Biosynthesis in vitro of sialyl(alpha 2-3)neolactotetraosylceramide by a sialyltransferase from embryonic chicken brain.

A sialyltransferase activity present in 7- to 12-day-old embryonic chicken brain catalyzes the transfer of sialic acid from CMP-sialic acid to the terminal galactose residue of [3H]nLcOse4Cer ([3H]Gal(beta 1-4).GlcNAc(beta 1-3)Gal(beta 1-4)Glc-Cer) to form NeuAc(alpha 2-3)-[3H]nLcOse4Cer (LM1 ganglioside). The product is sialidase-labile (96%), and the NeuAc group is linked to O-3 of the terminal galactose residue. The (alpha 2-3) linkage between sialic acid and the terminal galactose was determined on the basis of identification of 2,4,6-tri-O-methyl[3H]galactose obtained after hydrolysis of the permethylated enzymatic product. The CMP-sialic acid:nLcOse4Cer (alpha 2-3)sialyltransferase activity sediments (90%) at the junction of 1.2 M and 1.5 M on a discontinuous sucrose density gradient when still membrane bound (insoluble in 0.2% Triton X-100). The enzyme preparation also catalyzes the transfer of sialic acid from CMP-sialic acid to O-3 of GgOse4Cer (Gal(beta 1-3)GalNAc(beta 1-4)Gal(beta 1-4)Glc-Cer) to form NeuAc (alpha 2-3)GgOse4Cer (GM1b). Substrate inhibition studies indicate that these two reactions are probably catalyzed by the same enzyme.


sylceramide (Gal(/3l-4)GlcNAc(fll-3)Gal(/3l-4)Glc-Cer)
starting from lactosylceramide. At least three different sialyltransferases (23)(24)(25)(26)(27)(28)(29) are required in the biosynthesis of the major sialic acid-linked gangliosides of embryonic chicken * This work was supported by United States Public Health Service Grants NS-18005-01 and CA-14764-06. This is the 18th paper in a series dealing with biosynthesis in citro of blood group-related glycosphingolipids. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Recipient of grant-in-aid from Miles Laboratories, Elkhart, IN.
TO whom correspondence should be addressed.
The present studies are concerned with the biosynthesis of the LM1 ganglioside from the nLcOse4Cer. The reaction catalyzed by the sialyltransferase obtained from embryonic chicken brain is:

RESULTS
Requirements for Enzymatic Activity-As shown in Table  I (See Miniprint), the rate of reaction decreased markedly (89%) without addition of nLcOse4Cer. In the absence of detergent, the rate of sialyltransferase activity decreased by 58%. The reaction rate was fastest in the presence of either Triton CF-54 (1 mg/ml) or Triton X-100 (1.8 mg/ml) (Fig. 1 only 80%' as effective as Triton CF-54 or Triton X-100. Triton CF-54 was used in the present studies. Under these conditions, product formation was proportional to protein concentration up to 2.6 mg/ml and remained constant with time of incubation up to 1 hr. The incubation mixture contained 0.5 mM EDTA and 10 to 12 mM 2-mercaptoethanol to reduce hydrolysis (5%) of CMP-[14C]NeuAc by CMP-neuraminylhydrolase. MgClr (2.5 mM) stimulated sialyltransferase 3 activity by 5 to 10% and was used routinely in the incubation mixture.
Effect of Embryonic Age-The formation of ['%]NeuAc-nLcOse4Cer was measured using Pz membrane fractions of brains isolated from 7-to 21-day-old chicken embryos (Fig. 2, Miniprint). The highest specific activity of sialyltransferase 3 was observed in preparations from 7-day-old embryonic chicken brains. For the present work, however, 9-to ll-dayold embryonic chicken brains were used because they afforded larger amounts of brain tissues.
Distribution of Sialyltransferase 3 Activity-The distribution of CMP-NeuAc:nLcOse4Cer sialyltransferase activity was determined in different membrane fractions obtained from 11day-old embryonic chicken brains using a discontinuous sucrose gradient (Table 111, Miniprint). Fraction P.[ had the highest content of all three sialyltransferase activities as tested with specific substrates: LacCer; GM3; and nLcOse4Cer and GgOse+Cer. About 52% of sialyltransferase 3 activity was recovered from Fractions P, and P.1, whereas GM3 activity remained primarily in the Py and P:j fractions (31).
Acceptor Specificity-Apparent K,, and V,,,, values were determined using two different potential glycosphingolipid acceptors. The K", value of nLcOse4Cer (0.5 mM) was about 20-fold higher than that of GgOse4Cer (27 p~) . However, the difference in V,,,,, values was not significant (Fig. 3, Miniprint). Isolation a n d Characterization of the Radioactive Prod-u~t-[~~CC]NeuAc-nLcOse~Cer was isolated from a 30-fold increased incubation mixture (see "Materials and Methods"). The incubation mixture was applied to Whatman No. 3MM paper and developed with 1.0% borate to remove radioactive sialic acid and CMP-[14C]NeuAc. The radioactive product was eluted from the paper with chloroform/methanol/water (60:35:8, v/v/v). The [ 14C]NeuAc-nLcOse4Cer was separated from unreacted nLcOse,Cer (substrate) by DEAE-Sephadex A-50 column chromatography as described under "Materials and Methods." The radioactive product eluted with 0.1 M Na acetate in methanol was dialyzed and further purified (20,000 cpm) from preparative thin layer plates. The purified ["C] NeuAc-nLcOse4Cer moved as a single band a little ahead of nonradioactive NeuGc-nLcOse4Cer or GM1 ganglioside (Fig. 4,Miniprint) and behind the substrate nLcOse4Cer. The developing solvent system (chloroform/methanol/0.2% CaC12 in water, 55:45:10; v/v/v) also separated nLcOse,Cer from GgOse4Cer. The substrate nLcOse&er used for our enzymatic work was completely free from GgOse4Cer as was evident from the thin layer chromatographic plate (Fig. 4, Miniprint). The analyses of alditol acetate derivatives of the sugars derived from nLcOse4Cer showed the presence of galactose, glucose, and N-acetylglucosamine only.
Neuraminidase Treatment of the Enzymatic Product-The purified "C product (1000 cpm) was incubated with 0.05 unit of Clostridiumperfringens neuraminidase at 37 "C for 16 h in 50 pl of 0.1 M sodium acetate buffer, pH 5.0, containing 0.04% CaC12. The mixture was spotted on Whatman No. 3 " paper. When assayed by high voltage electrophoresis in 1% sodium tetraborate, 96% of "C-labeled sialic acid moved away from the origin and co-migrated with authentic unlabeled Nacetylneuraminic acid.

DISCUSSION
The present studies demonstrate transfer of ['4C]NeuAc from CMP-[I4C]NeuAc to neolactotetraosylceramide to form IV"NeuAc-nLcOse,Cer in the presence of a sialyltransferase present in 11-day-old embryonic chicken brain. T o our knowledge, the biosynthesis in vitro of a glycosphingolipid containing a terminal NeuAcu2-3Gal~l-4GlcNAc-carbohydrate chain has not been reported before. The transfer of NeuAc from CMP-NeuAc to lactose (GalDl-4Glc) and lactosamine (Galpl-4GlcNAc) was first reported by Jourdian et al. (40) and Carlson et al. (41) using a rat mammary gland particulate system. A marked difference in activity compared with goat colostrum sialyltransferase was observed by Roseman and associates when three different isomers of lactosamine were used as substrates (Dl-4 >> pl-3 > pl-6) (42). Hill and associates purified a sialyltransferase from bovine colostrum that catalyzes the synthesis in vitro of NeuAc(u2-6)lactosamine (43,44). Biosynthesis of a sialyl(a2-3)lactosamine-containing glycoprotein has been reported in fetal calf liver microsomes (45). Whether sialyltransferases from rat mammary gland, bovine colostrum, and fetal calf liver utilize nLcOse4Cer to form NeuAc(a2-3)nLcOse4Cer is not yet known.
The apparent K, value of nLcOse4Cer, 0.5 mM, was calculated from the saturation curve (Fig. 3, Miniprint). Under similar conditions, the apparent K,, value of GgOse4Cer was 27 p . Substrate competition studies were performed (Table   IV, Miniprint) in order to determine whether the transfer of [l'C]NeuAc to nLcOse4Cer and GgOse4Cer is due to one nonspecific or two different specific enzymes. The incorporation of [I4C]NeuAc into two specific substrates was first determined separately and then with a mixture of two substrates.
The quantity of ["'CINeuAc incorporated (13.7 nmol) into the mixture was close to the theoretical value (14.8 nmol) calculated for a single enzyme catalyzing two reactions simukaneously (46). The above results suggest that a single enzyme may be capable of transferring sialic acid to the terminal galactose of nLcOse4Cer or GgOse4Cer. Previously, NeuAc-GgOse4Cer from embryonic chicken brain was proved to contain (a2-3)-linked NeuAc (47). However, in each case our substrates and products have been characterized chemically. Further purification of this sialyltransferase 3 activity is under way (31). It also appears that this embryonic chicken brain sialyltransferase 3 activity is quite different from porcine submaxillary gland sialyltransferase (48), which prefers an 0glycosidically linked GalP1-3GalNAc carbohydrate chain and does not utilize asialo-cw,-glycoprotein. Whether the specificity of any sialyltransferase changes with the internal core structure of the carbohydrate chain is not yet known.

CF-54 lmhm and ~a a a , P h l l a d e l p h l a ) ; T r i t o n X-100 IPackard Instrument Co.);
The following materlala w e r e obtalned from COmmerclal ~ources: T r l t o n SOdlUm taurocholate ICalblochem); sodium taurodeoxycholate. 5'-UIP,

What-"
No. 31111 paper I s c l e n t l f i c P r o d u c t s ) . P e r t l l l r e d eggs were

' -A m . Clostrldzum Werfrlnqens s i a l i d a s e ISLgma Chermcal C 0 . I ; and
Obtazned from Ro~e.8 Hatchery ISOUth Bend. I N ) .
AccePfOrs -A l l the glycoephmgalzpid substrates used were l s o l a t e d I" our laboratory by previouely published methods and analyzed by cerdmlde was prepared by cleaving the terminal galactose from n w -GC mass spectrometry before use as sublltrates. Neolactotetraosyl-1CictOpentaosylCerade xsolated from rabbit erythrocytes [18,19). The methylated sugar-were applled to (Slllca gel GI a TLC p l a t e solutxon was dezonxred wlth Dowex 50 ( K f > md I M w x 1 IRCO-) re-ms. and developed wlth acetone-5 H a w n l a 1500:9). After radioautography. the standard methylated galactoses were Visualized wlth a n i l i n e p h t h a l a t e spray reagent.