0021-972X/92/7405-1062$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright 0 1992 by The Endocrine Society
Studies on the Biological Sulfate* STEPHEN W. SPAULDING, FAITH B. DAVIS, MEI-PING
Vol. 74, No. 5 f’rintrd in Lr,S.A.
Activity
of Triiodothyronine
TERRY J. SMITH, PATRICIA M. HINKLE, KUNG, AND JEROME A. ROTH
Department of Medicine, State University of New York School of Medicine, and Buffalo Veterans Administration Medical Center (S. W.S.), Buffalo, New York 14215; Department of Pharmacology and the Cancer Center, University of Rochester School of Medicine and Dentistry (P.M.H.), Rochester, New York 14642; Department of Medicine, Albany Medical College (T.J.S., F.B.D.), Albany, New York 12208; and the Department of Pharmacology, State University of New York School of Medicine (M.-P.K., J.A.R.), Buffalo, New York 14214
ABSTRACT. Hepatic microsomes and isolated hepatocytes in short term culture desulfate T, sulfate (T&SO,). We, therefore, wished to determine whether T,SO, could mimic the action of thyroid hormone in uitro. T:$O, had no thyromimetic effect on the activity of Ca’+-ATPase in human erythrocyte membranes at doses up to 10,000 times the maximally effective dose of Ts (lo-i0 mol/L). In GH,C, pituitary cells, T,SO, failed to displace [iz61]T3 from nuclear receptors in intact cells or soluble preparations. Thus, T3S01 was not directly thyromimetic in either an isolated human membrane system or a pituitary cell system in which nuclear receptor occupancy correlates with GH synthesis.
Thyroid hormones inhibit [3H]glycosaminoglycan synthesis by cultured human dermal fibroblasts, and TsSO, displayed about 0.5% the activity of T, at 72 h. Human fibroblasts contained roughly the same level of microsomal p-nitrophenyl sulfatase activity as that previously observed in hepatic microsomes. Propylthiouracil(50 pmol/L) did not affect the action of TaSO1, suggesting that deiodination was not important for this activity of T$O,. Thus, it appears T&SO4 has no intrinsic biological activity, but, under certain circumstances, may be reactivated by desulfation. (J Clin Endocrincl Metab 74: 1062-1067,1992)
L
brane calcium-stimulable magnesium-dependent ATP (Ca’+-ATPase; EC 3.6.1.3), is rapidly and directly stimulated by physiological concentrations of T4 and Ta (9, 10). The second system, the rat pituitary adenoma cell line GH&, displays saturable nuclear [lz51]T3 binding, and receptor saturation correlates closely with the induction of GH by thyroid hormone (11). The third system examined was the synthesis of glycosaminoglycan (GAG) by human skin fibroblasts in primary culture. Physiological concentrations of T8 or T4 inhibit the synthesis of hyaluronate, the most abundant GAG species elaborated by the cultured human fibroblast (12-15).
-TB IS conjugated with sulfate in the liver and possibly other mammalian tissues (l-3). Sulfation plays an important role in the metabolism of T:s, since TsSO, is inactivated by inner ring deiodination much more rapidly than unconjugated T3 (4). Although a large amount of TB appears to pass through this metabolic pathway in the rat, steady state levels of plasma T&SO4 are low unless agents such as propylthiouracil (PTU) are used to inhibit its deiodination (3). Crude microsomal fractions from several tissues of man and rat can desulfate TSSOl (5), although T3S04 previously had been assumed to be irreversibly inactivated (6-8). Isolated hepatocytes also desulfate T3S04 during short term incubations (5). To determine whether T,SO, displays thyromimetic actions in nonhepatic tissues, we examined its action in three different hormone-responsive in vitro models. The first system, isolated human red cell mem-
Materials
and Methods
Materials [“HIAcetic acid (2.8 Ci/mmol) and [1251]T3(2200 Ci/mmol) were obtained from New England Nuclear (Boston, MA). T,SO, was prepared from T,, using ClSOsH (5,16) and purified by chromatography through Sephadex LH-20 twice. All other chemicals used were the purest available from commercial sources. Thyroidectomized calf serum was obtained from Rockland Farms (Gilbertsville, PA). Heat-inactivated calf serum (Gibco, Grand Island, NY; 57 C; for 1 h) was resin stripped, as previously described (17).
Received September 10, 1991. Address requests for reprints to: Dr. Stephen W. Spaulding, Buffalo Veterans Administration Medical Center, 3495 Bailey Avenue, Buffalo, New York 14215. *This work was supported in part by funds from the Research Service of the V.A. (to S.W.S., T.J.S., and Paul J. Davis), NIH Grants DK-32847 (to P.M.H.) and NS-20530 and ES-04249 (to J.A.R.), and Cancer Center Core Research Grant CA-11098 (to P.M.H.). 1062
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
BIOLOGICAL Human
erythrocyte
membrane
Ca”‘-ATPase
ACTIVITY
activity
Cat+-ATPase activity in human erythrocyte membranes, prepared by hypotonic lysis, was determined in the presence or absence of 1 mM Ca’+. After exposure of the membranes to various concentrations of T&SO.,or TZIfor 60 min at 37 C, the quantity of inorganic phosphate generated from 1 mM ATP over a period of 90 min was assayed by the method of Fiske and Subbarow (18), as previously described (9). r’“I]T3
binding
to GH,C, cells
GH,C, cells were grown in Ham’s F-10 medium containing 15% horse serum and 2.5% fet.al calf serum under humidified 5% CO,. At least 24 h before an experiment, the medium was replaced with Ham’s F-10 medium that had been supplemented with 2.5% resin-stripped fetal bovine serum (11). To measure [lZ61]T3binding to intact cells, lo-“’ mol/L [‘“51] TB plus various concentrations of unlabeled iodothyronines were added to the medium for 2.5 h. In some experiments, unlabeled iodothyronines were added to the cultures for 4 days, and [iz51]T3was added for the final 2.5 h. Cultures were washed, and [rZ51]T3bound to the nuclear pellet and postnuclear supernatant was measured as previously described (11). Nonspecific nuclear binding was less than 5% of the total. Solubilized nuclear receptors were prepared by salt extraction of washed nuclei essentially as previously described (19). Soluble thyroid hormone receptors were incubated with 200 pmol/L [‘251]T3 and unlabeled hormone at 0 C for 2.5 h. Receptor-bound and free [‘*‘I]T:\ were separated on 0.5 mL Sephadex G-25 in plastic columns equilibrated in 0.25 M sucrose, 2omM Tris, l.lmM MgCla, 0.5% Triton X-100 pH 7.85 at 25 C. These columns were centrifuged for 5 min at 100 X g before use, and then 50-PL aliquots of the soluble receptor preparation were applied, and the columns were centrifuged again for 5 min at 2500 x g. The eluate contained receptorbound [‘251]T3, whereas free [‘“5I]T:1 was retained. Blanks were less than 5% of the added radioactivity. The nuclear radioreceptor assay is relatively insensitive because of the presence of 10% serum in the samples. Conditioned medium from dermal fibroblasts was tested for the ability to displace [12’I]T3 in a radioreceptor assay using the GH,C, cell system, as described above and in Fig. 2B, or in a double antibody RIA using a primary rabbit antiserum to T,, (Arnel, Parlin, NJ). This antiserum did not recognize T,SO, at concentrations up to 300 nmol/L, whereas Ts caused 50% displacement at 0.5 nmol/L. Medium that had been incubated with dermal fibroblasts for 72 h with or without T:, or T:,SO1 was diluted with control conditioned medium. Conditioned medium originally containing 300 nmol/L T&SO.,inhibited the binding of [‘*‘I]TS to nuclear receptors on GH4C1 cells by about 40%, equivalent to the inhibition caused by 2 nmol/L T:,. However, identical results were obtained when 300 nmol/L T3S04 were added to control conditioned medium and incubated for 3 days without exposure to cells. Medium containing T3S04 at concentrations below 100 nM had no effect on the immunoassay. It seems unlikely that significant conversion of T&SO, to Ts occurred in the fibroblast medium, because no T:+ was detected in the medium when the more sensitive RIA for T, was used. More likely, TSO, in the fibroblast medium was
OF T $0,
1063
causing nonspecific partial inhibition of GH,C, cell nuclear thyroid hormone receptors. Any Trr derived from intracellular metabolism of T.$O, was presumably either retained intracellularly or diluted to insignificant levels in the medium. Sulfatase
assay of crude microsomal
fraction
from fibroblasts
Fibroblasts were scraped in phosphate-buffered saline, sonicated, centrifuged at 100,000 X g for 1 h, and washed twice. The resulting microsomal pellet was assayed for its ability to remove the sulfate moiety from p-nitrophenyl sulfate after incubation at 37 C for 60 min in 200 FL 100 mmol/L potassium phosphate buffer, pH 8.0. Reactions were terminated by adding 0.8 mL ethanol and 0.5 mL 0.5 mol/L NaOH. After centrifugation, the optical absorption at 400 nm of the supernatant solution was determined as previously described (5). [‘H]GAG
synthesis
by human
fibroblasts
Normal human skin fibroblasts (American Type Culture Collection, Rockville, MD; CRL 1501) were grown to confluence on 60-mm plates at 37 C under humidified 5% C02, using Eagle’s Minimum Essential Medium plus 10% fetal calf serum and antibiotics (12). Cultures were then shifted to medium enriched with 10% thyroidectomized or resin-stripped calf serum to which T:,, T:,SO+ PTU (50 rmol/L), or equivalent concentrations of diluent (95% ethanol) were added. (Preliminary studies demonstrated that concentrations of PTU above 100 pmol/L inhibited basal [“HIGAG synthesis). After 48 h, the cultures were labeled with fresh medium containing [“HI acetic acid (2-10 &i/mL) for 24 h. Cell morphology, density, and total cell protein were not affected by any of the treatments. The incorporation of precursor into nondialyzable acid-soluble macromolecules was then determined, as previously described (12-15). Equivalent results were obtained using resin-stripped or thyroidectomized serum, and TSH has been shown not to affect GAG synthesis (Smith, T. J., unpublished observation). Approximately 80-90% of the [“HIGAG produced in this system is sensitive to Streptomyces hyaluronidase, defining the material as [“Hlhyaluronate (13). T, inhibits the synthesis of hyaluronate (12), and, to a lesser extent, that of dermatan sulfate, heparan sulfate, and chondroitin sulfate (13); the remainder of the GAGS synthesized by human skin fibroblasts are comprised of these sulfated molecules (12, 13). At least 1% serum must be present in the culture medium to maintain hyaluronate synthesis.
Results Human
erythrocyte
C’a’+-ATPase
activity
Red cell membrane Ca’+-ATPase increased with increasing doses of T:,; maximal stimulation occurred at a concentration of lo-“’ mol/L (Fig. l), with concentrations higher than this producing a progressively diminishing response, as previously reported (9, 10). T,SO,
failed to stimulate Ca’+-ATPase trations from lo-‘“-IO-’ reduce the stimulatory lo-“‘-lo-” mol/L.
over a range of concen-
mol/L. T,SO, (lo-“’ mol/L) did action of Tkl over the range of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
1064
SPAULDING
.lCE&M*1992
ET AL. intact Cell Nuclear
Receptors-2.5
h
0.06 -
0.04 -
0.02 -
O.OO-
J-7 .OOl
.Ol
.l
1
[lodothyronlne]
10
100
1000
nmoles/L
FIG. 1. Effects of T1 and T,SO, on human erythrocyte membrane Ca’+-ATPase activity. Ca*+-ATPase activity is expressed as micromoles of inorganic phosphate per mg/90 min. The change in enzyme activity after the addition of various concentrations of TJ (Cl), T&30, (O), or T, plus lo-“’ mol/L TklO, (A) is shown on the ordinate. Data shown are the results of three experiments, performed in duplicate (mean + SEM). T3 stimulated Ca?‘-ATPase activity (PC 0.001, by analysis of variance; F = 12.78), with a maximal effect at lo-‘” mol/L, while T$O, had no significant effect (F = 1.46). In the presence of 10-l” mol/L TSO,, the stimulation produced by T3 was reduced at all concentrations, but remained significant (P < 0.001; F = 21.47).
Binding of r2”I]T3
to GH,C, cells
Incubation of GH,C, cells with graded concentrations of unlabeled Ts for 2.5 h caused the expected displacement of [iz51]T8, with a half-maximal inhibitory concentration (IC& of 8 x lo-‘” mol/L (Fig. 2A). In contrast, T3S0, at concentrations up to 2.5 x 10e5 mol/L failed to reduce [‘251]T3 binding after 2.5 h. When the cells were incubated with TsSOI for 4 days before the addition of [lz51]T3, binding of the labeled T3 was only slightly reduced at the highest concentration of TaSOd (3 crmol/L). There was no change in total cellular protein at any of the iodothyronine concentrations shown. The ICso for T:, was less than 2 x 10-i” mol/L when cells were incubated with unlabeled hormone for 4 days before the addition of [lz51]Ta (Fig. 2B). The culture medium obtained from cells incubated with T&SO, for 4 days did not displace [‘?]T3 when tested on fresh cells (data not shown), indicating that T,SO, is stable under these incubation conditions. In neither experiment did TsS04 alter the amount of [1251]T3 associated with the nonnuclear fraction (data not shown); hence, it did not appear to alter [i2”I]T3 uptake or cytosol binding.
0
1 10 [IODOTHYRONINE]
1000 100 nmoles/L
FK:. 2. Effects of T, and T3S0, on nuclear [‘?]T3 binding in GH& cells and isolated nuclear receptors. A, Cultured GH,C, cells were incubated for 2.5 h with IO-” mol/L [“‘I]T3 and increasing concentrations of unlabeled T, (0) or T3SOI (0). Binding data are expressed as a percentage of the control value. B, Cultured cells were incubated for 4 days with unlabeled Ts or T3S0,, and [‘9]T3 was added during the last 2.5 h of incubation. C, Soluble salt-extracted nuclear T3 receptors were incubated with 2 x lo-‘” mol/L [‘“IIT, and unlabeled TI or T&30,. Receptor binding was measured as described in Materials and Methods. Data are normalized to the amount of cell protein in each fraction. The amount of [““I]T:, in the postnuclear supernatant fractions in whole cell experiments was not affected by the addition of unlabeled hormones. Values shown are the mean + range or SE of two or three dishes.
Pz51]T.J b’In d’zng to solubilized thyroid hormone receptors from GH,C, cells
The intact cell experiments suggested that either T:,S04 has negligible affinity for nuclear receptors or it cannot enter GH& cells. To distinguish between these mechanisms, the affinity of T$O, for soluble salt-extracted nuclear receptors was measured (Fig. 20. Unlabeled Ts displaced [1251]TRfrom solubilized receptors with an ICso of 10 nmol/L. Thyroid hormones bind to soluble receptors with a lower apparent affinity than to receptors in intact cells in most studies (191, as was found for GH4C, cells. TSO, again failed to displace [‘251]T3 at
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
BIOLOGICAL
concentrations up to 10 pmol/L, affinity for receptors.
indicating
ACTIVITY
no direct
Studies in human dermal fibroblasts Microsomal sulfatase activity. Sulfatase activity in the microsomal fraction from human fibroblasts appeared linear with increasing protein concentrations, as indicated by specific activities of approximately 1.3 nmol/min. mg microsomal protein when assayed at 0.2 mg/mL protein and approximately 1.2 nmol/min.mg with 0.4 mg/mL protein. This specific activity is similar to that found previously in human liver microsomes and about 10 times that found in rat brain microsomes (5). PHIGAG
synthesis
by human
fibroblasts
When human skin fibroblasts were cultured in 10% hypothyroid calf serum for 3 days (Fig. 3), they synthesized and accumulated about 30% more [3H]GAG than when T3 (100 nmol/L) was present in the same medium (12). We observed a maximal inhibitory effect of T3 at about lo-’ mol/L, confirming the findings of previous studies from this laboratory (12, 14) which show the dose-response curve to be steep, with a half-maximal inhibition at around 10-l’ mol/L. The response to T3 was equivalent whether cultures were incubated with
IOO-
Qo-
so-
70-
-t
-I .Ol
.I
I
IO
OF T,SO,
cultures were incubated with thyroidectomized or resinstripped calf serum (data not shown). When confluent fibroblasts were exposed to T3S04 for 3 days, [3H]GAG synthesis was inhibited in a dosedependent manner, with 3 x 10e7 mol/L T3S04 producing an inhibition that approached the maximal effect of Ts (Fig. 3), and half-maximal effects between 10-8-10-7 mol/ L T3S04. To determine whether the thyromimetic activity of T3S04 could be increased by inhibiting deiodination of this compound, we examined the effect of adding the deiodinase inhibitor PTU during fibroblast incubation. As shown in Table 1, 50 pM PTU failed to influence the inhibitory effect of T3S04 (10m7 mol/L) or T3 (10m7 mol/ L) on [3H]GAG synthesis. RIA of the medium from cultures incubated with T3S04 revealed no T3, suggesting that any T3 formed intracellularly was either retained in the cell layer or diluted to negligible concentrations in the medium. Discussion We have provided strong evidence that T3SOl is not directly thyromimetic, since neither the plasma membrane- nor nuclear T3 receptor-mediated system responds directly to the sulfated compound. The absence of an immediate response to T3S04 in these systems serves to confirm that there was little if any contamination of T3 in the T3S0, preparations used. Both the data obtained from the human erythrocyte membrane Ca”ATPase assay and those from the GH,C, cell assay indicate that any contamination of T3S04 by T3 must be substantially less than 0.0196, a level lower than that detectable by current high pressure liquid chromatographic methodology or the previously described T&SO* RIA (20,21). The inhibitory effect of 10-l’ mol/L T3S04 on T3-mediated stimulation in the Ca2+-ATPase system is similar to that previously described for tetraiodothyroacetic acid, another major thyronine metabolite (10, 22). GHJ!, ceils did not release any T&splacing activity into the medium after a 4-day incubation of GH& cells with 3 x lo-” mol/L T3S04, as detected by subsequent TABLE 1. Effect of PTU on the inhibition Tn and TSO,
of [3H]GAG synthesis by
100
Group
[lodothyronine]
1065
nmoledl
FIG. 3. Effect of Ta or T&O, on [3H]GAG synthesis in cultured human dermal fibroblasts. Confluent fibroblast cultures were shifted to thyroid hormone-depleted medium plus various concentrations of TS (0) and TaSO, (0) for 46 h. GAGS were labeled by shifting the cell cultures to fresh medium containing [SH]acetic acid (l-5 &/ml) as well as the respective iodothyronines for an additional 24 h. Samples were then processed as outlined in the text. Values shown are mean 2 range or SE, expressed as a percentage of the control value.
Mean dpm/plate + SEM ___174,900 -c 4,700 174,200 k 5,700 126,000 + 2,800” 115,700 f 4,200 137,100 + 1,300” 138,500 + 4,300”
% of
control
Control PTU 100 12 TB Ts + PTU 66 79 T&h 79 TSO, + PTU “P < 0.05 us. control and us. PTU alone (by analysis of variance; n = 5).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
1066
SPAULDING
bioassay with fresh GHICl cells. This confirms that T$SO, remains stable during the 4-day incubation in the medium, extending the previous report that T&SO4 is stable in human and rat serum for 18 h at 37 C (23). In human dermal fibroblast cultures, when T,SO, was present in the medium for 3 days, it appeared to have about 0.5% the activity of a maximally effective concentration of Tg. [The fraction of unbound T&O4 is probably less than that of the fractional unbound TS, since TSSOJ binds to rat and human serum binding proteins substantially more avidly than T3 does (3,23-25)]. This apparent thyromimetic activity of T&SO4 taken along with the demonstration of microsomal sulfatase activity in the fibroblast suggest that desulfation of T&SO, can occur in fibroblasts, as previously demonstrated in hepatocytes (5). Reports concerning the ability of human fibroblasts in culture to deiodinate Tq to Ts are conflicting (26-29), but we found no evidence that PTU (at 50 pmol/L) affected the ability of low7 mol/L T3S04 to inhibit GAG synthesis. It had been assumed that when Ts was sulfated, the compound was irreversibly inactivated (6-8), although the possibility that circulating T3S04 could serve as a reservoir that might permit regeneration of T3 in mammalian tissue had been discussed some 30 yr ago (1, 30). Recent data supporting the concept of reactivation of T&SO, have come from several directions. Intravenous administration of labeled T3S04 to rats rapidly raises the level of labeled Ttl in the blood (23); the authors of this study suggested that this appeared due to biliary excretion of the labeled TsS04, deconjugation of the TSSOd by colonic bacteria, and subsequent reabsorption of the intestinal Ts (23). However, even in animals whose intestinal contents had been decontaminated with antibiotics or whose biliary contents had been diverted, labeled Tg still appeared rapidly in the circulation (23). T&SO4 is also deconjugated in man (24). The authors of the latter study suggested that the deconjugation occurred in the renal tubule, since infusion of T3S04 apparently did not affect blood levels of TSH or T.?, but it is not clear the assays would have been sensitive enough to detect the level of increase in Ts that would be precdieted. These findings along with the previous finding of desulfation by hepatocytes in vitro (5) together with the current data all appear to suggest that direct in uiuo desulfation could provide an alternative pathway for the formation of TB from TSO,. It is possible that TB may undergo intracellular cycling between the sulfated and desulfated forms, as recently described for 4-methylumbelliferone in isolated rat hepatocytes (30). The results presented here indicate that T&SO4 has no intrinsic biological activity and that any thyromimetic action of T$SOl occurs after biotransformation. TS04 did inhibit GAG synthesis when present with human fibroblasts for 3 days, but it was considerably less potent
ET
JCE & M .199’2
AL.
than TB, findings most easily explained by postulating slow metabolism of T3S04 to T8. The inhibition produced by the iodothyronines was unaffected by PTU, suggesting that PTU-sensitive deiodinase activity was not involved in the response. The lack of effect of PTU is also consistent with the previous report that PTU fails to influence the binding of Ts to its nuclear receptor (32). We hypothesize that the apparent thyromimetic actions of T&SO4 in the fibroblast system are dependent upon its desulfation. Acknowledgments
We thank Mrs. Maureen Adolf for her assistance in performing the microsomal sulfatase assays, and Mr. Mike Cheavens for his assistance in the fibroblast studies. References 1. Roche J, Michel R, Closon J, Michel 0. Sur la sulfoconjugation hepatique de la 3,5,3’-triiodo+thyronine et la presence d’un ester sulfurique de cette hormone dans la bile et la plasma. Biochim Biophys Acta. 195$33:461-g. 2. Bollman JL, Flock EV. The role of the liver in the metabolism of P3’ thyroid hormones and analogs. In: Taylor W, ed. The biliary system. Oxford: Blackwell; 1965;345-65. 3. Rutgers M, Bonthuis F, Wouter W, Visser TJ. Accumulation of plasma triiodothyronine sulfate in rats treated with propylthiour acil. J Clin Invest. 1987;80:758-62. 4. Visser TJ, Mol JA, Otten MH. Rapid deiodination of triiodothyronine sulfate by rat liver microsomal fraction. Endocrinology. 1983;112:1547-9. 5. Kung M-P, Spaulding SW, Roth JA. Deeulfation of 3,5,3’-triiodothyronine sulfate by microsomes from human and rat tissues. Endocrinology. 1988;122:1195-200. 6. deHerder WW, Bonthuis F, Rutgers M, Otten MH, Haze&erg MP, Visser TJ. Effects of inhibition of type I iodothyronine deiodinase and phenol sulfotransferase on the biliary clearance of triiodothyronine in rats. Endocrinology. 1988,122:153-‘7. 7. Sekura RD, Sato K, Cahnmann HJ, Robbina J, Jakoby WB. Sulfate transfer to thyroid hormones and their analogs by hepatic aryl sulfotransferases. Endocrinology. 1981;108:454-6. 8. Mol JA, Visser TJ. Rapid and selective inner ring deiodination of thyroxine sulfate by rat liver deiodinase. Endocrinology. 1985:117:8-12. 9. Davis PJ, Bias SD. In vitro stimulation of human red blood cell Ca*+-ATPase by thyroid hormone. Biochem Biophys Res Commun. 1981;99:1073-80. 10. Davis FB, Cody V, Davis PJ, Borzynski LJ, Blas SD. Stimulation by thyroid hormone analogs of red blood cell Ca2+-ATPase activity in vitro: correlations between hormone structure and biological activity in a human cell system. J Biol Chem. 1983;258:1237317. 11. Kaji H, Hinkle PM. Epidermal growth factor decreases thyroid hormone receptors and attenuates thyroid hormone responses in GH&, cells. Endocrinology. 1987;120:537-43. 12. Smith TJ. Murata Y. Horwitz AL. Phil&on L. Refetoff S. Reaulation of glycosamindglycan synthesis by-thyroid hormone in vi&o. J Clin Invest. 1982;70:1066-73. 13. Smith TJ, Horwitz AL, Refetoff S. The effect of thyroid hormone on glycosaminoglycan accumulation in human skin fibroblasts. Endocrinology. 1981;108:2397-9. 14. Smith TJ. Dexamethasone regulation of glycosaminoglycan synthesis in human fibroblasts: similar effects of glucocorticosteroid and thyroid hormones. J Clin Invest. 1984;74:2157-63. 15. Murata Y. Refetoff S. Horwitz AL. Smith TJ. Hormonal reaulation of glycosaminoglycan accumulation in fibroblasts from patients with resistance to thyroid hormone. J Clin Endocrinol Metab.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
BIOLOGICAL
ACTIVITY
1983;57:1233-9. 16. Mol JA. Viseer TJH. Svnthesis and some nronerties of sulfate e&era and au&mates of iodothyroniies.’ Endocrinology. 1985;117:1-7. 17. Samuele HH, Stanley F, Cazanova J. Depletion of 3,3’,5-triiodothyronine and thyroxine in euthyroid calf serum for use in cell culture studies of the action of thyroid hormone. Endocrinology. 1979;106:80-5. 18. Fieke CH. Subbarow Y. The calorimetric determination of phoaphorue. JBiol Chem. 1925;66:375-400. 19. Samuele HH, Taai JS, Casanova J, Stanley F. Thyroid hormone action: in vitro characterization of eolubilized receptors from rat liver and cultured GH1 cells. J Clin Invest. 1974;54:853-65. 20. EeIkman Rooda SJ, Kaptein E, van Loon M, Viseer TJ. Development of an radioimmunoassay for triiodothyronine sulfate. J Immunoaeaay. 1988;9:126-35. 21. Rooda SJE, Kaptein E, Vieaer TJ. Serum triiodothyronine sulfate in man measured bv radioimmunoassav. J Clin Endocrinol Metab. 1989;69:552-6. 22. Davis PJ, Davie FB, Bias SD. Studies on the mechanism of thyroid hormone &mu&ion in vitro of human red cell Ca’+-ATPase activity. Life Sci. lQ82;3&675-82. 23. Rumra M, Heuadene FA, Bonthuis F, de Herder WW, Hazenberg MP, Viir TJ. Enterohepatic circulation of triiodothyronine (T3) in rate: importance of the microflora for the liberation and reabsorption of Ts from biliary conjugates. Endocrinology. lQ8%126:2822-30. -24. Lop&&i JS, Mizuno L, Nimalysuria A, Anderson KP, Spencer CA, Nicoloff JT. Characteristics of 3.5.3’-triiodothvronine sulfate metabolism in euthyroid man. J Cl& &docrinol tietab. 1991; 73:703-
OF T,SO,
1067
9. 25. Hays MJ, Hsu L. Equilibrium dialysis studies of plasma binding of thyroxine, triiodothyronine and their glucuronide and sulfate conjugates in human and cat plasma. Endocr Rec. 1988;14:51-8. 26. Ishikawa S-E, Saito T, Kuzuya T. The effect of glucose deprivation or 2-deoxy-D-glucose on the monodeiodination of thyroxine in human fibroblasts in culture. J Clin Bndocrinol Metab. 1985;61:252-7. 27. Refetoff S, Matalon R, Bigazzi M. Met&&m of L-thyroxine (T,) and L-triiodothyronine (Ta) by human fibroblaste in tieeue culture: evidence for cellular binding proteins and conversion of T, to T3. Endocrinology. 1972;91:934-47. 28. Refetoff S, DeGroot LJ, Benard B, DeWiid LT. Studies of a eibahip with apparent hereditary reei&ance to the intracellular action df thvroid hormone. Metabolism. 1972:21:723-56. 29. Kaplan MM, Pan C, Gordon PR, I& J-K, Gil&eat BA. Human epidermal keratinocytes in culture convert thyroxine to 3,5,3’triiodothyronine by type II iodothyronine deiodination: a normal endocrine function of the skin. J CIin Endocrinol Metab. 1988;66:815-21. 30. Roche J, Michel R, Closen J, Michel 0. Sur le metabolizme du sulfoconjugue de la 3,5,3’-triicdothyronine chez la rat. Biochim Biophys Acta. 1960;38:325-32. 31. Kauffman FC, Whittaker M, Anundi I, Thurman RG. FutiIe cycling of a sulfate conjugate by isolated hepatocytaa. Mol Pharmacol. 1991;39:414-20. 32. Eil C, Chestnut RY. The effects of radiographic contrast agents and other compounda on the nuclear binding of L-[‘~I]triiodothyronine in dieperaed human skin fibroblasta. J Clin Endocrinol Metab. 1985;60:548-52.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
Studies on the Biological Sulfate* STEPHEN W. SPAULDING, FAITH B. DAVIS, MEI-PING
Vol. 74, No. 5 f’rintrd in Lr,S.A.
Activity
of Triiodothyronine
TERRY J. SMITH, PATRICIA M. HINKLE, KUNG, AND JEROME A. ROTH
Department of Medicine, State University of New York School of Medicine, and Buffalo Veterans Administration Medical Center (S. W.S.), Buffalo, New York 14215; Department of Pharmacology and the Cancer Center, University of Rochester School of Medicine and Dentistry (P.M.H.), Rochester, New York 14642; Department of Medicine, Albany Medical College (T.J.S., F.B.D.), Albany, New York 12208; and the Department of Pharmacology, State University of New York School of Medicine (M.-P.K., J.A.R.), Buffalo, New York 14214
ABSTRACT. Hepatic microsomes and isolated hepatocytes in short term culture desulfate T, sulfate (T&SO,). We, therefore, wished to determine whether T,SO, could mimic the action of thyroid hormone in uitro. T:$O, had no thyromimetic effect on the activity of Ca’+-ATPase in human erythrocyte membranes at doses up to 10,000 times the maximally effective dose of Ts (lo-i0 mol/L). In GH,C, pituitary cells, T,SO, failed to displace [iz61]T3 from nuclear receptors in intact cells or soluble preparations. Thus, T3S01 was not directly thyromimetic in either an isolated human membrane system or a pituitary cell system in which nuclear receptor occupancy correlates with GH synthesis.
Thyroid hormones inhibit [3H]glycosaminoglycan synthesis by cultured human dermal fibroblasts, and TsSO, displayed about 0.5% the activity of T, at 72 h. Human fibroblasts contained roughly the same level of microsomal p-nitrophenyl sulfatase activity as that previously observed in hepatic microsomes. Propylthiouracil(50 pmol/L) did not affect the action of TaSO1, suggesting that deiodination was not important for this activity of T$O,. Thus, it appears T&SO4 has no intrinsic biological activity, but, under certain circumstances, may be reactivated by desulfation. (J Clin Endocrincl Metab 74: 1062-1067,1992)
L
brane calcium-stimulable magnesium-dependent ATP (Ca’+-ATPase; EC 3.6.1.3), is rapidly and directly stimulated by physiological concentrations of T4 and Ta (9, 10). The second system, the rat pituitary adenoma cell line GH&, displays saturable nuclear [lz51]T3 binding, and receptor saturation correlates closely with the induction of GH by thyroid hormone (11). The third system examined was the synthesis of glycosaminoglycan (GAG) by human skin fibroblasts in primary culture. Physiological concentrations of T8 or T4 inhibit the synthesis of hyaluronate, the most abundant GAG species elaborated by the cultured human fibroblast (12-15).
-TB IS conjugated with sulfate in the liver and possibly other mammalian tissues (l-3). Sulfation plays an important role in the metabolism of T:s, since TsSO, is inactivated by inner ring deiodination much more rapidly than unconjugated T3 (4). Although a large amount of TB appears to pass through this metabolic pathway in the rat, steady state levels of plasma T&SO4 are low unless agents such as propylthiouracil (PTU) are used to inhibit its deiodination (3). Crude microsomal fractions from several tissues of man and rat can desulfate TSSOl (5), although T3S04 previously had been assumed to be irreversibly inactivated (6-8). Isolated hepatocytes also desulfate T3S04 during short term incubations (5). To determine whether T,SO, displays thyromimetic actions in nonhepatic tissues, we examined its action in three different hormone-responsive in vitro models. The first system, isolated human red cell mem-
Materials
and Methods
Materials [“HIAcetic acid (2.8 Ci/mmol) and [1251]T3(2200 Ci/mmol) were obtained from New England Nuclear (Boston, MA). T,SO, was prepared from T,, using ClSOsH (5,16) and purified by chromatography through Sephadex LH-20 twice. All other chemicals used were the purest available from commercial sources. Thyroidectomized calf serum was obtained from Rockland Farms (Gilbertsville, PA). Heat-inactivated calf serum (Gibco, Grand Island, NY; 57 C; for 1 h) was resin stripped, as previously described (17).
Received September 10, 1991. Address requests for reprints to: Dr. Stephen W. Spaulding, Buffalo Veterans Administration Medical Center, 3495 Bailey Avenue, Buffalo, New York 14215. *This work was supported in part by funds from the Research Service of the V.A. (to S.W.S., T.J.S., and Paul J. Davis), NIH Grants DK-32847 (to P.M.H.) and NS-20530 and ES-04249 (to J.A.R.), and Cancer Center Core Research Grant CA-11098 (to P.M.H.). 1062
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
BIOLOGICAL Human
erythrocyte
membrane
Ca”‘-ATPase
ACTIVITY
activity
Cat+-ATPase activity in human erythrocyte membranes, prepared by hypotonic lysis, was determined in the presence or absence of 1 mM Ca’+. After exposure of the membranes to various concentrations of T&SO.,or TZIfor 60 min at 37 C, the quantity of inorganic phosphate generated from 1 mM ATP over a period of 90 min was assayed by the method of Fiske and Subbarow (18), as previously described (9). r’“I]T3
binding
to GH,C, cells
GH,C, cells were grown in Ham’s F-10 medium containing 15% horse serum and 2.5% fet.al calf serum under humidified 5% CO,. At least 24 h before an experiment, the medium was replaced with Ham’s F-10 medium that had been supplemented with 2.5% resin-stripped fetal bovine serum (11). To measure [lZ61]T3binding to intact cells, lo-“’ mol/L [‘“51] TB plus various concentrations of unlabeled iodothyronines were added to the medium for 2.5 h. In some experiments, unlabeled iodothyronines were added to the cultures for 4 days, and [iz51]T3was added for the final 2.5 h. Cultures were washed, and [rZ51]T3bound to the nuclear pellet and postnuclear supernatant was measured as previously described (11). Nonspecific nuclear binding was less than 5% of the total. Solubilized nuclear receptors were prepared by salt extraction of washed nuclei essentially as previously described (19). Soluble thyroid hormone receptors were incubated with 200 pmol/L [‘251]T3 and unlabeled hormone at 0 C for 2.5 h. Receptor-bound and free [‘*‘I]T:\ were separated on 0.5 mL Sephadex G-25 in plastic columns equilibrated in 0.25 M sucrose, 2omM Tris, l.lmM MgCla, 0.5% Triton X-100 pH 7.85 at 25 C. These columns were centrifuged for 5 min at 100 X g before use, and then 50-PL aliquots of the soluble receptor preparation were applied, and the columns were centrifuged again for 5 min at 2500 x g. The eluate contained receptorbound [‘251]T3, whereas free [‘“5I]T:1 was retained. Blanks were less than 5% of the added radioactivity. The nuclear radioreceptor assay is relatively insensitive because of the presence of 10% serum in the samples. Conditioned medium from dermal fibroblasts was tested for the ability to displace [12’I]T3 in a radioreceptor assay using the GH,C, cell system, as described above and in Fig. 2B, or in a double antibody RIA using a primary rabbit antiserum to T,, (Arnel, Parlin, NJ). This antiserum did not recognize T,SO, at concentrations up to 300 nmol/L, whereas Ts caused 50% displacement at 0.5 nmol/L. Medium that had been incubated with dermal fibroblasts for 72 h with or without T:, or T:,SO1 was diluted with control conditioned medium. Conditioned medium originally containing 300 nmol/L T&SO.,inhibited the binding of [‘*‘I]TS to nuclear receptors on GH4C1 cells by about 40%, equivalent to the inhibition caused by 2 nmol/L T:,. However, identical results were obtained when 300 nmol/L T3S04 were added to control conditioned medium and incubated for 3 days without exposure to cells. Medium containing T3S04 at concentrations below 100 nM had no effect on the immunoassay. It seems unlikely that significant conversion of T&SO, to Ts occurred in the fibroblast medium, because no T:+ was detected in the medium when the more sensitive RIA for T, was used. More likely, TSO, in the fibroblast medium was
OF T $0,
1063
causing nonspecific partial inhibition of GH,C, cell nuclear thyroid hormone receptors. Any Trr derived from intracellular metabolism of T.$O, was presumably either retained intracellularly or diluted to insignificant levels in the medium. Sulfatase
assay of crude microsomal
fraction
from fibroblasts
Fibroblasts were scraped in phosphate-buffered saline, sonicated, centrifuged at 100,000 X g for 1 h, and washed twice. The resulting microsomal pellet was assayed for its ability to remove the sulfate moiety from p-nitrophenyl sulfate after incubation at 37 C for 60 min in 200 FL 100 mmol/L potassium phosphate buffer, pH 8.0. Reactions were terminated by adding 0.8 mL ethanol and 0.5 mL 0.5 mol/L NaOH. After centrifugation, the optical absorption at 400 nm of the supernatant solution was determined as previously described (5). [‘H]GAG
synthesis
by human
fibroblasts
Normal human skin fibroblasts (American Type Culture Collection, Rockville, MD; CRL 1501) were grown to confluence on 60-mm plates at 37 C under humidified 5% C02, using Eagle’s Minimum Essential Medium plus 10% fetal calf serum and antibiotics (12). Cultures were then shifted to medium enriched with 10% thyroidectomized or resin-stripped calf serum to which T:,, T:,SO+ PTU (50 rmol/L), or equivalent concentrations of diluent (95% ethanol) were added. (Preliminary studies demonstrated that concentrations of PTU above 100 pmol/L inhibited basal [“HIGAG synthesis). After 48 h, the cultures were labeled with fresh medium containing [“HI acetic acid (2-10 &i/mL) for 24 h. Cell morphology, density, and total cell protein were not affected by any of the treatments. The incorporation of precursor into nondialyzable acid-soluble macromolecules was then determined, as previously described (12-15). Equivalent results were obtained using resin-stripped or thyroidectomized serum, and TSH has been shown not to affect GAG synthesis (Smith, T. J., unpublished observation). Approximately 80-90% of the [“HIGAG produced in this system is sensitive to Streptomyces hyaluronidase, defining the material as [“Hlhyaluronate (13). T, inhibits the synthesis of hyaluronate (12), and, to a lesser extent, that of dermatan sulfate, heparan sulfate, and chondroitin sulfate (13); the remainder of the GAGS synthesized by human skin fibroblasts are comprised of these sulfated molecules (12, 13). At least 1% serum must be present in the culture medium to maintain hyaluronate synthesis.
Results Human
erythrocyte
C’a’+-ATPase
activity
Red cell membrane Ca’+-ATPase increased with increasing doses of T:,; maximal stimulation occurred at a concentration of lo-“’ mol/L (Fig. l), with concentrations higher than this producing a progressively diminishing response, as previously reported (9, 10). T,SO,
failed to stimulate Ca’+-ATPase trations from lo-‘“-IO-’ reduce the stimulatory lo-“‘-lo-” mol/L.
over a range of concen-
mol/L. T,SO, (lo-“’ mol/L) did action of Tkl over the range of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
1064
SPAULDING
.lCE&M*1992
ET AL. intact Cell Nuclear
Receptors-2.5
h
0.06 -
0.04 -
0.02 -
O.OO-
J-7 .OOl
.Ol
.l
1
[lodothyronlne]
10
100
1000
nmoles/L
FIG. 1. Effects of T1 and T,SO, on human erythrocyte membrane Ca’+-ATPase activity. Ca*+-ATPase activity is expressed as micromoles of inorganic phosphate per mg/90 min. The change in enzyme activity after the addition of various concentrations of TJ (Cl), T&30, (O), or T, plus lo-“’ mol/L TklO, (A) is shown on the ordinate. Data shown are the results of three experiments, performed in duplicate (mean + SEM). T3 stimulated Ca?‘-ATPase activity (PC 0.001, by analysis of variance; F = 12.78), with a maximal effect at lo-‘” mol/L, while T$O, had no significant effect (F = 1.46). In the presence of 10-l” mol/L TSO,, the stimulation produced by T3 was reduced at all concentrations, but remained significant (P < 0.001; F = 21.47).
Binding of r2”I]T3
to GH,C, cells
Incubation of GH,C, cells with graded concentrations of unlabeled Ts for 2.5 h caused the expected displacement of [iz51]T8, with a half-maximal inhibitory concentration (IC& of 8 x lo-‘” mol/L (Fig. 2A). In contrast, T3S0, at concentrations up to 2.5 x 10e5 mol/L failed to reduce [‘251]T3 binding after 2.5 h. When the cells were incubated with TsSOI for 4 days before the addition of [lz51]T3, binding of the labeled T3 was only slightly reduced at the highest concentration of TaSOd (3 crmol/L). There was no change in total cellular protein at any of the iodothyronine concentrations shown. The ICso for T:, was less than 2 x 10-i” mol/L when cells were incubated with unlabeled hormone for 4 days before the addition of [lz51]Ta (Fig. 2B). The culture medium obtained from cells incubated with T&SO, for 4 days did not displace [‘?]T3 when tested on fresh cells (data not shown), indicating that T,SO, is stable under these incubation conditions. In neither experiment did TsS04 alter the amount of [1251]T3 associated with the nonnuclear fraction (data not shown); hence, it did not appear to alter [i2”I]T3 uptake or cytosol binding.
0
1 10 [IODOTHYRONINE]
1000 100 nmoles/L
FK:. 2. Effects of T, and T3S0, on nuclear [‘?]T3 binding in GH& cells and isolated nuclear receptors. A, Cultured GH,C, cells were incubated for 2.5 h with IO-” mol/L [“‘I]T3 and increasing concentrations of unlabeled T, (0) or T3SOI (0). Binding data are expressed as a percentage of the control value. B, Cultured cells were incubated for 4 days with unlabeled Ts or T3S0,, and [‘9]T3 was added during the last 2.5 h of incubation. C, Soluble salt-extracted nuclear T3 receptors were incubated with 2 x lo-‘” mol/L [‘“IIT, and unlabeled TI or T&30,. Receptor binding was measured as described in Materials and Methods. Data are normalized to the amount of cell protein in each fraction. The amount of [““I]T:, in the postnuclear supernatant fractions in whole cell experiments was not affected by the addition of unlabeled hormones. Values shown are the mean + range or SE of two or three dishes.
Pz51]T.J b’In d’zng to solubilized thyroid hormone receptors from GH,C, cells
The intact cell experiments suggested that either T:,S04 has negligible affinity for nuclear receptors or it cannot enter GH& cells. To distinguish between these mechanisms, the affinity of T$O, for soluble salt-extracted nuclear receptors was measured (Fig. 20. Unlabeled Ts displaced [1251]TRfrom solubilized receptors with an ICso of 10 nmol/L. Thyroid hormones bind to soluble receptors with a lower apparent affinity than to receptors in intact cells in most studies (191, as was found for GH4C, cells. TSO, again failed to displace [‘251]T3 at
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
BIOLOGICAL
concentrations up to 10 pmol/L, affinity for receptors.
indicating
ACTIVITY
no direct
Studies in human dermal fibroblasts Microsomal sulfatase activity. Sulfatase activity in the microsomal fraction from human fibroblasts appeared linear with increasing protein concentrations, as indicated by specific activities of approximately 1.3 nmol/min. mg microsomal protein when assayed at 0.2 mg/mL protein and approximately 1.2 nmol/min.mg with 0.4 mg/mL protein. This specific activity is similar to that found previously in human liver microsomes and about 10 times that found in rat brain microsomes (5). PHIGAG
synthesis
by human
fibroblasts
When human skin fibroblasts were cultured in 10% hypothyroid calf serum for 3 days (Fig. 3), they synthesized and accumulated about 30% more [3H]GAG than when T3 (100 nmol/L) was present in the same medium (12). We observed a maximal inhibitory effect of T3 at about lo-’ mol/L, confirming the findings of previous studies from this laboratory (12, 14) which show the dose-response curve to be steep, with a half-maximal inhibition at around 10-l’ mol/L. The response to T3 was equivalent whether cultures were incubated with
IOO-
Qo-
so-
70-
-t
-I .Ol
.I
I
IO
OF T,SO,
cultures were incubated with thyroidectomized or resinstripped calf serum (data not shown). When confluent fibroblasts were exposed to T3S04 for 3 days, [3H]GAG synthesis was inhibited in a dosedependent manner, with 3 x 10e7 mol/L T3S04 producing an inhibition that approached the maximal effect of Ts (Fig. 3), and half-maximal effects between 10-8-10-7 mol/ L T3S04. To determine whether the thyromimetic activity of T3S04 could be increased by inhibiting deiodination of this compound, we examined the effect of adding the deiodinase inhibitor PTU during fibroblast incubation. As shown in Table 1, 50 pM PTU failed to influence the inhibitory effect of T3S04 (10m7 mol/L) or T3 (10m7 mol/ L) on [3H]GAG synthesis. RIA of the medium from cultures incubated with T3S04 revealed no T3, suggesting that any T3 formed intracellularly was either retained in the cell layer or diluted to negligible concentrations in the medium. Discussion We have provided strong evidence that T3SOl is not directly thyromimetic, since neither the plasma membrane- nor nuclear T3 receptor-mediated system responds directly to the sulfated compound. The absence of an immediate response to T3S04 in these systems serves to confirm that there was little if any contamination of T3 in the T3S0, preparations used. Both the data obtained from the human erythrocyte membrane Ca”ATPase assay and those from the GH,C, cell assay indicate that any contamination of T3S04 by T3 must be substantially less than 0.0196, a level lower than that detectable by current high pressure liquid chromatographic methodology or the previously described T&SO* RIA (20,21). The inhibitory effect of 10-l’ mol/L T3S04 on T3-mediated stimulation in the Ca2+-ATPase system is similar to that previously described for tetraiodothyroacetic acid, another major thyronine metabolite (10, 22). GHJ!, ceils did not release any T&splacing activity into the medium after a 4-day incubation of GH& cells with 3 x lo-” mol/L T3S04, as detected by subsequent TABLE 1. Effect of PTU on the inhibition Tn and TSO,
of [3H]GAG synthesis by
100
Group
[lodothyronine]
1065
nmoledl
FIG. 3. Effect of Ta or T&O, on [3H]GAG synthesis in cultured human dermal fibroblasts. Confluent fibroblast cultures were shifted to thyroid hormone-depleted medium plus various concentrations of TS (0) and TaSO, (0) for 46 h. GAGS were labeled by shifting the cell cultures to fresh medium containing [SH]acetic acid (l-5 &/ml) as well as the respective iodothyronines for an additional 24 h. Samples were then processed as outlined in the text. Values shown are mean 2 range or SE, expressed as a percentage of the control value.
Mean dpm/plate + SEM ___174,900 -c 4,700 174,200 k 5,700 126,000 + 2,800” 115,700 f 4,200 137,100 + 1,300” 138,500 + 4,300”
% of
control
Control PTU 100 12 TB Ts + PTU 66 79 T&h 79 TSO, + PTU “P < 0.05 us. control and us. PTU alone (by analysis of variance; n = 5).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
1066
SPAULDING
bioassay with fresh GHICl cells. This confirms that T$SO, remains stable during the 4-day incubation in the medium, extending the previous report that T&SO4 is stable in human and rat serum for 18 h at 37 C (23). In human dermal fibroblast cultures, when T,SO, was present in the medium for 3 days, it appeared to have about 0.5% the activity of a maximally effective concentration of Tg. [The fraction of unbound T&O4 is probably less than that of the fractional unbound TS, since TSSOJ binds to rat and human serum binding proteins substantially more avidly than T3 does (3,23-25)]. This apparent thyromimetic activity of T&SO4 taken along with the demonstration of microsomal sulfatase activity in the fibroblast suggest that desulfation of T&SO, can occur in fibroblasts, as previously demonstrated in hepatocytes (5). Reports concerning the ability of human fibroblasts in culture to deiodinate Tq to Ts are conflicting (26-29), but we found no evidence that PTU (at 50 pmol/L) affected the ability of low7 mol/L T3S04 to inhibit GAG synthesis. It had been assumed that when Ts was sulfated, the compound was irreversibly inactivated (6-8), although the possibility that circulating T3S04 could serve as a reservoir that might permit regeneration of T3 in mammalian tissue had been discussed some 30 yr ago (1, 30). Recent data supporting the concept of reactivation of T&SO, have come from several directions. Intravenous administration of labeled T3S04 to rats rapidly raises the level of labeled Ttl in the blood (23); the authors of this study suggested that this appeared due to biliary excretion of the labeled TsS04, deconjugation of the TSSOd by colonic bacteria, and subsequent reabsorption of the intestinal Ts (23). However, even in animals whose intestinal contents had been decontaminated with antibiotics or whose biliary contents had been diverted, labeled Tg still appeared rapidly in the circulation (23). T&SO4 is also deconjugated in man (24). The authors of the latter study suggested that the deconjugation occurred in the renal tubule, since infusion of T3S04 apparently did not affect blood levels of TSH or T.?, but it is not clear the assays would have been sensitive enough to detect the level of increase in Ts that would be precdieted. These findings along with the previous finding of desulfation by hepatocytes in vitro (5) together with the current data all appear to suggest that direct in uiuo desulfation could provide an alternative pathway for the formation of TB from TSO,. It is possible that TB may undergo intracellular cycling between the sulfated and desulfated forms, as recently described for 4-methylumbelliferone in isolated rat hepatocytes (30). The results presented here indicate that T&SO4 has no intrinsic biological activity and that any thyromimetic action of T$SOl occurs after biotransformation. TS04 did inhibit GAG synthesis when present with human fibroblasts for 3 days, but it was considerably less potent
ET
JCE & M .199’2
AL.
than TB, findings most easily explained by postulating slow metabolism of T3S04 to T8. The inhibition produced by the iodothyronines was unaffected by PTU, suggesting that PTU-sensitive deiodinase activity was not involved in the response. The lack of effect of PTU is also consistent with the previous report that PTU fails to influence the binding of Ts to its nuclear receptor (32). We hypothesize that the apparent thyromimetic actions of T&SO4 in the fibroblast system are dependent upon its desulfation. Acknowledgments
We thank Mrs. Maureen Adolf for her assistance in performing the microsomal sulfatase assays, and Mr. Mike Cheavens for his assistance in the fibroblast studies. References 1. Roche J, Michel R, Closon J, Michel 0. Sur la sulfoconjugation hepatique de la 3,5,3’-triiodo+thyronine et la presence d’un ester sulfurique de cette hormone dans la bile et la plasma. Biochim Biophys Acta. 195$33:461-g. 2. Bollman JL, Flock EV. The role of the liver in the metabolism of P3’ thyroid hormones and analogs. In: Taylor W, ed. The biliary system. Oxford: Blackwell; 1965;345-65. 3. Rutgers M, Bonthuis F, Wouter W, Visser TJ. Accumulation of plasma triiodothyronine sulfate in rats treated with propylthiour acil. J Clin Invest. 1987;80:758-62. 4. Visser TJ, Mol JA, Otten MH. Rapid deiodination of triiodothyronine sulfate by rat liver microsomal fraction. Endocrinology. 1983;112:1547-9. 5. Kung M-P, Spaulding SW, Roth JA. Deeulfation of 3,5,3’-triiodothyronine sulfate by microsomes from human and rat tissues. Endocrinology. 1988;122:1195-200. 6. deHerder WW, Bonthuis F, Rutgers M, Otten MH, Haze&erg MP, Visser TJ. Effects of inhibition of type I iodothyronine deiodinase and phenol sulfotransferase on the biliary clearance of triiodothyronine in rats. Endocrinology. 1988,122:153-‘7. 7. Sekura RD, Sato K, Cahnmann HJ, Robbina J, Jakoby WB. Sulfate transfer to thyroid hormones and their analogs by hepatic aryl sulfotransferases. Endocrinology. 1981;108:454-6. 8. Mol JA, Visser TJ. Rapid and selective inner ring deiodination of thyroxine sulfate by rat liver deiodinase. Endocrinology. 1985:117:8-12. 9. Davis PJ, Bias SD. In vitro stimulation of human red blood cell Ca*+-ATPase by thyroid hormone. Biochem Biophys Res Commun. 1981;99:1073-80. 10. Davis FB, Cody V, Davis PJ, Borzynski LJ, Blas SD. Stimulation by thyroid hormone analogs of red blood cell Ca2+-ATPase activity in vitro: correlations between hormone structure and biological activity in a human cell system. J Biol Chem. 1983;258:1237317. 11. Kaji H, Hinkle PM. Epidermal growth factor decreases thyroid hormone receptors and attenuates thyroid hormone responses in GH&, cells. Endocrinology. 1987;120:537-43. 12. Smith TJ. Murata Y. Horwitz AL. Phil&on L. Refetoff S. Reaulation of glycosamindglycan synthesis by-thyroid hormone in vi&o. J Clin Invest. 1982;70:1066-73. 13. Smith TJ, Horwitz AL, Refetoff S. The effect of thyroid hormone on glycosaminoglycan accumulation in human skin fibroblasts. Endocrinology. 1981;108:2397-9. 14. Smith TJ. Dexamethasone regulation of glycosaminoglycan synthesis in human fibroblasts: similar effects of glucocorticosteroid and thyroid hormones. J Clin Invest. 1984;74:2157-63. 15. Murata Y. Refetoff S. Horwitz AL. Smith TJ. Hormonal reaulation of glycosaminoglycan accumulation in fibroblasts from patients with resistance to thyroid hormone. J Clin Endocrinol Metab.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
BIOLOGICAL
ACTIVITY
1983;57:1233-9. 16. Mol JA. Viseer TJH. Svnthesis and some nronerties of sulfate e&era and au&mates of iodothyroniies.’ Endocrinology. 1985;117:1-7. 17. Samuele HH, Stanley F, Cazanova J. Depletion of 3,3’,5-triiodothyronine and thyroxine in euthyroid calf serum for use in cell culture studies of the action of thyroid hormone. Endocrinology. 1979;106:80-5. 18. Fieke CH. Subbarow Y. The calorimetric determination of phoaphorue. JBiol Chem. 1925;66:375-400. 19. Samuele HH, Taai JS, Casanova J, Stanley F. Thyroid hormone action: in vitro characterization of eolubilized receptors from rat liver and cultured GH1 cells. J Clin Invest. 1974;54:853-65. 20. EeIkman Rooda SJ, Kaptein E, van Loon M, Viseer TJ. Development of an radioimmunoassay for triiodothyronine sulfate. J Immunoaeaay. 1988;9:126-35. 21. Rooda SJE, Kaptein E, Vieaer TJ. Serum triiodothyronine sulfate in man measured bv radioimmunoassav. J Clin Endocrinol Metab. 1989;69:552-6. 22. Davis PJ, Davie FB, Bias SD. Studies on the mechanism of thyroid hormone &mu&ion in vitro of human red cell Ca’+-ATPase activity. Life Sci. lQ82;3&675-82. 23. Rumra M, Heuadene FA, Bonthuis F, de Herder WW, Hazenberg MP, Viir TJ. Enterohepatic circulation of triiodothyronine (T3) in rate: importance of the microflora for the liberation and reabsorption of Ts from biliary conjugates. Endocrinology. lQ8%126:2822-30. -24. Lop&&i JS, Mizuno L, Nimalysuria A, Anderson KP, Spencer CA, Nicoloff JT. Characteristics of 3.5.3’-triiodothvronine sulfate metabolism in euthyroid man. J Cl& &docrinol tietab. 1991; 73:703-
OF T,SO,
1067
9. 25. Hays MJ, Hsu L. Equilibrium dialysis studies of plasma binding of thyroxine, triiodothyronine and their glucuronide and sulfate conjugates in human and cat plasma. Endocr Rec. 1988;14:51-8. 26. Ishikawa S-E, Saito T, Kuzuya T. The effect of glucose deprivation or 2-deoxy-D-glucose on the monodeiodination of thyroxine in human fibroblasts in culture. J Clin Bndocrinol Metab. 1985;61:252-7. 27. Refetoff S, Matalon R, Bigazzi M. Met&&m of L-thyroxine (T,) and L-triiodothyronine (Ta) by human fibroblaste in tieeue culture: evidence for cellular binding proteins and conversion of T, to T3. Endocrinology. 1972;91:934-47. 28. Refetoff S, DeGroot LJ, Benard B, DeWiid LT. Studies of a eibahip with apparent hereditary reei&ance to the intracellular action df thvroid hormone. Metabolism. 1972:21:723-56. 29. Kaplan MM, Pan C, Gordon PR, I& J-K, Gil&eat BA. Human epidermal keratinocytes in culture convert thyroxine to 3,5,3’triiodothyronine by type II iodothyronine deiodination: a normal endocrine function of the skin. J CIin Endocrinol Metab. 1988;66:815-21. 30. Roche J, Michel R, Closen J, Michel 0. Sur le metabolizme du sulfoconjugue de la 3,5,3’-triicdothyronine chez la rat. Biochim Biophys Acta. 1960;38:325-32. 31. Kauffman FC, Whittaker M, Anundi I, Thurman RG. FutiIe cycling of a sulfate conjugate by isolated hepatocytaa. Mol Pharmacol. 1991;39:414-20. 32. Eil C, Chestnut RY. The effects of radiographic contrast agents and other compounda on the nuclear binding of L-[‘~I]triiodothyronine in dieperaed human skin fibroblasta. J Clin Endocrinol Metab. 1985;60:548-52.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 00:17 For personal use only. No other uses without permission. . All rights reserved.
