Diethylenetriaminepentaacetic Acid Enhances Thyroid Hormone Action by a Transcriptional Mechanism

© Copyright 2004 by Humana Press Inc. All rights of any nature, whatsoever, reserved. 0163-4984/04/99(1–3)–0219 $25.00

Diethylenetriaminepentaacetic Acid Enhances Thyroid Hormone Action by a Transcriptional Mechanism MARIA P. SCIAUDONE,1 LILI YAO,1 MATTHEW SCHALLER,1 STEVEN A. ZINN,1 AND HEDLEY C. FREAKE*,2 Departments of 1Nutritional Sciences; 2Animal Science, University of Connecticut, Storrs, CT 06269-4017 Received September 25, 2003; Revised November 4, 2003; Accepted November 11, 2003

ABSTRACT Zinc is thought to be required as a structural component of the thyroid hormone (triiodothyronine, T3) receptor (TR). However, we have previously demonstrated that use of diethylenetriaminepentaacetic acid (DTPA) to restrict zinc availability to cultured cells actually potentiates rather than inhibits thyroid hormone action. In this article, the mechanisms underlying these effects of DTPA have been investigated. Treatment of GH3 rat pituitary tumor cells with DTPA in the presence of T3 resulted in twofold greater concentrations of growth hormone (GH) mRNA. Addition of actinomycin D to inhibit transcription showed that GH mRNA was actually less stable in the presence of DTPA, eliminating mRNA stabilization as a possible mechanism underlying this effect. Cycloheximide was able to block the induction by DTPA, showing a requirement for protein synthesis. Transient transfection of a GH promoter/luciferase reporter construct into GH3 cells revealed an inhibitory effect of DTPA on luciferase activity. However, when cells were stably transfected with the same construct, a T3dependent stimulation of luciferase activity by DTPA was observed, mimicking the effects seen with the endogenous mRNA. Thus, the GH promoter does mediate the effects of DTPA, but stable integration into chromosomal material is required. Index Entries: Zinc; thyroid hormone; diethylenetriaminepentaacetic acid; gene expression; growth hormone; transfection; GH3 cells.

*Author to whom all correspondence and reprint requests should be addressed. Biological Trace Element Research

219

Vol. 99, 2004

220

Sciaudone et al.

INTRODUCTION Zinc participates in a large and diverse array of biological processes. Its size and charge characteristics, coordination flexibility, and inability to undergo oxidation or reduction favor its incorporation into biological macromolecules, most notably enzymes and transcription factors (1). Within the nucleus, zinc is used by a wide array of transcription factors, usually to stabilize structures that permit binding to DNA (2). Included among these transcription factors are the members of the steroid/thyroid receptor superfamily that mediate the effects of a large number of hormone, nutrient, and other ligands on gene transcription (3). These receptors all contain nine invariant cysteine residues in their DNA-binding domain, eight of which coordinate two atoms of zinc and are essential for receptor function (4,5). The description of zinc biochemistry is becoming more complete and much is known of its physiology, including its requirement for growth and proper function of numerous organ systems (1). However, links between the biochemistry and physiology are tenuous at this time. For example, it is not clear which particular biochemical events lead to the growth failure of zinc deficiency (6). It has been suggested that transcription factors might bind zinc with lower affinity than enzymes and, under conditions of deficiency, might be more subject to zinc loss and, therefore, interrupted function (7). Consequently, we developed the hypothesis that some of the effects of zinc deficiency might be mediated by loss of zinc from nuclear receptors and, therefore, failure of these endocrine/nutrient signaling pathways. We tested this hypothesis using the thyroid hormone receptor (TR). Thyroid hormone, specifically its active metabolite triiodothyronine (T3), affects a diverse array of biological processes (8). Like zinc, its deficiency leads to failures in growth and development. GH3 rat pituitary tumor cells have been widely used to investigate T3 action (9). T3 stimulates growth hormone (GH) gene transcription in these cells and the details of this process, including TR interaction with the GH promoter, have been well described (10,11). Therefore, we tested the effects of limiting zinc availability to these cells by the addition of a chelator (diethylenetriaminepentaacetic acid, DTPA) to the media (12,13). Surprisingly, DTPA enhanced, rather than inhibited, induction of GH expression by T3. This effect was not seen in the absence of T3 and could be blocked by equimolar addition of zinc, although not other divalent cations. DTPA effects on GH mRNA were slow (24 h) and could be reversed by subsequent addition of zinc. Thus, it appeared likely that they were caused by an effect of zinc loss on the production or activity of a protein involved in the T3 signaling pathway. However, this protein did not appear to be the TR itself, because the T3 binding of this protein was not affected by either DTPA or zinc (12,13). Our objective in the work reported here was to probe the mechanisms underlying these surprising effects of zinc chelation on thyroid hormone Biological Trace Element Research

Vol. 99, 2004

DTPA Effects on Thyroid Hormone

221

action. We demonstrate that DTPA appears to act at the transcriptional level, although stable rather than transient transfections were required to see this effect.

MATERIALS AND METHODS Cell Culture GH3 rat pituitary tumor cells, obtained from the American Type Culture Collection (ATCC), were grown and maintained in monolayer culture with Ham’s F-10 medium containing penicillin (50,000 IU/ L) and streptomycin (50 mg/L) supplemented with horse serum (15%) and newborn calf serum (2.5%) as described previously (12).

Transcriptional Inhibition Studies GH3 cells were grown to confluence in 25-cm2 flasks. For the 48 h prior to treatment, standard medium was replaced with Ham’s F-10 containing antibiotics and 10% newborn calf serum, depleted of endogenous hormones by an ion-exchange resin (T3-free media) (14). Medium was replaced with fresh T3-free media and the flasks treated with 10 nM T3 with or without 50 µM DTPA for a second 48-h period. Media were then supplemented with the transcriptional inhibitor actinomycin D (1.4 µM) and RNA extracted at different times over the subsequent 16 h.

Translational Inhibition Studies GH3 cells were grown to confluence and pretreated with T3-free media for 48 h. Medium was replaced with fresh T3-free medium and the cells treated with 10 nM T3 with or without 50 µM DTPA. After 24 h, cycloheximide (25 µM) was added to one-half the flasks to inhibit protein synthesis, and incubation continued for a further 24 h, at which time cells were processed for RNA extraction.

RNA Extraction and Northern Analysis Total RNA was isolated using a modified Peppel/Baglioni protocol as reported previously (12). Equivalent amounts of RNA were size separated on 1% agarose–6% formaldehyde denaturing gels and transferred onto nitrocellulose membranes (MSI Nitropure, Westborough, MA). After ultraviolet (UV) crosslinking and baking, the membranes were hybridized separately with 32P-labeled GH and ribosomal protein L32 (RPL32) cDNA probes, synthesized using a random primer labeling kit (Life Technologies, Rockville, MD). Hybridization was carried out overnight at 42°C as previously described (12). After posthybridization Biological Trace Element Research

Vol. 99, 2004

222

Sciaudone et al.

washes, the membranes were autoradiographed at –80°C with an intensifying screen, and the resulting images were quantified using Molecular Analyst software for the GS-670 densitometer (Bio-Rad, Hercules, CA). The mRNA levels are expressed as the ratio between densitometric signals measured for GH and RPL32.

Transfection Assays The rat GH promoter (–530 basepairs; gift of H. Samuels, New York University, NY) (15) was subcloned into the pGL3-Basic luciferase reporter plasmid (Promega, Madison, WI). The pCMV-βGal plasmid (Stratagene, La Jolla, CA) was used to control for transfection efficiency. For transient transfection, GH3 cells were grown to confluence in 75-cm2 flasks (approx 17.5 × 106 cells). Cells were detached, pelleted, and resuspended in phosphate-buffered saline (PBS). Plasmid DNA (10 µg of each) was added and the cells incubated on ice for 10 min. Cells were transferred to a sterile cuvet, placed in the Gene Pulser II electroporator (Bio-Rad), and pulsed at 250 V and 975 µF for 10–16 ms. Cells were then resuspended in Ham’s F-10 medium containing 10% newborn calf serum without T3, aliquoted to a six-well plate, and incubated for 48 h in the presence of different combinations of T3 (10 nM), DTPA (50 µM), and/or zinc sulfate (40 µM). Cells were harvested for luciferase activity and β-galactosidase activity using the appropriate kits (Promega). Luciferase activity was determined using a TD-20/20 luminometer (Turner Designs, Sunnyvale, CA) and β-galactosidase measured in an Ultrospec Plus spectrophotometer (Pharmacia LKB Biotechnology, Piscataway, NJ). For stable transfections, the pGL3/GH plasmid was transfected into GH3 cells by electroporation as earlier, but replacing the pCMV-βGal with the pC1neo vector (2 µg; Promega). Cells were returned to culture flasks and treated with G418 (500 µg/mL; Promega). Parallel treatment of nontransfected cells demonstrated that this concentration was sufficient to kill all cells not expressing the neomycin transferase gene. Pooled cultures of stably transfected cells were developed and maintained with the addition of 200 µg/mL G418 to the growth medium. Cells were treated with combinations of T3, DTPA, and zinc sulfate and luciferase activity measured as for the transiently transfected cells. Protein concentration was measured in the cell lysates (16) and used to standardize luciferase activity.

Statistical Analysis Each treatment was replicated two to three times within an experiment and each experiment was repeated two to four times. Data were analyzed using one-way analysis of variance (ANOVA) followed by Scheffe’s post hoc test to determine group differences. The significance level was set at p