Suspensions of intact animal cells are capable of transferring the sugar moiety of exogenous nucleotide sugars to endogenous glycoprotein and glycolipid acceptors (10,12,13). It has frequently been assumed that incorporation by suspended cells incubated with nucleotidr sugars indicates the presence of an ectoglycosyltransferase. Ectoenzymes are defined as membrane-bound enzymes whose active sites are accessible from the outside of the cell (3).Several recent publications have challenged the suggestion that glycosyltransferases are ectoenzymes. Evans found a nucleotide pyrophosphatase on the plasma membrane of hepatocytes which could degrade exogenously added nucleotide-sugar substrates (5). Deppert et al. (4) have reported evidence that all the incorporation by baby hamster kidney (BHK) monolayers is due to hydrolysis and intraceilular utilization of the labeled galactose released by degradation of exogenous UDPgalactose.In this communication, we show that while suspensions of mouse spleen cells incorporate label from exogenous UDP-[l~C]galactose, there are no ectogalactosyltransferases. The spleen cell preparations degrade the nucleotide sugar, releasing galactose which is utilized for complex carbohydrate synthesis within the cell. In contrast, BALB/c 3T3 and simian virus 40-transformed 3T3 cells in suspension possess an ectogalactosyltransferase capable of transferring the carbohydrate moiety of exogenous UDP-galactose to endogenous acceptor molecules.
MATERIALS AND METHODS
Cell CulturesBALB/c 3T3 cells (clone A31) and simian virus 40-transformed 3T3 cells (SVT2) were the gift of Dr. G. Todaro, National Institutes of Health, Bcthesda, Md. The procedures for fibroblast growth and suspension using 0.01 M EDTA in saline for studies of r tosyltransferases have been described (10).Cells were teased out of spleens from BALB/c or C3H mice into Hanks' balanced salts solution with 0.6% dextran (HBSS + dex). Cells were dispersed by expressing through a 20-G needle. They were then washed twice with HBSS + dex and resuspcnded in RPMI 1640 medium supplemented with 5% fresh human serum, 2 mM glutamine, 40 U/ml penicillin, and 50 ~g/ml streptomycin. Suspensions were diluted to 5 • 106 mononuclear cells per milliliter and 2-ml aliquots were placed in 17 • 100-ram plastic culture tubes (Falcon Plastics, Oxnard, Calif.). Concanavalin A (Con A) 1 from Calbiochem (San Diego, Calif.) was added in 0.2-ml aliquots of a 100 ~g/ml solution (final concentration l0 /~g/ml). Incubation was carried out for 48 h at 37~ in a humidified atmosphere of 5% CO2 in air. The 48-h point coincides with the maximum rate of [SH]thymidinc incorporation. After Con A stimulation, cells were centrifuged, resuspended, and washed once with TBS.Measurement of galactose incorporation into acidprecipitable material was performed essentially as described (10). The incubations with lymphocyte or fibroblast cell suspensions contained 80 pmol UDP-[I'C]galactose, 80 pmol [SH]galactose, 0.5 ~mol MnCls, and 1-3 • 106 cells in a final volume of 0.06 ml of TBS. We hav...