The mechanism: a Golgi gate for Co/Ni in a crop
Where a nickel or cobalt ion ends up in a plant — locked in a root compartment or shipped to the shoot — is decided one membrane at a time. Wu & Nakanishi (2026), in Plant Molecular Biology (doi:10.1007/s11103-026-01701-1), identify OsFPN1 (Os06g0560000), a ferroportin-family metal transporter in rice (Oryza sativa cv. Nipponbare), as one such gate. In rice protoplasts the OsFPN1–sGFP fusion overlaps extensively with a Golgi marker and minimally with the plasma-membrane marker NAA60 — a predominantly Golgi address, though heterologous expression in onion epidermal cells shows a peripheral, plasma-membrane signal the authors concede may be a real minor pool.
The transport evidence is yeast-based but internally pointed. Expressing OsFPN1 in the Co/Ni-sensitive yeast mutant Δcot1 reduces growth on 1 mM Co or 1 mM Ni, while cells accumulate higher intracellular Co and Ni than vector controls (50 µM metal, 1 h, measured by ICP-OES, n = 3). In the Fe-uptake mutant Δfet3Δfet4, OsFPN1 improves growth on Fe-limited medium (50 µM BPDS; 5 µM BPDS in liquid culture), but it does not complement Mn- or Zn-deficiency mutants — so the substrate profile is Fe/Co/Ni, not generic divalent metal handling. Which way the transporter moves metal, and why more intracellular Co/Ni accompanies more sensitivity, the paper honestly leaves unresolved, framing itself as a re-examination of an unresolved paradox from Kan et al. (2022).
What the knockout does: metals leave the root
Two independent CRISPR alleles (osfpn1-1, +1 bp insertion; osfpn1-2, −21 bp deletion) make the whole-plant case. In hydroponics (100 µM Fe(III)-EDTA as the +Fe condition; 10 or 100 µM Co or Ni in a full factorial with ±Fe, 10-day assays, n = 3), the mutants under Co or Ni treatment consistently show less metal retained in roots and more delivered to shoots, alongside reduced shoot Fe. Under plain Fe deficiency, mutant shoots accumulate significantly less Fe while roots pile up Fe, Mn, Zn and Cu. The inferred job of OsFPN1: sequester excess Co/Ni into the endomembrane system of the root, keeping it out of the transpiration stream. All tissue-metal results are reported as directions — the paper gives no prose concentrations, and computes no translocation factors — but the pattern is coherent across both alleles and both metals.
The interactome points the same way. A split-ubiquitin yeast two-hybrid screen against a rice cDNA library retained 58 non-redundant candidate interactors; two were validated by point-to-point Y2H and bimolecular fluorescence complementation: OsPAR1 (Os05g0474400), a PRA1-family factor in Rab-dependent vacuolar trafficking, and Os12g0168900, a vacuolar V-ATPase subunit — both consistent with OsFPN1 plugging into endomembrane/vacuolar trafficking. Meanwhile the mutants over-activate the Fe-deficiency transcriptional program (OsIMA1, OsIMA2, and — in roots — OsNAS2, OsPAR1 and Os12g0168900), exactly the signature of a plant that cannot allocate Fe properly.
The iron interaction: the non-obvious risk
The result that should get attention is a triple interaction, not a single stress: under Fe deficiency plus 100 µM Ni, osfpn1 seedlings collapse after day 5 and are largely non-viable by day 10, while wild-type plants stay alive though stunted. Either stress alone is survivable; the combination, in a plant that cannot retain Ni in the root, is lethal. This matters because Fe deficiency is routine on the very paddy and upland soils where Ni (and Co) exposure is highest — laterite-derived, ultramafic-influenced or industrially contaminated soils are typically Fe-poor as well as Ni-rich. Rice already manages Fe and Ni through overlapping chemistry: what looks like Ni toxicity in the field may partly be Ni-worsened Fe starvation, and a genotype chosen for one trait could be silently vulnerable through the other. For food safety the mutant phenotype is the cautionary one — a checkpoint failure moves Co/Ni toward the grain-bearing shoot precisely when nutritional Fe stress is most likely.
The agromining mirror
For phytoremediation the same gene is a lever with a sign to choose. OsFPN1 is the retention arm of the Co/Ni economy: amplify it and you stabilize metals underground (phytostabilization, food-chain protection); disable it and you push Co/Ni into harvestable shoot biomass — the defining move of agromining. That is the architecture hyperaccumulators have already evolved: our earlier analysis of the tonoplast-aquaporin checkpoint in Arabidopsis showed the retention gate that a phytoextraction crop must evade or relocate to the shoot, and work on Odontarrhena chalcidica showed hyperaccumulator ferroportins doing the opposite job — loading Ni into the xylem in the root stele while sequestering it in the shoot. Wu & Nakanishi supply the crop-side counterpart of that map: in rice, the same family, pointed the other way. Editing a crop to hyperaccumulate is not the sane use; the realistic uses are subtler — diagnostic (Fe-deficient fields may fail at Ni retention first), and as a tolerance module to keep non-hyperaccumulating biomass crops alive on contaminated soils without letting metal reach the grain.
The honest limits
This is one cultivar (Nipponbare), hydroponics only, 10-day exposures, n = 3 throughout. Tissue-metal changes are directional, with absolute concentrations living in supplementary figures we could not audit. The headline lethality sits in supplementary figures too, is qualitative (“largely non-viable”), and has no survival-rate quantification. The transport direction of OsFPN1 — efflux, import into the Golgi, or trafficking between pools — is never established, and the yeast hypersensitivity-plus-accumulation paradox is carried over from Kan et al. (2022), not solved: this paper in fact contradicts Kan et al.'s conclusion that OsFPN1 is dispensable for Fe nutrition, so the field is unsettled. The two validated interactors are candidate links validated in tobacco cells, not rice, and the authors themselves call them preliminary.
Bottom line
OsFPN1 is a Golgi-level checkpoint that keeps Co and Ni out of rice shoots, and its function is conditional on iron supply — the first description of an Fe-dependent metal-retention checkpoint in the world’s most important crop species. For phytoremediation it sketches both a risk (Fe-deficient soils may silently defeat the root’s ability to keep Ni out of shoots) and an opportunity (the retention lever, moved or modulated, is a designed phytostabilization module). For agromining it is a reminder that hyperaccumulators did not just acquire loaders; crops evolved retainers, and the gap between those two states is precisely the engineering distance a phytoextraction crop would have to cross.
Provenance: based on the open-access full text (CC BY 4.0, Plant Molecular Biology 116:34, received 17 October 2025, accepted 19 February 2026, published 13 March 2026), read in full via Europe PMC (PMC12987785). Tissue-metal values are figure/supplement-only bars and are quoted here as directions, not numbers; no translocation factors exist in the paper. Every load-bearing claim was verified verbatim against the primary source by three independent reviewers (Claude/GPT/Kimi, each self-retrieving the full text); all three independently flagged the same limits (qualitative ICP-OES data, supplementary-only viability, unresolved transport direction, the Golgi-vs-plasma-membrane caveat, and the contradiction of Kan et al. 2022), which are disclosed above. See /methodology/.