Cesium Uptake by Rice Roots Largely Depends Upon a Single Gene, HAK1, Which Encodes a Potassium Transporter

Abstract

Incidents at the Fukushima and Chernobyl nuclear power stations have resulted in widespread environmental contamination by radioactive nuclides. Among them, ¹³⁷cesium has a 30 year half-life, and its persistence in soil raises serious food security issues. It is therefore important to prevent plants, especially crop plants, from absorbing radiocesium. In Arabidopsis thaliana, cesium ions are transported into root cells by several different potassium transporters such as high-affinity K⁺ transporter 5 (AtHAK5). Therefore, the cesium uptake pathway is thought to be highly redundant, making it difficult to develop plants with low cesium uptake. Here, we isolated rice mutants with low cesium uptake and reveal that the Oryza sativa potassium transporter OsHAK1, which is expressed on the surfaces of roots, is the main route of cesium influx into rice plants, especially in low potassium conditions. During hydroponic cultivation with low to normal potassium concentrations (0–206 µM: the normal potassium level in soil), cesium influx in OsHAK1-knockout lines was no greater than one-eighth that in the wild type. In field experiments, knockout lines of O. sativa HAK1 (OsHAK1) showed dramatically reduced cesium concentrations in grains and shoots, but their potassium uptake was not greatly affected and their grain yields were similar to that of the wild type. Our results demonstrate that, in rice roots, potassium transport systems other than OsHAK1 make little or no contribution to cesium uptake. These results show that low cesium uptake rice lines can be developed for cultivation in radiocesium-contaminated areas.

Introduction

Cesium (Cs) is a Group I alkali metal whose chemical properties are very similar to those of potassium (K). Cs can interfere with K uptake and metabolism in plants, but stable isotope Cs (¹³³Cs) concentrations in natural soils are generally low (up to 25 µg g^–1 soil) and not harmful to plants or human health (White et al. 2000). However, the release of radiocasium (¹³⁴Cs, ¹³⁷Cs) into the environment by incidents at nuclear reactors has led to serious environmental concerns, because these forms of Cs are rapidly incorporated into biological systems and result in internal exposure to β and γ radiation during their decay (White et al. 2000, Zhu and Smolder, 2000). Several radionuclides such as ¹³¹iodine (¹³¹I), ⁹⁰strontium (⁹⁰Sr),¹³⁴Cs and ¹³⁷Cs were discharged into the atmosphere after damage to the Fukushima Daiichi nuclear power plant on March 11, 2011 (Chino et al. 2011). Although a large amount of ¹³¹I was released, it was not detectable after several months due to its short half-life (8 d), and ⁹⁰Sr contaminated only a small area because of its low volatility. In contrast, radiocesiums are highly volatile and have long half-lives (¹³⁴Cs, 2.07 years; ¹³⁷Cs, 30.1 years) (Chino et al. 2011, Kinoshita et al. 2011). Therefore, radiocesium (especially ¹³⁷Cs) contamination is widespread and persistent (Yasunori et al. 2011). It is important to prevent radiocesium from entering the food chain, especially major crops (Smith et al. 2000, Zhu and Smolder, 2000)

After the Chernobyl accident, the environmental behaviors of radiocesium were investigated in detail (Fujiwara 2013). The vertical migration of radiocesium is very slow. It tends to remain in surface soils because it becomes fixed in the frayed edges of illite interlayers of the clay (Almgren et al. 2006, Smith 2011). However, stripping topsoil requires the transport and long-term management of the removed material, which is excessively costly for agricultural fields contaminated with low level radioactivity (Internatiol Atomic Energy Agency 1999, Fesenko et al. 2007). Another lower cost strategy to manage radiocesium contamination is to apply K fertilizer (Zhu et al. 2000), but this does not completely suppress ¹³⁷Cs uptake by major crops such as rice (Fujimura et al. 2013, Kato et al. 2015). Therefore, another approach is required.

There have also been many physiological studies on plant Cs uptake over the past 70 years. Those studies showed that K⁺ and Cs⁺ compete for influx into roots, suggesting that the same molecular mechanism underlies their uptake (White et al. 2000, Zhu and Shaw, 2000). Recently, K⁺ transporters (including channels) have been identified by heterologous expression analyses and homology searches of genomic databases (Lebaudy et al. 2007, Szczerba et al. 2009). In Arabidopsis, K⁺ transporters are encoded by 35 genes in five families (Mäser et al. 2001, Gomez-Porras et al. 2012). Molecular analyses of the plant Cs⁺ uptake pathway in Arabidopsis have shown that K⁺ is transported into roots through the co-operative action of the high-affinity K⁺ transporter AtHAK5 and the inward-rectifying K⁺ (KIR) channel AKT1 in normal soil conditions (Gierth et al. 2005, Rubio et al. 2008, Pyo et al. 2010, Rubio et al. 2010, Nieves-Cordones et al. 2016). athak5 mutants accumulated significantly less Cs than did the wild type (WT) when grown with a high concentration of Cs⁺ (300 µM). Heterologous expression analyses showed that AtHAK5 mediated Cs⁺ uptake into yeast cells. However, there was no significant difference in Cs⁺ influx rate between the WT and the athak5 mutant under high K⁺ (2 mM K⁺) or K⁺ deprivation conditions (0.5 µM K⁺). These findings indicated that AtHAK5 transports Cs⁺ and K⁺ in planta, but it is not the sole pathway of Cs⁺ uptake in Arabidopsis (Qi et al. 2008). Also, Cs⁺ influx was not suppressed in AKT1 channel knockout lines (Broadley et al. 2001, Qi et al. 2008). These results implied that Cs⁺ enters root cells via several K⁺ transporters including AtHAK5 in Arabidopsis (Smith et al. 2000); however, it was unknown whether a similar redundant Cs/K transport system exists in rice. Here, we isolated three mutant strains of rice that showed significantly reduced uptake of Cs⁺, compared with that of WT rice. Our results showed that in rice, most of the Cs⁺ influx into roots depends on the high-affinity K⁺ transporter OsHAK1.

Results

Mutagenesis of rice selection of lcs mutants

Mutations were induced in Japonica rice cv. Akitakomachi by using two chemical mutagens, sodium azide and N-methyl-N-nitrosourea (MNU). The mineral concentrations in de-husked M₄ seeds from the 8,027 individual plants were determined. We selected three mutants with less than one-fifth of the ¹³³Cs concentration and K:Cs ratio compared with those of the WT. Hence, lcs (low Cs accumulation)-1 was isolated from the population treated with sodium azide, and lcs-2 and -3 were isolated from the population treated with MNU.

Cultivation of lcs mutants in the field

When the lcs mutants were cultivated in a paddy field (Supplementary Table S1), their grain Cs concentrations were <10% of that in the WT (Fig. 1a) and their shoot Cs concentrations were also lower than that in the WT (Fig. 1c). The mutants showed similar growth and grain yields to those of the WT (although lcs-1 was a dwarf variety) (Fig. 1b, e, f). The shoot K level in lcs-3 was 68% of that in the WT, but the shoot K concentrations in the other two mutants were not significantly affected (Fig. 1d).

Radiocesium uptake imaging of lcs mutants

To visualize Cs uptake, seedlings were grown for 2 weeks under hydroponic conditions (Supplementary Table S2), and then treated with ¹³⁷Cs. After a further 18 h, whole plants were subjected to imaging analysis to visualize ¹³⁷Cs. Much less Cs accumulated in the roots of mutants than in those of the WT (Fig. 1g).

Identification of the causative gene for low Cs uptake in lcs mutants

To investigate whether the same gene was mutated in all three mutants, we conducted allelism tests. The root concentrations of Cs in the F₁ and F₂ seedlings were <10% of that in the WT. There were no significant differences in root Cs concentrations among the progeny of the mutants’ crosses (Supplementary Fig. S1a, b), suggesting that the low Cs uptake characteristics were attributable to the same gene in each lcs mutant.

We identified the gene responsible for low Cs uptake by gene mapping and whole-genome sequencing. The gene was mapped to the long arm of chromosome 4, between linkage map marker C975 (located at 52.6 cM) and marker S10628 (58.9 cM). Whole-genome sequencing analysis indicated that all three mutants carried the mutation in the OsHAK1 gene (locus LOC_Os04g32920) (Amrutha et al. 2007, Yang et al. 2009), but the affected nucleotides varied among the mutants. In lcs-1 and lcs-3, the same G→A point mutation in exon VIII of OsHAK1 resulted in an amino acid change from glycine to aspartic acid in OsHAK1 (Fig. 2a). This substitution was located in the intracellular domain of OsHAK1 (Fig. 2b), in which the K⁺ transporter activity is stimulated by phosphorylation (Szczerba et al. 2009, Li et al. 2014, Nieves-Cordones et al. 2016). In lcs-2, a G→A point mutation introduced a stop codon in exon II (Fig. 2a), resulting in premature termination of translation of the transmembrane region, and probably causing a complete loss of OsHAK1 function.

Genetic complementation of lcs mutants with full-length OsHAK1

To determine whether the mutant phenotype was caused by the loss of function of OsHAK1, we performed complementation analyses by introducing a genomic DNA fragment containing the OsHAK1 gene from WT rice [obtained from a bacterial artificial chromosome (BAC) clone, OSJNBb0012E08] (Amrutha et al. 2007) into lcs-3 via Agrobacterium tumefaciens-mediated transformation. In a hydroponic Cs uptake experiment, the Cs concentration in the roots of lcs-3–OsHAK1 transformants was the same as that in WT roots (Fig. 2c). Thus, OsHAK1 was proven to be the gene responsible for the defective Cs accumulation in the mutants.

Cs uptake by lcs mutants in solutions with different K concentrations

Since Cs⁺ uptake and K⁺ uptake are thought to be linked, we determined the relative levels of Cs⁺ and K⁺ influx in WT and lcs mutants. Two-week-old seedlings were transplanted and grown under hydroponic conditions with different concentrations of K⁺ (from 0 to 1.65 mM; Supplementary Table S2), and Cs⁺ at 10 µg l^–1. The Cs and K contents in roots were measured at various times up to 5 d after transplantation. At K⁺ concentrations of ≤0.55 mM, the root Cs concentrations in the lcs mutants were approximately a quarter those in the WT, and the rate of increase was also higher in the WT than in the mutants (Fig. 3a, b; Supplementary Fig. S2a, b). As the external K⁺ concentration increased, the root Cs concentration decreased in the WT, but only slightly decreased in the lcs mutants. The external K⁺ concentration had a much stronger effect on Cs uptake in the WT than in the mutants. At the external K⁺ concentration of ≥1.1 mM, Cs accumulated to the same levels in the WT and the lcs mutants (Fig. 3a–c; Supplementary Fig. S2). The Cs concentration was always lower in the shoots than in the roots, and there was a positive correlation between shoot and root Cs concentrations (Supplementary Fig. S3).

Next, we focused on K⁺ uptake in the lcs mutants. The root concentrations of K were significantly lower in the lcs mutants than in the WT at a K⁺ concentration of ≤1.1 mM (Supplementary Fig. S4). Although there were significant differences in shoot K concentrations between the lcs mutants and the WT, they were smaller than the differences in Cs accumulation between the lcs mutants and the WT (Supplementary Fig. S5). These results suggested that K⁺ was efficiently transported from roots to shoots and that the amount of K⁺ taken up into roots was sufficient for shoot growth in the lcs mutants.

The Cs⁺ uptake rates were measured in hydroponic solutions with different K⁺ concentrations (Fig. 3d). In the lcs mutants, the Cs⁺ uptake rates were low, and were not affected by the external K⁺ concentrations. In the range of 14–138 µM K⁺, which is equivalent to the K⁺ levels in normal soil solutions (Gomez-Porras et al. 2012), the Cs⁺ uptake rates in lcs mutants were only 5.9–12.5% of that in the WT (Fig. 3d). The K⁺ uptake rates into the roots in the lcs mutants were also analyzed in the different K⁺ conditions. The K⁺ uptake rates did not differ greatly between the lcs mutants and the WT (Supplementary Fig. S6). Our results showed that Cs⁺ influx into rice roots was mainly through OsHAK1, and that other (remaining) systems that take up K⁺ from the soil into roots contributed much less to Cs⁺ uptake in the lcs mutants (Fig. 3e, f). The residual Cs⁺ uptake in lcs mutants was not affected by the external K⁺ concentration, suggesting that this transport may be via mechanisms other than K⁺ transporters (Fig. 3d). It might be possible to reduce Cs⁺ uptake further by targeting these transport mechanisms.

Cultivation of lcs mutants in a ¹³⁷Cs-contaminated paddy field

To demonstrate their utility, we cultivated the three lcs mutants in a ¹³⁷Cs-contaminated paddy field (approximately 3,500 Bq kg^–1 soil) which was deficient in available K (79 mg kg^–1 exchangeable K, much lower than the recommended level of 200 mg kg^–1) to decrease ¹³⁷Cs accumulation in brown rice (Kato et al. 2015) (Supplementary Table S3). The ¹³⁷Cs levels in shoots were much lower in the lcs mutants than in the WT. The ¹³⁷Cs levels in the grain of the lcs mutants were less than one-tenth that in WT grain (Fig. 4a, b). The shoots of lcs mutants were slightly shorter and they showed delayed flowering. At harvest, the straw of the lcs mutants was softer than WT straw. The mature grain of the mutants could be harvested and their yields were similar to that of the WT (Fig. 4c).

Discussion

In rice, K⁺ transport systems are encoded by >50 genes in five families (Amrutha et al. 2007, Gomez-Porras et al. 2012). Among them, 27 genes encode HAK transporters (Yang et al. 2009). OsHAK1 is one of the major HAK transporters expressed in roots and is associated with high-affinity K⁺ uptake (Gupta et al. 2008, Chen et al. 2015). In rice, K⁺ is transported into the roots through the co-operative action of OsHAK1 and the KIR channel OsAKT1 (Nieves-Cordones et al. 2016), which correspond to AtHAK5 and AKT1, respectively, in Arabidopsis thaliana (Gierth et al. 2005, Lebaudy et al. 2007, Rubio et al. 2008). The influx of Cs⁺ is not greatly suppressed in either the athak5 or akt1 mutants in A. thaliana (Broadley et al. 2001, Qi et al. 2008). Several studies have shown that HAK transporters, KIR, outward-rectifying K⁺ channels and voltage-insensitive cation channels are all permeable by Cs⁺ (Maathuis et al. 1994, White et al. 2000, Hampton et al. 2005). The actions of multiple transporters in Arabidopsis enable Cs⁺ uptake to continue despite knockout of AtHAK5 and AKT1. In contrast, in rice plants, our results clearly showed that Cs⁺ influx into rice roots is mainly through OsHAK1. The Cs⁺ transport functions of HAK have been demonstrated in heterologous expression analyses. The yeast-expressed chimeric OsHAK1 (incomplete OsHAK1 cDNA fused with the first 48 bp of barley HvHAK1 encoding 16 amino acids) took up Cs⁺ with an estimated K_m of 11 µM (Bañuelos et al. 2002). Our results showed that other K⁺ transport systems besides OsHAK1 made little contribution to Cs uptake from soil to roots in the lcs mutants.

The contribution of OsHAK1 to K⁺ uptake increased in low external K⁺ conditions (50–55% in 50–100 µM and 30% in 1 mM K⁺). Previous studies reported that OsHAK1 expression was induced 8 ‐ to 12-fold under K⁺ starvation (Okada et al. 2008, Chen et al. 2015). This helps to explain why the application of K⁺ fertilizer reduces Cs⁺ uptake. The application of K fertilizer not only increases the K⁺:Cs⁺ ratio, but also maintains OsHAK1 expression at low levels. Furthermore, in field-grown plants, the relative differences in grain Cs concentrations between the lcs mutants and the WT were larger than those in shoot Cs concentrations between the lcs mutants and the WT. As reported in previous studies, OsHAK1 is expressed mainly in roots but also in shoots (Bañuelos et al. 2002, Chen et al. 2015). Transcripts of OsHAK1 were detected in the root–shoot junctions, stems, leaves and panicle axes (Chen et al. 2015). Therefore, it was suggested that OsHAK1 is involved in Cs⁺ influx into the root and also in the translocation of Cs⁺ from shoots to grains. If this is the case, then the accumulation of Cs in the grains would be suppressed at these two steps in the lcs mutants.

Recently, it was reported that Cs uptake was reduced by mutating OsSOS2, which encodes a kinase that phosphorylates OsSOS1 (an Na⁺/H⁺ antiporter) in root cells. OsSOS1 is activated by phosphorylation, so the knockout of OsSOS2 led to reductions in Na⁺ extrusion from the root cells by OsSOS1. The increased Na⁺ concentration in the root cells down-regulated the expression of genes encoding HAK transporters and AKT channels such as OsHAK1, OsHAK5, OsAKT1 and OsHKT2;1. The down-regulation of these K⁺ transporters also reduced Cs uptake into rice roots (Ishikawa et al. 2017).

Our results have identified OsHAK1 as the main mediator of Cs uptake among these K⁺ transporters. The lcs mutants in the present study had a loss of function of only OsHAK1, and accumulated much less Cs in the roots and shoots in low K⁺ (<0.2 mM) hydroponic conditions. Furthermore, oshak1 mutants did not require Na⁺ for the reduction of Cs uptake.

Finally, when the lcs mutants were cultivated in the field in ¹³⁷Cs-contaminated, low K soil, the ¹³⁷Cs contents in the grains were less than one-tenth that in WT grains. Naturally occurring radionuclides are present in the environment. The concentration of ⁴⁰K (mainly β^– decay; 1.31 MeV) in brown rice usually exceeds 50 Bq kg^–1. However, compared with ⁴⁰K, the concentration of ¹³⁷Cs (β^–, γ; 1.18 MeV) was much lower in the lcs mutants (<5 Bq kg^–1). This result showed that cultivation of lcs mutants in ¹³⁷Cs-contaminated soil with low level can be an effective substitute for, or can be combined with, other countermeasures such as decontamination by soil stripping or application of K fertilizers.

Soil contains several geochemically harmful elements that are toxic to human health and are ingested via crops. Rice is a staple crop for half of the world’s population however, it is a major dietary source of toxic cadmium (Cd) and arsenic (As) (Gilbert-Diamond et al. 2011, Rahman et al. 2011, Ishikawa et al. 2012). Countermeasures to reduce the concentrations of toxic compounds in crops such as soil removal or covering with uncontaminated soils are extremely expensive. Research on rice transporters has shown that Cd and As uptake into roots can be lowered by knocking out manganese and silicon transporters, respectively (Ma et al. 2008, Ishikawa et al. 2012).

We have shown that Cs uptake in rice was dramatically reduced by knocking out OsHAK1, which encodes a high-affinity K⁺ transporter, while other K⁺ uptake systems compensated to take up sufficient K⁺ for growth in the absence of OsHAK1. Breeding lines with low uptake of harmful elements is a promising strategy to create crops that can be grown in polluted soils at a low cost. In addition, these non-transgenic mutant lines may be more acceptable to the public than genetically engineered lines.

Materials and Methods

Mutagenesis of rice

Mutations were induced using sodium azide and MNU. Plants of O. sativa ssp. japonica cv. Akitakomachi were treated with 1.5 mM sodium azide in 100 mM sodium phosphate buffer (pH 3.0) for 6 h with aeration. Fertilized egg cells were treated with MNU. Blossoming flowers on panicles were marked and then treated with 1.5 mM MNU solution at 12:00 h on the day of flowering. After 1 h, the treated panicles were washed with running water for 10 h. The M₁ seeds were harvested from the treated panicles and grown in the field, and the M₂ generation seeds were harvested in bulk for each of the mutant populations.

Selection of lcs mutants

The M₃ seedlings were grown for about 1 month in the nursery and then transplanted into heavy-metal-contaminated paddy field soil containing a high concentration of ¹³³Cs. After 5 months, the rice plants were harvested separately. De-husked grains were decomposed with concentrated HNO₃, and minerals were quantified by inductively coupled plasma mass spectrometry (ICP-MS; X series II, Thermo Fisher Scientific).

Cultivation of lcs mutants in the field

Three mutant lines with low Cs uptake phenotypes were grown in a paddy field in Akita Prefecture, Japan, in 2015. The grains and shoots were harvested to analyze Cs accumulation. The paddy field was located in clay-rich reclaimed land (Supplementary Table S1).

Radiocesium uptake imaging of lcs mutants

The ¹³⁷Cs uptake and distribution in the seedlings was visualized with BAS Imaging plates. Two-week-old seedlings were transplanted into a 0.275 mM K⁺ hydroponic solution (Supplementary Table S2) containing 10 p.p.b. Cs and 100 Bq l^–1137Cs. After 18 h, the seedlings were placed in contact with a BAS Imaging plate (BAS IP MS 2040 E; GE Healthcare). After 16 h contact time, the plate was scanned using a Typhoon FLA 9500 scanner (GE Healthcare).

Identification of a causative gene for low Cs uptake in lcs mutants

Allelism tests. F₁ hybrids were created by cross-breeding between the three mutant lines. Some individuals in the F₁ generation and self-fertilized F₂ generation were subjected to allelism tests. To analyze Cs uptake, 2-week-old seedlings were grown in a 0.275 mM K⁺ hydroponic solution containing 10 p.p.b. Cs. After 48 h, Cs concentrations in the roots and shoots were determined.

Gene mapping of the causative gene for low Cs uptake. F₂ generations derived from crosses between each of the three mutant lines and the O. sativa ssp. indica cv. Kasalath were used to map the gene controlling Cs uptake. The concentrations of Cs in the roots of the F₂ generations were measured in hydroponic cultivation conditions as described above. Genomic DNA was extracted from the shoots of 82 seedlings identified as low Cs uptake plants when cultivated in a 0.275 mM K⁺ hydroponic solution containing 10 p.p.b. Cs. Mapping was carried out using sequence-tagged site and cleaved amplified polymorphic sequence markers, which mapped the locus of the gene responsible for low Cs uptake to the region between 52.6 and 58.9 cM on chromosome 4.

Whole-genome sequencing of lcs mutants. Genomic DNA was extracted from 10 seedlings of each of the three mutant lines after analyzing the Cs concentrations in the roots. Genomic DNA was also extracted from the Akitakomachi cultivar as a control. Whole-genome sequences were acquired with a HiSeq1000 instrument (Illumina), with a read length of 100 bp, and aligned to the reference sequence (IRGSP1.0) using the default conditions of the Burrows–Wheeler Aligner. The chromosomal region identified by mapping was visualized with the Integrative Genomics Viewer. We searched for genetic variations by comparing the nucleotide sequences of the mutants with that of the Akitakomachi cultivar.

Sanger sequencing of OsHAK1 cDNA from three mutants. For each mutant and for control cultivars (Akitakomachi, Nipponbare, Koshihikari and Kasalath), total RNA was extracted with a NucleoSpin^® RNA Midi kit (Macherey Nagel) from the roots of seedlings grown for 2 weeks in hydroponic conditions. Full-length first-strand cDNA was synthesized from total RNA with SuperScript^®III First-Strand Synthesis SuperMix (Thermo Fisher Scientific) and oligo(dT)₂₀ primers. The OsHAK1 cDNA was amplified by PCR with KOD-Plus-Neo DNA polymerase (Toyobo) and the primers HAK1_5primeF, 5′‐CCAGCCGGCGAGAGAGAGC-3′; and HAK1_3primeR, 5′‐AGCATGGACAACACACCACCAGTG-3′. These primers were designed according to the O. sativa ssp. japonica cv. Nipponbare OsHAK1 sequence deposited at the Rice Genome Annotation Project website (http://rice.plantbiology.msu.edu/index.shtml).

The PCR products were separated by electrophoresis on an agarose gel, and the amplified DNA band was purified using a MonoFas DNA purification kit I (GL Sciences). The cDNAs were sequenced with the BigDye Terminator Cycle Sequencing kit 3.1 (Applied Biosystems) using the primers shown in Supplementary Table S4, with an ABI3130xl genetic analyzer (Applied Biosystems).

Construction of OsHAK1 protein model

A ribbon representation of the three-dimensional model of OsHAK1 in lcs-3 was generated using the web-based Phyre2 protein fold recognition server (Kelley et al. 2015). The transmembrane domain, cytoplasmic domain and post-transcriptional activation domain are indicated in green, cyan and blue, respectively. The figure was prepared with PyMol (Schrödinger).

Genetic complementation of lcs mutants with OsHAK1

A genomic DNA fragment including OsHAK1 and its native promoter was excised with HindIII from the BAC clone OSJNBb0012E08, which was derived from the Nipponbare genome. The fragment was separated and purified by agarose gel electrophoresis, and inserted into the HindIII-digested, alkaline phosphatase-treated pPZP2 vector (Fuse et al. 2001).

The OsHAK1–pPZP2 recombinant vector was introduced into A. tumefaciens strain EHA101 for transformation into the low Cs uptake mutant lcs-3. Callus induction, infection of the callus with A. tumefaciens EHA101 harboring recombinant clones, selection of calli for resistance to 50 µg ml^–1 hygromycin and subsequent regeneration of calli into mature plants were conducted as described previously (Ozawa et al. 2012). The selected transformed calli were regenerated on re-differentiation medium. The resulting T₀ plants were grown and the T₁ seeds were harvested. The T₁ seedlings were grown in hydroponic conditions. After 2 weeks cultivation, the Cs concentrations in the roots of the transformants were determined 24 h after transfer to a 0.138 mM K⁺ hydroponic solution (Supplementary Table S2) containing 10 p.p.b. Cs.

Cs uptake of lcs mutants in solutions with different K concentrations

The Cs absorption characteristics of the lcs mutants (lcs-2 and -3) were determined using seedlings grown in hydroponic conditions. To generate these seedlings, seeds were sterilized and germinated for 3 d before transplanting onto floating nets. Seedlings were grown at 28°C under a 16 h light, 8 h dark photoperiod in partially modified Kimura B solution as the hydroponic medium (containing 0.55 mM K⁺; Supplementary Table S2). After 2 weeks of cultivation, the seedlings were transferred to hydroponic solutions containing 10 p.p.b. Cs (as cesium chloride) and K⁺ at various concentrations (Supplementary Table S2). The Cs levels in the seedlings were determined over the following 5 d. Seedlings were separated into roots and shoots, and Cs and K concentrations were determined by ICP-MS.

Cs⁺ uptake rate of the roots in oshak1 and the WT

To evaluate the Cs⁺ uptake rate into roots, 2-week-old seedlings of the lcs mutants (lcs-2 and -3) and the WT were transplanted into 10 different hydroponic solutions (Supplementary Table S2) containing 10 p.p.b. Cs. The root and shoot Cs concentrations were determined after 24 h growth in these solutions. The Cs⁺ uptake rate was calculated as follows: Cs uptake rate (µg g^–1 d^–1) = total Cs uptake (Cs in the roots + Cs in the shoots) ÷ fresh weight of roots.

Cultivation of lcs mutants in a ¹³⁷Cs-contaminated paddy field

The mutant strains lcs-2 and lcs-3 were grown in a ¹³⁷Cs-contaminated paddy field in Fukushima Prefecture, Japan, in 2016, and the grains and shoots were harvested. The paddy field contained ¹³⁷Cs at approximately 3,500 Bq kg^–1 soil (Supplementary Table S3). Accumulation of ¹³⁷Cs in the grain and shoots was analyzed with a germanium semi-conductor detector (GC2520-7500SL; Canberra Industries).

Statistical analysis

Statistical analyses were carried out with JMP software (SAS InstituteA). Statistical differences were assessed by Tukey’s or Dunnett’s tests. All P-values < 0.05 were regarded as statistically significant.

Supplementary data

Supplementary data are available at PCP online.

Funding

This work was supported by the Japan Society for the Promotion of Science (JSPS) [KAKENHI Grant-in-Aid for Scientific Research (C) (grant No. JP24780320)].

Abbreviations

Abbreviations
 
• As

  arsenic

 
• BAC

  bacterial artificial chromosome

 
• Cd

  cadmium

 
• Cs

  cesium

 
• HAK

  high-affinity K⁺ transporter

 
• HNO₃

  nitric acid

 
• I

  iodine

 
• ICP-MS

  inductivity coupled plasma mass spectrometry

 
• K

  potassium

 
• lcs

  low Cs accumulation

 
• MNU

  N-methyl-N-nitrosourea

 
• Sr

  strontium

 
• WT

  wild type

Acknowledgments

We thank Yukiya Kobayashi, Kentaro Yasuda and J.A. Akita-Shirakami for their support and for managing the experimental fields. We thank Junpei Takano for critically reading and revising the manuscript.

Disclosures

The authors have no conflicts of interest to declare.

References