Effects of the protein matrix on glycan processing in glycoproteins

In the biosynthesis of glycoproteins containing asparagine-linked glycans, a number of regulatory factors must be involved in converting the single glycan precursor into the variety of different final structures observed in different eukaryotic species. Among these factors are the kind of glycan-processing enzymes available in the Golgi apparatus of different cells, the specificity and regulatory properties of these enzymes, and the unique properties of the protein matrix in which a given glycan resides during the biosynthetic processing. In examining the role of this latter regulatory factor, we have considered a simplified model in which a few key steps are common to all cells, regardless of the nature of the processing enzymes available. The protein-bound oligomannose precursor Man8GlcNAc2-, arriving in the Golgi after the initial trimming in the endoplasmic reticulum (ER), first undergoes a series of preprocessing steps to yield Man5GlcNAc2- in animals and plants or Man13-15GlcNAc2- in yeast. At this stage the key commitment step--to process or not to process--determines whether the above intermediates will remain as unprocessed oligomannose structures or be initiated into a new series of reactions to yield processed structures characteristic of the organisms involved (complex or hybrid for vertebrates, polymannose for yeast, xylosylated glycans for plants and some invertebrates, or Man3GlcNAc2- structures for other invertebrates). It is proposed that this commitment step, along with the obligatory preprocessing steps, is regulated primarily by each glycan's unique exposure on its protein matrix. Subsequent processing steps leading to complex or hybrid structures, fucosylation, extent of branching, and specific structures at the nonreducing terminals are most likely determined primarily by the enzyme makeup of the individual processing machineries, but with the protein matrix still playing a significant role.     

The effect of the protein matrix on glycoprotein processing by oviduct Golgi enzymes

Using the avidin-biotinyl glycan system reported previously (Shao, M.-C., and Wold, F. (1987) J. Biol. Chem. 267, 2968-2972), we have compared the processing efficiency of oviduct enzymes acting on different glycan-(biotinyl)Asn and glycan-(6-biotinamidohexanoyl)Asn derivatives when they are free and bound to avidin. The glycans were selected to permit exploration of the individual processing steps, and the two different groups of derivatives were used to assess both the close (biotinyl) and more distal (biotinamidohexanoyl) display of the glycan relative to the avidin surface. The direct comparison of the free and avidin-bound glycans demonstrated that mannosidase I is strongly inhibited by avidin in both the close and distal complexes, whereas GlcNAc transferase I and mannosidase II are strongly inhibited only in the close complex. GlcNAc transferases III, IV, and V, which could only be assessed individually by indirect means using different substrates, did not appear to be affected in any major way by the protein matrix; the data suggest that transferase III is inhibited only to a minor extent in the close complex. Gal transferase activity showed a minor effect of the avidin matrix for both complexes in the hybrid processing pathways. The most significant consequence of the avidin effect on Gal transferase was the apparent abolishment of the incorporation of a 2nd Gal residue in the two avidin complexes. This survey of the protein matrix effects on glycan processing by oviduct enzymes appears to provide reasonable clues to the origin of the very different glycan structures observed in oviduct-processed glycoproteins. Thus, ovalbumin and avidin itself, containing a mixture of oligomannose and hybrid glycans at their single glycosylation sites, may well present they glycans to the processing enzymes in a display very similar to that of the avidin close complex observed here. The inhibition of mannosidase I and GlcNAc transferase I lead to preservation of oligomannose structures, whereas the strong inhibition of mannosidase II favors the incorporation of the bisecting GlcNAc by GlcNAc transferase III to yield hybrid structures as the most processed products. Ovomucoid, which contains multiantennary complex structures at all glycosylation sites, may on the other hand display its glycans, unencumbered by the protein surface, in conformations similar to either the free glycans or the distal complexes observed in this work.     

The regulation of glycan processing in glycoproteins. The effect of avidin on individual steps in the processing of biotinylated glycan derivatives

The effect of the protein matrix on glycan processing by rat liver Golgi enzymes has been evaluated by a direct comparison of substrate----products conversion of a free glycan and of the same glycan linked to a protein. The glycan substrates had the general structure R-glycan where R represented either biotinyl-Asn-GlcNAc2- or 6-(biotinamido)hexanoyl-Asn-Glc-NAc2- and the protein used was avidin; the extension arm in one of the glycan substrates permitted the additional comparison of two avidin-biotin-glycan complexes. By the use of different glycans as substrates, by the presence or absence of donor substrates (UDP-GlcNAc, UDP-Gal, and CMP-sialic acid (Sia) and/or the inhibitor, swainsonine, it was possible to dissect the individual steps involved in the conversion of R-Man6 (or R-Man5) to a biantennary complex glycan, R-Man3-GlcNAc2-Gal2-Sia2 or to the hybrid glycan R-Man5-GlcNAc-Gal-Sia. Using fast atom bombardment-mass spectrometry to identify and quantify the substrates and products of each parallel incubation of free and avidin-bound substrates, the following observations were made. With the substrate without the extension arm, avidin-binding inhibited mannosidase I, GlcNAc transferase I, and the second step of the reaction catalyzed by mannosidase II (R-Man4-GlcNAc----R-Man3-GlcNAc); the second step of the reaction catalyzed by Gal-transferase was also inhibited to a lesser extent. This inhibition was greatly reduced or absent with the substrates with the extension arm and was consequently referred to as the short range effect. A long range effect of avidin binding expressed by both substrates with and without extension arm was observed for Gal-transferase acting in the hybrid glycan pathway (R-Man5-GlcNAc----R-Man5-GlcNAc-Gal) in the presence of swainsonine and also for Sia-transferase in the catalysis of the incorporation of the second Sia residue into the complex product (R-Man3-GlcNA2-Gal2-Sia----R-Man3-GlcNAc2- Gal2-Sia2) and to a lesser extent in the hybrid pathway (R-Man5-GlcNAc-Gal----R-Man5-GlcNAc-Gal-Sia). GlcNAc transferase II did not appear to be affected by avidin. Based on the information available on the biotin-binding site in avidin, it is proposed that the short range effect reflects the masking of the core chitobiose unit in the avidin-glycan complexes in the absence of the extension arm, but not in the presence of the arm, and that the early processing enzymes thus may require a fully exposed chitobiose for full activity.(ABSTRACT TRUNCATED AT 400 WORDS)     

Protein matrix effects on glycan processing by mannosidase II and sialyl transferase from rat liver

The effect of the protein environment on the reaction sequence and the relative rates of two two-step reactions involved in the biosynthesis of complex glycans in glycoproteins has been explored by comparing the processing of biotinylated substrates either free or bound to avidin. By use of biotinyl and biotinamidohexanoyl derivatives, the display of the glycan in a proximal and distal association with the avidin surface could also be assessed. Mannosidase II removes two Man residues from the substrate GlcNAcMan5GlcNAc2-R to yield GlcNAcMAn3GlcNAc2-R. The NMR spectra of the substrate, intermediate, and product showed that the first Man is removed from the 6-arm of the substrate. The rate constants for the first and second step (estimated by direct analysis of the reactants by anion-exchange chromatography with a pulsed amperometric detector) were determined to be about 0.05 and 0.08 min-1, respectively, for the free substrates. In the proximal complex k1 was reduced 80-fold, and the k2 step could not be observed under the same conditions. In the distal complex both k1 and k2 were reduced about 8-fold. Sialyl transferases transfer Sia from CMP-Sia to the biantennary substrate Gal2GlcNAc2-Man3GlcNA2-R to yield the product Sia2Gal2-GlcNAc2Man3GlcNAc2-R with the Sia linked either 2-3 or 2-6 to the Gal residues. The NMR spectra showed that the first step involved the Gal on the 3-arm of the substrate and that both Sia residues were added 2-6.(ABSTRACT TRUNCATED AT 250 WORDS)     

Early effects of aminonucleoside on enzymes in the Golgi apparatus of rat liver.

Rats were injected with a single intravenous dose of aminonucleoside (AMN) and sacrificed 1-48 h later. The activity of several enzymes was assayed in the Golgi apparatus isolated from the liver. Galactosyltransferase activity showed little changes after the AMN, but both acid (EC 3.1.3.2) and alkaline phosphatase (EC 3.1.3.1) activities increased within the first hour and reached control levels only 5-24 h later. Thiamine pyrophosphatase and arylsulfatase A (EC 3.1.6.1) activities also increased and stayed at higher levels for the duration of the experiment. Arylsulfatase B (EC 3.1.6.1) activity decreased shortly after the AMN but later increased to above control levels. These findings support earlier results in which liver ultrastructural and biochemical changes were observed early before renal lesions and proteinuria.     

The effect of cycloheximide on galactosyltransferase of rat liver Golgi membranes.

An inhibitory effect of cycloheximide on the initial rate of galactosyltransferase of rat liver Golgi membranes has been demonstrated. Cycloheximide was effective in inhibiting the activity of the enzyme when added directly to the assay medium or after pre-incubation of the membranes with the drug. The inhibition observed with different concentrations of nucleotide sugar was shown to be competitive at higher concentrations of the nucleotide sugar (0.10-1.0 mumol). The inhibition observed with different concentrations of acceptor, N-acetylglucosamine was complex and could not be analysed further with the present data. Washing the Golgi membranes previously incubated with cycloheximide with water failed to reverse the inhibition. Washing with UDPgalactose partially reversed the inhibition only. These results, together with the observation that serum galactosyltransferase was not inhibited by cycloheximide supported the view that the cycloheximide effect may be primarily on the membrane system.     

Regulation of glycan processing by Golgi enzymes from red kidney bean (Phaseolus vulgaris) seedlings.

The effect of a protein matrix on the processing of glycoprotein glycans by Golgi enzymes from plant seedlings has been determined with an artificial glycoprotein system, comparing the processing rates of glycan-(biotinyl)Asn (or glycan-(biotinamidohexanoyl)Asn) substrates either free or bound to avidin. An analysis of the pooled glycoproteins from the seedlings suggested that the most common glycan structure is a complex one (GlcNAc-Man3GlycNAc2-protein), and consistent with this processing end-product, mannosidases I and II and GlcNAc transferases I and II were all found to be present in the seedling Golgi membrane preparations. The effect of the avidin matrix either in a proximal (biotinyl substrates) or distal (N-(biotinamido)hexonoyl substrates) association with the appropriate glycan substrate for these four enzymes was assessed from the direct comparison of the apparent first-order rate constants for the free and avidin-bound substrate-product conversions. All four plant enzymes were inhibited by the association of the glycan substrates with avidin, but the inhibition was much less pronounced than that observed with the corresponding enzymes from rat liver and hen oviduct. The rate effect shows a progression from 3- to 10-fold rate decreases in the proximal complexes and 2- to 3-fold in the distal complexes in going from the first (mannosidase I) to the fourth (GlcNAc transferase II) enzyme; with the mammalian and avian enzymes the largest effects were for the first ones and much larger absolute rate effects were observed. The results suggest that the nature of the processing enzymes in terms of this response to the avidin glycan substrates may differ in different organisms.     

GO-PROMTO illuminates protein membrane topologies of glycan biosynthetic enzymes in the Golgi apparatus of living tissues

The Golgi apparatus is the main site of glycan biosynthesis in eukaryotes. Better understanding of the membrane topology of the proteins and enzymes involved can impart new mechanistic insights into these processes. Publically available bioinformatic tools provide highly variable predictions of membrane topologies for given proteins. Therefore we devised a non-invasive experimental method by which the membrane topologies of Golgi-resident proteins can be determined in the Golgi apparatus in living tissues. A Golgi marker was used to construct a series of reporters based on the principle of bimolecular fluorescence complementation. The reporters and proteins of interest were recombinantly fused to split halves of yellow fluorescent protein (YFP) and transiently co-expressed with the reporters in the Nicotiana benthamiana leaf tissue. Output signals were binary, showing either the presence or absence of fluorescence with signal morphologies characteristic of the Golgi apparatus and endoplasmic reticulum (ER). The method allows prompt and robust determinations of membrane topologies of Golgi-resident proteins and is termed GO-PROMTO (for GOlgi PROtein Membrane TOpology). We applied GO-PROMTO to examine the topologies of proteins involved in the biosynthesis of plant cell wall polysaccharides including xyloglucan and arabinan. The results suggest the existence of novel biosynthetic mechanisms involving transports of intermediates across Golgi membranes.     

Kinetic characterization of calcium uptake by the rat liver Golgi apparatus

We carried out a kinetic characterization of the Ca(2+)active transport in the rat liver Golgi Apparatus (GA) membrane. Calcium accumulation by vesicles of a GA enriched fraction was found to be a function of both Ca(2+)and ATP-Mg concentrations, it was inhibited by 2 microm thapsigargin but not stimulated by 3 microm calmodulin. The kinetic parameter values obtained for the GA Ca(2+)pump were: J(max)of 3.96 nmol/mg min, K(m)for Ca(2+)of 0.150 microm and two K(m)'s for ATP of 1.14 microm and 519 microm. These results were almost identical to those obtained for the endoplasmic reticulum (ER) fraction, indicating that the GA Ca(2+)pump is a sarco/endoplasmic reticulum (SERCA) P-type, analogous-if not identical-to that present in the ER.     

Relationship of the glycan structure of glycoproteins to the desialylation process by rat liver endothelium

To investigate the variations in desialylation of glycoproteins by liver endothelium, we compared endothelial desialylation for 3 glycoproteins, human ceruloplasmin, human and rat transferrin. Radiolabeled glycoproteins were chased through purified rat liver endothelium and then fractionated by lectin affinity chromatography. Endothelium processed glycoproteins were fractionated by RCA120 chromatography into sialylated and desialylated components. The latter was then studied by Con A chromatography. Desialylation occurred only when the molecule contained at least a single triantennary chain of glycan. Desialylation was minimal in the case of human transferrin which contains mostly biantennary branching pattern. Thus, it appears that a single triantennary glycan chain is necessary and sufficient to trigger desialylation of glycoproteins by liver endothelium and this process is an all-or-none phenomenon.     

Biogenesis of microsomal membrane glycoproteins in rat liver. III. Release of glycoproteins from the Golgi fraction and their transfer to microsomal membranes

The presence in the Golgi fraction of glycoproteins destined to be incorporated into the microsomal membrane was investigated. When incubated in sucrose, washed Golgi vesicles released four major, weakly acidic glycoproteins, some of which could be incorporated into microsomal membranes by incubation. Double labeling with [3H]glucosamine and [14C]leucine demonstrated the incorporation of both protein and oligosaccharide moieties, and the main peak of radioactivity was associated with the 70,000 mol wt region after SDS-gel electrophoresis. The proteins that could be incorporated into microsomes were probably associated to a large extent with the outer surface of the Golgi membrane. Centrifugation of the proteins released from the Golgi in a KBr solution (p = 1.24) resulted in a separation of glycoproteins, those in the top layer most actively incorporated into microsomes. The lipoglycoproteins in the top layer that could be incorporated appeared in the 70,000 mol wt region after SDS-gel electrophoresis, as did the corresponding proteins isolated from the supernate. These results suggest that glycoproteins with completed oligosaccharide chains are released from the Golgi system to the cytosol and are subsequently transferred to microsomes as constitutive membrane components.     

The effect of protein depletion on the rate of protein synthesis in rat liver.

In order to understand the mechanism of decreased protein synthesis in the liver of rats fed a protein-free diet, the average polypeptide chain assembly time (tc) was measured by the method of Mathews et al. (J. Biol. Chem. (1973) 248, 1329). For rats fed a normal diet, tc in liver in vivo was 1.28 min. A 10-day period of protein depletion led to a value of tc = 2.08 min, corresponding to a 38% depression in polypeptide elongation rate. Protein depletion caused an extensive breakdown of hepatic polysomes and refeeding of a complete mixture of amino acids resulted in rapid recovery of polysomal profile. But tc in the liver of the refed animals gave still depressed value of 1.95 min. The amount and size distribution of poly(A)-containing mRNA in the liver, as determined by [3H]poly(U) hybridization, were the same for normal and depleted groups. These results suggest that both initiation and elongation steps of protein synthesis are depressed in the liver of protein-depleted rats. Refeeding of amino acid mixture rapidly restores initiation but not elongation activity.     

Evaluation of the role of rat liver Golgi endo-alpha-D-mannosidase in processing N-linked oligosaccharides

Golgi membranes from rat liver have been shown to contain an endo-alpha-D-mannosidase which can convert Glc1Man9GlcNAc to Man8GlcNAc with the release of Glc alpha 1----3Man (Lubas, W. A., and Spiro, R. G. (1987) J. Biol. Chem. 262, 3775-3781). We now report that this enzyme has the capacity to cleave the alpha 1----2 linkage between the glucose-substituted mannose residue and the remainder of the polymannose branch in a wide range of oligosaccharides (Glc3Man9GlcNAc to Glc1Man4GlcNAc) as well as glycopeptides and oligosaccharide-lipids. Whereas the tri- and diglucosylated species (Glc3Man9GlcNAc and Glc2Man9GlcNAc), which yielded Glc3Man and Glc2Man, respectively, were processed more slowly than Glc1Man9GlcNAc, the monoglucosylated components with truncated mannose chains (Glc1Man8GlcNAc to Glc1Man4GlcNAc) were trimmed at an increased rate which was inversely related to the number of mannose residues present. The endomannosidase was not inhibited by a number of agents which are known to interfere with N-linked oligosaccharide processing by exoglycosidases, including 1-deoxynojirimycin, castanospermine, bromoconduritol, 1-deoxymannojirimycin, swainsonine, and EDTA. However, Tris and other buffers containing primary hydroxyl groups substantially decreased its activity. After Triton solubilization, the endomannosidase was observed to be bound to immobilized wheat germ agglutinin, indicating the presence of a type of carbohydrate unit consistent with Golgi localization of the enzyme. The Man8GlcNAc isomer produced by endomannosidase action was found to be processed by Golgi enzymes through a different sequence of intermediates than the rough endoplasmic reticulum-generated Man8GlcNAc variant, in which the terminal mannose of the middle branch is absent. Whereas the latter oligosaccharide is converted to Man5GlcNAc via Man7GlcNAc and Man6GlcNAc at an even rate, the processing of the endomannosidase-derived Man8GlcNAc stalls at the Man6GlcNAc stage due to the apparent resistance to Golgi mannosidase I of the alpha 1,2-linked mannose of the middle branch. The results of our study suggest that the Golgi endomannosidase takes part in a processing route for N-linked oligosaccharides which have retained glucose beyond the rough endoplasmic reticulum; the distinctive nature of this pathway may influence the ultimate structure of the resulting carbohydrate units.     

The effect of wheat germ agglutinin on sialyl and galactosyltransferases of rat liver Golgi membranes.

The sialyltransferase and galactosyltransferase activities of the Golgi-rich fraction from rat liver were enhanced by the binding of wheat germ agglutinin (WGA). The sialytransferase was more sensitive than the galactosyltransferase to the WGA. Maximal stimulation of the galactosyltransferase activity resulted from the binding of 60--80 micrograms WGA to the Golgi membrane, while only 40 micrograms of WGA produced a maximal enhancement in the sialyltransferase activity. Within 5 min of WGA binding, the Golgi sialytransferase activity was doubled. After the initial binding of WGA to the Golgi fraction, the galactosyltransferase activity was decreased by 30%. However, in 15 min the activity was doubled by the binding of WGA. The activities of both enzymes were further enhanced by incubation for up to 90 min. The stimulation of both sialyltransferase and galactosyltransferase activities by WGA was reversed by N-acetyl-D-glucosamine (GlcNAc), the specific inhibitor of agglutination by WGA. Complete reversal of the enhanced activity was observed after 20--30 min in the presence of 1 micromol GlcNAc. The association constant for the binding of WGA to the Golgi membranes was calculated to be 4.16 X 10(-6) M from a Steck-Wallach plot. The 'n' value or mean binding sites was calculated as 5.26 X 10(-5) M/mg of Golgi membrane protein.     

The effect of turpentine-induced inflammation on rat liver glycosyltransferases and Golgi complex ultrastructure

The effect of vasopressin on the toad urinary bladder has been shown to be mediated by cyclic AMP. It has been assumed that, as demonstrated for other systems, this involves activation of cyclic AMP-dependent protein kinase. In order to test this hypothesis we investigated the effect of vasopressin on cyclic AMP-dependent protein kinases in epithelial cells of toad bladders. About 80% of protein kinase activity and cyclic AMP-binding capacity was found to be in the cytosol. DEAE-cellulose chromatography showed a pattern of 15--20% type I and 80--85% type II cyclic AMP-dependent protein kinase. Cytosolic kinase was activated 3--4-fold by cyclic AMP with half-maximal activation at 5 . 10(-8) M. Similarly, half-maximal binding of cyclic AMP occurred at 7 . 10(-8) M. Incubation of toad bladders in Ringer's solution containing 0.1 mM 3-isobutyl-1-methylxanthine, prior to homogenization and assay, showed stable cyclic AMP-binding capacity and protein kinase ratio --cyclic AMP/+cyclic AMP. Exposure of bladders to 10 mU/ml of vasopressin for 10 min caused intracellular activation of protein kinase and decrease in cyclic AMP-binding capacity that were maintained for at least 30 min. Incubation of bladders with increasing concentrations of vasopressin (0.5--100 mU/ml) resulted in a discrepancy between a progressive increase in cyclic AMP levels and a levelling off at 10 mU/ml of vasopressin for the changes in protein kinase ratio and cyclic AMP-binding capacity. The increase in kinase ratio was due to higher activity in the absence of exogenous cyclic AMP and was fully inhibitable by a specific protein kinase inhibitor. Using Sephadex G-25-CM50 column chromatography for separation of holoenzyme and free catalytic subunit we demonstrated that the activation of protein kinase in the vasopressin-treated bladders is due to intracellular dissociation of the kinase. These results show that the effect of vasopressin on the toad bladder involves activation of a cytosolic cyclic AMP-dependent protein kinase. The time course and the dose-response curve of the kinase activation closely parallel vasopressin's effect on osmotic water flow.     

The lipids of the Golgi apparatus subfractions from rat liver.

Golgi apparatus were isolated from untreated rat liver and separated into three fractions. One consisted mainly of vesicles, a second of tubular particles (dictyosomes) and the third was a mixed fraction. Large differences between these fractions could be seen in the electron microscope and by enzyme analysis. The total lipid content of the vesicles was 3.5-times greater than that of the dictyosomes and the neutral lipid value was 7-times greater. The ratio of phospholipids to protein was approximately the same in the three fractions. However, the phospholipid patterns differed between the vesicle and dictyosome fractions.     

Kinetic analysis of protein crystal nucleation in gel matrix

The effect of agarose on nucleation of hen egg white lysozyme crystal was examined quantitatively using a temperature-jumping technique. For the first time, to our knowledge, the inhibition of agarose during the nucleation of lysozyme was quantified in two respects: a), the effect of increasing interfacial nucleation barrier, described by the so-called interfacial correlation parameter f(m); and b), the ratio of diffusion to interfacial kinetics obtained from dynamic surface tension measurements. It follows from a dynamic surface tension analysis that the agarose network inhibits the nucleation of lysozyme by means of an enhancement of the repulsion and interfacial structure mismatch between foreign bodies and lysozyme crystals, slowing down the diffusion process of the protein molecules and clusters toward the crystal-fluid interface and inhibiting the rearrangement of protein molecules at the interface. Our results, based on ultraviolet-visible spectroscopy, also show no evidence of the supersaturation enhancement effect in protein agarose gels. The effects of nucleation suppression and transport limitation in gels result in bigger, fewer, and perhaps better quality protein crystals. The understandings obtained in this study will improve our knowledge in controlling the crystallization of proteins and other biomolecules.     

Kinetic analysis of rat liver sorbitol dehydrogenase

• 1.1. A steady state kinetic investigation was performed on an improved preparation of rat-liver sorbitol dehydrogenase (l-iditol: NAD-oxidoreductase, EC 1.1.1.14).
• 2.2. Data analyses indicate the enzyme follows a rapid equilibrium random mechanism in the direction of sorbitol oxidation and a random mechanism in the direction of fructose reduction.
• 3.3. Kinetic constants were: K_m^NAD 0.082 mM; K_m^sorbitol 0.38 mM; K_m^NADH 67 μm; K_m^fructose 136 μM.
• 4.4. Evidence is adduced to indicate the more rapid reverse (fructose reduction) reaction is susceptible to metabolic control by formation of abortive enzyme-fructose-NAD and enzyme-NADH-sorbitol complexes.