Tissue-Specific Expression from a Compound TATA-Dependent and TATA-Independent Promoter

Overview

Garrity and Wold (California Institute of Technology, Division of Biology) report that the mouse metallothionein-I (MT-I) gene promoter functions in an unusual compound manner, directing both TATA-dependent and TATA-independent modes of transcription in vivo. The TATA-dependent message originates from the previously characterised +1 transcription start site and is the predominant species in most cell types and the only species that responds to cadmium induction. TATA-independent initiation occurs at sites distributed across the ~160 bp upstream of +1, is present in all tissues examined but is the predominant species only in testis (specifically in pachytene-stage meiotic cells and early spermatids), and is refractory to metal induction. Site-directed mutagenesis of the TATA element (substituting TCGAGA for TATAAA) drastically alters the ratio of the two RNA classes in cells where the +1 transcript normally dominates; in TATA-minus mutants the upstream TATA-independent RNAs become the most prevalent species but remain refractory to metal induction. The wiki uses this source as mechanistic background for Cadmium on the molecular basis of cadmium induction of metallothionein and on the testicular MT-I “blind spot” that contributes to germ-cell vulnerability to cadmium toxicity. It carries no food-matrix occurrence data and no regulatory content.

Why this matters

• It is the foundational characterisation that the mouse MT-I gene uses two mechanistically distinct promoter modes whose ratio is tissue-specific. Most prior models of metallothionein induction treated the MT-I promoter as a single TATA-dependent element driven by metal-responsive enhancer elements (MREs); this paper shows that picture is incomplete, and that the upstream TATA-independent RNAs — which predominate in germ cells of the testis — never respond to cadmium. The mechanistic asymmetry has direct toxicological consequences: testis is the canonical cadmium (Cd) target organ in mice (Parizek 1960; Nordberg 1971; Meek 1959), and Garrity & Wold pin this in part on the non-inducibility of MT-I in testis germ cells, rather than on metallothionein protein dysfunction or Cd sequestration failures alone. Downstream metallothionein reviews on the wiki (The Role of Metallothionein in Oxidative Stress, Metallothionein: An overview, Metallothionein: A Comprehensive Review of Its Classification, Structure, Biological Functions, and Applications) cite this lineage of work when they describe MT-I regulation as transcriptionally controlled and as having tissue-restricted induction behaviour.
• It establishes that the MT-I induction response to Cd is solely at the +1 (TATA-dependent) start site. RNA from the TATA-dependent +1 site increased ≥40% and commonly ≥25-fold in all cell types and tissues surveyed except testis after Cd administration. Upstream (TATA-independent) RNAs were unaffected (or slightly repressed) by Cd. This bears on any HMI claim that “metallothionein induction is the cellular defence against Cd” — the qualifier is that the induction is mechanistically restricted to the TATA-dependent transcription class, and there exist mammalian tissues (notably testis germ cells) where this class is not the dominant MT-I species.
• It documents the canonical experimental Cd dosing regimes used to study MT-I induction in mice: single subcutaneous 10 mg Cd/kg body weight in 200 µL saline 3 h before sacrifice (acute induction); daily subcutaneous 0.25 mg Cd/kg for 12 days (chronic pretreatment to test tolerance); single 1 mg Cd/kg challenge after pretreatment (protection assay). These dose anchors are used as method comparators when later studies (cited in The Role of Metallothionein in Oxidative Stress) measure metallothionein induction kinetics; the wiki may surface them where dose-response comparisons across mechanism papers are made.
• It corroborates and extends the Cd-tolerance phenomenon first reported by Parizek (1960) and Nordberg (1971): a low-dose Cd pretreatment regime (0.25 mg/kg/day × 12 days) protected DBA/2J mice from testicular damage at a subsequent 1 mg/kg challenge, even though MT-I and MT-II RNAs in testis remained non-inducible by metal. The authors conclude that “metallothionein protein cannot account for Cd resistance in testis” and that additional resistance mechanisms (e.g., Zn-MT-mediated tolerance attributable to pre-existing Zn-loaded metallothionein in other organs, or zinc-driven tolerance at the level of Cd transport) are involved. This nuance matters when the wiki frames metallothionein as the universal Cd defence: there are tissues (testis) where it is not the relevant defence.

Key concepts and structure

The paper is a nine-page primary research article (pages 5646-5654 of Volume 10, Number 11 of Molecular and Cellular Biology, November 1990). It is organised in the canonical IMRaD pattern: abstract, materials and methods, results (with three result subsections), discussion, and 56 numbered references. Figures 1-5 carry the experimental data: Figure 1 maps the MT-I gene structure and shows RNase protection / primer extension across tissues (testis, kidney, brain, spleen, liver, intestine, muscle, L cells), Figure 2 quantitates the upstream transcript start site groupings (groups A through F), Figure 3 shows RNase protection with full-length and 3’-end cDNA probes confirming splicing and 3’-end processing of the upstream RNAs, Figure 4 maps upstream RNA expression across testicular cell types (spermatogonia, leptotene/zygotene, pachytene, round spermatids, residual body, W/Wv germ-cell-deficient testes), and Figure 5 shows the consequences of TATA and Sp1 site-directed mutations in stably transfected L-cell pools.

The MT-I promoter map (Figure 2B)

The cis-acting regulatory landscape of the MT-I promoter spans approximately -200 to +1 and includes a canonical TATA element at -25 to -30, two Sp1 binding sites (a distal one at ~-145 and a proximal one at ~-50), two G-box elements, an MLTF binding site, a potential AP-1 site, and six metal-responsive enhancer elements (MREs) labelled A through F. In vivo footprinting in L cells detects DNA-protein interactions at most of these sites; only the MRE interactions are dependent on metal treatment.

The six start-site groups (Figure 2A)

The transcription start sites were operationally grouped into six clusters labelled A through F (panel B), where A = +3 to -1, B = -25 to -35, C = -39 to -75, D = -51 to -119, E = -82 to -146, F = -133 to -146 (per Fig. 2 legend, p. 5649). Group A is the canonical TATA-dependent +1 RNA. Groups B-F are the upstream TATA-independent RNAs. Quantitation in panel A shows, for each of testis, kidney, and L cells, the total MT-I RNA (left bars) and the start-site distribution (right bars). In L cells and kidney the +1 (Group A) transcript dominates; in testis the upstream groups dominate, with Group A constituting only ~15% of total MT-I RNA.

Tissue distribution (Figure 1B, 1C)

Figure 1B (RNase protection with 5’ genomic probe) and Figure 1C (primer extension) survey eight tissues from DBA/2J mice ± Cd treatment. In L cells and kidney, the +1 RNA dominates and Cd treatment shifts the ratio toward +1 (longer transcripts decrease 40-60% after metal treatment in L cells). In testis, the upstream (-44, -28, -134, -162) RNAs dominate and Cd has no significant effect on the +1 RNA level (testis does not respond to Cd with MT-I induction). Brain, spleen, liver, intestine, and muscle show intermediate patterns — upstream RNAs are present in all tissues but not as predominant species except in testis.

Cell-type localisation in testis (Figure 4)

To identify which testicular cell type carries the upstream MT-I RNAs, the authors examined fractionated germ-cell populations: spermatogonia (8 days post-partum), leptotene/zygotene (17 days), pachytene (>40 days), round spermatids (~40 days, ~60 days, >60 days), residual body, and somatic-only testes from W/Wv mice (which lack germ cells). The upstream MT-I RNAs are most prominent in late meiotic prophase (pachytene-stage cells) and remain prominent in round spermatids, declining in residual bodies and cytoplasmic fragments cast off by condensing spermatids. The +1 RNA gradually declines as germ cells progress from spermatogonia to round spermatids. W/Wv mice (germ-cell-deficient) contained substantially lower upstream MT-I RNA, demonstrating that the upstream transcripts are a germ-cell-restricted feature.

Site-directed mutagenesis of the TATA element (Figure 5)

The wild-type TATAAA at -25 to -30 was replaced with TCGAGA (a sequence previously shown to disable TATA element function). A separate construct mutated the distal Sp1 site (GGGGCGG → GCCCGGG). The mutated promoter sequences were stably transfected into L cells with the CAT gene fused downstream of +65 bp of MT-I. In the wild-type pMTCAT pool, the +1 RNA dominated and was metal-inducible; in the TATA-mutant pool, the +1 RNA was nearly eliminated and the upstream RNAs became the most prevalent transcripts but remained refractory to metal induction. The Sp1 mutation had no major effect on either the +1 or the upstream RNA production. Thus the TATA element is necessary for +1 RNA production and for metal induction of +1 RNA, but is not necessary for upstream RNA production, which is mechanistically TATA-independent.

Independent regulation of the two classes

The paper’s central conclusion is that the upstream RNAs are mechanistically TATA-independent (unaffected by TATA mutation) while the +1 RNA is TATA-dependent (eliminated by TATA mutation). Both classes are properly spliced and polyadenylated, both are RNA polymerase II products, and they differ only at their 5’ site of initiation. The choice between TATA-dependent and TATA-independent modes is regulated in a tissue-specific manner: the testis germ-cell pattern (upstream RNAs predominate, metal refractory) is intrinsic to MT-I promoter architecture and is not a property of any single cell-type-restricted accessory factor. Mutation of the TATA element alone in L cells does not switch the L-cell pattern to a germ-cell pattern, implying additional layers of regulation beyond TATA-dependence.

Key numbers

• Cd induction of +1 MT-I RNA (across cells/tissues surveyed except testis): ≥40% increase, commonly ≥25-fold.
• Group A (TATA-dependent +1) RNA share of total MT-I RNA in testis: ~15%.
• Effect of TATA mutation on +1 RNA in stably transfected L cells: +1 RNA virtually eliminated; upstream RNAs become the most prevalent species.
• Decrease of longer (upstream) MT-I transcripts after Cd treatment in L cells: 40-60%.
• Acute Cd dose (mouse subcutaneous, sacrifice 3 h post-injection): 200 µL of H₂O or 0.9% NaCl with or without CdSO₄ at 10 mg Cd/kg body weight.
• Chronic Cd pretreatment dose: 0.25 mg Cd/kg/day (subcutaneous CdSO₄ in saline) for 12 days; saline control n=5; Cd pretreated n=6.
• High-dose Cd challenge after pretreatment: single 1 mg Cd/kg subcutaneous injection 13 days post-treatment.
• Effect of high-dose Cd after saline pretreatment: testes “weighed approximately half as much” as saline-control testes (n=6 high-dose Cd arm; n=5 saline-only control) at 13 days.
• Effect of high-dose Cd after Cd pretreatment: “no discernible decrease in testicular weight” (n=4); body and kidney weights unaffected in either arm.
• MT-I transcription start site groupings (numbering relative to the +1 start site, per Fig. 2 legend, p. 5649): A = +3 to -1; B = -25 to -35; C = -39 to -75; D = -51 to -119; E = -82 to -146; F = -133 to -146.
• Approximate position of canonical MT-I TATA element: -25 to -30 (TATAAA → TCGAGA in the TATA-mutant construct pMCT).
• Approximate positions of two Sp1 sites: distal ~-145 (mutated in pMCSPA: GGGGCGG → GCCCGGG); proximal ~-50.

Methods (brief)

Animals: male DBA/2J and C57BL/6J mice (Jackson Laboratories, Bar Harbor, Maine) and BDF₁ mice (DBA × C57BL/6J offspring), 8 to 15 weeks old. Strain comparison: no strain-related variation in MT RNA expression noted. Additionally 15-week-old DBA/2J mice were used for the testicular toxicity protection experiments.

Acute Cd induction: mice injected subcutaneously with 200 µL of H₂O or 0.9% NaCl ± CdSO₄ (10 mg Cd/kg body weight), sacrificed 3 h later by cervical dislocation, tissues flash-frozen in liquid nitrogen and stored at -80°C until RNA isolation.

Chronic Cd pretreatment (with reference 42 protocol, CdSO₄ substituted for CdCl₂): daily subcutaneous injection of saline (n=5) vs CdSO₄ at 0.25 mg Cd/kg/day (n=6) for 12 days; bodies, testes, and kidneys weighed; no body or organ weight differences seen at this dose. Subsequent single high-dose challenge: 1 mg Cd/kg per mouse (n=6 pretreated; n=4 saline-pretreated control), tissues examined 13 days later for testicular weight.

Cell lines: L-cell derivative P cells (described in reference 29) maintained in dialysed calf serum. Eight hours prior to harvesting for RNA, fresh medium with or without 6 µM CdSO₄ was added.

Transfections: cells plated at 3.7 × 10⁵ per 10-cm dish; each plate received 10 µg of test plasmid (pMTCAT, pMCT, pMCSPA, or pBluescript KS⁻ control), 10 µg of L cell DNA, and 1 µg of pY3 (hygromycin resistance) by calcium phosphate coprecipitation. Selection in hygromycin (200 µg/mL) for 17 days. More than 90 separate colonies were obtained on each plate. Each test plasmid was used to transfect three plates; the pools of colonies were maintained and analysed separately.

RNA preparation: from L cells per reference 11 with a second organic extraction; from mice per reference 11; testis cell-type and mutant mouse RNAs were provided by Kelly Thomas and Mel Simon and by Debra Wolgemuth (samples per references 53 and 54). LiCl precipitation; A₂₆₀ concentration determination. DNase treatment did not alter results.

RNase protection assays: per references 29 and 56 with modifications including bis-acrylylcystamine acrylamide gels (Bio-Rad). LKB XL densitometer for quantitation. Type III preparation of small yeast RNAs (Sigma R-7125) used as carrier RNA.

Primer extensions: per reference 27, using avian myeloblastosis virus reverse transcriptase (Life Sciences). 5’-end-labelled oligonucleotide primers, signal linearity confirmed by analysing dilutions of a single RNA sample.

Constructs and probes: four MT-I probes used (5’-end genomic probe pSP6-MT-I, full-length cDNA probe pMT-IΔi, 3’-end cDNA probe pMT-IBH, internal cDNA probe pMT-INT). Plasmid pMTCAT (described in reference 30) is a pBluescript KS⁻ derivative containing the MT-I promoter EcoRI-BglII fragment fused to the chloramphenicol acetyltransferase (CAT) gene of pSV2cat and SV40 splicing and poly(A) addition signals. Oligonucleotide-directed mutagenesis created pMCSPA (Sp1 distal site mutant: GGGGCGG → GCCCGGG) and pMCT (TATA mutant: TATAAA → TCGAGA). pBS⁻ was a Stratagene KS⁻ derivative.

Funded by NIH, the Rita Allen Foundation, the Lucille P. Markey Charitable Trust (to B.W.W.), and a National Research Service Award (5 T32 GM 07616-11) from the National Institute of General Medical Sciences (to P.G.).

Implications

• Microbiome: Not relevant. The paper is on mouse mammalian tissues and cell lines; no gut microbiome or microbial population is studied.

Limitations

• The TATA mutation construct uses substitution of TCGAGA for TATAAA; the authors note that the double mutant TAGAGA had been shown previously to disable TATA function in studies using yeast TFIID complementation of a mammalian extract, and that the binding affinity of yeast TFIID for the double mutant adenovirus major late promoter is similar to its affinity for nonspecific DNA. The TCGAGA substitution is thus a strong negative control but is not formally proved to eliminate all TATA function in the in vivo mouse-cell context of this paper; some residual +1 RNA persists in the TATA mutant (Figure 5 lanes 3 and 4), albeit with 5’-end heterogeneity characteristic of TATA-element ablation.
• The “TATA-independent” RNA classification is operationally defined by persistence after TATA mutation. The authors acknowledge that this does not prove the upstream RNAs are produced by a wholly distinct promoter complex; the discussion considers two alternative models (closely related DNA-protein complexes with distinct components determining TATA-dependence vs TATA-independence; or mutually exclusive accessory factor complexes). The data presented cannot discriminate between them.
• The conclusion that “metallothionein protein cannot account for Cd resistance in testis” rests on the observation that MT-I and MT-II RNAs are not detectably induced in testis by either acute Cd or 12-day low-dose pretreatment. The paper does not directly measure metallothionein protein levels in testis (only RNA), and it does not rule out that constitutive (uninduced) baseline metallothionein protein could be load-bearing for tolerance. The conclusion as written is consistent with the data but is a conservative interpretation.
• Sample sizes are small: n=4 to n=6 per arm for the testicular toxicity protection experiments; transfection pools are three independent transformant pools per construct. No formal statistical comparison between conditions is reported beyond the consistency of the pattern across pools and tissues.
• Densitometric quantitation of autoradiographs (Figure 5 legend) cautions that autoradiographic exposure necessary to see upstream RNAs caused the +1 RNAs of the wild-type and Sp1-mutant lanes to be underrepresented. Quantitative comparison across lanes is therefore qualitative-magnitude only.
• The work uses radiolabelled RNase protection and primer extension as the readouts. The Northern blot results are mentioned as confirmatory (no gross structural differences in the body of the RNA between upstream and +1 species) but the data are described as “data not shown” rather than presented. The conclusions about splicing and 3’-end processing of upstream RNAs rest on the full-length cDNA probe protection pattern (Figure 3 lanes 1-7) which shows complementarity to the predicted 3’ protection product.
• No occurrence data; no consumer exposure data; no regulatory implication.

Related evidence

• Phytochelatins, the heavy-metal-binding peptides of plants, are synthesized from glutathione by a specific gamma-glutamylcysteine dipeptidyl transpeptidase (phytochelatin synthase)
• In Vitro Evolution and Preliminary Characterization of a Cadmium-Resistant Population of Chlamydomonas reinhardtii

Update history

The five most recent substantive edits to this page, classified major (evidence or structure moved), correction (a published value or statement was wrong and has been fixed), or minor (narrative rewritten without changing the underlying evidence). Each description is derived from what the edit did to this page; the linked commit is the authoritative record, routine regeneration passes are excluded, and the full version history lives in git.