Nogo-B receptor is necessary for cellular dolichol biosynthesis and protein N-glycosylation.

Dolichol monophosphate (Dol-P) functions as an obligate glycosyl carrier lipid in protein glycosylation reactions. Dol-P is synthesized by the successive condensation of isopentenyl diphosphate (IPP), with farnesyl diphosphate catalysed by a cis-isoprenyltransferase (cis-IPTase) activity. Despite the recognition of cis-IPTase activity 40 years ago and the molecular cloning of the human cDNA encoding the mammalian enzyme, the molecular machinery responsible for regulating this activity remains incompletely understood. Here, we identify Nogo-B receptor (NgBR) as an essential component of the Dol-P biosynthetic machinery. Loss of NgBR results in a robust deficit in cis-IPTase activity and Dol-P production, leading to diminished levels of dolichol-linked oligosaccharides and a broad reduction in protein N-glycosylation. NgBR interacts with the previously identified cis-IPTase hCIT, enhances hCIT protein stability, and promotes Dol-P production. Identification of NgBR as a component of the cis-IPTase machinery yields insights into the regulation of dolichol biosynthesis.

Hypoglycosylation due to dolichol metabolism defects.

Dolichol phosphate is a lipid carrier embedded in the endoplasmic reticulum (ER) membrane essential for the synthesis of N-glycans, GPI-anchors and protein C- and O-mannosylation. The availability of dolichol phosphate on the cytosolic site of the ER is rate-limiting for N-glycosylation. The abundance of dolichol phosphate is influenced by its de novo synthesis and the recycling of dolichol phosphate from the luminal leaflet to the cytosolic leaflet of the ER. Enzymatic defects affecting the de novo synthesis and the recycling of dolichol phosphate result in glycosylation defects in yeast or cell culture models, and are expected to cause glycosylation disorders in humans termed congenital disorders of glycosylation (CDG). Currently only one disorder affecting the dolichol phosphate metabolism has been described. In CDG-Im, the final step of the de novo synthesis of dolichol phosphate catalyzed by the enzyme dolichol kinase is affected. The defect causes a severe phenotype with death in early infancy. The present review summarizes the biosynthesis of dolichol-phosphate and the recycling pathway with respect to possible defects of the dolichol phosphate metabolism causing glycosylation defects in humans.

A defect in dolichol phosphate biosynthesis causes a new inherited disorder with death in early infancy.

The following study describes the discovery of a new inherited metabolic disorder, dolichol kinase (DK1) deficiency. DK1 is responsible for the final step of the de novo biosynthesis of dolichol phosphate. Dolichol phosphate is involved in several glycosylation reactions, such as N-glycosylation, glycosylphosphatidylinositol (GPI)-anchor biosynthesis, and C- and O-mannosylation. We identified four patients who were homozygous for one of two mutations (c.295T-->A [99Cys-->Ser] or c.1322A-->C [441Tyr-->Ser]) in the corresponding hDK1 gene. The residual activity of mutant DK1 was 2%-4% when compared with control cells. The mutated alleles failed to complement the temperature-sensitive phenotype of DK1-deficient yeast cells, whereas the wild-type allele restored the normal growth phenotype. Affected patients present with a very severe clinical phenotype, with death in early infancy. Two of the patients died from dilative cardiomyopathy.

Identification and characterization of a cDNA encoding a long-chain cis-isoprenyltranferase involved in dolichyl monophosphate biosynthesis in the ER of brain cells.

A long-chain cis-isoprenyltransferase (cis-IPTase) located in the endoplasmic reticulum (ER) catalyzes the chain elongation stage in the pathway for the de novo biosynthesis of dolichyl monophosphate (Dol-P) in eukaryotic cells. In Saccharomyces cerevisiae, the ER-associated cis-IPTase is encoded by the RER2 gene. Mutations in the RER2 gene result in defects in growth and protein N-glycosylation. In this study a cDNA isolated from human brain (Accession No. AK023164.1), which has substantial homology to cis-IPTases from bacteria, Arabidopsis, and S. cerevisiae, has been shown to: (1) complement the growth defect; (2) restore cis-IPTase activity; dolichol and Dol-P synthesis; and (3) restore normal N-glycosylation of carboxypeptidase Y (CPY) in the yeast rer2Delta mutant. Consistent with a role in Dol-P biosynthesis, overexpression of the human cis-isoprenyltransferase (hCIT) cDNA also suppresses the temperature-sensitive growth and CPY hypoglycosylation phenotypes in sec59-1 cells which are defective in Dol-P biosynthesis due to a temperature-sensitive mutation in dolichol kinase. Overexpression of hCIT in Chinese hamster ovary (CHO) cells results in a modest increase in cis-IPTase activity associated with microsomal fractions and the appearance of a new 38kDa polypeptide that co-localizes with calnexin in the ER, the site of Dol-P biosynthesis, even though no transmembrane domains are predicted by a hydropathy plot.

Effects of peroxisome proliferators gemfibrozil and clofibrate on syntheses of dolichol and cholesterol in rat liver.

The effects of two peroxisome proliferators, gemfibrozil and clofibrate, on syntheses of dolichol and cholesterol in rat liver were investigated. Gemfibrozil did not affect the overall content of dolichyl phosphate, but it changed the chain-length distribution of dolichyl phosphate, increasing the levels of species with shorter isoprene units. Gemfibrozil suppressed synthesis of dolichyl phosphate from [(3)H]mevalonate and [(3)H]farnesyl pyrophosphate in rat liver. In contrast, clofibrate increased the content of dolichol (free and acyl ester forms). It remarkably enhanced dolichol synthesis from mevalonate, but did not affect dolichol synthesis from farnesyl pyrophosphate. Gemfibrozil elevated cholesterol synthesis from [(14)C]acetate, but did not affect the synthesis from mevalonate. Clofibrate suppressed cholesterol synthesis from acetate, but did not affect cholesterol synthesis from mevalonate. These results suggest that gemfibrozil suppresses synthesis of dolichyl phosphate by inhibiting, at the least, the pathway from farnesyl pyrophosphate to dolichyl phosphate. As a result, the chain-length pattern of dolichyl phosphate may show an increase in shorter isoprene units. Clofibrate may increase the content of dolichol by enhancing dolichol synthesis from mevalonate. Gemfibrozil may increase cholesterol synthesis by activating the pathway from acetate to mevalonate. Unlike gemfibrozil, clofibrate may decrease cholesterol synthesis by inhibiting the pathway from acetate to mevalonate.

The ins(ide) and out(side) of dolichyl phosphate biosynthesis and recycling in the endoplasmic reticulum.

The precursor oligosaccharide donor for protein N-glycosylation in eukaryotes, Glc3Man9GlcNAc(2)-P-P-dolichol, is synthesized in two stages on both leaflets of the rough endoplasmic reticulum (ER). There is good evidence that the level of dolichyl monophosphate (Dol-P) is one rate-controlling factor in the first stage of the assembly process. In the current topological model it is proposed that ER proteins (flippases) then mediate the transbilayer movement of Man-P-Dol, Glc-P-Dol, and Man5GlcNAc(2)-P-P-Dol from the cytoplasmic leaflet to the lumenal leaflet. The rate of flipping of the three intermediates could plausibly influence the conversion of Man5GlcNAc(2)-P-P-Dol to Glc3Man(9)GlcNAc(2)-P-P-Dol in the second stage on the lumenal side of the rough ER. This article reviews the current understanding of the enzymes involved in the de novo biosynthesis of Dol-P and other polyisoprenoid glycosyl carrier lipids and speculates about the role of membrane proteins and enzymes that could be involved in the transbilayer movement of the lipid intermediates and the recycling of Dol-P and Dol-P-P discharged during glycosylphosphatidylinositol anchor biosynthesis, N-glycosylation, and O- and C-mannosylation reactions on the lumenal surface of the rough ER.

Expression of the insulin-like growth factor 1 receptor (IGF-1R) in breast cancer cells: evidence for a regulatory role of dolichyl phosphate in the transition from an intracellular to an extracellular IGF-1 pathway.

In this study we provide evidence that the low expression of IGF-1R at the cell surface of estrogen-independent breast cancer cells is due to a low rate of de novo synthesis of dolichyl phosphate. The analyses were performed on the estrogen receptor-negative breast cancer cell line MDA231 and, in comparison, the melanoma cell line SK-MEL-2, which expresses a high number of plasma membrane-bound IGF-1R. Whereas the MDA231 cells had little or no surface expression of IGF-1R, they expressed functional (i.e., ligand-binding) intracellular receptors. By measuring the incorporation of [3H]mevalonate into dolichyl phosphate, we could demonstrate that the rate of dolichyl phosphate synthesis was considerably lower in MDA231 cells than in SK-MEL-2 cells. Furthermore, N-linked glycosylation of the alpha-subunit of IGF-1R was 8-fold higher in the melanoma cells. Following addition of dolichyl phosphate to MDA231 cells, N-linked glycosylation of IGF-1R was drastically increased, which in turn was correlated to a substantial translocation of IGF-1R to the plasma membrane, as assayed by IGF-1 binding analysis and by Western blotting of plasma membrane proteins. The dolichyl phosphate-stimulated receptors were proven to be biochemically active since they exhibited autophosphorylation. Under normal conditions MDA231 cells, expressing very few IGF-1R at the cell surface, were not growth-arrested by an antibody (alphaIR-3) blocking the binding of IGF-1 to IGF-1R. However, after treatment with dolichyl phosphate, leading to a high cell surface expression of IGF-1R, alphaIR-3 efficiently blocked MDA231 cell growth. Taken together with the fact that the breast cancer cells produce IGF-1 and exhibit intracellular binding, our data suggest that the level of de novo -synthesized dolichyl phosphate may be critical for whether the cells will use an intracellular or an extracellular autocrine IGF-1 pathway.

Alterations in the biosynthesis of cholesterol, dolichol and dolichyl-P in the genetic cholesterol homeostasis disorder, Niemann-Pick type C disease.

The biosynthesis of cholesterol, dolichol and dolichyl-P were investigated in a murine model of Niemann-Pick type C disease using both in vitro and in vivo systems. In vivo incorporation of [3H]mevalonate into squalene, dolichol and dolichyl-P decreased. The amount of dolichyl-P was elevated due to a decrease in the rate of degradation. Labeling of squalene and cholesterol of liver homogenates in vitro was decreased in the diseased mice and a lowering of microsomal activities of both HMG-CoA reductase and squalene synthase were also observed. In experiments with brain homogenate, decreased [3H]mevalonate labeling of squalene, cholesterol and dolichol was found in vitro. The decreases in cis-prenyltransferase and squalene synthase activities were observed at a very early phase of the disease. In contrast to the decreased biosynthesis of cholesterol observed in vitro, the labeling of total liver cholesterol was found to be increased in Niemann-Pick type C liver upon in vivo investigation, possibly due to the accumulation of this lipid as a result of a deficient transport process. In the brain, where in vivo labeling reflects only biosynthesis, a decreased rate of cholesterol synthesis was demonstrated.

Analysis of isoprenoid biosynthesis in peroxisomal-deficient Pex2 CHO cell lines.

ZR-78 and ZR-82 cells are two peroxisomal-deficient Chinese hamster ovary (CHO) cell mutants. These cells lack normal peroxisomes and show reduced levels of plasmalogen synthesis and other peroxisomal functions attributed to the deficiency of peroxisomal matrix enzymes. As we have recently identified two HMG-CoA reductase proteins in CHO cells, a 97 kDa reductase localized in the ER and a 90 kDa reductase protein localized in peroxisomes, this enabled us to study the two reductase proteins for the first time in peroxisomal-deficient CHO cells. In this study we report the results of a detailed analysis of the isoprenoid biosynthetic pathway in the peroxisomal-deficient CHO cell lines ZR-78 and ZR-82. We demonstrate that total HMG-CoA reductase activity is significantly reduced in the peroxisomal-deficient cells as compared to the wild type cells. Analysis of the two reductase proteins in permeabilized cells indicated that in the ZR-78 and ZR-82 cells the 90 kDa peroxisomal reductase protein was mainly localized to the cytosol. We further show that the rates of both sterol (cholesterol) and non-sterol (dolichols) biosynthesis were significantly lower in the peroxisomal-deficient cells, when either [3H] acetate or [3H] mevalonate was used as substrate. In contrast, the rate of dolichol biosynthesis in the peroxisomal-deficient cells was similar to that of the wild type cells when incubated with [3H] farnesol. In addition, we demonstrate that the peroxisomal-deficient cells exhibited increased rates of lanosterol biosynthesis as compared to wild type cells. The results of this study provide further evidence for the essential requirement of peroxisomes for cholesterol biosynthesis as well as for dolichol production.

Regulatory adaptation of isoprenoid biosynthesis and the LDL receptor pathway in fibroblasts from patients with mevalonate kinase deficiency.

In a search for the pathophysiologic mechanisms, we estimated isoprenoid synthesis and concentration, cellular growth, and the activity of the LDL receptor pathway in fibroblasts from patients with mevalonate kinase deficiency (MKD), a severe multisystemic disorder of cholesterol and non-sterol isoprenoid biosynthesis. In response to different concentrations of LDL and non-lipoprotein-bound cholesterol, MKD cells partially counteracted their enzyme defect by increased activities of 3-hydroxy-3-methylglutaryl (HMG)-CoA reductase (results from earlier studies) and the LDL receptor pathway, responses similar to the pharmacologic effects seen upon administration of HMG-CoA reductase inhibitors. Rates of N-linked protein glycosylation, estimated as the amount of [14C]galactose-labeled macromolecules secreted into cell culture medium, were significantly decreased in MKD fibroblasts in comparison with control cells which may indicate alterations in the dolichol or dolichol phosphate pool. In response to exogenous cholesterol, the major feedback inhibitor of isoprenoid biosynthesis, growth velocities of MKD fibroblasts declined in comparison with control cells, further suggesting an impairment of non-sterol isoprenoid biosynthesis in MKD. Our results suggest an imbalance in the multilevel regulation of the biosynthesis of cholesterol and non-sterol isoprenoids in MKD, representing an additional causative factor responsible for the pre- and postnatal pathology of MKD.

Long-term effect of cyclic AMP on N-glycosylation is caused by an increase in the activity of the cis-prenyltransferase.

Previously we have shown that long-term pretreatment of JEG-3 choriocarcinoma cells with 8-bromo-cAMP increases the capacity for N-glycosylation that was caused by an 8-10-fold enlargement of the dolichol pyrophosphoryl oligosaccharide (Dol-PP-oligosaccharide) pool [Konrad and Merz (1994) J. Biol. Chem. 269, 8659-8666]. The factors involved in the effect of cAMP on synthesis of Dol-PP-oligosaccharide are investigated here. The GlcNAc transfer to dolichol phosphate (Dol-P) was found to be unaffected by pretreatment with 8-bromo-cAMP. By measuring the uptake of [3H]mevalonate, a 20-fold increase in the incorporation of the label into Dol-P was observed in the cells treated with 8-bromo-cAMP. Under the same conditions, the synthesis of dolichol was enhanced 60-fold. However, the incorporation of the radioactivity into cholesterol was not increased in the JEG-3 cells pretreated with 8-bromo-cAMP, which suggests a specific stimulation of the dolichol/Dol-P pathway by cAMP. The cis-prenyltransferase activity was found to be increased 10-fold in cells pretreated with 8-bromo-cAMP. Dolichol kinase activity was unaffected by stimulation with 8-bromo-cAMP. The present study suggests that the larger glycosylation capacity in JEG-3 cells treated with 8-bromo-cAMP is caused by an increase in the microsomal cis-prenyltransferase activity.

Lipid composition in scrapie-infected mouse brain: prion infection increases the levels of dolichyl phosphate and ubiquinone.

The neutral and phospholipid composition of mouse brain infected with scrapie prions was investigated. During the later stages of this disease, the level of dolichol decreased by 30% whereas the level of dolichyl phosphate increased by 30%. In terminally ill mice, there was also a 2.5-fold increase in both total ubiquinone and its reduced form. Furthermore, alpha-tocopherol was elevated at this stage by 50%. In contrast, no changes were observed in phospholipid amount, in phospholipid composition, and in phosphatidylethanolamine plasmalogen content during the entire disease process. The fatty acid and aldehyde composition of individual phospholipids remained unaltered as well. No modifications could be detected in cholesterol content. Thus, the majority of membrane lipids in scrapie-infected mouse brain are modified in neither quantity nor structure, but specific changes occur to a few polyisoprenoid lipids. This specificity indicates that, although prions accumulate in lysosomes, the infection process is not associated with a general membrane destruction caused by lysosomal enzyme leakage.

The biosynthesis of dehydrodolichyl phosphates by rat liver microsomes.

Using improved conditions with rat liver microsomes in the presence of 20% glycerol and 2% Triton X-100 at pH 8.5 it was shown that dehydrodolichyl diphosphate and dehydrodolichyl phosphate were synthesized from isopentenyl diphosphate and farnesyl diphosphate. Small amounts of geranylgeranyl diphosphate and geranylgeranyl phosphate were also formed. The carbon chain lengths of the dehydrodolichyl diphosphate and dehydrodolichyl phosphate were identical (C80-C85). A kinetic study showed that dehydrodolichyl diphosphate formed from farnesyl diphosphate and isopentenyl diphosphate was subsequently hydrolyzed to dehydrodolichyl phosphate. As the concentration of isopentenyl diphosphate was increased from 1 to 50 microM, the chain-length distribution of dehydrodolichyl products shifted from C75-C80 to C80-C85. Addition of MgCl2 into the assay mixture decreased product formation, but did not affect the chain-length distribution (C80-C85). The shift of the chain-length distribution to the same as that observed in naturally occurring dolichol derivatives (C90-C95) was observed when Triton X-100 was omitted from the assay mixture, although deletion of the detergent decreased the enzyme activity. These results, which provide insight into optimal conditions for enzymatic synthesis of the dolichol chain, are discussed in the context of the in vivo pathway for dolichol biosynthesis.

Differential effect of inflammation and dexamethasone on dolichol and dolichol phosphate synthesis.

Inflammation and glucocorticoids stimulate hepatic glycoprotein synthesis, resulting in an increased secretion of serum glycoproteins. We now present evidence that the synthesis of dolichol and dolichol phosphate from mevalonate is increased in hepatocytes from inflamed rats. Also, in inflamed rats, the levels of dolichol and dolichol phosphate are increased in liver homogenates and microsomes. Dexamethasone treatment of the cells, however, does not increase the synthesis of dolichol and dolichol phosphate from mevalonate. The results suggest that the inflammation-induced dolichol-linked saccharide and glycoprotein synthesis is possibly mediated through an increase in the level of dolichol and dolichol phosphate in the liver. Since dexamethasone treatment does not increase the synthesis of dolichol and dolichol phosphate, its action on glycoprotein synthesis appears to be different and to affect the induction of enzymes in mannosyl phosphoryl dolichol- and dolichol-linked oligosaccharide synthesis.

CTP-dependent lipid kinases of yeast.

Membrane fractions from yeast Saccharomyces cerevisiae catalyzed a transfer of gamma-phosphate from [gamma-32P]CTP into membranous lipids. Phosphorylated compounds were identified as phosphatidic acid and dolichyl phosphate (DolP). The membrane fraction also catalyzed phosphorylation of the exogenous dolichol. The activity of the phosphorylating enzymes could be modified by the yeast growing conditions; i.e., the enzyme from yeast grown aerobically favored the synthesis of phosphatidate over dolichyl phosphate in the ratio of 3:1, whereas the membrane fraction from anaerobically grown yeast synthesized PA and DolP in the ratio of 0.5:1. The activity of the phosphorylating enzymes could also be modified by divalent cations and the concentration of detergents. Phosphorylation of lipids does not occur in the presence of [gamma-32P]ATP and is not influenced by the presence of UTP or GTP. This result points to the specific role of CTP as a gamma-phosphate donor for the synthesis of phosphatidate and dolichyl phosphates in the yeast system.

Galactosyltransferase activities in mitochondrial outer membrane: biosynthesis of dolichylmonophosphate-galactose.

Mitochondrial outer membranes were prepared from mouse liver homogenates by swelling purified mitochondria in phosphate buffer and were purified on a discontinuous sucrose gradient. Assays for marker enzymes and controls in electron microscopy confirmed the purity and homogeneity of this subfraction. Mitochondrial outer membranes had significant galactosyltransferase activity when incubated with UDP-[14C]galactose: 14C-labelling was found in products extractable with organic solvents and in a residual precipitate. Addition of exogenous dolichylmonophosphate loaded into phosphatidylcholine liposomes strongly enhanced the incorporation of [14C]galactose into chloroform/methanol (2:1, v/v) -extractable products. Thin-layer chromatography of these 2:1 extracts showed that the increase of [14C]galactose incorporation was attributable to the synthesis of a new galactosylated lipid, 'lipid L'. This 'lipid L' has been purified on silicic acid columns by elution with chloroform/methanol (1:1, v/v). The purified 'lipid L' was labile in acid and released [14C]galactose. It had the same chromatographic behaviour as dolichylmonophosphate-mannose in neutral, acid and alkaline solvent systems. Upon incubation in presence of [3H]dolichylmonophosphate and UDP-[14C]galactose, purified 'lipid L' contained both 3H- and 14C-labelling. 'Lipid L', synthesized by mitochondrial outer membranes, was therefore characterized as dolichylmonophosphate-galactose.

Characterization of the Saccharomyces cerevisiae cis-prenyltransferase required for dolichyl phosphate biosynthesis.

The prenyltransferase involved in the biosynthesis of dolichyl phosphate has been characterized in Saccharomyces cerevisiae. Although the enzyme is predominantly membrane-bound, a significant percentage was found in the soluble fraction. The prenyltransferase preferentially utilizes farnesyl pyrophosphate as the allylic substrate and isopentenyl pyrophosphate as cosubstrate with half-maximal velocities obtained at 25 and 6.7 microM, respectively. The enzymatic activity is sensitive to sulfhydryl reagents and is inhibited by all detergents tested, except 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate at concentrations less than 5 mM. The product of the reaction has been characterized as an alpha-unsaturated polyprenyl pyrophosphate, containing 12-15 isoprene units, approximately two isoprene units shorter than the endogenous yeast dolichyl phosphate. The stereochemistry of addition of isoprene units by the prenyltransferase was shown to be cis by a comparison of the HPLC retention time for a pentadecaprenyl phosphate derived from the in vitro reaction product with that for an authentic mixture of alpha-cis- and alpha-trans-pentadecaprenyl phosphates.

Cell-cycle dependence of dolichyl phosphate biosynthesis.

The cell-cycle dependence of dolichyl phosphate biosynthesis has been investigated in mouse L-1210 cells fractionated by centrifugal elutriation. Dolichyl phosphate levels increased linearly through the cell cycle, reaching a value in late S phase twice that of early G1. The cell-cycle dependences of four dolichyl phosphate metabolizing enzymes have been measured: cis-prenyltransferase, CTP-dependent dolichol kinase, dolichyl phosphatase, and dolichyl pyrophosphatase. The kinase, the cis-prenyltransferase, and the pyrophosphatase showed cell-cycle variations, increasing through G1 to a maximum in S phase while the monophosphatase activity was cell-cycle independent. The rate of accumulation of dolichyl phosphate was not affected by growing the cells in mevalonolactone-supplemented media. The evidence presented is consistent with models in which either the cis-prenyltransferase or the kinase/phosphatase couple (or both) regulates the levels of dolichyl phosphate in the cell.

Synthesis of dolichol derivatives in trypanosomatids. Characterization of enzymatic patterns.

We have previously described that in certain parasitic protozoa, namely the trypanosomatids, the dolichol-P-P-linked oligosaccharides synthesized in vivo and transferred to protein are devoid of glucose residues and contain 6, 7, or 9 mannose units depending on the species. We have now conducted a cell-free characterization of the enzymatic patterns responsible for these phenotypes. Microsomes from Trypanosoma cruzi, Crithidia fasciculata, Leishmania enriettii, and Blastocrithidia culicis were found to synthesize dolichol-P-[14C]Man but not dolichol-P-[14C]Glc when incubated with rat liver dolichol-P and GDP-[14C]Man or UDP-[14C]Glc, thus providing for an explanation to the absence of glucosylated dolichol-P-P derivatives. Formation of dolichol-P-P-oligosaccharides was assayed in incubation mixtures containing rat liver dolichol-P, GDP-[14C]Man, microsomes, and unlabeled Man5-8GlcNAc2-P-P-dolichol from bovine liver. Membranes from species synthesizing dolichol-P-P-linked Man6GlcNAc2 or Man7GlcNAc2 in vivo were found to synthesize the same compounds but not the higher homologues in the cell-free assay. Species forming Man9GlcNAc2-P-P-dolichol in vivo were found to synthesize lipid-linked Man7GlcNAc2, Man8GlcNAc2, and Man9GlcNAc2 in vitro. It is concluded that there are at least three and probably four different dolichol-P-Man-dependent enzymatic activities involved in the synthesis of dolichol-P-P-linked Man9GlcNAc2 and that microorganisms not forming that compound are devoid of all mannosyltransferases responsible for the addition of the missing residues and not only of the enzyme involved in the synthesis of the homologue higher than the oligosaccharide occurring in vivo by a single mannose unit.