0021-972X/92/7501-0116$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright 0 1992 by The Endocrine Society
Vol. 75, No. 1 Printed in US A
Effects of Thyroid Hormone on Sex Hormone-Binding Globulin Gene Expression in Human Cells* LORNA E. RAGGATT, JOHN W. BARLOW Ewen Downie
Metabolic
ROZANNE
B. BLOK,
Unit, Alfred Hospital,
Melbourne,
P. SHANE Victoria
ABSTRACT
HAMBLIN,
AND
3181, Australia the concentration of free T, available to cells was aunroximatelv 15% __ of that added initially. ” Sex hormone-binding elobulin (SHBG) was secreted bv HeaGP cells in the absence of TJ and was specifically stimulated by the addition of T,. After 4 days, maximum stimulation occurred with added T, concentrations of lo-’ M or greater, and half-maximal stimulation of SHBG secretion was observed at about 3 X 10-l’ M free T:,. No significant changes in total secreted protein or cellular DNA content were observed under similar conditions. Northern analysis of RNA extracted from HepGP cells revealed a SHBG mRNA of 2 kilobases, which was stimulated in a dose-responsive manner by TI. No stimulation of corticosteroid-binding globulin mRNA was seen. Stimulation of the SHBG gene in HepG2 cells may be a useful model for investigations of T, action in human cells. (J Clin Endocrinol Metab 75: 116-120, 1992)
We have used a human hepatoblastoma cell line to establish a model system for thyroid hormone (T3) action in human cells. HepG2 cells were grown for 3 days in Dulbecco’s Modified Eagle’s Medium containing fetal calf serum and were maintained in serum-free medium for experimental manipulations. [‘251]T, incubated with cells was bound by newly secreted protein and degraded. After 24-h exposure to HepG2 cells in Dulbecco’s Modified Eagle’s Medium, only 35-40% of the radioactivity was recovered as authentic T,. Degradation of hormone was neither time nor concentration dependent, and occurred to a greater degree in the absence of cells, suggesting an interaction between the hormone and the plastic culture dish. After 4 days, in the absence of fetal calf serum and considering hormone binding and degradation,
T
HE ACTION of thyroid hormones in mammalian cells is mediated by a family of transcriptional regulators that bind the hormone and subsequently bind to elements of DNA associatedwith responsive genes (1, 2). In the rat, a number of different genes have been shown to respond to T3 (3-6), and a thyroid hormone response element has been defined (7). In contrast, the DNA region in responsive genes that mediates TX action in human tissuesis lesswell understood. There seemslittle doubt that the hormone affects gene transcription by an interaction with various c-erbA oncogene receptor forms (1, 2). However, there is no clear evidence that information about the mechanism of T3 action in man can be directly inferred from observations made in the rat. For example, T3 positively stimulates the rat GH gene in cultured rat pituitary cells (S), but has a negative influence on the human GH gene promoter when gene constructs are transfected into the samerat cell line (9). Such investigations may be better addressedin a homologous system in which T3 regulation of a human gene is investigated in a human cell line. The gene for sex hormone-binding globulin (SHBG) has recently been cloned (10). It is expressed in a human liver cell line, HepG2 cells, and is known to be influenced by PRL, insulin, estrogen, and T3 (11, 12). In this study we sought to develop an experimental system in which the influence of T3 Received April 15, 1991. Address all correspondence and requests for reprints to: John W. Barlow, M.Sc., Ph.D., Ewen Downie Metabolic Unit, Alfred Hospital, Commercial Road, Melbourne, Victoria 3181, Australia (use airmail please). * This work was supported in part by the National Health and Medical Research Council of Australia and the Alfred Hospital Medical Research Advisory Committee.
on SHBG gene expression could be examined in isolation. Our results show that SHBG secretion responds slowly to T3 at physiological free T3 concentrations and that increased secretion is preceded by an increase in SHBG mRNA. Stimulation of the SHBG gene in HepG2 cells may be a useful model for investigations of T3 action in human cells. Materials and Methods Cells HepG2 cells (13) were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Commonwealth Serum Laboratories, Parkville, Victoria, Australia) with 10% fetal calf serum (FCS; Commonwealth Serum Laboratories) and maintained at 37 C in 95% air and 5% carbon dioxide at 95% relative humidity. Before each experiment, cells were washed once with serum-free medium, and then the experimental medium was added.
Medium
Ti3 concentration
The concentration of bioavailable TZ was determined by incubating [1*51]T3 (>3000 wCi/pg; Amersham International, Aylesbury, Buckinghamshire, United Kingdom) in the presence and absence of cells. The proportion of radioactivity in the medium recovered as authentic [“‘I]T) was determined by Sephadex LH-20 chromatography (14). The free TS concentration in cell-conditioned medium was determined by equilibrium dialysis (15). Total T3 was determined by RIA after ethanol extraction (16).
Protein
measurements
SHBG secreted into the medium was measured by an immunoradiometric assay (Farmos Diagnostica, Oulunsalo, Finland). The total protein concentration was measured using Bio-Rad (Richmond, CA) dyeacid concentrate, with BSA (Commonwealth Serum Laboratories) as standard. 116
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
T:j EFFECT Estimation
of
SHBG reuptake
Reuptake of SHBG by cells 67 nmol/L SHBG diluted from ard. Cells were incubated with 4 days, and the residual SHBG determined as described above.
Northern
ON SHBG
analysis
Medium
r’“I]Tzi
Degradation of [1251]T3 over a 4-day period, in the presence and absence of cells, was studied in DMEM with or without FCS. Incubation of cells in DMEM with 10-l’ M [1251]T3 resulted in degradation of tracer within the first 24 h of incubation (Fig. la), with 35.3 + 11.9% (mean + SD; n = 3)
2
free T3 concentration
We measured the concentration of free T3 in medium exposed to cells for 4 days (Table 1). In the presence of FCS, the T3 free fraction was low, confirming extensive T3 binding to proteins present in FCS. Because FCS also contains endogenous T3, the free T3 concentration in the absence of added hormone was within the normal range. Hypothyroid values could not be obtained without further dilution or pretreatment of the FCS. Adding T3 to medium containing FCS increased the free concentration to well above the physiological range. In the absence of FCS, a much higher T3 free fraction was observed. Even so, the results suggested that a
Results of
117
of the radioactivity recovered as authentic [‘251]T3 (fractions 19-24). This effect was neither time dependent, as no further degradation was observed even after 4 days of incubation (Fig. lb; T3 recovered, 44.2 + 12.5%), nor concentration dependent, since similar proportional degradation was observed with lop8 M [‘251]T3 (not shown). Surprisingly, incubation of [1251]T3 for 4 days in the absence of cells resulted in a higher degree of tracer degradation (Fig. lc) than when cells were present, suggesting that the T3 degradation we observed was not uniquely due to cellular metabolism. When similar experiments were performed in the presence of FCS, degradation of [‘251]T3 was almost completely inhibited (not shown). No degradation was observed when tracer was stored at 4 C in glass in the dark (Fig. Id).
was investigated by incubating cells with a SHBG immunoradiometric assay standand without added SHBG in DMEM for concentration in the culture medium was
Total RNA (30 Kg) was extracted from T,-treated cells (17), denatured, and electrophoresed in a denaturing formaldehyde system using a 1% agarose gel (18). Capillary blot transfer of the RNA to a nylon support membrane (Hybond-N, Amersham) was performed in 20 X SSC (3 M NaCl and 0.3 M Na, citrate) overnight according to manufacturer’s instructions (Amersham). Baked blots were hybridized with [a-3zP] deoxy-CTP-labeled SHBG and corticosteroid-binding globulin (CBG) probe DNA. Expression vectors containing the cDNA for SHBG (10) and CBG (19) were provided by Dr. G. L. Hammond, University of Western Ontario (London, Ontario, Canada). Probes were labeled by the random hexamer priming method using a kit (Multiprime labeling kit, Amersham).
Degradation
SECRETION
(a)
2r
(b)
1
I
6
12
18
24
6
FIG. 1. Representative chromatographic profiles on Sephadex LH-20 after incubation of 10-r’ M [‘251]Ta in plastic culture dishes with HepG2 cells in DMEM for 1 day (a), HepG2 cells in DMEM for 4 days (b), DMEM for 4 days (c), and untreated control (d). Free iodine elutes at fraction 6, and authentic T, at fraction
12
18
24
12 18 Fraction number
24
(4
23.
2
1
6
12 Fraction
18 number
24
u-LI
*
l
6.
. l *
l *a+
l
**-*ad_,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
RAGGATT
118
TABLE 1. Concentration of free T, in HepGP cell-conditioned medium (DMEM with or without FCS) after 4 days of incubation the presence of various concentrations of total Ta Ta free fraction” -FCS 0.443 0.423 0.495 0.394 0.400 +FCS 0.102 0.100 0.111 0.165
Total Ta (nmol/L)b 0.0001 0.001 0.01 0.1 1.0 0.13 4.5 30.5 248
Recovery of SHBG from medium
in
Residual SHBG present in cell-conditioned medium was determined after 4 days of incubation with added SHBG, as described in Materials and Methods. In the absenceof added SHBG, the cells secreted 16.4 nmol/L SHBG (Table 2). In the presence of 67 nmol/L added SHBG, 56.4 nmol/L SHBG were recovered, representing a total loss of 32%. A no-cells control showed that 14% of the added SHBG was lost, presumably due to nonspecific adsorption to the culture vessel. This suggeststhat uptake or attachment of SHBG by the cells accounted for the loss of 18% of SHBG from the culture medium.
Free Ta (pmwz
0.016 0.148 1.73 13.8 140 6.65 226 1,690 20,400
Data are the means of duplicates. a Determined by equilibrium dialysis. *Added exogenously and/or endogenous in FCS, as measured RIA. ’ Calculated from the free fraction and the degree of iodothyronine degration, as observed by Sephadex chromatography.
Northern by
low level of T3 binding was present even in the serum-free medium, probably due to binding proteins secreted de nova by the cells. The free T3 concentration was also corrected by 44.2% for hormone degradation, as described above. After these corrections, the concentration of free T3 available to cells in the absence of FCS was approximately 15% of the added T3 (Table 1). These data showed that by appropriate addition of T3 to serum-free medium, cells were exposed to a physiologically relevant range of free hormone concentrations. SHBG
in cell-conditioned
JCE & M. 1992 Vol75.Nol
ET AL.
medium
Cells were maintained for up to 4 days in DMEM with a range of T3 concentrations, and the concentration of SHBG secreted into the medium was determined (Fig. 2a). In the absenceof T3 or at very low concentrations, SHBG secretion increased approximately ‘/-fold with time. Stimulation of SHBG secretion by T3 was seen in the range of added T3 concentrations of 10-‘“-10-6 M after 3 and 4 days in culture. Maximal stimulation was observed on day 4 at an added T3 concentration between 1O-s-1O-7M. This represented a 17fold stimulation of SHBG secretion by T3 with time and a 2.5-fold increase above the unstimulated level on day 4. Total protein secretedinto the medium increased between 2and 3-fold with time, but was not influenced by T3 (Fig. 2b). Variations in protein concentration were not statistically significant when analyzed by one-way analysis of variance. These results suggest that T3 acts specifically to stimulate SHBG secretion, rather than by increasing total protein secretion in general. The cellular DNA concentration in the absence and presence of T3 was determined in order to estimate the effect of T3 on cell growth. No difference was seen when cells were treated for 4 days with 10m7M T3 (1.05 + 0.31 pLg/pL) compared with the control value (0.93 + 0.38 Kg/FL; mean + SD; n = 4).
analysis
Figure 3 shows the results of Northern analysis of RNA isolated from HepG2 cells cultured for 1 day in DMEM containing T3. Hybridization with [32P]deoxy-CTP-labeled SHBG cDNA (Fig. 3) shows that the transcript detected was responsive to a range of T3 concentrations up to 10m7M. Stripping and rehybridization of this filter with the cDNA for CBG (Fig. 3) showed no response of this transcript to increasing concentrations of T3. Discussion It is well known that T3 is one of a number of factors that influence the circulating concentration of SHBG (20). Others have shown that the human cell line HepG2 synthesizes and secretes SHBG and that this process is influenced to some degree by T3 (11, 12). It was our aim here to optimize the conditions under which T3 promotes SHBG secretion in order to establish an in vitro model for T3 action in human cells. T3 added to cells maintained in DMEM is both degraded and bound to newly secreted binding proteins. We found that more than 35% of exogenous T3 was degraded within the first 24 h of addition to the culture flask. We alsoobserved considerable degradation of T3 exposed to DMEM alone, suggestingthat this phenomenon may not be due to cellular metabolism of T3, but results from interaction between the hormone and the plastic culture dish. Binding proteins secreted by the cells into the medium further contributed to a reduction in the concentration of free T3 available to the cells. After 4 days, considering both degradation and extracellular binding, the concentration of T3 available to the cells was approximately 15% of that added initially. T3 treatment of HepG2 cells resulted in stimulation of SHBG secretion in a dose-responsive manner after 3 and 4 days in culture. Considering the degradation and binding of T3 in the culture medium, this response was maximal at a free T3 concentration of between 10-9-10-8 M and is halfmaximal at 3 x 1Ollh.1.We observed some decreasein measurable SHBG at high concentrations (-20%), so the amount of SHBG secreted by the cells may be slightly underestimated. Even so, the half-maximal stimulation is approximately the same as the affinity for T3 receptors described in other cells and is virtually identical to the affinity of T3binding sites we measured in the nuclei of HepG2 cells. In
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
Ta EFFECT
ON SHBG
SECRETION
125.
(b)
FIG. 2. Secretion of SHBG by HepG2 cells. Cells were incubated in serum-free medium and a range of T3 concentrations for 1 (O), 2 (O), 3 (D), or 4 (0) days. SHBG (a) and total protein (b) secreted into the medium over time were estimated as described in Materials and Methods. *, P < 0.05; **, P < 0.02; ***, P < 0.01; ****, P < 0.001 (compared with the no hormone control). Values are the mean k SEM (n = 4).
310
9 TS cont. added
8 ( -log
7 molll
--i
6
1
TABLE 2. Residual SHBG concentration in HepG2 cellconditioned medium after 4 days of incubation with or without added SHBG Added SHBC (nmol/L) 67
Recovered SHBG (nmol/L) 57.9 + 0.08” No cells 16.4 + 0.42* Cells 67 56.4 + 3.39’ Cells These results are the mean + SD of quadruplicate determinations. The pooled means and SDS of ’ and * are significantly different from those of’ (P < 0.001).
two separateexperiments (data not shown) we found a single classof T,-binding sites with an affinity (KJ of 3.6 X lo-” M and 8000 sites/cell. HepG2 cells are a continuous cell line derived originally from a human hepatoblastoma and are chromosomally abnormal (13). Southern analysis of HepG2 cell DNA and subsequent probing for the SHBG gene (data not shown) gave a DNA pattern identical to that obtained with normal human placental DNA (21), suggesting that the SHBG gene in HepG2 cells has not undergone any major deletions or insertions. Treatment of the cells with T3 stimulated the gene to increase SHBG mRNA production. On day 1, the cells produced a full-length mRNA for SHBG, which responded to T3 over a concentration range from O-10-’ M added T3. These data confirm the results of Mercier-Bodard et al. (22), who detected SHBG mRNA in H5A cells, a HepG2 cell derivative, treated with high concentrations of T3 in the presenceof serum. Our data also show that SHBG mRNA is responsive to different concentrations of free T3 and suggest that the stimulatory effect of the hormone is at least in part due to an effect at the level of gene transcription. In contrast,
SHBG-,
FIG. 3. Northern analysis of RNA from HepG2 cells. HepG2 cells were incubated for 1 day in serum-free DMEM containing 0, lo-“, 10e9, lo-‘, or lo-’ M Ta. Total RNA was extracted and subjected to Northern analysis for SHBG mRNA. The blot was then stripped and reprobed with a CBG-specific probe. The relative positions of the 28s and 18s ribosomal RNA bands are indicated. A representative blot is shown. Another experiment was virtually identical.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
RAGGATT ET AL.
120
the mRNA for CBG was unstimulated by TJ, in agreement
with other observations that circulating levels of CBG do not respond to changes in T3 (23). Perhaps the most salient finding in the present study is the slow responseof SHBG secretion to TX. We did not seea measurable response in SHBG secreted into the medium until 3 days of T3 treatment, in marked contrast to the speed of responsesto T3 documented in the rat and rat cell lines. The mRNA for S14, for example, responds within 20 min to T3 administration (24). Our studies suggest that the mRNA for SHBG may also respond rapidly to T3, and the slow secretion in our study may be a result of delays in processing and secretion of the final product. If this is so, then stimulation of the SHBG gene may occur through a direct interaction of the T3 receptor with T3 response elements located in the 5’-flanking regions of the gene. Preliminary examination of the published sequence of the gene (21) has not revealed a thyroid response element in the 2700 basepairs up-stream from the transcription initiation site. The location of this element is currently the subject of investigation in our laboratory. Acknowledgments We wish to thank Dr. G. L. Hammond who generously provided the SHBG and CBG probes, and Dr. J. Apostopoulos Department of Biochemistry, University of Melbourne, who provided the HepG2 cells.
References 1. Samuels HH, Forman BM, Horowitz ZD, Zheng-Sheng Ye. 1988 Regulation of gene expression by thyroid hormone. J Clin Invest. 81:957-67. 2. Evans RM. 1988 The steroid and thyroid hormone receptor superfamily. Science. 240:889-95. 3. Glass CK, Franc0 R, Weinberger C, Albert VR, Evans RM, Rosenfeld MG. 1987 A c-erbA binding site in rat growth hormone gene mediates trans-activation by thyroid hormone. Nature. 329:738-41. 4. Izumo S, Mahdavi V. 1988 Thyroid hormone receptor (Y isoforms generated by alternative splicing differentially activate myosin HC gene transcription. Nature. 334:539-42. 5. Petty KJ, Desvergne B, Mitsuhashi T, Nikodem VM. 1990 Identification of a thyroid response element in the malic enzyme gene. J Biol Chem. 265:7395-400. 6. Zilz ND, Murray MB, Towle HC. 1990 Identification of multiple thvroid hormone response elements located far upstream from the rat Si4 promoter. J Bib1 Chem. 265:8136-43. 7. Glass CK. Hollowav TM. Devarv OV. Rosenfeld MG. 1988 The thyroid hormone reckpior’binds with opposite transcriptional effects to a common sequence motif in thyroid hormone and estrogen response elements. Cell. 54:313-23.
JCE & M. 1992 Vol75.Nol
8. Seo H, Vassart G, Brocas H, Refetoff S. 1977 Triiodothyronine stimulates specifically growth hormone mRNA in rat pituitary tumor cells. Proc Nat1 Acad Sci USA. 74:2054-B. 9. Cattini PA, Anderson TR, Baxter JD, Mellon P, Eberhardt NL. 1986 The human growth hormone gene is negatively regulated by triiodothyronine when transfected into rat pituitary tumor cells. J Biol Chem. 261:13367-72. GL, Underhill DA, Smith CL, et al. 1987 The cDNA 10. Hammond deduced primary structure of human sex hormone-binding globulin and location of its steroid-binding domain. FEBS Lett. 215:100-4. 11. Plvmate SR, Matei LA, Tones RE, Fried1 KE. 1988 Inhibition of sex hormone-binding glob&n production in the human hepatoma (Hep G2) cell line bv insulin and orolactin. 1 Clin Endocrinol Metab. I 67:460-4. ’ 12. Luppa P, Oettrich K, Schwab I, Langmandel U, Neumeier D. 1989 Immunocytochemical localization of the sex hormone-binding globulin in a human hepatoma cell line. Acta Endocrinol (Copenh). 121:791-6. 13. Knowles BB, Howe CC, Aden DP. 1980 Human hepatocellular carcinoma cell lines secrete the major plasma proteins, hepatitis B surface antigen. Science. 209:497-9. 14. Otten MH, Mol JA, Visser TJ. 1983 Sulfation preceeding deiodination of iodothyronines in rat hepatocytes. Science. 221:81-3. 15. Stockigt JR, Topliss DJ, Barlow JW, White EL, Hurley DM, Taft P. 1981 Familial euthyroid thyroxine excess: an appropriate response to abnormal thyroxine binding associated with albumin. J Clin Endocrinol Metab. 53:353-9. 16. Stockigt JR, White EL, Petrou S, Taft P. 1979 The course of untreated mild T3 toxicosis. J Aust. 2:6-7. 17. Chomczynski P, Sacchi N. 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 162:156-9. J, Fritsch EF, Maniatis T. 1989 Molecular cloning-a 18. Sambrook laboratory manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor Laboratory. 19. Hammond GL, Smith CL, Goping IS, et al. 1987 Primary structure of human corticosteroid binding globulin, deduced from hepatic and pulmonary cDNAs exhibits, homology with serine protease inhibitors. Proc Nat1 Acad Sci USA. 84:5153-7. 20. Sarne DH, Refetoff S, Rosenfield RL, Farriaux JP. 1988 Sex hormone-binding globulin in the diagnosis of peripheral tissue resistance to thyroid hormone: the value of changes after short term triiodothyronine administration. J Clin Endocrinol Metab. 66:7406. 21. Hammond GL, Underhill DA, Rykse HM, Smith CL. 1989 The human sex hormone-binding globulin gene contains exons for androgen-binding protein and two other testicular messenger RNAs. Mol Endocrinol. 3:1869-76. 22. Mercier-Bodard C, Baville F, Bideux G, Binart N, Chambraud B, Baulieu E-E. 1989 Regulation of SBP synthesis in human cancer cell lines by steroid and thyroid hormones. J Steroid Biochem. 34:199-204. 23. Rosner W, Aden DP, Khan MS. 1984 Hormonal influences on the secretion of steroid-binding proteins by a human hepatoma-derived cell line. J Clin Endocrinol Metab. 59:806-B. 24. Jump DB, Narayan P, Towle H, Oppenheimer JH. 1983 Rapid effects of triiodothyronine on hepatic gene expression: hybridization analysis of tissue-specific regulation of mRNA-S14. J Biol Chem. 259:2789-97.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
Vol. 75, No. 1 Printed in US A
Effects of Thyroid Hormone on Sex Hormone-Binding Globulin Gene Expression in Human Cells* LORNA E. RAGGATT, JOHN W. BARLOW Ewen Downie
Metabolic
ROZANNE
B. BLOK,
Unit, Alfred Hospital,
Melbourne,
P. SHANE Victoria
ABSTRACT
HAMBLIN,
AND
3181, Australia the concentration of free T, available to cells was aunroximatelv 15% __ of that added initially. ” Sex hormone-binding elobulin (SHBG) was secreted bv HeaGP cells in the absence of TJ and was specifically stimulated by the addition of T,. After 4 days, maximum stimulation occurred with added T, concentrations of lo-’ M or greater, and half-maximal stimulation of SHBG secretion was observed at about 3 X 10-l’ M free T:,. No significant changes in total secreted protein or cellular DNA content were observed under similar conditions. Northern analysis of RNA extracted from HepGP cells revealed a SHBG mRNA of 2 kilobases, which was stimulated in a dose-responsive manner by TI. No stimulation of corticosteroid-binding globulin mRNA was seen. Stimulation of the SHBG gene in HepG2 cells may be a useful model for investigations of T, action in human cells. (J Clin Endocrinol Metab 75: 116-120, 1992)
We have used a human hepatoblastoma cell line to establish a model system for thyroid hormone (T3) action in human cells. HepG2 cells were grown for 3 days in Dulbecco’s Modified Eagle’s Medium containing fetal calf serum and were maintained in serum-free medium for experimental manipulations. [‘251]T, incubated with cells was bound by newly secreted protein and degraded. After 24-h exposure to HepG2 cells in Dulbecco’s Modified Eagle’s Medium, only 35-40% of the radioactivity was recovered as authentic T,. Degradation of hormone was neither time nor concentration dependent, and occurred to a greater degree in the absence of cells, suggesting an interaction between the hormone and the plastic culture dish. After 4 days, in the absence of fetal calf serum and considering hormone binding and degradation,
T
HE ACTION of thyroid hormones in mammalian cells is mediated by a family of transcriptional regulators that bind the hormone and subsequently bind to elements of DNA associatedwith responsive genes (1, 2). In the rat, a number of different genes have been shown to respond to T3 (3-6), and a thyroid hormone response element has been defined (7). In contrast, the DNA region in responsive genes that mediates TX action in human tissuesis lesswell understood. There seemslittle doubt that the hormone affects gene transcription by an interaction with various c-erbA oncogene receptor forms (1, 2). However, there is no clear evidence that information about the mechanism of T3 action in man can be directly inferred from observations made in the rat. For example, T3 positively stimulates the rat GH gene in cultured rat pituitary cells (S), but has a negative influence on the human GH gene promoter when gene constructs are transfected into the samerat cell line (9). Such investigations may be better addressedin a homologous system in which T3 regulation of a human gene is investigated in a human cell line. The gene for sex hormone-binding globulin (SHBG) has recently been cloned (10). It is expressed in a human liver cell line, HepG2 cells, and is known to be influenced by PRL, insulin, estrogen, and T3 (11, 12). In this study we sought to develop an experimental system in which the influence of T3 Received April 15, 1991. Address all correspondence and requests for reprints to: John W. Barlow, M.Sc., Ph.D., Ewen Downie Metabolic Unit, Alfred Hospital, Commercial Road, Melbourne, Victoria 3181, Australia (use airmail please). * This work was supported in part by the National Health and Medical Research Council of Australia and the Alfred Hospital Medical Research Advisory Committee.
on SHBG gene expression could be examined in isolation. Our results show that SHBG secretion responds slowly to T3 at physiological free T3 concentrations and that increased secretion is preceded by an increase in SHBG mRNA. Stimulation of the SHBG gene in HepG2 cells may be a useful model for investigations of T3 action in human cells. Materials and Methods Cells HepG2 cells (13) were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Commonwealth Serum Laboratories, Parkville, Victoria, Australia) with 10% fetal calf serum (FCS; Commonwealth Serum Laboratories) and maintained at 37 C in 95% air and 5% carbon dioxide at 95% relative humidity. Before each experiment, cells were washed once with serum-free medium, and then the experimental medium was added.
Medium
Ti3 concentration
The concentration of bioavailable TZ was determined by incubating [1*51]T3 (>3000 wCi/pg; Amersham International, Aylesbury, Buckinghamshire, United Kingdom) in the presence and absence of cells. The proportion of radioactivity in the medium recovered as authentic [“‘I]T) was determined by Sephadex LH-20 chromatography (14). The free TS concentration in cell-conditioned medium was determined by equilibrium dialysis (15). Total T3 was determined by RIA after ethanol extraction (16).
Protein
measurements
SHBG secreted into the medium was measured by an immunoradiometric assay (Farmos Diagnostica, Oulunsalo, Finland). The total protein concentration was measured using Bio-Rad (Richmond, CA) dyeacid concentrate, with BSA (Commonwealth Serum Laboratories) as standard. 116
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
T:j EFFECT Estimation
of
SHBG reuptake
Reuptake of SHBG by cells 67 nmol/L SHBG diluted from ard. Cells were incubated with 4 days, and the residual SHBG determined as described above.
Northern
ON SHBG
analysis
Medium
r’“I]Tzi
Degradation of [1251]T3 over a 4-day period, in the presence and absence of cells, was studied in DMEM with or without FCS. Incubation of cells in DMEM with 10-l’ M [1251]T3 resulted in degradation of tracer within the first 24 h of incubation (Fig. la), with 35.3 + 11.9% (mean + SD; n = 3)
2
free T3 concentration
We measured the concentration of free T3 in medium exposed to cells for 4 days (Table 1). In the presence of FCS, the T3 free fraction was low, confirming extensive T3 binding to proteins present in FCS. Because FCS also contains endogenous T3, the free T3 concentration in the absence of added hormone was within the normal range. Hypothyroid values could not be obtained without further dilution or pretreatment of the FCS. Adding T3 to medium containing FCS increased the free concentration to well above the physiological range. In the absence of FCS, a much higher T3 free fraction was observed. Even so, the results suggested that a
Results of
117
of the radioactivity recovered as authentic [‘251]T3 (fractions 19-24). This effect was neither time dependent, as no further degradation was observed even after 4 days of incubation (Fig. lb; T3 recovered, 44.2 + 12.5%), nor concentration dependent, since similar proportional degradation was observed with lop8 M [‘251]T3 (not shown). Surprisingly, incubation of [1251]T3 for 4 days in the absence of cells resulted in a higher degree of tracer degradation (Fig. lc) than when cells were present, suggesting that the T3 degradation we observed was not uniquely due to cellular metabolism. When similar experiments were performed in the presence of FCS, degradation of [‘251]T3 was almost completely inhibited (not shown). No degradation was observed when tracer was stored at 4 C in glass in the dark (Fig. Id).
was investigated by incubating cells with a SHBG immunoradiometric assay standand without added SHBG in DMEM for concentration in the culture medium was
Total RNA (30 Kg) was extracted from T,-treated cells (17), denatured, and electrophoresed in a denaturing formaldehyde system using a 1% agarose gel (18). Capillary blot transfer of the RNA to a nylon support membrane (Hybond-N, Amersham) was performed in 20 X SSC (3 M NaCl and 0.3 M Na, citrate) overnight according to manufacturer’s instructions (Amersham). Baked blots were hybridized with [a-3zP] deoxy-CTP-labeled SHBG and corticosteroid-binding globulin (CBG) probe DNA. Expression vectors containing the cDNA for SHBG (10) and CBG (19) were provided by Dr. G. L. Hammond, University of Western Ontario (London, Ontario, Canada). Probes were labeled by the random hexamer priming method using a kit (Multiprime labeling kit, Amersham).
Degradation
SECRETION
(a)
2r
(b)
1
I
6
12
18
24
6
FIG. 1. Representative chromatographic profiles on Sephadex LH-20 after incubation of 10-r’ M [‘251]Ta in plastic culture dishes with HepG2 cells in DMEM for 1 day (a), HepG2 cells in DMEM for 4 days (b), DMEM for 4 days (c), and untreated control (d). Free iodine elutes at fraction 6, and authentic T, at fraction
12
18
24
12 18 Fraction number
24
(4
23.
2
1
6
12 Fraction
18 number
24
u-LI
*
l
6.
. l *
l *a+
l
**-*ad_,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
RAGGATT
118
TABLE 1. Concentration of free T, in HepGP cell-conditioned medium (DMEM with or without FCS) after 4 days of incubation the presence of various concentrations of total Ta Ta free fraction” -FCS 0.443 0.423 0.495 0.394 0.400 +FCS 0.102 0.100 0.111 0.165
Total Ta (nmol/L)b 0.0001 0.001 0.01 0.1 1.0 0.13 4.5 30.5 248
Recovery of SHBG from medium
in
Residual SHBG present in cell-conditioned medium was determined after 4 days of incubation with added SHBG, as described in Materials and Methods. In the absenceof added SHBG, the cells secreted 16.4 nmol/L SHBG (Table 2). In the presence of 67 nmol/L added SHBG, 56.4 nmol/L SHBG were recovered, representing a total loss of 32%. A no-cells control showed that 14% of the added SHBG was lost, presumably due to nonspecific adsorption to the culture vessel. This suggeststhat uptake or attachment of SHBG by the cells accounted for the loss of 18% of SHBG from the culture medium.
Free Ta (pmwz
0.016 0.148 1.73 13.8 140 6.65 226 1,690 20,400
Data are the means of duplicates. a Determined by equilibrium dialysis. *Added exogenously and/or endogenous in FCS, as measured RIA. ’ Calculated from the free fraction and the degree of iodothyronine degration, as observed by Sephadex chromatography.
Northern by
low level of T3 binding was present even in the serum-free medium, probably due to binding proteins secreted de nova by the cells. The free T3 concentration was also corrected by 44.2% for hormone degradation, as described above. After these corrections, the concentration of free T3 available to cells in the absence of FCS was approximately 15% of the added T3 (Table 1). These data showed that by appropriate addition of T3 to serum-free medium, cells were exposed to a physiologically relevant range of free hormone concentrations. SHBG
in cell-conditioned
JCE & M. 1992 Vol75.Nol
ET AL.
medium
Cells were maintained for up to 4 days in DMEM with a range of T3 concentrations, and the concentration of SHBG secreted into the medium was determined (Fig. 2a). In the absenceof T3 or at very low concentrations, SHBG secretion increased approximately ‘/-fold with time. Stimulation of SHBG secretion by T3 was seen in the range of added T3 concentrations of 10-‘“-10-6 M after 3 and 4 days in culture. Maximal stimulation was observed on day 4 at an added T3 concentration between 1O-s-1O-7M. This represented a 17fold stimulation of SHBG secretion by T3 with time and a 2.5-fold increase above the unstimulated level on day 4. Total protein secretedinto the medium increased between 2and 3-fold with time, but was not influenced by T3 (Fig. 2b). Variations in protein concentration were not statistically significant when analyzed by one-way analysis of variance. These results suggest that T3 acts specifically to stimulate SHBG secretion, rather than by increasing total protein secretion in general. The cellular DNA concentration in the absence and presence of T3 was determined in order to estimate the effect of T3 on cell growth. No difference was seen when cells were treated for 4 days with 10m7M T3 (1.05 + 0.31 pLg/pL) compared with the control value (0.93 + 0.38 Kg/FL; mean + SD; n = 4).
analysis
Figure 3 shows the results of Northern analysis of RNA isolated from HepG2 cells cultured for 1 day in DMEM containing T3. Hybridization with [32P]deoxy-CTP-labeled SHBG cDNA (Fig. 3) shows that the transcript detected was responsive to a range of T3 concentrations up to 10m7M. Stripping and rehybridization of this filter with the cDNA for CBG (Fig. 3) showed no response of this transcript to increasing concentrations of T3. Discussion It is well known that T3 is one of a number of factors that influence the circulating concentration of SHBG (20). Others have shown that the human cell line HepG2 synthesizes and secretes SHBG and that this process is influenced to some degree by T3 (11, 12). It was our aim here to optimize the conditions under which T3 promotes SHBG secretion in order to establish an in vitro model for T3 action in human cells. T3 added to cells maintained in DMEM is both degraded and bound to newly secreted binding proteins. We found that more than 35% of exogenous T3 was degraded within the first 24 h of addition to the culture flask. We alsoobserved considerable degradation of T3 exposed to DMEM alone, suggestingthat this phenomenon may not be due to cellular metabolism of T3, but results from interaction between the hormone and the plastic culture dish. Binding proteins secreted by the cells into the medium further contributed to a reduction in the concentration of free T3 available to the cells. After 4 days, considering both degradation and extracellular binding, the concentration of T3 available to the cells was approximately 15% of that added initially. T3 treatment of HepG2 cells resulted in stimulation of SHBG secretion in a dose-responsive manner after 3 and 4 days in culture. Considering the degradation and binding of T3 in the culture medium, this response was maximal at a free T3 concentration of between 10-9-10-8 M and is halfmaximal at 3 x 1Ollh.1.We observed some decreasein measurable SHBG at high concentrations (-20%), so the amount of SHBG secreted by the cells may be slightly underestimated. Even so, the half-maximal stimulation is approximately the same as the affinity for T3 receptors described in other cells and is virtually identical to the affinity of T3binding sites we measured in the nuclei of HepG2 cells. In
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
Ta EFFECT
ON SHBG
SECRETION
125.
(b)
FIG. 2. Secretion of SHBG by HepG2 cells. Cells were incubated in serum-free medium and a range of T3 concentrations for 1 (O), 2 (O), 3 (D), or 4 (0) days. SHBG (a) and total protein (b) secreted into the medium over time were estimated as described in Materials and Methods. *, P < 0.05; **, P < 0.02; ***, P < 0.01; ****, P < 0.001 (compared with the no hormone control). Values are the mean k SEM (n = 4).
310
9 TS cont. added
8 ( -log
7 molll
--i
6
1
TABLE 2. Residual SHBG concentration in HepG2 cellconditioned medium after 4 days of incubation with or without added SHBG Added SHBC (nmol/L) 67
Recovered SHBG (nmol/L) 57.9 + 0.08” No cells 16.4 + 0.42* Cells 67 56.4 + 3.39’ Cells These results are the mean + SD of quadruplicate determinations. The pooled means and SDS of ’ and * are significantly different from those of’ (P < 0.001).
two separateexperiments (data not shown) we found a single classof T,-binding sites with an affinity (KJ of 3.6 X lo-” M and 8000 sites/cell. HepG2 cells are a continuous cell line derived originally from a human hepatoblastoma and are chromosomally abnormal (13). Southern analysis of HepG2 cell DNA and subsequent probing for the SHBG gene (data not shown) gave a DNA pattern identical to that obtained with normal human placental DNA (21), suggesting that the SHBG gene in HepG2 cells has not undergone any major deletions or insertions. Treatment of the cells with T3 stimulated the gene to increase SHBG mRNA production. On day 1, the cells produced a full-length mRNA for SHBG, which responded to T3 over a concentration range from O-10-’ M added T3. These data confirm the results of Mercier-Bodard et al. (22), who detected SHBG mRNA in H5A cells, a HepG2 cell derivative, treated with high concentrations of T3 in the presenceof serum. Our data also show that SHBG mRNA is responsive to different concentrations of free T3 and suggest that the stimulatory effect of the hormone is at least in part due to an effect at the level of gene transcription. In contrast,
SHBG-,
FIG. 3. Northern analysis of RNA from HepG2 cells. HepG2 cells were incubated for 1 day in serum-free DMEM containing 0, lo-“, 10e9, lo-‘, or lo-’ M Ta. Total RNA was extracted and subjected to Northern analysis for SHBG mRNA. The blot was then stripped and reprobed with a CBG-specific probe. The relative positions of the 28s and 18s ribosomal RNA bands are indicated. A representative blot is shown. Another experiment was virtually identical.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
RAGGATT ET AL.
120
the mRNA for CBG was unstimulated by TJ, in agreement
with other observations that circulating levels of CBG do not respond to changes in T3 (23). Perhaps the most salient finding in the present study is the slow responseof SHBG secretion to TX. We did not seea measurable response in SHBG secreted into the medium until 3 days of T3 treatment, in marked contrast to the speed of responsesto T3 documented in the rat and rat cell lines. The mRNA for S14, for example, responds within 20 min to T3 administration (24). Our studies suggest that the mRNA for SHBG may also respond rapidly to T3, and the slow secretion in our study may be a result of delays in processing and secretion of the final product. If this is so, then stimulation of the SHBG gene may occur through a direct interaction of the T3 receptor with T3 response elements located in the 5’-flanking regions of the gene. Preliminary examination of the published sequence of the gene (21) has not revealed a thyroid response element in the 2700 basepairs up-stream from the transcription initiation site. The location of this element is currently the subject of investigation in our laboratory. Acknowledgments We wish to thank Dr. G. L. Hammond who generously provided the SHBG and CBG probes, and Dr. J. Apostopoulos Department of Biochemistry, University of Melbourne, who provided the HepG2 cells.
References 1. Samuels HH, Forman BM, Horowitz ZD, Zheng-Sheng Ye. 1988 Regulation of gene expression by thyroid hormone. J Clin Invest. 81:957-67. 2. Evans RM. 1988 The steroid and thyroid hormone receptor superfamily. Science. 240:889-95. 3. Glass CK, Franc0 R, Weinberger C, Albert VR, Evans RM, Rosenfeld MG. 1987 A c-erbA binding site in rat growth hormone gene mediates trans-activation by thyroid hormone. Nature. 329:738-41. 4. Izumo S, Mahdavi V. 1988 Thyroid hormone receptor (Y isoforms generated by alternative splicing differentially activate myosin HC gene transcription. Nature. 334:539-42. 5. Petty KJ, Desvergne B, Mitsuhashi T, Nikodem VM. 1990 Identification of a thyroid response element in the malic enzyme gene. J Biol Chem. 265:7395-400. 6. Zilz ND, Murray MB, Towle HC. 1990 Identification of multiple thvroid hormone response elements located far upstream from the rat Si4 promoter. J Bib1 Chem. 265:8136-43. 7. Glass CK. Hollowav TM. Devarv OV. Rosenfeld MG. 1988 The thyroid hormone reckpior’binds with opposite transcriptional effects to a common sequence motif in thyroid hormone and estrogen response elements. Cell. 54:313-23.
JCE & M. 1992 Vol75.Nol
8. Seo H, Vassart G, Brocas H, Refetoff S. 1977 Triiodothyronine stimulates specifically growth hormone mRNA in rat pituitary tumor cells. Proc Nat1 Acad Sci USA. 74:2054-B. 9. Cattini PA, Anderson TR, Baxter JD, Mellon P, Eberhardt NL. 1986 The human growth hormone gene is negatively regulated by triiodothyronine when transfected into rat pituitary tumor cells. J Biol Chem. 261:13367-72. GL, Underhill DA, Smith CL, et al. 1987 The cDNA 10. Hammond deduced primary structure of human sex hormone-binding globulin and location of its steroid-binding domain. FEBS Lett. 215:100-4. 11. Plvmate SR, Matei LA, Tones RE, Fried1 KE. 1988 Inhibition of sex hormone-binding glob&n production in the human hepatoma (Hep G2) cell line bv insulin and orolactin. 1 Clin Endocrinol Metab. I 67:460-4. ’ 12. Luppa P, Oettrich K, Schwab I, Langmandel U, Neumeier D. 1989 Immunocytochemical localization of the sex hormone-binding globulin in a human hepatoma cell line. Acta Endocrinol (Copenh). 121:791-6. 13. Knowles BB, Howe CC, Aden DP. 1980 Human hepatocellular carcinoma cell lines secrete the major plasma proteins, hepatitis B surface antigen. Science. 209:497-9. 14. Otten MH, Mol JA, Visser TJ. 1983 Sulfation preceeding deiodination of iodothyronines in rat hepatocytes. Science. 221:81-3. 15. Stockigt JR, Topliss DJ, Barlow JW, White EL, Hurley DM, Taft P. 1981 Familial euthyroid thyroxine excess: an appropriate response to abnormal thyroxine binding associated with albumin. J Clin Endocrinol Metab. 53:353-9. 16. Stockigt JR, White EL, Petrou S, Taft P. 1979 The course of untreated mild T3 toxicosis. J Aust. 2:6-7. 17. Chomczynski P, Sacchi N. 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 162:156-9. J, Fritsch EF, Maniatis T. 1989 Molecular cloning-a 18. Sambrook laboratory manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor Laboratory. 19. Hammond GL, Smith CL, Goping IS, et al. 1987 Primary structure of human corticosteroid binding globulin, deduced from hepatic and pulmonary cDNAs exhibits, homology with serine protease inhibitors. Proc Nat1 Acad Sci USA. 84:5153-7. 20. Sarne DH, Refetoff S, Rosenfield RL, Farriaux JP. 1988 Sex hormone-binding globulin in the diagnosis of peripheral tissue resistance to thyroid hormone: the value of changes after short term triiodothyronine administration. J Clin Endocrinol Metab. 66:7406. 21. Hammond GL, Underhill DA, Rykse HM, Smith CL. 1989 The human sex hormone-binding globulin gene contains exons for androgen-binding protein and two other testicular messenger RNAs. Mol Endocrinol. 3:1869-76. 22. Mercier-Bodard C, Baville F, Bideux G, Binart N, Chambraud B, Baulieu E-E. 1989 Regulation of SBP synthesis in human cancer cell lines by steroid and thyroid hormones. J Steroid Biochem. 34:199-204. 23. Rosner W, Aden DP, Khan MS. 1984 Hormonal influences on the secretion of steroid-binding proteins by a human hepatoma-derived cell line. J Clin Endocrinol Metab. 59:806-B. 24. Jump DB, Narayan P, Towle H, Oppenheimer JH. 1983 Rapid effects of triiodothyronine on hepatic gene expression: hybridization analysis of tissue-specific regulation of mRNA-S14. J Biol Chem. 259:2789-97.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 25 November 2015. at 02:05 For personal use only. No other uses without permission. . All rights reserved.
