; BP, biological process; MF, molecular function.
Transcript abundance heatmap for Ectocarpus species 7 ribosomal genes.
Heatmap with log2(NormalisedCounts +1) for (from left to right) free-swimming male gamete, sporophyte initial cell (24 hr after gamete release), elongating sporophyte initial cell (48 hr after gamete release), sporophyte 2–5 cell stage, non-fertile adult sporophyte, fertile sporophyte, non-fertile female and male gametophyte, fertile female and male gametophyte Ectocarpus species 7 developmental stages for ribosomal protein annotated genes. Left annotation track: differential expression analysis results in the transitions between (1) free-swimming male gamete and sporophyte initial cell stage, and (2) adult sporophyte and gametophyte; right annotation track: WGCNA module colours.
Three-dimensional representations of a ribosome illustrating differential expression of ribosomal protein genes.
(A, B) Lateral views. (C) Top view. (D) Bottom view. (E and F) Intersubunit views of the small and large subunit, respectively. Small and large subunit ribosomal proteins are coloured in red and purple, respectively. Darker colour indicates that the gene encoding a protein was significantly upregulated between the gamete and sporophyte initial cell stages. Proteins coloured in white are present in the Triticum aestivum ribosome, which was used as a reference, but no equivalent was found in Ectocarpus species 7. RNAs are shown as ribbons, ribosomal RNAs in grey, messenger and transfer RNAs in green. A, P, and T indicate the aminoacyl, peptidyl, and exit tRNA binding sites, respectively. En, mRNA entry site; Ex, mRNA exit site; Tu, nascent peptide tunnel exit.
Conservation of life-cycle-related gene co-expression patterns between Ectocarpus species 7 and D. dichotoma.
(A) WGCNA density and connectivity median rank statistics indicating the degree of conservation of gene co-expression patterns when Ectocarpus species 7 genes in Ectocarpus species 7 co-expression modules were substituted with their D. dichotoma orthologues. Manually assigned general biological functions are indicated in boxes. SDR, sex-determining region. Module size indicates the number of one-to-one orthologues in each module. The modules ‘darkred’ and ‘darkturquoise’ were not included in this analysis because they had only one and zero orthologues, respectively (Supplementary file 6). TF, transcription factor; SDR, sex-determining region. (B) Correlation heatmap comparing three Ectocarpus species 7 modules between Ectocarpus species 7 and D. dichotoma: ‘midnightblue’ (very poor conservation), ‘purple’, and ‘royalblue’ (good conservation). The lower half of the heatmap was calculated based on the expression pattern of the D. dichotoma orthologue of each Ectocarpus species 7 gene in each module. (C) Counts of shared one-to-one orthologues between gene co-expression modules defined for Ectocarpus species 7 (x-axis) and D. dichotoma (y-axis). The colour code represents the log10(p-value) for Fisher’s exact test (red bar), which was applied to determine whether pairs of modules shared a greater number of one-to-one orthologues than expected from a random distribution. Numbers after the module names indicate the number of one-to-one orthologues in each module. The ‘dark turquoise’ module was not included in this analysis because it contained zero orthologues (Supplementary file 6). (D) Heatmap showing the expression levels (log2(NormalisedCounts +1)) of 20 selected one-to-one orthologous flagellum-related genes in Ectocarpus species 7 and D. dichotoma. Ectocarpus species 7 timepoints: free-swimming male gamete, sporophyte initial cell (24 hr after gamete release), elongating sporophyte initial cell (48 hr after gamete release), sporophyte 2–5 cell stage, non-fertile adult sporophyte, fertile sporophyte, non-fertile female and male gametophytes, fertile female and male gametophytes. D. dichotoma timepoints: sperm, egg cell, zygote, embryo, adult non-fertile sporophytes, and female and male gametophytes. (E) Computed z score based on log2(NormalisedCounts +1) values for three TALE homeodomain transcription factors over six developmental stages shared between Ectocarpus species 7 and D. dichotoma, namely sperm, zygote, early sporophyte, non-fertile sporophyte, female and male gametophyte. The Pearson correlation coefficient between the two expression datasets (cor) and the associated p-value is indicated.
D. dichotoma PCA and modules.
(A) Principal component analysis of D. dichotoma gene expression across seven life cycle stages. (B) Characterisation of D. dichotoma gene co-expression modules generated using WGCNA. Left to right: manually assigned functional categories based on GO term enrichment; number of genes in the module positively or negatively correlated with the module eigengene; average module gene expression profiles for WGCNA modules across sperm (1 hr after release), egg cell (15 min after release), zygote (1 hr after release), embryo (8 hr after fertilisation), fertile sporophyte, fertile female and male gametophyte life cycle stages; module eigengene dendrogram. The average module gene expression profiles were calculated using genes with a WGCNA MM >0.86.
Enriched GO terms D. dichotoma gene modules.
Statistically significant enriched GO terms in D. dichotoma gene modules. Enrichment is indicated as log2 of the ratio of the proportion of genes assigned to the GO term in the module divided by the proportion for the whole genome. padj, p-value adjusted for multiple testing based on the Benjamini-Hochberg false discovery rate.
Transcript abundance heatmap for one-to-one orthologous transcription factors between Ectocarpus species 7 and D. dichotoma.
Heatmap representing expression of one-to-one orthologous genes from Ectocarpus species 7 and D. dichotoma presented as the z score of the log2(NormalisedCounts +1) for the following equivalent developmental stages in Ectocarpus species 7 and D. dichotoma: sperm, zygote, early sporophyte, adult sporophyte, female and male gametophyte. The left-hand tracks show Pearson correlation coefficients and log10 Benjamini-Hochberg-adjusted p-values (log10 padj).
Plot of estimations of the mean percent shared identity between the orthologues within each module (based on mean BLASTp identity scores) against log10(pvalue) values obtained with the Fisher's exact test applied in Figure 6C to determine whether pairs of modules shared a greater number of one-to-one orthologues than expected from a random distribution.
Error bars indicate the standard deviation.
Additional files
-
Supplementary file 1
RNA-seq data used in this study.
GA, gametophyte; SP, sporophyte; pSP, partheno-sporophyte; GBG, generation-biased genes.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp1-v1.xlsx
-
Supplementary file 2
Generation-biased and generation-specific gene sets in 10 species of brown algae based on comparisons of adult sporophyte and gametophyte generations.
(A) Overview of the generation-biased and generation-specific gene sets of each species. (B) Complete proteomes of the 10 species indicating the results of the comparison of adult sporophytes and gametophytes, including log2(fold change), p-value, mean TPM and assigned generation-biased expression class. TPM, transcripts per kilobase million.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp2-v1.xlsx
-
Supplementary file 3
List of orthogroups for the 10 brown algal species analysed for generation-biased gene expression.
The genomes analysed (with ENA accession numbers or links to ORCAE in brackets) correspond to: Dictyota dichotoma strain ODC1387m (GCA_964200555), Ectocarpus species 7 strain Ec32 (https://bioinformatics.psb.ugent.be/orcae/overview/EctsiV2), Macrocystis pyrifera strain P11B4 (GCA_964200385), Myriotrichia clavaeformis strain Myr cla04 (GCA_964200105), Pylaiella littoralis strain U1.48 (GCA_964200295), Saccharina latissima strain SLPER63f7 (GCA_964200175), Saccharina japonica strain Ja (https://bioinformatics.psb.ugent.be/orcae/overview/Sacja), Saccorhiza polyschides strain SpolBR94m (GCA_964200605), Scytosiphon promiscuus strain Ot110409-Otamoi-16-male (GCA_964200365) and Sphacelaria rigidula strain Sph rig Cal Mo 4-1-68b (GCA_964200075). The one-to-one orthologues correspond to the 9,317 orthogroups identified using the relaxed criteria described in the Materials and methods.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp3-v1.xlsx
-
Supplementary file 4
Counts of RNA-seq reads mapped to Ectocarpus species 7 genes.
This data was used to calculate the gene expression levels reported in Supplementary file 5.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp4-v1.xlsx
-
Supplementary file 5
Functional and expression-related information for Ectocarpus species 7 genes.
List of all Ectocarpus species 7 genes with transcript abundance under each of the conditions analysed, measured as log2(NormalisedCounts +1), WGCNA module assignment (Colour), WGCNA module membership (MM) scores and associated p-values (pMM) for each gene in each module, the functional annotation for each gene (description), the manually assigned functional category (Manual.functional.category), the DESeq2 output for the comparison between free-swimming male gamete and sporophyte initial cell stage (gamete_initial), and adult sporophyte and gametophyte (sporophyte_gametophyte) with the corresponding differential expression (DE) annotation, the HECTAR output for each gene with the predicted targeting category and corresponding scores.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp5-v1.xlsx
-
Supplementary file 6
Overview of the gene co-expression modules for Ectocarpus species 7 and D. dichotoma.
The grey modules contain all the genes that could not be assigned to a co-expression module. The gold module was a random sample of 1000 Ectocarpus species 7 genes that provided a null distribution of gene expression patterns for the comparisons with D. dichotoma modules.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp6-v1.xlsx
-
Supplementary file 7
Orthofinder orthogroup analysis of Ectocarpus species 7 and D. dichotoma.
The genomes analysed (with ENA accession numbers or links to ORCAE in brackets) correspond to: Dictyota dichotoma strain ODC1387m (GCA_964200555) and Ectocarpus species 7 strain Ec32 (https://bioinformatics.psb.ugent.be/orcae/overview/EctsiV2).
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp7-v1.xlsx
-
Supplementary file 8
Conservation of life-cycle-related co-expression modules between Ectocarpus species 7 and D. dichotoma.
For each Ectocarpus species 7 module: preservation statistics computed by the WGCNA modulePreservation function (200 permutations), preservation statistics computed by the NetRep modulePreservation function (10,000 permutations) with the associated log10(p-value) (Bonferroni-corrected for WGCNA), and the silhouette score for each module in D. dichotoma. WGCNA medianRank statistic, aggregation of the densityMedianRank and connectivityMedianRank statistics; WGCNA – medianRankDensity.pres, WGCNA density median rank preservation statistic; WGCNA – medianRankConnectivity.pres, WGCNA connectivity median rank preservation statistic; WGCNA propVarExplained.pres, coherence: proportion of module variance explained by the module eigengene (=summaryprofile = eigenvector of the 1st principal component across all observations for every node composing the module); WGCNA – meanSignAwareKME.pres, average node contribution: average Pearson correlation coefficient to the module’s summary profile; WGCNA - meanSignAwareCorDat.pres, density of correlation structure: how strongly modules are correlated in the test dataset, average node correlation; WGCNA - meanAdj.pres, average edge weight: average connection strength between nodes; WGCNA – meanMAR.pres, average module adjacency ratio; WGCNA: cor.kIM, concordance of intramodular connectivity: similarity of the relative rank of each nodes’ weighted degree (intramodular connectivity) across datasets; WGCNA – cor.kME, concordance of node contribution: preservation of relative rank of nodes (ordered by Pearson correlation coefficient to the module’s summary profile) across datasets; WGCNA – cor.kMEall; concordance of node contribution: preservation of relative rank of nodes (ordered by Pearson correlation coefficient to the module’s summary profile) across datasets taking into account the kME of the gene for all modules; WGCNA - cor.cor, concordance of correlation structure: quantifies how similar the correlation structure is across datasets; WGCNA – cor.MAR, concordance of module adjacency ratio: concordance of the ratio between the maximum and the minimum adjacency in the module; WGCNA - separability.pres, separability preservation; NetRep - avg.weight, average edge weight: average connection strength between nodes; NetRep - coherence, coherence, i.e. proportion of module variance explained by the module eigengene (=summaryprofile = eigenvector of the 1st principal component across all observations for every node composing the module); NetRep – cor.cor, concordance of correlation structure: quantifies how similar the correlation structure is across datasets; NetRep - cor.degree, concordance of intramodular connectivity: similarity of the relative rank of each nodes’ weighted degree (intramodular connectivity) across datasets; NetRep - cor.contrib, concordance of node contribution: preservation of relative rank of nodes (ordered by Pearson correlation coefficient to the module’s summary profile) across datasets; NetRep - avg.cor, density of correlation structure: how strongly modules are correlated in the test dataset, average node correlation; NetRep - avg.contrib, average node contribution: average Pearson correlation coefficient to the module’s summary profile; silhouette.cluster.width, Silhouette coefficient; silhouette.distance.to.average, Silhouette coefficient. The ‘gold’ module was a random sample of 1000 Ectocarpus species 7 genes that provided a control distribution of gene expression patterns for comparisons with D. dichotoma modules. The ‘dark turquoise’ module was not included in this analysis because it contained zero orthologues (Supplementary file 6).
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp8-v1.xlsx
-
Supplementary file 9
Counts of RNA-seq reads mapped to D. dichotoma genes.
This data was used to calculate the gene expression levels reported in Supplementary file 10.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp9-v1.xlsx
-
Supplementary file 10
Functional and expression-related information for D. dichotoma genes.
List of all D. dichotoma genes with the log2(NormalisedCounts +1) under each condition used for the analysis, WGCNA module assignment (Colour), WGCNA module membership (MM) scores and associated p-values (pMM) for each gene in each module.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp10-v1.xlsx
-
Supplementary file 11
Correlation between the expression patterns of orthologous transcription factors in Ectocarpus species 7 and D. dichotoma.
Correlation was evaluated based on comparisons of six approximately equivalent stages for the two species (Supplementary file 1). The names of the Ectocarpus species 7 and D. dichotoma genes in each orthogroup are given in Figure 6—figure supplement 3. padj, p-value adjusted for multiple testing based on the Benjamini-Hochberg false discovery rate.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp11-v1.xlsx
-
Supplementary file 12
Manual re-annotation of the Ectocarpus species 7 ribosomal proteins.
The e-values and scores are for BLASTp queries of the Ectocarpus species 7 ribosomal proteins against the Arabidopsis thaliana ribosomal protein dataset.
- https://cdn.elifesciences.org/articles/107449/elife-107449-supp12-v1.xlsx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/107449/elife-107449-mdarchecklist1-v1.docx