Effects of UDP-glucose Addition on the Synthesis of Mannosyl Lipid-linked Oligosaccharides by Cell-free Fibroblast Preparations*

The pattern of mannosyl lipid-linked oligosaccharides synthesized by cell-free enzyme preparations from cultured fibroblasts is altered substantially when 0.2 FM UDP-glu-case is added to the incubation medium. Inclusion of UDP- glucose results in the appearance of a new labeled oligosaccharide, which is 1 or 2 glycose units larger than the lipid-linked oligosaccharide synthesized in the presence of only GDP-mannose (2 PM) and UDP-N-acetylglucosamine (20 FM). Label from UDP-[“HIglucose is incorporated into the same larger oligosaccharide size class. The results can be explained most easily by assuming that the new mannosyl lipid-linked oligosaccharide contains 1 or 2 glucose residues in addition to 5 to 6 mannose residues. The results are compatible with the recent finding by Spiro et al. (Spiro, M. J., Spiro, R. G., and Bhoyroo, V. D. (19’76) Fed. Proc. 35, 1375; Spiro, M. J., Spiro, R. G., and Bhoyroo, V. D. (1976)5. Biol. Chem. 251,

The pattern of mannosyl lipid-linked oligosaccharides synthesized by cell-free enzyme preparations from cultured fibroblasts is altered substantially when 0.2 FM UDP-glucase is added to the incubation medium. Inclusion of UDPglucose results in the appearance of a new labeled oligosaccharide, which is 1 or 2 glycose units larger than the lipidlinked oligosaccharide synthesized in the presence of only GDP-mannose (2 PM) and UDP-N-acetylglucosamine (20 FM). Label from UDP-["HIglucose is incorporated into the same larger oligosaccharide size class. The results can be explained most easily by assuming that the new mannosyl lipid-linked oligosaccharide contains 1 or 2 glucose residues in addition to 5 to 6 mannose residues. The results are compatible with the recent finding by Spiro  Chem. 251, 6420-6425) that naturally occurring mannosyl lipid-linked oligosaccharides contain glucose. In addition to being incorporated into lipid oligosaccharides, glucose residues are also incorporated into endogenous glycoproteins.
Incorporation of glucose into glycoproteins that give rise to pronase glycopeptides of the typical asparagine-linked size classes is almost completely dependent on the presence of GDP-mannose.
There is a generally held opinion that the complex glycoproteins of serum and plasma membrane lack glucose (see, for example Ref. 1). In his classical review on nucleotide sugars and transglycosylation reactions, Ginsburg (2) makes the point that "a teleological argument can be made for the absence Of D-&ICOSE! in that it is the only sugar found free in most tissue fluids. The efficiency of 'recognition surfaces' composed of  residues would be impaired by free n-glucose much as antigen-antibody interactions are inhibited by haptenes. Evolutionary selection against this impairment, how-ever slight, might thus be expected to eliminate n-glucose as a component of cell surfaces." Whatever the merits of this teleological argument and of published data concerning the presence or absence of glucose in cell surface glycoproteins there are a number of facts that at least indirectly implicate n-glucose in the synthesis of cell surface macromolecules.
These include numerous demonstrations that cell-free systems will catalyze the incorporation of glucose from UDP-glucose into lipid-linked structures with properties similar to those of the mannosyl lipid components which now appear to be intermediates in glycoprotein synthesis. Dolichol monophosphate-glucose has been synthesized with preparations from rat liver, hen oviduct, Z'etruhymena pyriformis, and cotton fibers, as well as other sources (3). It has also been shown that lipid-linked glucosyl oligosaccharides can be synthesized by many of the same preparations (for recent reviews see Ref. 3). In general, these glucose-containing lipid-linked oligosaccharides have not been characterized in detail.
Among characterized glycoproteins only collagen and related structural proteins have been clearly shown to contain glucose. These proteins have hydroxylysine residues which carry 24L~-D-glucosylgalactose disaccharides (4)(5)(6). Although unrelated to the glycoprotein question it should be pointed out that most complex glycosphingolipids contain a single glucosyl moiety attached directly to sphingosine.
A major new contribution to this area of research has been the discovery by Spiro et al. (4)(5)(6)(7)(8) of lipid-linked oligosaccharides containing both glucose and mannose. These materials have been isolated from thyroid, oviduct, kidney, thymus, liver, and pancreas. All appear to contain glucosamine and mannose, with small amounts of glucose. The glucose residue(s) has been demonstrated to be an integral part of the oligosaccharide structure. In the present paper we show that extremely small amounts of UDP-glucose can modify the nature of lipid-linked oligosaccharides synthesized from GDPmannose in a cell-free system, and that glucose is probably incorporated into these same structures.  (9). The cell lysate was centrifuged at 1000 x g for 7 min. The pellet was discarded and the supernatant fluid was centrifuged at 100,000 x g for 1 h. The high speed pellet was resuspended in 2 x 10e2 M Tris/Cl, 1 x 10-l M NaCl, pH 7.4 (Tris-buffered saline), and spun again at 100,000 x g for 1 h. The final pellet was once again resuspended in Tris-buffered saline, pH 7.4, and this material is referred to as the particulate enzyme preparation. All the solutions used and operations carried out were at O-4", except for the removal of the cells from the bottles, which was carried out at room temperature.

Assay: Conditions and Extraction Procedure
The final concentrations of the components in the standard incubation mixture (0.08 ml) were as follows: 2 x lo-* M TrislCl, pH 7.4, 1.5 x 10-l M NaCl, 2 x 10m4 M MnCl*, 2 x 10ms M UDP-iV-acetylglucosamine ( and pellet by centrifugation at 700 x g for 5 min. The pellet was re-extracted with 2 ml of the chloroform/methanol (2/l) mixture and the resultant supernatant was combined with the first extract.
The 4 ml of chloroform/methanol 2/l extract were partitioned into two phases by the addition of 1 ml of 0.9% NaCl; the solutions were mixed for 30 s chilled for 10 min, and centrifuged (700 x g) for 5 min at room temperature.
The upper phase was discarded, the lower phase, mostly chloroform, was again partitioned as described after the addition of 1.5 ml of 0.9% saline/methanol (2/l, by volume). The resultant lower phase (Fraction I), now washed free of non-lipid radioactivity, was either stored (-20") for later characterization or counted.
The amount of radioactivity present was determined by liquid scintillation counting after the evaporation of the chloroform, a potential quencher. The original pellet was dried thoroughly under a stream of nitrogen and then dispersed in 1 ml of water by sonication (Branson Sonic Power, Branson Instruments Inc., power level No. 1). After centrifugation at 700 x g for 5 min, the supernatant was removed. The pellet was washed by successive additions of water and centrifugations until the supernatant contained less than 200 cpm. Material soluble in mixtures of chloroform/methanol/water (l/1/0.3, by volume) was obtained by two successive additions of 3 ml of this solvent mixture to the washed pellet. The combined supernatant (Fraction II), after 5-min centrifugation at 700 x g, were either stored t-20") for later characterization or analyzed for radioactivity. After removal of the chloroform/methanol/water by evaporation, 1 ml of 1% sodium dodecyl sulfate was added to solubilize the radioactivity for counting. The final extracted pellet (Fraction III) was either stored C-20" under chloroform/methanol/water l/1/0.3) or dissolved in 1 ml of 10% sodium dodecyl sulfate (lOO", 3 min), and the amount of radioactivity determined.

Chromatography and Electrophoresis
Chromatography acid was added to a dry sample; 1 ymol each of mannose, galactose, and glucose, was added as internal standards.
The sample was heated at 110" for 1 h, cooled, and dried under nitrogen.
The sample was dissolved in water then applied to Whatman No. 3MM paper for descending chromatography as described above. Mild Acid Hydrolysis -The procedure used to prepare samples for gel filtration chromatography was as follows: the sample was dissolved in 0.2 ml of 0.5 N HCbtetrahydrofuran (l/4, by volume) and was heated at 50" for 2 h. After cooling the sample, 0.2 ml of 0.1 N NaOH in 0.02 N sodium phosphate was added. The solution was then evaporated to dryness under reduced pressure. The oligosaccharide was dissolved in water and applied to the column.

RESULTS
Most of the experiments reported in this paper were carried out with particulate enzyme preparations from NIL-8 fibroblasts. The particles were incubated with 2 pM GDP-[Y!]mannose and or 0.2 PM UDP-["HIglucose. All incubations also contained unlabeled UDP-N-acetylglucosamine (20 PM). The effect of varying nucleotide sugar concentrations is discussed below. The fractions isolated following the incubation were as follows. Fraction I: material soluble in chloroform/ methanol (211). This fraction contains dolichol monophosphate-glycose derivatives and other components; Fraction II: material soluble in chloroform/methanol/water (l/1/0.3). The principal labeled materials in this fraction are lipid-linked oligosaccharides, presumptive intermediates in glycoprotein synthesis; Fraction III: insoluble material. The bulk of the labeled material in this fraction is converted to compounds with the properties of glycopeptides by pronase digestion.
The monosaccharides released from Fractions II and III by strong acid hydrolysis were shown by paper chromatography to be exclusively [Y]mannose and lJHlglucose. Paper chromatography was carried out in a system which clearly separates gala&se, glucose, and mannose. Labeled galactose was not found. Incorporation Kinetics Fig. 1 presents the kinetics of incorporation of label into Fractions I, II, and III when the two nucleotide sugars were incubated separately or together. In all of these experiments the initial concentrations of GDP-mannose and UDP-glucose were 2 and 0.2 FM, respectively. All incubations contained 20 PM unlabeled UDP-N-acetylglucosamine.
As can be seen, the M), on the other hand strongly inhibits the incorporation of UDP-glucose into the lipid fraction, and stimulates its incorporation into the oligosaccharide-lipid and residue fractions. Since in every case the UDP-glucose incorporation was maximal at 5 to 10 min, the lo-min time point was chosen for the isolation and characterization studies described below. The relatively short duration of UDP-glucose incorporation may reflect complete utilization of the small amount of added substrate or other limiting factors. Higher concentrations of UDPglucose were not used since, in general, they led to heterogeneity in the number and size of the oligosaccharide products (see "Discussion"). These points are currently being investigated.

Fraction
I: Chloroform/Methanol (211) -When the chloroform/methanol (2/l) extract from doubly labeled particles is subjected to chromatography on DEAE-cellulose, several components are displayed. A component which elutes with chloroform/methanol alone contains only glucose label and could be glycosphingolipid or a glycosyl glyceride but has not been characterized. The components eluting at the beginning of the ammonium acetate gradient are in the position of dolichol monophosphate-glycose (mannose and glucose). A third pair of peaks eluting at higher ammonium acetate concentrations, could be oligosaccharide lipid derivatives. All of these materials are now being studied in detail and will be the subject of a future publication. (1 I1 IO.31 -When subjected to DEAE-cellulose chromatography, the material in this fraction behaves in the manner expected of oligosaccharide pyrophosphate lipid derivatives, as described by previous workers (4)(5)(6). Essentially all of the radioactivity is retained by the column and is eluted at approximately 50 mM ammonium acetate (Fig. 2).
Mild acid hydrolysis converts this material to water-soluble oligosaccharides which have been fractionated by chromatography on Bio-Gel P-6. A typical set of profiles for singly and doubly labeled materials prepared from samples incubated for 10 min is shown in Fig. 3. The striking result is the alteration in the mannose oligosaccharide profile produced by UDPglucose addition. Although the total yield of ["'Clmannose oligosaccharide material is the same in 3A and 3B (cf. also Fig. l), approximately half of the mannose label appears in a larger oligosaccharide when 0.2 pm UDP-glucose is added to the incubation mixture. Radioactive glucose from UDP-glucase is incorporated into this same larger oligosaccharide. Some 3H is incorporated into the larger oligosaccharide area in the absence of GDP-mannose but this incorporation is stimulated 2-to S-fold by GDP-mannose addition.
The incubation, extraction, and mild acid hydrolysis were carried out as described in Fig. 3 in a sealed tube in 1 N NaOH at 100" for 1 h. The sample was cooled, neutralized, and a portion treated with Escherichio coli alkaline phosphatase.
The treated (not shown) and untreated (shown) portions were streaked separately on Whatman No. 3MM paper and subjected to high voltage paper electrophoresis in pyridinium acetate. The sample was started at 0 and the anode is to the right.
The electrophoretograms were cut into Z-mm strips, eluted into 1 ml of water, and counted. The anionic region of electrophoretogram shown above contained materials completely sensitive to alkaline phosphatase.
Non-phosphatase-sensitive materials, presumably alkaline degradation products, occurred in the cationic and in more anionic areas. Other experiments have shown that UDP-N-acetylglucosamine stimulates glucose incorporation into the lipid oligosaccharide fraction only in the presence of GDP-mannose (data not shown). The Bio-Gel P-6 columns used for these experiments have been calibrated with glycopeptides of known molecular weight. Calculation based on these calibrations shows that the larger oligosaccharide contains 1 or 2 more glycose units than the smaller component.
The P-6 pattern of oligosaccharides derived from Fraction II changes with incubation time. Fig. 3 shows a typical lo-min profile. A typical 30-min profile for doubly labeled material is shown in Fig. 4. As may be seen by comparing Figs. 3 and 4, the major change produced by extended incubation is an increase in the size of the lower molecular weight oligosaccharide peak. In addition there may be a broadening of both the RH and "'C peaks which could reflect an increase in the complexity of the oligosaccharide mixture. In order to resolve the oligosaccharide mixture further, high voltage paper electrophoresis separations were carried out on the oligosaccharide phosphate esters derived by alkaline cleavage of the lipid oligosaccharides.
A typical electrophoresis pattern for doubly labeled material is shown in Fig. 5 . The resolution of these components was seen more clearly in other electrophoretograms (data not shown). All of these materials were converted to neutral species by alkaline phosphatase, and following phosphatase treatment the free oligosaccharides gave single sharp peaks in the approriate position on P-6 columns. Component A corresponds to the larger oligosaccharide and Component B to the smaller oligosaccharide (see Fig. 3). These electrophoretic separation studies are continuing.
Fraction III: Nonextractible Residue -Strong acid hydrolysis of the residue fraction releases L:'H]glucose and L'4Clmannose. No detectable label is present in galactose or other monosaccharides. Label in the residue fraction is largely solubilized by pronase digestion. Typical Bio-Gel P-6 profiles for pronase digests of singly and doubly labeled materials are shown in Fig. 6. The position of the exclusion volume (V,,) is marked as well as the calibrated positions of Sindbis virus glycopeptides S-2, S-3, and S-4 (13).
As is true for the lipid-linked oligosaccharides, the presence of UDP-glucose in incubation mixtures alters the pattern of mannose-labeled glycopeptides. There is stimulation of mannose incorporation into the oligosaccharides in the S-2 to S-4 size region. There is significant incorporation of glucose into glycopeptides of the same size class only in the presence of GDP-mannose.
In addition to the demonstration of incorporation into Fraction III, we have shown that glucose can be incorporated into the well characterized glycoproteins of Sindbis virus. These studies were carried out with membrane fractions of infected chick embryo fibroblasts. The data are presented in Table I. Both direct and indirect Sindbis specific antibody precipitates were isolated. As may be seen, the ratio of glucose to mannose iabel is the same in Fraction III and the specific precipitates. Sodium dodecyl sulfate-gel electrophoresis patterns of antibody precipitates prepared from these labeled materials show a single radioactive component with a mobility similar to that of the viral glycoproteins. '
Although the materials synthesized in our in vitro system are probably smaller by several glycose units than the oligosaccharides isolated by Spiro et al., this result may be explained by differences between the in vitro and i n viva synthesis systems, or may be a reflection of tissue specificity. We have not, as yet, examined the oligosaccharide lipids made by fibroblasts in viuo. We should emphasize that we have not yet proven, in our system, that mannose and glucose are present in the same molecule although the evidence points strongly in this direction. The relevant experiments are being carried out now and will be the subject of a separate publication.
Calculations based on glycopeptide calibrations of the P-6 columns indicate that the presumptive mannosylglucosyl oligosaccharide is 1 or 2 glycose units larger than the mannosyl oligosaccharide.
Such calculations are uncertain, however, since P-6 elution position depends on both molecular size and shape. Calculations of sugar ratios based on the specific activities of the nucleotide sugar precursors are even more precarious because of the possible presence of unlabeled nucleotide sugars in the enzyme preparation and the presence of some glucose label in the larger oligosaccharide peak area when