Schilsky et al. 1989 — Hepatocellular copper toxicity in HepG2 and its attenuation by zinc-induced metallothionein

Schilsky and colleagues (Albert Einstein College of Medicine, Marion Bessin Liver Research Center) report a mechanistic study in the HepG2 human hepatoblastoma cell line of acute copper toxicity and zinc-mediated protection. Confluent HepG2 cells exposed to copper chloride for 48 h showed a dose-dependent loss of viability with LD50 ≈ 750 µM CuCl₂. Preincubation with 200 µM zinc acetate for 2 h raised the LD50 to 1,250 µM CuCl₂ and protected against the copper-induced generalised reduction in protein synthesis seen at 1,000 µM Cu for 1 h (>80% reduction without zinc; preserved at 80±5% of control with zinc). The zinc protection persisted for ~30 h after the 2 h preincubation. The protective effect was not explained by inhibition of cellular copper uptake (essentially unchanged across 0/200/400 µM coincubated zinc at 30 and 60 min) but correlated tightly with zinc-induced synthesis of a 10-kD cysteine-rich protein identified as metallothionein by Western blot using sheep anti-rat metallothionein antibody (cross-reactive with human MT) and by metallothionein mRNA induction on Northern blot. HPLC gel-filtration of 100,000 g supernatants showed increased ⁶⁷Cu and [³⁵S]cysteine in the metallothionein-containing 10-30 kD fractions of zinc-pretreated cells. Cellular ATP was virtually unchanged at 1,000 µM Cu for 1 h, and ultrastructural injury (loss of intercellular adhesion, ER disruption, lysosomal dense bodies, mitochondrial vacuolisation) appeared only at 3-12 h, well after the protein-synthesis collapse, supporting a translational rather than energetic primary mechanism. The wiki uses this source as mechanistic background for copper (acute hepatocellular copper toxicity, the translational-machinery target, lysosomal sequestration of copper-metallothionein) and for zinc (the canonical zinc-induces-metallothionein hepatoprotection axis cited in Wilson’s disease therapeutic literature). It carries no food-matrix occurrence data and no regulatory content.

Why this matters

• It is among the foundational HepG2 demonstrations that copper’s earliest acute toxic effect is on the translational machinery — total RNA synthesis is increased 39±5% at 1,000 µM Cu for 1 h while protein synthesis is suppressed by more than 80% — placing copper’s primary acute lesion at translation rather than at transcription, mitochondrial bioenergetics (ATP unchanged), or ultrastructural disruption (no abnormalities at 1 h, only at 3-12 h). This anchors any wiki claim on the hierarchy of copper hepatotoxicity mechanisms.
• It provides direct experimental evidence that zinc’s protective effect against copper toxicity, in hepatocyte-like cells, is not mediated by reduced copper uptake (Table I: cell-associated copper at 30 min was 390, 410, 370 pg/mg protein at 0, 200, 400 µM coincubated zinc; at 60 min 550, 590, 430 pg/mg protein). This is consequential because the long-standing zinc-for-Wilson’s-disease rationale (Hoogenraad 1987; Hill 1987; Brewer) was built on intestinal copper-absorption blockade via enterocyte metallothionein induction (Cousins 1985). Schilsky et al. demonstrate a second mechanism: direct hepatocyte metallothionein induction sequestering copper intracellularly into the 10-30 kD low-molecular-mass fraction co-localising with metallothionein. Wiki framing of zinc as a copper-handling therapeutic should retain both axes.
• It establishes the temporal correlation between zinc protection and metallothionein biosynthesis (peak [³⁵S]cysteine incorporation into the 10-kD band at 2 h, returning to basal by 48 h) rather than metallothionein steady-state level (which rises through 48 h on Western blot). The protection window — ~30 h post-preincubation — tracks the biosynthesis peak. This nuance matters when downstream wiki claims invoke “metallothionein levels” as the protection metric: the data point to active de novo metallothionein synthesis (presumably to bind incoming copper into the nascent protein) as the load-bearing variable, not to total protein pool. The authors explicitly note this and reference earlier work from the same laboratory showing reduced copper incorporation into the low-molecular-mass protein fraction of HepG2 after cycloheximide (Stockert 1986).
• It documents the second copper- and cysteine-containing low-molecular-mass HPLC peak induced by zinc, identified by thin-layer chromatography as glutathione, but shows that the steady-state glutathione level was essentially unchanged (1.9±0.1 vs 2.13±0.04 µg/mg protein, zinc-treated vs control), excluding cellular glutathione as the principal mediator of the protective effect despite documented evidence that glutathione participates in heavy metal detoxification (Meister 1988; Freedman 1989).
• It corroborates the lysosomal sequestration pathway for copper-metallothionein complexes in hepatocytes by ultrastructural identification of electron-dense cytoplasmic bodies (visible at 3 h of 1,000 µM CuCl₂ exposure) ultrastructurally similar to the metallothionein-containing lysosomes documented in Bedlington terriers with inherited copper toxicosis (Johnson 1981; Lerch 1985) and in patients with advanced Wilson’s disease (Sasa 1986; Nartey 1987) and primary biliary cirrhosis (Hanaichi 1984; Janssens 1984). This places the HepG2 in vitro mechanism within the broader pathology lineage referenced by Wilson’s-disease and copper-toxicosis reviews on the wiki.

Key concepts and structure

The paper is a seven-page primary research article (pages 1562-1568 of Volume 84, J. Clin. Invest., November 1989), submitted 11 January 1988 and revised 31 May 1989. It is organised in the canonical IMRaD pattern: abstract, methods (with growth/survival, protein synthesis, glutathione, cellular ATP, RNA synthesis, copper uptake/metabolism, cell ultrastructure subsections), results (mirroring the methods subsections), and discussion. Figures 1-12 carry the experimental data: Figure 1 (HepG2 survival and growth across 0-1,000 µM Cu in exponential phase), Figure 2 (confluent-cell viability at 48 h across 500-1,500 µM Cu ± 2 h 200 µM Zn-acetate preincubation), Figure 3 (duration of zinc-mediated protection over 0-40 h post-preincubation), Figure 4 (rate of [³⁵S]methionine incorporation over 60 min at 1,000 µM Cu), Figure 5 (SDS-PAGE autoradiograph of [³⁵S]methionine-labelled proteins across five treatment conditions), Figure 6 (autoradiograph of [³⁵S]cysteine-labelled 10-kD band and metallothionein mRNA Northern blot), Figure 7 (Western blot time course of metallothionein steady-state level vs [³⁵S]cysteine biosynthesis), Figure 8 (HPLC gel-filtration profiles of [³⁵S]cysteine and copper in 100,000 g supernatants), Figure 9 (immunoprecipitated albumin and albumin mRNA Northern blot), Figures 10-12 (electron micrographs at 1 h, 3 h of 1,000 µM Cu and 12 h of 500 µM Cu). Table I tabulates total cell-associated copper across 0/200/400 µM coincubated zinc at 30 and 60 min.

HepG2 dose-response to copper (Figures 1, 2)

In the exponential growth phase (~1 × 10⁵ cells per dish seeded), copper concentrations up to 83 µM allowed near-control growth over days 2-5; 117 µM produced moderate growth inhibition; 330, 500, and 1,000 µM produced progressive growth arrest and cell death (Figure 1). Confluent monolayers (1 × 10⁶ cells per dish at day 5) exposed for 48 h to graded copper showed sigmoidal viability loss with LD50 ≈ 750 µM CuCl₂ (Figure 2). Pre-incubation in 200 µM zinc acetate for 2 h before the 48 h copper exposure shifted the dose-response curve right, raising the LD50 to ≈ 1,250 µM and enhancing survival at every copper concentration tested. Zinc alone (200 µM, 2 h) did not affect viability.

Duration of zinc protection (Figure 3)

To define how long the protective effect of a 2 h, 200 µM zinc-acetate preincubation persists, cells were returned to zinc-free medium for 0-40 h before challenge with 1,000 µM CuCl₂ for 48 h. Protection was maximal at 0-10 h post-preincubation (~80-95% of control viability), declined between 10 and 30 h, and was lost by ~30-40 h post-preincubation. The unpretreated 1,000 µM Cu condition produced only 25% survival at 48 h (cross-referenced to Figure 2).

Translational suppression as the early copper lesion (Figures 4, 5, 9)

[³⁵S]methionine incorporation into TCA-precipitable cell protein during 1,000 µM CuCl₂ exposure showed an initial rapid decline within the first 10 min, a brief plateau, then a continued progressive decline to ~20% of control by 60 min (Figure 4). SDS-PAGE of [³⁵S]methionine-labelled lysates at 1 h showed a generalised reduction across all bands in the 1,000 µM Cu lane (Figure 5e) — comparable in magnitude to the cycloheximide-treated lane (Figure 5d) — that was prevented by 2 h pre-treatment with 200 µM zinc acetate (Figure 5c, preserved banding at 80±5% of control TCA-incorporation). Zinc alone did not alter the banding pattern (Figure 5b). Immunoprecipitation of albumin from cells treated for 1 h with 1,000 µM CuCl₂ showed marked reduction of newly synthesised [³⁵S]methionine-albumin (Figure 9A) in parallel with the generalised reduction. Albumin mRNA on Northern blot was unaltered (Figure 9B), indicating that the lesion is at translation, not at albumin mRNA abundance.

Metallothionein induction by zinc (Figures 6, 7)

Pre-incubation with 200 µM zinc acetate for 2 h before [³⁵S]cysteine labelling increased synthesis of a 10-kD protein detected on autoradiograph (Figure 6A) and identified as metallothionein by Western blot using a sheep anti-rat metallothionein antibody (kindly provided by Dr I. Bremner, Aberdeen, Scotland) that the authors had separately verified to cross-react with human metallothionein (Figure 7A). Metallothionein-IIA mRNA on Northern blot (using probe from Dr Dean Hamer, NCI) rose substantially in zinc-treated cells (Figure 6B). The time course (Figure 7) showed [³⁵S]cysteine incorporation peaking at 2 h post-zinc preincubation and returning to basal by 48 h, while Western-blot steady-state metallothionein continued to accumulate over 48 h. The biosynthesis-vs-steady-state divergence is the central observation supporting the authors’ model that active de novo metallothionein synthesis (sequestering incoming copper into the nascent protein) drives protection, rather than total metallothionein pool size.

Copper distribution in the cytosol (Figure 8)

HPLC gel-filtration of 100,000 g supernatants on a Zorbax GF-250 column (with 50 µM Cu in the labelling medium to support quantitation) showed two zinc-induced peaks of [³⁵S]cysteine radioactivity (Figure 8a): fractions 22-23 (corresponding to purified metallothionein retention time) and fractions 26-28 (subsequently identified as glutathione by TLC). Copper-content profiles of the same supernatants (Figures 8b, 8c) showed increased copper in the 10-30 kD region (metallothionein-containing fractions) of zinc-pretreated cells relative to control. Copper was measured by graphite furnace atomic absorption spectroscopy using copper-free glassware and polypropylene tubes (Stockert 1986).

Copper uptake into whole cells is essentially unaltered by zinc (Table I)

To distinguish “zinc reduces copper uptake” from “zinc redistributes intracellular copper into metallothionein,” the authors coincubated confluent HepG2 with ⁶⁷Cu (carrier-free, Los Alamos National Laboratory) and 0, 200, or 400 µM zinc acetate for 30 or 60 min and measured total cell-associated radioactive copper. Total cell-associated copper at 30 min was 390 (0 µM Zn), 410 (200 µM Zn), and 370 (400 µM Zn) pg/mg protein; at 60 min, 550, 590, and 430 pg/mg protein respectively. Pre-incubation with 0-500 µM zinc acetate did not alter total cell-associated copper either (data not shown). The conclusion: zinc’s protection is mediated by intracellular handling, not by uptake blockade.

Ultrastructure (Figures 10-12)

At 1 h of 1,000 µM CuCl₂ exposure, only scattered cytoplasmic vacuoles were visible; intercellular adhesion and organelle morphology were essentially normal (Figure 10, ×3,400). At 3 h, cells became pleomorphic, intercellular adhesions were lost with widening of intercellular spaces, the endoplasmic reticulum was disrupted, cytoplasmic vacuoles became prominent, and dense bodies with the appearance of lysosomal granules appeared in the cytoplasm; mitochondria appeared unaltered (Figure 11, ×8,500). At 12 h of 500 µM CuCl₂ exposure, profound endoplasmic reticulum disruption, mitochondrial matrix condensation and vacuolisation, nuclear changes, and marked cellular pleomorphism were seen, with many cells in various stages of degeneration (Figure 12, ×6,800). The temporal sequence — translational collapse at 1 h, intercellular and ER disruption with lysosomal-dense-body appearance at 3 h, mitochondrial involvement only at 12 h — supports the authors’ conclusion that translation is the earliest hepatocellular target of copper toxicity.

Key numbers

• LD50 confluent HepG2 at 48 h, copper chloride alone: 750 µM CuCl₂.
• LD50 confluent HepG2 at 48 h, after 2 h preincubation with 200 µM zinc acetate: 1,250 µM CuCl₂.
• Duration of zinc protection after a 2 h, 200 µM zinc-acetate preincubation: ~30 h (Figure 3); 25% survival at 48 h for unpretreated 1,000 µM Cu condition (cross-reference Figure 2).
• Protein synthesis at 1 h of 1,000 µM CuCl₂ exposure (no zinc): reduced by >80% (TCA-precipitable [³⁵S]methionine; Figure 4).
• Protein synthesis at 1 h of 1,000 µM CuCl₂ after 2 h, 200 µM zinc-acetate preincubation: 80±5% of control levels.
• Total RNA synthesis at 1 h of 1,000 µM CuCl₂ (no zinc), by [³H]deoxyuridine incorporation: +39±5% vs control.
• Total RNA synthesis at 1 h of 1,000 µM CuCl₂ after 2 h, 200 µM zinc-acetate preincubation: 86±8% of control.
• Cellular ATP at 1 h of 1,000 µM CuCl₂ vs control: 32±1.2 vs 36.0±2.0 nmol/mg protein.
• Cellular reduced glutathione, zinc-treated vs control: 1.9±0.1 vs 2.13±0.04 µg/mg protein (essentially unchanged).
• Total cell-associated copper (⁶⁷Cu coincubation), 30 min, pg/mg protein: 390 (0 µM Zn), 410 (200 µM Zn), 370 (400 µM Zn).
• Total cell-associated copper (⁶⁷Cu coincubation), 60 min, pg/mg protein: 550 (0 µM Zn), 590 (200 µM Zn), 430 (400 µM Zn).
• Nonspecific cell-associated copper at 4°C control: <1% of copper bound at 30 min.
• Background copper in MEM+ growth medium: 1.4 µM.
• Initial seeding density for exponential-phase experiment: ~1 × 10⁵ cells per 35 mm dish.
• Confluent density at day 5: ~1 × 10⁶ HepG2 cells per dish (~1.5 × 10⁶ for Figure 2 confluent experiments).
• Copper concentrations tested in growth-phase experiment (Figure 1): 0, 83, 117, 330, 500, 1,000 µM CuCl₂.
• Copper concentration range in confluent-cell viability experiment (Figure 2): 500-1,500 µM CuCl₂ in 250 µM steps.
• Metallothionein protein band molecular mass (SDS-PAGE): ~10 kD.
• Metallothionein-containing HPLC gel-filtration fraction: fractions 22-23 (Figure 8a).
• Glutathione-containing HPLC gel-filtration fraction: fractions 26-28 (Figure 8a; identified by Rf match to reduced glutathione standard on silica-gel TLC in n-butanol-acetic acid-pyridine-water 15:3:10:12).
• Cycloheximide condition (positive control for translational arrest): 2 µg/mL for 1 h.
• Pulse-labelling conditions: [³⁵S]methionine 100 µCi/mL for 10 min (or 200 µCi/mL for 30 min in dialysed FBS for albumin immunoprecipitation); [³⁵S]cysteine 50 µCi/mL for 1 or 2 h in cysteine-free RPMI 1640 or MEM with dialysed FBS.

Methods (brief)

Cell culture: HepG2 human hepatoblastoma (originally from Drs Barbara P. Knowles and David P. Aden, Wistar Institute) in 35 mm plasticware dishes (Falcon Labware, Oxnard, CA) at 37°C under 5% CO₂/95% air in minimal essential medium (MEM) with 10% fetal bovine serum (FBS) — termed MEM+ — containing 1.4 µM background copper. Cultures fed on day 2 after seeding and studied at confluence on day 5 (~1 × 10⁶ cells per dish, or ~1.5 × 10⁶ per dish for the Figure 2 confluent-cell viability experiment). Viability scored by trypan blue exclusion (Boyse et al. 1964).

Copper toxicity: copper chloride added to MEM+ at 0-1,500 µM. Confluent-cell exposures of 48 h. Exponential-phase cells seeded at ~1 × 10⁵ per dish in MEM+ ± copper, scored on days 1-5. Zinc protection: 2 h preincubation in MEM+ with 200 µM zinc acetate before copper exposure; or, for duration experiment, 2 h preincubation then return to MEM+ for 0-40 h before 48 h, 1,000 µM CuCl₂ exposure.

Protein synthesis: cells washed with PBS pH 7.4, exposed to MEM+ with 1,000 µM CuCl₂ for 0-60 min, with some preparations preincubated in MEM+ with 200 µM zinc acetate (2 h) or 2 µg/mL cycloheximide (1 h). Pulse-labelled in methionine-free MEM with 100 µCi/mL [³⁵S]methionine for 10 min, washed in ice-cold MEM and PBS, harvested by scraping, suspended in PBS. Aliquots used for acid-insoluble [³⁵S]methionine TCA precipitation onto glass fibre filters; for total protein by Lowry et al. (1951); and for 10% SDS-PAGE on cell lysates in 1% SDS, 0.1 M DTT, 10% glycerol, 0.05 M Tris-Cl pH 6.7, phenol red. Gels fixed in 10% acetic acid/40% methanol, soaked in 3 vol Enhance (New England Nuclear), dried, and exposed to SB-5 x-ray film (Eastman Kodak) at −70°C.

Albumin immunoprecipitation: cells treated for 1 h with 1,000 µM CuCl₂ and labelled in methionine-free MEM with 10% dialysed FBS and 200 µCi/mL [³⁵S]methionine for 30 min, lysed in 0.15 M NaCl, 1% NP40, 2 µM PMSF, 1 mM EDTA, 10 mM Tris, 1 mg/mL BSA, pH 7.4. Rabbit anti-human albumin antibody (Cooper Biomedical, Inc., Malvern, PA) 10 µL, overnight 4°C, then 50 µL of 20% protein A-Sepharose for 2 h at 4°C. Pellets washed 4× in 0.15 M NaCl/50 mM Tris/1% Triton/0.1% SDS/0.5% deoxycholate, once in 25 mM Tris/0.15 M NaCl pH 7.4, resolved by 10% SDS-PAGE.

[³⁵S]cysteine labelling: cysteine-free RPMI 1640 (Gibco Laboratories, Grand Island, NY) with 10% dialysed FBS and 50 µCi/mL [³⁵S]cysteine for 1 h. Cell lysates extracted with 1 vol acetone, centrifuged at 10,000 g; supernatant treated with 3 vol acetone and recentrifuged (Johnson et al. 1981). Pellet dissolved in PBS with 10 mM DTT under nitrogen 4 h room temperature, treated with 20 mM iodoacetamide at 37°C for 30 min, then SDS-PAGE sample buffer with 0.1 M DTT. Resolved on 10-20% gradient SDS-PAGE.

Metallothionein detection: Western blot (Towbin et al. 1979) with sheep anti-rat metallothionein antibody (kindly provided by Dr I. Bremner, Aberdeen, Scotland), separately verified by the authors to cross-react with human metallothionein. Transfer to nitrocellulose in 48 mM Tris, 39 mM glycine, 0.04% SDS, 20% methanol (Novablot apparatus, LKB Instruments). Blocking in 2% non-fat dry milk. Detection with ¹²⁵I-protein A, exposure to XAR-5 film (Eastman Kodak) at −70°C.

HPLC gel filtration: confluent HepG2 incubated in MEM+ ± 200 µM zinc acetate for 2 h, labelled with 50 µCi/mL [³⁵S]cysteine for 2 h in cysteine-free MEM with 10% dialysed FBS and 50 µM CuCl₂. Homogenisation in 10 mM Tris pH 7.7; supernatants from 100,000 g for 1 h fractionated on a Zorbax GF-250 column (Dupont Co., Wilmington, DE) at 1 mL/min in 10 mM Tris/0.15 M NaCl pH 7.79. Fractions collected at 0.5 min intervals. Radioactivity measured on an LKB Instruments Rackbeta counter. Copper measured by graphite furnace atomic absorption spectroscopy (Stockert 1986) using copper-free glassware and polypropylene plastic tubes, with all solutions made in double-distilled deionised water.

Glutathione: Sekune and Ando (1972) maleimide method, using BACM (10-(maleimido)-2,3,6,7-tetrahydro-11-oxo-1H,5H,11H-(1)-benzopyrano[6,7,8-l]quinolizine; Polysciences, Inc., Warrington, PA) yielding a fluorescent thiol derivative. Fluorescence spectrophotometer (Perkin-Elmer Corp., Norwalk, CT), excitation 400 nm, emission 500 nm, slit 10 nm. Reduced-glutathione standard. Levels measured in TCA-precipitate supernatants of cell homogenates after MEM+ alone or MEM+ + 200 µM zinc for 2 h. Thin-layer chromatography on silica gel in n-butanol/acetic acid/pyridine/water (15:3:10:12) used to confirm glutathione identity in HPLC fractions 26-28 (Rf match vs reduced-glutathione standard; data not shown).

Cellular ATP: modification of Lamprecht and Trautschold method (Samuelson et al. 1988). HepG2 confluent in 25 cm² culture flasks, treated with MEM+ or MEM+ + 1,000 µM CuCl₂ for 1 h, washed in PBS, harvested into ice-cold 3% perchloric acid. Supernatant ATP measured in triplicate from duplicate experiments; protein by Lowry et al. (1951).

RNA synthesis: confluent HepG2 incubated in MEM+ with 100 µCi/mL [³H]deoxyuridine ± 1,000 µM CuCl₂ or 200 µM zinc acetate for 1 h. TCA precipitation. Total RNA extraction by guanidine thiocyanate/caesium chloride gradient (Chirgwin et al. 1979 modification; Zern et al. 1985). Albumin and metallothionein mRNA by Northern blot (10 µg total RNA per lane), with HSAF-47 human albumin cDNA (Dr R. Lawn, Genentech) (Lawn et al. 1981) or with MT-IIA cDNA (Dr Dean Hamer, NCI) (Schmidt et al. 1984), labelled with [³²P]dCTP by primer-extension kit (International Biotechnologies) to 2-5 × 10⁸ cpm/µg DNA. Filters washed stringently, exposed to XAR-5 film at −70°C.

Copper uptake: confluent cells in 96-well plates (Flow Laboratories, McLean, VA) preincubated in MEM+ ± 0-500 µM zinc acetate, washed, then incubated in MEM with carrier-free ⁶⁷Cu (Los Alamos National Laboratory; Stockert et al. 1986) or with the same zinc-acetate concentrations in MEM. At 30 or 60 min, medium aspirated, cells washed with 50 nM sodium phosphate, 0.15 M NaCl, pH 7.2, with 2 mM EDTA at 4°C. Lysed in 1.0 mL of 1 N NaOH for radioactivity and total protein. Nonspecific cell-associated copper measured by parallel 4°C incubation; <1% of copper bound at 30 min.

Electron microscopy: HepG2 confluent in MEM+ exposed to 1,000 µM CuCl₂ for 1 and 3 h or to 500 µM CuCl₂ for 12 h. Pellets fixed in chilled 1% osmium tetroxide / barbital buffer pH 7.4 for 1 h, embedded in Epon, sections stained with uranyl acetate and lead citrate, examined under an Elmiskop I electron microscope (Siemens-Allis Inc., Cherry Hill, NJ).

Funding: NIH grants DK-34668, DK-17702, DK-32972; National Center for the Study of Wilson’s Disease, Inc.; National Research Service Award DK-08151 (to M.L.S.); American Liver Foundation Postdoctoral Research Fellowships (to M.L.S. and M.J.C.); Irma T. Hirschl Career Scientist and Sinsheimer Foundation Awards (to M.A.Z.).

Implications

• Certification: The paper provides no occurrence data for any food matrix, no exposure data for any human population, and no regulatory threshold-setting information. It does not move any HMTc threshold or category. Its contribution to HMTc is indirect — it is part of the mechanistic case (cited downstream by metallothionein reviews and Wilson’s-disease therapeutic literature on the wiki) that hepatocyte metallothionein induction by zinc is a viable second axis (beyond the established intestinal absorption blockade) for handling internalised copper, and that copper’s earliest acute hepatocyte lesion is translational rather than energetic. This nuance is preserved when the wiki frames the toxicity-mitigation literature for copper.
• App: No routing to ingredient or product pages. The biological samples are a cultured human hepatoblastoma cell line; the work is mechanistic biochemistry/cell biology, not food or supplement testing. Empty ingredients, products, matrices, jurisdictions.
• Courses: Useful as a primary-literature anchor for the “early translational arrest, late ultrastructural injury” temporal sequence of acute copper hepatotoxicity, for the biosynthesis-vs-steady-state distinction in metallothionein-mediated protection, and for the demonstration that zinc protection in hepatocytes operates intracellularly (sequestration) rather than at uptake. Pair with garrity1990-mt1-tissue-specific-promoter (mouse MT-I transcriptional regulation) for the regulatory layer, and with later metallothionein reviews on the wiki for the broader synthesis. The paper is the canonical HepG2 demonstration cited in zinc-for-Wilson’s-disease therapeutic reviews.
• Microbiome: Not relevant. The paper is on a human hepatoblastoma cell line; no gut microbiome or microbial population is studied.

Limitations

• HepG2 is a hepatoblastoma-derived cell line, not a primary hepatocyte. While it maintains many differentiated hepatocyte functions (Knowles et al. 1980, cited in the paper), its metallothionein induction profile and copper-handling stoichiometry may not perfectly recapitulate primary adult human hepatocytes; the authors acknowledge HepG2 as a model that “maintains a number of differentiated physiologic functions” without claiming complete fidelity.
• Sample sizes per condition are not always specified. Figure 2 viability data are means±SD of duplicates; Figure 3 are means±SD of triplicates; Figure 4 protein-synthesis time course is means±SD of four experiments; cellular ATP is triplicate samples from duplicate experiments; Table I copper-uptake is means of at least four determinations with SD <10%. The paper does not present formal statistical comparisons (no p-values, no confidence intervals) — internal consistency and the magnitude of the effects (LD50 shift from 750 to 1,250 µM CuCl₂; >80% vs 80±5% protein-synthesis differential) carry the inference.
• The conclusion that zinc protection is mediated by metallothionein induction rests on temporal correlation (peak biosynthesis at 2 h post-preincubation; protection window ~30 h) and on co-localisation of zinc-induced [³⁵S]cysteine signal and copper in the 10-30 kD HPLC fractions. The paper does not perform a metallothionein-knockout or metallothionein-inhibitor experiment to formally demonstrate necessity. The cycloheximide control (Figure 5d) demonstrates that translational competence is required for the [³⁵S]methionine signal but does not establish metallothionein necessity per se.
• The glutathione observation is incomplete: an apparent zinc-induced [³⁵S]cysteine peak in the HPLC fraction co-eluting with reduced-glutathione standard suggested increased glutathione synthesis, yet the steady-state pool size was unchanged. The authors conclude glutathione is unlikely to account for the protection but do not measure cellular glutathione turnover (synthesis vs degradation) to formally support that conclusion.
• The ⁶⁷Cu uptake experiment used coincubation of zinc and ⁶⁷Cu rather than the protective protocol of 2 h zinc preincubation followed by copper exposure. The paper notes separately that “preincubation with media containing 0-500 µM zinc acetate did not alter total cell-associated copper (data not shown),” extending the conclusion to the protective protocol, but does not show the raw data for the preincubation arm.
• Copper concentrations used (500-1,500 µM in the viability experiments; 1,000 µM in the mechanism experiments) are pharmacological doses substantially above physiological liver copper concentrations. The authors do not present a low-dose copper viability curve; the protection by zinc may not generalise to chronic low-dose copper exposure scenarios. The discussion notes that the work concerns “acute copper toxicity.”
• No occurrence data; no consumer exposure data; no regulatory implication.

Wiki pages this source may touch

• copper
• zinc

Verification notes

Existing-page check. DOI grep (none, doi field null), raw_handle grep (MFK_34-hepatocellular-copper-toxicity-and-its-attenuat), and cite-key glob (schilsky*, *hepatocellular*, *copper-tox*) over wiki/sources/ on 2026-06-08 returned no matches. This is a NEW source page — no prior version to merge-enhance.

DOI provenance. The PDF first page shows the standard ASCI citation format J. Clin. Invest. © The American Society for Clinical Investigation, Inc. 0021-9738/89/11/1562/07 $2.00 Volume 84, November 1989, 1562-1568. The Journal of Clinical Investigation backfilled DOIs to legacy content under the pattern 10.1172/JCI<articleId>, but the article-specific JCI ID was not independently verified during ingest (no offline lookup table available; the cite-key search and the sibling JCI 1998 source (schett1998-hsp70-hsf1-rheumatoid-synovial) confirm this is the project pattern for unverified JCI DOIs). Leaving doi: null with no_doi_assigned: true per the 2026-05-13-doi-fallback-followup.md convention; a Google Scholar search URL is supplied as access_url until a future FIX-MISSING-DOIS pass can backfill the canonical 10.1172/JCI<articleId> resolver and the JCI archive URL https://www.jci.org/articles/view/<articleId>.

Evidence tier. B. This is a primary peer-reviewed laboratory study in the HepG2 human hepatoblastoma cell line. Sample sizes are small per condition (duplicates to triplicates to “at least four determinations”); no formal statistical comparisons are reported (no p-values; no confidence intervals); effect sizes are large and self-consistent across the figure panels. Tier B is appropriate per docs/conventions — A-tier is reserved for well-powered primary studies with formal statistics or authoritative agency monographs; this paper does not clear that bar. Its historical importance (HepG2-based mechanistic anchor cited in the zinc-for-Wilson’s-disease therapeutic literature and in metallothionein-biology reviews) does not elevate its evidence weight for HMI’s contamination-and-exposure-focused taxonomy.

Metals frontmatter. Cu, Zn. The paper uses CuCl₂ (and ⁶⁷Cu for uptake) as the toxicant and zinc acetate as the protective inducer; both are central to the paper’s experimental design and conclusions. No other metals are dosed or measured. The metallothionein-IIA isoform identity is established through the cDNA probe (MT-IIA, Dr Dean Hamer, NCI); the authors do not measure metallothionein-I-specific induction. No methylated, organic, or oxidation-state-specific Cu or Zn species are reported; the canonical HMI analyte symbols Cu and Zn apply.

Ingredients, products, matrices, jurisdictions frontmatter. All empty. The biological samples are HepG2 cells (a human hepatoblastoma line). No food matrix, supplement matrix, or personal-care matrix is sampled. The work is conducted at Albert Einstein College of Medicine (Bronx, NY, USA) but no regulatory or jurisdictional frame applies; the work is basic biochemistry/cell biology. jurisdictions: remains empty per the garrity1990-mt1-tissue-specific-promoter sibling-page pattern.

Sample size. null. The paper does not state a single sample-size count; per-experiment n values appear as duplicates (Figure 2), triplicates (Figures 1 and 3), four experiments (Figure 4), “at least four determinations” (Table I), and triplicate samples from duplicate experiments (cellular ATP). The sample_population field summarises the experimental units.

Brand firewall (Part 12). No commercial brand names appear in any contamination-value or product-evaluation context. Vendor mentions are limited to scientific-method context: Falcon Labware (Oxnard, CA) for plastic dishes; Gibco Laboratories (Grand Island, NY) for cysteine-free RPMI 1640; Flow Laboratories (McLean, VA) for 96-well plates; Cooper Biomedical, Inc. (Malvern, PA) for rabbit anti-human albumin antibody; Polysciences, Inc. (Warrington, PA) for BACM fluorescent maleimide reagent; Perkin-Elmer Corp. (Norwalk, CT) for the fluorescence spectrophotometer; Dupont Co. (Wilmington, DE) for the Zorbax GF-250 HPLC column; LKB Instruments (Gaithersburg, MD) for Novablot transfer and Rackbeta counter; Brinkmann Instruments Co. (Westbury, NY) for the Eppendorf centrifuge; New England Nuclear (Boston, MA) for Enhance fluorography enhancer; Eastman Kodak Co. (Rochester, NY) for SB-5 and XAR-5 x-ray film; Siemens-Allis Inc. (Cherry Hill, NJ) for the Elmiskop I electron microscope; International Biotechnologies, Inc. (New Haven, CT) for the primer-extension kit; Los Alamos National Laboratory for carrier-free ⁶⁷Cu; Wistar Institute (origin of HepG2 cells); Genentech, Inc. (South San Francisco, CA) — affiliation of Dr R. Lawn, source of human albumin cDNA probe HSAF-47; National Cancer Institute (Bethesda, MD) — affiliation of Dr Dean Hamer, source of MT-IIA cDNA probe; Jackson Laboratories not relevant here (HepG2 is not a mouse line). All vendor references are standard scientific-method vendor mentions allowed under the verification-checklist Exception 2 for analytical methodology context. No food, supplement, or personal-care brand is named. No firewall action required.

HMTc firewall (Part 2). The paper contains no HMTc-threshold language, no claims about HMI certification levels, no consumer-audience risk advisories, and no policy-relevant content. The discussion is entirely mechanistic biochemistry and cell biology, with references to Wilson’s-disease therapeutic context (Hill 1987; Hoogenraad 1987) framed as scientific rationale for the experimental design rather than as policy recommendation. No firewall action required.

Speciation note. The paper uses “Cu” and “Zn” generically; experimental doses are administered as CuCl₂ and zinc acetate; the uptake tracer is ⁶⁷Cu. No methylated, organic, or oxidation-state-specific species are reported. The HMI canonical analyte symbols for copper and zinc are Cu and Zn (no isotope/species distinction needed; see CLAUDE.md Part 14). Metallothionein isoform is MT-IIA per the cDNA probe used (Schmidt, Hamer & McBride 1984).

Date arithmetic. Received 11 January 1988; revised 31 May 1989; published November 1989 (Vol. 84, No. 5, pp. 1562-1568). Consistent with year: 1989 frontmatter.

Raw-handle stem. The MFK_34 handle stem MFK_34-hepatocellular-copper-toxicity-and-its-attenuat is taken from the Kimi-generated PDF filename 34_Hepatocellular_Copper_Toxicity_and_Its_Attenuation_by_ZincM.pdf (the trailing M in ZincM appears to be a Kimi indexing artefact — likely the start of “Mitigation” truncated mid-word, or a stray character in the filename build). The handle stem truncates at the 56-character convention boundary; no correction to the raw_handle is performed to preserve the audit trail back to the source folder.

Scope fit. The paper sits in the “Black Market Peptide Metal Survey / heavy_metals_peptides” Manual Fetch Kimi folder alongside garrity1990-mt1-tissue-specific-promoter, nagel1989-cadmium-resistant-chlamydomonas, schett1998-hsp70-hsf1-rheumatoid-synovial, and other peptide/metal mitigation papers. Per the 2026-06-02 scope commit 3f47f95 — scope: mitigation/remediation is in-scope, not a skip, peptide and metallothionein mechanism papers are in-scope as background for the mitigation-evidence chapter of the wiki. This 1989 HepG2 primary-experimental paper is foundational for the hepatocellular axis of zinc-induced metallothionein protection against copper toxicity that is referenced in Wilson’s-disease therapeutic literature and in copper-page mechanistic background.

Audit subagent (2026-06-08) verdict: PROMOTE → minor concern applied. Five checks returned ✅ on all five (Check 1 numerical fidelity, Check 2 slug vocabulary, Check 3 speciation/methods, Check 4 Part 12 brand firewall, Check 5 Part 2 HMTc firewall). One ⚠️ concern was raised in Check 3 (the Western-blot XAR-5 exposure temperature in Methods (brief) was written as “70°C” when the PDF p. 1563 has “−70°C” — film exposure for autoradiography is always at sub-freezing temperatures, +70°C would destroy the film, so this was a missing-minus-sign typo). Verified against PDF p. 1563 (“exposure to XAR-5 film (Eastman Kodak Co.) at −70°C”) and against the parallel mentions of “−70°C” for SB-5 film exposure on p. 1562 and the [³⁵S]cysteine sample storage on p. 1564, all consistent with the standard autoradiography practice. Applied the correction: “70°C” → “−70°C” in the Western-blot Methods entry. Subagent ID a6ec741a344df5657.

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.