Transforming growth factor-\g=b\1production in porcine thyroid follicular cells: regulation by intrathyroidal organic iodine A.

J. Cowin, J.

R. E. Davis and S. P.

Bidey

Department of Medicine, Stopford Building, University of Manchester, Oxford Road, Manchester Ml3 9PT, U.K. (Requests for offprints should be addressed to S. P. Bidey) received

1

May

1992

ABSTRACT

The present studies have demonstrated the production of transforming growth factor-\g=b\1(TGF-\g=b\1)by porcine thyroid follicular cells (TFCs) maintained in vitro as subconfluent monolayers, and have confirmed a stimulatory effect of iodide on thyroidal TGF-\g=b\1mRNA and peptide release. RNA extracted from TFCs maintained in the absence of iodide contained a 2\m=.\5kb transcript which hybridized specifically with a cDNA probe for human TGF-\g=b\1,and which showed an approximate doubling in intensity in cells exposed to 10 \g=m\mol NaI/l. In the presence of the anti-thyroid thionamide drug methimazole (MMI; 1 mmol/l), the action of iodide on TGF-\g=b\1 mRNA was attenuated, although MMI alone had no effect on the control level of TGF-\g=b\1 mRNA. The TGF-\g=b\1peptide content of TFC-conditioned media (TFC-CM) was assessed using the fetal mink lung cell line Mv1 Lu, in which activated TGF-\g=b\1specifically suppresses trichloroacetic acid-precipitable

INTRODUCTION

The maintenance of normal

thyroid function is dependent upon an adequate supply of inorganic iodide, which is metabolized through specific thyrotrophin (TSH)-dependent pathways leading to thyroglobulin iodination and thyroid hormone bio¬ synthesis. In addition, however, further intracellular

metabolites of iodide appear to be involved in the regulation of functional cell responsiveness to TSH (Ingbar, 1972; Pisarev, 1985). The iodide-dependent autoregulation of TSH-dependent iodide uptake (Wolff & Chaikoff, 1944) is known to reflect actions of organic iodine intermediates at the level of TSHdependent adenylate cyclase (Van Sande et al. 1975, Rapoport et al. 1976; Filetti & Rapoport, 1984; Van

[methyl- 3H]thymidine incorporation. Newly conditioned TFC-CM stimulated [methyl-3H]thymidine incorporation into Mv1Lu cells, but after heat treatment to inactivate growth stimulators and activate the latent TGF-\g=b\1 component this medium inhibited [methyl- 3H]thymidine incorporation. This inhibitory effect was prevented by immunoadsorption of TFC-CM with a TGF-\g=b\1-neutralizing antiserum, confirming the specificity of the inhibitory response. The inhibitory activity of TFC-CM was increased when the TFCs were preincubated with 10 \g=m\mol NaI/l, and lost when TFCs were exposed to MMI. In conclusion, TFCs produce TGF-\g=b\1 mRNA and TGF-\g=b\1peptide, which are both increased by iodide treatment in vitro. The anti-thyroid effects of MMI may, at least in part, be mediated by a decrease in TFC-derived TGF-\g=b\1 production.

Journal of Molecular Endocrinology (1992) 9,

197-205

Sande et al. 1985). Previous studies have shown that DNA synthesis in cultured porcine thyroid follicular cells (TFCs) is markedly inhibited by micromolar levels of iodide (Becks et al. 1988; Tramontano et al. 1989; Beere et al. 1991). It has been suggested that such events may involve the mediation of thyroidal iodolipids (Pisarev et al. 1988), among which iodolactone derivatives of arachidonic acid metabolites (Boeynams & Hubbard, 1980; Gartner et al. 1985; Dugrillon et al. 1990) and an iodohexadecanal (Pereira et al. 1990) have been identified as potential candidates. In-vitro studies with TFCs from a number of species have provided clear evidence for the elabor¬ ation and release of several peptide growth factors, which may act in an autocrine or paracrine manner

within the thyroid follicle. Those factors presently identified include insulin-like growth factor-1 (IGF-I) (Bachrach et al. 1988; Tode et al. 1989) and basic fibroblast growth factor (FGF) (Logan et al. 1992). The thyroidal production of trans¬ forming growth factor-ß (TGF-ß) has also been reported (Grubeck-Loebenstein et al. 1989), and it has been proposed that TGF-ß synthesis within the TFC may be dependent upon the action of intracellular iodide or a specific metabolite(s) thereof (Tsushima et al. 1988; Grubeck-Loebenstein et al. 1989). Evidence that TGF-ß may assume fundamental regulatory roles within the thyroid fol¬ licle has been forthcoming from the studies of Morris et al. (1988), demonstrating a TGFß-induced reduction in TSH-dependent cyclic AMP accumulation, while Tsushima et al. (1988) noted a diminished uptake of iodide in TFCs exposed to

TGF-ß. The TGF-ßl

isoform of the peptide is widespread in cells of mammalian origin, and in many epithelial cells, including TFCs (Tsushima et al. 1988), mem¬ bers of the TGF-ß family act as negative regulators of cell proliferation (Sporn et al. 1986; Massague, 1990; Miller et al. 1990). In support of a possible role for TGF-ß in counteracting the effects of growthstimulatory factors within the normal thyroid folli¬ cle, we have shown that TGF-ßl reduces the release of IGF-I peptide from porcine TFCs (Beere et al. 1991). This study also showed an enhanced TFC DNA synthesis in the presence of TGF-ßl antiserum, and that similar treatment partially reverses the inhibitory effect of iodide on TFC DNA synthe¬ sis (Beere et al. 1991). Further evidence of a link between thyroidal iodide availability, TGF-ß pro¬ duction and the local control of thyroid cell prolifer¬ ation has been obtained by Grubeck-Loebenstein et al. (1989), who demonstrated a reduced level of TGF-ß production in TFCs derived from post¬ operative thyroid tissue specimens from patients with iodine-deficiency goitre. The present study was designed to investigate the extent to which the regulatory actions of iodide on DNA synthesis may be reflected in the autocrine production of TGF-ßl in normal TFCs. The studies described were undertaken using Northern blot analysis of porcine thyroidal RNA, combined with an assessment of the negative growthregulating activity of culture media conditioned by iodide-treated TFCs, using a TGF-ß responsive cell line, MvlLu. Some of these studies were presented at the 11th Joint Meeting of British Endocrine Societies (Cowin et al. 1992).

MATERIALS AND METHODS

Preparation of thyroid cell cultures Freshly excised porcine thyroid glands were obtained from the abattoir within 1 h of slaughter. Under aseptic conditions, pooled collections of eight to ten glands were trimmed of fat and con¬

nective tissue. The follicular material was pro¬ cessed to 1—2 mm fragments, which were pooled and washed repeatedly in Ca and Mg -free Hank's balanced salt solution (mHBSS) until the supernatant was clear of red blood cells and con¬ nective tissue debris. A two-stage digestion pro¬ cedure was then used to obtain intact follicles, this being a modification of the method of Reader et al. (1985). Briefly, the process involved initial treat¬ ment of the tissue fragments with a mixture of 0-25% (w/v) Dispase II and 005% (w/v) collagenase in mHBSS at 37 °C. Enzymatic digestion was mech¬ anically assisted with a motor-driven Teflon paddle in a tight-fitting polystyrene container with a rota¬ tion speed of approximately 200 r.p.m. After an initial incubation for 20 min, the supernatant digest was aseptically removed and the partially digested tissue fragments were redispersed in a fresh portion of mHBSS. After a further digestion for 1 h 15 min, the supernatant, containing single cells together with broken and intact follicles, was centrifuged (200 g, 5 min) and resuspended in Eagle's minimal essential medium (EMEM) containing 5% (v/v) fetal calf serum (FCS). Intact follicles were allowed to sediment by gravity for 20 min, after which the follicle-enriched fraction was resus¬ pended in EMEM/5% (w/v) FCS supplemented with 1 mmol L-glutamine/1, 100 mU penicillin/1, 100 mg streptomycin/1 and 500 mg fungizone/1. Follicular yield was assessed using a standard haemacytometer and, after dilution of the suspen¬ sion to 10 follicles/ml, 20 ml portions were seeded into 75 cm tissue culture flasks (Primaria, Falcon Plastics, Oxnard, CA, U.S.A.). The cultures were maintained under an atmosphere of 5% CO2 in air at 37 °C. Following incubation for 24 h, media were removed from the cultures and replaced with an equal volume of serum-free Ham's F12 medium (Coon's modification; mF12) containing the sup¬ plements previously described. Where appropriate for individual experiments, Nal and/or the iodide organification-blocking thionamide compound methimazole (MMI) were also added. Following test incubation periods of 24 or 72 h, media were removed and retained for analysis of growthregulating activity using the MvlLu cell line, and the cells were prepared for RNA extraction. -

Post-incubation treatment of TFC-conditioned media Media conditioned for 72 h by TFCs, in the pres¬ absence of iodide (10 or 100p:mol/l) and/ or MMI (1 mmol/1), were aseptically removed and stored frozen at —20 °C for subsequent analysis. Where appropriate, conditioned media were heated to 85 °C for 10 min to activate latent TGF-ß (Piao et al. 1990) and to abolish the activity of peptide growth stimulators of follicular cell origin simul¬ taneously. In some instances, media were immunoadsorbed with a protein-A affinity-purified rabbit anti-porcine TGF-ßl neutralizing antiserum ence or

(2-5mg/l).

Assessment of endogenous growth-regulating activity in TFC-conditioned media The residual growth-regulating activities of condi¬ tioned, treated media derived from porcine TFCs were assessed using the fetal mink lung cell line, MvlLu (Henderson et al. 1974). The proliferation of this cell line is enhanced by recombinant IGF-I (Cowin et al. 1992) and inhibited by increasing doses of TGF-ß, as determined by the incorporation of [methyl- H]thymidine into newly synthesized DNA (Iwata et al. 1985; Cheifetz et al. 1987). Cells from stock MvlLu cultures were resuspended in EMEM containing 2% (v/v) FCS, plated out into 24-well plates (10 cells/well) and maintained at 37 °C in TFCa water-saturated incubator. After 24 h, conditioned media were added to the cells to a final level of 10% (v/v), and incubation was continued for a further 24 h. At this time, [methyl- Hjthymidine (1 mCi/1; Amersham International pic, Amersham, Bucks, U.K.) was added to each well. After 24 h at 37 °C, the media were removed and the cells rinsed briefly in 500 p:l ice-cold 10% (v/v) trichloroacetic acid (TCA). This was then discarded and replaced with a further 500 \\\ aliquot of TCA. After incuba¬ tion for approximately 3 h at 4 °C, the TCA supernatants were removed by aspiration and 250 |ff 1 mol NaOH/1 added to each well. After overnight solubilization, 100 p:l aliquots of the alkaline digests were mixed with 5 ml liquid scintillant and counted for H activity using a beta-scintillation counter. Comparison between tests and controls was made using unpaired Student's i-test. Northern blot

analysis

of RNA

For Northern blot analysis, total RNA was isolated using a modification of the procedure described by Auffray & Rougeon (1980). Briefly, 107-108 cells were scraped into HBSS, pelleted and resuspended

in 1 ml sodium acetate

(10 mmol/l):LiCl (3 mol/1):

sodium dodecyl sulphate (SDS):0T% antifoam agent. After lysis by sonication, digests were left at 4 °C for 12 h. Each lysate was then centrifuged for 15 min in a microfuge and the supernatant was removed. To each pellet were added 500 nl LiCl (4 mol/l):urea (8 mol/1) and after mixing and recentrifugation supernatants were removed and replaced with 500 |ff sodium acetate (100mmol/l) containing 01% SDS and 500 id phenol : chloroform : isoamyl alcohol (24:24:1). The solutions were centrifuged for 10 min in a microfuge. The aqueous phase was removed and added to 3 mol sodium acetate/1 to a final level of 10% (v/v). Two volumes of 100% ethanol were then added. After mixing, solutions were stored at —20 °C overnight. RNA was recovered by centrifugation in a microfuge for 15 min at 4 °C and, after removing the super¬ natant, was washed with 70% ethanol. RNA pellets were finally resuspended in 50 p:l distilled water. RNA (10p:g) was fractionated by electrophoresis on a 1% agarose—formaldehyde gel. The separated RNA fractions were transferred to a nylon hybrid¬ ization membrane (Hybond N; Amersham Interna¬ tional pic), which was then subjected to u.v. crosslinking before being baked at 80 °C for 2 h. After prehybridization, the membrane was hybrid¬ ized with a P-labelled cDNA probe to human TGF-ßl, then exposed to X-ray film at —70 °C using an intensifying screen. The intensities of bands on the autoradiographs were determined using a flying spot densitometer (Shimadzu Corporation, Kyoto, urea

(6 mol/l):0T%

Japan).

The radiolabelled

TGF-ßl

cDNA probe was sub¬ from the membrane using 1 mmol Tris/1, 01 mmol EDTA/1 and 01% (w/v) SDS for 15 min at 95 °C. After prehybridization, the membrane-bound RNA was hybridized with an oligonucleotide probe for 28S rRNA, end-labelled with [y- PJATP (Amersham International pic). The quantitation of hybridizing 28S rRNA was made from autoradiographs using a flying-spot densitometer.

sequently stripped

Sources of materials

mHBSS, penicillin, streptomycin, FCS and EMEM were obtained from Gibco-BRL Ltd, Uxbridge, Middlesex, U.K. Dispase II was purchased from Boehringer Mannheim, Lewes, East Sussex, U.K. and collagenase and MMI were obtained from Sigma Chemical Co., Poole, Dorset, U.K. mF12 (Coon's modification) was obtained from Imperial Laboratories Ltd, Salisbury, Wilts, U.K. Human TGF-ßl and an oligonucleotide antisense probe

for 28S rRNA

were

purchased

from

Cambridge

Bioscience, Cambridge, Cambs, U.K. The cDNA probe to human TGF-ßl encoded the complete sequence of TGF-ßl, and included 779 bp of 5' untranslated sequence and 179 bp of 3' untranslated sequence; it was obtained from the American Type Culture Collection, Rockville, MD, U.S.A. A neu¬ tralizing antiserum for porcine TGF-ßl was purch¬ ased from British Biotechnology Ltd, Oxford, Oxon, U.K.

RESULTS

Effect of iodide and MMI

on

TGF-ßl

mRNA

Northern blot analysis of the RNA extracted from 72-h control TFC cultures revealed a 2-5 kb band hybridizing specifically with the cDNA probe for TGF-ßl. A minor band of 1 -6 kb was also observed. Fig. la shows an autoradiograph, representative of three experiments, of a Northern blot hybridized with this probe, while Fig. lb shows the densitometric analysis of the 2-5 kb TGF-ßl transcripts on the autoradiograph, in relation to the intensity of hybridization of the same blots with a 28S rRNA

probe. Exposure of TFCs to lOjimol Nal/l led to an approximate doubling in intensity of the TGF-ßl 25 kb transcript, compared with control cells (Fig. 16). In TFCs simultaneously exposed to 10 ^.mol

Nal/l and

1 mmol MMI/1, the stimulation seen with Nal alone was partly prevented, while the level of TGF-ßl mRNA in TFCs exposed to a high dose of MMI alone (10mmol/l) did not differ from that found in control cells (Fig. lb).

Growth-regulating activity of porcine

TFC-

conditioned media: effect of iodide

When tested on the MvlLu cell line, medium condi¬ tioned by control TFCs gave a significant (P