Cardamine hupingshanensis, the Brassicaceae from the Se-rich belt of Enshi, Hubei, stores selenium the way hyperaccumulators do: not in protein, but as free selenoamino acids — chiefly methylselenocysteine (MeSeCys) and selenolanthionine (SeLan). That biochemical choice is the plant’s whole trick, because free SeLan and MeSeCys stay out of proteins and out of trouble. Duan, Du, Li, Zhang, Zhang and Yuan (Xi’an Jiaotong-Liverpool University and collaborators) now show, in an open-access paper in Food Chemistry: X 39:104410, that the standard laboratory method for measuring those compounds has been partially destroying them — and that a hyperaccumulator’s speciation profile can be almost an artefact of the extraction protocol.
The number that should unsettle the literature
The authors ran the conventional one-step protease extraction on hydroponically grown C. hupingshanensis (80 mg Se L⁻¹ as selenite, 14 days; leaves sampled) and recovered only 23.26 ± 3.63% of total Se as identified species — with SeLan at just 5.06 ± 0.28 µg Se g⁻¹. Yet a simple water leach of the same tissue released 72.06 ± 0.32% of total Se. Splitting the protocol into a water step (free selenoamino acids) followed by enzymatic digestion of the pellet (protein-bound Se) moved SeLan to 143.74 ± 2.70 µg g⁻¹ — a 28.41-fold increase, with SeLan found almost exclusively in the aqueous fraction. Total species recovery rose to 80.65 ± 8.51% (vs. 20.82% at the low end of the one-step range). For comparison, the wheat controls behaved differently in kind: grain from soil enriched with 40 mg Se kg⁻¹ as selenate (31 days, grown at Yangzhou University) held Se mainly as protein-tractable selenomethionine, and the winning protocol was a one-step enzymatic hydrolysis under anaerobic conditions, lifting SeMet from 98.73 ± 4.50 to 112.17 ± 7.87 µg g⁻¹ and total recovery from 34.11 ± 11.83% to 90.02 ± 5.93%.
Why the molecules disappear
Two mechanisms, one honest gap. First, the oxidation problem: under aerobic extraction, SeMet oxidises to selenomethionine oxide (SeOMet), which the authors show co-migrates with other species and distorts the chromatogram; sealing the digestion under anaerobic conditions suppresses that. Second, the buffer problem: in Tris-HCl buffer (pH 4.5), measured SeLan collapsed from 157.30 ± 8.14 µg g⁻¹ in plain water to 42.72 ± 3.74 µg g⁻¹, while wheat SeMet fell from 32.66 ± 3.23 at pH 4.5 to 4.09 ± 0.47 µg g⁻¹ at pH 8.5. The authors are candid that why enzymolysis and Tris-HCl destabilise these free Se-amino acids “remains unclear” and “requires further investigation” — pH, ionic strength and enzyme-substrate interactions during hydrolysis are suspects, not convicted mechanisms. They also flag that strict anaerobiosis is operationally hard to sustain (residual O₂ exposure “cannot be completely excluded”) and suggest reducing agents such as β-mercaptoethanol or dithiothreitol as practical alternatives.
The non-obvious implication: the method is the message for phytomonitoring
The analytical point generalises into a phytoremediation and agro-quality one. Speciation — not total Se — is what determines whether a Se-rich biomass is a bioavailable supplement, a feedstock for Se recovery, or a toxicity liability, and this paper demonstrates that speciation numbers are protocol-dependent at the level of a 28-fold swing in a named detoxification product. The authors cite prior work (Both et al., 2018) reporting ~60% of the plant’s Se as water-extractable, of which nearly 40% occurs as SeLan; a lab using a one-step enzymatic method on the same tissue would measure almost no SeLan at all and could conclude the pathway differs. The second implication is access: the orthogonal LC-AFS platform (anion-exchange SAX for selenite/selenate, cation- exchange SCX for the five organic species) matches ICP-MS-grade speciation at a fraction of the instrument cost, which matters for routine monitoring of Se-biofortified crops or of phytoremediation outputs in labs that will never own an ICP-MS. As a bonus, the SCX method picks up γ-glutamyl-MeSeCys in C. hupingshanensis — reported here for the first time, and one more non-protein Se sink to account for in mass balances.
The risks, stated plainly
These are hydroponics-and-grain bench results, not field data: the Cardamine plants saw a single 14-day selenite pulse at 80 mg L⁻¹, and the wheat was grown by a partner lab and supplied as grain. Two plant matrices were tested; the authors themselves frame the protocols as matrix-specific, so the protocols should be re-validated before being applied to other Se accumulators such as Stanleya or Astragalus. The SeOMet reference was qualitative only — no certified standard exists — so the oxidation correction is directional, not absolute. And the headline caveat stands: until the chemistry of SeLan loss in enzymatic digests is understood, the fix is empirical — water first, keep O₂ out, avoid Tris — rather than mechanistic.
Source: Duan, Du, Li, Zhang, Zhang & Yuan (2026), Food Chemistry: X 39:104410, DOI 10.1016/j.fochx.2026.104410 (CC BY-NC; published 2 September 2026). Full text read via Europe PMC (PMC13572759); every load-bearing number was verified verbatim by three independent reviewers retrieving the source through separate paths, with no disagreements. See /methodology/ for how we source and check analysis pieces.