sign in sign up
<<
Home , Glyceraldehyde, Nonose

Review: Biochemistry of blood group antigens

L.G. G ILLIVER AND S.M. H ENRY

This review presents the basics of the structural glycan (the largest size expected for a carbohydrate chemistry of blood group glycoconjugates, with special epitope). This is compared with only 46,656 reference to red cell serology. Its aim is to create an combinations possible with a chain made up of six appreciation of the inherent subtleties of the amino acids, more than seven orders of magnitude carbohydrate blood group antigens, which are lower than the diversity possible in glycans. 1 currently poorly understood within the field of blood Fortunately, despite the vast potential for complexity in transfusion. It is hoped that a better understanding of carbohydrate chains, there is generally only a limited the intricacies of the carbohydrate blood group range of structures representing the carbohydrate systems will lead to further contributions to the body blood group epitopes (glycotopes), although many of of knowledge within this growing field. the structural differences are subtle and some epitope surfaces may be common. These subtleties may represent different blood group antigens and in some Introduction In 1900, Karl Landsteiner discovered the cases, specific antibodies may recognize the common carbohydrate blood group system of ABO after mixing surfaces. For example, anti-A,B recognizes the the cells and serum of his colleagues. Over the next similarities between A and B antigens, not the subtleties 100 years many other blood group systems were which distinguish them. 2 Unfortunately, this potential discovered and later resolved at both the molecular and variation makes structural determination of carbo- the genetic level. It is now clearly recognized that most hydrate antigens complex, requiring combinations of blood group systems are either carbohydrate, protein, powerful techniques such as nuclear magnetic or protein-carbohydrate in nature, although they may resonance spectroscopy and mass spectrometry to associate with other components and complexes to be elucidate the structures of carbohydrate antigens recognized as specific blood group determinants. without ambiguity. 3,4 Progress in determining the Essentially, a blood group antigen is any polymorphic structures of carbohydrate blood group antigens has three-dimensional conformation expressed on the been slow, hindered by a combination of factors, such outer membrane of a RBC against which an antibody is as the fact that carbohydrate blood group antigens available. The concept of the three-dimensional shape are not the primary product of a gene—being rather is very important because this is what antibodies secondary biosynthetic products—and that the recognize, and the number of potential shapes that can isolation and purification of carbohydrate blood group be found in carbohydrate chains is enormous. The antigens in requisite quantities is difficult. As a result, variability in characteristics such as type, linkage we know only the basic structures of the major type, branch positioning, shape, charge, ring size, and glycotopes and antigens expressed in the carbohydrate epimeric (pairs of which differ only blood group systems. We have not yet characterized in the configuration about one atom are many of the phenotypic subtleties, such as those that epimers of each other e.g., D- and D- determine ABO subgroups. are epimers with respect to the C2 carbon atom) and Because carbohydrate blood group antigens are 12 anomeric configuration results in more than 1.05 × 10 important to transfusion medicine and are becoming possible structural combinations from a six -sugar more recognized in transplantation (especially xeno-

IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 33 L.G. G ILLIVER AND S.M. H ENRY

5 transplantation ) and studies of micoorganism binding Carbon numbering D-glucose L-glucose and interactions, 6 it is becoming more important for serologists to understand the biochemistry of carbohydrate blood group antigens. We presently recognize several blood group systems and collections 7 where the blood group glycotopes are comprised entirely of carbohydrate, such as ABO, Lewis, H, Ii, P, Sd, Cad, etc.; or others, such as MN, where carbohydrate moieties are essential to the structure of the epitope; or even others, where are associated with the character of the antigen—e.g., GPI-linked membrane proteins such as Cromer, Dombrock, Cartwright, and John Milton Hagen, etc. Fig. 1. Carbon numbering of linear and the D- and L- This review is not intended to cover the structures of glucose. The carbon numbering of a generic 6-car - bon sugar is shown on the left (with the chiral center hydro - carbohydrate blood group systems, rather it is intended gens and hydroxyl groups not shown). 2, 3, 4, and 5 to provide a basic understanding of carbohydrate are the chiral centers of 6-carbon sugars. The enantiomeric form is designated by the direction of the hydroxyl group on blood group chemistry, and in this regard we will use the highest numbered chiral carbon (arrows)—in the case of ABO antigens as examples. 6-carbon sugars, this is carbon 5. In D-glucose (middle), the hydroxyl group of carbon 5 points to the right, and in L-glucose (right), this hydroxyl group points to the left. Monosaccharides The common blood group–related sugars, namely total number of carbons in the structure). If this hydroxyl group points to the right when the carbonyl -D-glucose (Glc), N -acetyl - -D-glucosamine (GlcNAc), β β group is at the top, the is said to be in β-D- (Gal), N -acetyl -β-D-galactosamine the D configuration. The mirror image of this structure, (GalNAc), α-L- (Fuc), and D-mannose (Man) are all six -carbon sugars (). The carbonyl group with the hydroxyl group pointing to the left, is the L (C=O) of each residue is an (HC=O), and thus configuration. The D and L configurations of glucose they are all classified as aldohexoses. Another special are shown as open chain formulas monosaccharide which contains nine carbons is also (Fig. 1). In blood group glycoconjugates, sugars exist common in blood group antigens, with the most almost exclusively in the D configuration, except common blood group variation being N -acetyl - fucose, which is found in the L configuration. The chiral centers of saccharides can also be designated as α-D- (also referred to as NeuAc or sialic acid). There are various forms of sialic acid, 8,9 but these R or S under the Cahn-Ingold-Prelog system. This is a will not be reviewed here. more modern notation; however, it is not often used in Sugar molecules can have configurational carbohydrate chemistry because it causes confusion in which differ in their arrangement in space even though the classification of some monosaccharides. they have identical composition. The variation in the The seven common blood group sugars almost configuration of individual carbohydrates was first exclusively form pyran rings—six atoms forming the characterized in relation to the smallest chiral ring, compared with furan, which is a five-atom ring carbohydrate, , by German chemist Emil (Fig. 2). Fischer in the late 19th century. In extremely simple Numbering of the carbons in the ring terms, is “handedness”—that is, the existence formation of hexose and sugars is shown in of left/right apposition. For example, your left hand Figure 3. and right hand are mirror images and therefore“chiral.” The pyran ring comes about through the formation The enantiomeric (mirror image) configurations of an bridge between a carbon (carbon 1 in the Fischer described were designated D and L by Rosanoff common hexose blood group sugars, or carbon 2 in in 1906. 10 Today, all monosaccharides are classified sialic acid) and the oxygen of the hydroxyl group five according to this principle, which depends on the carbons away. 10 In hexoses, carbon 5 rotates so that its orientation of the hydroxyl group attached to the hydroxyl group apposes carbon 1. The double-bonded highest numbered chiral carbon (C (n -1) , where n is the oxygen on carbon 1 is reduced—taking the

34 IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 Review: carbohydrate blood group antigens

1 Generic furan ring Generic pyran ring

D-galactose 2

β-D-galactofuranose Open chain β-D-galactopyranose D-galactose 3 Fig. 2. Furan and pyran rings of D-galactose. The open chain formation of D-galactose is shown in the middle of the figure. The (lower left) and pyranose (lower right) forms of D-galactose ( β is used in this example) are formed from the open chain structure. The generic furan (upper left) and pyran (upper right) ring for - -D-galactopyranose mations are also shown. β Fig. 4. Scheme showing the formation of the pyran ring. Step 1— Carbon 5 (square) of the open chain form of D-galactose rotates so that its hydroxyl group apposes carbon 1 (circle). Step 2—The double-bonded oxygen on carbon 1 is reduced and the hydroxyl group of carbon 5 is oxidized, leaving carbon 1 and the oxygen on carbon 5 each with one less than their preferred number of bonds. Step 3—A bond forms between carbon 1 and the oxygen of carbon 5 to yield β-D-galactopyranose.

hexose sugars and carbon 2 in sialic acid. This carbon is termed the anomeric carbon, and the configurational difference occurs in the orientation of its hydroxyl group in relation to the oxygen on the anomeric- reference atom, which is carbon 5 in most hexoses.

Generic hexose Nonose sugar, i.e., sialic acid

Fig. 3. Carbon numbering in pyranose rings. Most of the common blood group sugars are 6-carbon (hexoses), except for sialic acid, which has 9 carbons (nonose). Carbons are numbered in a clockwise direction from the oxygen bridge. from the hydroxyl group of carbon 5, in the process oxidizing the hydroxyl group to an oxygen radical with only one bond. This leaves carbon 1 as a radical also, with only three bonds. This is one less than the preferred complement of bonds in the case of both atoms, enabling the formation of a bond between them (see Fig. 4 for the scheme of this reaction, using D- galactose as an example). The scheme is basically the same for sialic acid, except that the bridge is formed between carbon 2 and the oxygen on carbon 6. α-D-galactose β-D-galactose The two ring formations possible for each sugar are Fig. 5. Fischer linear ring projection showing the configurational differ - ence at the anomeric carbon hydroxyl group (top arrows) of α-D- either the α or β anomeric forms. These are two galactose (left) and β-D-galactose (right) in relation to the - isomeric forms that differ only in the configuration at ic reference oxygen (bottom arrows). The anomeric carbon hydroxyl group and the anomeric reference oxygen are on the the carbon involved in the formation of the oxygen same side of the carbon chain in the α-form, and on opposite bridge, which is carbon 1 in the common blood group sides of the carbon chain in the β-form.

IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 35 L.G. G ILLIVER AND S.M. H ENRY

When in the Fischer linear representation of the ring (Fig. 5), the α form occurs when the anomeric carbon hydroxyl is on the same side of the carbon chain as the oxygen on carbon 5. Accordingly, in the β form, the Amino group Acetyl group Acetamido group anomeric carbon hydroxyl and the oxygen on carbon 5 Fig. 7. The acetamido group of N-acetyl- D-galactosamine, N -acetyl- D- are on opposite sides. Due to the rotation of carbon 5 glucosamine and sialic acid. The acetamido group is made up of the amino group and the acetyl group. in hexoses (carbon 6 in sialic acid) and hence its hydroxyl group (Fig. 4), this rule cannot be applied to hydroxyl group of carbon two is replaced by an the Haworth projection of the ring. However, the acetamido group, which is an amino group with an anomeric form can be deduced in the Haworth acetyl group substituted for one of the projection—it is α when the anomeric hydroxyl group (Fig. 7). This important substitution differentiates is down, and β when the anomeric hydroxyl is up. blood group A from blood group B. Blood group A has The Haworth projections of the and forms of α β N-acetyl - -D-galactosamine, while blood group B has mannose and galactose are shown in Figure 6. The two α -D-galactose. Furthermore, modification of the cyclic and anomeric forms exist in equilibrium α α β blood group A sugar N -acetyl - -D-galactosamine by 11 α with the open chain structure, however, the open bacterial deacetylase can generate the chain configuration is present only in minute amounts. acquired B antigen (GalNAc to GalNH —Fig 6). This is Because these are interchangeable, the 2 where N -acetyl -α-D-galactosamine loses its acetyl anomerity of the sugar is really only of interest once it group but retains the nitrogen—thus, the acquired is made permanent by bonding with another B antigen is similar but not identical to the B saccharide. determinant α-D-galactose. N-acetyl -D-glucosamine most commonly adopts the β form in blood group glyco- conjugates, while N -acetyl -D- galactosamine most commonly adopts the α configuration— α-L-fucose α-D-galactose N-acetyl -α-D- Acquired B β-D-galactose although the form is also galactosamine determinant β seen. Sialic acid is unlike the other sugar residues discussed here in that it has nine carbons and the anomeric carbon is carbon 2 (Fig. 6). It is substituted at β-D-glucose N-acetyl -β-D- α-D-mannose β-D-mannose Sialic acid (N -acetyl -α-D- carbon 5 with an acetamido glucosamine neuraminic acid) group as described for carbon Fig. 6. Haworth projection formulae of the sugar residues commonly found in blood group–related glycoconjugates. 2 of N -acetyl -D-glucosamine and N -acetyl -D-galactosamine. Figure 6 also shows the other saccharides The carboxyl group (COOH) of carbon 1 gives sialic commonly found in blood group antigens. D-glucose acid its acid characteristics, and because of this, and D-galactose are most often found in the β form, glycoconjugates incorporating it are known as while L-fucose, which is almost identical in structure to gangliosides, which are a subset of acidic glyco- galactose (except that fucose does not have a hydroxyl sphingolipids. The acidic glycosphingolipid group also group on its sixth carbon, thus it is also known as encompasses uronoglycosphingolipids (containing 6-deoxygalactose) is found in the α form. D-mannose is uronic acid residues), sulfoglycosphingolipids a sugar residue commonly found in glycoproteins (containing sulfate ester groups), phosphoglyco- (exclusively N -linked ) and exists in sphingolipids (containing phosphate mono- or both the α and β configuration. diester groups), and phosphonoglycosphingolipids N-acetyl -D-glucosamine and N -acetyl -D-galactosa- (containing [2-aminoethyl]hydro-xyphosphoryl mine are examples of amino sugars (Fig. 6). The groups).

36 IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 Review: carbohydrate blood group antigens

Up to this point, the pyranose-forming mono- Bonds and Linkages saccharides have all been represented as having planar Sugar residues can be joined into chains by the rings with functional groups projecting above or synthesis of glycosidic bonds. As previously below. In reality this is not the case—the electron mentioned, the two monosaccharide cyclic forms— α orbitals of the carbons and the oxygen in the ring and and β—are in equilibrium with the open chain the large functional groups attached to the carbons formation when in aqueous solution. However, when prevent the rings from existing in a planar form. The joined together by glycosidic linkages, the residues result is that the ring contorts into one of three three- become fixed in ring formation except for the last dimensional shapes, two of which are suggestive of a sugar in the chain. 12 chair and one of which looks like a boat (Fig. 8). The Glycosidic bonds are the result of a condensation reaction between the hydroxyl group of carbons 2, 3, 4, 6, or 8 of one residue and the anomeric carbon hydroxyl group of a second sugar. The bond is considered to belong to the sugar whose anomeric carbon is involved, which for hexoses is carbon 1, and Chair Chair Boat is expressed as 1-2, 1-3, 1-4, or 1-6 depending on the point of attachment. Sialic acid, whose anomeric Fig. 8. Chair and boat conformations of pyranose rings. There are two variants of the chair conformation, which differ mainly in the carbon is carbon 2, forms 2-3, 2-6, or 2-8 bonds. axial versus equatorial orientations of the ring functional groups (X and Y). In the boat conformation, the functional groups Two types of are possible; this labeled A are in an axial position, while those labeled E are depends on the configuration of the anomeric carbon equatorial. hydroxyl group involved in the bond. The bond functional groups attached to the ring carbons are able to adopt either an axial position—perpendicular to the general plane of the ring—or an equatorial position— in line with the general plane of the ring. In this way a single saccharide can potentially assume three different three -dimensional shapes, although in practice the lowest energy conformation is preferred, i.e., the one which minimizes the proximity of large functional groups. The functional groups in the equatorial orientation are less crowded together than those in the axial position, and so the lowest energy conformation Fig. 10. Example of α and β linkages in the A type 1 molecule. This dia - occurs when as many as possible of the large hydroxyl gram shows A type 1 in two schematic forms; one structural and the other alphanumeric—both a simple form (lower right) and groups, rather than the relatively small hydrogens, are an expanded form associated with the structural diagram to iden - situated in the equatorial orientation. Only saccharides tify the appropriate sugars and bonds. R = the remainder of the molecule. with alternating chirality, such as β-D-glucose (Fig. 1), can position all the ring hydroxyl groups in the equatorial position, because crowding in the axial formed when the anomeric hydroxyl group is in the α orientation is reduced due to the ring hydrogens being configuration is designated an α bond. Likewise, a β able to alternate above and below the plane of the ring bond will result when the anomeric hydroxyl is in the (Fig. 9). β configuration. The A type 1 structure shown in Figure 10 has two bonds and three bonds. Fig. 9. β-D-glucopyranose in its α β preferred chair conforma- Gangliosides contain sialic acid residues. As tion. All the hydroxyl groups are in the equatori - discussed above, the anomeric carbon of sialic acid is al orientation. Note that carbon 2 (Figs. 3 and 6), and thus the glycosidic bond the hydrogens alternate occurs at the hydroxyl group of this carbon. Sialic acid above and below the plane of the ring in the can be added to a hexose residue at carbons 3 or 6; axial positions. α2-3 and α2-6 bonds between sialic acid and galactose are shown in Figure 11. Sialic acid usually occupies a

IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 37 L.G. G ILLIVER AND S.M. H ENRY

(R. Oriol, personal communi- cation). The actual structure and size of the poly- saccharide is determined by interactions of various poly- morphic blood group glyco- syltransferases and other monomorphic glycosyltrans- ferases. Glycosyltransferases Sialic acid ( α2-3) galactose Sialic acid ( α2-6) galactose Sialic acid ( α2-8) sialic acid are gene–encoded proteins Fig. 11. Sialic acid linkages. Left: N -acetyl -α-D-neuraminic acid (left) bonded to β-D-galactose (right) through (enzymes), and are respon- an ( α2-3) linkage. Center: N -acetyl -α-D-neuraminic acid (left) bonded to β-D-galactose (right) through an sible for the stepwise (α2-6) linkage. Right: two N -acetyl -α-D-neuraminic acid residues joined together through an ( α2-8) link - age. R = the remainder of the molecule. addition of substrate residues to acceptor structures. It is terminal (and sometimes also a sub-terminal) position. therefore clear that the biosynthesis of carbohydrate A terminal sialic acid residue prevents further blood group antigens is not a straightforward gene elongation of the chain except for the addition of equals antigen process. Instead, a gene must encode a another sialic acid residue through an α2-8 linkage functional protein (), which must be retained in (Fig. 11). the Golgi apparatus, and in the presence of appropriate Different glycoconjugate precursor chain precursors and substrates a carbohydrate blood group structures have been identified and categorized into six antigen can be synthesized (Fig. 12). Even this is overly groups based on the identity of the terminal sugar residues and the type of glycosidic Conceptual pathway Biosynthesis of A antigen genetic message (DNA, mRNA) A gene linkage that joins them together. 13 These terminal ↓↓ disaccharide structures of the six groups are shown in enzyme (glycosyltransferase) N-acetyl- α-D-galactosaminyl Table 1. transferase ↓↓ precursor + activated sugar H antigen + UDP-GalNAc Table 1. Six terminal disaccharide chain types. ↓↓ Terminal Disaccharide Structure carbohydrate antigen A antigen I I Type 1 Gal β1-3GlcNAc β1-R membrane/antigen presentation Type 2 Gal β1-4GlcNAc β1-R ↑ ↑ recognition by antibodies/lectins/receptors Type 3 Gal β1-3GalNAc α1-R Type 4 Gal 1-3GalNAc 1-R Fig. 12. Conceptual scheme for the formation of carbohydrate blood β β group antigens. Type 5 Gal β1-3Gal β1-R Type 6 Gal β1-4Glc β1-R simplistic, because the formation of a carbohydrate blood group antigen requires the interaction of several glycosyltransferases unrelated to blood groups before Biosynthesis of Saccharide Chains According to carbohydrate terminology, an the blood group determinant(s) can be added. consists of 3 to 10 sugar residues and a Furthermore, competition and sometimes cooperation contains more than 10 residues. 11 The occur between the various glycosyltransferases, some 14 blood group structures can exist in both linear and of which show redundancy and degeneration. branched forms and range from a few saccharide units Redundancy is said to occur when the same up to more than 60 sugar residues. Irrespective of the carbohydrate structure can be made by two separate size of any given polysaccharide, it is generally enzymes. For example, Se is capable of transferring a accepted that the actual surface recognized by fucose to the terminal β-D-galactose of both type 1 and antibody (glycotope) is usually no larger than 2–3 2 chains, while H can only utilize type 2 chains. Thus, sugars. However, sometimes glycotopes may look H type 2 can be made by two genetically different bigger (up to six sugars) because some extra sugars are enzymes. The presence of Se-formed H type 2 antigens needed to stabilize the three-dimensional configuration on Bombay RBCs is the basis of some para-Bombay of the sugars actually recognized by the antibody phenotypes. Degeneration refers to the situation

38 IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 Review: carbohydrate blood group antigens where one enzyme can make two different structures. Using the same example, Se shows degeneration in that it can make type 1 and type 2 H structures. Thus, it can be seen that the occurrence of specific glyco- conjugates is indirectly controlled by genes through the occurrence and subsequent synthetic actions of enzymes. After glycoconjugates are synthesized, they are transported to various destinations in and out of the cell, as well as being incorporated in the cell membrane where they can be detected by various antibodies. A more comprehensive review of the pathways leading Fig. 13. Structure of a sphingolipid (ceramide) and a monoglycosyl sphigolipid. Top: the sphingolipid is composed of a FA (nervonic to the expression of the blood group–related glyco- acid C24:1 15 —upper chain) and a LCB (4 -sphingenine d18:1 4— 13 sphingolipids can be found elsewhere. lower chain). Below: β-D-glucose is joined to the sphingolipid via a β-linkage between carbon 1 of the saccharide and ceramide. Glycolipids vs. Glycoproteins level of saturation (presence of double bonds), and the Glycans can be conjugated to protein or lipid number of hydroxyl groups—in fact, there are over 60 molecules in biological systems; such molecules are different variations of LCBs, 17,18 each of which can be called glycoproteins or glycolipids, respectively. The combined with one of 20 common variants of FAs. 10 It glycosylation pathways followed by glycoproteins and is speculated that some of the variants may be glycolipids are not always identical. This possibly important for presentation of the antigen. reflects the different biological functions of membrane–bound glycoconjugates (glycolipids and Glycoproteins glycoproteins) versus soluble structures (usually Glycans can be attached to proteins via nitrogen or 15,16 glycoproteins). oxygen—N– or O-glycosidically linked, respectively. N-glycosidically linked glycoproteins are found Glycolipids primarily in serum and cell membranes, while O- Glycosphingolipids are a class of glycolipids which glycosidically linked glycoproteins are located in are composed of a carbohydrate chain and a exocrine gland secretions and mucins, as well as in cell sphingolipid tail, the trivial name of which is ceramide. membranes, although this is less frequent. 7 Sphingolipids are common components of cell membrane lipid bilayers and are composed of a long- O-linked glycoproteins chain base (LCB) amino linked to a fatty acid (FA). Due An O -linkage occurs between the hydroxyl groups to the long hydrophobic hydrocarbon chains present of serine or threonine residues and the in LCBs and the FAs, ceramides are able to insert into anomeric carbon hydroxyl group of N -acetyl - - the lipid bilayer of RBCs. The hydroxyl group on α D-galactosamine, with the linkage type being α1 carbon 1 of the sphingolipid is oxidized to form (Fig. 14). Glycolipids are also O-linked, but the sugar the β-linkage with the anomeric carbon of β-D- Serine Threonine Asparagine glucose to form a mono- glycosyl sphingolipid (Fig. 13). This linkage is desig- nated β1-1 because the two hydroxyl groups involved belong to carbon 1 of their respective structures. There is considerable heterogeneity in ceramides, Fig. 14. Structure of glycoprotein O - and N -linkages. Left and center: O-linked glycoproteins where N -acetyl -D- mainly due to variations in galactosamine is linked through an α bond to the oxygen of the amino acids serine or threonine. Right: N-linked glycoprotein where N -acetyl -D-glucosamine is linked through a β bond to the nitrogen of hydrocarbon chain length, the amino acid asparagine. R 1 and R 2 represent the remainder of the polypeptide chain.

IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 39 L.G. G ILLIVER AND S.M. H ENRY

involved is β-D-glucose. Unlike the N -linked glyco- carbohydrate chemistry in order to develop an proteins, there is no single common oligosaccharide understanding of the complexities that exist within core structure for O -linked glycoproteins. Instead carbohydrate blood group systems. Traditionally, the there are six distinct di- and core antigens of these systems have been viewed structure types and a variety of terminal structures. 7 generically, without appreciation of the subtle O-linked glycoproteins range from the very simple differences between them or the impact this may have mono -glycosylated structures to highly complex on phenotyping or biological function. Not only is molecules containing large amounts of galactose and there a variety of glycotopes within an antigen N-acetylglucosamine, along with lesser amounts of designation, but there also is extensive variation in the N-acetylgalactosamine, fucose, and sialic acid. O -linked carbohydrate molecules which carry the glycotopes glycoproteins do not contain mannose residues. (Fig. 15). This range of variants can be further expanded by N-linked glycoproteins the polymorphic proteins and lipids which then An N-linkage occurs between the nitrogen of the support them. The presentation of the carbohydrate amino acid asparagine and the hydroxyl group of chain at the membrane surface may differ significantly N-acetyl - -D-glucosamine, with the linkage type being among the different types, as shown by molecular β 21 β1 (Fig. 14). modeling. The membrane orientation of saccharide N-linked glycoproteins have a common chains may vary from almost perpendicular (type 1), to 7 pentasaccharide structure consisting of Man 3GlcNAc 2. nearly parallel (types 2, 3, and 4). These different There are three main groups into which N -linked conformations of the carbohydrate chains are glycoproteins can be divided: high mannose type, important for antibody binding and for the complex type, and hybrid type. A comprehensive immunogenic and microbial-interaction characteristics. description of these types can be found elsewhere. 7 Some may actually determine phenotype characteristics, although this has not yet been determined. In addition, the carrier molecule may Red Cell Glycosylation and Serology influence the positioning of the carbohydrate chain on The outside of the human RBC is highly glycosylated, incorporating A type 1; A-6-1 A type 2; A-6-2 both glycoproteins and glycolipids. The major function of the glycoconjugate coat of the RBC is probably to make them more b y ALe ; A-7-2 ALe ; A-7-2 hydrophilic and to provide a “hard” but flexible protective coating. However, if this was the only function of red cell glycosylation, A type 4; Globo-A; A-7-4 A type 2; A-8-2 then it could be achieved by the expression of large amounts of simple carbohydrate chains, which does not account for the complex A type 3; Repetitive A; A-9-3 A-12-2 pattern of glycosylation that actually exists. Some of the glycoconjugates are very complex and bear genetically defined polymorphisms, for example ABH, Lewis, and P. Why A-14-2 these polymorphisms exist is unknown, but there is emerging evidence of evolutionary and biological function. 15,16,19,20 The intention of this review is to Fig. 15. Structure of nine examples of A glycolipid structures in the group A erythrocyte membrane. present basic concepts of Shorthand nomenclature according to Holgersson. 23

40 IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 Review: carbohydrate blood group antigens the cell membrane in relation to other membrane 9. Traving C, Schauer R. Structure, function and components. 22 of sialic acids. Cell Mol Life Sci 1998; From the perspective of blood transfusion, the 54:1330-49. current lack of depth of understanding of carbohydrate 10. Voet D, Voet JG. Biochemistry. New York: Wiley and blood group antigens has been accepted as adequate. Sons, 1998. However, as can be seen from other blood group 11. Atkins RC, Carey FA. Organic chemistry: a brief systems (e.g., Rh and HLA), there is value to be found in course. New York: McGraw-Hill, 1990. unraveling their intricacies. It is hoped that by 12. Pazur JH. Neutral . In: Chaplin MF, investigating the subtleties of the carbohydrate blood Kennedy JF, eds. Carbohydrate analysis: a practical group systems, important new biological information, approach. 2nd ed. Oxford: IRL Press, 1994:100. in particular microbial interactions, will be discovered. 13. Oriol R. ABO, Hh, Lewis, and secretion: serology, genetics, and tissue distribution. In: Cartron J, Rouger P, eds. Blood cell biochemistry: molecular References 1. Laine RA. Invited commentary. Glycobiology basis of human blood cell antigens. New York: 1994;4:759-67. Plenum Press, 1995:37-73. 2. Oriol R, Samuelsson BE, Messeter L. Workshop 1. 14. Oriol R, Le Pendu J, Mollicone R. Genetics of ABO, ABO antibodies. Serological behaviour and H, Lewis, X and related antigens. Vox Sang immuno-chemical characterization. J Immuno- 1986;51:161-71. genet 1990;17:279-99. 15. Karlsson KA. Meaning and therapeutic potential of 3. Samuelsson BE, Pimlott W, Karlsson KA. Mass microbial recognition of host glycosphingolipids. spectrometry of mixtures of intact glycosphingo- Mol Microbiol 1998;29:1-11. lipids. Methods Enzymol 1990; 193:623-46. 16. Karlsson KA. The human gastric colonizer 4. Karlsson H, Johansson L, Miller-Podraza H, Karlsson Helicobacter pylori: a challenge for host -parasite KA. Fingerprinting of large oligosaccharides linked glycobiology. Glycobiology 2000;10:761-71. to ceramide by matrix-assisted laser desorption- 17. Karlsson KA. On the chemistry and occurrence of ionization time-of-flight mass spectrometry. Highly sphingolipid long-chain bases. Chem Phys Lipids heterogeneous polyglycosylceramides of human 1970;5:6-43. erythrocytes with receptor activity for 18. Leray C. Cyberlipid Center: Your www site for fats Helicobacter pylori. Glycobiology 1999;9:765-78. and oils. Available at: http://www.cyberlipid.org/ 5. Oriol R, Barthod F, Bergemer AM, Ye Y, Koren E, amines/amin0001.htm#1. Accessed September 26, Cooper DK. Monomorphic and polymorphic 2001. carbohydrate antigens on pig tissues: implications 19. Henry SM. Review: phenotyping for Lewis and for organ xenotransplantation in the pig-to-human secretor histo-blood group antigens. model. Transpl Int 1994;7:405-13. Immunohematology 1996;12:51-61. 6. Henry SM, Samuelsson BE. ABO polymorphisms 20. Moulds JM, Moulds JJ. Blood group associations and their putative biological relationships with with parasites, bacteria, and viruses. Transfus Med disease. In: King M-J, ed. Human blood cells: Rev 2000;14:302-11. consequences of genetic polymorphisms and 21. Nyholm P-G, Samuelsson BE, Breimer M, Pascher I. variations. Imperial College Press, 2000:15-103. Conformational analysis of blood group A-active 7. Schenkel-Brunner H. Human blood groups: glycosphingolipids using HSEA-calculations. The chemical and biochemical basis of antigen possible significance of the core oligosaccharide specificity. Revised Edition. New York: Springer chain for the presentation and recognition of the A- Wien, 2000. determinant. J Mol Recognit 1989;2:103-13. 8. Reuter G, Gabius HJ. Sialic acids structure-analysis- metabolism-occurrence-recognition. Biol Chem Hoppe Seyler 1996;377:325-42.

IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003 41 L.G. G ILLIVER AND S.M. H ENRY

22. Hakomori S, Handa K, Iwabuchi K, Yamamura S, Lissa G. Gilliver, BSc, BAppSc(Hons), Project Leader, Prinetti A. New insights in glycosphingolipid Kiwi Ingenuity Ltd., Auckland, New Zealand, and function: “glycosignalling domain,” a cell surface PhD student, Auckland University of Technology, New assembly of glycosphingolipids with signal Zealand. Stephen M. Henry, PhD, Director, Kiwi transducer molecules, involved in cell adhesion Ingenuity Ltd., 19 Mount St., Auckland, New Zealand, coupled with signalling. Glycobiology 1998;8:11-9. and Director, Glycoscience Research Centre, Auckland 23. Holgersson J, Breimer ME, Samuelsson BE. Basic University of Technology, Private Bag 92006, biochemistry of cell surface carbohydrates and Auckland 1020, New Zealand, email: [email protected] aspects of the tissue distribution of histo-blood (corresponding author). group ABH and related glycosphingolipids. APMIS Suppl 1992;27:18-27.

Manuscripts : The editorial staff of Immunohematology welcomes manuscripts pertaining to blood group serology and education for consideration for publication. We are especially interested in case reports, papers on platelet and white cell serology, scientific articles covering original investigations, and papers on the use of computers in the blood bank. Deadlines for receipt of manuscripts for the March, June, September, and December issues are the first weeks in November, February, May, and August, respectively. Instructions for scientific articles and case reports can be obtained by phoning or faxing a request to Mary H. McGinnniss, Managing Editor, Immunohematology, at (301) 299-7443, or see “Instructions for Authors” in every issue of Immunohematology or on the Web. Include fax and phone numbers and e-mail address with your manuscript.

REMEMBER: THE PASSWORD IS “2000” For www.redcross.org/pubs/immuno Now, as a subscriber, you can enter the password, 2000, to access complete journal contents for the past 5 years. That means cover to cover! You will receive every article in total, every letter to the editor, every review, every ad, every notice, and every literature review! All of the other services will continue to be available on the Web page, including sending a letter to the editor, subscribing with a credit card on the secure order site, performing a literature search, reviewing instructions for authors, and linking to other important sites. Log on now to see this great service!

Free Classified Ads and Announcements Immunohematology will publish classified ads and announcements (SBB schools, meetings, symposia, etc.) without charge. Deadlines for receipt of these items are as follows:

Deadlines 1st week in January for the March issue 1st week in April for the June issue 1st week in July for the September issue 1st week in October for the December issue

Fax these items to Mary H. McGinniss, Managing Editor, (301) 299-7443.

42 IMMUNOHEMATOLOGY, VOLUME 19, NUMBER 2, 2003