Ca /Nicotinamide Adenine Dinucleotide Phosphate- Dependent H2 ...

0013-7227/01/$03.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 86(9):4339 – 4343 Copyright © 2001 by The Endocrine Society

Ca2ⴙ/Nicotinamide Adenine Dinucleotide PhosphateDependent H2O2 Generation Is Inhibited by Iodide in Human Thyroids LUCIENE C. CARDOSO, DENISE C. L. MARTINS, MARCIA D. L. FIGUEIREDO, DORIS ROSENTHAL, MARIO VAISMAN, ALICE H. D. VIOLANTE, AND DENISE P. CARVALHO Instituto de Biofı´sica Carlos Chagas Filho (L.C.C., M.D.L.F., D.R., D.P.C.); and Servic¸o de Endocrinologia (D.C.L.M., M.V., A.H.D.V.), Hospital Universita´rio Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil 21949-900 A calcium and NAD(P)H-dependent H2O2-generating activity has been studied in paranodular thyroid tissues from four patients with cold thyroid nodules and from nine diffuse toxic goiters. H2O2 generation was detected both in the particulate (P 3,000 g) and in the microsomal (P 100,000 g) fractions of paranodular tissue surrounding cold thyroid nodules (PN), with the same biochemical properties described for NADPH oxidase found in porcine and human thyroids. In PN tissues, the particulate NADPH oxidase activity (224 ⴞ 38 nmol H2O2䡠hⴚ1䡠mgⴚ1 protein) was similar to that described for the porcine thyroid enzyme. However, no NADPH oxidase activity was detectable in the particulate fractions from eight diffuse toxic goiter patients treated with iodine before surgery; all

but one also received propylthiouracil or methimazole in the preoperative period. Thyroid cytochrome c reductase (diffuse toxic goiters ⴝ 438 ⴞ 104 nmol NADPⴙ䡠hⴚ1䡠mgⴚ1 protein; PN ⴝ 78 ⴞ 10 nmol NADPⴙ䡠hⴚ1䡠mgⴚ1 protein) and thyroperoxidase (diffuse toxic goiters ⴝ 621 ⴞ 179 U䡠gⴚ1 protein; PN ⴝ 232 ⴞ 121 U䡠gⴚ1 protein) activities were unaffected by iodide. Thus, the human NADPH oxidase seems to be inhibited by iodinated compounds in vivo and probably is an enzyme involved in the Wolff-Chaikoff effect. Our findings reinforce the hypothesis that thyroid NADPH oxidase is responsible for the production of H2O2 necessary for thyroid hormone biosynthesis. (J Clin Endocrinol Metab 86: 4339 – 4343, 2001)

H

In human thyroid tissues, the presence of a Ca2⫹/ NAD(P)H-dependent H2O2 generator similar to the porcine thyroid NADPH oxidase was only recently characterized (15). More recently the human and porcine enzyme cDNA have been cloned (16, 17). However, further molecular identification of this thyroid enzyme has not yet been achieved. Iodide, apart from inhibiting thyroid hormone biosynthesis, has also been known for a long time for its properties in inhibiting thyroid hormone release and reducing thyroid gland vascularity (18, 19). So patients with diffuse toxic goiters (DTG) are almost routinely prepared for thyroidectomy with iodine in the preoperative period. The aim of the present study was to further characterize the human NADPH oxidase activity using paranodular tissues from cold thyroid nodules and to evaluate whether the use of iodine before surgery in patients with DTG could impair NADPH oxidase activity, leading to diminished thyroid hormone biosynthesis. We have found a human thyroid NADPH oxidase activity, in paranodular tissues, that is similar to the enzyme already characterized in porcine glands. In DTG patients treated with iodine for 10 –15 d, all thyroid serum hormone levels were decreased at the day of surgery, compared with serum hormone levels before treatment. Furthermore, there has been a marked inhibition of the Ca2⫹/ NADPH-dependent H2O2 generation in the particulate (P 3,000 g) fraction of thyroid samples obtained from iodinetreated patients. Thus, our results reinforce the concept that thyroid NADPH oxidase inhibition might occur during the Wolff-Chaikoff effect.

2O2

GENERATION IS a limiting step of thyroid hormones biosynthesis, as previously shown in dog thyroid slices (1, 2). In fact, the H2O2 generator system has been localized at the apical cell surface in rat and pig open follicles (3, 4) and in intact rat follicles (5). Biochemical studies have confirmed that in porcine thyroid glands H2O2 is generated by an NADPH:O2 oxidoreductase, the so-called thyroid NADPH oxidase (6 – 8). This enzyme is a flavoprotein dependent on calcium for activity (7–9) and is activated by ATP (7). We have previously shown that in pig thyrocytes NADPH oxidase activity is induced by TSH, an effect that is mimicked by forskolin, and cAMP analogs (10). In dog thyrocytes, the expression of the H2O2 generator was also shown to be induced by TSH via the cAMP cascade (11). The H2O2 generator, NADPH oxidase, therefore seems to be a thyrocyte differentiation marker (10, 11), just as thyroglobulin and thyroperoxidase are. Inhibition of protein iodination and thus of thyroid hormone biosynthesis by iodide or iodinated compounds seems to be owing to the inhibition of hydrogen peroxide generation, both in dog and in human thyrocytes (12, 13). Hence, the fact that thyroid NADPH oxidase is irreversibly inhibited in vitro by iodide and 2-iodohexadecanal, a thyroid iodocompound that probably mediates the Wolff-Chaikoff effect, is another evidence that this is the enzyme responsible for H2O2 generation associated with thyroid hormonogenesis (14). Abbreviations: DTG, Diffuse toxic goiters; FAD, Flavin adenine dinucleotide; HRP, horseradish peroxidase; SSKI, saturated solution of iodine; TPO, thyroperoxidase.

4339

4340

The Journal of Clinical Endocrinology & Metabolism, September 2001, 86(9):4339 – 4343

Materials and Methods Materials NADPH and lyophilized horseradish peroxidase (HRP, grade 1) were purchased from Boehringer (Mannheim, Germany); Scopoletin, cytochrome c, and Flavin adenine dinucleotide (FAD) were obtained from Sigma (St. Louis, MO).

Patients We studied paranodular thyroid tissue samples obtained from four female patients with cold nodules and normal serum T4, T3, and TSH levels who did not receive any treatment before surgery. Eight patients with DTG (females 27– 42 yr old) programmed for thyroidectomy received saturated solution of iodine (SSKI, 5 drops three times per day) for 10 –15 d before surgery. The patients received either propylthiouracil (PTU) (500 –900 mg/d, n ⫽ 6) or methimazole (MMI) (25–50 mg/d, n ⫽ 2) until the day before surgery. Another patient with DTG (21 yr old) was allergic to antithyroid drugs and received only SSKI (5 drops three times per day) for 7 d and propanolol (80 mg/d) before surgery. The patients gave their informed consent, and the study has been approved by the Institutional Human Research Committee. Thyroid tissue samples were obtained at thyroidectomy and either freshly processed for NADPH oxidase and cytochrome c reductase measurements or stored at ⫺20 C for further thyroperoxidase (TPO) extraction and activity evaluation.

Serum hormone levels Blood samples were collected before starting treatment (basal), 7 d afterward, and on the day of surgery. Total T4, total T3, and TSH (third generation) were measured using a solid-phase, chemiluminescent enzyme immunoassay (IMMULITE). Free T4 and reverse T3 were measured by RIA. All kits were purchased from Diagnostic Products Corp. (Los Angeles, CA).

Thyroid samples processing For NADPH oxidase and cytochrome c reductase preparations, fresh human thyroid tissue samples (1 g) were cleaned from fibrous tissue or hemorrhagic areas, minced, and homogenized in sodium phosphate buffer, pH 7.2, containing 0.25 m sucrose, 0.5 mm dithiothreitol (DTT), and 1 mm EGTA, using an Ultra-Turrax (Staufen, Germany). The homogenate was filtered through cheesecloth. The particulate fraction was collected by centrifugation at 3,000 g for 15 min at 4 C and resuspended in 3 ml 50 mm sodium phosphate buffer, pH 7.2, containing 0.25 m sucrose and 2 mm MgCl2 (buffer A). The pellet was washed twice with 3 ml of buffer A and centrifuged at 3,000 g for 15 min at 4 C. The last pellet (P 3,000 g) was gently resuspended in 1 ml buffer A. The supernatant of the first centrifugation was centrifuged at 100,000 g for 1 h at 4 C. The pellet (microsomal fraction, P 100,000 g) was washed twice in 2 ml buffer A, and gently resuspended in 0.5 ml buffer A. Protein concentrations were measured by the method of Bradford (20), using BSA as standard. The particulate fractions (P 3,000 g and P 100,000 g) were incubated with 1 N NaOH (30 min, 20 C) to dissolve particulates before protein determination. For TPO preparation, thyroid tissue samples (1 g) were cleaned from fibrous tissue or hemorrhagic areas, minced, and homogenized in 50 mm Tris-HCl buffer pH 7.2, containing 1 mm potassium iodide, using an Ultra-Turrax (Ika). The homogenate was centrifuged at 100,000 g for 1 h at 4 C, and the pellet was resuspended in 2 ml digitonin (1% wt/vol). The mixture was incubated at 4 C for 24 h and then centrifuged at 100,000 g for 1 h at 4 C. The supernatant containing solubilized TPO was used for the iodide-oxidation assays, as previously described (21).

Ca2⫹ and NADPH-dependent H2O2 generating system: NADPH oxidase activity H2O2 formation was measured by incubating samples of the thyroid particulate fractions (P 3,000 g and P 100,000 g) at 30 C in 1 ml 170 mm sodium phosphate buffer, pH 7.4, containing 1 mm sodium azide, 1 mm EGTA, 1 ␮m FAD, 1.5 mm CaCl2, as previously described (15). The reaction was started by adding 0.2 mm NADPH; aliquots of 100 ␮l were

Cardoso et al. • Iodide Inhibits Thyroid NADPH Oxidase

collected at intervals up to 20 min and mixed with 10 ␮l 3 N HCl to stop the reaction and destroy the remaining NADPH. Calcium-dependent thyroid NADH oxidase activity was also evaluated in the particulates of one DTG tissue sample, by adding 0.2 mm NADH to initiate the reaction. Initial rates of H2O2 formation were determined from eight aliquots of each assay by following the decrease in 0.4 ␮m scopoletin fluorescence in the presence of HRP (0.5 ␮g/ml) in 200 mm phosphate buffer, pH 7.8, in a Hitachi (Tokyo, Japan) spectrofluorimeter (F 4000), as previously described (9). The excitation and emission wavelengths were 360 and 460 nm, respectively. All measurements were performed on at least three samples from each particulate preparation and expressed as nmoles H2O2䡠h. Specific activities were expressed per milligram protein (nmoles H2O2䡠h⫺1䡠mg⫺1 protein) in the thyroid P 3,000 g and P 100,000 g fractions. To evaluate the activation of the enzyme by phosphate, the concentration of phosphate in the reaction mixture was increased from 50 to 200 mm. To determine the Ca2⫹-dependence of H2O2 generation, parallel samples were assayed without Ca2⫹, in the presence of 1 mm EGTA.

Thyroid NADPH-cytochrome c reductase activity Aliquots of human thyroid particulate fractions (P 3,000 g and P 100,000 g) were incubated, at 30 C, in 1 ml 50 mm sodium phosphate buffer (pH 7.2) containing 1 mm sodium azide, 1.2 mm EGTA, and 0.1 mm cytochrome c. The reaction was started by adding 0.1 mm NADPH. The initial NADPH concentration was measured at 340 nm in a U-3300 (Hitachi) double-beam spectrophotometer, using a molar absorption coefficient of 6.2 ⫻ 103 M⫺1 cm⫺1. Aliquots of 100 ␮l were taken at intervals and mixed with 10 ␮l 2 mm dithiothreitol and 100 ␮l 2% SDS to stop the reaction. The initial rates of cytochrome c-dependent NADPH oxidation were determined from eight aliquots of each assay by following the decrease in NADPH fluorescence at pH 8.0 in a Hitachi spectrofluorimeter (F 4000), as previously described (10). The excitation and emission wavelengths were 340 and 453 nm, respectively. NADPHcytochrome c reductase activity was expressed as nanomoles of NADPH oxidized per hour and milligram of protein in the thyroid P 3,000 g and P 100,000 g fractions (nmoles NADP⫹䡠h⫺1䡠mg⫺1 protein).

Thyroperoxidase iodide-oxidation activity Thyroid peroxidase iodide-oxidation assays were performed using 12 mm iodine in 50 mm phosphate buffer (pH 7.4), and glucose-glucose oxidase as the hydrogen peroxide (H2O2) generating system, as previously described (21, 22). The increase in absorbency at 353 nm (⌬A353) was followed for 4 min on a U-3300 (Hitachi) double-beam spectrophotometer. The TPO activity was estimated from the ⌬A353/min determined from the linear portion of the reaction curve. One unit of iodide oxidation activity is defined as ⌬A353/min (U) ⫽ 1.0, and activity was related to the protein concentration in the enzyme preparation (U/g⫺1 protein).

Statistical analysis Statistical analysis of intergroup enzyme activities [DTG and paranodular tissue surrounding cold nodules (PN)] was performed using the Mann-Whitney test. Phosphate effect on enzyme activity was evaluated by the Kruskal-Wallis ANOVA followed by the Dunn multiple comparison test. Serum hormone levels at different periods of iodine treatment were analyzed by ANOVA for repeated measures followed by the Bonferroni multiple comparison test. Results are expressed as mean ⫾ se.

Results Serum thyroid hormone levels in patients with DTG

Serum total and free T4, total T3, and rT3 levels in DTG patients receiving iodine for 7 d or 10 –15 d were compared with those found before iodine treatment. All serum hormonal levels were already significantly decreased after 7 d of iodine treatment, and persistently diminished until the


The Journal of Clinical Endocrinology & Metabolism, September 2001, 86(9):4339 – 4343 4341

TABLE 1. Serum thyroid hormone levels in patients with DTG treated with SSKI in the preoperative period

Basal 7d Surgery (10 –15 d)

Total T4 ␮g/dl (4 –13)a

Free T4 ng/dl (0.8 –2.0)a

T3 ng/dl (70 –210)a

rT3 ng/dl (0.09 – 0.35)a

14.5 ⫾ 1.7 11.0 ⫾ 1.6b 10.1 ⫾ 1.6d

2.7 ⫾ 0.5 2.0 ⫾ 0.4b 1.7 ⫾ 0.3d

226.1 ⫾ 25.6 164.2 ⫾ 19.8c 149.2 ⫾ 18.8d

0.63 ⫾ 0.15 0.40 ⫾ 0.10b 0.34 ⫾ 0.11c

Results are expressed as mean ⫾ SEM. a Normal range; b P ⬍ 0.05; c P ⬍ 0.01;

d

P ⬍ 0.001.

FIG. 1. Thyroid NADPH oxidase activity. A Ca2⫹- and NADPHdependent H2O2 generation was measured in human thyroid samples obtained from four paranodular to cold nodule tissues (PN) and from eight iodine-treated patients with diffuse toxic goiter (iodine). H2O2 generation was measured in 170 mM sodium phosphate buffer, pH 7.4, containing 1 mM sodium azide, 1 mM EGTA, 1 ␮M FAD, and 1.5 mM CaCl2. The reaction was started by adding 0.2 mM NADPH; aliquots of 100 ␮l were collected at intervals up to 20 min and mixed with 10 ␮l 3 N HCl to stop the reaction. Initial rates of H2O2 formation were determined from eight aliquots of each assay by following the decrease in 0.4 ␮M scopoletin fluorescence in the presence of HRP (0.5 ␮g/ml) in 200 mM phosphate buffer, pH 7.8, in a Hitachi spectrofluorimeter (F 4000, excitation ⫽ 360 and emission ⫽ 460 nm). A, Particulate 3,000 g fraction (P 3,000 g). B, Microsomal fraction (P 100,000 g). Enzyme activity is expressed as mean of at least two measurements in each particulate preparation.

day of surgery (10 –15 d after beginning of SSKI treatment; Table 1). Calcium and NADPH-dependent H2O2 generation activities in human thyroid tissues

In the absence of Ca2⫹, no NADPH-dependent H2O2 generating activity was found in any of the particulate fractions studied. In our assays, NADPH oxidase activity significantly increased in the presence of high concentrations (200 mm) of phosphate anions (456 ⫾ 47 nmol H2O2䡠h⫺1/ml⫺1), compared with the activity found in 50 mm phosphate (208 ⫾ 35 H2O2䡠h⫺1/ml⫺1, P ⬍ 0.05), confirming findings previously reported for pig and human thyroids (15). Therefore, in all following experiments, rates of H2O2 formation were measured using human thyroid particulate fractions in 170 mm phosphate buffer pH 7.4. NADPH/Ca2⫹-dependent H2O2 generation activities in paranodular thyroid tissues were found in both the P 3,000 g and P 100,000 g fractions, in contrast to pig thyroids in which no detectable activity is found in the P 100,000 g fraction.

scribed for porcine thyroid particulate, also enriched in thyroid plasma membranes. On the other hand, H2O2 generation was either undetectable or low in the P 3,000 g obtained from DTG patients treated with iodine (Fig. 1A). However, there were no significant differences between the microsomal fraction (P 100,000 g) NADPH oxidase activities in paranodular (200 ⫾ 36 nmol H2O2䡠h⫺1䡠mg⫺1 protein) or DTG samples (235 ⫾ 41 nmol H2O2䡠h⫺1䡠mg⫺1 protein), independent of iodine treatment received before surgery (Fig. 1B). NADH oxidase activity was also undetectable in the P 3,000 g obtained from one DTG sample and represented 50% of the NADPH oxidase activity found in the P 100,000 g of the same DTG (NADH ⫽ 77; NADPH ⫽ 139 nmol H2O2䡠h⫺1䡠mg⫺1 protein), as previously described (15). In the DTG patient who received neither PTU nor MMI before surgery, NADPH oxidase activity was also undetectable in the P 3,000 g fraction and was within the normal range in the P 100,000 g fraction (100 nmol H2O2䡠h⫺1䡠mg⫺1 protein). Cytochrome c reductase activity

Cytochrome c reductase is another enzymatic system, which was proposed as capable of generating H2O2 in the thyroid gland. So we evaluated the effect of iodine treatment on this enzyme activity. We found no effect of iodine treatment on cytochrome c reductase activity, either in the P 3,000 g fraction (DTG ⫽ 438 ⫾ 104 nmol NADP⫹䡠h⫺1䡠mg⫺1 protein; PN ⫽ 78 ⫾ 10 nmol NADP⫹䡠h⫺1䡠mg⫺1 protein, n ⫽ 2) or in the P 100,000 g fraction (DTG ⫽ 515 ⫾ 49 nmol NADP⫹䡠 h⫺1䡠mg⫺1 protein; PN ⫽ 216 ⫾ 84 nmol NADP⫹䡠h⫺1䡠mg⫺1 protein, n ⫽ 2) (Fig. 2A and Fig. 2B). In fact, the cytochrome c reductase activity seems to be higher in thyroid tissues obtained from DTG patients than in PN tissues. Thyroperoxidase activity

Total TPO iodide-oxidation activities in DTG tissues were not significantly different from that in paranodular tissues (DTG ⫽ 621 ⫾ 179 U/g⫺1 protein; PN ⫽ 232 ⫾ 121 U/g⫺1 protein, Fig. 2C). Furthermore, there was no significant difference between DTG-TPO activities in both the thyroid P 3,000 g (618 ⫾ 288 U/g⫺1 protein) and P 100,000 g fractions (1328 ⫾ 673 U/g⫺1 protein), indicating that iodine treatment does not block TPO activity, at least irreversibly. In the DTG patient who received neither PTU nor MMI before surgery, TPO was also within the normal range (695 U/g⫺1 protein). Discussion

NADPH oxidase activity is inhibited by iodine treatment

In paranodular thyroid tissues, NAPDH oxidase H2O2generating levels found in the P 3,000 g fraction (224 ⫾ 38 nmol H2O2䡠h⫺1䡠mg⫺1 protein) were similar to those de-

Iodine organification and thyroid hormone biosynthesis are dependent on thyroperoxidase and on H2O2 as cofactor. The H2O2 supply has been demonstrated to be a limiting step for thyroid hormone biosynthesis (2). The cDNA for the

4342

The Journal of Clinical Endocrinology & Metabolism, September 2001, 86(9):4339 – 4343

FIG. 2. Thyroid cytochrome c reductase and thyroperoxidase activities. NADPH oxidation by cytochrome c reductase and TPO iodide oxidation activities were measured in human thyroid samples obtained from four paranodular to cold nodule tissues (PN) and from eight iodine-treated patients with diffuse toxic goiter (iodine). Cytochrome c reductase activity was measured in the presence of 1 mM sodium azide, 1.2 mM EGTA, and 0.1 mM cytochrome c. The reaction was started by adding 0.1 mM NADPH; aliquots of 100 ␮l were taken at intervals and mixed with 10 ␮l 2 mM dithiothreitol and 100 ␮l 2% SDS to stop the reaction. NADPH oxidation was determined from eight aliquots of each assay by following the decrease in NADPH fluorescence at pH 8.0 in a Hitachi spectrofluorimeter (F 4000, excitation ⫽ 340 and emission ⫽ 453 nm). A, Particulate 3,000 g fraction (P 3,000 g). B, Microsomal fraction (P 100,000 g). All measurements were performed on at least two samples from each particulate preparation. C, TPO activity was measured using 12 mM KI in 50 mM phosphate buffer (pH 7.4), and glucose-glucose oxidase as the hydrogen peroxide (H2O2) generating system. The increase in absorbance at 353 nm (⌬A353) was followed for 4 min on a U-3300 (Hitachi) double-beam spectrophotometer, and enzyme activity was estimated from the ⌬A353/min determined from the linear portion of the reaction curve. One unit of iodide oxidation activity is defined as ⌬A353/min (U) ⫽ 1.0. Results are expressed as mean ⫾ SE.

enzyme responsible for H2O2 generation in the thyroid gland has only recently been cloned in porcine and human thyroids (16, 17). Nevertheless, some biochemical properties of the thyroid H2O2-generating enzyme, NADPH oxidase, have already been defined in porcine and more recently in human thyroids (15). Leseney et al. (15) reported a Ca2⫹ and NAD(P)H-dependent H2O2-generating activity in human thyroids, but a very low NADPH oxidase activity in the P 3,000 g fraction was found in their study. In the present study, however, the human thyroid NADPH oxidase activity found in the P 3,000 g fraction of PN tissues was similar to that of the porcine thyroid enzyme, probably because patients with cold thyroid nodules did not receive any iodine treatment before surgery (9, 15, 23). Furthermore, in pig thyroids NADPH oxidase activity is predominantly found in the P 3,000 g fraction, whereas in human thyroid tissues the enzymatic activity was present both in the microsomal and P 3,000 g fractions, as previously reported (15). It has previously been demonstrated that thyroid NADPH oxidase


activity (10) and mRNA expression (16) are significantly increased by TSH in porcine thyrocytes. Thus, in thyroid samples obtained from DTG, in which the TSH intracellular pathway is hyperstimulated, an increased NADPH oxidase activity could be found. Thyroid autoregulation by iodine involves the inhibition of several steps of thyroid metabolism, such as iodide transport, hormone secretion, and adenylate cyclase activity (24, 25). However, other mechanisms have been proposed to explain iodine inhibition of its own organification. Previous studies have shown that high, unphysiological, iodide concentrations can inhibit thyroperoxidase in vitro (26). However, other studies have shown that both in canine and human thyrocytes, the Wolff-Chaikoff effect could have been caused, at least in part, by impaired H2O2 generation, although the authors did not evaluate thyroid NADPH oxidase activity, an ill-defined enzymatic system at the time (12, 13). Then it was reported that the porcine NADPH oxidase could be irreversibly inhibited in vitro by iodide and 2-iodohexadecanal, a thyroid iodocompound probably involved in the thyroid autoregulation by iodine (14). These findings indicated the participation of NADPH oxidase inhibition in the decreased iodine organification caused by iodide. Our present findings of undetectable NADPH oxidase activity in the P 3,000 g fraction from thyroid tissues exposed to high in vivo levels of iodide show that a decreased H2O2 generation might be important for the hormone biosynthesis inhibition promoted by iodine treatment in human thyroids. NADPH oxidase inhibition occurred despite the fact that PTU was administered to the patients during the period of iodine treatment, suggesting that iodide could act directly on the enzyme in vivo, as previously described in vitro (14) and/or that PTU was not sufficient to prevent lipid iodination completely. NADH oxidase activity was also inhibited in the P 3,000 g obtained from one patient treated with iodine, indicating that thyroid NADPH and NADH oxidase activities could correspond to a single enzyme responsible for Ca2⫹-dependent H2O2 generation in the thyroid, as previously suggested (15). NADPH oxidase activity is modulated by calcium; however, enzyme synthesis is dependent on the cAMP cascade (10, 16, 27). Excess of iodide decreases cAMP in thyroid cells, but the strong inhibitory effect of iodide on thyroid NADPH oxidase, which was found only in the P 3,000 g fraction of DTG, is probably not owing to only decreased enzyme synthesis because microsomal fraction NADPH oxidase activity was similar in DTG and PN tissues. Furthermore, TPO whose synthesis is also regulated by cAMP was not inhibited, at least irreversibly, either in the P 3,000 g or in the microsomal fractions. Besides, cytochrome c reductase, a putative thyroid H2O2-generating system, was unaffected by iodine treatment, which shows, together with the previous findings, that this enzyme is not inducible by TSH (10), that cytochrome c reductase is not involved in H2O2 generation linked to thyroid hormonogenesis. Furthermore, we presume no direct irreversible effect of PTU or MMI on thyroid NADPH oxidase in vivo because this enzyme activity was undetectable in the P 3,000 g from a goiter that received only iodine and propanolol before surgery. Moreover, NADPH oxidase ac-


The Journal of Clinical Endocrinology & Metabolism, September 2001, 86(9):4339 – 4343 4343

tivity was normal in the microsomal fraction obtained from the DTG patients treated with either PTU or MMI. Thus, we believe there is a direct iodide/iodocompound effect on the thyroid NADPH oxidase. It has recently been demonstrated that in rats there is a decrease in the sodium/iodide symporter 24 h after excess iodide, which could explain the escape from iodine administration (28). In our patients, serum hormone levels were significantly decreased 7 d after the start of iodine treatment and remained diminished until the time of surgery, indicating that there was no escape from the iodine effect even after 15 d of treatment. Thus, it is possible that NADPH oxidase inhibition at the site for thyroid hormonogenesis might contribute to the maintained lower thyroid hormone serum level when iodine is administered to patients with DTG. In conclusion, our present findings reinforce the hypothesis that thyroid NADPH oxidase is the enzyme responsible for the production of H2O2 necessary for thyroid hormone biosynthesis and shows unequivocally its role in thyroid autoregulation. Acknowledgments Received July 7, 2000. Accepted May 7, 2001. Address all correspondence and requests for reprints to: Denise Pires de Carvalho, Instituto de Biofı´sica Carlos Chagas Filho, CCS-Bloco GCidade Universita´ria, Ilha do Funda˜o, Rio de Janeiro, Brazil 21949-900, Brasil. E-mail: [email protected].

9. 10.

11. 12.

13. 14. 15. 16. 17. 18. 19. 20.

References 1. Ahn CS, Rosenberg IN 1970 Iodine metabolism in thyroid slices: effects of TSH, dibutyryl cyclic 3⬘,5⬘-AMP, NaF and prostaglandin E1. Endocrinology 86:396 – 405 2. Corvilain B, Van Sande J, Laurent E, Dumont JE 1991 The H2O2-generating system modulates protein iodination and the activity of the pentose phosphate pathway in dog thyroid. Endocrinology 128:779 –785 3. Bjo¨rkman U, Ekholm R, Denef JF 1981 Cytochemical localization of hydrogen peroxide in isolated thyroid follicles. J Ultrastruct Res 74:105–115 4. Bjo¨rkman U, Ekhlom R 1984 Generation of H2O2 in isolated porcine thyroid follicles. Endocrinology 115:392–398 5. Labato MA, Briggs RT 1985 Cytochemical localization of hydrogen peroxide generating sites in the rat thyroid gland. Tissue Cell 17:889 –900 6. Virion A, Michot JL, De`me D, Kaniewski J, Pommier J 1984 NADPHdependent H2O2 generation and peroxidase activity in thyroid particulate fraction. Mol Cell Endocrinol 36:95–105 7. Nakamura Y, Ogihara S, Ohtaki S 1987 Activation by ATP of calciumdependent NADPH-oxidase generating hydrogen peroxide in thyroid plasma membranes. J Biochem 102:1121–1132 8. Dupuy C, Kaniewski J, De`me D, Pommier J, Virion A 1989 NADPH-depen-

21. 22. 23. 24. 25. 26. 27.

28.

dent H2O2 generation catalyzed by thyroid plasma membranes. Studies with electron scavengers. Eur J Biochem 185:597– 603 De`me D, Virion A, Aı¨t-Hammou N, Pommier J 1985 NADPH-dependent generation of H2O2 in a thyroid particulate fraction requires Ca2⫹. FEBS Lett 186:107–110 Carvalho DP, Dupuy C, Gorin Y, et al. 1996 The Ca2⫹- and reduced nicotinamide adenine dinucleotide phosphate-dependent hydrogen peroxide generating system is induced by thyrotropin in porcine thyroid cells. Endocrinology 137:1007–1012 Raspe´ E, Dumont JE 1995 Tonic modulation of dog thyrocyte H2O2 generation and I⫺ uptake by thyrotropin through the cyclic adenosine 3⬘,5⬘-monophosphate cascade. Endocrinology 136:965–973 Corvilain B, Laurent E, Lecomte M, Van Sande J, Dumont JE 1994 Role of the cyclic adenosine 3⬘,5⬘-monophosphate and the phosphatidylinositolCa2⫹ cascades in mediating the effects of thyrotropin and iodide on hormone synthesis and secretion in human thyroid slices. J Clin Endocrinol Metab 79:152–159 Corvilain B, Van Sande J, Dumont J 1988 Inhibition by iodide of iodide binding to proteins: the “Wolff-Chaikoff” effect is caused by inhibition of H2O2 generation. Biochem Biophys Res Comm 154:1287–1292 Ohayon R, Boeynaems JM, Braekman JC, Van den Bergen H, Gorin Y, Virion A 1994 Inhibition of thyroid NADPH-oxidase by 2-iodohexadecanal in a cellfree system. Mol Cell Endocrinol 99:133–141 Leseney AM, De`me D, Dupuy C, et al. 1999 Biochemical characterization of a Ca2⫹/NAD(P)H-dependent H2O2 generator in human thyroid tissue. Biochimie (Paris) 81:373–380 Dupuy C, Ohayon R, Valent A, Noel-Hudson MS, De`me D, Virion A 1999 Purification of a novel flavoprotein involved in the thyroid NADPH oxidase. Cloning of the porcine and human cDNAs. J Biol Chem 274:37265–37269 De Deken X, Wang D, Many MC, et al. 2000 Cloning of two human thyroid cDNAs encoding new members of the NADPH oxidase family. J Biol Chem 275:23227–23233 Robuschi G, Manfredi A, Salvi M, et al. 1986 Effect of sodium ipodate and iodide on freeT4 and free T3 concentration in patients with Graves’ disease. J Endocrinol Invest 9:287–291 Roti E, Robuschi G, Manfredi A, et al. 1985 Comparative effects of sodium ipodate and iodide on serum thyroid hormone concentrations in patients with Graves’ disease. Clin Endocrinol (Oxf) 22:489 – 496 Bradford MM 1976 A rapid and sensitive method for the quantification of microgram quantities of proteins utilizing the protein-dye binding. Anal Biochem 72:248 –252 Carvalho DP, Rego KGM, Rosenthal D 1994 Thyroid peroxidase in dyshormonogenetic goiters with organification and thyroglobulin defects. Thyroid 4:421– 426 Moura EG, Rosenthal D, Carvalho-Guimara˜es DP 1989 Thyroid peroxidase activity in human nodular goiters. Braz J Med Biol Res 22:31–39 Dupuy C, De`me D, Kaniewski J, Pommier J, Virion A 1988 Ca2⫹ regulation of thyroid NADPH-dependent H2O2 generation. FEBS Lett 233:74 –78 Wolff J 1989 Excess iodide inhibits the thyroid by multiple mechanisms. Adv Exp Med Biol 261:211–244 Pisarev MA 1985 Thyroid autoregulation. J Endocrinol Invest 8:475– 484 Pommier J, De`me D, Nunez J 1973 Effect of iodide concentration on thyroxine synthesis catalysed by thyroid peroxidase. Eur J Biochem 37:406 – 414 Gorin Y, Ohayon R, Carvalho DP, et al. 1996 Solubilization and characterization of a thyroid Ca2⫹-dependent and NADPH-dependent K3Fe(CN)6 reductase. Relationship with the NADPH-dependent H2O2 generating system. Eur J Biochem 240:807– 814 Eng PHK, Cardona GR, Fang SL, et al. 1999 Escape from the acute WolffChaikoff effect is associated with a decrease in thyroid sodium/iodide symporter messenger ribonucleic acid and protein. Endocrinology 140: 3404 –3410