Differences in mineral accumulation and gene expression profiles between two metal hyperaccumulators, Noccaea japonica and Noccaea caerulescens ecotype Ganges, under excess nickel condition

ABSTRACT

Here we compare mineral accumulation and global gene expression patterns between two metal hyperaccumulator plants – Noccaea japonica, originating from Ni-rich serpentine soils, and Noccaea caerulescens (ecotype Ganges), originating from Zn/Pb-mine soils – under excess Ni conditions. Significant differences in the accumulation of K, P, Mg, B, and Mo were explained by the expression levels of specific transporters for each mineral. We previously showed that total Ni accumulation in the whole plant is higher in N. caerulescens than in N. japonica. Here we found a similar tendency for Fe under excess Ni; however, the expression of iron-regulated transporter 1 (IRT1), which encodes the primary Fe uptake transporter and causes excess Ni uptake in Arabidopsis thaliana, was higher in N. japonica. NjIRT1 has a point mutation at Asp100, which is essential for Fe transport, and so might lack its Fe and possibly Ni transport function. Noccaea japonica might have lost its IRT1 function, which would prevent excess Ni uptake via IRT1 in Ni-rich soils, and come to rely on other transporters.

KEYWORDS:

• Metal hyperaccumulator
• nickel
• Noccaea
• RNA sequencing
• transporter

Some heavy metals, such as Zn and Ni, are essential for plant growth, but their excess accumulation causes severe growth inhibition. In metal-polluted soils, however, plant species that accumulate heavy metals in their shoots are found. They are called “metal hyperaccumulators”. A number of hyperaccumulators have been found in the genus Noccaea.¹ Metal accumulation and metal tolerance vary among species or accessions of Noccaea originating from different soil types,^2–4 but little is known about the genetic factors determining these variations. In a recent study, we found that Ni accumulation in shoots is significantly lower in Noccaea japonica originating from Ni-enriched serpentine soils than in Noccaea caerulescens ecotype Ganges (GA) originating from Zn/Pb-mine soils, on account of a lower efficiency of Ni translocation from roots to shoots in N. japonica.⁵ We also revealed a higher copy number of iron-regulated 2 (IREG2, also known as ferroportin 2), which encodes a Ni transporter that sequesters excess Ni into root vacuoles and suppresses the translocation of excess Ni from roots to shoots and, accordingly, higher IREG2 expression in N. japonica, suggesting that elevated IREG2 expression due to gene duplication decreases root-to-shoot Ni translocation in N. japonica. This alteration might be relevant to adaptation to Ni-rich soils. Here, we further compared mineral concentration patterns and global gene expression patterns between N. japonica and N. caerulescens GA.

Mineral concentrations in shoots and roots of plants grown under 0 or 25 µM NiCl₂ in hydroponic culture for 7 days were determined by inductively coupled plasma mass spectrometry (Figure 1, S1). Concentrations in Arabidopsis thaliana were used as reference. The plant samples were obtained in a previous study; A. thaliana showed severe chlorosis under 25 µM NiCl₂, but N. japonica and N. caerulescens GA showed no symptoms.⁵ Gene expression levels in the roots of plants under excess Ni were determined by de novo transcriptome analysis using RNA sequencing on the Illumina HiSeq platform (see Supporting Information for details: Fig. S2). The RNA samples, which were isolated from the roots of plants under 200 µM NiCl₂ in a previous study,⁵ were used. We confirmed that IREG2 expression was significantly higher in N. japonica than in N. caerulescens GA (Table 1),⁵ indicating effective de novo transcriptome analysis.

Table 1. Candidate genes involved in differences in mineral accumulation patterns between Noccaea japonica and Noccaea caerulescens.

Function	Arabidopsis homolog	Gene name	Log2 FC (Nj/Nc)	P-value	Description
Ni/Fe transport	At5g03570	IREG2	1.78	7.04E-03	Ni transporter involved in Ni sequestration in root vacuoles
	At4g19690	IRT1	2.73	1.86E-02	Major Fe uptake transporter in roots
Fe transport	At1g80830	NRAMP1	−3.09	8.60E-03	Fe uptake transporter in roots
	At2g23150	NRAMP3	−1.38	1.18E-03	Vacuolar Fe transporter involved in Fe homeostasis
	At4g19680	IRT2	−2.63	3.05E-02	Fe transporter involved in Fe homeostasis
Mg transport	At3g19640	MRS2-1	9.90	6.19E-03	Mg transporter
	At5g64560	MRS2-2	6.38	7.76E-05	Mg transporter
K transport	At1g32450	NPF7.3	−1.56	4.86E-03	K transporter for xylem loading involved in root-to-shoot K translocation
P transport	At5g43360	PHT1;3	−1.38	2.23E-03	Pi uptake transporter in roots
	At2g38940	PHT1;4	2.09	1.03E-02	Pi uptake transporter in roots
	At2g32830	PHT1;5	−6.35	3.59E-03	Pi transporter involved in Pi mobilization between organs
	At1g20860	PHT1;8	2.04	2.07E-02	Pi transporter involved in root-to-shoot Pi translocation
B transport	At1g80760	NIP6;1	−2.38	2.17E-03	B transporter involved in B distribution in shoots
Mo transport	At2g25680	MOT1	3.73	5.43E-03	Mo transporter involved in Mo uptake
	At1g80310	MOT2	2.79	5.22E-04	Mo transporter involved in vacuolar Mo export

FC, fold change. Differentially expressed genes with |log₂ FC| ≥ 1 and P < 0.05 (Welch’s t-test) were extracted. Nj, Noccaea japonica; Nc, Noccaea caerulescens.

Figure 1. Mineral profiles of Noccaea caerulescens ecotype Ganges, N. japonica, and Arabidopsis under excess Ni. Mineral concentrations in (a) shoots and (b) roots of plants exposed to 25 µM NiCl₂ in hydroponic culture for 7 days were determined. Half-strength MGRL solution was used for hydroponic culture. Detailed procedures for plant culture and sample preparation are described in a previous paper.⁵ Elemental analysis was conducted by inductively coupled plasma mass spectrometry (ELAN, Perkin Elmer Inc.). Log₂ geometric means of values relative to Arabidopsis are shown with SD (n = 4 biological replicates). Bars with the same letter within an element are not significantly different (P < 0.05, Bonferroni’s t-test) between species

There were significant differences in some minerals in roots or shoots between the two species. In shoots, K and B concentrations were higher and Mo concentration was lower in N. caerulescens GA (Figure 1a, S1A), and in roots, Mg and Mo concentrations were higher and P concentration was lower in N. japonica (Figure 1b, S1B), in both the Ni treatment and control. Differentially expressed genes, which we expected to cause the differential mineral accumulation, are summarized in Table 1. Differences in mineral profiles between the two species can be mostly explained by differences in the expression levels of specific mineral transporter genes: for example, higher K accumulation in the shoots of N. caerulescens GA is probably due to higher expression of nitrate transporter 1/peptide transporter family 7.3 (NPF7.3), which encodes the K transporter for xylem loading (which plays a major role in root-to-shoot K translocation in A. thaliana),⁶ in N. caerulescens GA.

High Mg concentration is characteristic of serpentine soils, and is another factor limiting plant growth.⁷ Root Mg was higher and shoot Mg tended to be lower in N. japonica, so the ratio of shoot Mg to root Mg was significantly lower in N. japonica (Fig. S3), indicating that translocation of Mg from roots to shoots is lower in N. japonica. The mitochondrial RNA splicing (MRS) family proteins have been identified as Mg transporters.⁸ Among MRS genes, MRS2-1 and MRS2-2 were highly expressed in N. japonica (Table 1), implying that the encoded proteins are involved in high Mg accumulation in roots in N. japonica. Genes considered necessary for Arabidopsis lyrata to adapt in serpentine soil include MRS2-2.⁹ High expressions of MRS2 genes in N. japonica might contribute to the adaptation to excess Mg in serpentine soils.

Fe accumulation in the roots tends to be higher in N. caerulescens GA in the Ni treatment (Figure 1b), but equivalent in the control (Fig. S1B). Ni treatment significantly increased root Fe uptake in N. caerulescens GA but not in N. japonica,⁵ suggesting that Fe transporter genes behave differentially between the species under excess Ni. Ni is taken up in A. thaliana by iron-regulated transporter 1 (IRT1); Ni-induced Fe deficiency induces IRT1 expression, promoting Ni uptake.^10–12 Ni accumulation in whole plants was significantly higher in N. caerulescens GA,⁵ and Fe accumulation tended to be higher (Fig. S4); we therefore hypothesized that more Ni is taken up under excess Ni in N. caerulescens GA because IRT1 is more highly expressed in N. caerulescens GA under Ni-induced Fe deficiency.⁵ The expression levels of natural resistance-associated macrophage protein (NRAMP) 1, NRAMP3, and IRT2, which are Fe transporter genes induced by Fe deficiency,^13–15 were significantly higher in N. caerulescens GA in the Ni treatment (Table 1), indicating that the Fe deficiency response was increased in N. caerulescens GA. However, contrary to expectations, the expression level of IRT1 was significantly higher in N. japonica than in N. caerulescens GA (Table 1, Fig. S5). Comparison of the amino acid sequences of IRT1 revealed that the negatively charged Asp residue at position 100, which is highly conserved among angiosperms, is substituted with an uncharged Asn residue in NjIRT1 (Figure 2, S6). Asp100 is located in the extracellular loop and is essential for Fe transport activities of AtIRT1;¹⁶ NjIRT1 may have lost its Fe transport activity, and possibly its Ni transport activity, through this mutation. Noccaea japonica might have evolved to use other Fe transporters (e.g., NRAMPs, YSLs) to prevent excess Ni accumulation via IRT1 in Ni-rich serpentine soils. Our results show that it is important to consider differences not only in abundance but also in the structure of transcripts for understanding mechanisms underlying intra- and inter-specific variation in metal hyperaccumulation.

Figure 2. Partial alignment of IRT1 sequences among angiosperms. Alignment of regions corresponding to AtIRT1 residues 95–105 are shown. Shading: black, identical; gray, similar. Arrow indicates residue essential for Fe transport activity.¹⁶ Nj, Noccaea japonica; Nc, Noccaea caerulescens ecotype GA; At, Arabidopsis thaliana; Nt, Nicotiana tabacum; Sl, Solanum lycopersicum; Os, Oryza sativa; Zm, Zea mays; Gm, Glycine max. Full alignment is shown in Fig. S6. Full-length sequences of NjIRT1 and NcIRT1 were validated by Sanger sequencing

Disclosure of conflicts of interest Statement

No potential conflicts of interest were disclosed.

Data deposition

The RNA-Seq data obtained in this study have been deposited in the DNA Data Bank of Japan (DDBJ) Sequence Read Archive under the accession number DRA011914.

Acknowledgments

We thank Matthew Stevens from ELSS group (https://www.elss.co.jp/en/) for editing the manuscript.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This work was supported by the KAKENHI Grants JP18KK0426 and JP19K05756. Some computations were performed on the NIG supercomputer at the ROIS National Institute of Genetics.