Synthesis of retinylphosphate mannose in yeast and its possible involvement in lipid-linked oligosaccharide formation

Biochimica et Biophysica Acta, 757 (1983) 77-84

77

Elsevier BiomedicalPress BBA21439

S Y N T H E S I S OF R E T I N Y L P H O S P H A T E M A N N O S E IN YEAST AND ITS P O S S I B L E I N V O L V E M E N T IN LIPID-LINKED OLIGOSACCHARIDE F O R M A T I O N LUDWIG LEHLE, ANTON HASELBECKand WIDMAR TANNER lnstitut fiir Botanik, Universitiit Regensburg, Universiti~tstrasse31, 8400 Regensburg (F.R.G.)

(Received October 4th, 1982) (Revised manuscript receivedDecember29th, 1982)

Key words: Retinylphosphate mannose," Oligosaccharide-lipid; (YeasO

A membrane fraction from S a c c h a r o m y c e s cerevisiae as well as a mannosyltransferase purified therefrom was shown to catalyze the transfer of mannose from GDPmannose to retinyl phosphate. The product formed has chromatographic and chemical properties characteristic for retinylphosphate mannose. The enzyme requires divalent cations. Mg 2+ is more effective than Mn 2+ with an optimum concentration around 25 mM. Amphomyein at a concentration of 0.1 m g / m l inhibits the reaction to 50%. Glycosyl transfer was specific for mannose residues from GDPmannose and did not occur with dolichylphosphate mannose nor with UDP galactose; UDPglucose is a poor donor. Formation of retinylphosphate mannose is inhibited by dolichyl phosphate. This observation as well as similarities between retinylphosphate mannose and dolichylphosphate mannose synthesis in respect to ion requirement, inhibition by amphomycin are suggestive that both reactions are catalyzed by one and the same enzyme. In experiments studying the glycosyl donor specificity in the assembly of lipid-linked oligosaecharide intermediates involved in N-glycosylation of proteins, it could be demonstrated that retinylphosphate mannose can replace dolichylphosphate mannose in the final steps of mannosylation.

Introduction 'Lipid intermediates' of the dolichyl type are involved in the biosynthesis of glycoproteins in eucaryotic cells (for review, see Refs. 1-3). A similar function has been discussed [4] for retinyl phosphate, ever since the synthesis of a mannosylated derivative of the compound has been reported [5,6]. Whereas the dolichol pathway for the N-glycosylation of proteins most likely is identical in yeast [7-9], higher plants [10,11], and animal cells [1-3], the formation of retinyl phosphate mannose so far has only been observed in animal cells. If retinyl phosphate plays a general role in glycoprotein metabolism it might do so also in lower eucaryotic cells. In the present study it was examined, therefore,

whether yeast membranes catalyze a mannosyl transfer from G D P mannose to retinyl phosphate and whether retinylphosphate mannose itself is able to serve as donor in mannosyl transfer reactions. Furthermore the enzymatic activity responsible for retinylphosphate mannose formation has been compared with the activity leading to dolichylphosphate mannose.

Materials and Methods Materials

GDP[14C]mannose (spec. act. 199 Ci/mol) and the other radiochemicals were obtained from Amersham. Various polyprenyl phosphates were a gift from Dr. G. Palamarczyk (Polish Academy of

0304-4165/83/0000-0000/$03.00 © 1983 ElsevierSciencePublishers

78 Sciences, Warsaw). Retinyl phosphate was synthesized by C.S, Silverman-Jones according to Ref. 12 and was a gift from Dr. L.M. De Luca, NIH, Bethesda.

Enzyme preparation Preparation of particulate fraction from S. cerevisiae X2180 cells was carried out as described previously [7] Soluble mannosyl transferase was purified according to [13]. Assay procedures Transfer of mannose from GDP mannose to retinyl phosphate and other polyprenyl phosphates. Retinyl phosphate (7.5 nmol) or the appropriate polyprenyl phosphates as indicated and 2 #mol MgEDTA were dried under nitrogen. Then the other components of the reaction mixture were added to a total volume of 0.07 ml: 10 mM TrisHCI (pH 7.5), 5 mM MgC12, 0.04% Triton X-100, 7/~M GDP [14C]mannose and enzyme (0.5-1 mg protein in case of particulate fraction; 5-10 /~g protein, when solubilized enzyme was used). After incubation at room temperature for the indicated times the reaction was stopped by addition of chloroform/methanol (3:2, by vol.) centrifuged and the supernatant fraction was analyzed by thin-layer chromatography on precoated silica gel G plates (Merck) in the solvent system chloroform/methanol/water (60 : 35 : 6, by vol.). All procedures were performed in a dark room under yellow light. Radioactive spots detected by scanning were scraped off the plate and counted by liquid scintillation with an efficiency of 75%. The results given are representative of a number of determinations. Preparation of retinylphosphate mannose was achieved by a 5-fold scaled up standard assay and purifying the obtained lipophilic fraction from several incubations by column chromatography on DEAE-cellulose as described earlier for dolichylphosphate mannose [14]. Transfer of mannose from retinylphosphate mannose and dolichylphosphate mannose, respectively, to lipid-linked oligosaccharide. Retinylphosphate [laC]mannose (30000 counts/min) or dolichylphosphate [14C]mannose (40000 counts/rain) was dried under nitrogen, resuspended in 0.075% Triton X-100, 10 mM Tris-HC1 (pH 7.5) 5 mM

MgC12 and incubated with particulate enzyme (2.5 rag) in a final volume of 0.08 ml for 45 min at room temperature. The reaction was stopped with 1 ml chloroform/methanol (3:2, by vol.) and washed twice with each 1 ml of the same solvent. Lipid-linked oligosaccharides were extracted twice with each 1.5 ml chloroform/methanol/water (10:10:3, by vol.), hydrolyzed by mild acid and the oligosaccharide fraction thus obtained was subsequently analyzed by gel filtration as described in Ref. 7. The residual pellet was washed two more times with 1 ml 50% methanol and counted for radioactivity to determine the incorporation into polymer. Results

Formation of retinylphosphate mannose The enzyme catalyzing the formation of dolichylphosphate mannose from GDPmannose and dolichyl phosphate was solubilized from yeast membranes and subsequently purified to almost homogeneity [13]. As shown in Fig. 1 such an enzyme preparation also transfers mannosyl residues to exogeneously added retinyl phosphate. The radioactive product formed partitions on thin-layer chromatography as authentic retinylphosphate mannose, chromatographs on a DEAE-cellulose column like polyprenylmonophosphate sugars and the radioactivity is rendered water soluble by mild acid hydrolysis. Retinylphosphate mannose is also formed using the crude membrane fraction as enzyme source (Fig. 1). In this case in addition to retinylphosphate mannose radioactive dolichylphosphate mannose arises during the incubation due to the presence of dolichyl phosphate in the membranes; variable amounts of free mannose are also found originating from the cleavage of GDPmannose and to a minor extent from the breakdown of retinylphosphate mannose on TLC. As shown in Table I the presence of Triton X-100 (0.075%) or bovine serum albumin (7 mg/ml) stimulate the formation of retinylphosphate mannose. Moreover mannosyl transfer to retinyl phosphate seems to compete slightly with synthesis of dolichylphosphate mannose, which possibly indicates that both reactions are catalyzed by the same enzyme (see also below). The rate of retinylphosphate mannose formation with the crude

79

Q

TABLE I

Ret-P-Man

INFLUENCE OF TRITON X-100 AND BOVINE SERUM ALBUMIN ON THE FORMATION OF RETINYLPHOSPHATE MANNOSE BY YEAST MEMBRANES Transfer of mannose from GDPmannose to retinyl phosphate was carried out in (A) the presence (0.075%) or (B) the absence of Triton X-100 or (C) without Triton but with bovine serum albumin (7 mg/ml BSA). Incubation condition

(A) minus Triton X-100 Retinyl phosphate + Retinyl phosphate (B) plus Triton X-100 Retinyl phosphate + Retinyl phosphate (C) plus BSA (minus Triton) + Retinyl phosphate -

c 0

Man

o

iJ

-

Formation of Retinylphosphate mannose (cpm)

Dolichylphosphate mannose a (cpm)

25 3 752

1979 1780

12 6 010

2 933 2 453

8 307

1545

a Dolichylphosphate mannose arises from transfer to dolichyl phosphate endogenously present in the membranes.

J

Distance moved

Fig. 1. Thin-layer chromatography of chloroform/methanolsoluble radioactivity. Incubations were carried out according to the standard assay for 10 rain. (A) Incubation with purified mannosyl transferase in the presence of retinyl phosphate. (B) Incubation with purified transferase in the presence of dolichyl phosphate. (C) Incubation using particulate membrane fraction with retinyl phosphate as acceptor.

m e m b r a n e fraction is a b o u t 10 p m o l / m i n p e r mg o f p r o t e i n a n d with the purified enzyme 750 p m o l / m i n p e r m g of protein. W h e r e a s n o r a d i o a c t i v e p r o d u c t b e h a v i n g like r e t i n y l p h o s p h a t e m a n n o s e was d e t e c t a b l e with the purified e n z y m e in the absence of retinyl phosp h a t e the i n c u b a t i o n s in the presence of high a m o u n t s m e m b r a n e s gave rise to a small a m o u n t o f a m a n n o l i p i d ( 3 - 5 % of the total [14C]glycolipid formed), which h a d the same m o b i l i t y on T L C as

r e t i n y l p h o s p h a t e m a n n o s e ( d a t a n o t shown). This might i n d i c a t e the existence of e n d o g e n o u s retinyl p h o s p h a t e in yeast. However, further e x p e r i m e n t s are required to s u p p o r t this conclusion. T h e time course of the reaction with three different a m o u n t s of retinyl p h o s p h a t e is r e p r e s e n t e d in Fig. 2. T h e reaction is s a t u r a b l e with increasing a m o u n t s of retinyl p h o s p h a t e , 0.2 m M of which gave h a l f - m a x i m u m rate of transfer. T h e a p p a r e n t K m value for G D P m a n n o s e was d e t e r m i n e d from the Lineweaver-Burk p l o t to b e 40 # M . I n c o m parison, for the m a n n o s y l a t i o n of dolichyl phosp h a t e u n d e r the same c o n d i t i o n s tested, the K ~ values were f o u n d to be 60 # M for dolichyl phosp h a t e a n d 8 / ~ M for G D P m a n n o s e , respectively.

Cofactors and inhibitors In Fig. 3 the d e p e n d e n c e on ions for the formation of r e t i n y l p h o s p h a t e m a n n o s e a n d dolichyl p h o s p h a t e m a n n o s e is illustrated. Both reactions reveal a similar ion requirement. D i v a l e n t cations e n h a n c e the transfer. M g 2+ is m o r e effective than M n 2÷ with an o p t i m u m c o n c e n t r a t i o n a r o u n d

80 1/*

@

12

20.

10 ¸

100

@

Q

Q. (3.

6

~50

•.~ 10-

g

Ib

k~

/.

oo t3

t-r" Or"

O

4

2-

lb Time (rain)

2'0

30

~b

50

Divfllent Cfltions (rnM)

01 012 6.3 0'.~ 65 6.6 Amphomyci n (mg/ml)

Fig. 2. Time course of mannosyl transfer from GDPmannose to retinyl phosphate. Purified mannosyltransferase was incubated with (-I ) 0.7/xg, (A A) 1.4 #g or (11 II) 2.8 /~g retinyl phosphate under standard conditions. Fig. 3. Ion dependence for the mannosyl transfer to retinyl phosphate and dolichyl phosphate. Incubations were carried out using purified mannosyltransferase under standard conditions but varying Mg 2~ and Mn 2+ ion concentration. Formation of retinylphosphate mannose (O O, Mg; A ZX, Mn); formation of dolichylphosphate mannose (O O, Mg; • A Mn). Fig. 4. Inhibition of retinylphosphate mannose and dolichylphosphate mannose formation by amphomycin. Mannosyl transfer to retinyl phosphate (O O) and dolichyl phosphate (0 o) was carried out with varying amounts of amphomycin and purified enzyme under the standard incubation. The mixture was preincubated with the inhibitor for 8 min and then the reaction initiated by addition of GDPmannose. The incubation time was 2 rain for dolichyl phosphate and 8 rain for retinyl phosphate.

20-25 mM. Ca 2+ is ineffective with both reactions. Studies by Kang et al. [15] have shown that amphomycin inhibits the formation of dolichylphosphate mannose. As can be seen from Fig. 4 this antibiotic prevents mannosyl transfer to retinyl phosphate to a similar extent. 50% inhibition is observed at around 0.1 mg/ml.

small extent with UDPglucose (Table II). No transfer of galactose from UDPgalactose was observed. Dolichylphosphate mannose did not replace GDPmannose as mannosyl donor. Using the purified enzyme fraction exclusively GDPmannose worked as donor. Obviously, the glucosyltransferase is separated from the mannosylation activity during the purification.

Specificity for sugar nucleotide donors From previous studies in rat liver it appeared that retinyl phosphate is a carrier highly specific for mannose [16,17]. On the other hand Helting and Peterson [18] reported on the formation of retinylphosphate galactose with a membrane fraction from mouse mastocytoma. In the case of yeast membranes glycosylation of retinyl phosphate occurred with GDPmannose as donor and only to a

Specificity of polyprenyl phosphates Although retinol shares certain similarities with dolichols, it is different in chain length and in that retinol contains an additional double-bound per isoprene unit, has an allylic phosphate and an ionon-ring structure at the opposed terminal end. It was of interest, therefore, to compare various polyprenyl phosphates for their ability to accept

81 TABLE II GLYCOSYL D O N O R SPECIFICITY Yeast membranes were incubated for 15 rain with the glycosyl donors (0.1 /zCi) as indicated using the standard assay as described in Materials and Methods. Besides retinylphosphate mannose also dolichylphosphate mannose arises due to the presence of endogenous dolichyl phosphate in the membranes. Sugar nucleotide

GDPmannose Dolichylphosphate mannose UDPglucose UDPgalactose

Transfer to Dolichyl phosphate (endogenous) (cpm)

Retinyl phosphate (cpm)

1 868 1410 0

10 554 0 320 a 0

a This product has not been characterized further; but it is absent, when retinyl phosphate is omitted. Therefore it is tentatively called retinylphosphate glucose.

mannosyl residues. As can be seen from Table III retinyl phosphate is a much poorer acceptor than all polyprenyl phosphates tested. The calculated V values for retinyl phosphate and dolichyl phosphate would differ by a factor of 5. Since also the unsaturated (C35)-polyprenyl phosphate is a better substrate than retinyl phosphate, the ring structure of retinol might be responsible for the decrease in

the rate of transfer. In a previous publication [19] a minimal chain length of about seven isoprene units has been observed for polyprenyl phosphates to be active in mannosyl transfer. This effect cannot be seen with the purified enzyme. In fact, polyprenols of shorter chain length reveal even a slight increase in acceptor activity as compared to longer ones. The discrepancy is not understood at the moment and has to await further clarification. Recently Lrtoublon et al. [20] have performed studies on the specificity of polyprenyl phosphates for mannosyl transfer in microsomal fractions from Aspergillus niger. In this extract no transfer to retinyl phosphate could be demonstrated, although retinyl phosphate inhibited dolichyl phosphate mannosylation.

Influence of dolichyl phosphate on the formation of retinylphosphate mannose In order to get more information, whether dolichyl phosphate and retinyl phosphate are competing for the same enzyme a competition experiment was carried out. Retinylphosphate mannose formation was followed at a fixed amount of retinyl phosphate (27.5 /~M) and at varying concentrations of dolichyl phosphate. As can be seen from Table IV retinylphosphate mannose formation was progressively inhibited with increasing amounts of dolichyl phosphate.

TABLE III

TABLE IV

T R A N S F E R OF MANNOSE FROM GDP-MANNOSE TO POLYPRENYL PHOSPHATES

INHIBITION OF FORMATION OF RETINYIPHOSPHATE MANNOSE BY DOLICHYL PHOSPHATE

Transfer of mannose from GDPmannose to various polyprenyl phosphates (150 ~M) was carried for 1 min using purified enzyme under the standard assay conditions.

Purified mannosyltransferase was incubated with a fixed amount of retinyl phosphate (27.5 #M) and increasing amounts of dolichyl phosphate as indicated. Incubation time was 4 min. Formation of retinylphosphate mannose was determined according to the standard assay procedure.

Substrate added

Retinyl phosphate (C20) C20 dolichyl phosphate C35 dolichyl phosphate C35 polyprenyl phosphate C95 dolichyl phosphate

Incorporation of mannose (cpm) 1668 19 581 15 516 13 864 15 218

Dolichyl phosphate stands for polyprenyl phosphates with an a-saturated isoprene unit. Polyprenyl phosphate has an a-unsaturated isoprene residue.

Addition of dolichyl phosphate (R,M)

Formation of retinylphosphate mannose (cpm)

30 60 120 180

2430 2456 2033 1597 1468

82 Retinylphosphate mannose as mannosyl donor for dolichyl phosphate-linked Glc s Man 9GlcNA c2 oligosaceharide ' Several recent reports have indicated that the mannose residues of the lipid-linked Glc3Man 9 GlcNAc 2 oligosaccharide, which serves as the donor for glycosylation of asparagine residues of

1/.0 120 100 8060-

J

/*020i

i

I

i

i

T

i

I

I

60-

:~ 20

'1

I

i

i

c 60

/.0

20

i

120

i

i

i

1/.0

i

160

i

i

180

i

i

200

Froction number

Fig. 5. Gel filtration profile of [mannosyl-~4C]oligosaccharide. Lipid-linked oligosaccharide were synthesized under the conditions described in Materials and Methods with dolichylphosphate mannose (A), retinylphosphate mannose (B) or retinylphosphate mannose plus 1 m M GDP. About 2% of mannose was transferred from retinylphosphate mannose and 4% in case of dolichylphosphate mannose, respectively. The oligosaccharide were released by mild acid and analyzed on Bio-Gel P4. The arrows indicate the position of marker oligosaccharides: M v - M 9, Man7_gGlcNAc2; G 1 - G 3, Glcl_3Man9GlcNAc2.

glycoproteins are donated by both GDPmannose and dolichylphosphate mannose [7,21-23]. The latter donor is involved in the late steps of oligosaccharide assembly (mannose residues 6 through 9). It was of interest, therefore, to investigate whether retinylphosphate mannose can be substituted for dolichylphosphate mannose. Yeast m e m b r a n e s were i n c u b a t e d either with dolichylphosphate [14C]mannose or retinylphosphate [14C]mannose, the oligosaccharides subsequently released from the newly formed glycolipid by mild acid hydrolysis and analyzed by gel filtration. As presented in Fig. 5 an oligosaccharide profile is obtained from retinylphosphate mannose, which is very similar to that from dolichylphosphate mannose. The major compound migrated with the Man9GlcNAc 2 marker and minor amounts of radioactivity were incorporated into Mans/7GlcNAc 2. The species eluting like the Glcl_3-oligosaccharide markers are oligosaccharides with nine mannose residues extended by 1-3 glucose units due to the presence of unlabelled endogenous dolichylphosphate glucose as observed previously [7]. No radioactivity in the region of the oligosaccharide fraction appears from a zero-time control incubation (data not shown). To guard against the eventuality that the lipid-oligosaccharide synthesized with retinylphosphate mannose as donor was formed via GDPmannose (this could arise during the incubation from retinylphosphate mannose, if endogenous G D P were present), a further control experiment was run with retinylphosphate mannose plus GDP. Fig. 5C shows that the same pattern is obtained as in the absence of GDP. Moreover in separate experiments it was tried to demonstrate a possible formation of GDPmannose both from dolichylphosphate mannose and retinylphosphate mannose in the presence of GDP. The reverse reaction only worked with the purified enzyme (not with membranes under the conditions used above) and only with dolichylphosphate m a n n o s e (96% of dolichylphosphate mannose was converted to GDPmannose in 30 min); it did not work with retinylphosphate mannose (data not shown). Thus the lipid-oligosaccharide formation from retinylphosphate mannose in the membrane fraction does not seem to occur indirectly via GDPmannose.

83

Discussion In the early 1970's two polyisoprenoids, dolichyl phosphate [24,25] and retinyl phosphate [5,6] were suggested to play a role in sugar transfer reactions. Since then glycosyl phosphoryl derivatives of vitamin A have been characterized in several systems in vitro and in vivo [26-29]. Whereas for dolichol by now a clear picture has emerged as far as its function in N- and O-glycosylation of proteins is concerned, the mode of action of retinyl phosphate remains to be elucidated. The observation [30] that retinylphosphate mannose functions as a mannosyl donor to endogenous acceptor protein of rat liver membranes distinct to the transfer from dolichylphosphate mannose is suggestive of a role of vitamin A different from the role of dolichol derivatives. The present study has shown that both a yeast membrane fraction and a mannosyltransferase purified therefrom are able to catalyze the transfer of mannose from GDPmannose but not from dolichylphosphate mannose to retinyl phosphate. The product formed has chromatographic and chemical properties characteristic for retinylphosphate mannose. The formation of retinylphosphate mannose was compared to that of dolichylphosphate mannose. From the similarity in ion requirement, from the inhibition studies with amphomycin and from the competition of retinylphosphate mannose synthesis by dolichyl phosphate one may conclude that both reactions are catalyzed by one and the same enzyme. In addition indication has been obtained, that both activities copurify during the purification procedure. With rat liver microsomes on the other hand, Shidoji et al. [31] recently obtained indirect evidence for the existence of two separate enzyme activities: a higher affinity mannosyltransferase responsible for dolichylphosphate mannose synthesis, while a lower affinity enzyme catalyzed retinylphosphate mannose formation. This system also revealed an extreme sensitivity towards detergents [17]. Thus in the absence of Triton X-100 and the presence of bovine serum albumin a 100-fold higher rate of mannose transfer was observed. Such effects were not measurable with yeast membranes (Table I). Furthermore our results indicate for the first

time that retinylphosphate mannose could replace dolichylphosphate mannose as a donor in the final steps of the assembly of the lipid-oligosaccharide. The oligosaccharide fraction obtained from the incubation with retinylphosphate mannose revealed an almost identical radioactivity pattern as with dolichylphosphate mannose. From control experiments it seems unlikely that the mannose residues are transferred indirectly via GDPmannose or dolichylphosphate mannose by conceivable reverse reactions, since none of this compounds was formed during the incubation. Recently, Rosso et al. [32] have shown that vitamin A-deficient rats accumulate an oligosaccharidelipid of the composition MansGlcNAc 2 instead of the Glc3Man9GlcNAc 2 derivative in normal rats. The effect could be reversed by vitamin A administration. The observation made here that retinylphosphate mannose can participate in the elongation of the oligosaccharide-lipid could partly explain this finding. Whether the reactions described here are of physiological relevance for the yeast remains unknown at the moment till the occurrence of vitamin A in this organism is clearly demonstrated.

Acknowledgements Thanks are due to Mrs. Marianne Neueder for excellent technical assistance and Dr. DeLuca for the gift of retinylphosphate. This work was supported by the Deutsche Forschungsgemeinschaft (SFB 43).

References 1 Waechter, C.J. and Lennarz, W.J. (1976) Annu. Rev. Biochem. 45, 95-112 2 Parodi, A.J. and Leloir, L.F. (1979) Biochim. Biophys. Acta 559, 1-37 3 Hubbard, S.C. and Ivatt, R.J. (1981) Annu. Rev. Biochem. 50, 555-583 4 De Luca, L.M. (1977) Vit. Horm. 35, 1-57 5 De Luca, L.M., Rosso, G.C. and Wolf, G. (1970) Biochem. Biophys. Res. Commun. 41,615-620 6 De Luca, L.M. Maestri, N , Rosso, G.C. and Wolf, G. (1973) J. Biol. Chem. 248, 641-648 7 Lehle, L. (1980) Eur. J. Biochem. 109, 589-601 8 Lehle, L., Schulz, I. and Tanner, W. (1980) Arch. Microbiol. 127, 231-237 9 Parodi, A.J. (1977) Eur. J. Biochem. 75, 171-180

84 10 Staneloni, R.J. Tomalsky, M.E., Petriella, C., Ugalde, R.A. and Leloir, L.F. (1980) Biochem. J. 191,257-260 11 Lehle, L. (1981) FEBS Lett. 123, 63-66 12 Bhat, P.V., De Luca, L.M. and Wind, M.L. (1980) Anal. Biochem. 102, 243-248 13 Haselbeck, A. and Tanner, W. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 1520-1524 14 Lehle, L. and Tanner, W. (1978) Eur. J. Biochem. 83, 563-570 15 Kang, M.S., Spencer, J.P. and Elbein, A.D. (1978) J. Biol. Chem. 254, 8860-8866 16 Bergman, A., Mankowski, T., Chojnacki, T., De Luca, L.M., Peterson, E. and Dallner, G. (1978) Biochem. J. 172, 123-127 17 Shidoji, Y. and De Luca, L.M. (1981) Biochem. J. 200, 529-538 18 Helting, T. and Peterson, P.A. (1972) Biochem. Biophys. Res. Commun. 46, 429-436 19 Palamarczyk, G., Lehle, L., Mankowski, T., Chojnacki, T. and Tanner, W. (1980) Eur. J. Biochem. 105, 517-523 20 L6toublon, R., Mayer, B., Frot-Coutaz, J. and Got, R. (1982) J. Lipid Res. 23, 1053-1057 21 Chambers, J., Forsee, W.T. and Elbein, A.D. (1977) J. Biol. 252, 2498-2506

22 Datema, R. and Schwarz, R.T. (1979) Biochem. J. 194, 113-123 23 Chapman, A., Fujimoto, K. and Kornfeld, S. (1980) J. Biol. Chem. 255, 4441-4446 24 Behrens, N.H. and Leloir, L.F. (1970) Proc. Natl. Acad. Sci. U.S.A. 66, 153-159 25 Tanner, W. (1969) Biochem. Biophys. Res. Commun. 35, 144-150 26 Rosso, G.C., De Luca, L.M., Warren, C.D. and Wolf, G. (1975) J. Lipid. Res. 16, 235-243 27 Barr, R.M. and De Luca, L.M. (1974) Biochem. Biophys. Res. Commun, 60, 355-363 28 Masushige, S., Schreiber, J.B. and Wolf, G. (1978) J. Lipid Res. 19, 619-627 29 Martin, H.G. and Thorne, K.J.K. (1974) Biochem. J. 138, 281-289 30 Sasak, W. and De Luca, L.M. (1980) FEBS Lett. 114, 313-318 31 Shidoji, Y., Sasak, W., Silverman-Jones, C.S. and De Luca, L.M. (1981) Ann. N.Y. Acad. Sci. 359, 345-359 32 Rosso, G.C., Bendrick, C.J. and Wolf, G. (1981) J. Biol. Chem. 256, 8341-8347