THE JOURNAL 01 B~L~CICAL CHEMISTRY
Vol. 253, No. 2, Issue of January 25, pp. 377-379, 1973
Printed in U.S.A.
Human Blood Group Glycosyltransferases
I. PURIFICATION OF N-ACETYLGALACTOSAMINYLTRANSFERASE”
(Received for publication, May 13, 1977)
MASAKO NAGAI,$ VIBHA DAvB, BRUCE E. KAPLAN, AND AKIRA YOSHIDA
From the Department ofBiochemical Genetics, City ofHope Medical Center, Duarte, California 91010
An N-acetylgalactosaminyltransferase, which converts A persons specifically transfers the sugar residue from UDP-
blood group 0 red blood cells to A cells, was purified to N-acetylgalactosamine to the terminal gala&se of the H-
homogeneity from plasma of blood group A, subjects. The substance of the 0 red cell, producing blood type A substance,
enzyme was adsorbed on Sepharose QB, and after washing while galactosyltransferase of blood group B persons exclu-
out the impurities, the enzyme was eluted with UDP. This sively transfers gala&se from UDP-gala&se to the acceptor
Downloaded from www.jbc.org by guest, on October 22, 2011
procedure resulted in a 70,000- to lOO,OOO-fold increase in sites of the H-substance, producing blood type B substance
specific activity with recovery of about 80%. Further purifi- (7-9).
cation of the enzyme was achieved by Bio-Gel P treatment. In attempts to elucidate the nature of the glycosyltransfer-
The final enzyme preparation showed a single protein band, ases, we first examined two enzymes which are responsible
which coincided with enzyme activity, on acrylamide gel for the synthesis of blood group A and B substances; i.e. N-
electrophoresis, and revealed a single protein band on so- acetylgalactosaminyltransferase (A-enzyme) and n-galactosyl-
dium dodecyl sulfate-gel electrophoresis. Judging from the transferase (B-enzyme). The existence of these two enzymes
molecular weight (90,000 to lOO,OOO), which was estimated has been demonstrated both in the soluble form in human
by Sephadex gel filtration, and the subunit size estimated serum (lo-12), in human milk (13, 14), and in ovarian cyst
by sodium dodecyl sulfate-gel electrophoresis, the enzyme fluid (15, 16) and in the membrane bound form in human
is presumably in a dimeric form. The enzyme required submaxillary glands (17) and in gastric mucosa (18). Wbite-
Mn”+ and had optimum activity at pH 6.5 to 7.0. head et al. (19) attempted to purify N-acetylgalactosaminyl-
transferase from human plasma, but their preparation was
not homogeneous. The membrane-bound N-acetylgalactosa-
Cell surface carbohydrates and sialic acids play an impor- minyltransferase was recently purified from A-positive por-
tant role in determining the biological and immunological cine submaxillary _ glands (20). This paper describes the puri-
- __
activities of cells. The interaction of cells with other cells, fication of the A-enzyme from human plasma.
foreign organisms, and the environment is partially a result EXPERIMENTAL PROCEDURES AND RESULTS
of the composition and organization of cell surface carbohy-
drates. For example, the structure of carbohydrates on the Details of experimental procedures and results are pre-
surface of red cells which is related to the ABO and Lewis sented in the miniprint supplement which follows.’
blood group systems has been well characterized. The en-
DISCUSSION
hanced agglutinability and peculiar growth pattern of malig-
nant cells are presumably related to the structure of surface Whitehead et al. (19) found that UDP-N-acetylgalactosami-
carbohydrates (l-31, and considerable alterations in the activ- nyltransferase of human plasma of blood type A, was adsorbed
ities of glycosyltransferases have been reported in transformed on Sepharose 4B, and they purified the enzyme to about lOOO-
cells (4-6). fold by adsorption of the enzyme to Sepharose 4B and elution
Varieties of glycosyltransferases, some of which are soluble with UDP. They claimed that the enzyme prepared by their
and others of which are associated with cellular components, method was homogeneous. However, the purity of their prep-
are responsible for transferring a particular carbohydrate aration was proved to be very low. In the present work,
residue to terminal acceptor sites of surface glycoproteins and adsorption properties of the enzyme with Sepharose 4B was
glycolipids. These transferases have extremely high specificity examined, and the optimum condition for adsorption and
with respect to a donor and acceptor of a carbohydrate residue. elution of the enzyme was established. In contrast to their
Thus, the N-acetylgalactosaminyltransferase of blood group purification method, the present method reduced the amount
of Sepharose 4B needed for purification (10 g/2 liters of plasma
* This work was supported by Research Grant HL-15125 and HL-
20301 from the National Institutes of Health. The costs of publication 1 Portions of this paper (including Figs. 1 to 4 and Table I) are
of this article were defrayed in part by the payment of page charges. presented in the mininrint at the end of this naner. Full size
This article must therefore be hereby marked %duertisement” in photocopies are available from the Journal of Biological Chemistry,
accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 9650 Rockville Pike, Bethesda, Md. 20014. Request Document 77M-
$ present address, Biological Laboratory, School of Paramedicine, 731, cite authors, and include a check or money order for $1.35 per
Kanazawa University, Kanazawa, Japan. set of photocopies.
377
Blood Group N-Acetylgalactosaminyltransferase
in contrast to 50 g), and also reduced the concentration of acceptor can be fully accounted for by the higher concentration
UDP in the elution buffer (20 PM in contrast to 5 mM). Thus, of the soluble acceptor (12 x 1Ol7 molecules/ml) than that of H
this method minimized nonspecific adsorption and nonspecific sites in the reaction mixture (lOI sites/ml).
elution of contaminating proteins. This step alone resulted in
REFERENCES
an increase in specific activity to 70,000- to lOO,OOO-fold with
recovery of about 80% in contrast to about lOOO-fold in their 1. Yogeeswaran, G., Sheinin, R., Wherrett, J. R., and Murray, R.
K. (1972) J. Biol. Chem. 247, 5146-5158
method. 2. Hakomori, S. (1970) Proc. Nutl. Acad. Sci. U. S. A. 67, 1741-
The mechanism of adsorption of the enzyme to Sepharose 1747
4B is not clear, and adsorption capacity is different from lot to 3. Vheri, A., and Ruoslahti, E. (1974)Znt. J. Cancer 13, 579-586
lot. It was necessary to select an adequate lot for the enzyme 4. Grimes, W. J. (1970) Biochemistry 9, 5083-5092
5. Bosmann, H. B. (1972) Biochem. Biophys. Res. Commun. 49,
purification. For example, Lot 4X-0181 cannot be used due to 1256-1262
its irreversible adsorption, and Lot 3325 cannot be used due 6. Fishman, P. H., McFarland, V. W., Mora, P. T., and Brady, R.
to lack of adsorption capacity. The partially purified enzyme 0. (1972) Biochem. Biophys. Res. Commun. 48, 48-57
was adsorbed to Bio-Gel P, and the enzyme can be eluted by 7. Watkins, W. M., and Morgan, W. T. J. (1959) VOX Sang. 4, 97-
119
0.2 M NaCl.
8. Poretz, R. D., and Watkins, W. M. (1972) Eur. J. Biochem. 25,
The enzyme purified by Bio-Gel P treatment was apparently 455-462
homogeneous. An’overall yield was about 10%. The loss of the 9. ‘Puppy, H., and Schenkel-Brunner, H. (1969) Eur. J. Biochem.
enzyme activity at the Bio-Gel P step was presumably due to 10, 152-157
the partial inactivation of the enzyme during the treatment, 10. Kim, Y. S., Perdamo, J., Bella, A., and Nordberg, J. (1971)
Proc. Natl. Acad. Sci. U. S. A. 68, 1753-1756
since the partially purified enzyme was very unstable at 4” 11. Schachter, H., Michaelis, M. A., Tilley, C. A., Crook&on, M.
(Fig. 4). On account of such inactivation, the real specific C., and Crookston, J. H. (1973) Proc. Natl. Acad. Sci. U. S.
activity of the purified enzyme should be several times higher A. 70, 220-224
than that given in Table I. Judging from the molecular 12. Ko, G. K. W., and Raghupathy, E. (1972) Biochem. Biophys.
weight as estimated by gel filtration, and subunit size, as Res. Commun. 46, 1704-1712
Downloaded from www.jbc.org by guest, on October 22, 2011
13. Kobata, A., and Ginsburg, V. (1970) J. Biol. Chem. 245, 1484-
estimated by sodium dodecyl sulfate-acrylamide gel electro- 1490
phoresis, the enzyme is probably in a dimeric form. The 14. Pacuszka, T., and KoScielak, J. (1972) Eur. J. Biochem. 31, 574-
enzyme remained close to the top of the acrylamide gel at pH 577
8.3, suggesting that the enzyme is a basic protein, as was 15. Hearn, V. M., Race, C., and Watkins, W. (1972) Biochem.
Biophrs. Res. Commun. 46, 498-956
reported with the crude plasma enzyme (21). 16. Nelson, J. D., Jato-Rodriguez, J. J., and Mookerjea, S. (1973)
From the number of antigenic sites of A,-erythrocytes Biochem. Biophys. Res. Commun. 55, 530-536
(about 106 sites per cell) (22), and from the agglutinability of 17. Hearn, V. M., Smith, Z. G., and Watkins, W. M. (1968) Biochem.
the transformed 0 red cells after the enzyme reaction, it can J. 109. 315-317
be calculated that 1 ml of crude outdated Al-plasma can 18. Schenkel-Brunner, H., and ‘Puppy, H. (1973) Eur. J. Biochem.
34, 125-128
transfer about 2 x lOi molecules of N-acetylgalactosamine to 19. Whitehead, J. S., Bella, A., Jr., and Kim, Y. S. (1974) J. Biol.
the H-substance of the 0 red cell surface in 1 h at 37”. The Chem. 249, 3442-3447
same amount of plasma can transfer about 2 x 1Ol5 molecules 20. Schwyzer, M., and Hill, R. L. (1977) J. Biol. Chem. 252, 2338-
of the sugar to fucosyllactose2 under the assay condition. The 2345
21. Topping, M. D., and Watkins, W. M. (1975) Biochem. Biophys
higher transfer rate of the sugar to fucosyllactose as an Res. Commun. 64, 89-96
22. Economidou, J., Hughes-Jones, N. C., and Ciradner, B. (1967)
* The abbreviation used is: fucosyllactose, O-a-cfucopyranosyl- VOX Sung. 12, 321-328
(1-2~-0-~-D-galactopyranosyl-~l-4~-D-glucopyranose. Additional references are found on p. 379.
Downloaded from www.jbc.org by guest, on October 22, 2011
379
N-Acetylgalactosaminyltransferase
Group
Blood