 |
|
|
Research Background and Significance
Glycosylation of proteins represents
the most complex post-translational modification of proteins known to occur.
Carbohydrates are attached to proteins in two broad groups: linked to the
nitrogen of Asn in the consensus sequence -NXT- (N-linked), and linked
to the oxygen of Ser or Thr in Ser/Thr-rich regions (O-linked). Knowledge
is being accumulated on the structure and metabolism of O-linked glycoproteins,
but N-linked glycoproteins are much better known. The following discussion
reflects this bias of knowledge, relating predominantly to the N-linked
glycoproteins.
By the use of seven different
monosaccharides in two anomeric configurations, and several linkage sites
per monosaccharide, a plethora of oligosaccharide structures can potentially
be created. The range of structures which is observed to occur in
nature is considerably smaller than theoretically possible, and these structures
are highly conserved among species (1). However, there are often
a number of slightly different carbohydrate structures observed on a single
protein, most often differing in sialic acid content or number of branches,
a phenomenon referred to as microheterogeneity. This pattern of glycoforms
is thought to be protein-specific, site-specific and tissue- or cell-specific
(2). More recently the concept has expanded to include time-specific
(3).
Alterations in oligosaccharide
structure have been reported to occur in infertility (4) and in disease
states (5-9). In addition, immature T cells express the hexasaccharide
NeuNAca2-3Galßl-3(NeuNAca2-3Galßl-4GlcNAcßl-6)GalNAc
(10). Resting mature T cells lose expression of this hexasaccharide,
but activated T cells express the motif. These data strongly suggest
that changes in oligosaccharide structure must have biological ramifications
on protein structure, function, or turnover. The changes which occur
in disease states are poorly understood, both in terms of mechanism and
reason for the change, and in terms of the ramifications of the change.
A number of constituent
elements of oligosaccharide chains have been shown to influence biological
activity and protein half-life. Lysosomal proteins express mannose-6-phosphate
and are directed to lysosomes via the Man-6-P receptor, a C-type lectin
in bovine and murine species (11). The asialoglycoprotein receptor
(ASGPR), present on liver Kupffer cells, binds glycoproteins with exposed
galactose residues and removes them from the circulation (12). Glycoprotein
hormones expressing the carbohydrate motif SO4-GalNAc-GlcNAc-Man
are rapidly cleared from the circulation by a separate Kupffer cell receptor
(13), and this clearance is thought to be necessary for pulsatility of
lutropin levels during the menstrual cycle (3). It is also known
that Kupffer cells contain a receptor which binds fucosylated proteins
(14), although this is perhaps a less well-defined process. The rat
Kupffer cell receptor for fucose was originally isolated as a result of
studies examining the turnover of fucolipids (15) and fucoproteins (16).
The first observation
that fucose may play a role in protein turnover was that lactoferrin, a
fucose-containing protein, was rapidly cleared from blood into the liver
(16). This clearance was not altered by glycoproteins which are cleared
via Gal or Man/GlcNAc residues (17,18). Work performed in Hill's
laboratory (14,19-22) went on to show that the receptor has a molecular
weight of 88,000, and a 77 kDa proteolytic degradation product can be observed
which is the active lectin (21). The receptor is a C-type lectin,
requiring Ca2+ for binding, and has a pH optimum between 7.6
and 8.6 (20). The receptor is responsible for the endocytosis of
BSA modified to contain covalently bound fucose. At least two other
reports have indicated fucose affects protein turnover (23,24). Tollefson
showed that the percent of protein-bound 125I decreased rapidly
when radiolabeled a-amylase was injected intravenously
into rats. In contrast to early observations, however, nonfucosylated
a-amylase was degraded at a faster rate (t1/2
= 9.5 min) than fucosylated species (t1/2 = 12 min)(23).
The reason for this difference is unknown. More recently, Hajjar
and Reynolds have shown that degradation of tPA by Hep G2 cells is decreased
by EGF and fucose (24), although this finding has generated some controversy
(25), see below. t-PA is one of several proteins (26-29) which contains
O-linked fucose in an EGF domain. Therefore, the presence of terminal
fucose may be a general mechanism for influencing the degradation of proteins.
Fucose is a monosaccharide
normally present in the serum and is the only levorotatory sugar synthesized
and utilized by mammalian systems (30). The metabolism of fucose
is only partially understood. It is known that fucose can be synthesized
by mammals (31), but how much is synthesized and by what tissues has not
been fully defined. Yorek et al. have shown that feeding rats a diet
containing fucose causes the serum unbound fucose level to increase (32,33),
implying intestinal absorption of fucose, but intestinal absorption of
fucose has not been shown directly. As a post-doctoral research associate
in the Yorek laboratory, I demonstrated that fucose could be accumulated
by cultured cells via a specific transport protein (34). Whether
or not fucose can be degraded by mammals is debatable. Segal and
Topper (35) reported that when [14C]fucose is administered to
man intravenously, 34% was excreted as 14CO2 after
6 h. On the other hand, Bocci and Winzler found that 14CO2
was produced when [14C]fucose was administered to rats orally
but not intravenously, and therefore suggested that intestinal bacteria
were responsible for the oxidation (36). To my knowledge no one else
has examined these observations to determine what differences may exist
in the two models. In my hands, a large amount of the fucose fed
to rats is excreted in the urine in a concentration-dependent fashion (Wiese
and Yorek, unpublished observations), although the amount of free and bound
fucose increases in the serum of fucose-fed rats (32,33).
Most of the fucose
accumulated by mammalian cells in culture is incorporated into protein,
the majority of which are secreted. Only 2% is incorporated into
lipid, and little remains unbound (34). In known fucose-containing
structures, fucose is found incorporated into protein in an a1-6
linkage to the proximal GlcNAc of the chitobiose core, a1-2
on terminal galactose residues and in an a1-3
or a1-4 linkage to the penultimate GlcNAc residue
(1). An unusual difucosylated biantenary oligosaccharide has recently
been isolated from the venom of honeybees (37). In this case, both
fucose residues are attached to the proximal GlcNAc, one in an al-3-
and the second in an a1-6-configuration.
O-glycosidic linkages for Fuc also exist. As described above, Kentzler
has found fucose O-linked to threonine in the EGF domain of pro-urokinase
(26). Subsequent reports have demonstrated fucose O-linked to serine
or threonine of tPA (27), human and bovine factor VII (38), factor IX (28),
factor XII (39), and vampire bat salivary plasminogen activator (29).
Whereas fucose is usually thought of as a terminal sugar residue, one recent
report states that a tetrasaccharide containing Gal, Fuc, GlcNAc and NeuAc
is bound to serine 61 of human factor IX, with fucose being the reducing
sugar (28). As Ser-61 of factor IX is the residue normally decorated
with a single fucose residue (29), the tetrasaccharide may represent a
minor microheterogeneity species. A second report (40) has also shown
that an oligosaccharide is produced at very low level, linked through fucose.
However, in this report it is shown that the majority of O-linked fucose
is linked as a monosaccharide.
A large number of
studies have reported that glycoprotein metabolism is altered in cancer
and other diseases. Hakomori makes the point that in pathological
conditions such as cancer, changes in carbohydrate structure of glycoproteins
can almost always be observed (41). Some of these changes involve
changes in fucose-containing structures. Hakomori has shown changes
in fucolipid expression of dimeric Lewisa (Lea) in Colo 205
tumor cells grown in nude mice (42). Lewisa has the structure NeuNAca2-3Galßl-3(Fucal-4)GlcNAc-R.
Tumor cells are enriched with the carbohydrate epitope sialyl Lewis x (sLex)
(10), which has the structure NeuNAca2-3Galßl4(Fucal-4)GlcNAc-R.
Turner has published data indicating an increase in fucosylation of haptoglobin
in ovarian cancer (7), heavy smoking and drinking (43) and rheumatoid arthritis
(9). Although no structural data has ever been reported, Turner also
states that this increase is in fucose linked to subterminal GlcNAc in
an a1-3 linkage (8). It is important that
structural data eventually be obtained in studies of this sort. It
is probable that an increase in fucosylation of the chitobiose core will
have completely different effects than increased fucosylation of the terminal
galactose residues. It is also possible that O-linked fucose will
have different effects on protein turnover than N-linked fucose.
If fucose is demonstrated to influence protein turnover, then understanding
fucose metabolism in greater depth will allow manipulation of the fucosylation
index of proteins, thereby manipulating protein half-lives.
One of the complexities
of glycobiology is the number of glycoproteins produced by a given cell.
It is commonly held, for example, that the majority of secreted proteins
are glycoproteins. In addition to the possibly differing effects
of O- vs N-glycosylation, and a1,6 fucosylation
as opposed to other types, it is possible but not as likely that different
proteins would display different effects upon altered glycosylation.
For these reasons, I have chosen transferrin as an initial choice of protein
to study for the following reasons. [1] Transferrin has been extensively
studied due to its important role in iron transport, leading to a well-defined
pattern of turnover. [2] Numerous manuscripts have been produced examining
the structure of transferrin oligosaccharides, with varying degrees of
rigor. The consensus however, is that the glycoprotein structure
of normal transferrin consists of two identical complex biantennary structures.
[3] Transferrin has been examined as a possible marker of carbohydrate-deficient
glycoprotein syndrome and other diseases (44)
Although an altered
fucose metabolism has been demonstrated in several disease states, the
biological ramifications of the change(s) has not been adequately defined.
Because of the impact oligosaccharide chains have on protein structure,
function, and turnover; and the observation that fucose specifically can
affect turnover, it is worthwhile examining proteins for the effect of
altered fucose index, so that predictions can be made as to what biological
effects will occur if glycoprotein metabolism changes in disease.
Furthermore, it would be desirable to examine individual changes, such
as altered a1-6 fucosylation, to determine with
precision the effects of changes.
Besides its roles
in protein turnover and antigenicity, the fucose moiety has demonstrated
importance in three other areas. First, Helicobacter pylori,
the causative agent for peptic ulcer disease, has been shown to adhere
to stomach wall through fucosylated blood group antigens (45). Second,
several works have shown an involvement of fucosylated proteins in long-term
memory formation. Initial reports indicated a participation of fucosylated
macromolecules in the maintenance of long-term potentiation (46).
Fucose administration enhances this process, and is attached to newly synthesized
proteins of undefined identity (47) in an a1-2
linkage (48). Third, a lectin from Pseudomonas aeruginosa has recently
been shown to stop cillary beating in cystic fibrosis patients (49) and
it was suggested that fucose be added to nebulizers to abrogate this effect.
Given the importance of fucose in these various processes, a better understanding
of fucose metabolism would be beneficial. It seems reasonable to
assume, for example, that some of the fucose added to nebulizers would
be absorbed and incorporated into macromolecules. What effect this
would have on other processes is at this time hard to predict. The
studies outlined in this proposal would increase our ability to predict
the effect(s).
Last Modified September 5, 2005
|
|
|
|
