Enzymatic deglycosylation of human thyroglobulin: fluorescence studies

Enzymatic deglycosylation of human thyroglobulin: fluorescence studies

Biochimica et Biophysica Acta, 957 (1988) 105-110 105 Elsewer BBA33210 Enzymatic deglycosy|afion of human thyrogiobulin: fluorescence studies Sett...

457KB Sizes 0 Downloads 78 Views

Biochimica et Biophysica Acta,

957 (1988) 105-110

105

Elsewer BBA33210

Enzymatic deglycosy|afion of human thyrogiobulin: fluorescence studies Settimio Grimaldi a, Dcleana Pozzi b, Roberto Verna c, Serafino Lio Gabriella Giganti d, Roberto De Pirro e and Fabrizio Monaco :

a,

a lstituto di Medicina Sperimentale del CNR, Roma, h Dipartimento di Medicina Sperimentale, Universit~ 'La Sapienza', Roma, c Cattedra di Chimica e Microscopia Clinica, Universitd di L "Aquila, Aquila, d Dipartimento di Medicina Sperimentale, i i Universit~ di Roma, Tot Vergata, Roma, e Cattedra di Endocrinologia, Universitd di Ancona, Ancona. / Caitedra di Endocrinologia, Universit~ di Chieti, Chieti (Italy)

(Received27 January 1988) (Revisedmanuscriptreceived20 June 1988)

Key words: Thyroglobulin;Proteinfluorescence;Deglycosylation;Proteinconformation;(Human) The interaction between the carbohydrate and the amino acid residues in human thyroglobu|in has been studied. Previous reports showed that the removal of the two terminal carbohydrates of the complex chains leads to an increase in thyroglobulin binding to thyroid membranes. In our study, after enzymatic release w i ~ glycosJdases of the sugar moieties from thyroglobulin, a time.dependent decrease in tryptophan fluorescence has been observed. This decrease was also associated with a shift in the emission peak from 335 to 340 nm. The strong quenching of tryptophan emission was also accompanied by a decrease in the exposure of tryptophan residues, as shown by a Stern-Volmer analysis with the neutral quencher acrylamide. These data, together with the increase in fluorescence of the dansylated deg|ycosylafed thyroglobolin, strongly suggest that a significant conformational change of thyrogiobulin follows the deg|ycosylation of the protein.

Introduction

Thyroglobulin, the specific thyroid prohormone is a glycoprotein with a molecular mass of 660 kDa with a sedimentation coefficient of 19 S, in which thyroid hormone formation occurs [1]. About 10% of the molecular weight of human thyroglobulin is represented by carbohydrates distinct in two chains, the simple one is called Unit A and the complex one Unit B. It has been demon-

Abbreviation: DNS, dansyl(5-dimethylamino-l-naphthalenesulphonyl). Correspondence:F. Monaco,Cattedra di Endocrinologia,Universit~tdi Chieti.OspedaleSan CamilloDe Lellis,66100Chieti, Italy.

strated that the glycosylation of thyroglobulin is a prerequisite for thyroid hormone formation [2], since formation of the carbohydrate chains allows, during exocytosis, the molecule to migrate from the Golgi apparatus to the apical cell membrane [3,4] where iodination and thyroid hormone formation occur [5]. It has been also reported that enzymatic deglyo cosylation of thyroglobulin modifies the immunochemical properties of the molecule [6] with a modification in the binding of the aglycothyroglobulin to thyroid membranes [7,8]. Furthermore, several studies indicate that the modification of the carbohydrate chains of thyroglobulin results in structural modifications of the deglycosylated protein with the formation of iodoprotein aggregates [8,9].

016%4838/88/$03.50 © 1988 ElsevierSciencePublishers B.V.(BiomedicalDivision)

106

Aim of the present study was to verify whether the complete enzymatic deglycosylation of human thyroglobulin could modify some of the structural .properties of the molecule, and therefore interfere with the iodotyrosine-iodotyrosine interaction, thus affecting the hormonogenetic efficiency of the glycoprotein prohormone [7,8].

Ultraviolet fluorescence Fluorescence spectra and intensities were obtained with a Perldn-Elmer 650-40 spectrofluorometer at 20 ° C; excitation and emission wavelength were 280 and 340 nm for tryptophan, and 340 and 490 nm for DNS, respectively [11].

Acrylamide quenching analysis Materials and Methods

Acrylamide fluorescence quenching data were analysed according to the Stern-Volmer equation:

Thyroglobulin purification

F/Fo=1+ K~(Q)

Tissue specimen from normal thyroid were obtained from patients subjected to neck surgery, who were not affected by thyroid diseases. Thyroglobulin was purified as described previously [10]. Briefly, the thyroid tissue was gently homogenized in 0.1 M Tris/0.25 M sucrose (pH 7.4) at 4 ° C. Thyroglobulin was isolated from the 1.4-1.8 M (NH4)~SO4 fractionation; the 1.4-1.8 M pellet containing the thyroglobulin fraction was subsequently purified by sucrose gradient ultracentrifugation (Fig. 1). The purity of the protein was then confirmed electrophoretically and by analytical uitracentrifugation.

were F and Fo are the tryptophanyl intensities in the absence and presence of a quencher (Q); K~v is the dynamic quenching constant [12].

Enzymatic deglycosylation A solution of thyroglobulin (40 #g/ml in 0.1 M phosphate buffer (pH 6.0)), was incubated at 37 ° C for 2 h with 1/~1 of mixed glycosidases (5 mg/ml) and mannosidase (1 unit/ml) in the same buffer. Glycosidases were kindly provided by Dr. G. Ashwell (NIADDK, NIH). They were extracted from Diplococcus Pneumoniae and were free of any proteinase activity, as determined by proteinase assays. The reaction was followed in the fluorometer as described [13].

Carbohydrates analysis Sialic acid, that is released from thyroglobulin by enzymatic digestion, was measured by the thiobarbituric acid assay [14]. Hexosamines were determined by a Durum analyser [15] and neutral sugars were identified and quantitated by partition chromatography, as described in Ref. 15. The carbohydrate content of native and deglycosylated thyroglobulin is reported in Table I.

0.6 0,5

f0 J~

I.

,13

oa 275

0.2

Analytical ultracentrifugation

O,1

1

5

10 15 Fraction

20

25

Fig, 1, Elution profile at 280 nm of native thyroglobulin purified by preparative ultracentrifugation on a 10-40~ sucrose gradient at 4 ° C for 24 h at 27000 rpm.

Analytical ultracentrifugation was performed in a double sector cell at 22°C in a Spinco Model E analytical ultracentrifuge equipped with Schlieren optics and a temperature-control unit. The sedimentation coefficients were calculated following the standard procedure, using the correction for water at 22°C and calculating the Johnston-Ogston effect [16].

107 TABLE I CARBOHYDRATE CONTENT OF NATIVE AND DEGLYCOSYLATED THYROGLOBULIN Sugars

Sialic acid Mannose Galactose Hex6samine

Residues per moles of protein Tg a

dTg b

35 100 60 90

0 8 10 6

a Thyroglobulin. b Deglycosylatedthyroglobulin.

Polyacrylamide gel electrophoresis Slab-gel electrophoresis was performed on a 4% acrylamide slab gel in a discontinuous pH gradient (upper buffer Tris 0.1 M/glycine 0.1 M (pH 8.9); lower buffer Tris 0.1 M/glycine 0.1 M pH (8.1)) at 4 ° C using a constant power of 35 mA.

DNS labelling Microliter amounts of DNS in acetone were added to a protein solution in 0.1 M Na2HCO3 (pH 8.9) at 4 ° C in the dark as previously reported [15]. After 1 h incubation with occasional shaking, the unbound DNS was removed by gel filtration on a Sephadex G-25 column. The number of moles of bound DNS was 5.0, as determined by ab~orbance measurements at 340 nm using 3360 as the molar absorption coefficient of the dansyl group [15].

~100

,,

t~

~

&

,$

80

g 70

~o 6o ~- 5 0

._~ o~

40 , I

60

q

I

18o

i Time

3~'0

I

4~o

I

I

I

~4o

6'6o

(min)

Fig. 2. Rate of loss in tryptophan fluorescenceintensity of thyroglobulin at 340 nm with or without glycosidases. A, the effect of the mixed glycosidaseson thyroglobulinat 37 o C; A, thyroglobulin without enzymes.The reaction was performed at pH 6.0 in 0.10 M phosphate at 37 o C. The concentrationof the protein was 40 #g/ml.

3). The increase in quantum yield of the DNS fluorophore could be related to a change in hydrophobicity of the microenvironment of the probe labelled on the deglycosylated protein. [11].

Effect of deglycosylation on the tryptophan emission spectra of thyroglobulin The loss in fluorescence of thyroglobulin following deglycosylation was accompanied by a shift in the emission spectrum from 335 to 340 rim, (Fig. 4). This result suggests that the loss in fluorescence of thyroglobulin following deglycosylation can originate from a molecular rearrangement at

,

Results

5O

Fluorescence changes accompanying enzymatic deglycosylation of thyroglobulin Intrinsic emission: Tryptophan. The rate of the

40

enzymatic deglycosylation of thyroglobulin was accompanied by a decrease in thyroglobulin emission intensity. The decrease in tryptophan emission intensity with enzymatic deglycosylation of thyroglobulin at p H 6.0 in 0.1 M phosphate buffer is shown in Fig. 2. The decrease observed in several preparations of thyroglobulin was close to 50% after 2 of incubation. Extrinsic emission: DNS. Deglycosylation of thyroglobulin covalently labelled with D N S produced a 20% increase in DNS fluorescence (Fig.

,,

~ 9o

i

i

,

s

,

,

I ~

8 ~ 30

*~ 20

Time

(rain)

Fig. 3. Rate of change in fluorescenceof DNS-labelled thyroglobulin digested with mixed glycosidases (4) and withou! enzymes (A). Conditions: 0.1 M phosphate buffer (pH 6.0) at 37°C. The protein concentration was 40 #g/m]. I tool of DNS was covalently labelled to thyroglobulin.Wavelengthsof excitations and emissionwere 340 and 490 nm respectively.

108 I

I

I

I

I

10

60

(V =- 6

40 I.

0

0

~ 3o _o G; re 20 i/

I

10 I

300

I

I

I

I

3 t 0 320 3 3 0 3 4 0 Wovelenght (riM)

I

I

350

360

Fig. 4. Emission spectra of native thyrogiobulin ( ) and deglycosylated thyroglohulin ( - - - - - - ) Conditions: 0.10 Tris (pH 8.0) at 24 ° C. Excitation wavelength, 280 nm.

the tryptophan residues in deglycosylated thyroglobulin. After normalization and subtraction of the two spectra, a peak at 320 nm was found (Fig. 5), demonstrating that the fluorescence of the less-expected tryptophan residues is not contributing to the overall fluorescence after deglycosylation. The loss in quantum yield of tryptophan emitting at short wavelengths increases the relative contribution at lower wavelengths. Since thyroglobulin is an iodoprotein containing a significant amount of iodotyrosine residues, the 5 nm red-sift, together with the fluorescence decrease, can be interpreted in terms of a conformational change of the molecule. In this conformational change some tryptophan residues, emitting at lower wavelengths, could come closer to the iodothyronine molecules where the heavy iodine atoms can quench the tryptophan fluorescence [11].

Acrylamide quenchingof the tryptopna, fluorescence In order to determine whether deglycosylation of thyroglobulin modifies the structure of the

,

310

320

Wovelenght

330

(riM)

Fig. 5. Difference emission spectra after normalization of the emission spectra of native thyroglobulin and deglycosylated thyroglobulin. Other conditions as in Fig. 4.

thyroglobulin molecule, the quenching effect of the acrylamide on the tryptophan emission was measured (Fig. 6). The value of Ksv for thyroglobulin is 14.6 M-1 compar~t to that of deglycosylated thyroglobulin that is ~ M-]. The decrease in collisional quenching constant in deglycosylated thyroglobulin suggests that after deglycosylation the tryptophan residues in deglycosylated

I

I

I 0.25

I

I

I

I

I

I

i

i 1.75

FOF 3.C



1.0 I I I 0.75 1.25 Acrylomi(:le (raM)

Fig. 6. Quenching of fluorescence ( F ) of native thyroglobulin (A) and de~ycosylated thyroglobulin (z~) by acrylamide. Conditions: 0.1 M Tris, (pH 8.0) at 24 ° C. Fo is the fluorescence intensity in the absence of acrylamide. Emission and excitation wavelengths were 280 and 340 nm, respectively. The concentration of protein was 40/~g/ml.

109

thyroglobulin are less exposed to the solvent than in the native protein [11].

Hydrodynamic studies The sedimentation velocity pattern of native and deglycosylated thyroglobulin showed a sedimentation coefficient of 19 S for the native protein, and a decrease of 1 Svedberg unit for the deglycosylated protein (data not shown). Since deglycosylation leads to a significant increase in isoelectric point [18], resulting from the loss of sialic acid, we can exclude deglycosylation as a cause of aggregation of thyroglobulin to greater than 19 S. Discussion

The functional role of the carbohydrate units of glycoproteins has received considerable attention in the last years [19-22]. Recently it has been suggested that the o~igosaccharide chains of thyroglobulin could be involved directly in the secretion of the newly synthesized protein, and indirectly in thyroid hormonogenesis [2,6]. The enzymatic deglycosylation of the two ultimate sugars, sialic acid and galactose, with the exposure of N-acetylglucosamine residues as the terminal sugar, has been shown to increase the binding of the protein to thyroid membranes [7,8]. Furthermore, it has been recently demonstrated that the completion of the carbohydrate units of thyroglobulin is important for the protein secretion, since in tunicamycin-treated thyroid cells thyroglobulin secretion and iodination were clearly inhibited [4]. It has therefore been postulated that the formation of carbohydrate chains of thyroglobulin are involved in thyroglobulin secretion and iodination [2,4]. In this study the interaction of the carbohydrate chains of thyroglobulin with its apoprotein has been investigated. Twe parameters are influencing the hydrodynamicity and the fluorescent behaviour of deglycosylated thyroglobulin: the first is the loss of 10% of the mass of thyroglobulin due to the removal of oligosaccharide chains, the second is the modification in the structure of the apoprotein as revealed by the change in the fluorescence quantum yield of the tryptophan and DNS chromophores. The

decrease in Ksv, obtained by the Stern-Volmer analysis of the acrylamide quenching c,n tryptophanyl fluorescence of deglycosylated thyroglobulin, suggests that the release of the carbohydrate chains of thyroglobulin, obtained by complete enzymatic deglycosylation, implies some molecular rearrangements with a decrease in the accessibility of the tryptophan residues. This indicates that the tryptophan chromophore becomes less exposed with deglycosylation and is therefore less hydrated. The quantum yield of tryptophan chromophores decreases significantly when almost all the carbohydrates are enzymatically removed from thyroglobulin. The decrease in tryptophanyl fluorescence, following deglycosylation, is accompanied by a 5 nm red-shift, from 335 to 340 nm, in the tryptophan emission spectrum. A concomitant increase in the quantum yield of the extrinsic DNS probe is also observed. Since the DNS fluorescence rises with the hydrophobicity of the environment [11], the DNS molecule would be expected to reside after deglycosylation in a region of the thyroglobulin molecule that is more excluded from the solvent. The difference in the tryptophanyl emission spectra, obtained by subtracting the quenched by the unquenched spectra, indicates that with deglycosylation the tryptophan residues emitting at lower wavelength are more quenched than the others. Recently it has been demonstrated that modifications of the carbohydrate chains of thyroglobulin result in a decrease in antigenicity [6,9,11] and in a modification of the binding of the asialoagalacto-thyroglobulin to the cell membrane [8,1!]. In the present study it has been shown that the removal of the carbohydrates from thyroglobulin results in a conformational change, suggesting that a similar change in structure could be related to the addition ~f the sugar moiety to the newly synthesized thyroglobulin. It has to be emphasized that recently the addition of carbohydrates to the protein has been proposed to be a prerequisite for thyroglobulin iodination and thyroid hormone formation [3,5]. In conclusion it is suggested that the carbohydrate chains of thyroglobulin play a significant role both in the conformational arrangement of the protein as well as in: (i) its immunogenic properties, (ii) iodination processes, and (iii) in its

110

migration to the apical border of the cell where iodination and thyroid hormone formation occur.

Acknowledgements This work was supported in part by CNR special project 'Oncologia' contract No. 87.0148.144 and by M.P.I. (40~ and 60~ funds) to R.V.

References 1 Bggo, M,C. and Burrow, G.N. (Eds.) (1985) Prog. Endocr. Res. Ther. 2, 1-341. 2 Monaco, F., Salvatore, G. and Robbins, J. (1975) J. Biol. Chem. 250, 1575-1599. 3 Monaco, F. and Dominici, R. (1985) Progr. Endocr. Res. Ther. 2, 115-126. 4 Monaco, F., Dominici, R., Carlucci, C., De Pirro, R., Felli, R. and Roche, J. (1981) C.R. Soc. Biol. 175, 446-451. 5 Monaco, F., Grimaldi, S., Dominici, R. and Robbins, J. (1975) Endocrinology 97, 347-351. 6 Monaco, F. and Andreoli, M. (1976) Acta Endocrinol. (Kbh) 83, 93-98. 7 Shifrin, $., Consiglio, E., Laccetti, P., Salvatore, G. and Kohn, L.D. (1982) J. Biol. Chem. 257, 9539-9547. 8 Shifrin. S., Consigfio, E. and Kohn, L.D. (1983) J. Biol. Chem. 258, 3780-3786.

9 Grimaldi, S., Fusco, A., Olivieri, A., loppolo, A., Iacovacci, P., Carlini, F., Monaco, F. and Roche, J. (1986) C.R. Soc. Biol. 180, 284-289. 10 Monaco, F. and Robbins, J. (1972) Biochim. Biophys. Acta 279, 118-128. 11 Grimaldi, S., Edelhoch, H. and Robbins, J. (1982) Biochemistry 21, 145-151. 12 Davoli, C., Grimaldi, S., Rusca, G., Andreoli, M. and Edelhoch, H. (1982} Biochim. Biophys. Acta 705, 243-248. 13 Grimaldi, S., Edelhoch, H. and Robbins, J. (1985) Biochemistry 2, 3771-3776. 14 Warren, L. (1959) J. Biol. ,::'hem. 234, 1971-1975. 15 Cheng, S.Y., Morrone, S. and Robbins, J. (1979) J. Biol. Chem. 254, 8830-8833. 16 Andreoli, M., Sena, L., Edelhoch, H. and Salvatore, G. (1969) Arch. Biochem. 134, 242-246. 17 Monaco, F., Grimaldi, S., $cuncio, G. and Andreoli, M., (1974) Acta EndocrinoL 77, 517-526. 18 Consiglio, E., Shifrin, S., Yavin, Z., Ambesi impiombato F.$., Rail, J.E., Salvatore, (3. and Kohn, L.D. (1981) J. Biol. Chem. 256, 10592-10599. 19 Morell, A.G., Gregoriadis, G., Sheinberg, H., Hickman, J. and Ashw¢ll, G. (1971) J. Biol. Chem. 246, 1461-1467. 20 Winterburn, P.J. and Phelps, C.F. (1972) Nature 236, 147-151. 21 Marianna, T., Perini, J.M., Houvena Ghel, M.C., Tramu, C., Lamblin, G. and Ronnel, R. (1986) Carbohydr. Res. 151, 7-19. 22 Shifrin, S., Consiglio, E. and Kohn, L.D. (1983) J. Biol. Chem. 258, 3780-3786.