308
chapter
16
Carbohydrate Metabolism
111
: Glycoproteins, Glycolipids, GPI Anchors, Proteoglycans, and Peptidoglycans
donor sugar and a particular acceptor molecule. This speci-
ficity has led to the hypothesis of a unique glycosyltrans-
ferase for every linkage between two monosaccharides in
an oligosaccharide. Although not strictly true, the con-
cept of “one linkage-one enzyme” is useful in understand-
ing glycoconjugate synthesis. Many glycosyltransferases
are highly specific, using only one particular nucleotide
sugar as donor and requiring as acceptor an oligosac-
charide with the correct terminal and penultimate sugars
in the proper linkage. Other enzymes are less specific.
For example, /3-1,4-galactosyltransferase attaches galac-
tose to different acceptors, provided the terminal sugar is
N-acetylglucosamine.
Synthesis of glycoconjugates, in contrast to that of
DNA, RNA, and protein, is not directed by a template
but depends on the specificity of the enzymes that cat-
alyze addition of each sugar residue. Thus, oligosac-
charide chain biosynthesis may be less predictable than
DNA, RNA, or protein synthesis and more often results
in incomplete oligosaccharide chains. Furthermore, be-
cause there is no template from which to deduce the
oligosaccharide size or sequence in a given glycocon-
jugate, the synthetic reactions can best be described by
sequencing the product. The oligosaccharide reflects its
biosynthetic history, and understanding of the biosynthetic
pathways has hinged on structural analysis of oligosaccha-
rides.
The recent advances in the determination of oligosac-
charide structure using high-performance liquid chroma-
tography (HPLC), gas chromatography-mass spectrom-
etry (GC-MS), and nuclear magnetic resonance (NMR)
have increased the capacity to separate and sequence sug-
ars and their linkages in oligosaccharides. Analytical pro-
cedures must overcome several unique sequencing prob-
lems that include:
1. Lack of unique characteristics (such as catalytic
activity) that can be used to follow purification;
2. Presence of more than one type of oligosaccharide
side chain in a particular glycoprotein;
3. Presence of more than one type of glycosidic bond
within an oligosaccharide, in contrast to the peptide
bond that universally links amino acids to
proteins;
4. Chemical similarity of one sugar to another, which
leads to poor separation in a number of analytical
systems and makes identification of a particular sugar
more difficult. Except for the purification steps,
determination of oligosaccharide structure is now
relatively routine.
Carbohydrate
structures,
whether
attached
to
N-
glycans,
O-glycans,
or glycosphingolipids,
generally
have a core structure with arms containing terminal sugar
sequences (Figure 16-2). In some cases the arms may
consist of repeating Gal and GlcNAc referred to as poly-
lactosamine. Figure 16-1 shows the structures of several
oligosaccharides found in different types of glycoconju-
gates, together with the abbreviations used in writing these
structures (see also Chapters 10 and 11).
The high degree of specificity of the glycosyltrans-
ferases necessitates a detailed scheme for their nomen-
clature. Criteria used to classify glycosyltransferases are
the nucleotide-sugar, the transferred sugar, the monosac-
charide at the acceptor site, and the linkage. In addition,
transferases can show specificity for the location of the
acceptor site, i.e., the protein or lipid to which the accep-
tor is attached or the exact location of the acceptor on the
macromolecule.
This specificity can be indicated in one of two ways:
1. by specifying the sugar acceptor, linkage, and sugar
donor (e.g., GlcNAc-//1,4-galactosyltransferase) or
2. when there are a number of glycosyltransferases with
similar acceptor and linkages, a Roman numeral is
placed after the name derived from the above criteria
(e.g., a family of UDP-N-acetylglucosaminyltrans-
ferases are numbered I-VI based on the particular
acceptor sites to which they transfer GlcNAc
(Figure 16-3).
Uridine diphosphate (UDP) is the most common nu-
cleotide carrier for sugars, although cytidine monophos-
phate (CMP)) and guanosine diphosphate (GDP) are also
used. The most abundant nucleotide-sugar (and the first
discovered) is UDP-glucose. Its structure and a list of
other nucleotide-sugar types are given in Figure 16-4. See
Chapter 15 for a discussion of the synthesis of UDP-
glucose and its role in glycogen synthesis. Figure 16-5
summarizes biosynthetic pathways for UDP-glucose and
other nucleotide-sugars important for glycoconjugate syn-
thesis. The sugar donors can be synthesized from glu-
cose, provided that the requisite enzymes are present,
and the interconversion of fucose to other sugars does
not occur. Similarly, N-acetylmannosamine
in vivo
seems
to be entirely metabolized to N-acetylneuraminic acid
(sialic acid), despite the existence of alternative reaction
pathways, as shown in Figure 16-5). As a result, radioac-
tively labeled fucose and N-acetylmannosamine, when ad-
ministered experimentally in animals, can mark specific
glycosylation sites within cells.
Several sugar-nucleotide pool defects have been iden-
tified:
Type I carbohydrate-deficient glycoprotein syn-
drome
(CDGs type I), a defect in phosphomannomu-
tase, one of the enzymes responsible for converting
glucose to GDP-mannose, results in an absence of
