Peng et al. 2023 — Al and F stresses altered organic acid and secondary metabolism in tea plants

Summary

This is a controlled hydroponic experiment on Camellia sinensis seedlings, not a survey of commercial tea or matcha products. The authors exposed two-year-old tea seedlings to graded aluminum (0.4 mM, 2.5 mM as Al₂(SO₄)₃) and fluoride (10 mM as NaF) treatments, alone and in combination, then tracked organic-acid secretion, tricarboxylic-acid-cycle and ascorbate-pathway gene expression, catechin and proanthocyanidin profiles, caffeine and theanine accumulation, and Al/F tissue concentrations over 48 h to 10 d. The mechanistic finding is that tea roots secrete oxalic, tartaric, citric, and shikimic acids under Al and F stress, that Al up-regulates and F down-regulates most TCA-cycle and amino-acid-biosynthesis genes, and that high F (10 mM) suppresses catechin, caffeine, and theanine accumulation in young leaves while increasing tissue F to roughly 5 mg/g fresh weight at day 10. For the Heavy Metal Index, the relevance is contextual: the paper documents that tea is an Al hyperaccumulator and characterizes the plant-physiology basis for that accumulation. It does not measure Al, F, or any other element in market tea, matcha, or any consumer-form product, and its tissue concentrations are produced by artificial stress treatments well above environmental exposures rather than by ordinary cultivation.

Key numbers

All values from the paper’s figures (Figure 1, Figure 7) and Introduction. The paper does not report tabulated numerical means for Al/F tissue accumulation; values below are taken from Figure 7 axis labels and bar heights at day 10 (n = 3 biological replicates, two-tailed Student’s t-test versus the day-1 baseline for each treatment group; * p < 0.05, ** p < 0.01).

Aluminum content in young tea leaves (apical bud + 1st leaf), Figure 7A, mg/g fresh weight, day 1 → day 10:

• 0.4 mM Al treatment: ~17 → ~20 mg/g FW (* p < 0.05)
• 2.5 mM Al treatment: ~17 → ~42 mg/g FW (** p < 0.01)
• 10 mM F alone (in 0.4 mM Al baseline): ~17 → ~12 mg/g FW (F antagonizes Al uptake)
• 2.5 mM Al + 10 mM F combination: ~19 → ~30 mg/g FW (** p < 0.01)

Fluoride content in young tea leaves, Figure 7B, mg/g fresh weight, day 1 → day 10:

• 0.4 mM Al treatment: ~0.5 → ~0.5 mg/g FW (no significant change; F was not supplied)
• 2.5 mM Al treatment: ~0.5 → ~0.5 mg/g FW (** p < 0.01 but values remain low; F was not supplied)
• 10 mM F alone (in 0.4 mM Al baseline): ~0.5 → ~5 mg/g FW (** p < 0.01)
• 2.5 mM Al + 10 mM F combination: ~0.5 → ~3.5 mg/g FW (** p < 0.01; Al partially antagonizes F uptake)

Caffeine content in young leaves, Figure 7D, mg/g fresh weight, day 1 → day 10:

• 0.4 mM Al: ~16 → ~17 mg/g FW (* p < 0.05, slight increase)
• 2.5 mM Al: ~16 → ~14 mg/g FW (no significant change)
• 10 mM F alone: ~17 → ~13 mg/g FW (** p < 0.01 reduction)
• 2.5 mM Al + 10 mM F: ~16 → ~13 mg/g FW (** p < 0.01 reduction)

Theanine content in young leaves, Figure 7E, mg/g fresh weight, day 1 → day 10:

• 0.4 mM Al: ~2.4 → ~3.0 mg/g FW (** p < 0.01 increase)
• 2.5 mM Al: ~2.5 → ~2.8 mg/g FW (* p < 0.05 increase)
• 10 mM F alone: ~2.5 → ~2.0 mg/g FW (** p < 0.01 reduction)
• 2.5 mM Al + 10 mM F: ~2.5 → ~2.0 mg/g FW (** p < 0.01 reduction)

Catechin contents in young leaves, Figure 7C, mg/g fresh weight: EGCG dominated all treatments (~33 mg/g FW), followed by EGC (~13–15 mg/g FW), with smaller amounts of GC, EC, C, and ECG. The 2.5 mM Al treatment slightly increased EGC and EC at day 10; the 10 mM F and 2.5 mM Al + 10 mM F treatments inhibited EC and EGCG accumulation in young leaves; most catechin contents did not change drastically across treatments. Exact bar values are not tabulated in the paper.

Introduction-stated background ranges (citing references [7]–[11], not measured in this study): tea plant leaves can accumulate up to 20,000 mg Al/kg and 1000–3000 mg F/kg in mature and old leaves, and more than 600 mg Al/kg and 100–300 mg F/kg in young leaves; tea plantation soils in China average 540 mg F/kg, ranging 190–1100 mg F/kg in most areas and exceeding 2000 mg F/kg in polluted and mining-area soils such as in southwest provinces.

Organic-acid secretion into hydroponic media (Figure 1, summary): under Al + F co-treatment at 12 h, oxalic acid in media reached the highest measured level (~310 µg/mL, ** p < 0.01 versus 0 h baseline of ~10 µg/mL). Tartaric acid in media reached ~0.55 µg/mL under Al + F at 12 h. Shikimic acid in media reached ~0.25 µg/mL at 48 h under Al + F. These values are root-exudate concentrations into the hydroponic medium, not tea-product concentrations.

Evidence Fitness

EF-4 Context only. This is a mechanistic hydroponic study on tea seedlings exposed to Al and F concentrations (0.4–2.5 mM Al, 10 mM F) that are well above field-cultivation soil-solution levels and that are imposed as instantaneous step changes rather than as long-term cultivation. The tissue Al concentrations reported (up to ~42 mg/g FW in young leaves at day 10 of 2.5 mM Al exposure) are not representative of values in commercial tea or matcha products, and the paper does not measure any consumer-form product. The source supports: (a) qualitative confirmation that tea is an Al and F hyperaccumulator, (b) mechanistic characterization of root organic-acid exudation and TCA-cycle gene response under Al/F stress, and (c) the directional finding that high F suppresses theanine and caffeine accumulation in young leaves. The source does NOT support: pooled occurrence percentiles for tea or matcha Al, market-product threshold-setting, hazard-quotient calculations against consumer intake, or any contamination-survey claim. For HMI synthesis, use only as supporting context for the upstream-of-product Al/F accumulation pathway; do not pool these tissue values into any product-category contamination_profile.

Methods (brief)

Plant material and culture: Two-year-old seedlings of Camellia sinensis (L.) O. Kuntze cv. Shuchazao and cv. Chuyeqi grown in hydroponic Shigeki Konishi (SK) nutrient solution in a greenhouse at 20–25 °C, light intensity 1300 µmol m⁻² s⁻¹, 18 h photoperiod / 6 h dark, until new tender roots emerged. Healthy seedlings were transferred to fresh hydroponic solutions containing 0, 0.4 mM Al, 2.5 mM Al (as Al₂(SO₄)₃), 10 mM F (as NaF, in 0.4 mM Al baseline), or 2.5 mM Al + 10 mM F. Hydroponic media sampled at 0, 12, 24, 48 h for organic-acid measurement; roots and young leaves (apical bud + 1st leaf) sampled at the same timepoints for RNA and metabolite analysis, and at day 1 vs day 10 for tissue Al, F, catechin, caffeine, and theanine accumulation. Five individual plants per biological replicate, three biological replicates per timepoint per treatment.

Organic acid quantification: HPLC on Agilent 1100 (Agilent Technologies, Santa Clara, CA, USA) with BEH-C18 analytical column (2.1 mm × 50 mm, 1.7 µm). Sample preparation: 5 mL media lyophilized, re-dissolved in 0.25 mL HCl (pH 1.0), filtered through 0.45 µm membrane. Mobile phase: 0.01 mol/L potassium dihydrogen phosphate. UV detector G4212-60008 diode array, 210 nm. Flow 1 mL/min, column temperature 30 °C, injection 20 µL. Reference standards: oxalic, tartaric, citric, malic, succinic, and shikimic acids (each 0.01 mol/L K₂HPO₄ solvent). Each measurement in triplicate; ANOVA used for analysis.

Transcriptome analysis: Plant Total RNA Extraction Kit (Tiangen Biotech, Beijing). mRNA assessed on NanoDrop 2000 spectrophotometer and RNA analyzer (Thermo Scientific, Wilmington, DE, USA). Libraries built with Illumina TruseqTm RNA Sample Prep Kit; sequencing on Illumina HiSeq2500 in triplicate. Reads mapped to the tea plant genome (~80 % mapping rate) using TopHat2 v2.1.0. Expression quantified as FPKM. NCBI BioProject PRJNA748249. Heatmap visualization with TBtools v1.108.

qRT-PCR validation: cDNA synthesized with SweScript RT I First Strand cDNA Synthesis Kit (Takara, Dalian, China). House-keeping genes CsACTIN (TEA019484) and CsGAPDH (TEA003029) as internal references. SYBR Green qPCR Premix (Universal) in 20 µL reactions (ddH₂O 8.2 µL, SYBR Premix 10 µL, forward and reverse primers 0.4 µL each, cDNA 1 µL). Each sample in triplicate; relative expression by 2^−ΔΔCt. SPSS v18 for analysis. Primer sequences in Supplemental Table S4.

Catechins, caffeine, and theanine: Amino acids (theanine, glutamine, glutamate, ornithine, etc.) extracted and quantified on L-8900 high-speed amino acid analyzer (Hitachi, Ibaraki, Japan) following She et al. 2023 (cited as ref [33]). Amino acid standards from Aladdin (Shanghai) and Wako Pure Chemical Industries (Osaka). Catechins (EC, EGC, EGCG, C, CG, GC) and caffeine quantified by HPLC following published protocols (refs [68], [69]); reference standards from Sigma-Aldrich (Steinheim, Germany).

Statistical analysis: Student’s two-tailed t-test for paired comparisons (treatment vs control or vs 0 h baseline) and ANOVA multiple range test at p ≤ 0.05 for multi-group comparisons. Confidence limits stated as 95 % or 99 %. Significance markers: * p < 0.05, ** p < 0.01.

Limitations stated by the authors and inferred from design: the Al and F concentrations used (0.4–2.5 mM Al = ~10.8–67.4 mg Al/L solution; 10 mM F = ~190 mg F/L solution) are imposed as step-change hydroponic exposures, not as steady-state field soil-solution concentrations. Tissue concentrations are recorded after 10 days of continuous exposure to these levels. The paper does not measure soil-solution chemistry in field tea-growing systems and does not report Al or F concentrations in harvested commercial tea or matcha. The two cultivars (Shuchazao, Chuyeqi) are research cultivars; no statement is made about how representative their Al/F uptake is of commercial tea cultivars used for matcha or other tea products.

Implications

For Heavy Metal Index synthesis: the paper contributes plant-physiology context for tea as an Al hyperaccumulator and for the antagonism between Al and F uptake at the root, but it does not contribute occurrence data usable for product-category pooled distributions. Its Al/F tissue values are produced by experimental hydroponic stress, not by field cultivation of commercial-grade tea, and the paper itself does not present them as representative of consumer-form tea or matcha. Sources that do measure Al in consumer-form tea infusions or matcha powder are the appropriate primary inputs for pooled Al occurrence in tea products; cross-link to ozturk2024-al-tea-infusion-teapot for measured Al in Turkish black, green, and white tea infusions across teapot materials and brewing times, and to szymczykowska2026-matcha-elemental-composition for the elemental composition of two organic Japanese matcha products (note: szymczykowska2026 explicitly did not measure Al).

For courses: the paper supports a teaching point that tea plants secrete organic acids (oxalic, tartaric, citric, shikimic) under Al and F stress as a detoxification mechanism, that Al up-regulates most TCA-cycle and catechin-pathway genes while high F represses them, and that high F suppresses theanine and caffeine accumulation in young leaves. These mechanistic points are useful background for explaining why tea Al content varies with growing conditions and why F co-exposure modifies tea quality. The teaching frame should be plant physiology, not consumer safety, since this study does not measure consumer exposure.

For app: no contamination_profile updates are warranted from this source. The Al values are tissue concentrations under artificial stress, not market-product concentrations. The app’s Al input for tea should come from market-product measurement sources.

Microbiome: not addressed in this paper. The paper measures root organic-acid exudation into hydroponic medium but does not characterize rhizosphere microbiota or any host-microbiome interactions, and so is not a federation candidate for WikiBiome.

Provenance notes

Open-access MDPI article published under CC BY 4.0 on 27 February 2023 (received 31 January 2023; revised 21 February 2023; accepted 23 February 2023). Raw PDF retrieved via the discover autopilot skill into raw/Manual Fetch Discovery/peng2023-matcha-heavy-metal.pdf. SHA-256 verified at ingest. Supplementary materials (Supplemental Figures S1–S6, Supplemental Tables S1–S4, including the qRT-PCR primer list, full pathway figures, and per-replicate organic-acid concentrations) are hosted at https://www.mdpi.com/article/10.3390/ijms24054640/s1 and were not retrieved into the corpus; tissue Al/F values were read off Figure 7 axis bars directly, and exact per-bar numerical values would require Supplemental Table S1 (organic-acid concentrations) and the figure-source numerical tables, which the paper does not embed in the body text. NCBI BioProject PRJNA748249 holds the transcriptome data.

Wiki pages updated on ingest

• tea
• aluminum
• tea

Verification notes

Fresh ingest 2026-06-08 (Claude Opus 4.7, autonomous /ingest-next-manual-fetch-pdf cycle from raw/Manual Fetch Discovery/).

• Filename–content mismatch documented. The source PDF was auto-fetched by the discover skill into raw/Manual Fetch Discovery/peng2023-matcha-heavy-metal.pdf with the suffix -matcha-heavy-metal, which appears to be a generic discover-skill template suffix rather than an accurate paper descriptor. The paper does not measure matcha. The paper is a hydroponic plant-physiology study on Camellia sinensis seedlings under controlled Al and F stress. The cite_key peng2023-al-fluoride-tea-stress describes the actual paper; the raw_handle MFD_peng2023-matcha-heavy-metal preserves the original auto-fetch filename for traceability per the existing MFD ingest pattern (cf. el-daouk2020-aluminum-food-lebanon and masri2025-toxicant-foods-california, which followed the same convention for similarly-templated MFD filenames).
• Routing scope kept broad. Frontmatter declares products: [tea] (umbrella) and ingredients: [tea] (umbrella) only. The paper studies tea plant tissue in general, not any specific tea product form (black, green, white, oolong, matcha, kombucha, supplement). Per CLAUDE.md Part 5b, broad scope is the correct routing-layer input for a paper of this kind; the routing layer is responsible for any downstream fan-out, not this source page.
• Fluoride not declared in metals:. Fluoride is not a heavy metal and is not in the HMTc 10-analyte slate. The paper measures fluoride extensively but per CLAUDE.md Part 14 metal-name vocabulary, only Al is declared. The fluoride findings are preserved verbatim in Key numbers, Methods, and Implications because they are integral to interpreting the Al data (Al–F antagonism is a central finding of the paper).
• Evidence Fitness set to EF-4 Context only. The Al/F tissue values are produced by 10-day hydroponic step-change exposures at 0.4–2.5 mM Al and 10 mM F, not by ordinary field cultivation. They are not representative of consumer-form tea or matcha. The Evidence Fitness section explicitly states what this source can and cannot support, per the EF-X schema. This page should not be pooled into any tea or matcha contamination_profile.
• No brand naming. Methods sections name instrument vendors (Agilent 1100, NanoDrop 2000, Hitachi L-8900, Illumina HiSeq2500, Tiangen Biotech, Takara, Sigma-Aldrich, Aladdin, Wako, TBtools v1.108, SPSS v18) per CLAUDE.md Part 12 Exception 2 (scientific-method vendor/material allowance). No consumer-product brand names appear in the source.
• No new pages proposed. The source’s slugs (tea ingredient, tea product, aluminum metal, China jurisdiction, tea-leaves matrix) all exist in the current taxonomy. No ingredient or product scaffold-creation triggers fired during this ingest.

Phase 2 audit application 2026-06-08 (fresh-context Agent subagent, verdict PROMOTE):

• All five audit checks returned ✅ clean (numerical fidelity, slug vocabulary, speciation and methods, Part 12 brand firewall, Part 2 wiki/HMTc firewall).
• False positive rejected: alleged “SwScript” vs “SweScript” typo in Methods. The auditor’s only ⚠️ finding was that the Methods section “says SwScript RT (Takara) in one place but SweScript RT I First Strand cDNA Synthesis Kit (Takara, Dalian, China) elsewhere.” Verified by grep against the committed page: the page uses “SweScript” once and contains no instance of “SwScript”. The auditor appears to have misread the single SweScript occurrence as two distinct strings. No correction applied; no false-positive correction necessary because the typo does not exist in the page.
• No content edits applied. Audit-application commit records the verdict and the false-positive disposition only.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.