Elucidation of Spartina dimethylsulfoniopropionate synthesis genes enables engineering of stress tolerant plants

Abstract

The organosulfur compound dimethylsulfoniopropionate (DMSP) has key roles in stress protection, global carbon and sulfur cycling, chemotaxis, and is a major source of climate-active gases. Saltmarshes are global hotspots for DMSP cycling due to Spartina cordgrasses that produce exceptionally high concentrations of DMSP. Here, in Spartina anglica, we identify the plant genes that underpin high-level DMSP synthesis: methionine S-methyltransferase (MMT), S-methylmethionine decarboxylase (SDC) and DMSP-amine oxidase (DOX). Homologs of these enzymes are common in plants, but differences in expression and catalytic efficiency explain why S. anglica accumulates such high DMSP concentrations and other plants only accumulate low concentrations. Furthermore, DMSP accumulation in S. anglica is consistent with DMSP having a role in oxidative and osmotic stress protection. Importantly, administration of DMSP by root uptake or over-expression of Spartina DMSP synthesis genes confers plant tolerance to salinity and drought offering a route for future bioengineering for sustainable crop production.

Fig. 1. Spartina anglica produces DMSP via SDC and DOX.

a Spartina anglica at Stiffkey saltmarsh, May 2021. b DMSP synthesis pathway in S. anglica converts methionine to DMSP via four enzyme activities (green). c HPLC trace of SDC assay for recombinant S. anglica SDC (SaSDC, S-methylmethionine decarboxylase) incubated with SMM, leading to production of DMSP-amine, alongside an SMM-only control. d GC result of DOX assay performed on enzyme extracts from S. anglica leaf tissue, recombinant S. anglica copper amine oxidase 1 (SaCAO1) and recombinant S. anglica DOX (SaDOX, DMSP-amine oxidase). Enzymes were incubated with DMSP-amine or DMSP-amine and competing substrate putrescine in excess, leading to production of DMSP-aldehyde that spontaneously degrades to produce DMS. Pig kidney diamine oxidase used as a positive control (n = 3 per protein and substrate). e RT-qPCR of S. anglica DMSP synthesis genes in denoted tissues (n = 4 for leaf, rhizome, root and stem, n = 3 for inflorescence). p-values indicate statistical significance after one-way ANOVA followed by Tukey test; n.s. denotes no statistically significant difference. f HPLC traces of SDC assays performed on enzyme extracts prepared from denoted S. anglica tissues. Retention times of SMM (6.5 min) and DMSP-amine (7.6 min) are indicated in blue and red, respectively. g DOX assays performed on enzyme extracts prepared from denoted S. anglica tissues, normalised to 1 µg of total extracted protein and with activities shown relative to leaf (n = 3, p = 0.00000373). Letters indicate statistical significance after one-way ANOVA followed by Tukey test. h DMSP accumulation in denoted S. anglica tissues (n = 4 for inflorescence, leaf and root, n = 3 for rhizome and stem, p = 0.041686 for leaf vs stem, p = 0.005 for leaf vs rhizome, p = 0.00002820 for leaf vs root, p = 0.0000584 for leaf vs inflorescence). Letters indicate statistical significance after one-way ANOVA followed by Tukey test. Data represent mean ± one standard error. Source data are provided as a Source Data file.

Fig. 2. DMSP accumulation in Spartina anglica is associated with stress responses.

a Satellite image of Spartina anglica sampling sites at Stiffkey saltmarsh, taken from Google Maps. b Variability in DMSP accumulation in individual clumps of S. anglica harvested from the sites denoted in panel a. Samples identified as ‘highest accumulators’ (red) and ‘lowest accumulators’ (blue) were subjected to RNA-sequencing (n = 3 per clump per site, p = 8.3e-14 in one-way ANOVA). Letters indicate statistical significance after one-way ANOVA followed by Tukey test. c Volcano plot demonstrating contrasting gene expression profiles between highest and lowest DMSP accumulating clumps of S. anglica. Each dot represents a gene, with red dots indicating genes that are more highly expressed in highest accumulators, blue dots indicating genes that are lower, and grey dots denoting genes that are not significantly differentially expressed. Dotted lines indicate significance thresholds from DESeq2 analysis. Labelled dot refers to Ethylene Responsive Element Binding Protein 1 (EREBP1), a key stress responsive transcription factor in grasses. d Gene ontology enrichment analysis of differentially expressed genes from (c), demonstrating an enrichment of stress-responsive genes in highest DMSP accumulators. e Heatmap of pairwise Pearson correlations between measured variables (abundance of indicated elements and amino acids in S. anglica leaves), demonstrating negative linear relationship between DMSP and levels of nitrogen and proline, and positive linear relationship between DMSP and molybdenum (Mo). Crosses indicate that a relationship is not significant. Data represent mean ± one standard error. Source data are provided as a Source Data file.

Fig. 3. Spartina anglica produces high concentrations of DMSP due to its unique SDC activity.

a DMSP measurements showing accumulation of the compound in a large diversity of different plant species (n = 3 per species). b Phylogenetic tree of ODC/SDCs in plant species related to S. anglica. c HPLC traces of SDC assays showing that the recombinant S. anglica SDC produces DMSP-amine from SMM, whilst enzyme homologs from other plant species do not have this activity. d Phylogenetic tree of DOX homologs in plant species related to S. anglica. e DOX assays performed with selected enzyme homologs from S. anglica (Sa), Saccharum offinarum (So), Setaria viridis (Sv) and Solanum lycopersicum (Sl) (n = 3 per recombinant enzyme, p = 0.7506270 for SaDOX vs SoDOX, p = 0.2102275 for SaDOX vs SvDOX, p = 0.0004608 for SaDOX vs SlDOX). Letters indicate statistical significance after one-way ANOVA followed by Tukey test. f Number of reads mapping to homologs of denoted DMSP synthesis genes in each species, normalised to the number of reads mapping to ACT8 homologs (square rooted for clarity), showing levels of DOX are much higher in S. anglica, despite equivalent enzymology. EF1A and GAPDH are included as reference gene controls, to demonstrate expression differences of other genes are not attributable to read-depth artefacts (n = 3 libraries for Spartina anglica, Solanum lycopersicum and Nicotiana benthamiana, n = 2 libraries for Arabidopsis thaliana and Hordeum vulgare). Data represent mean ± one standard error. Source data are provided as a Source Data file.

Fig. 4. DMSP protects plants from abiotic stress.

a Image of representative tomato plants from each treatment group, demonstrating that biomass loss caused by NaCl is rescued by DMSP treatment. b Fresh weight measurements of tomato plants from each treatment group, demonstrating that biomass loss caused by NaCl is rescued by DMSP treatment (n = 9, p < 0.05 in one-way ANOVA followed by Tukey tests for shoot and root, separately). c Volcano plot demonstrating contrasting gene expression profiles of tomato plants treated with salt stress alone (NaCl) or salt stress and DMSP (NaCl + DMSP). Each dot represents a gene, with red dots indicating genes that are more highly expressed in NaCl treatment, blue dots representing downregulation, and grey dots denoting genes that are not significantly differentially expressed. Dotted lines indicate significance thresholds from DESeq2 analysis. d Clustered heatmap of average normalised count numbers of differentially expressed genes of interest from panel c, showing that samples from Control and NaCl + DMSP-treated plants cluster together away from samples from NaCl-treated plants. e DMSP measurements of Nicotiana benthamiana leaves transformed with different S. anglica DMSP synthesis genes, showing that all three genes (MMT, SDC and DOX) are needed for high DMSP accumulation (n = 15 for negative control, MMT SDC, SDC DOX, and MMT SDC DOX, n = 5 for MMT, SDC, DOX). p-value indicates statistical significance after one-way ANOVA followed by Tukey test; n.s. denotes no statistically significant difference. f DMSP measurements of Arabidopsis thaliana over-expressing S. anglica MMT, SDC and DOX showing highly elevated accumulation in four independent transgenic lines (n = 3 × 18 plants). g Images of wildtype (WT, Col-0) and transgenic A. thaliana seedlings (Line 2) growing in the presence and absence of NaCl, demonstrating over-expression of S. anglica DMSP synthesis genes (MMT, SDC and DOX) convey tolerance to salt stress. Scale bar = 1 cm. h Fresh weight measurements of plants in (g), demonstrating biomass rescue is statistically significant (n = 6 × 9 plants, total 54 per condition). p-values indicate statistical significance after one-way ANOVA followed by Tukey test. Data represent mean ± one standard error. Source data are provided as a Source Data file.