Defective Thyroglobulin Synthesis in an Experimental Rat Thyroid Tumor: Iodination and Thyroid Hormone Synthesis in Isolated Tumor Thyroglobulin FABRIZIO MONACO, SETTIMIO GRIMALDI, ROBERTO DOMINICI, AND JACOB ROBBINS* Centro della Tiroide do Ha Clinica Medica dell'Universita, Policlinico Umberto 1°, Rome, Italy, and *Clinical Endocrinology Branch, National Institute of Arthritis, Metabolism, and Digestive Diseases, National Institutes of Health, Bethesda, Maryland 20014 ABSTRACT. Purified rat tumor thyroglobulin from the experimental rat thyroid tumor, line 1-1C2, was studied for its thyroid hormone content after in vivo and in vitro iodination and compared with normal and desialylated normal rat thyroglobulin. Tumor thyroglobulin had a very low sialic acid and iodine content; after in vivo iodination it contained only small amounts of triiodothyronine (T3) and no detectable thyroxine (T4). After in vitro iodination with m I it showed a distribution of T3 and T4 very similar to that of normal and desialylated normal
thyroglobulin iodinated in vitro. In vitro iodination dissociated tumor and desialylated normal thyroglobulin to a greater extent than normal thyroglobulin. Tumor tissue, on the other hand, showed considerable iodinating activity in the 105,000 x g pellet when studied with exogenous acceptors. These results are compatible with a role for sialic acid in the maturation and migration of thyroglobulin to the iodination site, provided that the intracellular distribution of the iodinating enzymes are normal. (Endocrinology 97: 347, 1975)
nr^HE function of carbohydrates in glyco-L proteins is still poorly understood (1-6); however, it has been possible to attribute a physiological role to sialic acid since recently it has been demonstrated that this carbohydrate is important in preventing tissue uptake of circulating glycoproteins (3). We have studied an experimental rat thyroid tumor (Wollman line 1-1C2) to establish whether carbohydrates in thyroglobulin1 could play a role in thyroid hormone formation. This tumor has been shown to have a defect in thyroglobulin secretion (7) associated with an absence of sialic acid incorporation resulting from a lack of membrane-bound sialyltransferase activity (8). Recently we have isolated and
purified the tumor asialo-thyroglobulin and shown that it has a sedimentation coefficient and immunological properties similar to normal and enzymatically desialylated thyroglobulin (9). On the other hand tumor thyroglobulin has a lower electrophoretic mobility than normal thyroglobulin and a very low iodine content. In the present work we have examined the thyroid hormone content after in vivo and in vitro iodination of tumor thyroglobulin in comparison with normal and desialylated rat thyroglobulin to establish whether the presence of sialic acid in thyroglobulin plays a direct or indirect role in thyroid hormone formation. Our results suggest that the intracellular incorporation of sialic acid into thyroglobulin in vivo is important and may be necessary for thyroid hormone formation.
Received December 16, 1974. Supported in part by the Public Health Service Research Grant no. 1R01 AM 16932 from the National Institute of Arthritis, Metabolism, and Digestive Diseases, U.S. Public Health Service. Reprint requests should be sent to: Dr. Fabrizio Monaco, Centro della Tiroide c/o Ha Clinica Medica dell'Universita, Policlinico Umberto 1°, 00100 Rome, Italy. 1 The abbreviations used are: TG, thyroglobulin; T4, thyroxine; T3, triiodothyronine; DIT, diiodotyrosine; MIT, monoiodotyrosine.
Materials and Methods Tumor thyroglobulin, from the experimental rat thyroid tumor 1-1C2, was isolated by ammonium sulfate fractionation followed by sucrose gradient ultracentrifugation (9); briefly the tissue was mildly homogenized in Tris-HCl sucrose buffer and the homogenate centrifuged at 105,000 x g for 1 h at 4 C (7). The 105,000 x g
347
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
Endo • 1975 Vol 97 • No 2
MONACO, GRIMALDI, DOMINICI AND ROBBINS
348
supernatant after exhaustive dialysis against Tris-HCl buffer was fractionated twice with ammonium sulfate; the 1.4-1.8M ammonium sulfate fraction, dissolved in Tris-HCl buffer, was dialyzed and separated by sucrose density gradient ultracentrifugation (8). The fractions sedimenting in the same region as 19S thyroglobulin were collected, concentrated, and dialyzed against Tris-HCl buffer. Tumor thyroglobulin was homogeneous in analytical ultracentrifugation and showed no contaminating protein either in polyacrylamide gel electrophoresis or immunoelectrophoresis (9). Normal rat thyroglobulin was similarly purified (9), and desialylated with neuraminidase from Vibria cholerae (8), from Boehringer/Mannheim. In vivo iodination of tumor, normal and desialylated rat thyroglobulin was examined 24 and 48 h after ip injection of 200 /nCi of 125 I~(Serin, Saluggia). The in vitro enzymatic iodination of tumor, normal and desialylated rat thyroglobulin was performed with lactoperoxidase, as described (10). Lactoperoxidase was from Calbiochem, San Diego (25 IU/mg and A412/A280 = 0.589). Glucose oxidase was purchased from Boehringer and S., GmbH, Mannheim. The incubation mixture contained per ml of solution (in 0.5M phosphate buffer, pH 6.0) 2 /u,g lactoperoxidase, 1 nmole thyroglobulin, 50-100 nmol iodide, 125I~ as tracer and an hydrogen peroxide generating system consisting of 0.75 mg glucose and 2.0 /xg glucose oxidase. After 30-60 min the reaction was stopped by dialysis against 0.02M phosphate buffer-O.lM KC1, pH 7.4. Protein recovery was estimated by A280 measurement before and after iodination. The fraction of the total 125I bound to protein was determined by paper chromatography (11); the 125I remaining at the origin of the chromatogram was considered as protein-bound. Hydrolysis of 125I-labeled thyroglobulin was performed with pronase and leucylaminopeptidase as described (12) and the product analyzed by descending paper chromatography in butanol-acetic acid-water (78:5:17) and TABLE
127J
Sialic acid
1. Sialic acid and iodine content of purified rat thyroglobulin Tumor
Desialylated normal
Normal
g/lOOg
g/lOOg
g/100g
0.02 0.15
0.32 0.20
0.34 1.3
The values are the means of three determinations.
TABLE 2. Distribution of IMI iodoamino acids in purified rat thyroglobulin Atoms I/mol thyroglobulin
Origin
MIT
DIT
Ts
T,
1
2
40
51
3
0
Desialylated normal
16.2
4
28
52
1
9
Normal
17.0
3
30
48
2
10
Tumor
32.4t
3
29
44
2
10
Desialylated normal
27.lt
2
33
49
1
8
Normal
28.6t
4
28
46
2
8
% of total '"I in chromatogram
In vivo iodination' Tumor
In vitro iodination
The values are the means of 2 experiments in duplicate. * 24 h after 200 /xCi '"I, ip. f Sum of newly introduced and endogenous iodine.
t-pentanol-dioxane-lM NH4OH (2:2:1) or by thin layer chromatography in pentanol-dioxaneNH4OH (2:2:1) (13). The iodinating activity of tumor tissue was performed as described (14); briefly, tumor tissue and normal thyroid for control were gently homogenized, centrifuged at 105,000 x g and the pellet suspended in Tris-HCl buffer, pH 7.4, at 37 C in the presence of exogenous acceptors and 1 /u.Ci of 125I" for 1 h. After incubation the mixtures were centrifuged, precipitated with TCA, and the precipitate was measured for radioactivity. The DNA content of the tissue was measured with the diphenylamine reagent using deoxyadenosine as standard (15). The protein content of the 105,000 x g pellet was determined by the method of Lowry et al. Sialic acid was determined by the thiobarbituric acid assay (17) and iodine according to the method of Zak et al. (18). Low iodine thyroglobulin was isolated from a human nontoxic goiter, by ammonium sulfate fractionation followed by density gradient ultracentrifugation (9).
Results The recovery of thyroglobulin from the tumor was 7.5 mg/g wet tissue compared to 60-70 mg/g for normal thyroid. The iodine content of tumor TG was very low compared to normal and desialylated
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
-8
~
E E o CO CVJ
.1 -
-2
10
15
FRACTION
FIG. 1. Density gradient ultracentrifugation patterns of normal (A), desialylated normal (B) and tumor (C) rat thyroglobulin. Sucrose linear density gradient 10 to 40% in 0.1M Tris-HCl, pH 7.4, rotor SW 27 small buckets; 26,000 rpm for 20 h, 20 C, in a Spinco Beckman L2-50 ultracentrifuge. T indicates the top of the tube and B indicates the bottom. The zone labeled 19 S was determined with normal rat thyroglobulin in the same centrifuge run. D Profile of the protein absorbance at 28Q nm before iodination. • Profile of the 125I radioactivity after iodination.
10
15
30
25
20
T
NUMBER
20
25
30 T
FRACTION NUMBER
10
15
FRACTION
20 NUMBER
30
T
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
350
MONACO, GRIMALDI, DOMINICI AND ROBBINS
TG, as shown in Table 1. The sialic acid content of tumor TG, (Table 1) was also very low and similar to that of enzymatically desialylated TG. Table 2 shows the labeled iodoaminoacid distribution in the three thyroglobulins 24 h after in vivo iodination with 200 /LtCi of 125I" injected ip. It is evident that tumor TG contained very few iodoaminoacid residues, compared to the normal and desialylated TG. Thyroxine was not detected in the tumor TG whereas 3% of the label was present in the region of triiodothyronine. The T3 peak was clearly evident in those chromatogram systems which separated T3 from T4. Similar results were obtained at 48 hours after in vivo iodination with 125I~. Preliminary experiments on in vitro iodination were performed in order to obtain the same degree of iodination in each of the three TG preparations since they differed in initial iodine content as well as in the rate of iodine incorporation. The recovery of tumor TG after in vitro iodination was 43% of starting material compared to 58% for desialylated TG and 97% for normal TG. The instability of tumor and desialylated TG was also shown by ultracentrifugal analysis of the product. As seen in Fig. 1, there was a relatively increased proportion of 12 S and 3-8 S components in the iodinated thyroglobulin compared to the starting material and also compared to normal TG. Iodoaminoacid analysis was done on the total protein in each case. As shown in Table 2, the three TG preparations contained similar amounts of all iodoaminoacids including thyroxine. TABLE 3. Iodinating activity of normal or tumor participate fraction m
I cpm incorporated in TCA precipitate
Acceptor
/g wet tissue
/mg protein
/fig DNA
Normal
TG BSA
20,188,000 7,592,000
4,800 2,550
6,700 3,500
Tumor
TG BSA
3,171,000 1,354,000
2,100 800
840 220
The values are the means of 2 experiments in duplicate.
Endo • 1975 Vol 97 • No 2
A control experiment was then carried out to see whether the decreased ability of tumor tissue to form thyroid hormones in vivo could be attributed to a defective iodinating activity of tumor tissue. As shown in Table 3 the tumor 105,000 x g pellet exhibited considerable iodinating activity, using as acceptors either low iodine human thyroglobulin or bovine serum albumin, although the activity was significantly lower than that in normal tissue. In a separate experiment, iodination of thyroglobulin by the 105,000 x g pellet of normal rat thyroid was tested in the presence of tumor 105,000 x g supernatant. The reaction mixture contained 1 mg of normal particulars protein, 2 mg of tumor supernatant protein and 0.5 mg of low iodine human thyroglobulin. 125I incorporation was 87,700 and 90,500 cpm in the presence and absence of tumor supernatant, respectively, indicating that the decreased iodination in the tumor in vivo was not caused by a soluble inhibitor. Discussion The experimental rat thyroid tumor, Wollman line 1-1C2, synthesizes very little soluble thyroglobulin; most of the protein is membrane-bound and can be released by vigorous homogenization or detergent (7). Moreover the tumor is unable to incorporate sialic acid into thyroglobulin because of a defective sialyltransferase activity (8). The recent purification of the tumor asialothyroglobulin allowed us to study its chemical content of iodine and sialic acid and to see whether this protein was able in vitro to form thyroid hormones. Tumor thyroglobulin, in fact, has very low sialic acid and iodine content, and tumor tissue in vivo is unable to iodinate thyroglobulin completely. When the tumor bearing rat is given an ip injection of 125I~ its thyroglobulin contains only a minute amount of T3 and no detectable T4. In order to establish whether the low hormone synthesis was caused by a structural defect in the thyroglobulin, in vitro iodination of purified TG was examined. It was shown that the tumor
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
SIALIC ACID AND IODINE IN THYROGLOBULIN thyroglobulin was able to form both T3 and T4 in normal amounts. That sialic acid does not play a role directly in iodination of thyroglobulin is confirmed by the finding that enzymatically desialylated normal rat thyroglobulin shows a normal formation of T3 and T4 when iodinated in vitro. We have, however, to emphasize that sialic acid may play a role in the stability of thyroglobulin because the recovery of 19 S thyroglobulin, after in vitro iodination of tumor and desialylated thyroglobulin, is much less than normal TG. On the other hand we cannot exclude the possibility that tumor thyroglobulin has a structural defect in the polypeptide chain, although it is unlikely that neuraminidase treatment altered the polypeptide chain of normal thyroglobulin. There remains the possibility that tumor tissue was unable to iodinate thyroglobulin because of a defective iodinating activity. Evidence against this possibility was obtained in the demonstration that the 105,000 x g pellet of the tumor was able to iodinate exogenous acceptor proteins. Although the activity was lower than in normal thyroid, prior autoradiographic studies on this very cellular tumor have shown that iodination occurs only in small scattered loci in the tumor. On the other hand, it is known that peroxidase activity of normal thyroid can be demonstrated in many parts of the cell (19) whereas only that located at the apical membrane may be important for in vivo iodination of thyroglobulin. We have no information on the intracellular localization of iodinating activity in the tumor. In conclusion our results demonstrate that the experimental rat thyroid tumor synthesizes thyroglobulin with very low sialic acid and iodine content. Tumor thyroglobulin iodination in vivo is accompanied by little or no thyroid hormone synthesis whereas in vitro it shows hor-
351
mone formation very similar to that shown by normal and desialylated normal thyroglobulin. This is compatible with a role for sialic acid in the maturation and migration of thyroglobulin to the iodinating site of the cell. Acknowledgments The technical assistance of Mr. C. Santelamazza, I. De Ros and Mrs. M. Minu is gratefully acknowledged; we are indebted to Mrs. A. Di Girolamo and Mrs. L. Perry for preparation and revision of the manuscript.
References 1. Eylar, E. H . J Theor Biol 10: 84, 1966. 2. Winterburn, P. J., and C. F. Phelps, Nature 236: 147, 1972. 3. Morell, A. G., G. Gregoriadis, I. H. Scheinburg, J. Hickman, and G. AshwellJ Biol Chem 246: 1461, 1971. 4. Braatz, J. A., and E. C. Heath, / Biol Chem 249: 2536, 1974. 5. Cumar, F. A., R. O. Brady, E. H. Kolodny, V. W. McFarland, and P. T. Mora, Proc Natl Acad Sci USA 67: 757, 1970. 6. Den, H., A. M. Schultz, M. Basu, and S. Roseman, J Biol Chem 246: 2721, 1971. 7. Monaco, F., and J. Robbins, Biochim Biophys Ada 272: 118, 1972. 8. , and ,J Biol Chem 248: 2328, 1973. 9. , S. Grimaldi, G. Scuncio, and M. Andreoli, Acta Endocrinol (Kbh) 77: 517, 1974. 10. Pommier, J., L. Sokoloff, and J. Nunez, Eur J Biochem 38: 497, 1973. 11. Lamas, L., M. L. Dorris, and A. Taurog, Endocrinology P0: 1417, 1972. 12. Rolland, '*!., R. Aquaron, and S. Lissitzky, Anal Biochev. 33: 307, 1970. 13. Cahnm&Tin, H. J., In Berson, S. A., J. E. Rail, and I. J. Kop.n (eds.), Methods in Investigative and Diagnostic Endocrinology, vol. I, part I, North Holland Publ. Co., Amsterdam, 1972, p. 27. 14. Monaco, F., G. Salvatore, and J. Robbins, J Biol Chem 250: 1595, 1975. 15. Burton, K., Biochem J 62: 315, 1956. 16. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. RandallJ Biol Chem 193: 265, 1951. 17. Warren, L . J Biol Chem 234: 1971, 1959. 18. Zak, B., A. H. Willard, G. B. Myers, and A. J. Boyle, Anal Chem 24: 1345, 1952. 19. Tice, L. W., and S. H. Wollman, Endocrinology 94: 1555, 1974.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
thyroglobulin iodinated in vitro. In vitro iodination dissociated tumor and desialylated normal thyroglobulin to a greater extent than normal thyroglobulin. Tumor tissue, on the other hand, showed considerable iodinating activity in the 105,000 x g pellet when studied with exogenous acceptors. These results are compatible with a role for sialic acid in the maturation and migration of thyroglobulin to the iodination site, provided that the intracellular distribution of the iodinating enzymes are normal. (Endocrinology 97: 347, 1975)
nr^HE function of carbohydrates in glyco-L proteins is still poorly understood (1-6); however, it has been possible to attribute a physiological role to sialic acid since recently it has been demonstrated that this carbohydrate is important in preventing tissue uptake of circulating glycoproteins (3). We have studied an experimental rat thyroid tumor (Wollman line 1-1C2) to establish whether carbohydrates in thyroglobulin1 could play a role in thyroid hormone formation. This tumor has been shown to have a defect in thyroglobulin secretion (7) associated with an absence of sialic acid incorporation resulting from a lack of membrane-bound sialyltransferase activity (8). Recently we have isolated and
purified the tumor asialo-thyroglobulin and shown that it has a sedimentation coefficient and immunological properties similar to normal and enzymatically desialylated thyroglobulin (9). On the other hand tumor thyroglobulin has a lower electrophoretic mobility than normal thyroglobulin and a very low iodine content. In the present work we have examined the thyroid hormone content after in vivo and in vitro iodination of tumor thyroglobulin in comparison with normal and desialylated rat thyroglobulin to establish whether the presence of sialic acid in thyroglobulin plays a direct or indirect role in thyroid hormone formation. Our results suggest that the intracellular incorporation of sialic acid into thyroglobulin in vivo is important and may be necessary for thyroid hormone formation.
Received December 16, 1974. Supported in part by the Public Health Service Research Grant no. 1R01 AM 16932 from the National Institute of Arthritis, Metabolism, and Digestive Diseases, U.S. Public Health Service. Reprint requests should be sent to: Dr. Fabrizio Monaco, Centro della Tiroide c/o Ha Clinica Medica dell'Universita, Policlinico Umberto 1°, 00100 Rome, Italy. 1 The abbreviations used are: TG, thyroglobulin; T4, thyroxine; T3, triiodothyronine; DIT, diiodotyrosine; MIT, monoiodotyrosine.
Materials and Methods Tumor thyroglobulin, from the experimental rat thyroid tumor 1-1C2, was isolated by ammonium sulfate fractionation followed by sucrose gradient ultracentrifugation (9); briefly the tissue was mildly homogenized in Tris-HCl sucrose buffer and the homogenate centrifuged at 105,000 x g for 1 h at 4 C (7). The 105,000 x g
347
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
Endo • 1975 Vol 97 • No 2
MONACO, GRIMALDI, DOMINICI AND ROBBINS
348
supernatant after exhaustive dialysis against Tris-HCl buffer was fractionated twice with ammonium sulfate; the 1.4-1.8M ammonium sulfate fraction, dissolved in Tris-HCl buffer, was dialyzed and separated by sucrose density gradient ultracentrifugation (8). The fractions sedimenting in the same region as 19S thyroglobulin were collected, concentrated, and dialyzed against Tris-HCl buffer. Tumor thyroglobulin was homogeneous in analytical ultracentrifugation and showed no contaminating protein either in polyacrylamide gel electrophoresis or immunoelectrophoresis (9). Normal rat thyroglobulin was similarly purified (9), and desialylated with neuraminidase from Vibria cholerae (8), from Boehringer/Mannheim. In vivo iodination of tumor, normal and desialylated rat thyroglobulin was examined 24 and 48 h after ip injection of 200 /nCi of 125 I~(Serin, Saluggia). The in vitro enzymatic iodination of tumor, normal and desialylated rat thyroglobulin was performed with lactoperoxidase, as described (10). Lactoperoxidase was from Calbiochem, San Diego (25 IU/mg and A412/A280 = 0.589). Glucose oxidase was purchased from Boehringer and S., GmbH, Mannheim. The incubation mixture contained per ml of solution (in 0.5M phosphate buffer, pH 6.0) 2 /u,g lactoperoxidase, 1 nmole thyroglobulin, 50-100 nmol iodide, 125I~ as tracer and an hydrogen peroxide generating system consisting of 0.75 mg glucose and 2.0 /xg glucose oxidase. After 30-60 min the reaction was stopped by dialysis against 0.02M phosphate buffer-O.lM KC1, pH 7.4. Protein recovery was estimated by A280 measurement before and after iodination. The fraction of the total 125I bound to protein was determined by paper chromatography (11); the 125I remaining at the origin of the chromatogram was considered as protein-bound. Hydrolysis of 125I-labeled thyroglobulin was performed with pronase and leucylaminopeptidase as described (12) and the product analyzed by descending paper chromatography in butanol-acetic acid-water (78:5:17) and TABLE
127J
Sialic acid
1. Sialic acid and iodine content of purified rat thyroglobulin Tumor
Desialylated normal
Normal
g/lOOg
g/lOOg
g/100g
0.02 0.15
0.32 0.20
0.34 1.3
The values are the means of three determinations.
TABLE 2. Distribution of IMI iodoamino acids in purified rat thyroglobulin Atoms I/mol thyroglobulin
Origin
MIT
DIT
Ts
T,
1
2
40
51
3
0
Desialylated normal
16.2
4
28
52
1
9
Normal
17.0
3
30
48
2
10
Tumor
32.4t
3
29
44
2
10
Desialylated normal
27.lt
2
33
49
1
8
Normal
28.6t
4
28
46
2
8
% of total '"I in chromatogram
In vivo iodination' Tumor
In vitro iodination
The values are the means of 2 experiments in duplicate. * 24 h after 200 /xCi '"I, ip. f Sum of newly introduced and endogenous iodine.
t-pentanol-dioxane-lM NH4OH (2:2:1) or by thin layer chromatography in pentanol-dioxaneNH4OH (2:2:1) (13). The iodinating activity of tumor tissue was performed as described (14); briefly, tumor tissue and normal thyroid for control were gently homogenized, centrifuged at 105,000 x g and the pellet suspended in Tris-HCl buffer, pH 7.4, at 37 C in the presence of exogenous acceptors and 1 /u.Ci of 125I" for 1 h. After incubation the mixtures were centrifuged, precipitated with TCA, and the precipitate was measured for radioactivity. The DNA content of the tissue was measured with the diphenylamine reagent using deoxyadenosine as standard (15). The protein content of the 105,000 x g pellet was determined by the method of Lowry et al. Sialic acid was determined by the thiobarbituric acid assay (17) and iodine according to the method of Zak et al. (18). Low iodine thyroglobulin was isolated from a human nontoxic goiter, by ammonium sulfate fractionation followed by density gradient ultracentrifugation (9).
Results The recovery of thyroglobulin from the tumor was 7.5 mg/g wet tissue compared to 60-70 mg/g for normal thyroid. The iodine content of tumor TG was very low compared to normal and desialylated
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
-8
~
E E o CO CVJ
.1 -
-2
10
15
FRACTION
FIG. 1. Density gradient ultracentrifugation patterns of normal (A), desialylated normal (B) and tumor (C) rat thyroglobulin. Sucrose linear density gradient 10 to 40% in 0.1M Tris-HCl, pH 7.4, rotor SW 27 small buckets; 26,000 rpm for 20 h, 20 C, in a Spinco Beckman L2-50 ultracentrifuge. T indicates the top of the tube and B indicates the bottom. The zone labeled 19 S was determined with normal rat thyroglobulin in the same centrifuge run. D Profile of the protein absorbance at 28Q nm before iodination. • Profile of the 125I radioactivity after iodination.
10
15
30
25
20
T
NUMBER
20
25
30 T
FRACTION NUMBER
10
15
FRACTION
20 NUMBER
30
T
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
350
MONACO, GRIMALDI, DOMINICI AND ROBBINS
TG, as shown in Table 1. The sialic acid content of tumor TG, (Table 1) was also very low and similar to that of enzymatically desialylated TG. Table 2 shows the labeled iodoaminoacid distribution in the three thyroglobulins 24 h after in vivo iodination with 200 /LtCi of 125I" injected ip. It is evident that tumor TG contained very few iodoaminoacid residues, compared to the normal and desialylated TG. Thyroxine was not detected in the tumor TG whereas 3% of the label was present in the region of triiodothyronine. The T3 peak was clearly evident in those chromatogram systems which separated T3 from T4. Similar results were obtained at 48 hours after in vivo iodination with 125I~. Preliminary experiments on in vitro iodination were performed in order to obtain the same degree of iodination in each of the three TG preparations since they differed in initial iodine content as well as in the rate of iodine incorporation. The recovery of tumor TG after in vitro iodination was 43% of starting material compared to 58% for desialylated TG and 97% for normal TG. The instability of tumor and desialylated TG was also shown by ultracentrifugal analysis of the product. As seen in Fig. 1, there was a relatively increased proportion of 12 S and 3-8 S components in the iodinated thyroglobulin compared to the starting material and also compared to normal TG. Iodoaminoacid analysis was done on the total protein in each case. As shown in Table 2, the three TG preparations contained similar amounts of all iodoaminoacids including thyroxine. TABLE 3. Iodinating activity of normal or tumor participate fraction m
I cpm incorporated in TCA precipitate
Acceptor
/g wet tissue
/mg protein
/fig DNA
Normal
TG BSA
20,188,000 7,592,000
4,800 2,550
6,700 3,500
Tumor
TG BSA
3,171,000 1,354,000
2,100 800
840 220
The values are the means of 2 experiments in duplicate.
Endo • 1975 Vol 97 • No 2
A control experiment was then carried out to see whether the decreased ability of tumor tissue to form thyroid hormones in vivo could be attributed to a defective iodinating activity of tumor tissue. As shown in Table 3 the tumor 105,000 x g pellet exhibited considerable iodinating activity, using as acceptors either low iodine human thyroglobulin or bovine serum albumin, although the activity was significantly lower than that in normal tissue. In a separate experiment, iodination of thyroglobulin by the 105,000 x g pellet of normal rat thyroid was tested in the presence of tumor 105,000 x g supernatant. The reaction mixture contained 1 mg of normal particulars protein, 2 mg of tumor supernatant protein and 0.5 mg of low iodine human thyroglobulin. 125I incorporation was 87,700 and 90,500 cpm in the presence and absence of tumor supernatant, respectively, indicating that the decreased iodination in the tumor in vivo was not caused by a soluble inhibitor. Discussion The experimental rat thyroid tumor, Wollman line 1-1C2, synthesizes very little soluble thyroglobulin; most of the protein is membrane-bound and can be released by vigorous homogenization or detergent (7). Moreover the tumor is unable to incorporate sialic acid into thyroglobulin because of a defective sialyltransferase activity (8). The recent purification of the tumor asialothyroglobulin allowed us to study its chemical content of iodine and sialic acid and to see whether this protein was able in vitro to form thyroid hormones. Tumor thyroglobulin, in fact, has very low sialic acid and iodine content, and tumor tissue in vivo is unable to iodinate thyroglobulin completely. When the tumor bearing rat is given an ip injection of 125I~ its thyroglobulin contains only a minute amount of T3 and no detectable T4. In order to establish whether the low hormone synthesis was caused by a structural defect in the thyroglobulin, in vitro iodination of purified TG was examined. It was shown that the tumor
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
SIALIC ACID AND IODINE IN THYROGLOBULIN thyroglobulin was able to form both T3 and T4 in normal amounts. That sialic acid does not play a role directly in iodination of thyroglobulin is confirmed by the finding that enzymatically desialylated normal rat thyroglobulin shows a normal formation of T3 and T4 when iodinated in vitro. We have, however, to emphasize that sialic acid may play a role in the stability of thyroglobulin because the recovery of 19 S thyroglobulin, after in vitro iodination of tumor and desialylated thyroglobulin, is much less than normal TG. On the other hand we cannot exclude the possibility that tumor thyroglobulin has a structural defect in the polypeptide chain, although it is unlikely that neuraminidase treatment altered the polypeptide chain of normal thyroglobulin. There remains the possibility that tumor tissue was unable to iodinate thyroglobulin because of a defective iodinating activity. Evidence against this possibility was obtained in the demonstration that the 105,000 x g pellet of the tumor was able to iodinate exogenous acceptor proteins. Although the activity was lower than in normal thyroid, prior autoradiographic studies on this very cellular tumor have shown that iodination occurs only in small scattered loci in the tumor. On the other hand, it is known that peroxidase activity of normal thyroid can be demonstrated in many parts of the cell (19) whereas only that located at the apical membrane may be important for in vivo iodination of thyroglobulin. We have no information on the intracellular localization of iodinating activity in the tumor. In conclusion our results demonstrate that the experimental rat thyroid tumor synthesizes thyroglobulin with very low sialic acid and iodine content. Tumor thyroglobulin iodination in vivo is accompanied by little or no thyroid hormone synthesis whereas in vitro it shows hor-
351
mone formation very similar to that shown by normal and desialylated normal thyroglobulin. This is compatible with a role for sialic acid in the maturation and migration of thyroglobulin to the iodinating site of the cell. Acknowledgments The technical assistance of Mr. C. Santelamazza, I. De Ros and Mrs. M. Minu is gratefully acknowledged; we are indebted to Mrs. A. Di Girolamo and Mrs. L. Perry for preparation and revision of the manuscript.
References 1. Eylar, E. H . J Theor Biol 10: 84, 1966. 2. Winterburn, P. J., and C. F. Phelps, Nature 236: 147, 1972. 3. Morell, A. G., G. Gregoriadis, I. H. Scheinburg, J. Hickman, and G. AshwellJ Biol Chem 246: 1461, 1971. 4. Braatz, J. A., and E. C. Heath, / Biol Chem 249: 2536, 1974. 5. Cumar, F. A., R. O. Brady, E. H. Kolodny, V. W. McFarland, and P. T. Mora, Proc Natl Acad Sci USA 67: 757, 1970. 6. Den, H., A. M. Schultz, M. Basu, and S. Roseman, J Biol Chem 246: 2721, 1971. 7. Monaco, F., and J. Robbins, Biochim Biophys Ada 272: 118, 1972. 8. , and ,J Biol Chem 248: 2328, 1973. 9. , S. Grimaldi, G. Scuncio, and M. Andreoli, Acta Endocrinol (Kbh) 77: 517, 1974. 10. Pommier, J., L. Sokoloff, and J. Nunez, Eur J Biochem 38: 497, 1973. 11. Lamas, L., M. L. Dorris, and A. Taurog, Endocrinology P0: 1417, 1972. 12. Rolland, '*!., R. Aquaron, and S. Lissitzky, Anal Biochev. 33: 307, 1970. 13. Cahnm&Tin, H. J., In Berson, S. A., J. E. Rail, and I. J. Kop.n (eds.), Methods in Investigative and Diagnostic Endocrinology, vol. I, part I, North Holland Publ. Co., Amsterdam, 1972, p. 27. 14. Monaco, F., G. Salvatore, and J. Robbins, J Biol Chem 250: 1595, 1975. 15. Burton, K., Biochem J 62: 315, 1956. 16. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. RandallJ Biol Chem 193: 265, 1951. 17. Warren, L . J Biol Chem 234: 1971, 1959. 18. Zak, B., A. H. Willard, G. B. Myers, and A. J. Boyle, Anal Chem 24: 1345, 1952. 19. Tice, L. W., and S. H. Wollman, Endocrinology 94: 1555, 1974.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 November 2015. at 04:04 For personal use only. No other uses without permission. . All rights reserved.
