Salt stress is causing discoordination of root and shoot growth. Arabidopsis HapMap population was screened for salt stress-induced changes in root:shoot ratio.

The increase in the projected area of shoot and root (Fig. S2) was used to estimate growth rate of (A) shoot and (B) root using an exponential function. (C) The root:shoot growth rate ratio was calculated per genotype. The histograms represent the number of accessions acros three studied conditions (0, 75 and 125 mM NaCl), whereas the population average is indicated using the dashed line. Additionally, the distribution of the genotypes within each treatment was additionally visualized using the error plots (lower panel), where population average and standard error is indicated using a colored point and a line respectively. Individual grey points represent individual genotypes.

SR3G is associated with salt stress-induced changes in shoot-per-total seedling area ratio.

The salt induced changes in root-to-shoot and shoot-per-total seedling area ratios were used as an input for GWAS. Manhattan plots representing the associations of shoot-per-total seedling area recorded 6 days after transfer to 125 mM NaCl and (A) 250 k SNPs and (B) 4 M SNPs. (C) Significant associations were found with the SNPs forming a locus on chromosome 3 in and around SR3G (AT3G50160), encoding Domain of Unknown Function 247 (DUF247). The natural variation in the LD region was studied in 162 accessions sequenced by 1001 Genomes Project. The upper panel represents portion of the missing data, upper middle panel represents deletions relative to Col-0, while lower middle panel represents the sequence similarity compared to Col-0. The bottom panel represents the open reading frames (ORFs) within the LD, and the location of associated SNPs is indicated with the dashed lines. Red dashed lines represent associations above Bonferroni threshold in 250k SNP mapping, while grey lines represent associations with -log10(p-value) > 5 in the 4M SNP mapping. (D) The haplotype analysis performed using SNPs located within the coding region of the SR3G revealed significant differences in shoot-per-total seedling area recorded 6 days after transfer to 125 mM NaCl between Haplotype groups 1 (represented by 46 accessions, including Col-0) and 11 (represented by 11 accessions, including Blh-1). The significant differences between individual haplotype groups were tested using ANOVA with Tukey HSD to identify significantly different groups. (E) Upon further sequencing of SR3G in accessions from Haplotype 1 and 11, revealed two 200 bp insertions within SR3G exons in 4 tested accessions belonging to haplotype 11 group. These 200 bp insertions resulted in a premature STOP-codon within the DUF247 domain (after Gly-215).

Phylogeny and positive selection analysis of SR3G and its orthologs from other species.

The orthologous genes of SR3G (AT3G50160) from other seven species were identified using GoGeBlast (E-value < 1.00E-10). (A) The protein sequences of the 50 homologous genes were aligned by MUSCLE and then an unrooted phylogenetic tree was reconstructed by a Maximum -likelihood (ML)-derived style by RAxML (Bootstrap number: 100). (B) A positive selection analysis was conducted by branch-site model A test: Specifically, we hypothesized the differential or constant dN/dS (ω) substitution rate among the closely distant branches leading to DUF247 (AT3G50160) (See Method & Materials for details). These branches were marked as ω1 (light brown branch), ω2 (DUF247-specific branch), ω3 (light blue branch), and ω-rest (light purple branch). (C) The gene structure of DUF247 (AT3G50160) was illustrated by wide arrows (grey arrow: 5 and 3’ UTR, yellow arrow: coding-region). The regions used for cloning (Fragment A, B, and C) were marked by narrow arrows along with “Coil-Coil Domain”, “Putative Ca2+ binding fomain” and “Super-polar Fragment” marked on Fragment A and B, respectively. Lastly, the Bayes Empirical Bayes (BEB) was performed to test the probability of sites along with ω > 1 over the DUF247 (AT3G50160). Three sites were generated with a posterior probability > 0.90: 124C (P = 0.992), 206A (P = 0.951), and 209A (P = 0.988). These three sites were marked with an orange asterisk and assumed to have been undergoing positive selection.

Transcript abundances for the two closely related DUF247s.

Expression of (A) SR3G (AT3G50160) (B) and its closest homologue DUF247-150 (AT3G50150) were measured in Col-0 seedlings grown with and without salt stress. RT-qPCR analyses were conducted using seedlings grown on 1/2x MS for 4 d and then followed by transferring to the treatment plates with or without 75 mM NaCl for one week. Mean values are shown ± SE, with number of replicates (n) shown in each graph. AT4G04120 was used as a reference gene for normalization. Significance was determined by the Tukey–Kramer HSD test in JMP.

SR3G full-length protein resides in the plasma membrane while removal of its transmembrane domain results in protein relocation to nucleus and cytosol.

SR3G full-length protein (A-B) or truncated versions (C-E) fused to mVenus at the N-terminus were agro-infiltrated into the Nicotiana benthamiana leaves for transient expression. Shown are a bright-field image of transfected leaf cells and mVenus-SR3G-mediated fluorescence as well as GFP-based subcellular marker. (A) Localization of full-length SR3G alone or (B) in combination with the known plasma membrane dye, FM4-64. Localization of truncated versions of SR3G are shown for (C) SR3G-Fragment A, (D) SR3G-Fragment B, and (E) SR3G-Fragment AB. (F) Free GFP was used as a nuclear and cytosolic marker. Scale bar = 25 µm.

sr3g mutant displays reduced growth rate for the lateral root length and Na+ content in the root.

Root system architecture analysis of Col-0 and sr3g-5 genotypes under various concentrations of NaCl are shown here. (A) Shown are representative images of 13-day-old Col-0 and sr3g-5 genotypes that experienced 9 days of salt treatment at indicated concentrations as well as growth rate for lateral root length. (B) Salt Tolerance Index (STI) for the main root length, average lateral root length, and lateral root number at 75 mM NaCl. The STI was calculated by dividing the growth rate measured under salt stress by the growth rate measured under control condition for each genotype. (C) Na+/K+ ratio in root and shoot of Col-0 and sr3g-5 after two weeks on 75 mM salt are shown. Each dot in (A) and (B) represents individual replicate per genotype. Lines in (A) and (B) graphs represent median and average, respectively. In (A) and (B), the asterisks above the graphs indicate significant differences between Col-0 and the sr3g-5 mutant, as determined by the Student ‘s t-test: *P < 0.05. Statistical analysis in (C) was done by comparison of the means for all pairs using Tukey–Kramer HSD test. Levels not connected by the same letter are significantly different (P < 0.05). Root system architecture and ICP-AES analyses of Col-0 and sr3g-5 mutant are shown in details in Figure S17.

sr3g mutant has higher growth rate and larger rosette area while displaying less ion leakage and Na+ accumulation in shoot under salt stress.

The salt stress responses of 2-week-old (referred to as “early salt stress”) or 3-week-old (referred to as “late salt stress”) soil-grown Col-0 and sr3g-5 plants that were exposed to a final concentration of 100 mM NaCl were examined here. (A) Rosette growth rate and (B) area were monitored over a period of two weeks, during which the “early salt stress” group experienced two weeks of salt stress while the “late salt stress” group experienced one week. Representative images of 4-week-old Col-0 and sr3g-5 genotypes were shown in (A). (C) Ion leakage percentage, (D) Na+ and (E) K+ contents in shoot were measured in 4-week-old plants. Each line in (A) and (B) represents the trajectory per genotype through time, where red and blue lines indicate Col-0 and sr3g-5 plants grown under control or salt-stress conditions, respectively. The dashed lines represent the mean values of the genotype per condition, whereas the grey band represents the confidence interval. The asterisks above the graphs indicate significant differences between Col-0 and sr3g-5 plants, as determined by one-way analysis of variance: *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. Statistical analysis in (C) to (E) was done by comparison of the means for all pairs using Tukey–Kramer HSD test. Levels not connected by the same letter are significantly different (P < 0.05).

sr3g mutant root is more suberized than Col-0.

(A) Representative images of fluorol yellow (FY) staining of four root sections (# 1 to 4, from root tip to root base as shown by white rectangles across the seedling) were shown for Col-0 and sr3g-5 mutant with or without 75 mM NaCl. Seeds were germinated on the 1/2 MS plates for 4 d and then seedlings were transferred to the treatment plates with or without 75 mM NaCl for 2 more days. (B) Quantification of FY signal intensity for each root section. The FY quantification was done using ImageJ. Scale bar is equal to 500 µm in all. (C) Suberin monomers detected using Gas Chromatography–Mass Spectrometry (GC-MS) in the Col-0 and sr3g-5 mutant roots with or without 75 mM NaCl included fatty acid (FA), fatty alcohol (OH-FA), α,ω-dicarboxylic acids (DCA), and ω-hydroxy fatty acid (ωOH-FA). The significant differences between Col-0 and sr3g-5 mutant were evaluated using Student ‘s t-test.

WRKY75 binds to the promoter region of the SR3G but not to its neighboring DUF247.

(A) In silico searches in CisCross (https://plamorph.sysbio.ru/ciscross/CCL_index.html) shows potential transcription factors with binding sites on the main SR3G (At3g50160) promoter region. A black-arrow points to the WRKY75 (AT5G13080). RT-qPCR showing expression of SR3G (AT3G50160) (B) and WRKY75 (AT5G13080) (C) genes, respectively, in Col-0, sr3g-5 mutant, and two different mutant alleles of the wrky75, i.e., wrky75-1 and wrky75-3. RT-qPCR analyses were conducted using seedlings grown on 1/2x MS for 4 d and then followed by transferring to the treatment plates with or without 75 mM NaCl for one week. Mean values are shown ± SE, with three biological replicates used in each condition and genotype. AT4G04120 (transposable_element_gene) was used as a reference gene for normalization. Significance was determined by the Tukey–Kramer HSD test in JMP. Levels not connected by same letter are significantly different. Yeast one-hybrid (Y1H) assay showing WRKY75 (438 bp) binds to the promoter region of the main SR3G (AT3G50160, 953 bp) (D) but not to its neighboring promoter, i.e. DUF247-150 (At3g50150, 631 bp) (E). pB42AD (AD) and pLacZ were used as effector and reporter construct, respectively. Effector and reporter constructs were co-transformed into yeast strain EGY48. Transformants were selected and grown on SD/-Trp-Ura medium. The interactions were observed on SD/- Trp/-Ura + X-Gal medium. Empty vector expressing AD domain and pLacZ were used as negative control. The two positive controls are NIGT1.4-GolS2 and NIGT1.4-GAE1. The oligo sequences used for Y1H and luciferase assay were provided in Table S4.

wrky75/sr3g double mutants roots exhibit reduced sensitivity to salt stress and accumulate lower levels of Na+ in their shoots.

(A-C) Root system architecture analysis for the Col-0, sr3g-5, wrky75-3, and two double mutants under various concentrations of NaCl are shown here. Main root growth rate (A), Lateral root number increase (B), and Lateral root length increase (C) are shown for the indicated genotypes at various salt concentrations. (D) Na+/K+ ratio in root and shoot of indicated genotypes after two weeks on treatment plates. Each dot represents individual replicate per genotype. The asterisks above the graphs (A-C) indicate significant differences between Col-0 and the genotype, as determined by the Student ‘s t-test: *P < 0.05 and **P < 0.01. ns denotes no statistically significant. Statistical analysis in (D) was done by comparison of the means for all pairs using Tukey–Kramer HSD test. Levels not connected by the same letter are significantly different (P < 0.05).