Biology Reference
In-Depth Information
are very different. The sugar directly bound to asparagine is N-acetylglucosamine. O-linked
sugars tend to be shorter (mono-, di- or tri-saccharides) and more diverse than are the
N-linked sugars that contain much longer polysaccharides. For most membrane glycopro-
teins, N-linked sugars are more common on dynamic enzymatic proteins while O-linked
sugars are predominant on structural proteins. Since a large protein will normally have
numerous possible amino acids that could theoretically have sugar attachments, there are
a number of different locations where sugars will be found. In addition there may be a variety
of different sugars that can be attached at a single location. This complexity leads to what is
known as
microheterogeneity
. For example, ovalbumin has only one site of glycosylation, yet
more than a dozen different attached sugars have been identified at this site.
In the late 1990s, computer-based sequencing searches were applied to membrane glyco-
sylation by Nathan Sharon and colleagues
[19,20]
. In one study of known membrane
proteins, the question was addressed of how common glycoproteins are and how the sugars
are attached to the proteins. Complete sequences of 1823 integral membrane proteins with
extracellular features were identified. Of these, 1676 glycoproteins (92% of the total integral
proteins) were identified, and of these 1630 or 97.3%were N-glycosylated. This leaves only 46
(2.8%) possible O-glycosylated proteins. Not glycosylated were 116 multiple span proteins,
leaving only 14 possible single span proteins, at most, that were not glycosylated. From
this study, two major conclusions can be drawn. The vast majority of plasma membrane inte-
gral proteins are glycosylated and, of these, there are many times more N-glycosylations than
O-glycosylations.
D. GPI-ANCHORED PROTEINS
A highly unusual family of membrane integral proteins is GPI (glycosylphosphatidylino-
sitol)-anchored (Chapter 6, and
[21,22]
). These proteins are all anchored to the membrane
outer leaflet and have no trans-membrane segment. Therefore GPI-anchored proteins are
also ecto-proteins. Membrane anchorage is primarily through the two acyl chains of PI
and so binding is much weaker than trans-membrane proteins. GPI-anchored proteins
are linked at their C-terminus to the membrane PI via a phosphoethanolamine to a tetra-
saccharide composed of three mannoses and one glucosamine. The structure is shown in
Figure 6.13.
GPI-anchored proteins are found in most organisms, but not bacteria. Although widely
distributed, GPI-anchored proteins are not abundant. In yeast, where all GPI proteins are
known, only 1 in 400 proteins have GPI-anchors and even these are very similar to one
another. In man, the 45 known GPI-anchored proteins are most abundant in neurons where
they appear to be concentrated in lipid rafts (see Chapter 8). Here they have functions in
receptor-mediated cell signaling pathways. Being outer surface proteins, they also function
in cell-adhesion and as cell surface antigens. GPI-anchored proteins can be released from
the cell surface by phospholipase C which also produces diacylglycerol and by phospholi-
pase D which generates PA. Diacylglycerol and PA are both highly reactive, membrane-
bound compounds (see Chapter 5).
Several human genetic disorders have been linked to defects in GPI-anchored proteins.
Although the most common of these is paroxysmal nocturnal hemoglobinuria (UPNH),
