Effects of carbohydrate depletion on the structure, stability and activity of glucose oxidase from Aspergillus niger

Biochimica et Biophysiea Acta, 1080( 1991) 138-142 ~ 19tq ElsevierSciencePublishersB.V. All rightsreserved0167-4838/91/$03.50 ADONIS 0167483891003140

138

BBAPRO 341124

Effects of carbohydrate depletion on the st:ucture, stability and activity of glucose oxidase from Aspergillus niger Henryk M. Kalisz l, Hans-Jiirgen Hecht 2, Dietmar Schomburg 2 and Rolf D. Schmid i t Department of Enzyme Technology, G B F - Gesellschaft fiir Biotechnologische Forschung, flraunschweig (F.R.G.) amt e Department of Molecular Structural Research, GBF - Gesellsehaft fur Biotechnologische Forschung, Braunschweig (F.R G.)

(Received 7 January 1991) (Revised manuscriptreceived 7 June 1991)

Key words: Glucoseoxidase;Deglycosylation,Property:(A. niger)

Glucose oxidase from Aspergillus niger was purified to homogeneity by hydrophobic interaction and ion-exchange chromatography. Approx. 95% of the carbohydrate moiety was cleaved from the protein by incubation of glucose oxidase with endoglycosidase H and a-mannosidase. Cleavage of the carbohydrate moiety effected a 2 3-30% decrease in the molecular weight and a reduction in the number of isoforms of glucose oxidase. No si~atificant changes were observed in the circular dichroism spectra of the deglyeosylated enzyme. Other properties, such as thermal stability, pH and temperature optima of glucose oxidase activity and substrate specificity were not affected. However, removal of the carbohydrate moiety marginally affecled the kinetics of glucose oxidation and stability at low pH. From these results it appears that the carbohydrate chain of glucose oxidase does not contribute significantly to the structure, stability and activity of glucose oxidase.

Introduction The flavoprotein glucose oxidase (GOD) (fl-D-glucose: oxygen-oxidoreductase, EC 1.1.3.41 catalyses the oxidation of fl-D-glucose by molecular oxygen to D-glucono-tS-lactone and hydrogen peroxide. GOD is of considerable commercial importance [1]. The enzyme is used in food processing, in the production of gluconic acid [2] and for the quantitative determination of l> glucose in samples such as blood, food and fermentation products [3,4]. The primary structure of the Aspergillus niger GOD has recently been deduced and its gene isolated and cloned [5,6]. The successful growth of crystals of an enz3qnatically deglycosylated GOD amenable to X-ray diffraction analysis [7] means the elucidation of its three-dimensional structure is imminent. As a first step in the elucidation of the three-dimensional structure of the deglycosylated GOD the effects of carbohydrate depletion on the properties of GOD are described.

Abbreviations:GOD, glucoseoxidase;CD, circulardiehroism;Endo H. endoglyc,vsidaseH. Correspondence: H. Kalisz, Department of Enzyme Technology, GBF, Mascheroder Weg I. W-3300Braunschweig,F.R.G.

Several functions have been proposed for the carbohydrate moiety of glycoproteins, including correct targeting of proteins [8], transport through membranes [9,10], biological function [11,12], immune response [12-14], and stabilisation of the three-dimensional structure of the protein moiety [15-17]. However, removal of the carbohydrate moiety from a variety of glycoproteins had no apparent effect on their enzymatic or structural properties [18,19].

Materials and Methods Materials

GOD from A. niger (Type VII) was purchased from Sigma (Munich, F.R.G.). Peroxidase, endoglycosidase H, a-mannosidase and 2,2'-azino-di-[3-ethylbenzthiazoline-6-sulphonic acidl were from Boehringer Mannheim (Mannheim, F.R.G.). Liquid chromatography media and Phast gels were purchased from Pharmacia LKB (Freiburg, F.R.G.). All other chemicals were obtained from Merck (Darmstadt, F.R.G.). E n z y m e assay

GOD activity was assayed at 420 nm as described previously [20]. One GOD unit is defined as the amount of enzyme that catalyses the oxidation of 1 /~mol

139 glucose to gluconolactone and H202 in 1 min at 25 ° C and oH 6. Protein wa~ determined by the method of Bradford [21] using Coomassie brilliant blue-G reagent (Bio-Rad) with bovine serum albumin as standard. A multiplying factor of 1.30 was used for the glycosylated GOD to correct for the consistently lower values obtained by the Bradford method in comparison to estimates of the dry weight and extinction at 280 nm using t h e value of Swoboda and Massey [22]. Correction for the deglycosylated enzyme was not necessary.

The amino acid composition of GOD was determined with a Biotronit LC5001 amino acid analyser (Munich, F.R.G.) following hydrolysis in 6 M HCI at 105°C for 24-72 h. Cysteine was measured as cystic acid .:~in~ the method of Ref. 26. Tryptophan was measured spectrophotometrically [27]. Automated Edman degradation was performed with an Applied Biosystems Model 470 A gas phase sequencer with an on-line C-18 reverse phase HPLC.

Gel electrophoresis

Circular dichroism

PAGE was performed on a Pharmacia Phast System (Pharmacia LKB Biotechnoiogy AB) according to the manufacturer's recommendations. The isoelectric point of GOD was determined by isoelectric focusing using the Pharmacia Phast System in the pH range 4.0-6.5 according to [23]. Gels were stained for protein with silver by the method of Ref. 24. Activity staining was done according to Ref. 25 using N-methyldibenzopyrazine methyl sulphate salt ('phenazine methosulphate') and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide.

Enzyme purification Commercial GOD was purified to homogeneity by hydrophobic interaction chromatography and by ionexchange chromatography using a Pharmacia FPLC unit. GOD was dissolved in the starting buffer (20 mM phosphate buffer (pH 7.5) with 1 M ammonium sulphate) and applied to a PhenyI-Sepharose column (1.6 x 15 cm). Unbound substances were washed from the column using 5 column volumes of starting buffer. Adsorbed samples were eluted at a flow rate of 2 ml m i n - t with 20 mM phosphate buffer (pH 8). Active fractions were pooled and concentrated as described previously [7] then applied to a Mono-Q HR 10/10 column equilibrated with 20 mM phosphate buffer (pH 8.5). GOD was eluted at a flow rate of 6 ml m i n - t with a linear gradient of increasing salt concentration (1 M NaCI). Fractions containing GOD activity were pooled desalted and concentrated as described above.

Amino acid composition and sequence analysis

Circular dichroism spectra were recorded at 20°C on a Jasco J-600 spectropolarimeter. Spectra in the near UV (300-250 nm) and far UV (250-200 nm) were recorded in cells of 5 and 1 mm pathlength and at enzyme concentrations of 0.90 and 0.45 mg/ml, respectively. Results

Purification and deglycosylation Commercial GOD lrom A. niger was purified to homogeneity (Fig. 1) by hydrophobic interaction and ion-exchange chromatography. The purification steps and yields of GOD are summarised in Table 1. Incubation of the 7urified GOD with a-mannosidase and Endo H effected a decrease in its molecular weight and in the width of the protein band to a single sharp band (Fig. 2). Analysis by gas chromatography demonstrated the presence of about 5% of the original carbohydrate moiety. The ratio (and number of residues) of mannose to N-acetylglucosamine decreased from 12 : 1 (190:16 residues) to 4:1 (8:2 residues). This GOD is 2

2

3

•

5

iS

(kJ)a) 440

-->

232

--~.

3.40

--:b

6?

--->

m u m

Deglycosylation Purified GOD (1 mg) was incubated for 24 h at 3 7 ° C with 6 U a-mannosidase (EC 3.2.1.24) and 20 mU endoglycosidase H (EC 3.2.1.96) in 30 mM potassium phosphate buffer (pH 5). The degree of deglycosylation was assessed electrophoretically. The deglycosylated GOD was purified on a Sephadex G-25 column and on a Mono S column at pH 4.2. Active fractions were pooled, desalted and concentrated by ultrafiltration. As control 1 mg GOD was incubated under identical deglycasylation conditions but in the absence of the glycohydrolases.

Fig. I. Native PAGE of Asperg///usniger GOD. Samplesobtained after each purification step were subjectedto PAGE under non-dissociatingconditionson an 8-25% gradient gel usingthe Pharmacia Phast System. Sampleswere detected by silver staining(lanes 1-3) and activitystainingfor GOD (lanes 4-6). Lanes 1 and 4, commercial GOD; lanes2 and 5, PhenyI-Sepharosepurified GOD; lanes 3 and 6, Phen),l-Sepharoseand Mono-Qpurified GOD. Marker prc~ reins were: ferritin(440 kDa), catalase (232 kDa), lactate debydrogenase(140kDa), bovineserumalbumin(67 kDa).

140 x,

(kOa)

440

-->

232

-->

140

-->

lnculmtion period (h)

TABLE 1

Purification of a commercialpreparation of GOD from Aspergillus n~er

m m EID

;III

One GOD unit is defined as the amount of enzyme that catalyses the oxidation of 1 #mol glucose to gluconolactone and H202 in 1 min as 25 ° C and pH 6.

D

i

Total protein (mg) Commercial 78.9 Pbenyl Sepharose 58.3 Mono-Q 38.5 0

0.5

1

3

5

7

10

GOD activity (U/ml) 19650 17431 13664

Specific Purification activity factor (U/mg) 249 1.00 299 1.20 355 i.43

Yield (%) 100 89 70

24

Fig. 2. Effect of deglycosylatio:t time on the molecular weight of GOD. Samples of GOD were removed from the deglycosylation incubation mixture at the times indicated and were subjected to PAGE under non-dissociating conditions as described in the legend to Fig. !. in lanes 2-8, the upper protein band corresponds to a-mannosidase. referred to as deglycosylated G O D . The specific activity decreased from 355 U mg -I to 315 U mg - t upon deglycosylation. However, as a similar degree of inactivation was observed with the control G O D , inactivation was most probably due to the deglycosylation conditions rather than carbohydrate depletion.

Electrophoretic properties A molecular mass of 157 and 80 kDa was estimated for the glycosylated G O D by P A G E under non-dissociating and dissociating conditions, respectively. The deglycosylated enzyme migrated under identical conditions as a single sharp band with an apparent molecular weight of 110 and 61 kDa, respectively. A molecular mass of 61.15 kDa was estimated from the amino acid analysis. These values were in good agreement with the molecular weight of 63.2 kDa determined from the D N A sequence of G O D [5,6]. The observed decrease of 24-30% in the molecular weight of G O D indicates a

significantly higher carbohydrate content than previously reported [22,28,29]. lsoelectric focusing under native conditions revealed 5 bands between p I 3.97 and 4.16 for native G O D and only 2 bands at p l 4.12 and 4.16 for the deglycosylated enzyme. The decrease in the number of isoforms confirms previous results [29] that variations in the protein-bound carbohydrate content contribute to the isoelectric raultiplicity of glycosylatcd G O D .

Amino acid composition o f N-terminal sequence ana~sis Amino acid compositions and N-terminal sequences of the glycosylated and deglycosylated G O D s were identical and showed good correlation with the composition derived from the D N A sequence [5,6] and with previous analyses [28,30].

Circular dichroism spectra No significant differences were observed in the C D ~pectra of native and deglycosylated G O D ; the elliptic~y of the latter decreasing by 8% at 212 nm and increasing by 8% at 272 nm (Fig. 3).

Enzyme activity and stability The p H and temperature optima were unaffected by the deglycosylation process. More than 90% of the 5 0

.

.

.

'!i 200

210

220

w ~

230

(m)

240

250

25O

~

35O

400

450

"

Fig. 3. Circular dichroism spectra of native (1), control (2) and deglycosylated (3) GOD in (a) far ultraviolet and (b) visible and near ultraviolet region. Spectra were recorded at 20 o C in 50 mM potassium phosphate buffer (pH 6) in 1 and 5 mm cuvettes, respectively. The respective protein concentrations were 0.45 and 0.90 mg/ml.

141 TABLE It Effect of pH on the stability of GOD GOD was incubatedin triplicateat each pH for up to 15 monthsat 4°C and residual activitywas measured under standard conditions of assay as described in Materials and Methods. The results are expressed as a percentageof the initialactivity.Data in parentheses ind;£atesthe time required for 50% inactivation(in days). Buffer(50 raM)

pH

Residualactivity(%) glycosylated deglycosylated GOD GOD 57 37 (234) 73 61

Glycine-HCI

3.0 4.0

Acetate

4.0 5.0 6.0 6.0 7.0 8.0

60 71 84 80 88 41 (305)

48 (360) 50 75 83 87 44 (335)

7.0 8.0 9.0

72 10 (88) <2 (13)

70 7 (74) <2 (12)

36 (287) 5 (25)
29 (I 76) <5 (17)
Phosphate

Tris-HCI

Glycine-NaOH

8.0 9.0 10.0 !1.0 12.0

maximum activity of both enzymes was observed between pH 4.0-7.0, with an optimum at pH 5.5-6.0. Outside this range activity decreased rapidly. The activity of glycosylated and deglycosylated GOD increased only 2-fold from a minimum at 25 ° C to a maximum at 55 ° C. Increases in activity were lower than those observed by Gibson et al. [31], but were similar to those measured by Nakamura and Ogura [32]. The relatively small increases in activity with temperature may be attributed to the decreased solubility of oxygen and increased K,, for oxygen [31] at higher temperatures. Deglyeosylation had no marked effect on the stability of GOD above pH 5 (Table ll). However, the deglycosylated GOD was less stable at low pH. At alkaline pH, particularly at and above pH 9.0, both GODs were relatively unstable. The thermal stability of GOD was also unaffected by the depletion of carbohydrate. The enzyme was stable up to 50 ° C. At 50 *C GOD was inactivated by 30% over a period of 11 h. At 60 ° C the Tt/2 was 60 rain; at 70°C, 5 min. The activation energy for the destruction reaction, E d, was 280 kJ mol-~ (67 kcal mol - m). Substrate specifwity The substrate specificity of GOD was unaffected by the deglycosylation; D-glucose was oxidised at a much

faster rate than 2-deoxy-D-glucose anO D-mannose, whereas L-glucose, o-galactose, D-arabinose, D-xylose were not oxidised. Kinetics Deglycosylation marginally affected the kinetics of glucose oxidation. The Michaclis-Menten constant, Kin, for fl-D-glucose increased by 10% from 30 to 33 mM on deglycosylation. The catalytic constant, k¢~t, decreased from 920 s - ' to 815 s -~. The specificity, or apparent second-order, constant, kcat//Km, was reduced from 30.7 to 24.7 m M - ' s - ~. The Vmx of the deglycosylated GOD (407 U mg -~) was !1% lower than that of the glycosylated enzyme (458 U mg-~). Discussion

GOD from Aspergillus niger is a highly glycosylated protein comprising 190 mannose and 16 N-acetylglucosamine residues. The carbohydrate content accounts for about 24% of its molecular weight. Previous deglycosylation attempts were only partially successful, despite the use of denatured GOD [6] or extremely high amounts of an endoglycosidase [33]. Similar results were obtained in our laboratory when treating GOD with a single glycohydrolase (unpublished results). However, incubation with both Endo H and a-mannosidase enabled the cleavage of about 95% of the carbohydrate moiety of GOD. The remaining 5% carbohydrate moiety, accounting for about 2 kDa of the molecular weight of GOD, comprises 8 mannose and 2 N-acetylglucosamine residues. Hence, since GOD is composed of two identical subunits, only one of the 8 potential N-linked glycosylation sites per monomer, probably at position 388 [6], appears to be glycosylated. The mannose residues may form an unusual type of O-linkage to serine or threonine of GOD [33]. However, direct evidence for this type of linkage in GOD does not exist. Detailed analysis of the carbohydrate moiety and elucidation of the three-dimensional structure of GOD should provide useful information about the types of sugar-protein linkages in GOD. The carbohydrate moiety of various glycoproteins is believed to play an important role in the folding of proteins into a specific configuration [15-17]. However, carbohydrate depletion had no effect on the structures of several glycoproteins [ 18,19]. Deglycosylation did not signific~.ntly affect the three-dimensional structure of GOD, although the small changes of about 8% in its ellipticity indicate a possible loss of secondary structure and an increase in the compactness of its tertiary structure on deglycosylation. However, although the protein-bound carbohydrate moiety may cause a restraining effect on the configuration of GOD, structural changes are likely to occur on the surface of the

142 enzyme and should not influence the secondary structure. Minor structural modifications may account for the changes in the kinetics of glucose oxidation and decreased stability az low pH which accompanied deglycosylation. However, a direct in~uence of the carbohydrate moiety on the kinetics and stability at low pH cannot be ruled out. Other properties, such as thermal stability, pH and temperature optima of GOD activity and substrate specificity were not affected. Thus, the carbohydrate moiety of GOD, like that of other glycoproteins, does not appear to contribute significantly to the biological properties of the enzyme. Moreover, the protein-bound carbohydrate does not appear to play an important role in the folding of the enzyme into a specific configuration.

Acknowledgements We thank Claudia Stein-Kemmesies for her invaluable technical assistance, Rita Getzlaff for performing the amino acid and N-terminal sequence analyses, Dr. Manfred Nimtz for analysis of the protein-bound carbohydrate content, and Hedwig Schrader and Dr. Roll-Joachim Miiller for help with the CD spectrum analysis. The generous gift of Endo H and a-mannosidase from Boehringer Mannheim is greatly appreciated. References 1 Crneger, A. and Crneger, W. (1984) in Biotechnology, Vol. 6a (Rehm, H.-J. and Reed, G., eds,), pp. 421-457, Verlag Chemic, Weinheim. 2 R6hr, M., Kubieek, C.P. and Kominek, J. (1983) in Biotechnology, Vol. 3 (Rehm, H.-J. and Reed, G., eds.), pp. 455-465, Veda8 Chemie, Weinheim. 3 Turner, A.P.F., Karube, I. and Wilson, G.S. (1987) Editors of Biosensors - Fundamentals and Applications, Oxford University Press, Oxford. 4 Schmid, R.D. and Karube, 1. (1988) in Biotechnology, Vol. 6b (Rehm. H.-J. and Reed, G., eds.), pp. 317-365, Verlag Chemie, Weinheim. 5 Kriechbaum, M., Heilmann, H.J., Wientjes, F.J., Hahn, M., Jan¥, 1C-D., Gassen, H.G., Sharif, F. and Alaeddinoglu, G. (1989) FEELS Lett. 255, 63-66.

6 Frederick, K.R., Tang, J., Emerick, R.S., Masiarz, F.R., Chamberlain, S.H., Vasavada, A., Rosenberg, S., Chakraborty, S., Schopfer, L.M. and Massey, V. (1990) J. Biol. Chem. 265, 37933802. 7 Kalisz, HM., Hecht, H.-J., Schomburg, D. and Schmid, R.D. (1990) J. Mol. Biol. 213, 207-209. 8 Neufeld, E.F. and Ashwell, G. (1980) in The Biochemistry of Glycoproteins and Proteoglycans (Lennarz, W.J., ed.), pp. 241266, Plenum Press, New York. 9 Tabas, I., Schlesinger, S. and Kornfeld, S. (1978) J. Biol. Chem. 253, 716-722. 10 Fischer, H.D., Gonzalez-Noriega, A., Sly, W.S. and Morre, D.J. (1980) J. Biol. Chem. 255, 9608-9615. 11 Kalyan, N.K. and Bahl, O.P. (1983) J. Biol. Chem. 258, 67-74. 12 Rademacher. T.W., Parekh, R.B. and Dwek, R.A. (1988) Annu. Rev. Biochem. 57, 785-838. 13 Pazur, J.H. and Aronson, N.N., Jr. (1972) Adv. Carbohydr. Chem. Biochem. 27, 301-41. 14 Knight, P. f1989) Biotechnology 7, 35-40. 15 Wang, F.F.C. and Hirs, C.H.W.(1977) J. Biol. Chem. 252, 83588364. 16 Chu, F.K., Trimble, R.B. and Maley, F. (1978)J. Biol. Chem. 253, 8691-8693. 17 Barbaric, S., Mrsa, V., Ries, B. and Mildner, P. (1984) Arch. Biochem. Biophys. 234, 567-575. 18 Tarentino, A.L, Plummer, T.H., Jr. and Maley, F. (1974) J. Biol. Chem. 249, 818-824. 19 Trimble, R.B. and Maley, F. (1977) Biochem. Biophys. Res. Commun. 78, 935-944. 20 Kalisz, H.M., Hendle, J. and Schmid, R.D. (1990) J. Chromatog. 521,245-250. 21 Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. 22 Swoboda, B.E.P. and Massey, V. (1965) J. Biol. Chem. 240, 2209-2215. 23 OIsson, I., Axio-Fredriksson, U., Degerman, M. and Oisson, B. (1988) Electrophoresis 9, 16-22. 24 Butcher, L.A. and Tomkins, J.K. (1985) Anal. Biochem. 148, 384-388. 25 Sock, I. and Rohringer, R. (1988) Anal. Biochem. 171,310-319. 26 Moore, 8. (1963) J. Biol. Chem. 238, 235-237. 27 Barman, T.E. and Koshland, D.E., Jr. (1967) J. Biol. Chem. 242, 5771-5776. 28 Tsuge, H., Natsuaki, O. and Ohashi, K. (1975) J Biochem. 78, 835 -843. 29 Hayashi, S. and Nakamura, S. (1981) Biochim. Biophys. Acta 657, 40-51. 30 Jones, M.N., Manley, P. a~d Wilkinson, A. (1982) Biochem. J. 203, 285-291. 31 Gibson, Q.H., Swoboda, B.E.P. and Massey, V. 0964) J. Biol. Chem. 239, 3927-3934. 32 Nakamura, S. and Ogura, Y. (1968) J. Biochem. 63, 308-316. 33 Takegawa, K, Fujiwara, K., lwahara, S., Yamamoto, K. and Tochikura, T. (1989) Biochem. Cell Biol.67, 460-464.