Improved synthesis of CMP-sialates using enzymes from frog liver and equine submandibular gland

Enzymatic synthesis of nucleotide sugars

The present review gives a survey on the biosynthetic pathways of nucleotide sugars which are important for the in vitro synthesis of mammalian glycoconjugates. With respect to the use of these enzymes in glycotechnology the availability as recombinant enzymes from different sources, the large-scale synthesis of nucleotide sugars and their in situ regeneration in combination with glycosyltransferases are summarized and evaluated.

Applications of enzymes in synthetic carbohydrate chemistry

By photosynthesis 200 billion tons of glucose are formed per annum most of which remains in the sugar state. Thus, the well-known properties of saccharides as the basis for structural materials and energy stores is well established. Increasingly, focus is being addressed to their functions as recognition molecule. Therefore, carbohydrate chemistry became en vogue again as documented by classical syntheses of complex hetero-oligosaccharides employing protecting and activating group chemistry. In recent years enzymatic methods for acylation and deacylation in saccharide chemistry came into use. Novel saccharides could be approached following aldolase-catalyzed de novo syntheses. The very complex task of stereospecific glycosylation was addressed by use of enzymes. With glycohydrolases via transglycosylation as well as with glycosyltransferases including cofactor regenerating circles advances were made. Modified acceptor and also donor substrates allowed access to oligosaccharides in preparative quantities. This paper will discuss these modern developments and will focus on some selected examples from others' and our own laboratory.

Catalytic properties of the CMP-N-acetylneuraminic acid hydroxylase from the starfish Asterias rubens: comparison with the mammalian enzyme

The biosynthesis of N-glycolylneuraminic acid (Neu5Gc) was investigated in cell-free extracts of the starfish Asterias rubens, which is one of the evolutionarily least-advanced species known to possess Neu5Gc-containing glycoconjugates. As in higher animals, Neu5Gc is synthesised in Asterias rubens by the action of a CMP-Neu5Ac hydroxylase. Enzyme activity was detected in all starfish tissues tested, the turnover being the greatest in the gonads. The enzyme from this tissue has a temperature optimum between 25 and 33 degrees C and a pH optimum between pH 6.0 and 6.4. This hydroxylase exhibits many characteristics in common with the mammalian enzyme. For example, the enzyme is extracted in a predominantly soluble form. Oxygen and a reduced pyridine nucleotide are necessary for activity, with NADH being the most effective cofactor. Furthermore, the activation of the hydroxylase by exogenously added iron salts and the potent inhibitory effects of several iron ligands point to the involvement of a non-haem iron cofactor. The enzyme has a high affinity for the substrate CMP-Neu5Ac, the apparent Km being 18 microM. In contrast to the mammalian enzyme, the hydroxylase from Asterias rubens is not inhibited by increased ionic strength and cannot be activated by non-ionic detergents. Moreover, the CMP-Neu5Ac turnover increased linearily with increasing protein concentration. In accordance with other enzymes in starfish, seasonal changes in the CMP-Neu5Ac hydroxylase activity were also observed.

Enzyme-catalyzed oligosaccharide synthesis

Cell-surface carbohydrates and their conjugates are involved in many types of molecular recognition. This review describes recent developments in enzyme-catalyzed oligosaccharide synthesis, with particular focus on glycosyltransferase and glycosidase reactions. With the increasing availability of glycosyltransferases via recombinant DNA technology, glycosyltransferase-catalyzed glycosylation with in situ regeneration of sugar nucleotides appears to be the most effective method for large-scale stereocontrolled oligosaccharide synthesis.

A study on the regulation of N-glycoloylneuraminic acid biosynthesis and utilization in rat and mouse liver

The relative contribution of N-glycoloyl-beta-D-neuraminic acid (Neu5Gc) to total sialic acids expressed in mouse and rat liver glycoconjugates was found to be 95% and 11%, respectively. This considerable difference in sialic acid composition made these two tissues suitable models for a comparative investigation into the regulation of Neu5Gc biosynthesis and utilization. An examination of the CMP-glycoside specificity of Golgi-associated sialyltransferases using CMP-N-acetyl-beta-D-neuraminic acid (CMP-Neu5Ac) and CMP-Neu5Gc revealed no significant tissue-dependent differences. The Golgi membrane CMP-sialic acid transport system from rat liver did, however, exhibit a slightly higher internalisation rate for CMP-Neu5Ac, though no preferential affinity for this sugar nucleotide over CMP-Neu5Gc was observed. In experiments, where Golgi membrane preparations were incubated with an equimolar mixture of labelled CMP-Neu5Ac and CMP-Neu5Gc, no significant tissue-dependent differences in [14C]sialic acid composition were observed, either in the luminal soluble sialic acid fraction or in the precipitable sialic acid fraction, results which are consistent with the above observations. From this experiment, evidence was also obtained for the presence of a Golgi-lumen-associated CMP--sialic acid hydrolase which exhibited no apparent specificity for either CMP-Neu5Ac or CMP-Neu5Gc. The specific activity of the CMP-Neu5Ac hydroxylase, the enzyme responsible for the biosynthesis of Neu5Gc, was found to be 28-fold greater in high-speed supernatants of mouse liver than of rat liver. No hydroxylase activity was detected in the Golgi membrane preparations. It is therefore proposed that the cytoplasmic ratio of CMP-Neu5Ac and CMP-Neu5Gc produced by the hydroxylase, remains largely unmodified after CMP-glycoside uptake into the Golgi apparatus and transfer on to growing glycoconjugate glycan chains. The close relationship between the total sialic acid composition and the sialic acid pattern in the CMP-glycoside pools of the tissues lends considerable weight to this hypothesis.

The side-chain conformations of N-acetyl-7-,8-,9-deoxy-, and -4,7-dideoxy-neuraminic acid and their effect on the activation of CTP:N-acylneuraminic acid cytidylyltransferase

The conformations of the deoxy side-chain analogs of N-acetylneuraminic acid and of N-acetyl-4-deoxyneuraminic acid have been studied by n.m.r. spectroscopy, which allowed us to distinguish between local minima conformations suggested by hard-sphere calculations. The conformations were correlated with the activity with CTP: N-acetylneuraminic acid cytidylyltransferase (CMPsialate synthase; EC 2.7.7.43).

Transport of CMP-N-glycoloylneuraminic acid into mouse liver Golgi vesicles

CMP-Neu5Gc has been shown to be transported into mouse liver Golgi vesicles by a specific carrier the characteristics of which were investigated in detail. In the system employed, CMP-Neu5Gc enters the Golgi vesicles within 2 min; transport was saturable with high concentrations of the sugar-nucleotide and was dependent on temperature. A kinetic analysis gave an apparent Km of 1.3 microM and a maximal transport velocity of 335 pmol/mg protein per min. Almost identical values were obtained with CMP-Neu5Ac, under the same incubation conditions. Furthermore, the uptake of CMP-Neu5Gc was inhibited by CMP-Neu5Ac, a substrate analogue. Conversely, the uptake of CMP-Neu5Ac was inhibited by CMP-Neu5Gc to the same extent, leading to the conclusion that the transport of CMP-Neu5Ac and CMP-Neu5Gc is mediated by the same carrier molecule. This transport system for CMP-Neu5Gc involves both CMP and CMP-Neu5Gc since intravesicular CMP induced the entry of external CMP-Neu5Gc.

Activation and transfer of novel synthetic 9-substituted sialic acids

In this report several NeuAc analogues differently modified at position C-9 were tested as substrates for CMP sialic acid synthase from bovine brain: the hydroxy group at C-9 was replaced by an amino, acetamido, benzamido, hexanoylamido and azido group. The synthase was partially purified by chromatography on CDP-hexanolamine--Sepharose. CMP-glycosides synthesized were measured by analytical HPLC at 275 nm. Each NeuAc analogue was activated to the respective CMP-glycoside: Km-values varied from 0.8 mM to 4.6 mM, the Km for NeuAc was 1.4 mM. Thus affinity of the enzyme was influenced only moderately by chemical modification at C-9. CMP-glycosides were synthesized on a preparative scale with good yield and characterized by analytical HPLC. In addition, 500-MHz 1H-NMR data of CMP-9-amino-NeuAc and CMP-9-acetamido-NeuAc were obtained. Each CMP-activated NeuAc analogue was a suitable donor substrate for Gal beta 1-4GlcNAc alpha 2,6-sialyltransferase from rat liver. Transfer was determined by the thiobarbituric acid method and by analytical HPLC at 200 nm. The results demonstrate that synthetic, not naturally occurring, non-labelled NeuAc analogues can be incorporated into glycoprotein with high yield.

Studies on acylneuraminate cytidylytransferase from human placenta

The presence of acylneuraminate cytidylyltransferase has not been shown in human placenta so far. In this paper, evidence is put forward to demonstrate the presence of this enzyme in human placenta. The soluble enzyme was partially purified 30-fold using ammonium sulphate fractionation (30-60%) and DEAE Sephadex A-50 chromatography. This enzyme preparation had a specific activity of 0.75 mkat/kg protein. The pH optimum of the reaction was 9.0. The Km values for N-acetylneuraminic acid and cytidine triphosphate were 2.2 mM and 4 mM respectively. It was also found that the presence of magnesium is necessary for the enzyme activity.

Chemical behaviour of cytidine 5'-monophospho-N-acetyl-beta-D-neuraminic acid under neutral and alkaline conditions

The chemical behaviour of CMP-N-acetylneuraminic acid under neutral and different alkaline conditions has been investigated. The products formed were isolated by ion-exchange chromatography and gel filtration and analysed by colorimetric methods, thin-layer chromatography, combined gas-liquid chromatography/mass spectrometry and/or 360-MHz 1H-NMR spectroscopy. A maximum stability of CMP-N-acetylneuraminic acid was observed at pH8-11. In the tested pH range of 6-13, CMP and N-acetylneuraminic acid were formed in variable amounts as decomposition products. 2-Deoxy-2,3-dehydro-N-acetylneuraminic acid was produced at pH greater than 7; the amount of this substance increased with increasing pH. In anhydrous triethylamine its yield was 50%. A new neuraminic acid derivative, N-acetyl-beta-D-neuraminic acid 2-phosphate, could be isolated from the mixture of alkaline decomposition products of CMP-N-acetylneuraminic acid. The yield of this compound was maximum 22% in anhydrous triethylamine. Because 2-deoxy-2,3-dehydro-N-acetylneuraminic acid was formed under simulated physiological conditions, it is assumed that this compound, which occurs in tissues and fluids of man and animals, is derived from CMP-N-acetylneuraminic acid non-enzymically also under conditions in vivo.

Isolation and characterization of acylneuraminate cytidylyltransferase from frog liver

Frog liver (Rana esculenta) is a rich source of acylneuraminate cytidylyltransferase. The soluble enzyme was purified 250-fold almost to purity with 25% yield and a specific activity of 9 mkat/kg protein (0.54 U/mg protein) using DEAE Sephadex and Sepharose 6B chromatography, followed by preparative polyacrylamide gel electrophoresis. The molecular weight of the cytidylyltransferase was determined to be 163 000 with the aid of Sepharose 6B chromatography and gel electrophoresis, with or without dodecyl sulphate or urea. No subunits were found. The isoelectric point of the enzyme is at pH 6. Optimum reaction rate was observed at pH 9, 37 degrees C, 50mM Mg2 or Ca2 and ImM mercaptoethanol. The Km values for N-acetylneuraminic acid, N-glycoloylneuraminic acid and CTP are 1.6mM, 2.3 mM and 0.6mM, respectively. O-Acetylated sialic acids are inactive with the cytidylyltransferase from frog liver. Enzyme activity can be inhibited by SH reagents and CMP (Ki = 0.5mM).

Metabolism of 4-O-methyl-N-acetylneuraminic acid a synthetic sialic acid

1. 4-O-Methyl-N-acetylneuraminic acid shows a strong positive periodate-thiobarbiturate reaction. The mechanism of dye formation in this test for sialic acids is discussed in view of the studies already published. 2. An efficient preparation of a tritium-labelled 4-O-methyl-N-acetylneuraminic acid, with high specific radioactivity, by an oxymercuration-demercuration procedure is presented. 3. Sialytransferase activities in microsomal fractions of equine liver using desialylated fetuin are studied. The enzyme activity, assayed in a radioactive procedure, shows an apparent Km value for CMP-N-acetylneuraminic acid of 0.7 mM, whereas this value is 3.4 mM for CMP-4-O-methyl-N-acetylneuraminic acid. Differences are also observed in the maximal velocity for the two substrates. 4. The equine liver system can be used to prepare substantial amounts of fetuin containing radioactive N-acetylneuraminic acid or 4-O-methyl-N-acetylneuraminic acid. The isolated reaction products show similar sialic acid release by treatments with acid or fowl-plague virus neuraminidase. In contrast, 4-O-methyl-N-acetylneuraminic acid-fetuin displays a marked resistance to desialylation by Vibrio cholerae neuraminidase. 5. Free 4-O-methyl-N-acetylneuraminic acid is completely resistant to the action of acylneuraminate pyruvate-lyase. It does not inhibit the enzymic cleavage reaction of N-acetylneuraminic acid. 6. The influence of a substitution at C-4 neuraminic acid on the enzymatic reaction mechanisms is discussed.

The specificity of sialytransferases using glycosylated lysozyme derivatives as substrates

Galactose, lactose, N-acetylgalactosamine, N-acetylglucosamine and fibrinoglycopeptides were bound to lysozyme by different linkages. These glycosylated lysozymes were tested as N-acetylneuraminic acid acceptors using particular sialytransferase preparations from frog and bovine liver and from bovine and porcine submandibular glands. Desialylated fetuin served as reference compound. Galactose residues of desialo-fetuin and lysozyme-lactose are sialylated by all four sialytransferases tested, galactose bound to lysozyme via a phenylazo group is inactive with the enzyme from bovine submandibular gland, and galactose bound directly to lysozyme serves as substrate only for the frog liver sialytransferase. Lysozyme-phenylazo-N-acetylgalactosamine is active only with the sialytransferase from bovine sumbandibular gland. N-Acetylglucosamine derivatives of lysozyme are inactive with all sialytransferases tested. These observations are discussed in the light of the natural substrates for the sialytransferases investigated.