The mechanism: a hyperaccumulation program that is switched on, not induced
Noccaea caerulescens is unusual even among Brassicaceae metal specialists: most accessions hyperaccumulate zinc and cadmium, but some also hyperaccumulate nickel. Belloeil et al. produced a chromosome-scale genome of the Ni-hyperaccumulating Cira accession and compared it with the non-accumulating relative Microthlaspi perfoliatum and with calamine accessions of N. caerulescens that have lost the Ni trait.
The first surprise is how little the Ni-hyperaccumulating accession responds to nickel at the transcriptome level. When Firmiensis (Ncfi) was grown with 37.5 µM Ni, only 4 DEGs in roots and 7 in shoots met the |log₂FC| ≥ 1, padj ≤ 0.05 threshold. The non-accumulator M. perfoliatum responded far more strongly: 104 root and 98 shoot DEGs under the same conditions (Belloeil et al., 2026).
Instead of an inducible stress response, Ni hyperaccumulation correlates with constitutively high expression of just a few transporters. Compared with M. perfoliatum, Ni-hyperaccumulating N. caerulescens accessions showed:
- NcHMA4 expression 20–70× higher in roots and ~30× higher in shoots;
- NcHMA3 expression 25–200× higher in shoots;
- NcIREG2 expression ~30× higher in shoots.
These genes therefore form a small, pre-loaded transport module rather than a metal-activated defence network.
The root is the bottleneck
Reciprocal grafting between the hyperaccumulating Cira accession (Ncci) and the non-accumulating La Calamine accession (Nclc) showed that the root controls the phenotype. When Nclc shoots were grafted onto Ncci roots, leaves accumulated 16,930 ± 240 µg g⁻¹ Ni, matching self-grafted Ncci (16,950 ± 510 µg g⁻¹ Ni). In the reciprocal combination, Ncci shoots on Nclc roots reached only 640 ± 40 µg g⁻¹ Ni, close to self-grafted Nclc (580 ± 4 µg g⁻¹ Ni) (Belloeil et al., 2026).
Root Ni content followed the same pattern: Nclc roots held 330 ± 40 µg g⁻¹ Ni, whereas Ncci roots held 4,870 ± 500 µg g⁻¹ Ni. The shoot therefore cannot hyperaccumulate nickel if the root fails to take it up efficiently.
IRT1 mutations explain the loss of the trait
Resequencing identified the defective step in Nclc roots: a 10-bp deletion in the third exon of NcIRT1 that knocks out the iron-regulated transporter. Two other non-accumulating calamine accessions, Plombières (Ncpb) and Prayon (Ncpr), carry a large DNA insertion in the first exon of NcIRT1 instead. A separate frameshift mutation in NcZIP10 was also found in Nclc, but it is not shared with Ncpb/Ncpr, so it is unlikely to be the recurrent cause of the lost trait.
Yeast assays confirmed the functional difference. The Firmiensis allele NcfiIRT1 complemented the fet3fet4 iron-uptake mutant and increased nickel sensitivity and accumulation, whereas the La Calamine variant NclcIRT1 did neither. Expressing a functional NcfiIRT1–TagRFP fusion in Nclc roots raised shoot Ni from 480 ± 60 µg g⁻¹ to 1,060 ± 330 µg g⁻¹, an approximately 2-fold increase that was positively correlated with transgene expression (r² = 0.87, P < 0.01). Yet that restoration still fell well short of wild-type Ncci shoots (3,560 ± 600 µg g⁻¹ Ni) or Ncci hairy-root controls (4,330 ± 1,862 µg g⁻¹ Ni) (Belloeil et al., 2026).
NcIRT1 is therefore necessary for full Ni hyperaccumulation, but it is not sufficient on its own: the high-expression HMA and IREG2 background is also required.
What this means for agromining and breeding
For nickel agromining, the paper adds a practical genotype screen. Accessions intended for Ni phytoextraction or phytomining should carry an intact NcIRT1 allele and the constitutive transporter-expression module. The La Calamine, Prayon, and Plombières accessions show that N. caerulescens germplasm cannot be assumed to hyperaccumulate Ni simply because the species is famous for Zn/Cd hyperaccumulation; the same species can be a poor Ni accumulator when the root uptake step is broken.
The evolutionary picture is also important. Because Ni-hyperaccumulating accessions occur in multiple genetic units across Western Europe, the authors argue that the trait is likely ancestral in N. caerulescens and has been repeatedly lost in calamine populations by independent NcIRT1 lesions. That implies conservation value for ultramafic populations that still retain the functional alleles, not just because they are metallophytes, but because they preserve the complete transport module.
Caveats
The findings are based on controlled hydroponic and grafting experiments using a limited set of accessions. Whether the same NcIRT1 lesions explain non-accumulating populations outside the Belgian calamine belt remains to be tested. The transgenic complementation restored only part of the hyperaccumulation phenotype, so other root and shoot transporters — including the constitutively expressed NcHMA3, NcHMA4, and NcIREG2 — remain essential for full performance.
Provenance: primary source verified through the publisher DOI, the WUR open-access PDF, Crossref, OpenAlex, and an independent three-vendor fact-check; see /methodology/ for the hyperaccumulate.com editorial process.