Conserved histidine of metal transporter AtNRAMP1 is crucial for optimal plant growth under manganese deficiency at chilling temperatures

Authors

  • Anna Ihnatowicz,

    Corresponding author
    1. Laboratory of Plant Protection and Biotechnology, Intercollegiate Faculty of Biotechnology of University of Gdansk and Medical University of Gdansk, Gdansk, Poland
    2. Department of Plant Breeding and Genetics, Max Planck Institute for Plant Breeding Research, Cologne, Germany
    Search for more papers by this author
  • Joanna Siwinska,

    1. Laboratory of Plant Protection and Biotechnology, Intercollegiate Faculty of Biotechnology of University of Gdansk and Medical University of Gdansk, Gdansk, Poland
    Search for more papers by this author
  • Andrew A. Meharg,

    1. Institute for Global Food Security, Queen's University Belfast, David Keir Building, Belfast, UK
    Search for more papers by this author
  • Manus Carey,

    1. Institute for Global Food Security, Queen's University Belfast, David Keir Building, Belfast, UK
    Search for more papers by this author
  • Maarten Koornneef,

    1. Department of Plant Breeding and Genetics, Max Planck Institute for Plant Breeding Research, Cologne, Germany
    2. Laboratory of Genetics, Wageningen University, PE Wageningen, the Netherlands
    Search for more papers by this author
  • Matthieu Reymond

    Corresponding author
    1. Department of Plant Breeding and Genetics, Max Planck Institute for Plant Breeding Research, Cologne, Germany
    2. Department of Plant Cell Wall, Function and Utilization, Institut Jean-Pierre Bourgin, INRA Centre de Versailles-Grignon, Versailles Cedex, France
    Search for more papers by this author

Summary

  • Manganese (Mn) is an essential nutrient required for plant growth, in particular in the process of photosynthesis. Plant performance is influenced by various environmental stresses including contrasting temperatures, light or nutrient deficiencies. The molecular responses of plants exposed to such stress factors in combination are largely unknown.
  • Screening of 108 Arabidopsis thaliana (Arabidopsis) accessions for reduced photosynthetic performance at chilling temperatures was performed and one accession (Hog) was isolated. Using genetic and molecular approaches, the molecular basis of this particular response to temperature (G × E interaction) was identified.
  • Hog showed an induction of a severe leaf chlorosis and impaired growth after transfer to lower temperatures. We demonstrated that this response was dependent on the nutrient content of the soil. Genetic mapping and complementation identified NRAMP1 as the causal gene. Chlorotic phenotype was associated with a histidine to tyrosine (H239Y) substitution in the allele of Hog NRAMP1. This led to lethality when Hog seedlings were directly grown at 4°C.
  • Chemical complementation and hydroponic culture experiments showed that Mn deficiency was the major cause of this G × E interaction. For the first time, the NRAMP-specific highly conserved histidine was shown to be crucial for plant performance.

Introduction

Plant photosystems are sensitive to both temperature and to metal ion homeostasis. Chilling temperatures reduce photosystem II (PSII) function, possibly by oxidative damage to thylakoids (Tambussi et al., 2004), by inducing the inhibition of downstream electron transport (Zhou et al., 2004) and conformational changes of PSII (Parvanova et al., 2004). Manganese (Mn) clusters form the catalytic centre of PSII (Marschner, 2012). In chloroplasts, Mn deficiency induces PSII photoinhibition and the loss of the PSII-D1 subunit binding Mn cluster, and produces reactive oxygen species (ROS), while iron (Fe) starvation mostly affects photosystem I (PSI) and causes a decrease in chlorophyll synthesis (Nouet et al., 2011).

An antagonistic relationship between Fe and Mn in plants is well established (van der Vorm & van Diest, 1979) and it may occur during their uptake by the root or during their translocation from roots to shoots. Soil availability of Fe and Mn is dependent on pH, soil redox potential, aeration, water retention and the availability of a range of other ions (Moosavi & Ronaghi, 2011; Marschner, 2012). Mn deficiency can be a major cause of growth and yield reduction in crops (Donald & Prescott, 1975; Hebbern et al., 2005, 2009; Jiang, 2006; Schmidt et al., 2013). It is problematic on alkaline soils where characteristic interveinal leaf chlorosis is often observed due to changes in thylakoid structure and the promotion of chlorophyll degradation (Simpson & Robinson, 1984; Jiang, 2006; Papadakis et al., 2007). When Mn availability and accumulation is in excess, decreased plant growth, presence of brown spots on mature leaves, severe chlorosis and even necrosis, can all occur (Marschner, 2012). Arabidopsis thaliana (Arabidopsis) plants grown under excess Mn exhibited increased susceptibility to PSII photoinhibition and significantly lower abundance of PSI reaction centre polypeptides compared with the control (Millaleo et al., 2013).

A significant inhibition of Mn accumulation by plants was correlated with the excess of Fe and cadmium (Cd) accumulation (Roomizadcha & Karimian, 1996; Hernández et al., 1999), suggesting that Mn transporters have broad selectivity. Homologues of the conserved Natural Resistance Associated Macrophage Protein (NRAMP) family of divalent-metal transporters plays a critical role in Mn2+ acquisition and homeostasis along with Fe2+, Co2+, Cd2+, Cu2+, Ni2+ and Zn2+ (Cellier et al., 1995; Nelson, 1999; Agranoff et al., 2005; Nevo & Nelson, 2006; Oomen et al., 2009). There are six NRAMP transporters in Arabidopsis, belonging to two subfamilies: AtNRAMP1 and AtNRAMP6 form the first group, and AtNRAMP2–5 comprise the second group (Maser et al., 2001). AtNRAMP3 and AtNRAMP4 are essential for seed germination under low Fe conditions (Thomine et al., 2003; Lanquar et al., 2005), while AtNRAMP6 contributes to Cd toxicity (Cailliatte et al., 2009). Importantly, the vacuolar export of Mn by AtNRAMP3 and AtNRAMP4 were shown to be required for optimal photosynthesis and growth under Mn deficiency (Lanquar et al., 2010). Arabidopsis NRAMP1 is a high-affinity Mn transporter (Cailliatte et al., 2010).

Screening of Arabidopsis revealed one accession (Hog) that exhibited differential chlorosis at lower temperatures that was dependent on the soil used in screening. This interaction between chlorosis and chilling was investigated here to elucidate the underlying molecular basis of this phenomenon, establishing that it was related to Mn and Fe homeostasis, and the environmental regulation of this nutritional × temperature interaction.

Materials and Methods

Plant material

The Arabidopsis thaliana (L.) Heynh. accessions studied were obtained from ABRC (http://arabidopsis.org/) and NASC (http://arabidopsis.info/) stock centres and are listed in Supporting Information Table S1. The F2 and the advanced-backcrossed mapping populations were derived from crosses between Landsberg erecta (Ler originates from Poland – used as a recurrent parent line) and Hodja-Obi-Garm accessions (Hog collected at c. 1800 m a.s.l. in Tajikistan). The Hog (N6179) line used in these crosses (named Hog hereafter) is a descendant line and was derived originally from several selfings of the Hog (N922) line by the NASC stock centre (named Hog-N922 hereafter). T-DNA insertion mutants were identified in the Salk collection (http://signal.salk.edu/; Alonso et al., 2003), which consist of flank-tagged ROK2 T-DNA lines (Col-0 background), by searching the insertion flanking database SIGnAL (http://signal.salk.edu/cgi-bin/tdnaexpress).

Growth conditions and treatments

For stratification, Arabidopsis seeds were sown in Petri dishes on water-saturated Whatman paper followed by a cold treatment for 4 d at 4°C, and then planted into different soil mixes. In all described experiments, time is referred to as days after sowing (DAS). The Arabidopsis accessions, monitored with an automated measurement system detecting alterations in the effective quantum yield of PSII (ΦII), were grown in growth chambers under a photoperiod of 16 h light (intensity 120 μmol m−2 s−1) at 28°C and 8 h dark at 20°C or under a photoperiod of 16 h light (350 μmol m−2 s−1) at 16°C and 8 h dark at 14°C. A set of nine accessions was used in the temperature shift experiment, in which plants were grown for 16 d under a photoperiod of 16 h light (120 μmol m−2 s−1) at 22°C and 8 h dark at 20°C and subsequently transferred to various environmental scenarios as described in Table S2. A set of various soil mixes derived from commercial products (commercial soil 1-CS1, commercial soil 2-CS2, commercial de-acidified moss-CDM), different proportions of vermiculite (3–6 mm in diameter) and/or soils originating from four natural habitats in the north of Poland (Table S3) were used. Soil pH was measured with CP-105 waterproof pocket pH meter. Chemical analysis of soils (Table S4) was conducted by the Regional Agro-Chemical Station (OSCh-R, http://www.oschrgdansk.pl/) in Gdansk, Poland. The hydroponic system was derived from the modified system of Heeg et al. (2008). Arabidopsis seeds were surface sterilized in a hypochloride solution (5% NaOCl) and after rinsing three times in autoclaved millipore water were sown on tip-cut, 0.65% agar-filled tubes that hung through holes in a plate into a 96-tips box with nutrient solution. Nutrient solutions contained: 2 mM KNO3, 2 mM Ca(NO3)2, 0.5 mM MgSO4, 0.25 mM KH2PO4, 0.5 mM NH4NO3, 40 μM FeSO4-Na-EDTA, 25 μM H3BO3, 2 μM ZnSO4, 0.5 μM CuSO4, 50 μM KCl, 0.075 μM (NH4)6Mo7O24, 0.15 μM CoCl2 and 2 μM MnCl2 (for Mn control condition) or no Mn (for Mn depleted condition). The pH was adjusted to 5.9 with KOH, and MES was used as a buffering agent. Plants were grown hydroponically in a controlled environment (35 μmol m−2 s−1 for 16 h at 20°C and 8 h dark at 18°C) for 4 wk and subsequently were transferred to a cold environment (20 μmol m−2 s−1 for 16 h at 6°C and 8 h dark at 4°C).

Genetic mapping of locus involved in the observed G × E interaction

The F2 mapping population (= 88 individuals), derived from a cross between the Hog and Ler accessions, was genotyped with 36 simple sequence length polymorphisms (SSLP) markers. Subsequently, an advanced-backcrossed population strategy was followed by crossing F2 plants showing chlorosis at 4°C with Ler as a nonchlorotic recurrent parent. Among 648 BC1F2 plants genotyped with 62 SSLP markers, two lines with a Ler genetic background for most of the genome were selected and crossed again with Ler. A set of 544 BC2F2 plants derived from these last crosses was genotyped with 25 SSLP markers. The density of the genetic maps was further increased with 85 single nucleotide polymorphisms (SNPs) using a Illumina Golden Gate SNP Assay as described by Huang et al. (2012). In total 172 SSLP and SNP markers (Tables S5, S6) were used to genotype this BC2F2 segregating population. As a result, an isogenic line (IL) was selected containing a unique small introgression of Hog alleles (0.82 Mb) at the bottom of chromosome one in an otherwise genetic background of Ler.

For fine mapping, a BC3F2 population was generated by backcrossing the selected chlorotic IL (BC2F2) with Ler. For this purpose, a set of new polymorphic markers in the mapped region (Table S6, highlighted in bold) was developed. Newly developed SSLP and cleaved amplified polymorphic sequences (CAPS) markers were designed based on DNA sequencing of a large number of 1-kb Hog amplicons from the region. The primers used to amplify Hog genomic DNA were designed based on the Col-0 database (http://www.arabidopsis.org) using Gene Runner version 3.05 (Hastings, NY, USA). In total, 1500 BC3F2 plants were genotyped and a subset of lines with different recombinant events was selected. These recombinant lines were phenotyped, which together with genotyping data enabled us to delimit the interval of the causal locus to 96 kb, covering 32 genes (Table S7). Among them, 20 candidates were selected (Table S8), for which T-DNA insertion mutants were ordered (http://signal.salk.edu/) and grown at low temperatures. Genomic DNA used for mapping was isolated using the MagAttract 96 DNA plant core kit (Qiagen). DNA used for sequencing was isolated with the QIAGEN Plant DNA kit. Homozygous mutants nramp1-1 (SALK_053236) and pde18 (pigment defective 318, SALK_145367) were identified by PCR using gene-specific and T-DNA primers (Table S9). The annotated insertion sites in both mutant lines were confirmed by sequencing.

Genetic complementation by allelism test

For the genetic complementation, chlorotic mutant lines (nramp1-1 and pde18) were crossed with the selected isogenic line (IL) from the advanced-backcrossed population (BC2F3 generation) and with Hog accession. Additionally, control crosses were performed with the Col-0 accession. Subsequently, the possible reversion of the chlorotic phenotype was monitored in F1 hybrids grown under low temperature conditions.

Sequence analysis

DNA sequences were determined using the Abi Prism 377, 3100, 3730 sequencers and the 3730xl DNA analyzer and 3130xl Genetic Analyzer (Applied Biosystems) using BigDye terminator v3.1 chemistry. All sequences were aligned using CLUSTALW (Thompson et al., 1994). The graphical presentation of the genotype of the selected IL was created with 172 polymorphic SSLP, CAPS and SNP (Ilumina) markers (Tables S5, S6) using GGT Software (www.wageningenur.nl; van Berloo, 2007).

Quantitative real-time PCR analysis

Total RNA was extracted from both chlorotic (Hog, IL, nramp1-1) and nonchlorotic plants (Hog-N922, Ler and Col-0) grown at chilling temperatures. The RNeasy® Plant Mini Kit (Qiagen) was used following the instructions of the manufacturer and including on-column DNA digestion step with the RNase-Free DNase Set (Qiagen) to eliminate genomic DNA contamination. 1 μg of RNA was used for reverse transcription by SuperScript® III Reverse Transcriptase (Invitrogen) with oligo(dT)20 primers. qPCR was performed using LightCycler® 480 Real-Time PCR System (Roche) and LightCycler® FastStart DNA Master SYBR Green I (Roche), using the following gene-specific primers: for ACT2 (ACTIN2) 5′-CTTGCACCAAGCAGCATGAA-3′ and 3′-CCGATCCAGACACTGTACTTCCTT-5′ (Czechowski et al., 2005); for NRAMP1 5′-CAAACGGGAGCTCAAAGGAT-3′ and 3′-TGCTCTCATCGGTGGTTCAG-5′. Primer specificity for NRAMP1 was confirmed by the analysis of the melting curves. Relative transcript level (RLT) of the NRAMP1 was normalized to the transcript level of the house-keeping ACTIN2 gene (At3g18780) and calculated as follows: RLT = N0 × 2Cp (N0, initial copy number; Cp, crossing point).

Chlorophyll fluorescence measurement

The procedure used to identify accessions showing changes in ΦII was described before by Varotto et al. (2000). In-vivo Chla fluorescence was carried out with a PAM (pulse amplitude modulation) 101/103 fluorometer (Walz, Effeltrich, Germany). Plants used for measurements were either grown at optimal condition at 22°C under a photoperiod of 16 h light (120 μmol m−2 s−1) or first grown under optimal condition for 16 d and then placed at 4°C under a photoperiod of 16 h light (20 μmol m−2 s−1). Measurements were performed at room temperature on single leaves of the average plants from each genotype. For plants carrying a mutated NRAMP1 allele, a chlorotic area was used for measurement. Plants were dark-adapted for 30 min before each measurement. Pulses (0.8 s) of white light (6000 μmol m−2 s−1) were used to determine the maximum fluorescence (FM) and the ratio (FM – F0)/FM = FV/FM. A 5 min illumination with actinic light (80 μmol m−2 s−1) served to drive electron transport before ΦIIII = (FM′ – F0′)/FM′) and photochemical quenching (qP = (FM′ – FS)/FM′ – F0) were measured (Bonardi et al., 2005). Different actinic light intensities in the range from 80 to 1200 μmol m−2 s−1 were used to measure the ΦII and qP parameters and the nonphotochemical quenching (NPQ) in a light-intensity-dependent manner.

Chlorophyll measurement

Chlorophyll extraction with acetone was done by the method given by Porra et al. (1989). The same FW of rosette leaves were frozen in liquid nitrogen and disrupted in a mortar. Homogenized samples were taken into centrifuge tubes, to which 1 ml 80% acetone was added. The tubes were kept in the dark for 10 min at 4°C and subsequently a 10 min centrifugation was done at 34,100 g at 4°C. The absorbances of the diluted supernatants were taken at 750.0, 663.6 and 646.6 nm on a Beckman DU 640 Spectrophotometer (CA, USA). After measurement a special formula was used to convert absorbance measurements to mg of Chl (Porra et al., 1989).

Chemical rescue of chlorosis

Before sowing, all soil mixes were treated with general-purpose fertilizer. Subsequently, plants were watered with general-purpose fertilizer once per week and with Fe-EDDHMA (YaraVita Rexolin®M48, Oslo, Norway) or with Mn-EDTA (Florovit, Susz, Poland) once per week. The compositions of fertilizers used are described in Table S10.

Trace element analysis

Arabidopsis shoots were first milled before microwave (MARS) assisted digestion in concentrated nitric acid (Aristar). Metal concentrations were determined by Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) (Thermo iCAP Q, Bremen, Germany).

Statistical analysis

All treatments included at least three biological replicates. Data processing and statistical analyses (pairwise comparisons using t-tests) were carried out using Excel (Microsoft Excel 2010). Error bars representing standard deviation (SD) are shown in the figures; the data presented are means. The significance level used here is < 0.05. In the figures, the data points whose means were found to be significantly different are linked with asterisk(s).

Results

Chlorosis and growth on temperature reduction

Arabidopsis accessions were grown at two temperatures (16 and 28°C). An automated screening system detecting alterations in ΦII has been used to quantify photosynthetic performance. Under high temperature, no significant differences in photosynthetic yield between accessions were observed, whereas under lower temperature, only one accession Hodja-Obi-Garm (Hog) showed pale green leaves and a reduction in photosynthetic yield (Fig. 1). Based on this observation a temperature shift experiment was performed. Hog plants together with a set of eight other accessions were grown at optimal condition (120 μmol m−2 s−1 for 16 h at 22°C and 8 h dark at 20°C) for 16 d (Fig. 2a) and subsequently sets of plants were distributed to various environmental scenarios with different combinations of light intensities, temperatures and photoperiods (Table S2). Interestingly, only the Hog accession showed an induction of a severe leaf chlorosis and impaired growth after transfer to chilling temperatures. This phenotype was the most striking at 4°C under a photoperiod of 16 h light (20 μmol m−2 s−1) occurring 2–3 wk after transfer to low temperature (Fig. 2b). Interestingly, the recovery of cold-induced chlorosis was observed just 5 d after transfer back to optimal temperature (Fig. 2c).

Figure 1.

Frequency distribution of chlorophyll fluorescence measured as the effective quantum yield of photosystem II (ΦII = (FM' – F0′)/FM') for 108 Arabidopsis thaliana accessions. The average values of ΦII obtained for Hog (arrow with diamond) and Ler (arrow with dot) are indicated. Averages are based on three independent measurements. Plants used for screening were grown under a photoperiod of 16 h light (intensity 350 μmol m−2 s−1) at 16°C and 8 h dark at 14°C. The chlorophyll fluorescence measurement was performed at room temperature by using the automated measurement detecting alterations in ΦII.

Figure 2.

Phenotypic appearance of two Arabidopsis thaliana accessions (Hog and Ler), which showed contrasting responses to low temperatures and were selected as parents for the advanced-backcrossed population. (a) 16-d-old plants (16 days after sowing, DAS) grown at optimal temperature (22°C) before transfer to cold growth room (4°C). (b) Induction of chlorotic phenotype of Hog plants grown at 4°C for next 45 d. (c) 61 DAS plants were transferred back from 4°C to 22°C. The first signs of chlorosis recovery were seen 5 d after transfer to 22°C (66 DAS).

Genetic mapping of the chlorotic phenotype

In order to identify the number and the chromosomal position of the loci involved in the observed G × E interaction, an F2 population was derived from a cross between Hog (chlorotic at low temperatures only) and Ler (green whatever the temperature). Two-week-old F2 seedlings were transferred to 4°C under a photoperiod of 16 h light (20 μmol m−2 s−1). The first signs of chlorosis were observed 7–14 d after transfer. A 3 : 1 ratio (the frequency of normal: chlorotic plants in F2) fitted the data (χ² = 0.0607, > 0.85), indicating that a single locus was responsible for the effect of the observed G × E and that the allele from Hog leading to chlorosis is recessive. In order to map the chromosomal position of this locus, the F2 plants were genotyped with 36 SSLP markers uniformly distributed throughout the genome. Hence, the locus was mapped to the bottom of the chromosome 1 linked to the marker Msat1.2 (28.5 Mb, < 0.001). Subsequently, successive backcrosses from the chlorotic F2 plants and using Ler as a recurrent line were performed in order to introgress Hog alleles at the bottom of chromosome 1 into the genetic background of Ler. At each generation obtained, chlorotic plants were selected and genotyped in order to select those with a higher number of Ler alleles among the tested markers spread over the genome. Among the BC1F2 population, two lines with the ‘cleanest’ Ler genetic background were selected and backcrossed again with Ler. In the next generation, the selection procedure resulted in the identification of a single line (IL for isogenic line) containing only a small introgression of Hog alleles (0.82 Mb) at the bottom of the chromosome 1 (Fig. 3a), which still remained chlorotic when grown at 4°C. In order to reduce the number of genes present in the introgressed region and to select candidate genes involved in the highlighted G × E interaction, the selected IL was once more backcrossed with the recurrent parent Ler providing a BC3F2 generation of which 1500 individuals were genotyped using polymorphic markers flanking the introgressed region. Only the progeny of plants with a recombinant event in this region were phenotyped at 4°C. SSLP and CAPS polymorphic markers within the introgressed region were developed (Table S6) and used to precisely map the interval carrying the causal gene, by analysing co-segregation of the allelic values of markers and the occurrence of the chlorotic phenotype. The region involved in the chlorotic phenotype was thereby fine-mapped to a 96-kb interval, between the CAPS030 and CAPS900 markers (physical position: 30 302 133 and 30 400 280 bp, respectively) (Fig. 3b).

Figure 3.

Genetic and fine-mapping of the region encoding for the chlorotic phenotype and identification of the Arabidopsis thaliana NRAMP1 locus. (a) Graphical representation of the genotype of the selected introgression line (IL) constructed with 172 polymorphic simple sequence length polymorphisms (SSLP), cleaved amplified polymorphic sequences (CAPS) and single nucleotide polymorphism (SNP) markers (Illumina) (Supporting Information Tables S5, S6). The physical positions of polymorphic markers (in Mb) are indicated above green bars and by the corresponding names below. IL carries an introgression of Hog alleles (highlighted in red) into the genetic background of Ler (highlighted in green). (b) Schematic view of IL and a population of lines with different recombinant events selected in BC3F2 population. Ler alleles, highlighted in green; Hog alleles, red; heterozygous alleles, blue. +, recombinant lines that are chlorotic at low temperatures; −, recombinant lines remaining green at low temperatures. (c) 32 genes located within the fine-mapped region (96 kb) and the physical position of the NRAMP1 locus (At1 g80830) (www.arabidopsis.org).

Identification of NRAMP1 as the causal gene

Among the 32 genes present in the fine-mapped region (Fig. 3c), 20 candidate genes were identified according to their described function in TAIR (www.arabidopsis.org). For all candidate genes, T-DNA insertion lines were obtained. Among the homozygous mutants tested, two lines (nramp1-1 and pde18) showed a chlorotic pattern similar to the one observed for the selected IL and the Hog accession when grown at 4°C. Subsequently, an allelism test was performed to determine whether NRAMP1 and/or PDE18 are alleles of the same gene causing the observed chlorosis at low temperatures when mutated. Because the allele from Hog involved in the G × E is recessive, F1 plants derived from a cross between the selected IL or Hog and the mutant altered in the gene responsible for the effect of the G × E was expected to show a chlorotic pattern when allelic and when grown at 4°C. The F1 offspring of crosses between nramp1-1 mutant line with IL and Hog were chlorotic (Fig. S1). Under the same environmental conditions, control crosses of the nramp1-1 homozygous mutant line with the nonchlorotic wild-type accession Col-0 were green. The F1 hybrids of all crosses with the second candidate mutation (pde18) were all green, suggesting that PDE18 is not the gene involved in the chlorotic response to low temperature observed in the Hog accession. Therefore, the allelism tests clearly revealed that NRAMP1 is the causal gene involved in the chlorotic phenotype at low temperatures present in the Hog accession.

Photosynthesis at low temperatures

In order to perform more detailed characterization of observed chlorosis, phenotypic variation for photosynthetic performance traits of plants grown under two contrasted environments was characterized by measuring parameters of chlorophyll fluorescence. Spectroscopic analyses using pulse amplitude modulation (PAM) were performed on leaves of IL and parental lines (Hog, Ler), as well as on the nramp1-1 homozygous mutant line and mutant's genetic background (Col-0). When grown at 22°C, no clear differences in the photosynthetic performance were detected between tested lines (Fig. 4a,c). By contrast, significant changes in the measured parameters were detected, when plants grown at 4°C were measured (Fig. 4b,d). Under this condition, all chlorotic plants (IL, Hog and nramp1-1) showed significant reductions in the effective quantum yield of PSII (ΦII), the photochemical quenching (qP) and the maximum quantum yield of PSII (FV/FM) when compared to the corresponding controls (Ler and Col-0) (< 0.05). These data indicate a severe alteration in the photosynthetic electron flow in plants showing chlorosis at 4°C. When photosynthetic parameters of IL and both parental accessions (Hog, Ler) were performed in a light-intensity-dependent manner, additional differences between green and chlorotic plants grown at 4°C were detected (Fig. S2a–c), in particular the reduction of nonphotochemical quenching (NPQ) of Hog plants (Fig. S2c).

Figure 4.

Chlorophyll fluorescence parameters of Arabidopsis thaliana plants grown at two contrasted temperatures (22 and 4°C). The chlorophyll fluorescence measurement was performed at room temperature on (a,b) parental (Hog and Ler) and introgression (IL) lines and on (c,d) nramp1-1 T-DNA mutant line in comparison to its genetic background (Col-0). ΦII, effective quantum yield of photosystem II (PSII); qP, photosynthetic quenching; FV/FM, maximum quantum yield of PSII. Error bars, ± SD, from three measurements. *, < 0.05; **, < 0.01.

Molecular basis of G × E interaction

In order to reveal the causal polymorphism(s) at the nucleotide level involved in the different responses to low temperatures, the NRAMP1 gene of the IL, Hog and Ler plants were sequenced. A single nucleotide substitution was detected in the NRAMP1 coding region of the Hog and IL plants, leading to an amino acid change (histidine to tyrosine, respectively, referred to as a H239Y substitution) when compared with the reference accessions: Col-0 and nonchlorotic parent Ler. At this site, a highly conserved histidine (His239) in the transmembrane domain 6 (TMD6) is present, which is a part of the conserved peptide motif Met-Pro-His present in this gene in all eukaryotes and bacteria (Fig. S3). The 1001 Genomes Project database (www.1001genomes.org) does not mention any polymorphism at His239. Original seed stock from NASC (Hog-N922) was ordered and phenotypically compared with a descendant line (Hog), which was used as a parental line in our mapping populations. By contrast to this Hog line, original seed stock (Hog-N922) did not show any sign of chlorosis at low temperatures and importantly did not contain the histidine to tyrosine substitution at position 239. So, we hypothesize that the histidine to tyrosine substitution might be a spontaneous mutation that happened either during multiplication in the stock centres or in our laboratory.

Chlorosis, growth reduction and Mn deficiency at chilling temperatures

During fine-mapping experiments plants with a mutated allele of the NRAMP1 gene (IL, nramp1-1, Hog) became chlorotic at chilling temperatures only when grown on particular soils. To investigate this, chlorotic (Hog, IL, nramp1-1) and nonchlorotic (Ler, Col-0, Hog-N922) lines were grown at two contrasting temperatures (22°C and below 8°C) on various soils (Table S3). The soils varied in Fe and Mn content, as well as in Fe: Mn ratio (Fig. S4). At chilling temperatures, plants with the H239Y mutation and nramp1-1 line showed a severe chlorosis on some soils (Fig. 5a (control), S4). The observed chlorosis and impaired plant growth was not caused directly by low Mn concentration in soil, but rather by an excess of external Fe leading to Mn deficiency. Importantly, this chlorotic phenotype could be reversed by watering plants with chelated Mn-fertilizer (Fig. 5a, +Mn). The link between the chlorotic phenotype of plants carrying mutated NRAMP1 alleles and Mn deficiency was further confirmed by growing all six genotypes in a hydroponic system at 20 and 4°C. Plants were cultivated in controlled conditions either under optimal Mn concentration (2 μM) or without Mn. Under optimal Mn concentration no phenotypic differences were observed, even at 4°C (Fig. S5a). However, under the Mn-deficient condition all plants carrying the mutated NRAMP1 allele showed a growth reduction and paler pigmentation already at optimal temperature. When plants grown in hydroponic media without Mn were transferred to low temperature conditions, chlorosis of Hog, IL and nramp1-1 plants seemed to be further induced (Fig. S5b).

Figure 5.

Characterization of six genotypes (Ler, IL (introgression line), Col-0, nramp1-1, Hog-N922 and Hog) grown at low temperatures in manganese(Mn)-deficient soil (no. 4 in Supporting Information Tables S3, S4; Fig. S4). (a) Phenotypic appearance of Arabidopsis thaliana plants watered with general purpose fertilizer (control), after chemical rescue of chlorosis by watering plants once per week with chelated manganese fertilizer (+Mn), and watered with chelated iron (Fe) fertilizer (+Fe). Bars, 1 cm. (b) Trace element content (Mn and Fe) of fully expanded plant rosettes of green and chlorotic plants. Mn and Fe concentrations shown in ppm (mg kg−1) were determined by Inductively Coupled Plasma-Mass Spectrometry (ICP-MS). (c) Chlorophyll b content (Chla + Chlb) and Chla: Chlb ratio of fully expanded plant rosettes of green and chlorotic plants. Chlorophyll parameters were determined spectrophotometrically. Error bars represent the SD from three (chlorophyll) and from nine (mineral content) measurements. *, < 0.05; **, < 0.01; ***, < 0.001.

Chlorosis is linked with a significant decrease in Mn and Chl content

The trace element analysis of plants grown at low temperature indicated that the Mn content of the chlorotic plants (IL, nramp1-1, Hog) was significantly lower (< 0.01) in comparison to the corresponding controls (Ler, Col-0 and Hog-N922, respectively) (Fig. 5b). The Fe content was significantly lower in the Hog and nramp1-1 lines (< 0.01) when compared with Hog-N922 and Col-0, respectively, but not in the IL line in comparison to the Ler accession (Fig. 5b). A range of other trace elements were significantly changed in the chlorotic plants (IL, nramp1-1, Hog) in comparison to the corresponding controls (Ler, Col-0 and Hog-N922, respectively), but the only element significantly changed in all chlorotic lines was Mn (Table S11). The Mn- and temperature-dependent chlorosis of plants carrying mutated alleles of the NRAMP1 gene was associated with a significant decrease in the chlorophyll content (< 0.05) and a slight increase in the Chla: Chlb ratio (Fig. 5c), indicating the particular reduction of PSII. As shown by a qPCR result, observed chlorosis was not linked with altered level of NRAMP1 transcript (Fig. 6).

Figure 6.

The relative quantification of the NRAMP1 transcript level in Arabidopsis thaliana plants grown on manganese(Mn)-deficient soil at low temperatures. NRAMP1 gene expression level determined by qPCR was normaliszed to the reference gene ACT2. Plants becoming chlorotic at low temperatures carry mutated alleles of the NRAMP1 gene. Three lines carrying NRAMP1 wild-type alleles remain green at 4°C. Five tested genotypes Col-0, Ler, Hog-N922, Hog and introgression line (IL) (excluding T-DNA KO mutant line nramp1-1) show similar NRAMP1 gene expression level. Error bars, ± SD, from three measurements.

Discussion

A spontaneous loss-of-function mutation in the NRAMP1 gene has been identified and we showed that it is responsible for the impaired plant growth and chlorosis present under particular conditions in one Arabidopsis accession. The mutated allele has only been detected in the line derived from a single plant. Spontaneous laboratory mutations have been reported previously in Arabidopsis: HUA2 (Doyle et al., 2005) and TZP (Loudet et al., 2008). As shown in our experiments, a single mutation of a highly conserved NRAMP-specific residue has a great impact on plant performance. Nonsynonymous mutations, which cause a change from one amino acid to another, can have dramatic consequences for the performance of an organism, particularly when altering critical positions in a protein's structure (Lenski, 2001).

Several site-directed mutagenesis studies have investigated the structure–function relationship in the NRAMP protein family. The Escherichia coli Nramp ortholog, MntH, was used as a model to study the structure–function relationship in the Nramp protein family (Courville et al., 2008) where the conserved peptide motifs Asp-Pro-Gly (TMD1) and Met-Pro-His (TMD6), forming antiparallel ‘TM helix/extended peptide’ boundaries, are key for transport function. In another study, two highly conserved histidines were identified in the MntH transmembrane domain 6 (TMD6), for which an essential role in transport activity was proposed (Haemig et al., 2010). Identification of more mutation-sensitive and highly conserved residues in TMD4 and TMD6 led to the conclusion that both regions are essential for transporter activity, probably by participating in the formation of a selective pathway through which protons and metal ions pass (Nevo & Nelson, 2006). Mutant analysis performed by Chaloupka et al. (2005) suggested the importance of four Nramp family-specific residues within TMDs 1, 6 and 11 in metal uptake and metal-dependent H+ transport. These residues are candidate sites of functional divergence of the Nramp family.

New mutations that influence fitness in an environment-dependent manner might be a primary basis for the evolution of ecological specificity (Kavanaugh & Shaw, 2005). They could contribute to the ecological specificity by one of two alternative modes proposed by Fry et al. (1996): (1) enhancing fitness in one environment while reducing in a second environment or (2) not enhancing fitness in either environment but reducing it to a different degree depending on the environment. The results presented here are in accordance with the hypothesis that new, spontaneous mutations reduce fitness more severely in more stressful environments (Kondrashov & Houle, 1994; Korona, 1999; Szafraniec et al., 2001; Fry & Heinsohn, 2002). It was shown here that the H239Y substitution in the TMD6 of Arabidopsis NRAMP1 has a dramatic effect on plant performance at chilling temperatures, leading to seedling lethality when mutants were grown directly at low temperature. Our results confirm the proposition of Chaloupka et al. (2005) that changing the outgroup Tyr to the corresponding Nramp-specific His residue could have contributed to the functional divergence of the Nramp family, and that this site is structurally and functionally important. It can be concluded that the basic amino acid group of His239 may be required for AtNRAMP1 function. The effects of mutations of the corresponding conserved histidine residue pair in TMD6 (His267 equivalent to E. coli His211, and His272 equivalent His216) have also been analysed in the mouse Nramp2 homologue (Lam-Yuk-Tseung et al., 2003). An involvement in the pH regulation of transport was suggested for this histidine pair by the same authors. By contrast, the study of Haemig et al. (2010) did not strongly support a role of the conserved histidine residues (His211 and His216) with regard to pH regulation of the transporter, nor in its metal binding properties. Instead, it was suggested that the conserved residues of TMD6 may be critical for protein conformational changes during transport. Conserved residues in the transmembrane domains are essential for all transport functions; some mutations have also changed the metal ion specificity (Cohen et al., 2003). Although our results clearly show an important role for His239 in AtNRAMP1 function, without further studies it is not possible to answer how H239Y substitution affects the transport process. Does it affect the enzymatic function of the NRAMP1 transporter? Or, rather, does it affect the protein biogenesis, sub-cellular targeting or stability of AtNRAMP1? Further experiments including measurement of Mn uptake of Hog plants transformed with functional NRAMP1, measurement of Mn uptake defects in nramp1-1 plants complemented with Hog NRAMP1, or studies using the Xenopus oocytes and Saccharomyces cerevisiae heterologous expression systems (Nevo & Nelson, 2006) are needed to answer these questions.

The exact physiological roles of NRAMP transporters in planta are not well understood. In this study, the combined effects of low temperature and nutritional stresses on the plant performance were analysed. We showed that conditional mutation of the NRAMP1 gene had a severe effect on plant growth and optimal photosynthesis under Mn deficiency and chilling temperatures. The ratio of variable to maximum fluorescence (FV/FM) was significantly reduced under this condition, which might be linked to the reduction of quantum efficiency of the PSII. A parallel reduction in ΦII and qP parameters suggests lower efficiency of photochemical energy conversion and inhibition of PSII electron transport in plants carrying mutated alleles of NRAMP1. Low temperatures cause a decrease in membrane fluidity that further stimulates photoinhibition under Mn-deficient conditions. Mn deficiency may lead to a complete loss of crops during winter, a phenomenon called ‘winterkill’ (Schmidt et al., 2013). The ‘winterkill’ complex is a result of weak growth, increased disease susceptibility and frost intolerance of Mn-deficient plants. One could expect that applying the combination of Mn deficiency and other environmental stresses will affect plant performance more severely. The observed chlorosis and impaired plant growth was not caused by low external Mn concentration per se, but rather by excess of external Fe leading to Mn deficiency. It is commonly known that soils high in available Fe or with high Fe applications can reduce Mn uptake. Indeed, significantly lower Mn content in the leaves of plants carrying a mutated allele of NRAMP1 was confirmed by ICP-MS analysis. There are different mechanisms that could possibly explain this effect of Fe excess: (1) a reduction of Mn concentration by the dilution effect; (2) an antagonistic effect of Fe on Mn absorption by plant roots; (3) a reduction in root-to-shoot translocation; and (4) a toxic effect of Fe over-accumulation due to Fenton's reaction (Mortvedt, 1991, 1994; Moosavi & Ronaghi, 2010, 2011).

Even though AtNRAMP1 was clearly shown to be responsible for Mn uptake in low-Mn conditions and to be marginally associated in Fe uptake from the soil, its specific contribution to Fe nutrition remains unclear (Cailliatte et al., 2010). The NRAMP1 relative affinity to Fe and its expression profile in response to environmental factors might be the reason for that (Thomine & Vert, 2013). As shown previously, the NRAMP1 overexpressor lines were hypertolerant to toxic Fe (Curie et al., 2000). Higher tolerance of overexpressor lines, which have increased Mn content, was suggested to be an indirect effect (Cailliatte et al., 2010). The reason for that could be either the competition of Mn with tissue Fe content resulting in minimizing Fe toxicity for the plants, or protection against Fenton's reaction. The selectivity of NRAMP1 for Mn and Fe might be linked with their availability in the root apoplast or, alternatively, the Fe activity of NRAMP1 could be masked by other transporters (Cailliatte et al., 2010). AtNRAMP1 was shown by Cailliatte and coworkers to be essential but not absolutely crucial for plants under particular environmental conditions, as we showed in this study.

By analysing the growth of plants on different soils derived from both natural ecosystems and agricultural systems, which contain various combinations of essential micronutrient, it is possible to achieve a more integrated picture of how plants manage their ionome (Palmer & Guerinot, 2009). Because Mn and Fe homeostasis, and their bioavailability, are closely linked to those of other micronutrients, our understanding of the roles of various minerals needs to be comprehensively integrated (Kobayashi & Nishizawa, 2012). Our results confirm the importance of Mn homeostasis, which if often underestimated. Further molecular and biochemical analysis, including site-directed mutagenesis, are necessary to gain a better understanding of the contribution of plant NRAMP transporters to Me2+ metabolism and homeostasis. It might be beneficial for the nutritional enhancement of crop plants.

Plants respond to a specific combination of stresses in a nonadditive manner (Atkinson & Urwin, 2012) and, in this case, impacts on plant performance cannot be predicted simply from the study of individual stresses (Mittler, 2006). The results presented here show that plant responses to a combination of various environmental stresses must be taken into account in the functional analysis of metal transporters. Moreover, due to the possible redundancies and specialized functions of the particular classes of transition metal transporters, elucidating their respective functions in planta is not straightforward. The identification and characterization of mutants with altered Mn phenotypes not only increases the understanding of the mechanism of Mn toxicity and tolerance in plants, but also allows the identification of novel components of Mn homeostasis (Pittman, 2005). Such knowledge could be used in breeding plants for low or enhanced Mn content, which is particularly important in developing countries.

Acknowledgements

This research was supported by the Max Planck Society and the University of Gdansk (Gdansk University Grant 538-M031-B145-13). We thank Ms Sigi Effgen, Ms Barbara Eilts and Ms Regina Gentges for their assistance in the handling of plants.

Ancillary