Cell Tissue Res (1987) 250: 149-156 Cell andTlSSUe Reseatdl © Springer-Verlag 1987

Localization of the incorporation of 3H-galactose and 3H -sialic acid into thyroglobulin in relation to the block of intracellular transport induced by monensin Studies with isolated porcine thyroid follicles

P. Ring, U. Bjorkman, and R. Ekholm Department of Anatomy, University of Goteborg, Goteborg, Sweden

Summary. The Na + /K + ionophore monensin is known to arrest the intracellular transport of newly synthesized pro­teins in the Golgi complex. In the present investigation the effect of monensin on the secretion of 3H -galactose-labeled and 3H-sialic acid-labeled thyroglobulin was studied in open thyroid follicles isolated from porcine thyroid tissue.

Follicles were incubated with 3H-galactose at 20° C for 1 h; at this temperature the labeled thyroglobulin remains in the labeling compartment (Ring et al. 1987a). The folli­cles were then chased at 37° C for 1 h in the absence or presence of 1 ~M monensin. Without monensin substantial amounts of labeled thyroglobulin were secreted into the medium, whereas in the presence of the ionophore secretion was inhibited by 80%. Since we have previously shown (Ring et al. 1987b) that monensin does not inhibit secretion of thyroglobulin present on the distal side of the monensin block we conclude that galactose is incorporated into thyro­globulin on the proximal side of this block.

Secretion was also measured in follicles continuously incubated with 3H-galactose for 1 hat 37° C in the absence or presence of monensin. In these experiments secretion of labeled thyroglobulin was inhibited by about 85% in the presence of monensin. Identically designed experiments with 3H-N-acetylmannosamine, a precursor of sialic acid, gave similar results, i.e., almost complete inhibition ofsecre­tion of labeled thyroglobulin in the presence of monensin. The agreement between the results of the galactose and sialic acid experiments indicates that sialic acid, like galac­tose, is incorporated into thyroglobulin on the proximal side of the monensin block.

Considering observations made in other cell systems the present results suggest that galactosylation and sialylation of thyroglobulin are completed within the Golgi complex.

Key words: Thyroid gland (porcine) - Thyroglobulin - 3H_ Galactose - Sialic acid - Monensin - Intracellular transport - Secretion

Thyroglobulin, the major secretory protein of the thyroid gland, is a glycoprotein containing about 10% carbohy­drates. Almost all the monosaccharides occur in two dis­tinct carbohydrate units, both N-glycosidically linked to the polypeptide chain (Spiro 1965; Fukuda and Egomi

Send offprint requests to: Dr. R. Ekholm, Dept. of Anatomy, Univ. of Goteborg, Box 33031, S-40033 Goteborg, Sweden

1971; Arima and Spiro 1972). One carbohydrate unit (A) consists only of N-acetylglucosamine and mannose; the other unit (B) has an inner core of the same monosacchar­ides but contains, in addition, fucose, N-acetylglucosamine, galactose and sialic acid.

The initial step in glycosylation of thyroglobulin com­prises transfer en bloc of an oligosaccharide containing N­acetylglucosamine, mannose and glucose (G1c3 Mang G1c­Nac2) from a dolichol pyrophosphate carrier to the nascent protein (Spiro et al. 1976a-c; Spiro et al. 1979). During the subsequent processing all glucose and most mannose residues are removed (Godelaine et al. 1981). The peripher­al monosaccharides of the complex (B) unit are then incor­porated.

Concerning the intracellular location of the different steps in thyroglobulin glycosylation it is clear that the at­tachment of the core oligosaccharides to the polypeptides takes place in the rough endoplasmic reticulum of the folli­cle cells. There is also evidence that in the processing of the carbohydrate units, all glucose residues and the first mannose residues are removed in the rough endoplasmic reticulum, while additional mannose residues are removed in the Golgi complex (Godelaine et al. 1981).

The localization of the incorporation of the peripheral monosaccharides into the B unit of thyroglobulin is based on autoradiographic studies after administration of labeled fucose (Haddad et al. 1971), galactose (Whur et al. 1969; Haddad 1972) and sialic acid precursor (Bennet et al. 1981); by demonstration of galactosyl- and sialyltransferases in thyroid subcellular fractions (Chabaud et al. 1974); and by immunohistochemical localization of galactosyltransferase (Bouchilloux et al. 1981). All these studies show that the Golgi complex is the major site of the terminal glycosylation of thyroglobulin. However, in these studies it has not been possible to localize the incorporation of the peripheral sug­ars in a subcompartment of the Golgi complex or to ex­clude that some incorporation of these sugars occurs apical to the Golgi complex. The latter possibility is of special interest with respect to observations that terminal glycosy­lation of thyroglobulin influences its binding to thyroid membranes. According to suggestions advanced in these studies (Consiglio et al. 1979; Consiglio et al. 1981; Van den Hove et al. 1982) newly synthesized thyroglobulin mol­ecules, with incomplete carbohydrate B-chains, would be bound to the apical plasma membrane where they should be concentrated near special iodination sites. At these sites sialic acid (and possibly galactose as well) would be added

150

and iodination initiated; sialylation would allow the mole­cules to be released into the follicle lumen.

In the present study the carboxyl ionophore monensin was used to elucidate the location of incorporation of galac­tose and sialic acid into thyroglobulin in isolated pig thyroid follicles. Monensin binds Na +, K + and protons and me­diates exchange of cations and protons across biological membranes resulting in partial equilibration of Na + and K + (Pressman and Fahim 1982; Tartakoff 1983a, b). Mon­ensin arrests intracellular transport of newly synthesized secretory and membrane proteins in the Golgi complex; a structural expression of this transport inhibition is dila­tion of Golgi cisternae (Tartakoff 1983 b).

Materials and methods

Porcine thyroid glands were obtained from the local abat­toir within 10 min after slaughter. They were transported to the laboratory in ice-cold Tyrode salt solution saturated with oxygen, buffered with 20 mM Hepes buffer (pH 7.4) and supplemented with antibiotics.

Isolation of follicles. Follicles were isolated as described in previous papers (Denef et al. 1980; Ring et al. 1987a). The procedure comprised cutting the tissue into small pieces, incubation with collagenase in Tyrode solution and me­chanical disintegration by pi petting. Follicles were collected by centrifugation, resuspended and filtered through a nylon mesh. They were kept in Tyrode solution, gassed with O2 /

CO2 and supplemented with amino acids and 0.5% bovine serum albumin for 1 h before use. This isolation procedure yields open follicles, devoid of colloid, and follicle segments (Bjorkman and Ekholm 1984).

Incubation of isolated follicles with 3H-galactose, 3H-N-ace­tyl-mannosamine and 3 H-Ieucine. The incubation medium was Tyrode solution, pH 7.4, supplemented with amino acids (Eagle 1959), 0.2 J.lg/ml DNAse and 50 IV/ml benzyl­penicillin and with glucose replaced by pyruvate. Follicles were either incubated with 3H-galactose or 3H-N-acetyl­mannosamine (a precursor of sialic acid, see Bennet and O'Shaughnessy 1981) at 20° C and then chase-incubated at 37° C or continuously incubated with the labeled sugars at 37° C. In the former experiments an incubation volume of 100 J.ll contained 200 J.lCi 3H-galactose or 1 mCi 3H-N­acetylmannosamine and follicles obtained from 2 g thyroid tissue. When follicles were labeled continuously for 1 h the same amount of follicles was incubated in 300 J.ll medium containing 75 J.lCi 3H-galactose and 500 J.lCi 3H-N-acetyl­mannosamine, respectively. Incubation of follicles with 3H_ leucine was performed in Tyrode solution with the same supplements as above but without leucine. Follicles from 2 g thyroid tissue were incubated in 250 J.ll containing 25 J.lCi 3H-Ieucine.

In chase experiments, the follicles were washed twice after labeling and then resuspended and chased in the pres­ence of 20 mV/ml TSH (which stimulates secretion of thy­roglobulin, cf. Ring et al. 1987 a) and in the presence or absence of 1 J.lM monensin as indicated in Results. At the end of the chase, the follicles were separated from the medi­um by centrifugation. The follicle pelletes were washed twice, resuspended in a hypotonic Tyrode solution, frozen and thawed and homogenized in a Potter-Elvehjem homo­genizer. After centrifugation at 105000 x g for 1 h the super-

natant, containing the soluble proteins, was dialysed against phosphate-buffered saline, pH 6.8; the chase medium was dialysed in the same way.

Continuous labeling experiments performed in the ab­sence or presence of monensin were terminated by separat­ing follicles from medium by means of centrifugation. The medium was dialysed and the pellet was treated in the same way as the pellets in the chase experiments.

Proteins in the final follicle pellets were precipitated with trichloroacetic acid (TCA, final conc. 10%). Precipitates were washed once in 10% TCA, twice in 95% ethanol and once in ether. The residues were dissolved in 0.4 M NaOH and used for protein and DNA determination. Radioactivi­ty was determined by liquid scintillation; protein was esti­mated according to Lowry et al. (1951) and DNA according to McIntire and Sproull (1957).

Radioactivity in thyroglobulin was determined after im­munoprecipitation. The secretion of thyroglobulin was ex­pressed as labeled thyroglobulin in the medium in per cent of total labeled thyroglobulin (soluble labeled thyroglobulin from the follicles + labeled thyroglobulin in the medium).

Immunoprecipitation of thyroglobulin. Samples were mixed with antithyroglobulin immunoglobulin, obtained from rabbits, in the predetermined antibody excess range. The samples, containing 1 % Triton X-IOO, were incubated for 2 h at 24° C. Antiserum to rabbit immunoglobulin was add­ed and the incubation was continued for 1 h at 24° C and then for 24 h at 4° C. The precipitates were collected by centrifugation, washed and solubilized in 0.5 ml 1 M NaOH. Radioactivity was determined in pellet and superna­tant by liquid scintillation.

Chemicals. Collagenase (type II, 135 U/mg) was obtained from Worthington Biochemical Corporation, Freehold, NJ, USA; bovine pancreas DNAse, type I, and soybean trypsin inhibitor, type 1-s, from Sigma Chemical Co., St. Louis, MO, USA; bovine serum albumin fraction V from Miles Biochemicals, Slough, England; monensin from Calbio­chern-Behring Corp., La Jolla, CA, USA. Bovine TSH (30 IU /mg) was a gift from National Hormone and Pituitary Program, Baltimore, MD, USA. 3H-Ieucine, 130-190 Ci/ mmol, 3H-galactose, 5-20 Ci/mmol, and 3H-N-acetylman­nosamine, 23 Ci/mmol, were purchased from the Radio­chemical Centre, Amersham, England.

Results

Properties of follicle preparations. The follicle preparations used in the present study consist of open follicles and follicle segments (Fig. 1). As previously documented (Denef et al. 1980; Bjorkman and Ekholm 1984) and shown in Fig. 2, the follicle cells have preserved polarity and normal ultra­structural features. The follicles synthesize and release thy­roglobulin into the incubation medium by a true secretion, and unspecific leakage of thyroglobulin into the medium is minute during incubation periods used in the present study (Bjorkman and Ekholm 1982). The average basal se­cretion rate of thyroglobulin is about 2 J.lg per h in follicle samples isolated from 1 g thyroid tissue and containing about 200 J.lg DNA; the secretion rate is doubled during the first hours after addition of 20 mU/ml TSH (Ring et al. 1987a).

Fig. 1. Light micrograph of a typical preparation of open follicles and follicle segments isolated from porcine thyroid tissue. x 700

The incorporation rate of labeled compounds into thy­roglobulin varied due to different incubation conditions. To make the estimation of secretion insensitive to the rate of thyroglobulin synthesis, secretion was expressed as the amount of labeled thyroglobulin in the medium in percent

151

of the totallabe1ed thyroglobulin (intracellular + extracellu­lar).

Effect of monensin on the secretion of 3 H-galactose-Iabeled thyroglobulin. In a series of experiments follicles were la­beled by incubation with 3H-galactose at 20° C for 1 hand then washed and chase-incubated for 1 h at 37° C in the presence of TSH and in the absence or presence of monen­sin. Figure 3 shows that a substantial amount of labeled thyroglobulin was released from the follicles during the chase in the absence of monensin; in the presence of the ionophore this release was inhibited by 80%.

In another series of experiments the release of labeled thyroglobulin was measured in follicles continuously incu­bated with 3H-galactose for 1 h at 37° C in the absence or presence of monensin. In contrast to the preceding set of experiments, the incorporation of 3H-galactose into thy­roglobulin was strongly reduced (by about 50%, Fig. 6) in the presence of monensin. The secretion of labeled thyro­globulin, expressed as in the previous experiment in percent of total labeled thyroglobulin, was inhibited by monensin to about the same degree as in the chase experiments (Fig. 4, B).

Effect of monensin on the secretion of 3 H-sialic acid-labeled thyroglobulin. In a series of experiments follicles were la­beled by incubation with 3H-N-acetylmannosamine at 20° C for 15 min and then washed and chase-incubated at

Fig. 2. Electron micrograph of a portion of a follicle segment after incubation in Tyrode solution at 37° C for 2 h. The follicle cells are well polarized. The apical cell surface bears microvilli. Like the apical plasma membrane, the basal plasma membrane is in direct contact with the incubation medium, the basal lamina being removed during the isolation procedure. The organization of the cytoplasm and the structure of the organelles appear to be the same as seen after in vivo fixation of thyroid tissue. x 7500

152

50

40 Cl ~

co ~

0 30

'0 "0

'" <J)

'" Q) 20 ~ Cl ~

*' 10

monensin

Fig. 3. Follicles were labeled with 3H-galactose for 1 h at 20° C, washed and chased for 1 hat 37° C in the presence ofTSH (20 mUj ml) and in the absence or presence of monensin (1 ~M). The dia­gram shows the secretion of labeled thyroglobulin during the chase. Each bar represents the mean of duplicate values. - In this and the experiments shown in Figs. 4 and 5 the amount of labeled thyroglobulin secreted into the medium is expressed as a percentage of total labeled thyroglobulin (soluble thyroglobulin from the folli­cles + thyroglobulin in the medium). The graph represents one of 8 similar experiments giving essentially the same results. Calculated on all experiments, the release of labeled thyroglobulin was reduced by 79% (SD = 7.5) in the presence of monensin as compared with the controls. - The total thyroglobulin-bound radioactivity was approximately 2 x 104 DPM per sample

25

A B

1 20

Cl ~

co 0 15 ~

'0 "0 Q) <J)

'" 10 '" ~ Cl ~

*' 5

monensin monensin

3H-N-acetyl mannosamine 3H-Galactose

Fig. 4. Secretion of labeled thyroglobulin in follicles incubated with 3H-N-acetylmannosamine and 3H-galactose, respectively, for 1 h at 37° C in the absence or presence of monensin; TSH was present during the last 0.5 h of incubation. Each bar represents the mean of duplicate values. - The total thyroglobulin-bound radioactivity in the sialic acid-labeled samples was approximately 5 x 103 DPM in controls and 2.5 x 103 DPM in monensin-exposed follicles. The corresponding values for the galactose-labeled samples were 18 x 104 DPM and 6 x 104 DPM, respectively

50

A B

40 Cl ~

~ (5

30 '0 "0

'" <J)

'" '" '"

20

Cl ~

*' 10

monensin monensin

3 H _ N - acetylmannosamine 3H-Galactose

Fig. 5. Follicles were labeled with 3H-N-acetylmannosamine and 3H-galactose, respectively, for 15 min at 20° C and chased for 1 h in the presence of TSH and in the absence or presence of monensin. The diagram shows the secretion of labeled thyroglobulin during the chase. Each bar represents the mean of duplicate values. -The total thyroglobulin-bound radioactivity in the sialic acid-la­beled samples was approximately 3 x 104 DPM in the controls and 1 x 104 DPM in the monensin-exposed follicles. In the galactose experiments the total thyroglobulin-bound radioactivity was about 7.5 x 105 DPM in both controls and monensin-exposed samples

Table 1. Two groups of four follicle samples were labeled with 3H-N-acetylmannosamine and 3H-galactose, respectively, for 0.5 h at 20° C. Two samples from each group were washed and total thyroglobulin-bound radioactivity (DPM) was determined. The re­maining samples were chase incubated for 0.5 h at 37° C and total thyroglobulin radioactivity was measured

Labeling Labeling + Chase

3H -N -acetylmannosamine

1800 13200

3H-galactose

18700 19100

37° C for 1 h. Fig. 5 (A) shows that the follicles released a considerable amount of 3H-sialic acid-labeled thyroglobu­lin in the absence of monensin and a very small amount in the presence of the ionophore. The results correspond well with those obtained after 3H-galactose-labeling in an identical experiment run in parallel (Fig. 5, B) and with those shown in Fig. 3.

It was found, however, that in these experiments the incorporation of 3H-sialic acid in the monensin-exposed follicles was much lower than in control follicles despite the fact that monensin was not present during the labeling but only during the chase. To elucidate this reduced incor­poration of 3H-sialic acid, the total amount of labeled thy­roglobulin (thyroglobulin in the medium + thyroglobulin in the follicles) was measured immediately after incubation with 3H-N-acetylmannosamine and after wash and chase incubation. Table 1 shows that, in contrast to the incorpo­ration of 3H-galactose, practically no incorporation of sialic acid occurred during incubation with sialic acid precursor at 20° C but took place during the chase at 37° C. This

50

40

30 en

l-e

':' ;: 20

:;;: 0.. 0

10

monensin 1h 2h

- monensin 1h

monensin

1h 2h

153

Fig. 6. Control follicles were preincubated for 1 h at 37° C, washed and incubated with 3H-galactose, 3H_N_ acetylmannosamine or 3H-leucine for another hour. A second set of follicles was similarly treated but monensin was present during the incubation with the labeled compounds (" monensin 1 h "). In a third set of follicles monensin was present both during the preincubation and the labeling (" monensin 2 h")

3H - Galactose 3H·Sialic acid 3 H - Leucine

inhibition of 3H-sialic acid incorporation may be due to retarded conversion of 3H-N-acetylmannosamine into 3H_ sialic acid at 20° C

Secretion experiments were also performed in which fol­licles were continuously incubated with 3H -N -acetylmannos­amine for 1 h at 37° C (Fig. 4,A). The results agree well with corresponding galactose experiments run in parallel: There was a considerable secretion of labeled thyroglobulin in control follicles and a very small release in the presence ofmonensin. The incorporation of 3H-sialic acid into thyro­globulin was reduced by about 50% (Fig. 6) in the presence of monensin.

Effect of monensin on the incorporation of 3 H-leucine, 3 H­galactose and 3 H-sialic acid into thyroglobulin. Incorpora­tion into thyroglobulin of both 3H-galactose and 3H-sialic acid was strongly and similarly reduced in experiments in which monensin was present during incubation at 3r C for 1 h with the labeled sugars. Since it seemed possible that this reduction was secondary to inhibition of protein synthesis the effect of monensin on the incorporation of 3H-Ieucine into thyroglobulin was compared with the effect of monensin on the incorporation of 3H-galactose and 3H_ sialic acid. In these experiments (Fig. 6) follicles were prein­cubated at 37° C for 1 h in the presence or absence of mon­ensin, washed and incubated with 3H-Ieucine, 3H-galactose or 3H-N-acetylmannosamine for 1 hat 37° C with or with­out monensin; total thyroglobulin-bound radioactivity (fol­licle+medium) was measured. The results show that 3H_ leucine incorporation was not affected when monensin was present only during the labeling hour but was reduced by about 40% when the follicles were preincubated with the ionophore for 1 h. The incorporation of 3H-galactose and 3H-sialic acid was reduced by about 50% when monensin was added after preincubation and by 70% (galactose) when present also during preincubation.

Discussion

Two recent studies from our laboratory, performed on the same system of isolated follicles as used in the present study,

are of importance for the interpretation of our observa­tions. In one of these studies (Ring et al. 1987b) we exam­ined the effect of monensin on the secretion of 3H-Ieucine­labeled thyroglobulin. The observations were in agreement with the concept that monensin blocks the intracellular transport of thyroglobulin in the Golgi complex and showed furthermore that monensin does not, on the other hand, inhibit the transport of thyroglobulin from the distal side of the block to the apical cell surface or the exocytosis of thyroglobulin. The other study (Ring et al. 1987 a) showed that intracellular transport and secretion of thyro­globulin is almost completely but reversibly blocked when the temperature is lowered to 20° C. The protein is retained in the compartment where it was when the temperature was lowered. On the other hand, the incorporation of 3H_ galactose into thyroglobulin is only moderately reduced at 20° C Thus, it is possible to label thyroglobulin efficiently with 3H-galactose without the risk of movement of labeled thyroglobulin out of the labeling compartment.

The experiments in the present study in which follicles were labeled with 3H-galactose at 20° C should be judged with respect to the findings in these two previous studies. Chase incubation at 37° C of labeled follicles resulted in a substantial secretion of labeled thyroglobulin. When mon­ensin was present during the chase the secretion was drasti­cally reduced. The obvious interpretation of this finding is that galactose is incorporated into thyroglobulin in a compartment proximal to the monensin-induced block. Moreover, the degree of secretion inhibition in these experi­ments was similar to the inhibition observed after pulse­labeling with 3H-Ieucine and chase for 4 h in the presence of 1 11M monensin (Ring et al. 1987b). Since this inhibition should represent the maximal reduction of thyroglobulin transport by 1 11M monensin, the same inhibition in the present galactose experiments indicates that galactose incor­poration into thyroglobulin occurs solely on the proximal side of the monensin block.

In similarly designed prelabeling-chase experiments with 3H-sialic acid precursor, monensin inhibited secretion of labeled thyroglobulin to approximately the same degree as in the galactose experiment. It turned out, however, that

154

in these experiments almost no incorporation of 3H-sialic acid occurred during the incubation at 20° C but took place during the chase at 37° C. Experiments were therefore un­dertaken in which the secretion of labeled thyroglobulin was measured in follicles that were continuously incubated, in the absence or presence of monensin, with 3H-galactose or 3H-sialic acid precursor at 37° C for 1 h. It was found that in the presence of monensin the total incorporation into thyroglobulin of 3H-galactose and 3H-sialic acid, re­spectively, was similarly inhibited. The secretion (calculated in percent of total labeled thyroglobulin) of 3H-galactose­labeled thyroglobulin was inhibited by monensin to about the same degree as in the prelabeling-chase experiments and the secretion of 3H-sialic acid-labeled thyroglobulin to the· same degree as the secretion of 3H-galactose-Iabeled thyro­globulin.

The conformity of the results of the experiments in which follicles were continuously incubated with 3H-galac­tose and 3H-N-acetylmannosamine, respectively, indicates that sialic acid, like galactose, is incorporated into thyro­globulin proximal to the monensin block. However, since readily interpretable prelabeling-chase experiments similar to those with 3H-galactose could not be performed with 3H-sialic acid precursor, the possibility should be consid­ered that a significant proportion of sialic acid might be incorporated into thyroglobulin in a compartment distal to the monensin block. As previously shown, there is a continuous discharge of thyroglobulin from the post-block compartments in the presence of monensin (Ring et al. 1987b); this would reduce the amount of thyroglobulin available for sialic acid incorporation and, hence, also re­duce the absolute amount of secreted labeled thyroglobulin. However, labeling in a post-block compartment can hardly explain the observed strong inhibition of secretion, since secretion was calculated as the amount of labeled thyroglo­bulin in the medium at the end of incubation relative to the amount of labeled thyroglobulin remaining in the cells at the end of incubation + the released labeled thyroglobu­lin. Thus, it seems justified to conclude that most, and possi­bly all, sialic acid is incorporated into thyroglobulin in a compartment proximal to the monensin block.

The incorporation of galactose and sialic acid into thy­roglobulin was considerably reduced within 1 h in the pres­ence ofmonensin. The mechanism of this effect ofmonensin is not clear. Lack of substrate due to decreased thyroglobu­lin synthesis cannot be the cause since incorporation of leucine into thyroglobulin was measurably affected only after 2 h exposure to the ionophore. Also, lack of substrate due to location of the incorporation of galactose and sialic acid in a post-block Golgi compartment, from which thyro­globulin is co.ntinuously drained, can be excluded for reasons discussed above. A third cause of substrate deficiency could be decreasing transfer of thyroglobulin from the RER when the Golgi compartments become dilated by monensin; this alternative is to some extent supported by observations in­dicating accumulation of protein in the RER cisternae dur­ing prolonged monensin treatment (Burditt et al. 1985). An­other possibility, suggested by Strous et al. (1983), could be that the interaction between the transferase and the thy­roglobulin molecules is disturbed because of changed spa­tial interrelations in the highly dilated Golgi cisternae.

Numerous studies have been published concerning the location of terminal glycosylation of glycoproteins in rela­tion to the site of the monensin block. Two major groups

of results can be discerned: In one group the observations have been interpreted to show that the site of terminal gly­cosylation is distal to the monensin block (e.g., Tartakoff and Vassalli 1979; Tartakoff et al. 1981; Griffiths et al. 1983; Edwardson 1984), in the studies belonging to the other group terminal glycosylation has been found to occur proximal to the monensin block (e.g., Crine and Dufour 1982; Pesonen and Kiiiiriiiinen 1982). It should also be not­ed that secretory and membrane proteins may be terminally glycosylated at different sites in relation to the monensin block in the same cell (Strous and Lodish 1980), that a certain virus protein may acquire its terminal sugars at dif­ferent sites in relation to the monensin block in different virus-infected cells (cf. Pesonen and Kiiiiriiinen 1982; Grif­fiths et al. 1983) and that the site of terminal glycosylation may differ between two virus membrane proteins in the same cell (Kuismanen et al. 1985). Some of these discrepan­cies could be due to differences in experimental design, but other certainly represent true dissimilarities. It seems there­fore, as pointed out by Griffiths et al. (1983), that observa­tions on the location of terminal glycosylation obtained with monensin in one system cannot directly be applied to another system. Consequently, the present observations per se do not allow a precise description in structural terms of the Golgi subsite of the monensin block in thyroid follicle cells or the subsite(s) of incorporation of galactose and sialic acid into thyroglobulin.

However, recent studies have elucidated the location of the terminal glycosylation reactions in some cells by im­munocytochemical demonstration of glycosyltransferases and identification of sugar residues by specific lectins. Con­cerning galactose, Roth and Berger (1982) and Slot and Geuze (1983) have shown in HeLa cells and hepatoma cells, respectively, that galactosyltransferase is located in trans­Golgi cisternae, and Griffiths et al. (1982) have demon­strated that viral membrane proteins acquire galactose in trans-Golgi cisternae in kidney cells of the baby hamster. Sialyltransferase has been localized in trans-cisternae and connected "trans-tubular network" in liver hepatocytes and sialic acid residues have been demonstrated in the same compartments (Roth et al. 1985). (Observations by indirect immunofluorescence by Berger and Hesford (1985) suggest that sialyltransferase in cultured kidney cells and fibroblasts may be located in peripheral Golgi vesicles and "trans­tubular network" distal to the trans-Golgi cisternae con­taining galactosyltransferase.) These recent observations seem to justify the notion that the trans-Golgi cisternae may be the loci of galactosylation and sialylation of thyro­globulin. (A" trans-tubular network" has not been identified in thyroid follicle cells.) If this assumption is valid it would imply that in thyroid follicle cells the monensin block is located between the trans-Golgi cisternae and the exocytic vesicles since we have shown that discharge of thyroglobulin from these vesicles is not inhibited by monensin (Ring et al. 1987b).

The exocytic vesicles in the follicle cells carry not only thyroglobulin but also membrane from the trans-Golgi cisternae to the apical cell surface. The membrane of the vesicles bears enzymes destined to exert their activity at the apical cell surface after fusion of the vesicle membrane with the plasma membrane. An enzyme belonging to this category is peroxidase (Bjorkman et al. 1976), a reactant in thyroglobulin iodination, a process that takes place on the apical plasma membrane (Ekholm and Wollman 1975).

However, the present study indicates that no incorporation of galactose or sialic acid into thyroglobulin occurs in the exocytic vesicles or at the apical cell surface; moreover, this notion is supported by the lack of immunocytochemi­cally demonstrable galactosyltransferase and sialyltransfer­ase in membranes outside the Golgi complex in other cell systems (Roth and Berger 1982; Slot and Geuze 1983; Roth et al. 1985). This would imply that domains of the mem­brane of the trans-Golgi cisternae that contain these trans­ferases are excluded from the membrane portions that form exocytic vesicles. It appears therefore that galactosylation and sialylation of thyroglobulin are completed within the Golgi complex. Consequently, the observations in the pres­ent study do not lend support to the hypothesis presented by Consiglio et al. (1979) and Consiglio et al. (1981) imply­ing that a receptor for asialothyroglobulin may serve in directing asialothyroglobulin from the Golgi complex to the apical cell surface and retain thyroglobulin molecules on this membrane until sialylation is completed and iodina­tion initiated.

References

Arima R, Spiro RG (1972) Studies on the carbohydrate units of thyroglobulin. J Bioi Chern 247:1836-1848

Bennet G, O'Shaughnessy D (1981) The site of incorporation of sialic acid residues into glycoproteins and the subsequent fates of these molecules in various rat and mouse cell types as shown by radioautography after injection of 3H-N-acetylmannosa­mine. I. Observations in hepatocytes. J Cell Bioi 88: 1-15

Bennet G, Kan FWK, O'Shaughnessy D (1981) The site of inc or po­ration of sialic acid residues into glycoproteins and the subse­quent fates of these molecules in various rat and mouse cell types as shown by radioautography after injection of [3H]-N­acetylmannosamine. II. Observations in tissues other liver. J Cell Bioi 88: 16-28

Berger EG, Hesford FJ (1985) Localization of galactosyl- and sia­lyltransferase by immunofluorescence: Evidence for different sites. Proc Natl Bioi Sci USA 82:4736-4739

Bjorkman U, Ekholm R (1982) Effect of tunicamycin on thyro­globulin secretion. Eur J Biochem 125: 585-591

Bjorkman U, Ekholm R (1984) Generation of H 2 0 2 in isolated porcine thyroid follices. Endocrinology 115: 392-398

Bjorkman U, Ekholm R, Ericson LE, Ofverholm T (1976) Trans­port of thyroglobulin and peroxidase in the thyroid follicle cell. Mol Cell Endocrinol 5:3-17

Bouchilloux S, Gabrion J, Delori P (1981) Antibodies to {3-1,4-galactosyltransferase from porcine thyroid microsomes. Immu­nohistochemical observation on thyroid tissue and cells. Bioi Cell 40:143-146

Burditt LJ, Ratcliffe A, Fryer PR, Hardingham TE (1985) The intracellular localisation of proteoglycans and their accumula­tion in chondrocytes treated with monensin. Biochim Biophys Acta 844:247-255

Chabaud 0, Bouchilloux S, Ronin C, Ferraud M (1974) Localiza­tion in a Golgi-rich thyroid fraction of sialyl-, galactosyl- and N -acetyl-glucosaminyltransferases. Biochimie 56: 119-130

Consiglio E, Salvatore G, Rail EJ, Kohn LD (1979) Thyroglobulin interaction with thyroid plasma membranes. J Bioi Chern 254:5065-5076

Consiglio E, Shifrin S, Yavin Z, Ambesi-Impiombato FS, Rail EJ, Salvatore G, Kohn LD (1981) Thyroglobulin interactions with thyroid membranes. J Bioi Chern 256: 10592-10599

Crine P, Dufour L (1982) Effects of monensin on the processing of pro-opiomelanocortin in the intermediate lobe of the rat pituitary. Biochem Biophys Res Commun 109: 500-506

Denef J-F, Bjorkman U, Ekholm R (1980) Structural and func-

155

tional characteristics of isolated thyroid follicles. J Ultrastruct Res 71: 185-202

Eagle H (1959) Amino acid metabolism in mammalian cell cultures. Science 130:432-437

Edwardson JM (1984) Effects of monensin on the processing and intracellular transport of influenza virus haemagglutinin in in­fected MDCK cells. J Cell Sci 65: 209-221

Ekholm R, Wollman SH (1975) Site of iodination in rat thyroid gland deduced from electron microscopic autoradiographs. En­docrinology 97: 1432-1444

Fukuda M, Egomi F (1971) The structure of a glycopeptide puri­fied from porcine thyroglobulin. Biochem J 123:407-415

Godelaine D, Spiro MJ, Spiro RG (1981) Processing of the carbo­hydrate units of thyroglobulin. J Bioi Chern 256: 10161-10168

Griffiths G, Brands R, Burke B, Louvard D, Warren G (1982) Viral membrane proteins acquire galactose in trans Golgi cister­nae during intracellular transport. J Cell Bioi 95: 781-792

Griffiths G, Quinn P, Warren G (1983) Dissection of the Golgi complex. I. Monensin inhibits the transport of viral membrane proteins from medial to trans Golgi cisternae in baby hamster kidney cells infected with Semliki forest virus. J Cell Bioi 96:835-850

Haddad A (1972) In vivo incorporation of 3H-galactose by thyroid follicular cells of the rat, as shown by electron microscopic radioautography. J Histochem Cytochem 20:220-224

Haddad A, Smith MD, Herscovics A, Nadler NJ, Leblond CP (1971) Radioautographic study of in vivo and in vitro incorpo­ration of fucose- 3H into thyroglobulin by rat thyroid follicular cells. J Cell Bioi 49: 856-882

Kuismanen E, Saraste J, Pettersson RE (1985) Effect of monensin on the assembly of Uukuniemi virus in the Golgi complex. J Virol 55: 813-822

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with Folin phenol reagent. J Bioi Chern 193:265-275

McIntire FC, Sproull MF (1957) A simple method for determina­tion of desoxypentosenucleic acid in tissue cultures. Proc Soc Exp Bioi Med 95:458-461

Pesonen M, Kiiiiriiiinen L (1982) Incomplete complex oligosacchar­ides in Semliki forest virus envelope proteins arrested within the cell in the presence ofmonensin. J Mol Bioi 158:213-230

Pressman BC, Fahim M (1982) Pharmacology and toxicology of the monovalent carboxylic ionophores. Ann Rev Pharmacol Toxicol 22: 465-490

Ring P, Bjorkman U, Ekholm R (1987a) Effect of cooling on intracellular transport and secretion of thyroglobulin. Cell Tis­sue Res 247: 505-513

Ring P, Bjorkman U, Ekholm R (1987b) The effect of monensin on thyroglobulin secretion. Studies with isolated follicles from pig thyroids. Cell Tissue Res 248: 153-160

Roth J, Berger EG (1982) Immunocytochemical localization of galactosyltransferase in HeLa cells: codistribution with thia­mine pyrophosphate in trans-Golgi cisternae. J Cell Bioi 93:223-229

Roth J, Douglas JT, Lucocq JM, Weinstein J, Paulson JC (1985) Demonstration of an extensive trans-tubular network continu­ous with the Golgi apparatus stack that may function in glyco­sylation. Cell 43: 287-295

Slot JW, Geuze HJ (1983) Immunoelectron microscopic explora­tion of the Golgi complex. J Histochem Cytochem 31: 1049-1056

Spiro RG (1965) The carbohydrate units of thyroglobulin. J Bioi Chern 240:1603-1610

Spiro MJ, Spiro RG, Bhoyroo VD (1976a) Lipid-saccharide inter­mediates in glycoprotein biosynthesis. I. Formation of an oligo­saccharide-lipid by thyroid slices and evaluation of its role in protein glycosylation. J Bioi Chern 251: 6400-6408

Spiro MJ, Spiro RG, Bhoyroo VD (1976 b) Lipid-saccharide inter­mediates in glycoprotein biosynthesis. II. Studies on the struc­ture of an oligosaccharide-lipid from thyroid. J Cell Bioi 251:6409-6419

156

Spiro MJ, Spiro RG, Bhoyroo VD (1976c) Lipid-saccharide inter­mediates in glycoprotein biosynthesis. III. Comparison of oligo­saccharide-lipids formed by slices from several tissues. J Cell Bioi 251:6420-6425

Spiro MJ, Spiro RG, Bhoyroo VD (1979) Glycosylation of proteins by oligosaccharide-lipids. J Bioi Chern 254: 7668-7674

Strous G, Lodish H (1980) Intracellular transport of secretory and membrane proteins in hepatoma cells infected by vesicular sto­matitis virus. Cell 22: 709-717

Strous GJAM, Willemsen R, van Kerkhof P, Slot JW, Geuze HJ, Lodish HF (1983) Vesicular stomatitis virus glycoprotein, albu­min and transferrin are transported to the cell surface via the same Golgi vesicles. J Cell Bioi 97: 1815-1822

Tartakoff AM (1983a) Perturbation of vesicular traffic with the carboxylic ionophore monensin. Cell 32: 1026--1028

Tartakoff AM (1983 b) Perturbation of the structure and function of the Golgi complex by monovalent carboxylic ionophores. Methods EnzymoI98:47-59

Tartakoff A, Vassalli P (1979) Plasma cell immunoglobulin M mol­ecules. J Cell Bioi 83: 284-299

Tartakoff A, Hoessli D, Vassalli P (1981) Intracellular transport of lymphoid surface glycoproteins. J Mol Bioi 150: 525-535

Van den Hove M-F, Couvreur M, DeVisscher M, Salvatore G (1982) A new mechanism for the reabsorption of thyroid iodo­proteins: Selective fluid pinocytosis. Eur J Biochem 122:415-422

Whur P, Herscovics A, Leblond CP (1969) Radioautographic visu­alization of the incorporation of galactose-3H and mannose- 3 H by rat thyroids in vitro in relation to the stages of thyroglobulin synthesis. J Cell Bioi 43:289-311

Accepted February 19, 1987