Research Adaptation to coastal soils through pleiotropic boosting of ion and stress hormone concentrations in wild Arabidopsis thaliana Silvia Busoms1,2 , Joana Teres1 , Levi Yant3 , Charlotte Poschenrieder1 and David E. Salt2 1 Plant Physiology Laboratory, Bioscience Faculty, Universitat Autonoma de Barcelona, Carrer de la Vall Moronta s/n Bellaterra, Barcelona E-08193, Spain; 2Future Food Beacon and School of Biosciences, University of Nottingham, Nottingham, NG7 2RD, UK; 3Future Food Beacon and School of Life Sciences, University of Nottingham, Nottingham, NG7 2RD, UK Summary Authors for correspondence:  Local adaptation in coastal areas is driven chiefly by tolerance to salinity stress. To survive Charlotte Poschenrieder high salinity, plants have evolved mechanisms to specifically tolerate sodium. However, the Email: Charlotte.poschenrieder@uab.es pathways that mediate adaptive changes in these conditions reach well beyond Na+. David E. Salt  Here we perform a high-resolution genetic, ionomic, and functional study of the natural Email: David.Salt@nottingham.ac.uk variation in Molybdenum transporter 1 (MOT1) associated with coastal Arabidopsis thaliana accessions. We quantify the fitness benefits of a specific deletion-harbouring allele (MOT1DEL) Received: 26 January 2021 present in coastal habitats that is associated with lower transcript expression and molybdenum Accepted: 16 June 2021 accumulation.  Analysis of the leaf ionome revealed that MOT1DEL plants accumulate more copper (Cu) and less sodium (Na+) than plants with the noncoastal MOT1 allele, revealing a complex inter- New Phytologist (2021) 232: 208–220 doi: 10.1111/nph.17569 dependence in homeostasis of these three elements. Our results indicate that under salinity stress, reduced MOT1 function limits leaf Na+ accumulation through abscisic acid (ABA) sig- nalling. Enhanced ABA biosynthesis requires Cu. This demand is met in Cu deficient coastal Key words: adaptation, ionome, salinity, soils through MOT1DEL increasing the expression of SPL7 and the copper transport protein stress signalling, structural variation. COPT6.  MOT1DEL is able to deliver a pleiotropic suite of phenotypes that enhance salinity tolerance in coastal soils deficient in Cu. This is achieved by inducing ABA biosynthesis and promoting reduced uptake or better compartmentalization of Na+, leading to coastal adaptation. is controlled by variation in MOT1, and at least some of this vari- Introduction ation may be adaptive to soil Mo concentrations (Poormoham- Our understanding of the genomic basis of adaptive variation is mad Kiani et al., 2012). MOT1 belongs to the sulphate increasing, but there is still very little known about the transporter superfamily, is localized to mitochondria, and regu- pleiotropic consequences of adaptive variation. Powerful models lates whole-plant Mo homeostasis in both Arabidopsis (Tomatsu of plant adaptation to environmental variation in mineral nutri- et al., 2007; Baxter et al., 2008) and rice (Huang et al., 2019). By ent and trace elements offer promising starting points from which dissecting the variance heterogeneity of MOT1 in detail, Forsberg to improve our understanding of these pleiotropic consequences, et al. (2015) characterized two major polymorphisms in MOT1: which arise from the interconnectedness of ion homeostasis net- a deletion in the MOT1 promoter (MOT1DEL) and a duplication works (Busoms et al., 2015, 2018; Teres et al., 2019). In particu- inside of a previous transposable element (MOT1DUP). The lar, ionomic approaches enable global contrasting of mineral duplication increases the expression of MOT1 and results in nutrient and trace element accumulation between species and higher accumulation of Mo in shoots; the deletion reduces environments, allowing broad insight into the consequences of expression of the MOT1 gene and causes a reduction in whole- adaptive variation at particular alleles. Combined with genome- plant Mo content. wide association (GWA) studies, ionomics has allowed a far- We have identified plants harbouring the reference-like MOT1 reaching assessment of adaptive, naturally evolved changes in allele (Col-0-like; MOT1C) and plants with the two structural mineral nutrient uptake and storage (Huang & Salt, 2016). variants, MOT1DEL and MOT1DUP in numerous natural demes Using such approaches, Baxter et al. (2008) identified Molyb- of Arabidopsis in the north-east of Spain (Busoms et al., 2015), denum Transporter 1 (MOT1) as the causal gene driving reduced allowing a closer look at the mechanism underlying these obser- shoot molybdenum (Mo) in two accessions of Arabidopsis vations. Interestingly, the MOT1DEL allele occurs primarily in thaliana (Van-0 and Ler-0). Forsberg et al. (2015) showed that coastal environments, and based on this distribution we suggest variation in shoot Mo content across 283 Arabidopsis accessions that MOT1 loss-of-function may control tolerance to elevated 208 New Phytologist (2021) 232: 208–220 Ó 2021 The Authors www.newphytologist.com New Phytologist Ó 2021 New Phytologist Foundation This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. New Phytologist Research 209 soil salinity at the coast. Several studies have shown that the Soil and leaf elemental analysis application of Mo and copper (Cu) to soil increases fitness under salt stress, reducing sodium (Na+) content and increasing both For soil analysis, we performed three independent soil analyses abscisic acid (ABA) biosynthesis and the ability of plants to take per site. To characterize the elemental composition of soil, analy- in essential macronutrients such as potassium (K+) (e.g. Eskan- ses were performed on the 2-mm fraction samples following the dari & Mozaffari, 2014; Wu et al., 2017). Forsberg et al. (2015) extraction method described in Busoms et al. (2018). Plants from detected an association with variation in leaf Mo, independent of the field or the laboratory were sampled by removing 2–3 leaves MOT1, c. 600 kb upstream of MOT1. Using a T-DNA knockout (1–5 mg dry weight) and washed with ultrapure (18 MΩ) water allele, they proposed Copper Transporter 6 (COPT6) as a promis- before being placed in Pyrex digestion tubes. The digestion ing candidate gene for this association, suggesting a concrete method and calibration are described in Busoms et al. (2018). genetic link between Mo and Cu homeostasis. A further connection between Cu and Mo homeostasis is given Single nucleotide polymorphism (SNP) genotyping and by the fact that Cu is essential for ABA biosynthesis, because Cu sequencing of MOT1 is needed for the synthesis of the molybdenum cofactor Moco (Schwarz & Mendel, 2006). Abscisic acid plays a key role in plant DNA from leaf material was extracted following the method responses to abiotic stress such as drought and salinity (Zhang detailed in Supporting Information Methods S1. An SSR marker et al., 2006). Under stress conditions, tolerant plants accumulate (Simple Sequence Repeat) was developed based on the deletion ABA, inducing stress-responsive genes. The main site of Na+ tox- in the promoter of MOT1 in Van-0 plants (CS1584) (Baxter icity is usually the leaf blade; therefore, the activation of genes et al., 2008) with forward primer 50 -CTCCGGTTATCGC promoting the compartmentalization of Na+ before it reaches the GTTGTAT-30 and reverse primer 50 -ACTGTCGCCATCA leaves is crucial for salinity tolerance (Munns & Tester, 2008). AGGTTTT-30 . We visualized, aligned, and blasted sequences We hypothesized that MOT1 DEL has a role in local adaptation from the NCBI database with GENEIOUS v.10.0.9 (Kearse et al., to coastal saline habitats by enhancing COPT6 expression, Cu 2012). uptake, Moco and ABA biosynthesis, resulting in less leaf Na+ accumulation, thus allowing salinity stress resilience. To test this, Field-based reciprocal common garden experiments we performed high-resolution, multiyear experiments in the field and under lab-controlled conditions, focusing on the complex After genotyping plants from 2012 and 2013 collections, we interplay between Mo-Cu-Na+ and their transporters. This set of selected two demes containing plants with the MOT1C allele and laboratory and field-based experiments has led us to conclude plants with the MOT1DEL allele named T9 and LLO2. In March that the naturally occurring MOT1DEL allele has an adaptive 2014 and March 2015, 100 seeds of each deme/MOT1 variant advantage in coastal conditions, providing enhanced salinity tol- (T9C, T9DEL; LLO2C, LLO2DEL) were sown at Blanes coastal erance (NaCl) through the pleiotropic enhancement of both Cu field (lat. 41°400 37.64″, long. 2°480 3,86″) and at Santa Coloma and ABA concentrations. de Farners inland field (lat. 41°500 41.04″, long. 2°400 36.13″) into 16 9 16 cm squares. We studied the fitness of 10 plants for each deme/MOT1 variant at each site, and the other 10 plants Materials and Methods were harvested in April 2014 and April 2015 to analyse the leaf ionome. Rosette diameters were measured every wk for 2 months Collection of plant and soil material and the number of siliques was counted at maturity. Only seeds Thirty-six demes of Arabidopsis thaliana were selected in Catalo- containing siliques were considered. The 2013 and 2014 recipro- nia (north-east Spain). A ‘deme’ is defined as a small group of cal transplants are described in Busoms et al. (2015). Data were Arabidopsis plants growing in relatively homogeneous ecological reanalysed, classifying the plants as MOT1DEL (shoot Mo2+ con- conditions and separated from other groups by at least 35 m. tent < 3.5 µg g1); MOT1DUP (shoot Mo content > 9 µg g1) or Each year (from 2013 to 2015) at each site, samples were col- MOT1C (Dataset S1). lected such that half of each collected plant was taken for ionomic analysis and the other half was used to extract DNA for MOT1 Salinity tolerance assays genotyping. Seeds of each individual were collected directly in the field and stored in packets over silica gel in a sealed box until Seeds of the mot1-2 (SALK 069683) and copt6-1 (SALK 083438) they were used. For the soil elemental composition analysis, we T-DNA insertion alleles and aba2-1 (CS 156) were obtained collected three soil samples (c. 50 g of soil from the first 10 cm from the Arabidopsis Biological Resource Center (Alonso et al., depth) at each site during the first wk of May of each year. Soil 2003). The same demes selected for the field common garden samples from the coastal and inland field sites were collected experiment, together with Col-0 and the mot1-2 mutant plants, twice per month during Arabidopsis reciprocal transplants (from were used to perform irrigation, hydroponics and salt spray February to June of 2013, 2014 and 2015). Soil was air-dried experiments with NaCl to test salinity tolerance and analyse the under laboratory conditions, passed through a 2-mm sieve, and leaf ionome (procedure and growth conditions specified in stored dry. Methods S2, S3). Ó 2021 The Authors New Phytologist (2021) 232: 208–220 New Phytologist Ó 2021 New Phytologist Foundation www.newphytologist.com New 210 Research Phytologist Genotype data of the 1135 Arabidopsis strains were downloaded or multivariate ANOVA was used to test for significant differences from http://1001genomes.org/ (1135_snp-short-indel.vcf.gz). between means of fitness, elemental contents of soil and leaf mate- MOT1 allele type classification and normalized leaf values of 18 rial, gene expression and geostatistical variables. To test for correla- ions for the 1135 strains cultivated in common soil under control tions between two variables, a bivariate fit was applied. To conditions can be found in Datasets S2, S3. Moreover, a world- perform multiple comparisons of group means we used Tukey’s wide set of 306 natural accessions was phenotyped for leaf Na+ HSD (Dataset S5) in SAS software JMP v.16.0 (https://www.jmp. content. All the accessions are listed in Dataset S4 and MOT1 type com/es_es/home.html, 2016). is indicated. The experimental design is detailed in Methods S4. For the complementation study, we quantified the shoot Mo, Results and discussion Na+ and Cu content (µg g–1 dry weight) of six plants of each type: Col-0, mot1-2, copt6-1, T9C, T9DEL, LLO2C and LLO2DEL and Previous work has provided evidence that Arabidopsis popula- F1s from crosses between T9C 9 T9DEL, LLO2DEL 9 LLO2C and tions from north-east Spain harbour substantial genetic variabil- T9DEL 9 LLO2DEL cultivated in common soil irrigated with half- ity and adaptive variation to elevated salinity in coastal soils at strength Hoagland solution (pH 6.0) with 50 mM NaCl for 4 wk. high geographic and genomic resolution (Busoms et al., 2015, 2018). Here we build on these works, focussing on novel struc- tural variation in Molybdenum Transporter MOT1 and the adap- Expression (quantitative real-time polymerase chain tive consequences of this variation. reaction (qRT-PCR)) analysis RNA from leaf and root material was extracted following the method MOT1 allelic variation and distribution described in Methods S1. Primers used for MOT1 (At2g25680), COPT6 (At2g26975), SPL7 (At5g18830), CNX6 (At2g43760), Genome-wide association studies indicate that the MOT1 locus CNX1 (At5g20990) and ABA3 (At1g16540) transcript quantifica- exhibits contrasting effects on leaf Mo accumulation through dif- tion are detailed in Table S1. For normalization across samples, the ferent natural alleles (Forsbert et al., 2015; Baxter et al., 2008). expression of the ACTIN 1 gene (At2g37620) was analysed. For each Tomatsu et al. (2007) and Baxter et al. (2008) first detected a 53 sample, the average value from triplicate real-time PCRs was used to bp deletion on the MOT1 promoter associated with reduced estimate transcript abundance. Data were analysed using the SDS MOT1 expression and decreased shoot Mo content (MOT1DEL). v.1.0 software (Applied Biosystems 7900HT Fast Real-Time PCR Forsberg et al. (2015) also found independent polymorphisms System, Thermo Fisher Scientific - ES; https://www.thermofisher.c which were associated with increased MOT1 expression and ele- om/es/es/home/brands/applied-biosystems.html). Threshold cycle vated shoot Mo concentration: two forms of a 330-bp-long (Ct) values were determined based on efficiency of amplification. duplication in a transposable element sequence, AT2TE47050 The mean Ct values were normalized against Actin1 gene and Ct val- (MOT1DUP); and two SNP markers, one located c. 25 kb down- ues, calculated as (CtGene–CtActin1). The root and leaf expression of stream (MOT1-SNP1) and another located c. 600 kb upstream of MOT1 was calculated using the 2DCt method. The relative expres- the MOT1 coding locus (MOT1-SNP2). sion of each target gene was calculated using the 2DDCt method To determine which MOT1 alleles are present in wild popula- (NaCl treatment – Control). tions of Arabidopsis adapted to high soil salinity and naturally heterogenous soils on the north-east coast of Spain (Busoms et al., 2015, 2018), we performed a high-resolution survey over 3 Abscisic acid quantification consecutive years (2013–2015). We collected, genotyped and Abscisic acid concentration (nmol l1) was quantified by indirect quantified shoot Mo content from 36 demes (small groups of enzyme-linked assay (ELISA) using a Phytodetek ABA Test Kit plants growing in relatively homogeneous ecological conditions; (Agdia, Elkhart, IN, USA) on three plants per treatment from T9 see the sub-section ‘Collection of plant and soil material’, above) (Mix), T1 (MOT1DEL) and PA10 (MOT1C) demes, along with in a region c. 60 km across (Fig. 1a). This yielded 25 demes con- Col-0 and mot1-2, copt6-1 and aba2-1 mutants cultivated in taining only plants with the MOT1 Col-0 reference-like allele common soil irrigated with 0 and 50 mM of NaCl for 2 wk. MOT1C (Fig. 1a: black dots, coastal; grey dots, inland), while the rest of the demes harboured either mixtures of the other alleles or only MOT1DEL (Fig. 1a: green dots). Additionally, there were Statistical analysis three demes located in the north of the area that contained mix- For the hierarchical clustering, we generated a progressive align- tures of plants with either MOT1C or MOT1DUP alleles (Fig. 1a: ment of 262 SNPs for the MOT1 gene tree and 37 574 SNPs for red/black dots). The overall distribution of MOT1 alleles was the whole-genome tree from 74 plants from Busoms et al. (2018). nonrandom across the study area: while we did not detect a corre- Pairwise genetic distance between individuals and between demes lation with soil Mo concentrations (r2 = 0.087), we detected a was calculated using the Maximum Likelihood statistical method significant association between the presence of MOT1DEL and and the Jukes–Cantor substitution model using JMP v.13.0 (SAS, the distance to the sea (ChiSq = 324,83; P < 0.001), with the 2016). Mean-standardized values (1< value >1) of elemental demes containing only MOT1DEL allele being close to the coast contents of soil and leaf material were used to represent the radar (Fig. 1b), suggesting a potential adaptive value of this allele in plots and compare MOT1 allelic variants (Dataset S2). One-way coastal environments. Moreover, we did not detect MOT1DEL New Phytologist (2021) 232: 208–220 Ó 2021 The Authors www.newphytologist.com New Phytologist Ó 2021 New Phytologist Foundation New Phytologist Research 211 (a) (b) V3 0.12 2 V1 PO1 R = 0.087 Soil [Mo] (mg g–1) RO2 0.08 A1 A5 0.04 LM2 I LG7 SCF 0 LG5 T13 PA10 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 T9 SFG9 C T6 Log (distance to sea, m) JBB T11 T2 C - Mot1 C I - Mot1 C LLO Mot1 DEL Mot1 DUP 0 km 10 (c) 1093: 3255 4350 4465 4561 4814 5136 DEL Mot1 D429V I68T V33L Del DUP Mot1 Dup mot1-2 10935152 tDNA ins C Mot1 MOT1 CDS AT2TE47050 (d) PA10-1 PA10-2 PA10-3 PA10-4 LM2-1 LM2-2 LM2-3 LM2-4 PO1-3 PO1-4 PO1-2 PO1-1 RO2-3 RO2-1 RO2-2 RO2-4 LLO-2 LLO-3 LLO-4 LLO-1 SFG-1 SFG-2 SFG-3 SFG-4 LG7-1 LG7-2 LG7-3 LG7-4 LG5-1 LG5-2 LG5-3 LG5-4 JBBT1 JBBT2 SCF-1 SCF-3 SCF-4 SCF-2 T11-1 T11-3 T11-4 T11-2 T13-3 T13-1 T13-2 T13-4 A1-2 A1-4 A5-1 A5-2 A5-3 A5-4 A1-1 A1-3 V3-2 V3-3 V3-4 V1-2 V1-3 V1-1 V1-4 V3-1 T6-1 T6-2 T6-3 T6-4 T2-1 T2-2 T2-3 T2-4 T9-1 T9-2 T9-4 (e) PA10-1 PA10-2 PA10-4 PA10-3 LM2-2 LM2-3 LM2-1 PO1-1 PO1-3 PO1-4 PO1-2 RO2-1 RO2-2 RO2-4 RO2-3 LLO-2 LLO-3 LLO-4 LLO-1 LG5-2 LG5-3 LG5-1 LG5-4 SFG-1 SFG-2 SFG-4 SFG-3 LG7-1 LG7-2 LG7-4 LG7-3 JBBT1 JBBT2 SCF-3 SCF-4 SCF-2 SCF-1 T11-1 T11-2 T11-3 T11-4 T13-3 T13-2 T13-1 T13-4 LM2- A1-2 A1-4 A1-1 A1-3 A5-1 A5-3 A5-2 A5-4 V1-4 V1-1 V3-1 V3-2 V3-3 V3-4 V1-2 V1-3 T6-1 T6-2 T6-3 T6-4 T2-1 T2-2 T2-3 T2-4 T9-4 T9-1 T9-2 T9-3 * * Fig. 1 Sampling, MOT1 alleles studied, genetic structure, and relationship with soil molybdenum (Mo) concentrations. (a) Geographic location in Catalonia, north-east Spain of the 36 demes of Arabidopsis studied, with the 19 sequenced demes named. (b) Distribution of each MOT1 allele as a function of distance to the sea and soil Mo concentration. The correlation between soil Mo and distance to the sea (dotted line) is not significant (r2 = 0.087). Black dots, coastal demes consisting of MOT1C; grey, inland demes consisting of MOT1C; green, coastal demes consisting of MOT1DEL; black/green, demes harbouring both MOT1C and MOT1DEL; black/red, demes harbouring both MOT1C and MOT1DUP. The rhombus symbols represent the two field sites, Blanes (coastal, blue rhombus) and Santa Coloma de Farners (inland, orange rhombus). (c) Sequence alignment of the MOT1 locus for MOT1C, MOT1DEL, MOT1DUP alleles and mot1-2 mutant (SALK_069683). Genomic positions, MOT1 CDS, AT2TE47050 and polymorphisms are indicated. (d, e) Neighbour-joining cladograms of the MOT1 allele (based on 262 SNPs) (d) and the whole-genome fourfold degenerate (putatively neutral) sites (37,574 SNPs) (e) among 74 Arabidopsis individuals from our study region in northeast Spain. In MOT1, the highest probability for the value of K = 3 classifies the three versions of the MOT1 allele: MOT1C (black), MOT1DEL (green), MOT1DUP (red). The same colours have been used on the whole-genome cladogram, with asterisks (*) denoting individuals with the MOT1C allele from demes containing plants with the MOT1DEL allele. Ó 2021 The Authors New Phytologist (2021) 232: 208–220 New Phytologist Ó 2021 New Phytologist Foundation www.newphytologist.com New 212 Research Phytologist plants in any inland deme, suggesting a possible cost of the allele alternative alleles at the inland site (Fig. 2b), confirming coastal- under inland conditions. specific adaptation of MOT1DEL harbouring plants. We have We analysed the MOT1 locus in 74 resequenced individuals been unable to identify plants harbouring the MOT1DEL in our from 19 demes chosen to represent the full range of ecotypic and inland sampling, suggesting a potential fitness cost for this allele genetic variation over this area (Dataset S1 in Busoms et al. in inland environments. However, given our benign common (2018)). We also performed MOT1 transcript expression analysis garden field conditions, it is perhaps not surprising that we do and quantified leaf Mo concentrations across selected demes not detect a fitness cost over a single generation. That said, we do (Fig. S1). In MOT1DEL individuals we observed a 53 bp deletion see differences in fitness proxies between plants harbouring each 13 bp upstream of the transcription start site; we also detected an MOT1 allele as a function of salinity challenge (Fig. 3a,b), with a additional nonsynonymous change in the coding region (Fig. 1c), lack of sodium challenge associated with lower fitness in the and we confirmed the low leaf Mo content and the reduced MOT1DEL allele, supporting the idea of a fitness effect. These expression of MOT1 in both roots and leaves (Fig. S1). Individu- results underscore a strongly environmental context-dependence als with the MOT1DUP allele have a 322 bp duplication inside of on fitness estimations of naturally evolved alleles. the transposable element AT2TE47050 (Fig. 1c) and a higher During the 3 yr of the reciprocal transplants, we monitored expression of MOT1, especially in roots, and the plants accumu- the soil mineral nutrient and trace element content at each field late higher levels of Mo in their leaves (Fig. S1). Although Mo site (Dataset S2). Inland and coastal soils consistently differed leaf content in MOT1DEL individuals is reduced, plants do not in Na+, magnesium (Mg), phosphorus (P), zinc (Zn), Cu, Mo show symptoms of Mo deficiency: Mo concentrations are within and rubidium (Rb) concentrations (Fig. 2c). However, the accu- the normal range (0.4–1.0 µg g1) for Brassicaceae species (Men- mulation of these elements in plant tissues did not fully reflect gel et al., 2001). the soil quantities. Instead, a difference emerged specifically To gain an understanding of the population genomic relation- related to Mo. At all sites MOT1DUP plants had high shoot Mo, ships among plants harbouring each of the three major MOT1 while MOT1DEL plants had not only lower Mo, but also alleles, we generated a neighbour-joining (NJ) cladogram of the decreased Na+ (Fig. 2d). MOT1DEL plants were also able to MOT1 locus from our 74 resequenced individuals and compared uptake more Cu, P and K+ than MOT1C plants under coastal it to the genome-wide cladogram. This group consisted of 22 conditions (Fig. 2d,e). These results indicate that the allelic vari- individuals with MOT1DEL; six with MOT1DUP; and 46 with the ation present in MOT1 not only results in reduced Mo accumu- MOT1C allele (Fig. 1d,e). As predicted, the MOT1 locus clado- lation in the leaves, but is also linked to mechanisms that allow gram differentiates the MOT1 alleles into three distinct clades MOT1DEL plants to exclude Na+. Furthermore, we can affirm (Fig. 1d). Except for the LLO-1 individual that genetically clus- that this low Mo concentration in MOT1DEL plants does not ters with the individuals from the A5 deme (suggesting a rare case adversely affect their normal development, as these plants had of migration due to the demes being > 40 km apart, as previously better or equal growth and reproductive fitness compared to discussed in Busoms et al. (2018)), Fig. 1(e) shows that genomes MOT1C plants when cultivated under natural field conditions, harbouring MOT1DEL or MOT1DUP are nested among MOT1C as measured by silique number (Fig. 2b). sisters, indicating that different MOT1 alleles have recombined onto common genetic backgrounds. MOT1 interaction with Na+ and Cu homeostasis To confirm that the MOT1DEL allele can mediate a salinity resis- MOT1 influence on coastal adaptation tance phenotype whilst maintaining low Mo and Na+ contents Despite our dense sampling, we could only find the MOT1DEL and high Cu content in leaves, we performed controlled environ- allele in demes located < 3 km from the coast, suggesting that this ment experiments. We modulated NaCl concentrations in both allele may have a role in adaptation to coastal environments. To soil and hydroponics using the same demes we had cultivated in test this idea, we reanalysed data from reciprocal transplants per- the field (T9C, T9DEL; LLO2C, LLO2DEL). Without the addition formed in coastal and inland field sites in 2013 and 2014 of NaCl, plants containing MOT1C or MOT1DEL alleles grew (Busoms et al., 2015), considering the status of the MOT1 allele similarly. After treatment with 50 mM or 100 mM NaCl for 3 version in each individual plant. Fig. 2(a) shows that plants har- wk, growth of all plants was reduced. However, plants with the bouring MOT1DEL are consistently better adapted to coastal con- MOT1DEL allele and the mot1-2 mutant grew significantly better ditions, often producing over twice as many siliques when grown than plants with the MOT1C allele in high NaCl conditions (Fig. at the coast and outperforming the rest of the coastal demes. a). This enhanced salinity tolerance linked to low MOT1 expres- To isolate the effect of the MOT1 allele and homogenise sion was also reflected in fitness. Both the mot1-2 mutant and genetic backgrounds, we selected two mixed demes with plants MOT1DEL plants cultivated in the same soil produced more harbouring MOT1C or DEL alleles (T9C, T9DEL; LLO2C, siliques than Col-0 and MOT1C plants when they were irrigated LLO2DEL) for further reciprocal transplant experiments at the with NaCl at any concentration (Fig. 3b). Ionomic data from same coastal and inland sites. In both years (2014 and 2015), these experiments confirmed that plants with the MOT1DEL plants with the MOT1C allele from both demes were less fit than allele subjected to salinity stress accumulate more Cu and less plants with the MOT1DEL allele when grown at the coast. How- Na+ in their leaves than plants with the MOT1C allele under the ever, no difference was observed between plants with the two same conditions (Fig. S2; Dataset S2). New Phytologist (2021) 232: 208–220 Ó 2021 The Authors www.newphytologist.com New Phytologist Ó 2021 New Phytologist Foundation New Phytologist Research 213 (a) (b) 2014 2015 a 2013 2014 450 500 b a LLO2 a 300 * Silique number * Silique number 400 b * b a a * a 150 a b a aa a 300 * c * * 0 b c c c a 200 a 450 b T9 100 cc 300 b * * a a * 150 a b a Coastal Inland Coastal Inland a 0 Coastal Inland Coastal Inland (c) Soil 2013 Soil 2014 Soil 2015 Field site Li Coastal Cd 0.5 Na* Cd 0.5 Li Na* Cd 0.5 Li Na* Inland Mo* 0.2 Mg* Mo* 0.2 Mg* Mo* 0.2 Mg* Sr -0.1 P* -0.1 -0.1 Sr* P* Sr* P* -0.4 -0.4 -0.4 Rb K Rb K Rb K* -0.7 -0.7 -0.7 Se Ca Se* Ca Se* Ca As Mn As Mn As Mn Zn Fe Zn* Fe Zn* Fe Cu* Co* Cu* Co Cu* Co Ni Ni Ni (d) Leaf 2013 Leaf 2014 (e) Leaf 2014 Leaf 2015 Cd 0.5 Li Na* 0.5 Li* Na* Li Li Cd Cd 0.5 Na* Cd 0.5 Na* Mo* 0.3 Mg Mo* 0.3 Mg 0.3 0.3 Mo* Mg Mo* Mg 0.0 0.0 0.0 0.0 Sr P* Sr P* Sr P* Sr P* -0.3 -0.3 -0.3 -0.3 Rb K* Rb -0.5 K* Rb K Rb K -0.5 -0.5 -0.5 Se Ca Se* Ca Se Ca Se Ca As Mn As Mn As Mn As* Mn Zn Fe* Zn Fe Zn Fe Zn Fe* Cu* Co Cu* Co Cu* Co Cu* Co Coastal Coastal 0.5 Li* Na Li Li Li Cd Cd 0.5 Na* Cd 0.5 Na* 0.5 Na* 0.3 Cd Mo* Mg Mo* 0.3 Mg 0.3 0.3 Mo* Mg Mo* Mg 0.0 0.0 Sr P Sr P 0.0 0.0 -0.3 Sr P Sr P -0.3 -0.3 -0.3 Rb -0.5 K Rb K Rb K -0.5 -0.5 Rb -0.5 K Se Ca Se Ca Se Ca Se Ca As Mn* As Mn Mn As As Mn Zn Fe Zn Fe* Zn Fe Zn Fe Cu* Co Cu* Co Cu* Co Inland Inland Cu* Co Fig. 2 Local adaptation of MOT1 alleles and global elemental accumulation analysis. Fitness (mean  SE of silique number) of (a) Arabidopsis plants from inland demes harbouring the MOT1C (grey) allele, and from coastal demes harbouring MOT1C (black), MOT1DEL (green) or MOT1DUP (red) alleles, cultivated at coastal and inland common gardens in 2013 (n = 359) and 2014 (n = 400); (b) plants from LLO2 and T9 demes harbouring the MOT1C (black) or MOT1DEL (green) alleles cultivated at coastal and inland common gardens in 2014 (n = 40) and 2015 (n = 60). Significant differences are indicated by letters (between alleles) and asterisks (between sites). Normalized difference of 17 elements in (c) soils from coastal (blue) and inland (orange) common gardens collected in the 3 yr of cultivations (n = 60); (d) leaves of plants harbouring the MOT1C (black), MOT1DEL (green) or MOT1DUP (red) alleles from 2013 (n = 293) and 2014 (n = 390) reciprocal transplants; and (e) leaves of plants from LLO2 and T9 demes harbouring the MOT1C (black) or the MOT1DEL (green) alleles from 2014 (n = 40) and 2015 (n = 40) reciprocal transplants. Elements exhibiting significant differences (according to a t-test) are marked with an asterisk (*, P < 0.05). Ó 2021 The Authors New Phytologist (2021) 232: 208–220 New Phytologist Ó 2021 New Phytologist Foundation www.newphytologist.com New 214 Research Phytologist (a) Fig. 3 Interplay of MOT1 and salinity tolerance. (a) Fresh biomass (g) of Arabidopsis plants being exposed to either 0, 50 or 100 mM of NaCl in LLO2 T9 Col-0 hydroponic solution for 3 wk and (b) fitness (number of siliques produced) 3.0 Fresh biomass (g) MOT1 DEL MOT1 DEL WT of plants cultivated in potting mix soil under controlled conditions and MOT1 C MOT1 C *** mot1-2 2.3 irrigated with 0, 50 and 100 mM of NaCl until maturity. Data represent the * mean  SE of plants with the MOT1C (black) allele, plants with the MOT1DEL (green) allele from LLO2 and T9 demes, and mot1-2 (yellow) 1.5 *** and Col-0 (blue) individuals (10 plants per accession). Asterisks indicate ** ** significant differences (t-test; *, P < 0.05; **, P < 0.005; ***, P < 0.0005). ** *** 0.8 (c, d) Mean  SE of (c) relative dry weight (DW treatment/DW control) and (d) survival (in days) after a high salinity treatment of plants with the 0.0 MOT1C allele (black), plants with the MOT1DEL allele (green), and Col-0 0 50 100 0 50 100 0 50 100 (blue) and mot1-2 (yellow) plants cultivated in potting mix soil irrigated NaCl (mM) with no NaCl (Control), 75 mM NaCl (Salt) or 150 mM NaCl (High-Salt) (b) and sprayed with no NaCl (CS) or 150 mM NaCl (SS) for 2 wk. Bars represent the mean  SE and boxplots represent the median and LLO2 T9 Col-0 interquartile range of six plants per accession. Letters indicate significant 140 MOT1 DEL MOT1 DEL ** WT differences (Tukey’s HSD, P < 0.05). Silique number 120 MOT1 C MOT1 C mot1-2 100 80 were interested in understanding whether the MOTDEL can also 60 ** ** confer elevated tolerance to salinity in the form of exposure to 40 *** NaCl applied as a spray, to simulate salt spray from the sea (Du ** ** 20 & Hesp, 2020). We therefore cultivated the same demes from 0 the previous experiments in potting mix soil, and 3 wk after sow- 0 50 100 0 50 100 0 50 100 ing we started five different treatments: Control-CS (plants irri- NaCl (mM) gated and sprayed with 0 mM of NaCl), Salt-CS (plants irrigated (c) with 75 mM and sprayed with 0 mM of NaCl), Control-SS 1.0 (plants irrigated with 0 mM and sprayed with 150 mM of NaCl), a a Salt-SS (plants irrigated with 75 mM and sprayed with 150 mM Relative dry weight a a a 0.8 of NaCl), and High-Salt-SS (plants irrigated with 150 mM and b b ab sprayed with 150 mM of NaCl). After 2 wk of treatment, plants 0.6 b b b b bc bc from the first four treatments were harvested, dried and weighed. c c Plants from the fifth treatment were visually inspected to record 0.4 c d survival. Compared to control plants, coastal demes with the 0.2 MOT1DEL allele showed high tolerance to soil salinity, salt spray and the combination of both (Fig. 3c,d). Col-0 plants were 0.0 highly sensitive to salt spray, dying a few days after the start of Salt-CS Control-SS Salt-SS the High-Salt-SS treatment. By contrast, mot1-2 plants grew less than MOT1DEL individuals but always performed better than High-Salt-SS Col-0 in all the salinity treatments, exhibiting a similar tolerance (d) to plants with the MOT1C allele (Fig. 3c,d). a 20 To functionally confirm the association between the MOT1 a allele and salinity in a broad diversity panel, we challenged a set Days of survival 15 b b of 306 natural accessions (14 with the MOT1DEL allele) with 50 c mM NaCl 5 wk after sowing, and further added 25 mM NaCl each following wk for 2 wk. We measured leaf Na+ content 10 before and after the treatment. This confirmed for a much d broader sample that MOT1DEL plants accumulate less Na+ in 5 aerial tissues when Na+ solution concentrations are high (Fig. 4a). LLO2C LLO2DEL T9 C T9 DEL Col-0 mot1-2 To further test whether the MOT1 deletion is related to Na+ exclusion, we analysed the leaf ionome profiles of 953 accessions from the 1135 Arabidopsis collection (Campos et al., 2021) after We have previously established that coastal adaptation in the growth in nonsaline potting mix (Dataset S2). We classified them Arabidopsis demes we are studying here is driven primarily by into three groups (MOT1C: black; MOT1DEL: green; MOT1DUP: adaptation to elevated soil salinity by assessing the fitness of red) and located them on an online map (Fig. S3a). Geographical plants grown in a controlled environment in excavated coastal variables (‘Country’ and ‘Distance to sea’), and 18 estimated soil and inland soil (Busoms et al., 2015). However, in addition, we parameters based on Land Use and Coverage Area frame Survey New Phytologist (2021) 232: 208–220 Ó 2021 The Authors www.newphytologist.com New Phytologist Ó 2021 New Phytologist Foundation New Phytologist Research 215 (a) (b) Leaf ionome 306 A. thaliana accessions 953 A. thaliana accessions Leaf [Na] (μg g–1 DW) ** Li 6000 Cd 0,3 Na* Mo* 0,1 Mg 4500 Sr -0,1 P 3000 Rb K* -0,3 1500 Se Ca 0 As Mn Pre-NaCl Post-NaCl Zn Fe Cu Co (c) Leaf ionome, 50 mM NaCl a –1 DW) 10 b b 8 6 c c 4 d 2 d d d d 0 a bcd abc –1 DW) 2000 abc ab bcde 1500 cde de de e 1000 500 14 a –1 DW) a 12 ab bc ab 10 c c bc 8 c 6 c 4 L T9 C O2 C L C L C L 1 DE 2 DE -0 2 DE E 6- T9 D T9 D 1- O2 l T9 Co pt ot O2 T9 L LL LO co xL m LL L x DE xL E L O2 LL Fig. 4 Allelic effects on ion content. (a) Mean  SE of sodium (Na+) content in leaves of plants from 306 European Arabidopsis accessions (14 MOT1DEL (green); 292 MOT1C (black)) cultivated in control conditions for 5 wk, and leaves from the same plants after being irrigated with NaCl for 4 wk. Asterisks indicate significant differences between MOT1 alleles (t-test; **, P < 0.005). (b) Normalized difference of 17 elements in leaves of 953 accessions from the Arabidopsis world-wide collection of 1135, classified into plants harbouring the MOT1C (black), the MOT1DEL (green) or the MOT1DUP (red) alleles. (c) Mean  SE of shoot molybdenum (Mo), Na+ and copper (Cu) content (µg g1 dry weight) of Col-0, T9C, T9DEL, LLO2C and LLO2DEL parental plants, F1 plants from crosses between T9C 9 T9DEL, LLO2DEL 9 LLO2C and T9DEL 9 LLO2DEL and mot1-2 and copt6-1 mutants cultivated in potting mix soil irrigated with 50 mM of NaCl for 4 wk. Boxplots represent the median and interquartile range of six plants per accession, and letters indicate significant differences (F-test, P < 0.05). (LUCAS) 2009/2012 topsoil data (Panagos et al., 2012) were conditions (Fig. 4b), indicating a world-wide effect of the pres- tested. We could discern no clear association between the pres- ence of these MOT1 alleles on the accumulation of Mo and Na+. ence of MOT1 allelic variation and these variables (Fig. S3b). We also observed elevated K+ associated with MOT1DEL. However, leaf Mo and Na+ content were significantly lower in The main interaction between Mo and Cu occurs via the syn- plants with the MOT1DEL allele when cultivated under common thesis of the Mo containing cofactor Moco, where COPT Ó 2021 The Authors New Phytologist (2021) 232: 208–220 New Phytologist Ó 2021 New Phytologist Foundation www.newphytologist.com New 216 Research Phytologist transporters play a role. In particular, the Cu transporter COPT6 concentrations and an enhanced ABA response. We therefore has been associated with MOT1 (Billard et al., 2014; Forsberg quantified the expression of the main genes involved in Moco et al., 2015). To examine whether polymorphisms in MOT1 and biosynthesis, a critical cofactor in ABA biosynthesis, and ABA COPT6 independently affect Mo and Cu concentrations in concentrations in plants with various MOT1 alleles submitted to planta, we compared Mo and Cu concentrations in mot1-2 and salt stress. copt6-1 plants cultivated in common soil irrigated with 50 mM The formation of active Mo containing enzymes depends not NaCl for 3 wk (Fig. 4c). In both mot1-2 and copt6-1 mutants, only on the availability of Mo but also on the presence of the two leaf Mo contents are low in comparison to the wild-type Col-0. metals iron (Fe) and Cu (Vigani et al., 2017). It is assumed that Moreover, the mot1-2 mutant exhibited elevated leaf Cu levels, Cu binds to the dithiolate group of molybdopterin (MPT) at the suggesting a potential link between these two genes (Fig. 4c). end of step 2 of Moco biosynthesis (Mendel, 2007). In this sec- To further confirm that the MOT1 locus is responsible for the ond step, MPT is initially formed by the incorporation of two ionomic changes detected in the Catalonian demes from north- sulfur atoms into cPMP in a reaction catalyzed by MPT synthase, east Spain, we performed a complementation test, crossing a heterotetrameric complex of Cnx6 and Cnx7 subunits. The two LLO2DEL 9 LLO2C, T9C 9 T9DEL and T9DEL 9 LLO2DEL. F1 small Cnx7 subunits are responsible for S transfer into the com- and parental plants were cultivated together with mot1-2 and plex, but the MPT formation is mainly regulated by CNX6 (Ide copt6-1 mutants. F1 plants from the LLO2DEL 9 LLO2C, T9C et al., 2011). In the next step, MPT is adenylated by Mo insertase 9 T9DEL crosses contained intermediate concentrations of shoot CNX1, forming MPT-AMP. In the last step, MPT-AMP is Mo, Na+ and Cu relative to the parents (Dataset S2), indicating transferred to the Cnx1 E-domain, where Mo is inserted by complementation. Only the F1s from the T9DEL 9 LLO2DEL replacing Cu, forming Moco, which is finally allocated to Mo cross had significantly lower shoot Mo and Na+ (P < 0.005) and apoenzymes (Kaufholdt et al., 2013), such as aldehyde oxidase higher Cu, similar to the MOT1DEL parents (Fig. 4c), confirming (AO). ABA3 is an enzyme required for the enzymatic activity of that T9DEL and LLO2DEL are allelic with the recessive loss of AO, the enzyme essential for the biosynthesis of bioactive ABA function MOT1 allele. (Watanabe et al., 2018). We observed that salt stress increased the expression of CNX6, CNX1 and ABA3 in all genotypes, but they were strongly upregulated in the mot1-2 mutant and MOT1 link to abscisic acid signalling and salinity tolerance MOT1DEL plants (Fig. 5c). Consequently, ABA concentrations Soil ionome analyses showed that our coastal soils, as well as were also higher in MOT1DEL plants after salinity treatment being rich in Na+ due to the marine influence (Busoms et al., (Fig. 5d), supporting the hypothesis that the MOT1DEL allele 2015), contain lower concentrations of Cu compared to inland provides an enhanced ABA response to saline challenge. soils (Fig. 5a). Such levels are not considered critical for normal We found that ABA concentrations in copt6-1 mutants were plant growth but we detected that leaf Cu concentrations in the significantly lower in all treatments (Fig. 5d), supporting the role majority of plants with the MOT1C allele growing at the coast is of Cu in ABA biosynthesis. Under salinity stress, plants harbour- low (between 1 and 7 lg g1), concentrations that may indicate ing the MOT1DEL allele showed the highest induction of ABA Cu deficiency (Marschner, 1995). However, coastal plants with biosynthesis (Fig. 5d). This may allow MOT1DEL plants to main- the MOT1DEL allele are able to accumulate more Cu inside the tain their normal growth, even after being treated with 100 mM plant (Fig. 5b). Expression of SPL7 and COPT6 are elevated in of NaCl (Fig. S4b). plants containing MOT1DEL, suggesting that these genes are At the cellular level, both Mo and Cu are required for Moco playing a role in this elevated leaf Cu (Fig. 5c). This is also sup- biosynthesis, but the amount of each element can act either posi- ported by the enhanced leaf Cu and SPL7 expression we observed tively or negatively during this process (Pe~narrubia et al., 2015). in the mot1-2 mutant (Figs 4c, 5c). SPL7 activates the transcrip- It has been reported that CNX1 binds to the cytoskeleton (Sch- tion of multiple genes involved in Cu homeostasis in Arabidopsis warz et al., 2000) and could have a role as a Mo or Cu sensor, (Yamasaki et al., 2009; Jung et al., 2012), and the COPT6 plasma interacting with MOT1 or COPT transporters to regulate Mo membrane transporter has previously been shown to be induced and Cu concentrations in the cell (Tejada-Jimenez et al., 2009). under Cu deficiency conditions (Garcia-Molina et al., 2013; Sulfurylated Moco is a cofactor of ABA–AO that functions in the Puig, 2014; Pe~ narrubia et al., 2015). As a consequence, last step of ABA biosynthesis, and increased ABA concentrations MOT1DEL plants can maintain leaf Cu concentrations within an being involved in responses to salinity has been well documented adequate range despite growing in soils with lower Cu concentra- (e.g. Park et al., 2016). Therefore, the crosstalk between Mo-Cu- tions, in contrast to MOT1C plants. Na+ via ABA synthesis may explain the greater tolerance to salin- COPT6 is localized mainly in the vasculature of green tissues ity of our plants harbouring the MOT1DEL allele (Fig. 6). where it can facilitate Cu redistribution (Garcia-Molina et al., 2013) and also plays an important role in ABA biosynthesis. Conclusions Abscisic acid regulates SPL7 and subsequently affects the expres- sion of its COPT targets (Carrio-Seguı et al., 2016). Noting the Complex interactions occur among the homeostatic controls of potential relationship between COPT6 and MOT1 (Forsberg various nutrients in plants (Kearse et al., 2012). However, a com- et al., 2015), we hypothesized that MOT1DEL may mediate a prehensive understanding of the interactions between pathways degree of salinity tolerance at the coast due to increased Cu integrating the homeostasis of key elements is still lacking, New Phytologist (2021) 232: 208–220 Ó 2021 The Authors www.newphytologist.com New Phytologist Ó 2021 New Phytologist Foundation New Phytologist Research 217 (a) (mg g–1) (mg g–1) 100 10 0.05 (mg g–1) a ab a a ab 80 b 8 0.04 ab b b 60 6 0.03 c c 40 4 0.02 c 20 2 0.01 (b) 0 0 0 (μg g–1) 1600 a a 14 a (μg g–1) b (μg g–1) 10 ab 12 1200 ba b 8 a 10 a 6 c 8 800 b 6 b 4 b 400 4 c 2 2 c 0 0 0 MOT1 C MOT1 DEL MOT1DUP MOT1 C MOT1 C MOT1 DEL MOT1DUP MOT1 C MOT1 C MOT1 DEL MOT1DUP MOT1 C Coastal Inland Coastal Inland Coastal Inland (c) 14 a Relative gene expression 12 (2 -ΔΔCt NaCl / 2 -ΔΔCt Control) a 10 a a 8 a a b ab b 6 b b c 4 c b c bc bc d cd d c c 2 d d c 0 SPL7 COPT6 CNX6 CNX1 ABA3 (d) 50 a Col-0 45 ab a T9 MOT1 C [ABA] (nmol l–1) b b PA10 b mot1-2 a a 40 ab ab T1 ab MOT1 DEL b T9 35 c c c copt6-1 d aba2-1 30 0 mM NaCl 50 mM NaCl + Fig. 5 Stress signalling and MOT1. (a) Mean  SE of soil sodium (Na ), copper (Cu) and molybdenum (Mo) content of nine soil samples per deme, and (b) mean  SE of leaf Na+, Cu and Mo content of Arabidopsis plants collected from their natural habitat, classified as coastal demes harbouring the MOT1C allele (black), coastal demes harbouring the MOT1DEL allele (green), demes harbouring the MOT1DUP allele (red) and inland demes harbouring the MOT1C allele (grey). (c) Relative transcript expression (treatment (50 mM NaCl) vs control (0 mM NaCl)) of SPL7, COPT6, CNX6, CNX1 and ABA3 in shoots of triplicate biological replicates of plants cultivated in hydroponics under control conditions or exposed to 50 mM of NaCl for 2 wk. Data represent the mean  SE of three plants per accession. (d) Abscisic acid (ABA) content (nmol l1) of plants cultivated in common soil irrigated with 0 and 50 mM of NaCl for 2 wk. Boxplots represent the median and interquartile range of three plants per accession. Letters indicate significant differences (Tukey’s HSD, P < 0.05). Ó 2021 The Authors New Phytologist (2021) 232: 208–220 New Phytologist Ó 2021 New Phytologist Foundation www.newphytologist.com New 218 Research Phytologist Fig. 6 Model for the salinity tolerance strategy of MOT1DEL plants adapted to coastal conditions. ABA, abscisic acid. especially as they apply to adaptive contexts in nature. Here we generating the leaf ionome data for the 1 135 Arabidopsis acces- focused on allelic variation detected on the Arabidopsis Molybde- sions. We are grateful to Jardı Botanic Marimurtra de Blanes and num Transporter MOT1, in particular on the most prevalent vari- ECAF Santa Coloma de Farners for providing field sites. CP was ant, MOT1DEL, which is characterized by low MOT1 expression funded by Spanish Government grants BFU2016-75176-R and and reduced leaf Mo content. We explored the adaptive value of PID2019-104000-RB-100. DES was funded by US National this low expression in the context of the linked phenotype of Institutes of Health (NIH) grants 2R01GM078536 and altered Na + and Cu homeostasis. Within our study area in 2P4ES007373-19A1, European Commission grant PCIG9-GA- north-east Spain, Arabidopsis plants harbouring the MOT1DEL 2011-291798, and BBSRC grant BB/L000113/1. This work was allele could only be found near the salt-rich coast. Reciprocal also supported by the European Research Council (ERC) under transplant experiments demonstrated that plants with the the European Union’s Horizon 2020 research and innovation MOT1DEL allele are better adapted to these coastal conditions programme (grant no. ERC-StG 679056 HOTSPOT), via a than those with the normal-functioning MOT1C allele. More- grant to LY; and the Biotechnology and Biological Sciences over, salt stress challenge experiments demonstrated that the Research Council (grant no. BB/P013511/1), via a grant to the MOT1DEL allele is associated with enhanced salinity tolerance. John Innes Centre (LY). Apart from Na+, leaf Cu content was the only element that was consistently elevated under salinity challenge. This underscores Author contributions the relationship between COPT6 and MOT1 and the importance of Cu in Moco and ABA biosynthesis. We accordingly suggest SB, CP and DES conceived the study. SB and JT performed field that MOT1DEL has an adaptive advantage in coastal conditions, collections and laboratory experiments. SB and LY performed due to pleiotropic enhancement of both Cu and ABA levels. analyses. SB wrote the manuscript with primary input from all authors. All authors edited and approved the final manuscript. Acknowledgements ORCID We thank Xin-Yuan Huang for helping with MOT1 genotyping, and Ana Carolina A. L. Campos and William van Dijk for Silvia Busoms https://orcid.org/0000-0002-9917-7118 New Phytologist (2021) 232: 208–220 Ó 2021 The Authors www.newphytologist.com New Phytologist Ó 2021 New Phytologist Foundation New Phytologist Research 219 Charlotte Poschenrieder https://orcid.org/0000-0002-3818- Kaufholdt D, Gehl C, Geisler M, Jeske O, Voedisch S, Ratke C, Bollh€oner B, 0874 Mendel RR, H€a nsch R. 2013. Visualization and quantification of protein interactions in the biosynthetic pathway of molybdenum cofactor in David E. Salt https://orcid.org/0000-0003-0283-0991 Arabidopsis thaliana. Journal of Experimental Botany 64: 2005–2016. Joana Teres https://orcid.org/0000-0003-4136-9931 Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, Buxton Levi Yant https://orcid.org/0000-0003-3442-0217 S, Cooper A, Markowitz S, Duran C et al. 2012. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28: 1647–1649. Marschner H. 1995. Mineral nutrition of higher plants, 2nd edn. London, UK: References Academic Press. Alonso JM, Stepanova AN, Leisse TJ, Kim CJ, Chen H, Shinn P, Stevenson Mendel RR. 2007. Biology of the molybdenum cofactor. Journal of Experimental DK, Zimmerman J, Barajas P, Cheuk R et al. 2003. Genome-wide insertional Botany 58: 2289–2296. mutagenesis of Arabidopsis thaliana. Science 301: 653–657. Mengel K, Kirkby EA, Kosegarten H, Appel T. 2001. Molybdenum. In: Mengel Baxter I, Muthukumar B, Park HC, Buchner P, Lahner B, Danku J, Zhao K, K, Kirkby EA, Kosegarten H, Appel T, eds. Principles of plant nutrition. Lee J, Hawkesford MJ, Guerinot ML et al. 2008. Variation in molybdenum Dordrecht, the Netherlands: Springer. content across broadly distributed populations of Arabidopsis thaliana is Munns R, Tester M. 2008. Mechanisms of salinity tolerance. Annual Review of controlled by a mitochondrial molybdenum transporter (MOT1). PLoS Genetics Plant Biology 59: 651–681. 4: e1000004. Panagos P, Van Liedekerke M, Jones A, Montanarella L. 2012. European Soil Billard V, Ourry A, Maillard A, Garnica M, Coquet L, Jouenne T, Cruz F, Data Centre: Response to European policy support and public data Garcia-Mina JM, Yvin JC, Etienne P. 2014. Copper-deficiency in Brassica requirements. Land Use Policy 29: 329–338. napus induces copper remobilization, molybdenum accumulation and Park HJ, Kim WY, Yun DJ. 2016. A new insight of salt stress signaling in plant. modification of the expression of chloroplastic proteins. PLoS ONE 9: Molecules and Cells 39: 447. e109889. Pe~ narrubia L, Romero P, Carrio-Seguı A, Andres-Borderıa A, Moreno J, Sanz Busoms S, Paajanen P, Marburger S, Bray S, Huang XY, Poschenrieder C, Yant A. 2015. Temporal aspects of copper homeostasis and its crosstalk with L, Salt DE. 2018. Fluctuating selection on migrant adaptive sodium hormones. Frontiers in Plant Science 6: 255. transporter alleles in coastal Arabidopsis thaliana. Proceedings of the National Poormohammad Kiani S, Trontin C, Andreatta M, Simon M, Robert T, Salt Academy of Sciences, USA 115: E12443–E12452. DE, Loudet O. 2012. Allelic heterogeneity and trade-off shape natural Busoms S, Teres J, Huang XY, Bomblies K, Danku J, Douglas A, Weigel D, variation for response to soil micronutrient. PLoS Genetics 8: e1002814. Poschenrieder C, Salt DE. 2015. Salinity is an agent of divergent selection Puig S. 2014. Function and regulation of the plant COPT family of high-affinity driving local adaptation of Arabidopsis to coastal habitats. Plant Physiology 168: copper transport proteins. Advances in Botany 2014: 1–9. 915–929. SAS Software. 2016. JMP [WWW document] URL https://www.jmp.com/es_es/ Campos ACAL, van Dijk WFA, Ramakrishna P, Giles T, Korte P, Douglas A, home.html [accessed 8 July 2021]. Smith P, Salt DE. 2021. 1,135 ionomes reveals the global pattern of leaf and Schwarz G, Mendel RR. 2006. Molybdenum cofactor biosynthesis and seed mineral nutrient and trace element diversity in Arabidopsis thaliana. The molybdenum enzymes. Annual Review of Plant Biology 57: 623–647. Plant Journal 106: 536–554. Schwarz G, Schulze J, Bittner F, Eilers T, Kuper J, Bollmann G, Nerlich A,  Romero P, Sanz A, Pe~ Carrio-Seguı A, narrubia L. 2016. Interaction between Brinkmann H, Mendel RR. 2000. The molybdenum cofactor biosynthetic ABA signaling and copper homeostasis in Arabidopsis thaliana. Plant and Cell protein Cnx1 complements molybdate-repairable mutants, transfers Physiology 57: 1568–1582. molybdenum to the metal binding pterin, and is associated with the Du J, Hesp PA. 2020. Salt spray distribution and its impact on vegetation cytoskeleton. Plant Cell 12: 2455–2471. zonation on coastal dunes: a review. Estuaries and Coasts 43: 1885–1907. Tejada-Jimenez M, Galva n A, Ferna ndez E, Llamas A. 2009. Homeostasis of the Eskandari S, Mozaffari V. 2014. Interactive effect of soil salinity and copper micronutrients Ni, Mo and Cl with specific biochemical functions. Current application on growth and chemical composition of pistachio seedlings. Opinion in Plant Biology 12: 358–363. Communications in Soil Science and Plant Analysis 45: 688–702.  Teres J, Busoms S, Perez Martın L, Luıs-Villarroya A, Flis P, Alvarez-Fern a ndez Forsberg SK, Andreatta ME, Huang XY, Danku J, Salt DE, Carlborg O. € 2015. A, Tolra R, Salt DE, Poschenrieder C. 2019. Soil carbonate drives local The multi-allelic genetic architecture of a variance-heterogeneity locus for adaptation in Arabidopsis thaliana. Plant, Cell & Environment 42: 2384–2398. molybdenum concentration in leaves acts as a source of unexplained additive Tomatsu H, Takano J, Takahashi H, Watanabe-Takahashi A, Shibagaki N, genetic variance. PLoS Genetics 11: e1005648. Fujiwara T. 2007. An Arabidopsis thaliana high-affinity molybdate transporter Garcia-Molina A, Andres-Cola s N, Perea-Garcıa A, Neumann U, Dodani SC, required for efficient uptake of molybdate from soil. Proceedings of the National Huijser P, Penarrubia L, Puig S. 2013. The Arabidopsis COPT6 transport Academy of Sciences, USA 104: 18807–18812. protein functions in copper distribution under copper-deficient conditions. Vigani G, Di Silvestre D, Agresta AM, Donnini S, Mauri P, Gehl C, Bittner F, Plant and Cell Physiology 54: 1378–1390. Murgia I. 2017. Molybdenum and iron mutually impact their homeostasis in Huang XY, Liu H, Zhu YF, Pinson SR, Lin HX, Guerinot ML, Zhao FJ, Salt cucumber (Cucumis sativus) plants. New Phytologist 213: 1222–1241. DE. 2019. Natural variation in a molybdate transporter controls grain Watanabe S, Sato M, Sawada Y, Tanaka M, Matsui A, Kanno Y, Hirai MY, Seki molybdenum concentration in rice. New Phytologist 221: 1983–1997. M, Sakamoto A, Seo M. 2018. Arabidopsis molybdenum cofactor sulfurase Huang XY, Salt DE. 2016. Plant ionomics: from elemental profiling to ABA3 contributes to anthocyanin accumulation and oxidative stress tolerance environmental adaptation. Molecular Plant 9: 787–797. in ABA-dependent and independent ways. Scientific Reports 8: 1–4. Ide Y, Kusano M, Oikawa A, Fukushima A, Tomatsu H, Saito K, Hirai MY, Wu S, Hu C, Tan Q, Xu S, Sun X. 2017. Nitric oxide mediates molybdenum- Fujiwara T. 2011. Effects of molybdenum deficiency and defects in molybdate induced antioxidant defense in wheat under drought stress. Frontiers in Plant transporter MOT1 on transcript accumulation and nitrogen/sulphur metabolism Science 8: 1085. in Arabidopsis thaliana. Journal of Experimental Botany 62: 1483–1497. Yamasaki H, Hayashi M, Fukazawa M, Kobayashi Y, Shikanai T. 2009. Jung HI, Gayomba SR, Rutzke MA, Craft E, Kochian LV, Vatamaniuk OK. SQUAMOSA promoter binding protein–like7 is a central regulator for copper 2012. COPT6 is a plasma membrane transporter that functions in copper homeostasis in Arabidopsis. Plant Cell 21: 347–361. homeostasis in Arabidopsis and is a novel target of SQUAMOSA promoter- Zhang J, Jia W, Yang J, Ismail AM. 2006. Role of ABA in integrating plant binding protein-like 7. Journal of Biological Chemistry 287: 33252–33267. responses to drought and salt stresses. Field Crops Research 97: 111–119. Ó 2021 The Authors New Phytologist (2021) 232: 208–220 New Phytologist Ó 2021 New Phytologist Foundation www.newphytologist.com New 220 Research Phytologist Fig. S3 Map and geostatistical analysis of the 1212 world-wide Supporting Information Arabidopsis accessions classified by MOT1 allele type. Additional Supporting Information may be found online in the Supporting Information section at the end of the article. Fig. S4 Growth of plants with different MOT1 alleles exposed to soil salinity and salt spray. Dataset S1 MOT1 allele type classification and leaf molybdenum (Mo) content of the plants cultivated in the reciprocal trans- Methods S1 DNA and RNA extractions for MOT1 genotyping plants. and target gene expression. Dataset S2 Soil and leaf ionome analysis from all the experi- Methods S2 Irrigation and hydroponic experiment procedures. ments conducted in this study. Methods S3 Salt spray and survival experiment design. Dataset S3 MOT1 allele type classification and coordinates of the world-wide set of 1135 Arabidopsis natural accessions. Methods S4 Salinity test design of 306 Arabidopsis world-wide accessions. Dataset S4 MOT1 allele type classification and leaf sodium (Na+) content of a world-wide set of 306 Arabidopsis natural Table S1 Primers list for transcript quantification of target genes. accessions. Please note: Wiley Blackwell are not responsible for the content Dataset S5 Statistical analysis of all the comparisons performed or functionality of any Supporting Information supplied by the in this study. authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Fig. S1 Leaf Mo content and root and shoot MOT1 expression. Fig. S2 Ionomic profiles of MOT1 alleles exposed to salinity New Phytologist is an electronic (online-only) journal owned by the New Phytologist Foundation, a not-for-profit organization dedicated to the promotion of plant science, facilitating projects from symposia to free access for our Tansley reviews and Tansley insights. Regular papers, Letters, Viewpoints, Research reviews, Rapid reports and both Modelling/Theory and Methods papers are encouraged. We are committed to rapid processing, from online submission through to publication ‘as ready’ via Early View – our average time to decision is <26 days. There are no page or colour charges and a PDF version will be provided for each article. The journal is available online at Wiley Online Library. Visit www.newphytologist.com to search the articles and register for table of contents email alerts. If you have any questions, do get in touch with Central Office (np-centraloffice@lancaster.ac.uk) or, if it is more convenient, our USA Office (np-usaoffice@lancaster.ac.uk) For submission instructions, subscription and all the latest information visit www.newphytologist.com New Phytologist (2021) 232: 208–220 Ó 2021 The Authors www.newphytologist.com New Phytologist Ó 2021 New Phytologist Foundation