Carbohydrate moieties of glycoproteins a re-evaluation of their function

Glycosylation-deficient mutants and inhibitors of glycosulation have been used to investigate the biological function of the carbohydrate moieties of glycoproteins. These and other experimental findings are reviewed and critically evaluated in the present treatise.An hypothesis is proposed to explain the biological significance of the covalent attachment of carbohydrate to protein. It is proposed that the carbohydrate acts as a chemical ‘tag’ which, upon interaction with a specific intracellular membrane receptor, directs glycoproteins to specific cellular organelles following synthesis on the rough endoplasmic reticulum. It is further proposed that secretion of extracellular products is a more general phenomenon which appears not to be absolutely carbohydrate dependent. Also, data are presented which support the view that the carbohydrate moeity is required for the proteolytic or conformational stabilization of the protein component of glycoproteins, but not for the mediation of protein specific biological activity. A model is presented which suggests, for the first time, that the localization function of carbohydrates is not restricted to lysosomal enzymes.     

A benzoboroxole‐based affinity ligand for glycoprotein purification at physiological pH

Developing ligands capable of carbohydrate recognition has become increasingly important as the essential roles of glycoproteins and glycolipids in a diverse array of cellular signaling, pathophysiology, and immune response mechanisms are elucidated. Effective ligands for the glycan portion of glycoproteins and glycolipids are needed for pre‐enrichment proteomics strategies, as well as for the purification of individual glycoproteins from complex biological milieu encountered both in biochemistry research and bio‐pharmaceutical development. In this work, we developed a carbohydrate specific affinity ligand for glycoprotein purification using a one‐pot, multi‐component synthesis reaction (Ugi synthesis) and an amine‐functionalized benzoboroxole moiety immobilized on agarose beads. Benzoboroxoles are unique boronic acid derivatives that have recently been found to bind specifically to the cis‐diol groups of carbohydrates at physiological pH, with superior affinity to any other Wulff‐type boronic acid. The solid‐phase affinity ligand developed herein specifically binds the carbohydrate moiety of the glycoprotein glucose oxidase, as well as a fluorescein isothiocyanate‐dextran, as shown through deglycosylation binding studies. Additionally, the ligand is able to purify glucose oxidase from crude Escherichia coli lysate, at physiological pH, equitably to commercially available boronic acid‐functionalized agarose beads that required alkaline pH conditions. Thus, this affinity ligand is a marked improvement on current, commercially available boronic acid‐based glycoprotein enrichment matrices and has the potential to exhibit high individual glycoprotein specificity because of the additional functional groups available for variation on the Ugi scaffold. Copyright © 2015 John Wiley & Sons, Ltd.     

Carbohydrate synthesis and biosynthesis technologies for cracking of the glycan code: Recent advances

The glycan code of glycoproteins can be conceptually defined at molecular level by the sequence of well characterized glycans attached to evolutionarily predetermined amino acids along the polypeptide chain. Functional consequences of protein glycosylation are numerous, and include a hierarchy of properties from general physicochemical characteristics such as solubility, stability and protection of the polypeptide from the environment up to specific glycan interactions. Definition of the glycan code for glycoproteins has been so far hampered by the lack of chemically defined glycoprotein glycoforms that proved to be extremely difficult to purify from natural sources, and the total chemical synthesis of which has been hitherto possible only for very small molecular species. This review summarizes the recent progress in chemical and chemoenzymatic synthesis of complex glycans and their protein conjugates. Progress in our understanding of the ways in which a particular glycoprotein glycoform gives rise to a unique set of functional properties is now having far reaching implications for the biotechnology of important glycodrugs such as therapeutical monoclonal antibodies, glycoprotein hormones, carbohydrate conjugates used for vaccination and other practically important protein–carbohydrate conjugates.     

Sugar residues on proteins

Glycoproteins have become increasingly important in the structure and function of many different mammalian systems; for example, membrane glycoproteins and glycoprotein hormones. It is, therefore, important to understand their chemistry, which would include an understanding of both the carbohydrate and protein parts of the molecule. Since the chemical characterization of the protein moiety has been extensively examined and the techniques for its characterization are well worked out, only the carbohydrate portion of glycoproteins will be reviewed in this article. The chemical nature of the carbohydrate moiety of glycoproteins will be examined. First, the types of monosaccharides present in animal systems, especially those in the mammalian systems, will be described. Next, various types of simple and complex carbohydrate chains will be discussed to establish the diversity, size, and number of chains present in the carbohydrate units in different glycoproteins. Then, the type of linkages of the carbohydrate to the protein will be examined to determine if the primary sequence of protein is important in determining the size and type of carbohydrate chains present in glycoproteins. Finally, the current methods of structural elucidation such as monosaccharide sequence, intersugar bonds, and anomeric linkages in the carbohydrate moiety of glycoproteins will be reviewed. These methods include the techniques of periodate oxidation, methylation, partial acid hydrolysis, and specific glycosidase digestion of glycoproteins, as well as the latest techniques using micromethods of carbohydrate quantitation and characterization involving gas chromatography and mass spectrometry. The function of the carbohydrate in glycoproteins will also be considered. First, hormone glycoproteins will be discussed in their relationship to the immunological and biological function of the glycoprotein when the carbohydrate is sequentially removed. Next, the function of the carbohydrate in the turnover of glycoproteins will be discussed. These topics will be considered in order to develop an understanding of a specific function(s) of the carbohydrate in glycoproteins.     

A hypothesis on the biological role of ABH,lewis and P blood group determinant structures in glycosphingolipids and glycoproteins

A hypothesis is presented that glycosphingolipids of circulating erythrocytes are membrane-packing substances providing for an energetically cheap carbohydrate protective coat at the cell surface. The glycosphingolipids should cover the membrane surface not occupied by functional glycoproteins. This role is envisaged for the globo series of glycosphingolipids which are P^k and P antigens of human blood. Glycosphingolipids of the neolacto series terminated with non-informative A, B, H. Lewis, P₁ antigenic structures as well as with sialic acid residues should serve the same purpose. These carbohydrate structures may be also used for conferring biological inertness on otherwise functionally active carbohydrate structures and provide protection for circulatory and membrane glycoproteins from proteolysis, denaturation and recognition of potentially antigenic sites of protein moieties by the immunosurveillance system of the body. At the external body surface the same carbohydrate structures may protect cells from the action of pathogenic microorganisms and other environmental factors. The roles of the above mentioned carbohydrate sequences on glycosphingolipids and glycoproteins in the development, tumorigenesis and evolution of blood group polymorphism are discussed.Abbreviations GP glycoprotein - GSL glycosphingolipid - GC glycoconjugate     

Maturation of Borna disease virus glycoprotein

The maturation of Borna disease virus (BDV) glycoprotein GP was studied in regard to intracellular compartmentalization, compartmentalization signal-domains, proteolytic processing, and packaging into virus particles. Our data show that BDV-GP is (i) predominantly located in the endoplasmic reticulum (ER), (ii) partially exists in the ER already as cleaved subunits GP-N and GP-C, (iii) is directed to the ER/cis-Golgi region by its transmembrane and/or cytoplasmic domains in CD8-BDV-GP hybrid constructs and (iv) is incorporated in the virus particles as authentic BDV glycoprotein exclusively in the cleaved form decorated with N-glycans of the complex type. Downregulation of BDV-glycoproteins on the cell surface, their limited proteolytic processing, and protection of antigenic epitopes on the viral glycoproteins by host-identical N-glycans are different strategies for persistent virus infections.     

A galactose-rich, cell-wall glycoprotein from styles of Nicotiana alata

A basic, galactose-rich style glycoprotein (GaRSGP) encoded by a previously characterized style-specific cDNA (NaPRP4) has been isolated from the styles of Nicotiana alata and structurally characterized. The glycoprotein is associated with cell walls in the transmitting tract and is composed of approximately 25% (w/w) protein and 75% (w/w) carbohydrate. The purified glycoprotein appears as a smear of between 45–120 kDa on SDS—PAGE; the deglycosylated protein backbone has an apparent molecular weight of approximately 30 kDa. The glycoprotein is rich in the amino acids lysine, proline, and hydroxyproline and in the monosaccharides galactose and arabinose. It is one of only a few proline/hydroxyproline-rich glycoproteins (P/HRGPs) to be characterized both as a cDNA-clone and protein. Glycans are attached to the protein backbone through both O - and N -glycosidic linkages with the majority of the carbohydrate being O -linked and consisting of short, highly branched chains terminating primarily in galactose residues. A carbohydrate epitope(s) is found on both GaRSGP and another style-specific glycoprotein but not on glycoproteins from other tissues. This finding provides further evidence for the existence of a style-specific carbohydrate epitope(s) which may play a role in style function.     

Approaches to the study of N-linked glycoproteins in human plasma using lectin affinity chromatography and nano-HPLC coupled to electrospray linear ion trap--Fourier transform mass spectrometry

In this publication, we will describe the combination of lectin affinity chromatography with nano high performance liquid chromatography (HPLC) coupled to a linear ion trap Fourier transform mass spectrometer (capillary LC-LTQ/FTMS) to characterize N-linked glycosylation structures in human plasma proteins. We used a well-characterized glycoprotein, tissue plasminogen activator (rt-PA), which is present at low levels in blood, as a standard to determine the dynamic range of this approach. N-linked glycopeptides derived from rt-PA could be characterized at a ratio of 1:200 in human plasma (rtPA: total plasma protein, w/w) by accurate mass measurement in the FTMS and fragmentation (MS(n)) in the linear ion trap. We demonstrated that this platform has the potential to characterize the general N-linked glycosylation structures of abundant glycoproteins present in human plasma without the requirement for antibody-based purification, or additional carbohydrate analytical protocols. This conclusion was supported by the determination of carbohydrate structures for three glycoproteins, IgG, haptoglobin, and alpha-1-acid glycoprotein, at their natural levels in a human plasma sample, but only after the lectin enrichment step.     

Use of specific monoclonal and polyclonal antibodies to define distinct antigens of the porcine zonae pellucidae

Specific monoclonal and polyclonal antibodies to solubilized porcine and rabbit zonae pellucidae (ZP) and to purified ZP glycoprotein components have been used to define distinct ZP antigens. These studies demonstrate that the individual ZP glycoproteins contain both unique and shared determinants. One monoclonal antibody (R5) has been used to demonstrate that the major porcine ZP glycoprotein, which has multiple charge species ranging in molecular weight from 42,000 to 120,000, is composed of two distinct polypeptide antigens unique to this glycoprotein class. These distinct antigens can be differentiated by immunoblotting after high-resolution two-dimensional polyacrylamide gel electrophoretic separation of trypsin-treated or deglycosylated glycoproteins. The two polypeptides also differ in their staining properties with the silver-based color stain and in their susceptibility to proteolysis. A second monoclonal antibody (PSI) has been used to define a determinant shared by all three major porcine ZP glycoprotein classes. This determinant appears to involve either a carbohydrate moiety or some other molecular feature related to post-translational modification, since the antibody recognizes only the acidic species of each glycoprotein class, and does not recognize the deglycosylated forms of the proteins. This work demonstrates that there are both unique and shared antigenic determinants present in the individual components of the ZP, but that the immunodominant determinants appear to be unique to each glycoprotein.     

Use of carbohydrate probes in conjunction with fluorescence-activated cell sorting to select mutant cell lines of Chlamydomonas with defects in cell surface glycoproteins

Two carbohydrate-binding probes (the lectin concanavalin A and the anti-carbohydrate monoclonal antibody FMG-1) have been utilized in conjunction with fluorescence-activated cell sorting to select cell lines of Chlamydomonas reinhardtii that contain defects in cell surface-exposed glycoproteins. Two very different selection strategies (sorting cells with the lowest binding for the FMG-1 monoclonal antibody or the highest binding of concanavalin A) yield a class of mutant cells that exhibit a total lack of binding of the monoclonal antibody to cell wall and plasma membrane glycoproteins along with an increased affinity for concanavalin A. Detailed characterization of one such mutant cell line, designated L-23, is provided. The subtle glycosylation defect exhibited by this cell line does not alter the ability of the affected glycoproteins to be targeted to the flagellar membrane and does not affect the expression of flagellar surface motility, a phenomenon that appears to involve the major concanavalin A-binding glycoprotein of the flagellar membrane. This approach has general applicability for dissecting the role of carbohydrate epitopes in the targeting and function of any cell surface glycoprotein for which suitable carbohydrate probes are available.     

Glycosidase inhibitors: inhibitors of N-linked oligosaccharide processing.

The biosynthesis of the various types of N-linked oligosaccharide structures involves two series of reactions: 1) the formation of the lipid-linked saccharide precursor, Glc3Man9(GlcNAc)2-pyrophosphoryl-dolichol, by the stepwise addition of GlcNAc, mannose and glucose to dolichyl-P, and 2) the removal of glucose and mannose by membrane-bound glycosidases and the addition of GlcNAc, galactose, sialic acid, and fucose by Golgi-localized glycosyltransferases to produce different complex oligosaccharide structures. For most glycoproteins, the precise role of the carbohydrate is still not known, but specific N-linked oligosaccharide structures are key players in targeting of lysosomal hydrolases to the lysosomes, in the clearance of asialoglycoproteins from the serum, and in some cases of cell:cell adhesion. Furthermore, many glycoproteins have more than one N-linked oligosaccharide, and these oligosaccharides on the same protein frequently have different structures. Thus, one oligosaccharide may be of the high-mannose type whereas another may be a complex chain. One approach to determining the role of specific structures in glycoprotein function is to use inhibitors that block the modification reactions at different steps, causing the cell to produce glycoproteins with altered carbohydrate structures. The function of these glycoproteins can then be assessed. A number of alkaloid-like compounds have been identified that are specific inhibitors of the glucosidases and mannosidases involved in glycoprotein processing. These compounds cause the formation of glycoproteins with glucose-containing high mannose structures, or various high-mannose or hybrid chains, depending on the site of inhibition. These inhibitors have also been useful for studying the processing pathway and for comparing processing enzymes from different organisms.     

Estimating glycoprotein carbohydrate chain structures by lectin reactivities in polyacrylamide gels

Complex mixtures of cellular glycoproteins contain a myriad of different carbohydrate chains that cannot be easily analyzed without rigorous purification of each individual glycoprotein. We have analyzed the carbohydrate chains in complex mixtures of cellular glycoproteins by separation using sodium dodecyl-sulfate polyacrylamide gel electrophoresis and interacting the gels with several 125I-labeled lectins. By use of in situ chemical modifications of the glycoproteins after their electrophoretic separation together with the known carbohydrate-binding specifities of several lectins, it has been possible to estimate glycoprotein carbohydrate chain structures. As an example we have examined the cellular glycoproteins of a ovary-colonizing metastatic variant of B16 melanoma and report the types of carbohydrate chains that are found on various melanoma glycoproteins.     

Use of lambda gt11 and monoclonal antibodies to map the genes for the six major glycoproteins of equine herpesvirus 1.

To localize the genes for the major glycoproteins of equine herpesvirus 1 (EHV-1), a library of the EHV-1 genome was constructed in the lambda gt11 expression vector. Recombinant bacteriophage expressing EHV-1 glycoprotein epitopes as fusion products with beta-galactosidase were detected by immunoscreening with monoclonal antibodies specific for each of six EHV-1 glycoproteins. Seventy-four recombinant lambda gt11 clones reactive with EHV-1 monoclonal antibodies were detected among 4 X 10(5) phage screened. Phage expressing determinants on each of the six EHV-1 glycoproteins were represented in the library. Herpesviral DNA sequences contained in lambda gt11 recombinants expressing epitopes of EHV-1 glycoproteins were used as hybridization probes for mapping insert sequences on the viral genome. Genes for five EHV-1 glycoproteins (gp2, gp10, gp13, gp14, and gp21/22a) mapped to the genome L component; only one EHV-1 glycoprotein (gp17/18) was expressed from the unique S region of the genome where genes of several major glycoproteins of other herpesviruses have been located. Two glycoproteins of EHV-1, gp13 and gp14, mapped to positions colinear with genes of major glycoproteins identified in several other alphaherpesviruses (gC- and gB-like glycoproteins, respectively). The genomic locations of other EHV-1 glycoproteins indicated the existence of major glycoproteins of EHV-1 (gp2, gp10, and gp21/22a) for which no genetic homologs have yet been detected in other herpesviruses. The results confirm the general utility of the lambda gt11 expression system for localizing herpesvirus genes and suggest that the genomic positioning of several high-abundance glycoproteins of EHV-1 may be different from that of the prototype alphaherpesvirus, herpes simplex virus.     

Mucin-specific bark lectin from elderberry Sambucus sieboldiana and its applications to the affinity chromatography of mucin

Three bark lectins were isolated from elderberry Sambucus sieboldiana using fetuin-Sepharose 4B and mucin-Sepharose 4B, and were studied comparatively for their binding to glycoprotein and to clarify various physicochemical features. For each, a unique pattern on isoelectric focusing was noted and their affinity toward various glycoproteins differed, indicating the structures of their carbohydrate binding sites possibly differ. One bark lectin showed specific binding toward porcine mucin. The purity of mucin from a crude porcine stomach mucin or an extract of porcine submaxillary glands could be improved by affinity chromatography on immobilized lectin having binding specificity toward mucin.     

Structure of a major yolk glycoprotein and its processing pathway by limited proteolysis are conserved in echinoids

To study the fate of the yolk glycoproteins found in eggs and embryos of the sea urchin, Strongylocentrotus purpuratus, a polyclonal antibody to a 90-kDa polymannose glycoprotein found in the embryo was prepared. Immunoblot analysis of total proteins over the course of development showed that this antibody recognized a family of glycoproteins. Concomitant with the disappearance of the major 160-kDa yolk glycoprotein of the egg during embryogenesis, glycoproteins with a lower molecular mass appeared. These glycoproteins (115, 108, 90, 83, and 68 kDa) were purified from S. purpuratus and analyzed by limited proteolysis and peptide mapping. This analysis revealed that these glycoproteins were cleavage products derived from the major yolk glycoprotein. The antibody to the 90-kDa glycoprotein in S. purpuratus embryos was used to identify a homologous set of yolk glycoproteins with similar molecular masses in the embryos of three other species in the class Echinoidea: Arbacia punctulata, Lytechinus pictus, and Dendraster excentricus. However, eggs from other echinoderm classes and from Xenopus laevis, Drosophila melanogaster, and the chicken did not contain any cross-reactive molecules. Cross-reactivity within the class Echinoidea was not due to a common carbohydrate epitope, because the antibody recognized the glycoproteins even after the N-linked carbohydrate side chains were enzymatically removed. The major yolk glycoprotein (160-170 kDa) from each of the three sea urchin species was purified and analyzed. Comparison of the physical and chemical properties of these glycoproteins revealed striking similarities in pI and in amino acid and monosaccharide composition. The results of peptide mapping also supported the conclusion that the 160- to 170-kDa glycoproteins from the four echinoids are structurally homologous glycoproteins containing N-linked polymannose chains. Immunolocalization by electron microscopy in S. purpuratus showed that the yolk glycoproteins remained within the yolk platelet throughout development, and that externalization of the 160-kDa glycoprotein or its cleavage products was not detectable.     

Effect of tunicamycin on herpes simplex virus glycoproteins and infectious virus production.

The antibiotic tunicamycin, which blocks the synthesis of glycoproteins, inhibited the production of infectious herpes simplex virus. In the presence of this drug, [14C]glucosamine and [3H]mannose incorporation was reduced in infected cells, whereas total protein synthesis was not affected. Gel electrophoresis of [2-3H]mannose-labeled polypeptides failed to detect glycoprotein D or any of the other herpes simplex virus glycoproteins. By use of specific antisera we demonstrated that in the presence of tunicamycin the normal precursors to viral glycoproteins failed to appear. Instead, lower-molecular-weight polypeptides were found which were antigenically and structurally related to the glycosylated proteins. Evidence is presented to show that blocking the addition of carbohydrate to glycoprotein precursors with tunicamycin results in the disappearance of molecules, possibly due to degradation of the unglycosylated polypeptides. We infer that the added carbohydrate either stabilizes the envelope proteins or provides the proper structure for correct processing of the molecules needed for infectivity.     

Multivalent protein-carbohydrate interactions. A new paradigm for supermolecular assembly and signal transduction

Many biological recognition processes involve the binding and clustering of ligand-receptor complexes and concomitant signal transduction events. Such interactions have recently been observed in human T cells in which binding and cross-linking of specific glycoprotein counter-receptors on the surface of the cells by an endogenous bivalent carbohydrate binding protein (galectin-1) leads to apoptosis [Pace, K. E., et al. (1999) J. Immunol. 163, 3801-3811]. Importantly, different counter-receptors associated with specific phosphatase or kinase activities were shown to form separate clusters on the surface of the cells as a result of galectin-1 binding to the carbohydrate moieties of the respective glycoproteins. This suggests that the unique separation and organization of signaling molecules that results from galectin-1 binding is involved in delivering the signal to die. The ability of galectin-1 to induce the separation of specific glycoprotein receptors was modeled on the basis of molecular and structural studies of the binding of multivalent carbohydrates to lectins that result in the formation of specific two- and three-dimensional cross-linked lattices. These latter studies have been recently highlighted by X-ray crystallographic results showing that a single tetravalent lectin forms distinct cross-linked complexes with four different bivalent oligosaccharides [Olsen, L. R., et al. (1997) Biochemistry 36, 15073-15080]. In this report, binding and cross-linking of multivalent carbohydrates with multivalent lectins is shown to be a new paradigm for supermolecular assembly and signal transduction in biological systems.     

Carbohydrate composition of purified serum glycoproteins in Mucolipidosis II and Mucolipidosis III

Summary Mucolipidosis II (I-cell disease) and Mucolipidosis III (ML III) are inherited disorders in which the molecular defect may involve an abnormality in a common post-translational modification step (possibly glycosylation) shared by lysosomal hydrolases. We tested whether such an alteration might be a generalized defect in glycoprotein biosynthesis and, thus, be reflected in an abnormal carbohydrate composition of non-lysosomal glycoproteins. The apoprotein of low density lipoprotein (apo-LDL) and immunoglobulin G (IgG) were purified to apparent homogeneity. Gas liquid chromatographic (glc) analysis of the carbohydrate content of these glycoproteins from ML II, ML III and normal sera revealed no differences in the relative ratios and total amounts of mannose, galactose, N-acetylglucosamine and sialic acid. These results suggest that if the postulated post-translational defect in these disorders involves changes in carbohydrate composition, it is not a general defect in glycosylation and may be specific for lysosomal hydrolases.     

The role of N-linked oligosaccharides in glycoprotein function.

N-linked glycoproteins include such biologically important molecules as cell-surface receptors, cell-adhesion molecules, immunoglobulins and other serum proteins, and tumor antigens. Investigating the role of carbohydrate in glycoprotein function has included the use of glycosylation inhibitors or site-directed mutagenesis of specific glycosylation sites to prevent the addition of carbohydrate, or glycosylation processing inhibitors or animal cell glycosylation mutants to alter carbohydrate structure. In some proteins, glycosylation plays an important role in recognition, while in others, it may stabilize and/or control the conformation of the protein. The cloning of genes in bacteria or lower eukaryotes--with the goal of producing biologically active proteins for biotechnological purposes--necessitates a better understanding of the role of specific carbohydrate structures.     

Distribution and metabolism of glycoproteins and glycosaminoglycans in subcellular fractions of brain.

The distribution, carbohydrate composition, and metabolism of glycoproteins have been studied in mitochondria, microsomes, axons, and whole rat brain, as well as in various synaptosomal subfractions, including the soluble protein, mitochondria, and synaptic membranes. Approximately 90% of the brain glycoproteins occur in the particulate fraction, and they are present in particularly high amounts in synaptic and microsomal membranes, where the concentration of glycoprotein carbohydrate is 2-3% of the lipid-free dry weight. Treatment of purified synaptic membranes with 0.2% Triton X-100 extracted 70% of the glycoprotein carbohydrate but only 35% of the lipid-free protein residue, and the resulting synaptic membrane subfractions differed significantly in carbohydrate composition. The glycoproteins which are not extracted by Triton X-100 also have a more rapid turnover, as indicated by the 80-155% higher specific activity of hexosamine and sialic acid 1 day after labeling with [3H]glucosamine in vivo. The specific activity of sialic acid in the synaptosomal soluble glycoproteins 2 hr after labeling was greater than 100 times that of the synaptosomal particulate fraction, whereas the difference in hexosamine specific activity in these two fractions was only twofold, and by 22 hr there was little or no difference in the specific activities of sialic acid and hexosamine in synaptosomal soluble as compared to membrane glycoproteins. These data indicate that sialic acid may be added locally to synaptosomal soluble glycoproteins before there is significant labeling of nerve ending glycoproteins by axoplasmic transport. Fifty to sixty percent of the hyaluronic acid and heparan sulfate of brain is located in the various membranes comprising the microsomal fraction, whereas half of the chondroitin sulfate is soluble and only one-third is in microsomal membranes. When microsomes are subfractionated on a discontinuous density gradient over half of the hyaluronic acid and chondroitin sulfate are found in membranes with a density less than that of 0.5 M sucrose (representing a six- to sevenfold enrichment over their concentrations in the membranes applied to the gradient), whereas half of the heparan sulfate is present in membranes with a density greater than that of 0.8 M.