- Email: [email protected]
Cytochemical localization of hydrogen peroxide in isolated thyroid follicles
JOURNAL OF ULTRASTRUCTURE RESEARCH 74, 105-115 (1981)
Cytochemical Localization of Hydrogen Peroxide in Isolated Thyroid Follicles U. BJORKMAN, R.
E K H O L M , 1 AND
J.-F.
DENEF 2
Department of Anatomy, University of Gbteborg, S-400 33 G6teborg 33, Sweden Received November 3, 1980 The localization of H202 generation was studied in isolated rat thyroid follicles with a cytochemical technique. Reaction product (cerium perhydroxide) was regularly found on the apical plasma membrane in open follicles incubated with NADH or NADPH. Without these substrates no reaction product was formed. In the presence of NAD(P)H the apical reaction was time and temperature dependent, it was inhibited by prefixation of the follicles, by catalase, by anaerobic incubation conditions, by parachloromercuribenzenesulfonate, and by lowering the pH to 6.5; it was not inhibited by KCN. The observations are interpreted to show that H202 was generated on the apical surface of the follicle cells by NAD(P)H oxidase in the apical plasma membrane.
The formation of thyroid hormones comprises binding of iodine to tyrosyl residues in the thyroglobulin molecule (iodination) and coupling of iodinated tyrosyl residues to the iodothyronine hormones. The iodination involves oxidation of iodide, which is concentrated in the thyroid from the circulation. This oxidation is catalyzed by thyroperoxidase in a reaction in which hydrogen peroxide acts as electron acceptor. Mechanisms by which H2Oz may be generated in the thyroid have attracted great interest and different systems capable of forming hydrogen peroxide have been described. The specific enzymes involved in these systems are NADPH-cytochrome c reductase, NADH-eytochrome b5 reductase, monoamine oxidase, and xanthine oxidase (for a review see DeGroot and Niepomniszcze, 1977). However, the same H2Oz-producing enzyme systems have been demonstrated in other organs and it is not known which of these systems--if any--is involved in the iodination reaction. Moreover, these enzyme systems have generally
been studied in poorly defined fractions of the thyroid and their true subcellular location has not been shown. Knowledge about the generation site of the hydrogen peroxide involved in iodination would contribute to get a final answer to the question of the site of thyroglobulin iodination. For H~O2--being a highly reactive substance and also easily decomposed by catalase--is probably formed close to the site at which it is required in the iodination reaction. Autoradiographic studies in the electron microscope strongly indicate that the site of iodination of thyroglobulin is the apical (luminal) surface of the thyroid epithelium (Ekholm and Wollman, 1975). In the present cytochemical study on H202 localization in the thyroid gland the interest was therefore focused on the apical cell surface. MATERIALS AND METHODS Follicles from Sprague-Dawley rat thyroids were isolated by a combination of collagenase digestion and mechanical disintegration, as described in a previous paper (Denef et al., 1980). With this technique 7080% of the follicles were open while the remaining follicles were closed and filled with colloid. After isolation, the follicle preparations were kept for 30-60 min at 37°C under continuous gassing with O~-CO2 (95% 02, 5% CO2) in Tyrode solution supplemented
1 To whom requests for reprints should be addressed. z Present address: Laboratory of Histology, Louvain Medical School, U.C.L., B-1200 Brussels, Belgium. 105
0022-5320/81/010105-11 $02.00/0 Copyright O 1981 by Academic Press, Inc. All rights of reproduction in any form reserved.
106
BJORKMAN, EKHOLM, AND DENEF
with amino acids according to Eagle (1959) and containing 0.5% albumin and 2/zg/ml DNAse. For cytochemical demonstration of hydrogen peroxide, the cerium technique introduced by Briggs et al. (1975) was used. The principle in this technique is that cerium ions, in the presence of hydrogen peroxide, form an electron-dense precipitate presumably consisting of cerium perhydroxide. The incubation of the follicle samples was generally performed in a medium consisting of the salts of Tyrode solution except bicarbonate and phosphate, buffered with 0.01 M Tris-maleate, pH 7.5, supplemented with amino acids (according to Eagle, 1959) and 0.5% albumin, and containing 10 mM aminotriazole and 1 mM CeC13. Before incubation the follicle preparations were washed and preincubated for 10 min at 37°C in the same solution but without CeCIl. (In some experiments the amino acid and albumin-supplemented salt solution was replaced by 0.1 M Tris-maleate buffer, pH 7.5, with 7.5% sucrose as used by Briggs et al. (1975)). The incubations were performed at 37°C under constant gassing with O~; the incubation time was generally 1 hr but 20 min and 2 hr were also tested. Each tube contained follicles derived from two to three thyroid lobes in a volume of 3.6 ml. Follicle samples were incubated without substrate or with 0.8 mM NADH, 0.8 mM NADPH (+1 mM Mn2+), or 2.0 mM xanthine. In some incubations 1.0 mM KCN or 0.5 mM parachloromercurihenzenesulfonate (PCMBS) was included in the incubation medium together with NADH. Catalase at a concentration of 0.05 or 0. I% was added to the preincubation and incubation media in some experiments with NADH; in these incubations aminotriazole was omitted since aminotriazole inhibits catalase. Some incubations were performed with NADH at pH 6.5. Follicle preparations were also prefixed for 10 rain in 2% glutaraldehyde or 2% formaldehyde buffered with 0.05% sodium cacodylate, pH 7.3, before incubation with NADH. Some specimens were incubated with NADH at 0°C under gassing with oxygen and at 37°C under gassing with nitrogen. At the end of the incubation the samples were centrifuged at 50 g for 5 min. The pellets were fixed in 3% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.3, for 1 hr. After prefixation the pellets were resuspended in 0.1 M cacodylate, pH 6.0, and kept at 4°C for 1 hr in order to dissolve cerium hydroxide possibly formed during incubation (cf. Briggs et al., 1975). After centrifugation and wash with 0.1 M cacodylate, pH 7.3, the specimens were postfixed in 1% osmium tetroxide in 0.1 M cacodylate, pH 7.3, for 1 hr. The samples were then dehydrated in ethanol and embedded in Epon as previously described (Denef et al., 1980). Sections were stained with uranyl acetate and lead citrate and examined in a Philips 300 electron microscope.
RESULTS
The pattern of distribution of precipitate was principally the same whether the follicle preparations were incubated with CeC13 in the modified, amino acid-supplemented Tyrode solution or in sucrose buffered with Tris-maleate as used by Briggs et al. (1975). However, since the structural preservation of the follicles was superior in the modified Tyrode, this medium was generally used. Qualitatively the distribution of precipitate was similar after incubation for 20 rain and 1 hr. However, the amount of precipitate was larger after 1 hr incubation and the report below accounts for the observations at this incubation time. Open Follicles Incubation without substrate. After incubation without substrate, precipitate was found only in the intercellular spaces (Fig. 1). The amount of precipitate at this site was varying but it was present only in narrow portions of the intercellular spaces. No reaction product was seen on the apical or basal cell surfaces and there was no intracellular precipitate. Incubation with substrate. After incubation in media containing either NADH or NADPH, reaction product was regularly found on the apical (luminal) surface of the follicle cells (Fig. 2). The amount of apical reaction product was similar in incubations with NADH and NADPH but considerable variations were observed between specimens and between follicles in the same section. The precipitate formed either a continuous layer of rather even thickness coating both the microvilli and the plasma membrane between the microvilli (Fig. 3) or, more often, the layer of reaction product was discontinuous and of varying thickness (Fig. 4). In all cases the reaction product was closely associated with the plasma membrane. The only site except the apical cell surface where precipitate was observed after incubation with NAD(P)H was the in-
HYDROGEN PEROXIDE IN THYROID FOLLICLES
107
Fro. 1. Part of an open thyroid follicle incubated without substrate for 1 hr. There is no reaction product on the apical cell surface but precipitate is seen in the intercellular spaces, x 20 000. t e r c e l l u l a r s p a c e s b u t , like t h e s i t u a t i o n aft e r i n c u b a t i o n w i t h o u t s u b s t r a t e , this d e position of precipitate was varying. There w a s no c o r r e l a t i o n b e t w e e n t h e a m o u n t o f apical reaction product and the amount of
intercellular precipitate. A regular feature was, however, that the precipitate was confined to n a r r o w p o r t i o n s o f t h e i n t e r c e l l u l a r s p a c e s . P r e c i p i t a t e w a s n o t f o u n d in a n y other locations than those described. Thus,
108
B J O R K M A N , E K H O L M , AND D E N E F
the basal cell surface did not exhibit any deposits; intracellular reaction product was not observed after 1 hr incubation with NAD(P)H. Incubation with xanthine as substrate resulted in the same distribution of precipitate as did incubation without substrate: precipitate, in varying amounts, was only found in intercellular spaces. Incubation with N A D H + inhibitors. When KCN was present in the incubation medium together with NADH the distribution pattern of precipitate was the same as in the absence of KCN, i.e., reaction product was regularly present on the apical cell surface and, in varying amounts, in the intercellular spaces. The amount of reaction product on the apical cell surface appeared to be not less than in the absence of KCN. Presence of PCMBS in the NADH-supplemented medium caused a drastic reduction of the apical reaction product (Fig. 5). In general only small patches of reaction product were found on the apical cell surface and in many follicles the apical surface was quite bare. Intercellular deposits were present to the same extent as without PCMBS. There were no basal or intracellular precipitates. When catalase in concentrations of 0.05 or 0.1% was added to NADH-supplemented media precipitates were sometimes formed in the media, probably by interaction between catalase and cerium. In these cases, no evaluation of the deposition of reaction product was made. In incubations with catalase where no such precipitation could be detected the amount of apical reaction product was clearly reduced and even absent (Fig. 6). There were no basal or intraceUular precipitates but intercellular deposits were seen to the same extent as without catalase. Incubations with NADH at 0°C, with
NADH under anaerobic conditions (Fig. 7), and with NADH at pH 6.5 resulted in similar pattern of precipitate distribution: Practically no apical reaction product was found while intercellular deposits were present as in the other incubations with NADH. There were no basal or intracellular precipitates. Incubation o f fixed follicles. Prefixation of the follicle specimens for 10 min in either glutaraldehyde or formaldehyde resulted in a pronounced reduction of reaction product on the apical cell surface in incubations with NADH (Fig. 8). At most, small spots of reaction product were seen along the apical plasma membrane. As in the incubations reported above no precipitate was found on the basal cell surface or intracellularly, but varying amounts in the intercellular spaces.
Closed Follicles Closed follicles, their lumen filled with colloid, were present in all incubations. Irrespective of the incubation conditions, i.e., also in the presence of NADH or NADPH, these follicles never exhibited any reaction product on the apical cell surface (Fig. 9). Like open follicles, they also never showed any deposits intracellularly or on the basal cell surface. Varying amounts of precipitate were present in the intercellular spaces but never in, or apically to, the tight junctions. DISCUSSION
Reaction product was regularly found on the apical surface of the follicle cells in open follicles after incubation in the presence of NADH or NADPH. No apical reaction product was formed without these substrates or when xanthine was tested as substrate. In incubations with NAD(P)H the amount of apical reaction product formed was time and temperature depen-
FIG. 2. Part o f an open thyroid follicle incubated with N A D H for I hr. Reaction product is seen on the apical (luminal) cell surface and small patches o f precipitate are present in narrow portions of the intercellular spaces. There is no precipitate on the basal cell surface or intracellularly, x 14 000.
HYDROGEN PEROXIDE IN THYROID FOLLICLES
109
] l0
BJ()RKMAN, EKHOLM, AND DENEF
H Y D R O G E N P E R O X I D E IN T H Y R O I D F O L L I C L E S
lll
FIG. 3. Part of an open follicle incubated with N A D H for 1 hr. T h e luminal surface of the cells is coated with an almost continuous layer of reaction product. × 21 000. Fro. 4. Part of an open follicle incubated with N A D P H for 1 hr. Reaction product is seen on the luminal surface of the cells. Precipitate is also p r e s e n t in intercellular spaces. × 20 000. Fro. 5. Apical part of follicle cells in an open follicle incubated with N A D H and P C M B S for 1 hr. Only small p a t c h e s of reaction product are seen on the apical cell surface w h e r e a s the narrow portions of the intercellular spaces are filled with precipitate. × 13 000. Fie. 6. Apical part of follicle cells in an open follicle incubated with N A D H and 0.1% catalase for 1 hr. N o apical reaction product. Precipitate in the intercellular spaces. × 20 000.
112
BJORKMAN, EKHOLM, AND DENEF
H Y D R O G E N P E R O X I D E IN T H Y R O I D F O L L I C L E S
dent and was drastically reduced by prefixation of the follicles with aldehydes. Finally, no apical reaction occurred in anaerobic incubations with NAD(P)H. These observations indicate that the apical reaction product was formed as the result of an enzymatic reaction utilizing reduced pyridine nucleotides as substrate and requiring molecular oxygen. The specificity of the cerium reaction for HzO2 was demonstrated by Briggs e t al. (1975) by inhibition of the reaction with 0.01% catalase. In the present study catalase was used in concentrations of 0.05 and 0.1% since lower concentrations do not inhibit iodination in isolated follicles incubated with N A D H and iodide (unpublished observations). As pointed out, the incubations with catalase sometimes resulted in the formation of precipitate in the medium. Although the statement under Results that catalase reduced the amount of apical reaction product is based on observations in incubations without visible precipitate in the medium, some complexing of cerium and catalase cannot be excluded even in these cases. Such complexing would mean that the concentrations of both catalase and cerium were lower than indicated by the amounts of these substances added to the medium. This could imply that a reduced amount of apical reaction was due to reduced amount of cerium rather than an inhibitory effect of catalase. However, control experiments showed that the amount of apical reaction product after incubation with 1 mM CeC13 + catalase was invariably smaller than after incubation with 0.5 mM CeC13 without catalase which supports the inference that catalase did inhibit the for-
113
mation of apical reaction product. That the apical reaction product was the result of H202 formation is also indirectly borne out by the observed absence of apical reaction after incubation at pH 6.5; at this acidic pH, the specific product considered to result from the reaction between H202 and cerium ions (cerium perhydroxide) cannot be formed. In conclusion, we feel justified to assume that the reaction product on the apical cell surface was due to the presence of
H202. No effort was made in the present study to characterize the enzyme(s) involved in the H202 production resulting in the apical reaction product. However, the observation that the reaction was not inhibited by KCN is interesting with respect to the demonstration in the leukocyte plasma membrane of cyanide-insensitive N A D H oxidase (Briggs e t al., 1975) and NAD(P)H oxidase activities (Takanaka and O'Brien, 1975a). Furthermore, it has been reported (Takanaka and O'Brien, 1975b) that although cyanide inhibits the peroxidase activity of myeloperoxidase it does not affect the NAD(P)H oxidase activity of myeloperoxidase and the oxidase activity is also not inhibited by aminotriazole. This seems worthy of attention considering the suggestion by Klebanoff e t al. (1962) that thyroperoxidase may serve as its own H202 generating system in an aerobic oxidation of reduced pyridine nucleotides. Apart from the reaction product on the apical cell surface, only the intercellular spaces showed deposits of precipitate. These deposits, characterized by being restricted to narrow portions of the intercellular spaces, were present in all specimens
FIG. 7. Apical part of follicle cells in an open follicle incubated with N A D H under anaerobic conditions for 1 hr. No reaction product is seen on the apical cell surface but precipitate is present in the intercellular spaces. × 14 000. Fro. 8. Apical part of a follicle cell in an open follicle prefixed with glutaraldehyde for 10 rain and incubated with N A D H for 1 hr. Small spots of reaction product are seen on the luminal cell surface, z 36 000. Fro. 9. Apical part of follicle ceils in a closed follicle incubated with N A D H for 1 hr. Note the dense content of the follicle lumen. No apical reaction product is seen but the intercellular space contains precipitate. × 20 000.
114
BJORKMAN, EKHOLM, AND DENEF
but in varying amounts. The precipitate ed by observations on leukocytes (Briggs was formed in all types of experiment, i.e., et al., 1975) showing reaction product in in incubations with and without substrate, phagosomes within 20 rain of incubation (a at 0°C, in the presence of catalase and labeling that was not due to internalization PCMBS, under anaerobic conditions, at pH of plasma membrane deposits since it oc6.5, and after prefixation of the tissue. This curred also after formaldehyde prefixation complete independence of incubation con- of the cells). Against this background, the ditions shows that the precipitate was not observed absence of apical reaction in the result of H202 formation but a non- closed follicles incubated with NAD(P)H specific product. The nature of the precip- was probably due to the lack of NAD(P)H itate is not known. However, it is evidently on the apical cell surface. not cerium hydroxide, which could be exFrom the above reasoning it seems juspected to be formed during the incubation tified to assume that, under the present exat alkaline pH (cf. Briggs et al., 1975), since perimental conditions, H202 is generated it was found also in incubations at pH 6.5. on the apical cell surface by NAD(P)H oxReaction product was found on the apical idase residing in the apical plasma memcell surface in open follicles incubated with brane. The present experimental conditions NAD(P)H but was never found on the basal imply that NAD(P)H was directly accessicell surface in spite of the fact that this sur- ble to the external surface of the apical face was as freely accessible to cerium ions plasma membrane. Moreover, the inhibiand NAD(P)H as the apical surface. This tion by PCMBS points to the possibility restriction of the reaction product to one that the oxidase is an ectoenzyme. On the cell surface indicates per se that this sur- other hand, under physiological conditions face has special properties for H202 gen- an oxidase in the apical plasma membrane eration. The strong inhibition of the apical must depend on intracellular NAD(P)H. reaction by PCMBS, a sulfhydryl reagent This dependence does not however invalithat does not penetrate the plasma mem- date the present observations. For exambrane (Sutherland et al., 1967), supports ple, a transmembrane oxidase may react the interpretation that H202 is formed on with the impermeable NAD(P)H on the inthe apical cell surface. side of the membrane and 02 on the outside The possibility that the apical reaction and be associated with coupled electron was due to H202 formed in the cytoplasm transport. Similar mechanisms have been and diffusing to the apical surface requires, suggested to operate in the erythrocyte first, that the intracellular site of peroxide membrane (L6w et al., 1979). generation be close to the apical cell surIt should be pointed out that the present face, otherwise reaction product should observations do not exclude that H202 is also occur on the basal cell surface. Sec- also formed in the cytoplasm of the follicle ond, since apical reaction was observed cells. In fact, after prolonged incubations only in the presence of NAD(P)H and no (2 hr) small amounts of precipitate were reaction product was present in the cyto- observed intracellularly both in the presplasm after 1 hr incubation, it requires that ence and absence of NAD(P)H and in living NAD(P)H but not cerium ions could enter as well as prefixed follicles (observations to the follicle cells. However, data from stud- be published). The present observations are important ies on leukocytes indicate that the plasma membrane is virtually impermeable to for establishing the site of thyroglobulin ioNAD(P)H (Takanaka and O'Brien, 1975a). dination. Of the four components involved On the other hand, cerium ions are not ex- in the iodination reaction, thyroglobulin cluded by the plasma membrane as indicat- and iodide pass across the apical cell sur-
HYDROGEN PEROXIDE IN THYROID FOLLICLES
face and accumulate in the follicle lumen. The third component, peroxidase, has been located in the apical plasma membrane cytochemically (Tice and Wollman, 1974). This study now presents evidence that the fourth component, hydrogen peroxide, is generated on the apical cell surface. Consequently, the apical cell surface is a meeting place of all four components, hence representing a potential site of iodination. The present results thus corroborate previous autoradiographic observations (Ekholm and Wollman, 1975) which were interpreted to show that iodination of thyroglobulin takes place on the apical surface of the follicle cells. This work was supported by grants from the Swedish Medical Research Council (B80-12X-537) and the National Institutes of Health (AM-18842). We are grateful to Mrs. Gunnel Bokhede for excellent technical assistance and to Mrs. Maria Ekmark and Mrs. Karin Nilsson for generous assistance in preparing the manuscript.
115
REFERENCES BRIGGS, R. T., DRATH, D. B., KARNOVSKY, M. L., AND KARNOVSKY,M. J. (1975) J. Cell Biol. 67, 566586. DEGROOT, L. J., AND NIEPOMNISZCZE, H. (1977) Metabolism 26, 665-718. DENEF, J.-F., BJORKMAN, U., AND EKHOLM, R. (1980) J. Ultrastruct. Res. 71, 185-202. EAGLE, H. (1959) Science 130, 432--437. EKHOLM, R., AND WOLLMAN, S. H. (1975) Endocrinology 97, 1432-1444. KLEBANOFF, S. J., YIN, C., AND KESSLER, D. (1962) Biochim. Biophys. Acta 58, 563-574. L6w, H., CRANE, F. L., GREB~N6, C., HALL, K., AND TALLY, M. (1979) in WALDHAUSL, W. K. (Ed.), Diabetes, Proceedings of the 10th Congress of the International Diabetes Federation, Vienna, Austria, Excerpta Med. Int. Congr. Ser. No. 500, pp. 209-213. SUTHERLAND,R. M., ROTHSTEIN,A., AND WEED, R. I. (1967) J. Cell Physiol. 69, 185-198. TAKANAKA, K., AND O'BRIEN, P. J. (1975a) Arch. Biochem. Biophys. 169, 436-442. TAKANAKA,K., AND O'BRIEN, P. J. (1975b) Biochem. Biophys. Res. Commun. 62, 966-971. TICE, L. W., AND WOLEMAN, S. H. (1974) Endocrinology 94, 1555-1567.
Cytochemical Localization of Hydrogen Peroxide in Isolated Thyroid Follicles U. BJORKMAN, R.
E K H O L M , 1 AND
J.-F.
DENEF 2
Department of Anatomy, University of Gbteborg, S-400 33 G6teborg 33, Sweden Received November 3, 1980 The localization of H202 generation was studied in isolated rat thyroid follicles with a cytochemical technique. Reaction product (cerium perhydroxide) was regularly found on the apical plasma membrane in open follicles incubated with NADH or NADPH. Without these substrates no reaction product was formed. In the presence of NAD(P)H the apical reaction was time and temperature dependent, it was inhibited by prefixation of the follicles, by catalase, by anaerobic incubation conditions, by parachloromercuribenzenesulfonate, and by lowering the pH to 6.5; it was not inhibited by KCN. The observations are interpreted to show that H202 was generated on the apical surface of the follicle cells by NAD(P)H oxidase in the apical plasma membrane.
The formation of thyroid hormones comprises binding of iodine to tyrosyl residues in the thyroglobulin molecule (iodination) and coupling of iodinated tyrosyl residues to the iodothyronine hormones. The iodination involves oxidation of iodide, which is concentrated in the thyroid from the circulation. This oxidation is catalyzed by thyroperoxidase in a reaction in which hydrogen peroxide acts as electron acceptor. Mechanisms by which H2Oz may be generated in the thyroid have attracted great interest and different systems capable of forming hydrogen peroxide have been described. The specific enzymes involved in these systems are NADPH-cytochrome c reductase, NADH-eytochrome b5 reductase, monoamine oxidase, and xanthine oxidase (for a review see DeGroot and Niepomniszcze, 1977). However, the same H2Oz-producing enzyme systems have been demonstrated in other organs and it is not known which of these systems--if any--is involved in the iodination reaction. Moreover, these enzyme systems have generally
been studied in poorly defined fractions of the thyroid and their true subcellular location has not been shown. Knowledge about the generation site of the hydrogen peroxide involved in iodination would contribute to get a final answer to the question of the site of thyroglobulin iodination. For H~O2--being a highly reactive substance and also easily decomposed by catalase--is probably formed close to the site at which it is required in the iodination reaction. Autoradiographic studies in the electron microscope strongly indicate that the site of iodination of thyroglobulin is the apical (luminal) surface of the thyroid epithelium (Ekholm and Wollman, 1975). In the present cytochemical study on H202 localization in the thyroid gland the interest was therefore focused on the apical cell surface. MATERIALS AND METHODS Follicles from Sprague-Dawley rat thyroids were isolated by a combination of collagenase digestion and mechanical disintegration, as described in a previous paper (Denef et al., 1980). With this technique 7080% of the follicles were open while the remaining follicles were closed and filled with colloid. After isolation, the follicle preparations were kept for 30-60 min at 37°C under continuous gassing with O~-CO2 (95% 02, 5% CO2) in Tyrode solution supplemented
1 To whom requests for reprints should be addressed. z Present address: Laboratory of Histology, Louvain Medical School, U.C.L., B-1200 Brussels, Belgium. 105
0022-5320/81/010105-11 $02.00/0 Copyright O 1981 by Academic Press, Inc. All rights of reproduction in any form reserved.
106
BJORKMAN, EKHOLM, AND DENEF
with amino acids according to Eagle (1959) and containing 0.5% albumin and 2/zg/ml DNAse. For cytochemical demonstration of hydrogen peroxide, the cerium technique introduced by Briggs et al. (1975) was used. The principle in this technique is that cerium ions, in the presence of hydrogen peroxide, form an electron-dense precipitate presumably consisting of cerium perhydroxide. The incubation of the follicle samples was generally performed in a medium consisting of the salts of Tyrode solution except bicarbonate and phosphate, buffered with 0.01 M Tris-maleate, pH 7.5, supplemented with amino acids (according to Eagle, 1959) and 0.5% albumin, and containing 10 mM aminotriazole and 1 mM CeC13. Before incubation the follicle preparations were washed and preincubated for 10 min at 37°C in the same solution but without CeCIl. (In some experiments the amino acid and albumin-supplemented salt solution was replaced by 0.1 M Tris-maleate buffer, pH 7.5, with 7.5% sucrose as used by Briggs et al. (1975)). The incubations were performed at 37°C under constant gassing with O~; the incubation time was generally 1 hr but 20 min and 2 hr were also tested. Each tube contained follicles derived from two to three thyroid lobes in a volume of 3.6 ml. Follicle samples were incubated without substrate or with 0.8 mM NADH, 0.8 mM NADPH (+1 mM Mn2+), or 2.0 mM xanthine. In some incubations 1.0 mM KCN or 0.5 mM parachloromercurihenzenesulfonate (PCMBS) was included in the incubation medium together with NADH. Catalase at a concentration of 0.05 or 0. I% was added to the preincubation and incubation media in some experiments with NADH; in these incubations aminotriazole was omitted since aminotriazole inhibits catalase. Some incubations were performed with NADH at pH 6.5. Follicle preparations were also prefixed for 10 rain in 2% glutaraldehyde or 2% formaldehyde buffered with 0.05% sodium cacodylate, pH 7.3, before incubation with NADH. Some specimens were incubated with NADH at 0°C under gassing with oxygen and at 37°C under gassing with nitrogen. At the end of the incubation the samples were centrifuged at 50 g for 5 min. The pellets were fixed in 3% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.3, for 1 hr. After prefixation the pellets were resuspended in 0.1 M cacodylate, pH 6.0, and kept at 4°C for 1 hr in order to dissolve cerium hydroxide possibly formed during incubation (cf. Briggs et al., 1975). After centrifugation and wash with 0.1 M cacodylate, pH 7.3, the specimens were postfixed in 1% osmium tetroxide in 0.1 M cacodylate, pH 7.3, for 1 hr. The samples were then dehydrated in ethanol and embedded in Epon as previously described (Denef et al., 1980). Sections were stained with uranyl acetate and lead citrate and examined in a Philips 300 electron microscope.
RESULTS
The pattern of distribution of precipitate was principally the same whether the follicle preparations were incubated with CeC13 in the modified, amino acid-supplemented Tyrode solution or in sucrose buffered with Tris-maleate as used by Briggs et al. (1975). However, since the structural preservation of the follicles was superior in the modified Tyrode, this medium was generally used. Qualitatively the distribution of precipitate was similar after incubation for 20 rain and 1 hr. However, the amount of precipitate was larger after 1 hr incubation and the report below accounts for the observations at this incubation time. Open Follicles Incubation without substrate. After incubation without substrate, precipitate was found only in the intercellular spaces (Fig. 1). The amount of precipitate at this site was varying but it was present only in narrow portions of the intercellular spaces. No reaction product was seen on the apical or basal cell surfaces and there was no intracellular precipitate. Incubation with substrate. After incubation in media containing either NADH or NADPH, reaction product was regularly found on the apical (luminal) surface of the follicle cells (Fig. 2). The amount of apical reaction product was similar in incubations with NADH and NADPH but considerable variations were observed between specimens and between follicles in the same section. The precipitate formed either a continuous layer of rather even thickness coating both the microvilli and the plasma membrane between the microvilli (Fig. 3) or, more often, the layer of reaction product was discontinuous and of varying thickness (Fig. 4). In all cases the reaction product was closely associated with the plasma membrane. The only site except the apical cell surface where precipitate was observed after incubation with NAD(P)H was the in-
HYDROGEN PEROXIDE IN THYROID FOLLICLES
107
Fro. 1. Part of an open thyroid follicle incubated without substrate for 1 hr. There is no reaction product on the apical cell surface but precipitate is seen in the intercellular spaces, x 20 000. t e r c e l l u l a r s p a c e s b u t , like t h e s i t u a t i o n aft e r i n c u b a t i o n w i t h o u t s u b s t r a t e , this d e position of precipitate was varying. There w a s no c o r r e l a t i o n b e t w e e n t h e a m o u n t o f apical reaction product and the amount of
intercellular precipitate. A regular feature was, however, that the precipitate was confined to n a r r o w p o r t i o n s o f t h e i n t e r c e l l u l a r s p a c e s . P r e c i p i t a t e w a s n o t f o u n d in a n y other locations than those described. Thus,
108
B J O R K M A N , E K H O L M , AND D E N E F
the basal cell surface did not exhibit any deposits; intracellular reaction product was not observed after 1 hr incubation with NAD(P)H. Incubation with xanthine as substrate resulted in the same distribution of precipitate as did incubation without substrate: precipitate, in varying amounts, was only found in intercellular spaces. Incubation with N A D H + inhibitors. When KCN was present in the incubation medium together with NADH the distribution pattern of precipitate was the same as in the absence of KCN, i.e., reaction product was regularly present on the apical cell surface and, in varying amounts, in the intercellular spaces. The amount of reaction product on the apical cell surface appeared to be not less than in the absence of KCN. Presence of PCMBS in the NADH-supplemented medium caused a drastic reduction of the apical reaction product (Fig. 5). In general only small patches of reaction product were found on the apical cell surface and in many follicles the apical surface was quite bare. Intercellular deposits were present to the same extent as without PCMBS. There were no basal or intracellular precipitates. When catalase in concentrations of 0.05 or 0.1% was added to NADH-supplemented media precipitates were sometimes formed in the media, probably by interaction between catalase and cerium. In these cases, no evaluation of the deposition of reaction product was made. In incubations with catalase where no such precipitation could be detected the amount of apical reaction product was clearly reduced and even absent (Fig. 6). There were no basal or intraceUular precipitates but intercellular deposits were seen to the same extent as without catalase. Incubations with NADH at 0°C, with
NADH under anaerobic conditions (Fig. 7), and with NADH at pH 6.5 resulted in similar pattern of precipitate distribution: Practically no apical reaction product was found while intercellular deposits were present as in the other incubations with NADH. There were no basal or intracellular precipitates. Incubation o f fixed follicles. Prefixation of the follicle specimens for 10 min in either glutaraldehyde or formaldehyde resulted in a pronounced reduction of reaction product on the apical cell surface in incubations with NADH (Fig. 8). At most, small spots of reaction product were seen along the apical plasma membrane. As in the incubations reported above no precipitate was found on the basal cell surface or intracellularly, but varying amounts in the intercellular spaces.
Closed Follicles Closed follicles, their lumen filled with colloid, were present in all incubations. Irrespective of the incubation conditions, i.e., also in the presence of NADH or NADPH, these follicles never exhibited any reaction product on the apical cell surface (Fig. 9). Like open follicles, they also never showed any deposits intracellularly or on the basal cell surface. Varying amounts of precipitate were present in the intercellular spaces but never in, or apically to, the tight junctions. DISCUSSION
Reaction product was regularly found on the apical surface of the follicle cells in open follicles after incubation in the presence of NADH or NADPH. No apical reaction product was formed without these substrates or when xanthine was tested as substrate. In incubations with NAD(P)H the amount of apical reaction product formed was time and temperature depen-
FIG. 2. Part o f an open thyroid follicle incubated with N A D H for I hr. Reaction product is seen on the apical (luminal) cell surface and small patches o f precipitate are present in narrow portions of the intercellular spaces. There is no precipitate on the basal cell surface or intracellularly, x 14 000.
HYDROGEN PEROXIDE IN THYROID FOLLICLES
109
] l0
BJ()RKMAN, EKHOLM, AND DENEF
H Y D R O G E N P E R O X I D E IN T H Y R O I D F O L L I C L E S
lll
FIG. 3. Part of an open follicle incubated with N A D H for 1 hr. T h e luminal surface of the cells is coated with an almost continuous layer of reaction product. × 21 000. Fro. 4. Part of an open follicle incubated with N A D P H for 1 hr. Reaction product is seen on the luminal surface of the cells. Precipitate is also p r e s e n t in intercellular spaces. × 20 000. Fro. 5. Apical part of follicle cells in an open follicle incubated with N A D H and P C M B S for 1 hr. Only small p a t c h e s of reaction product are seen on the apical cell surface w h e r e a s the narrow portions of the intercellular spaces are filled with precipitate. × 13 000. Fie. 6. Apical part of follicle cells in an open follicle incubated with N A D H and 0.1% catalase for 1 hr. N o apical reaction product. Precipitate in the intercellular spaces. × 20 000.
112
BJORKMAN, EKHOLM, AND DENEF
H Y D R O G E N P E R O X I D E IN T H Y R O I D F O L L I C L E S
dent and was drastically reduced by prefixation of the follicles with aldehydes. Finally, no apical reaction occurred in anaerobic incubations with NAD(P)H. These observations indicate that the apical reaction product was formed as the result of an enzymatic reaction utilizing reduced pyridine nucleotides as substrate and requiring molecular oxygen. The specificity of the cerium reaction for HzO2 was demonstrated by Briggs e t al. (1975) by inhibition of the reaction with 0.01% catalase. In the present study catalase was used in concentrations of 0.05 and 0.1% since lower concentrations do not inhibit iodination in isolated follicles incubated with N A D H and iodide (unpublished observations). As pointed out, the incubations with catalase sometimes resulted in the formation of precipitate in the medium. Although the statement under Results that catalase reduced the amount of apical reaction product is based on observations in incubations without visible precipitate in the medium, some complexing of cerium and catalase cannot be excluded even in these cases. Such complexing would mean that the concentrations of both catalase and cerium were lower than indicated by the amounts of these substances added to the medium. This could imply that a reduced amount of apical reaction was due to reduced amount of cerium rather than an inhibitory effect of catalase. However, control experiments showed that the amount of apical reaction product after incubation with 1 mM CeC13 + catalase was invariably smaller than after incubation with 0.5 mM CeC13 without catalase which supports the inference that catalase did inhibit the for-
113
mation of apical reaction product. That the apical reaction product was the result of H202 formation is also indirectly borne out by the observed absence of apical reaction after incubation at pH 6.5; at this acidic pH, the specific product considered to result from the reaction between H202 and cerium ions (cerium perhydroxide) cannot be formed. In conclusion, we feel justified to assume that the reaction product on the apical cell surface was due to the presence of
H202. No effort was made in the present study to characterize the enzyme(s) involved in the H202 production resulting in the apical reaction product. However, the observation that the reaction was not inhibited by KCN is interesting with respect to the demonstration in the leukocyte plasma membrane of cyanide-insensitive N A D H oxidase (Briggs e t al., 1975) and NAD(P)H oxidase activities (Takanaka and O'Brien, 1975a). Furthermore, it has been reported (Takanaka and O'Brien, 1975b) that although cyanide inhibits the peroxidase activity of myeloperoxidase it does not affect the NAD(P)H oxidase activity of myeloperoxidase and the oxidase activity is also not inhibited by aminotriazole. This seems worthy of attention considering the suggestion by Klebanoff e t al. (1962) that thyroperoxidase may serve as its own H202 generating system in an aerobic oxidation of reduced pyridine nucleotides. Apart from the reaction product on the apical cell surface, only the intercellular spaces showed deposits of precipitate. These deposits, characterized by being restricted to narrow portions of the intercellular spaces, were present in all specimens
FIG. 7. Apical part of follicle cells in an open follicle incubated with N A D H under anaerobic conditions for 1 hr. No reaction product is seen on the apical cell surface but precipitate is present in the intercellular spaces. × 14 000. Fro. 8. Apical part of a follicle cell in an open follicle prefixed with glutaraldehyde for 10 rain and incubated with N A D H for 1 hr. Small spots of reaction product are seen on the luminal cell surface, z 36 000. Fro. 9. Apical part of follicle ceils in a closed follicle incubated with N A D H for 1 hr. Note the dense content of the follicle lumen. No apical reaction product is seen but the intercellular space contains precipitate. × 20 000.
114
BJORKMAN, EKHOLM, AND DENEF
but in varying amounts. The precipitate ed by observations on leukocytes (Briggs was formed in all types of experiment, i.e., et al., 1975) showing reaction product in in incubations with and without substrate, phagosomes within 20 rain of incubation (a at 0°C, in the presence of catalase and labeling that was not due to internalization PCMBS, under anaerobic conditions, at pH of plasma membrane deposits since it oc6.5, and after prefixation of the tissue. This curred also after formaldehyde prefixation complete independence of incubation con- of the cells). Against this background, the ditions shows that the precipitate was not observed absence of apical reaction in the result of H202 formation but a non- closed follicles incubated with NAD(P)H specific product. The nature of the precip- was probably due to the lack of NAD(P)H itate is not known. However, it is evidently on the apical cell surface. not cerium hydroxide, which could be exFrom the above reasoning it seems juspected to be formed during the incubation tified to assume that, under the present exat alkaline pH (cf. Briggs et al., 1975), since perimental conditions, H202 is generated it was found also in incubations at pH 6.5. on the apical cell surface by NAD(P)H oxReaction product was found on the apical idase residing in the apical plasma memcell surface in open follicles incubated with brane. The present experimental conditions NAD(P)H but was never found on the basal imply that NAD(P)H was directly accessicell surface in spite of the fact that this sur- ble to the external surface of the apical face was as freely accessible to cerium ions plasma membrane. Moreover, the inhibiand NAD(P)H as the apical surface. This tion by PCMBS points to the possibility restriction of the reaction product to one that the oxidase is an ectoenzyme. On the cell surface indicates per se that this sur- other hand, under physiological conditions face has special properties for H202 gen- an oxidase in the apical plasma membrane eration. The strong inhibition of the apical must depend on intracellular NAD(P)H. reaction by PCMBS, a sulfhydryl reagent This dependence does not however invalithat does not penetrate the plasma mem- date the present observations. For exambrane (Sutherland et al., 1967), supports ple, a transmembrane oxidase may react the interpretation that H202 is formed on with the impermeable NAD(P)H on the inthe apical cell surface. side of the membrane and 02 on the outside The possibility that the apical reaction and be associated with coupled electron was due to H202 formed in the cytoplasm transport. Similar mechanisms have been and diffusing to the apical surface requires, suggested to operate in the erythrocyte first, that the intracellular site of peroxide membrane (L6w et al., 1979). generation be close to the apical cell surIt should be pointed out that the present face, otherwise reaction product should observations do not exclude that H202 is also occur on the basal cell surface. Sec- also formed in the cytoplasm of the follicle ond, since apical reaction was observed cells. In fact, after prolonged incubations only in the presence of NAD(P)H and no (2 hr) small amounts of precipitate were reaction product was present in the cyto- observed intracellularly both in the presplasm after 1 hr incubation, it requires that ence and absence of NAD(P)H and in living NAD(P)H but not cerium ions could enter as well as prefixed follicles (observations to the follicle cells. However, data from stud- be published). The present observations are important ies on leukocytes indicate that the plasma membrane is virtually impermeable to for establishing the site of thyroglobulin ioNAD(P)H (Takanaka and O'Brien, 1975a). dination. Of the four components involved On the other hand, cerium ions are not ex- in the iodination reaction, thyroglobulin cluded by the plasma membrane as indicat- and iodide pass across the apical cell sur-
HYDROGEN PEROXIDE IN THYROID FOLLICLES
face and accumulate in the follicle lumen. The third component, peroxidase, has been located in the apical plasma membrane cytochemically (Tice and Wollman, 1974). This study now presents evidence that the fourth component, hydrogen peroxide, is generated on the apical cell surface. Consequently, the apical cell surface is a meeting place of all four components, hence representing a potential site of iodination. The present results thus corroborate previous autoradiographic observations (Ekholm and Wollman, 1975) which were interpreted to show that iodination of thyroglobulin takes place on the apical surface of the follicle cells. This work was supported by grants from the Swedish Medical Research Council (B80-12X-537) and the National Institutes of Health (AM-18842). We are grateful to Mrs. Gunnel Bokhede for excellent technical assistance and to Mrs. Maria Ekmark and Mrs. Karin Nilsson for generous assistance in preparing the manuscript.
115
REFERENCES BRIGGS, R. T., DRATH, D. B., KARNOVSKY, M. L., AND KARNOVSKY,M. J. (1975) J. Cell Biol. 67, 566586. DEGROOT, L. J., AND NIEPOMNISZCZE, H. (1977) Metabolism 26, 665-718. DENEF, J.-F., BJORKMAN, U., AND EKHOLM, R. (1980) J. Ultrastruct. Res. 71, 185-202. EAGLE, H. (1959) Science 130, 432--437. EKHOLM, R., AND WOLLMAN, S. H. (1975) Endocrinology 97, 1432-1444. KLEBANOFF, S. J., YIN, C., AND KESSLER, D. (1962) Biochim. Biophys. Acta 58, 563-574. L6w, H., CRANE, F. L., GREB~N6, C., HALL, K., AND TALLY, M. (1979) in WALDHAUSL, W. K. (Ed.), Diabetes, Proceedings of the 10th Congress of the International Diabetes Federation, Vienna, Austria, Excerpta Med. Int. Congr. Ser. No. 500, pp. 209-213. SUTHERLAND,R. M., ROTHSTEIN,A., AND WEED, R. I. (1967) J. Cell Physiol. 69, 185-198. TAKANAKA, K., AND O'BRIEN, P. J. (1975a) Arch. Biochem. Biophys. 169, 436-442. TAKANAKA,K., AND O'BRIEN, P. J. (1975b) Biochem. Biophys. Res. Commun. 62, 966-971. TICE, L. W., AND WOLEMAN, S. H. (1974) Endocrinology 94, 1555-1567.