-
The Chemical Neurobiology of Carbohydrates
Heather E. Murrey and Linda C. Hsieh-Wilson*
Division of Chemistry and Chemical Engineering and Howard Hughes
Medical Institute, California Institute of Technology,Pasadena,
California 91125
Received February 4, 2008
Contents
1. Introduction 17082. Sialic Acids 17092.1. Structure 17092.2.
Neurobiological Functions 17102.2.1. R(2-3)-Sialic Acid and
Myelin-Associated
Glycoprotein1710
2.2.2. Polysialic Acid 17102.3. Chemical Neurobiology of Sialic
Acid 17112.3.1. Synthetic Sialic Acid Derivatives: Probing
the Specificity of MAG Interactions1711
2.3.2. Development of MAG Antagonists withTherapeutic
Potential
1711
2.3.3. Synthetic Mimics of R(2-8)-Linked PSAfor Nerve
Regeneration
1712
2.3.4. Metabolic Labeling To RemodelCell-Surface Sialic Acid
Interactions
1712
3. R-L-Fucose 17143.1. Structure and Biosynthesis 17143.2.
Neurobiological Functions 17143.2.1. Neuronal Development
17143.2.2. Learning and Memory 1715
3.3. Chemical Approaches for Studying L-Fucose 17153.3.1.
Deoxygalactose Analogues 17153.3.2. Glycopolymers and Imaging
Probes 17163.3.3. Metabolic Labeling Using Alkynyl or Azido
Fucose Analogues1716
3.3.4. Summary of Fucosyl Oligosaccharides 17164. O-GlcNAc
Glycosylation 17164.1. Structure and Biological Functions 17164.2.
Neurobiological Functions of O-GlcNAc 17174.3. Chemical Tools To
Study O-GlcNAc
Glycosylation1717
4.3.1. Chemoenzymatic Labeling of O-GlcNAcProteins
1717
4.3.2. Metabolic Labeling of O-GlcNAc Proteins 17194.3.3.
Methods for Mapping Exact Glycosylation
Sites1719
4.3.4. Monitoring O-GlcNAc Dynamics 17205. Glycosaminoglycans
17225.1. Structure and Diversity 17225.2. Neurobiological Functions
17235.2.1. Neuronal Development 17235.2.2. Axon Guidance 17235.2.3.
Spinal Cord Regeneration 1723
5.3. Challenges to the Study of GAGs 17235.4. Synthetic
Molecules for Probing
Structure-Activity Relationships1723
5.4.1. Synthesis of Glycosaminoglycans 17235.4.2. Effects of HS
and DS Molecules on
Neuronal Growth1724
5.4.3. Neuroactive Small-Molecule ChondroitinSulfates
1724
5.5. Carbohydrate Microarrays for StudyingGAG-Protein
Interactions
1724
5.5.1. Oligosaccharide Microarrays 17245.5.2. Polysaccharide
Microarrays 1725
5.6. Glycosaminoglycan-Based Therapeutics 17255.6.1. Prion
Diseases 17255.6.2. Alzheimers Disease 17255.6.3. Future Challenges
1726
6. Summary and Future Directions 17267. Acknowledgments 17268.
References 1726
1. IntroductionThe cell surface displays a complex array of
oligosaccha-
rides, glycoproteins, and glycolipids. This diverse mixtureof
glycans contains a wealth of information, modulating awide range of
processes such as cell migration, proliferation,transcriptional
regulation, and differentiation.15 Glycosyla-tion is one of the
most ubiquitous forms of post-translationalmodification, with more
than 50% of the human proteomeestimated to be glycosylated.6
Glycosylation adds anotherdimension to the complexity of cellular
signaling andexpands the ability of a cell to modulate protein
function.The structural complexity of glycan modifications
rangesfrom the addition of a single monosaccharide unit
topolysaccharides containing hundreds of sugars in branchedor
linear arrays.7 This chemical diversity enables glycans toimpart a
vast array of functions, from structural stability andproteolytic
protection to protein recognition and modulationof cell signaling
networks.8,912
Emerging evidence suggests a pivotal role for glycans
inregulating nervous system development and function. Forinstance,
glycosylation influences various neuronal processes,such as neurite
outgrowth and morphology, and maycontribute to the molecular events
that underlie learning andmemory.7,13,14 Glycosylation is an
efficient modulator of cellsignaling and has been implicated in
memory consolidationpathways.1518 Genetic ablation of glycosylation
enzymesoften leads to developmental defects and can
influencevarious organismal behaviors such as stress and
cognition.1924Thus, the complexity of glycan functions help to
orchestrate
* To whom correspondence should be addressed. E-mail:
[email protected].
Chem. Rev. 2008, 108, 170817311708
10.1021/cr078215f CCC: $71.00 2008 American Chemical
SocietyPublished on Web 05/02/2008
-
proper neuronal development during embryogenesis, as wellas
influence behaviors in the adult organism.
The importance of glycosylation is further highlighted bydefects
in glycan structures that often lead to human disease,as exhibited
by congenital disorders of glycosylation(CDG).2529 These are
usually inherited disorders resultingfrom defects in glycan
biosynthesis, which are accompaniedby severe developmental
abnormalities, mental retardation,and difficulties with motor
coordination. Such disordershighlight the importance of glycan
biosynthesis in humanhealth and development. Because therapeutic
treatments arecurrently limited, investigations into the
structure-activityrelationships of glycans, as well as
disease-associatedalterations to glycan structure, are crucial for
developingstrategies to combat these diseases.
Understanding the structure-function relationships ofglycans has
been hampered by a lack of tools and methodsto facilitate their
analysis. In contrast to nucleic acids and
proteins, oligosaccharides often have branched structures,
andtheir biosynthesis is not template-encoded. As such,
thecomposition and sequence of oligosaccharides cannot beeasily
predicted, and genetic manipulations are considerablyless
straightforward. Analytical techniques for
investigatingoligosaccharide composition, sequence, and tertiary
structureare still undergoing development and are far from
routine,unlike methods for DNA and protein analysis. Lastly,
glycanstructures are not under direct genetic control and, thus,
areoftenheterogeneous.Thisheterogeneitycomplicatesstructure-function
analyses by traditional biochemical approaches thatrely on the
isolation and purification of glycans from naturalsources.
The problems associated with oligosaccharide analysishave
hindered efforts to understand the biology of oligosac-charides yet
have given chemists a unique opportunity todevelop new methods to
overcome these challenges. Thedevelopment of chemical tools for the
analysis of glycanstructure and function is essential to advance
our understand-ing of the roles of glycoconjugates in regulating
diversebiological processes. In this review, we will highlight
theemerging area of glyconeurobiology with an emphasis oncurrent
chemical approaches for elucidating the biologicalfunctions of
glycans in the nervous system.
2. Sialic Acids
2.1. StructureSialic acids participate in a multitude of
biologically
interesting phenomena, including cell-cell recognition,adhesion,
and intracellular signaling events.3032 Originallyknown as
neuraminic acid (Neu) and its derivatives, sialicacids are a family
of R-keto acids containing a nine-carbonbackbone.32 The most
well-known members of the sialic acidfamily include
N-acetylneuraminic acid (Neu5Ac), N-gly-colylneuraminic acid
(Neu5Gc), and deaminoneuraminic acid(KDN) (Figure 1). In addition
to these basic forms, morethan 50 distinct sialic acid structures
have been identified innature, arising from acetylation,
methylation, lactylation,sulfation, and phosphorylation of the C-4,
C-5, C-7, C-8, orC-9 hydroxyl groups.
Sialic acids exist predominantly as terminal monosaccha-rides
linked to galactose residues in glycan chains throughR(2-3)- or
R(2-6)-linkages. They can also form a ho-mopolymer of R(2-8)-linked
sialic acid in mammals, termedpolysialic acid (PSA).33,34 As
discussed below, each glyco-form dictates a unique function to the
glycoproteins andglycolipids expressing these sugars. Sialic acids
have histori-cally received much attention due to their
participation incell-cell recognition events and the pathogenesis
of diseasessuch as cancer,3537 inflammatory disease,3840 and
viralinfection.4144 The development of sialic acid analogues
asinhibitors or probes for biomedical research has led to
Heather E. Murrey received a B.A./M.S. degree in biochemistry
fromBrandeis University in 2000. She then conducted research at
theWhitehead Institute for Biomedical Research in the laboratories
of PeterS. Kim and Harvey F. Lodish. She is currently pursuing a
Ph.D. at Caltechunder the direction of Linda C. Hsieh-Wilson. Her
graduate studies havefocused on the role of fucosyl
oligosaccharides in neuronal communicationand development.
Linda C. Hsieh-Wilson received her B.S. degree in chemistry from
YaleUniversity in 1990. She then earned her Ph.D. degree in 1996
from theUniversity of California at Berkeley, where she worked with
Peter G.Schultz. After completing postdoctoral studies in
neurobiology at TheRockefeller University with Nobel Laureate Paul
Greengard in 2000, shejoined the faculty at the California
Institute of Technology, where she iscurrently Associate Professor
of Chemistry. In 2005, she was appointeda Howard Hughes Medical
Institute Investigator. Her research group worksat the interface of
organic chemistry and neurobiology to study the rolesof
carbohydrates and their associated proteins in transcription,
neuronalsignaling, and brain development.
Figure 1. Common structures of sialic acid derivatives:
neuraminicacid (Neu), N-acetylneuraminic acid (Neu5Ac),
N-glycolylneuraminicacid (Neu5Gc), and deaminoneuraminic acid
(KDN).
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1709
-
significant advances in our understanding of this
importantfamily of carbohydrates. Here, we will discuss some of
theroles of sialic acids in neurobiology and chemical
approachesthat have provided insight into their functions.
2.2. Neurobiological Functions2.2.1. R(2-3)-Sialic Acid and
Myelin-AssociatedGlycoprotein
Sialic acid is often expressed as R(2-3)-linked sialic acidin
the nervous system, a carbohydrate motif recognized bythe Siglec
(sialic acid-binding immunoglobulin-like lectin)family of proteins.
Human Siglecs include at least 13members, each containing a common
V-set immunoglobulindomain that interacts with sialic acid.45 One
interaction thathas been extensively studied is the binding of
myelin-associated glycoprotein (MAG; also known as Siglec-4)
withR(2-3)-sialic acid. MAG is a 100-kDa integral
membraneglycoprotein that is expressed myelinating glia cells.46,47
Itis involved in regulating the formation and maintenance
ofmyelin48 and has been suggested to inhibit nerve regenerationin
the adult central nervous system (CNS).4951 Micedeficient in MAG
display delayed myelination,52 defects inthe organization of
periaxonal space,53 and subtle morpho-logical abnormalities of
myelin sheaths.52 The interactionsof MAG with sialic
acid-containing glycosphingolipids,known as gangliosides, have been
extensively studied andhave contributed to our understanding the
role of MAG inmyelin formation and neural regeneration.
MAG preferentially binds the glycan structure
Neu5AcR-(2-3)Gal(1-3)GalNAc,54 which is expressed on cell-surface
gangliosides and O-glycans of glycoproteins.47Gangliosides
represent the major source of sialic acidexpression in the brain.
MAG binds with high affinity andspecificity to the major brain
gangliosides GD1a and GT1b,as well as the polysialoganglioside
GQ1bR, a minor gan-glioside expressed on cholinergic neurons
(Figure 2). Diges-tion of gangliosides purified from bovine brain
with neuramini-dase, an enzyme that cleaves sialic acid residues,
eliminatedthe binding of MAG to these gangliosides,
demonstratingthe importance of the sialic acid moiety in
mediatingMAG-ganglioside interactions.5557
Studies suggest that the association of MAG with
sialicacid-containing gangliosides plays an important
functionalrole in neuronal growth. The ability of MAG to inhibit
neuriteoutgrowth in Vitro is blocked by treatment of
cerebellargranule neurons with neuraminidase or with the
glucosyl-
ceramide synthase inhibitor P4, which prevents synthesis ofall
glycosphingolipids.55 Moreover, mice lacking the
glyco-syltransferase gene GalNAcT
(UDP-N-acetylgalactosamine:GM3/GD3
N-acetylgalactosaminyltransferase) do not expresscomplex
gangliosides such as GD1a and GT1b and, as aconsequence, exhibit
axon degeneration and gross dysmyeli-nation.58,59 These mice also
display progressive behavioralabnormalities consistent with
neurodegenerative disease, suchas defects in balance, reflexes, and
motor coordination.59Thus, detailed knowledge of MAG and its
interactions withsialylated glycans may enhance our understanding
of my-elinating disorders such as multiple sclerosis and
provideopportunities to enhance axon regeneration after CNS
injuryor disease.
2.2.2. Polysialic Acid
In the brain, PSA is expressed primarily on the proteinneural
cell adhesion molecule (NCAM).6062 NCAM playscritical roles in both
nervous system development andmemory formation, regulating
processes such as cell adhe-sion, axon targeting and fasciculation,
neuronal migration,synaptic plasticity, and
synaptogenesis.60,61,6370 PSA-NCAMis highly expressed in the
embryonic brain7173 and is foundin the adult brain in areas that
retain a high degree ofplasticity and neurogenesis, such as the
hippocampus,olfactory bulb, and hypothalamus.7477
Although the molecular mechanisms underlying PSAfunction are not
well understood, PSA is thought to modulatecell-cell adhesion by
attenuating homophilic NCAM-NCAMinteractions. The large steric bulk
and hydration shell of thecarbohydrate chain increase the
intercellular space by 10-15m, reducing trans NCAM-NCAM
interactions acrossapposing cells.78 In addition, PSA modulates the
interactionsof NCAM with other proteins, such as heparan
sulfateproteoglycans involved in the formation and remodeling
ofhippocampal synapses.79 The PSA chains on NCAM havealso been
proposed to play a role in some neuropsychiatricdisorders. For
example, expression of PSA-NCAM issignificantly reduced in the
hippocampus of schizophrenicpatients and may contribute to the
complex symptomsassociated with the disease.8082 Moreover, PSA has
beenimplicated in the etiology of Alzheimers disease,
asPSA-NCAM-positive granule cells are increased in thehippocampus
of Alzheimers patients and are associated withdisorganization of
PSA-positive fibers.83 Finally, PSA mayalso regulate neuronal
function through NCAM-independent
Figure 2. Structures of gangliosides that bind to MAG. Neu5Ac )
N-acetylneuramic acid; Gal ) galactose; GalNAc )
N-acetylgalactosamine;Glc ) glucose; Cer ) ceramide.
1710 Chemical Reviews, 2008, Vol. 108, No. 5 Murrey and
Hsieh-Wilson
-
mechanisms. For example, PSA has been suggested to actas a
competitive antagonist of the NMDA receptor, anionotropic glutamate
channel involved in synaptic transmis-sion,84 thereby preventing
glutamate-induced excitotoxicity.85
Despite intriguing roles for sialic acid-containing glycans,the
molecular mechanisms underlying their diverse functionsin the brain
remain largely unknown. As described below,chemical approaches to
access and manipulate sialic acidstructures have expanded our
understanding of the neuro-biological roles of sialic acid and
promise to continue toadvance the field.
2.3. Chemical Neurobiology of Sialic Acid2.3.1. Synthetic Sialic
Acid Derivatives: Probing theSpecificity of MAG Interactions
Synthetic sialic acid analogues have been used to elucidatethe
molecular determinants important for MAG-gangliosideinteractions.
The C-9 hydroxyl group represents a keyrecognition element:
substitution of this group with hydrogen,halogen, or thiol groups
attenuated the association of MAGwith Neu5Ac (Figure 3, compounds
1-5). Interestingly, anamino group at C-9 enhanced binding to MAG
by 3-fold,suggesting the importance of a hydrogen donor at
thisposition (compound 6).86 The C-5 N-acetyl group of Neu5Acwas
also found to be critical for MAG binding, although itis not always
required for interaction with other Siglecs.Replacement of this
group with an N-propanoyl, N-ami-noacetyl, or N-thioacetyl moiety
enhanced binding of sialicacid to MAG by up to 4-fold (compounds
7-9). Thecorresponding halogenated derivatives were all found
toincrease the binding to MAG (compounds 10-13), with
themonofluorinated derivative achieving a 17-fold increase
inpotency. In contrast, amino substitution at the C-5
positionsignificantly attenuated binding to MAG.86 Together,
thesestudies highlight key interactions between MAG and the
C-9hydroxyl and C-5 N-acetyl groups of sialic acid.
In addition to probing monosaccharide variants,
numerousoligosaccharide derivatives have been synthesized and
tested
for binding to MAG. These structures mimic naturallyoccurring
ganglioside structures such as GD1a (Figure 2).Consistent with
previous studies, substitution of the C-9hydroxyl of Neu5Ac with a
methyl group within thetrisaccharide Neu5AcR(2-3)Gal(1-4)Glc
attenuated bind-ing to MAG by 5-fold, again highlighting the
importance ofthe glycerol side chain.87 These results are
consistent withan X-ray crystal structure of the Siglec
sialoadhesin com-plexed with sialyllactose, in which the C-9
hydroxyl groupof NeuAc forms a hydrogen bond with the amide
backboneof Leu-107.88 Although these proteins are distinct, it
isconceivable that their mode of binding to sialic acid wouldbe
conserved across Siglec family members. In contrast toC-9, the C-7
and C-4 hydroxyls do not appear to contributesubstantially to the
binding energy of MAGsialic acidinteractions.87 The C-7 deoxy
derivative of Neu5AcR(2-3)-Gal((1-4)Glc-2-azidoethyl exhibited only
slightly en-hanced binding to MAG (1.5-fold), whereas the C-4
deoxyderivative showed slightly decreased binding (2-fold).
How-ever, both the C-7 and C-4 hydroxyls appeared to be criticalfor
binding when placed in the context of a polyvalentarray.57 Thus,
valency and cell-surface presentation mayreflect another facet of
the complex regulation and specificityof Siglec-ganglioside
interactions.
Synthetic oligosaccharide derivatives have also providedinsight
into the importance of specific glycosidic linkagesand other
residues within the structure. MAG was found tobind 5-fold better
to R(2-3)-linked Neu5Ac than to R(2-6)-linked Neu5Ac in synthetic
trisaccharides.87 Interestingly,replacement of Neu5Ac in a
pentasaccharide structure withthe naturally occurring sialic acid
KDN led to a 6.5-foldincrease in MAG binding,87 suggesting that
other sialic acidforms may bind MAG in ViVo. In addition to
contacts withterminal sialic acid residues, internal sugars were
also foundto be important for MAG interactions. For instance,
substitu-tion of the C-4 hydroxyl group of galactose in
Neu5AcR(2-3)-Gal(1-4)Glc with a hydrogen atom enhanced binding
toMAG by 2.3-fold. Changing this residue to GalNAc, addingan
O-methyl substituent at C-6, or exchanging the ringoxygen to an
N-methyl or N-butyl functionality decreasedthe potency of the
trisaccharide.87 Modifications of the thirdglucose residue to
N-acetylglucosamine (GlcNAc) alsodecreased the binding properties
of the molecules. Varioussubstitutions of the N-acetyl group, such
as N-phthaloyl orN-octanoyl substituents, increased the potency of
the com-pounds, which reflects the potential for a
hydrophobicinteraction with MAG at this site.87 Lastly,
pentasaccharidesof the structure
Neu5AcR(2-3)Gal(1-4)AllNAc(1-3)-Gal(1-4)Glc-2-(trimethylsilyl)ethyl
(AllNAc ) N-acetyl-allosamine) were found to increase binding above
thetrisaccharide Neu5AcR(2-3)Gal(1-4)Glc by 6-fold, sug-gesting
even more extensive contacts between MAG and theinterior residues
of large glycan structures.87
Together, studies using synthetic analogues have illustratedhow
subtle perturbations to the sialic acid core structure canhave
significant effects on protein binding. As describedbelow, such
studies may facilitate the design of novelsynthetic inhibitors of
MAG function with therapeuticpotential.
2.3.2. Development of MAG Antagonists with
TherapeuticPotential
The importance of MAG-ganglioside interactions fornerve
regeneration and myelination has inspired the design
Figure 3. Synthetic sialic acid analogues tested for binding
toMAG. Positions important for MAG interactions are shown in
red.
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1711
-
and synthesis of small molecules capable of disrupting
thoseinteractions. Such molecules have the potential to
enhancenerve regeneration by blocking the inhibitory effects of
MAGon neurite outgrowth. Below, we provide some examples ofsmall
molecule antagonists that exhibit activity in cellularregeneration
models.
Paulson and co-workers examined the interactions ofmonovalent
sialic acid derivatives with MAG and otherSiglec family members.89
Over 25 derivatives representingmost of the major sialic acid
structures found on glycopro-teins and glycolipids were tested. The
most potent inhibitorof MAG-ganglioside interactions was the
disialyl
structureNeu5AcR(2-3)Gal(1-3)[Neu5AcR(2-6)]GalNAcR-O-ThrOCH3
(Figure 4A), which exhibited an IC50 value of 0.3M. This compound
showed greater than 12000-fold en-hanced potency relative to Neu5Ac
for inhibiting MAG-sialic acid interactions.89
The disialyl structure above and other potent inhibitorssuch as
Neu5AcR(2-3)Gal(1-3)GalNAc were subse-quently tested for their
ability to attenuate MAG-mediatedinhibition of neurite outgrowth.90
When rat cerebellar granuleneurons (CGN) are cultured on a
substratum of myelin-extracted proteins, they project fasciculated
axons and clustertogether, leaving the majority of the substrata
bare. This formof neuronal growth inhibition is mediated primarily
by MAG.The sialosides relieved the MAG-dependent inhibition ofCGN
neurons, enhancing nerve regeneration in a dose-dependent manner
and proportional to their relative bindingaffinities for MAG.90 The
most potent compound, the disialylstructure, completely reversed
the inhibition induced byMAG. Thus, synthetic glycans can
effectively enhance neuriteoutgrowth in Vitro and, when used in
combination with othertreatments, may provide a means to improve
functional recoveryafter neuronal injury. The ability to compare
various Siglecfamily members against a large number of sialoside
structureshas also revealed the specificity of Siglecs for
differentcarbohydrate epitopes and may help to fine-tune the
develop-ment of selective MAG antagonists.
Many oligosaccharide-based inhibitors are
syntheticallychallenging to produce and can suffer from poor
pharma-cokinetics. As an alternative to this approach, Ernst and
co-workers generated structurally simplified mimics of
theganglioside GQ1bR. In particular, the Gal and GalNAcresidues in
the trisaccharide Neu5AcR(2-3)Gal(1-3)GalNAcwere replaced with an
R-linked benzyl ether moiety, andaromatic residues were positioned
on the glycerol side chain(Figure 4B). Despite its smaller size,
this compound dis-played a remarkable 1000-fold enhanced binding
affinityrelative to the trisaccharide
Neu5AcR(2-3)Gal(1-3)Gal-NAc-2-(trimethylsilyl)ethyl. Although the
compound wasnot tested in cellular regeneration assays, it was
anticipatedto have improved pharmacokinetic properties due to its
lowermolecular weight and favorable Clog P value.9193
Similarapproaches may yield additional therapeutic leads with
thedesired inhibitory potency and pharmacokinetics for thetreatment
of demyelinating disorders.
2.3.3. Synthetic Mimics of R(2-8)-Linked PSA for
NerveRegeneration
PSA expression is generally considered a permissivedeterminant
in areas of neuronal growth and plasticity,making it a potential
therapeutic target for neuronal regen-eration. In fact, expression
of PSA has been shown topromote functional recovery and provide a
favorable envi-ronment for axonal regeneration in animal models of
spinalcord injury.94,95 In these studies, PSA-NCAM was ectopi-cally
expressed in spinal cord astrocytes in ViVo,94 or PSA-expressing
Schwann cell grafts were employed.95 Althoughthe use of PSA oligo-
and polysaccharides may be viablealternatives, PSA isolated from
natural sources is oftenheterogeneous in length and can be
contaminated with othercell-surface glycans. In addition, PSA
adopts a helicalconformation96 and forms filament bundles,97 thus
exhibitingdifferent structural elements that may have distinct
functions.
To circumvent these challenges, Rougon, Schachner, andco-workers
screened a large peptide library to identifypotential PSA
mimetics.98 Two cyclic peptides were identi-fied that recapitulated
the properties of endogenous PSA.Both compounds stimulated the
outgrowth and defascicula-tion of mouse dorsal root ganglion (DRG)
neurons andpromoted neuronal migration in Vitro and in ViVo. In
addition,one peptide enhanced the migration of transplanted
neuronalprogenitor cells in the murine olfactory bulb in ViVo via
apathway known to be regulated by PSA.98 Thus, syntheticmimics may
provide novel alternatives to PSA for neuronalregeneration.
2.3.4. Metabolic Labeling To Remodel Cell-Surface SialicAcid
Interactions
The metabolic labeling of glycan chains with unnaturalsugars has
played a key role in expanding the knowledge ofsialic acid function
in the nervous system. Early studies byReuttar and colleagues
demonstrated that unnatural chemicalfunctionalities could be
incorporated into cell-surface sia-lylglycoconjugates by the
addition of N-acetylmannosamineanalogues (ManNAc; Figure 5A) to
cells.99103 ManNAc isthe first committed intermediate in the sialic
acid biosyntheticpathway, and the enzymes in this metabolic pathway
arepromiscuous for some unnatural substrates.104106 As de-scribed
below, the ability to alter the structures of sialyl-
Figure 4. Structure of (A) a potent disialyl MAG inhibitor
and(B) a simplified mimic of the ganglioside GQ1bR with
enhancedbinding affinity to MAG relative to
Neu5AcR(2-3)Gal(1-3)-GalNAc.
1712 Chemical Reviews, 2008, Vol. 108, No. 5 Murrey and
Hsieh-Wilson
-
glycoconjugates has provided key insights into the roles
ofsialic acid in neuronal migration and proliferation.
2.3.4.1. Metabolic Labeling of Neurons with ElongatedN-Acyl
Derivatives of Sialic Acid. Elongated N-acyl deriva-tives of ManNAc
have been incorporated into sialylglyco-conjugates of PC12 cells,
oligodendrocyte progenitor cells,microglia, astrocytes, and neurons
from cerebellar microex-plant cultures.101,107 In these studies,
cells were treated withN-propanoylmannosamine (ManNProp), wherein
the N-acetylsubstituent of Neu5Ac is replaced with a longer
N-propanoylgroup (Figure 5A). ManNProp was found to stimulate
theproliferation of microglia relative to cells treated with
thenatural sialic acid precursor, ManNAc.107 ManNProp alsoinduced
the migration of oligodendrocyte progenitor cells,the precursors to
oligodendrocyte cells, which play key rolesin myelin formation and
become functionally impaired inneurological diseases such as
multiple sclerosis.108112Interestingly, treatment with ManNProp
prolonged expressionof a sialylated ganglioside involved in cell
migration, theA2B5 epitope,113 revealing a potential mechanism for
itsfunctional effects.
In other studies, Reutter and co-workers investigatedwhether
ManNProp modulates signaling pathways withinoligodendrocytes.114
Treatment of these cells with ManNPropand the inhibitory
neurotransmitter -aminobutyric acid(GABA) induced GABA-dependent
oscillations in intracel-lular calcium. Calcium is an important
second messenger inthe nervous system, and calcium oscillations are
believed tocontribute to a highly plastic signaling system
underlyingthe communication between neurons and glia.114
Interest-ingly, ionotropic GABA receptors are modified by
sialicacid,115,116 suggesting that extended N-acyl substituents
mayalter the functional properties of this receptor.
However,ManNProp undoubtedly perturbs the expression of
multiplesialylglycoconjugates at the cell surface, and direct
evidencethat altered sialylation of the GABA receptor is
responsiblefor the observed response is lacking. In the future, it
will beinteresting to uncover the precise molecular mechanisms
bywhich these modifications to sialic acid structure elicit
theireffects on intracellular signaling.
ManNProp has also been shown to promote neuronalgrowth in
various contexts. For instance, ManNProp inducedthe neurite
outgrowth of small rat CGN, PC12 cells, andchick DRG
neurons.117,118 Moreover, treatment with Man-NProp promoted
reestablishment of functional connections
in the perforant pathway, which consists of projections fromthe
entorhinal cortex into the dentate gyrus of the hippoc-ampus, in
coculture experiments.117 Although the particularglycoconjugates
responsible for these effects were notelucidated, several cytosolic
proteins implicated in neuriteoutgrowth were found to be
differentially expressed afterthe ManNProp treatment, including
unc-33 like phosphop-rotein (ULIP), various heat shock proteins,
and 14-3-3, aprotein that associates with both GABA receptors and
theR(2-3)-sialyltransferase IV.117,119,120
Bertozzi and colleagues have explored the influence ofvarious
ManNAc derivatives on PSA biosynthesis. N-Butanoylmannosamine
(ManNBut, Figure 5A), but notManNProp, was shown to significantly
inhibit PSA expres-sion in a dose-dependent manner in the NT2
neuroblastomacell line. Moreover, both human polysialytransferases
re-sponsible for PSA biosynthesis (STX and PST) displayedreduced
kinetic efficiencies for transfer of ManNBut andManNPent (Figure
5A), whereas ManNProp was transferredat a rate sufficient for
biosynthesis.118,121 Thus, elongationof the N-acyl side chain of
sialic acid may interfere withrecognition of the growing PSA chain
by polysialyltrans-ferases. However, findings by Jennings and
co-workerssuggest that both ManNBut and ManNProp may be
partiallyincorporated into sialylglycoconjugates, as detected by
flowcytometry using a monoclonal antibody that
recognizesN-propanoyl- and N-butanoyl-PSA.122,123 Consistent with
aninhibitory effect on PSA biosynthesis, ManNBut
blockedpolysialylation of NCAM in both chick DRG neurons118 andNT2
cells124 and decreased the outgrowth of DRG neu-rons.118 The
effects on neurite outgrowth were comparableto those elicited by
treatment of cells with endoneuramini-dase, an enzyme that cleaves
PSA residues.
2.3.4.2. Metabolic Labeling with ManNGcPA. Metaboliclabeling of
neurons with unnatural sugars has also beenexploited to alter
protein recognition events at the cellsurface. Treatment of
neuroblastoma-glioma hybrid cellswith the sialic acid metabolic
precursor N-glycolylman-nosamine pentaacetate (ManNGcPA; Figure 5A)
convertedcell-surface sialylglycoconjugates from expressing
Neu5Acto expressing Neu5Gc,125 a sialic acid form that is
notnormally found in humans.126 Whereas Neu5Ac
sialylgly-coconjugates displayed on neuronal cells bound
efficientlyto MAG, the binding of MAG to cells expressing
Neu5Gcsialylglycoconjugates was significantly inhibited.127
Thesestudies demonstrate the potential of metabolic labeling
toserve as a useful tool for perturbing specific
glycan-proteininteractions.
2.3.4.3. Chemoselective Labeling of Sialylated Cell-Surface
Glycoconjugates. The ability to incorporate un-natural sugar
analogues into cell-surface glycoconjugatesallows for the
introduction of reactive chemical functional-ities onto
glycoproteins and glycolipids, such as ketone,azide, or alkyne
groups. These functionalities allow forselective labeling of
proteins with reporter groups such asaffinity tags and fluorescent
dyes or for the delivery oftoxins.128131 Bertozzi and co-workers
have exploited N-levulinoylmannosamine (ManLev), which contains a
ketonefunctionality appended to the N-acyl side chain (Figure
5A),to label neuroblastoma cells.129 Incubation of the cells
withManLev resulted in incorporation of the ketone moiety
intosialylated glycans in a concentration-dependent
manner.Subsequent reaction with a biotin hydrazide
derivative(Figure 5B) enabled visualization of sialylglycans by
fluo-
Figure 5. (A) Mannosamine derivatives used for metabolic
labeling(R ) H or Ac) and (B) chemoselective labeling reaction
aftertreatment of cells with ManLev (R ) biotin).
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1713
-
rescence microscopy, revealing their presence along the cellbody
and neuronal processes.132 Although the specificsialyltransferases
involved are not fully understood, ManLevwas successfully
incorporated into PSA, suggesting thatR(2-8)-polysialyltransferases
are capable of utilizing ketone-modified precursors for PSA
synthesis.132 These studiesprovide a powerful means to modulate the
structure of PSAand potentially other sialylglycans with a wide
variety ofchemical groups.
2.3.4.4. Summary of Sialic Acid Metabolic Labeling.Cumulatively,
studies have demonstrated that unnaturalManNAc derivatives can be
exploited to manipulate thestructure of sialylated glycans on
neuronal cell surfaces.These studies have revealed that subtle
alterations in sialicacid structure can have striking consequences
for PSAbiosynthesis and biological phenomena such as
neuriteoutgrowth, cell proliferation, and migration. In the
future,these versatile chemical tools could be employed
forvisualization of dynamic neuronal processes in ViVo, suchas
activity-dependent changes in the expression or localiza-tion of
sialylglycans. The ability to engineer the glycancomposition of
cell surfaces and to selectively label sialylatedglycans for
imaging or other applications provides a powerfulcomplementary
approach to genetics and biochemistry.
3. r-L-Fucose
3.1. Structure and BiosynthesisR-L-Fucose (6-deoxy-L-galactose;
Fuc) is generally ex-
pressed as a terminal monosaccharide on N- and
O-linkedglycoproteins and glycolipids. As such, it often serves as
animportant molecular recognition element for proteins. Fucoseis
distinct from other naturally occurring sugars because itis a
deoxyhexose sugar that exists exclusively in theL-configuration
(Figure 6). A structurally diverse array offucosylated glycans has
been identified with fucose oftenlinked to the C-2, C-3, C-4, or
C-6 positions of thepenultimate galactose in glycoconjugates or to
the coreGalNAc residue of N-linked glycans.1 O-Fucosylation,
thedirect modification of serine and threonine residues byfucose,
has also been observed on epidermal growth factor(EGF) repeats of
glycoproteins such as Notch, a proteininvolved in cell growth and
differentiation.133 While fucoseis not elongated in N-linked and
O-linked glycans, O-fucosecan be elongated by other sugars.1
Given the structural diversity of fucosylated glycans, it
isperhaps not surprising that more than a dozen differenthuman
enzymes are involved in the formation of Fuc
linkages.1 Two enzymes, FUT1 and FUT2, are dedicated tothe
synthesis of FucR(1-2)Gal glycans, an epitope foundon the ABO blood
group antigens134136 that has also beenimplicated in synaptic
plasticity.13,137,138 A gene homologousto FUT1 and FUT2, called
Sec1, contains translationalframeshifts and stop codons that
interrupt potential openreading frames and thus appears to be a
pseudogene.134 FUT3catalyzes the synthesis of both R(1-3)- and
R(1-4)-fucosylated glycans and can transfer fucose to both Gal
andGlcNAc in an oligosaccharide chain, whereas FUT4-7 formonly
R(1-3)-fucosylated glycans.139,140 FUT8 and FUT9generate
FucR(1-6)GlcNAc structures, with FUT8 generallycatalyzing
attachment of this structure to the core asparagineresidue of
N-linked oligosaccharides141 and FUT9 catalyzingits attachment to a
distal GlcNAc of polylactosamine chains.142FUT10 and FUT11 are
putative fucosyltransferases that arereported to synthesize
R(1-3)-fucosylated glycans based onsequence homology, although no
functional studies have yetbeen performed.1 Finally, POFUT1 and
POFUT2, alsoknown as O-fucosyltransferase 1 and
O-fucosyltransferase2, catalyze the direct fucosylation of serine
and threonineresidues within epidermal growth factor
repeats.143,144
3.2. Neurobiological FunctionsFucosylated glycans play important
roles in various
physiological and pathological processes, including
leukocyteadhesion,145,146 host-microbe interactions,147,148 and
neu-ronal development.149,150 They are prevalent on the
gly-colipids of erythrocytes, where they form the ABO bloodgroup
antigens that distinguish specific blood types.136Aberrant
expression of fucosylated glycoconjugates has beenassociated with
cancer,151154 inflammation,145,155157 andneoplastic
processes.158,159 For instance, the fucosylatedantigens, sialyl
LewisX, sialyl LewisY, and sialyl LewisB,are up-regulated in
certain cancers and have been associatedwith advanced tumor
progression and poor clinical progno-sis.160163 Moreover,
deficiency in fucose leads to a con-genital disorder of
glycosylation type IIc in humans, alsoknown as leukocyte adhesion
deficiency type II (LAD II).This disorder results in the impairment
of leukocyte-vascularepithelium interactions and is characterized
by immunode-ficiency, developmental abnormalities, psychomotor
difficul-ties, and deficits in mental capabilities.164
Although their roles in the brain are less well
understood,fucosylated glycans have been implicated in neural
develop-ment, learning, and memory. Here, we will highlight
aspectsof their biosynthesis and functional roles in the
nervoussystem.
3.2.1. Neuronal Development
Fucose has been reported to play an important role inneural
development. O-Fucosylation is essential for theactivity of Notch,
a transmembrane receptor protein thatcontrols a broad range of
cell-fate decisions duringdevelopment., 165169 Studies suggest that
fucose modulatesNotch signaling either by inducing a conformational
changein the protein or by interacting directly with Notch
ligands.168Notch signaling is believed to be involved in
neuronalprogenitor maintenance, and governs the cell-fate
decisionbetween neuronal and glial lineages. Notch signaling
mayalso contribute to the behavior of differentiated neurons
andneuronal migration.170 Genetic deletion of the POFUT1 geneis
embryonic lethal in mice and causes developmental defects
Figure 6. Structures of various fucose derivatives and
2-dGal.
1714 Chemical Reviews, 2008, Vol. 108, No. 5 Murrey and
Hsieh-Wilson
-
similar to those observed upon deletion of Notch
receptors,including abnormal vasculogenesis, somitogenensis,
andneurogenesis.171,172 These studies demonstrate the impor-tance
of fucose in proper neuronal development and implicateNotch
fucosylation as an important mediator of these events.
3.2.2. Learning and Memory
Multiple studies have suggested a role for fucosylation
inlearning and memory. For instance, the incorporation offucose
into glycoconjugates in the brain was significantlyenhanced by
task-dependent learning in both chicks andrats.173176 Rats were
trained in a brightness discriminationtask, in which animals
learned to enter a bright chamberwhile avoiding a dark one. Trained
animals demonstratedan increase in [3H]-labeled fucose
incorporation into glyco-conjugates at synapses, the specialized
sites of communica-tion between neurons.175 Moreover, exogenous
applicationof L-fucose or 2-fucosyllactose (Figure 6) enhanced
long-term potentiation (LTP), an electrophysiological model
forlearning and memory, both in ViVo and in
hippocampalslices.177,178
Fucose is highly enriched at neuronal synapses,13,179,180where
the majority of the fucosylated glycoconjugates existas complex
N-linked structures.181 Studies indicate that theactivity of
fucosyltransferases increases during synaptoge-nesis182 and upon
passive-avoidance training in animals.183Moreover, the cellular
machinery involved in protein gly-cosylation can be found within
dendrites,184,185 raising theintriguing possibility that local
protein synthesis and fuco-sylation may be occurring at synapses in
response to neuronalstimulation.
Further studies have specifically implicated
FucR(1-2)Gallinkages in neuronal communication processes. For
instance,2-deoxy-D-galactose (2-dGal; Figure 6), which competes
withnative galactose for incorporation into glycan chains and
thusprevents the formation of FucR(1-2)Gal linkages,186 hasbeen
shown to induce reversible amnesia in animals.138,186,187In
contrast, other small molecule sugars such as 2-deoxy-D-glucose,
Gal, or Glc had no effect, suggesting a uniquefunction for
FucR(1-2)Gal saccharides. 2-dGal has also beenreported to interfere
with the maintenance of LTP, both inVitro and in ViVo.188,189
Furthermore, a monoclonal antibodyspecific for FucR(1-2)Gal190
significantly impaired memoryformation in animals, presumably by
blocking formation ofthe FucR(1-2)Gal epitope.137
3.3. Chemical Approaches for Studying L-FucoseDespite intriguing
evidence linking FucR(1-2)Gal sugars
to neuronal communication and memory storage, the mo-lecular
mechanisms by which these sugars exert their effectsare not well
understood. Recently, however, chemical toolshave been developed
that are beginning to shed light on theroles of FucR(1-2)Gal
lectins and glycoproteins in the brain.
3.3.1. Deoxygalactose Analogues
Hsieh-Wilson and co-workers investigated the effects ofthe
amnesic compound 2-dGal and other fucosylation inhibi-tors on
cultured hippocampal neurons. Inhibition of FucR-(1-2)Gal linkages
using 2-dGal led to stunted neuriteoutgrowth in young neurons
lacking functional synapses(Figure 7).14 In contrast,
3-deoxy-D-galactose (3-dGal), whichinhibits fucose incorporation at
the C-3 position of galactose,
had no effect on neurite growth, suggesting that specificfucose
linkages are important for the neuritogenic activity.The effects of
2-dGal could be successfully rescued by theaddition of excess D-Gal
to the media, suggesting that theinhibition can be reversed by the
de noVo synthesis ofFucR(1-2)Gal sugars.
Interestingly, 2-dGal also exerted dramatic effects on
themorphology of older neurons, even after axonal differentia-tion
and synaptogenesis had begun to occur.13 Applicationof 2-dGal led
to a remarkable retraction of dendrites andcollapse of synapses,
whereas 6-dGal had no effect. However,D-Gal was only partially able
to rescue the effects of 2-dGal,which may reflect the decreased
plasticity of older neurons.Thus, fucosylated glycans and, in
particular, FucR(1-2)Galglycoconjugates appear to be important for
modulatingneuronal morphology and maintaining functional
neuronalconnections.
To gain insight into the molecular mechanisms
involved,Hsieh-Wilsonandco-workerssought to
identifyFucR(1-2)Galglycoproteins in the hippocampus.13 Using a
gel-based massspectrometry approach, they identified synapsins Ia
and Ibas the predominant FucR(1-2)Gal glycoproteins in
olderhippocampal cultures and in the adult rat brain. The
synapsinsare synaptic vesicle-associated proteins that play
importantroles in neurotransmitter release and
synaptogenesis.191,192Fucosylation of synapsin I was found to have
significanteffects on synapsin expression in neurons, protecting it
fromproteolytic degradation by the calcium-activated
proteasecalpain. Moreover, studies using 2-dGal and synapsin
I-deficient mice showed that synapsin fucosylation contributesto
the profound effects of 2-dGal on neurite outgrowth andsynapse
formation. However, other unknown FucR(1-2)Galglycoproteins were
also involved in the process. Thesestudies provide the first
evidence that FucR(1-2)Gal gly-coproteins are directly involved in
neurite outgrowth andunderscore the importance of identifying the
FucR(1-2)Galproteome of the brain.
Figure 7. Inhibition of FucR(1-2)Gal linkages with 2-dGal
leadsto stunted neurite outgrowth in hippocampal neurons cultured
for4 days in Vitro (DIV). D-Gal is able to rescue the effects of
2-dGal.3-dGal has no effect. White bar indicates 45 m. Images
courtesyof C. Gama.
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1715
-
3.3.2. Glycopolymers and Imaging Probes
Fucose often occupies a terminal position on glycan chains,and
as such, it serves as an important molecular recognitionelement for
lectins. A well-studied example is the bindingof L-selectin to the
fucosylated glycan sialyl LewisX, aninteraction known to be
critical for leukocyte adhesion.1 Toinvestigate whether
FucR(1-2)Gal lectins exist in themammalian brain, a small molecule
probe was designed andsynthesized that contained the FucR(1-2)Gal
epitope and abiotin moiety for imaging potential lectin receptors
in thebrain (Figure 8).14 Rat hippocampal neurons were
incubatedwith the small molecule probe, and the bound probe
wasvisualized on the cells using a streptavidin-dye
conjugate(Figure 8). Strong fluorescent staining of the cell body
andneuronal processes was observed, consistent with the pres-ence
of fucose-binding lectin receptors.
To investigate whether the association of FucR(1-2)Galwith these
receptors would elicit a neuronal response, Hsieh-Wilson and
colleagues treated cultured neurons with poly-acrylamide-based
polymers displaying multiple copies of theFucR(1-2)Gal epitope.14
The FucR(1-2)Gal polymerspromoted neurite outgrowth by more than
75%, and thepotency of the compounds was dramatically enhanced
withincreasing polymer concentration or carbohydrate
valency.Importantly, polymers bearing other carbohydrates
moieties,such as GlcNAc, Gal, FucR(1-3)GlcNAc, or only Fuc, hadno
appreciable effects, indicating that the observed neurito-genic
activity was specific for FucR(1-2)Gal. Together, thesestudies
provide the first evidence that FucR(1-2)Gal lectinreceptors are
found in the brain, and they identify a novelcarbohydrate-mediated
pathway for the regulation of neuronalgrowth. This work also
highlights the power of chemicalprobes to explore the biological
effects of specific glycansand their associated receptors. It will
be important in thefuture to identify the lectins involved and to
elucidate thespecific mechanisms and pathways leading to
neuronalgrowth.
3.3.3. Metabolic Labeling Using Alkynyl or Azido
FucoseAnalogues
Recently, the Bertozzi and Wong groups independentlydemonstrated
that alkynyl- or azido-containing fucose ana-logues could be
exploited to selectively label and imagefucosylated glycans in
mammalian cells.193,194 Their strategyexploits the fucose salvage
pathway to convert unnaturalfucose sugars into the corresponding
GDP-fucose analogues,which then serve as donors for
fucosyltransferases. Once theazido or alkynyl fucose analogue is
incorporated into glycans,it can be reacted with fluorescent dyes,
biotin, or peptidesvia Staudinger ligation or [3 + 2] azide-alkyne
cycloaddi-tion chemistry. Bertozzi and co-workers synthesized
fucosederivatives with azido groups at the C-2, C-4, and
C-6positions.193 Only the C-6 azido fucose analogue (Figure 6)was
successfully incorporated into the glycans of the JurkatT
lymphocyte cell line, consistent with earlier observationsthat some
fucosyltransferases tolerate substitutions at the C-6position of
the pyranose ring. Wong and colleagues dem-onstrated that both
azido- and alkynyl-modified fucosederivatives (Figure 6) could be
incorporated into the glycansof hepatoma cells, allowing for
fluorescent imaging offucosylated glycoconjugates.194,195
Interestingly, the alkynylfucose analogue was shown to be
significantly less toxic tocells than the azido fucose analogue.194
Future applicationof these powerful approaches to neurons should
facilitateproteomic studies to identify fucosylated glycoproteins
andmay allow for the dynamic imaging of protein fucosylationin
ViVo.
3.3.4. Summary of Fucosyl Oligosaccharides
Cumulatively, studies using chemical probes have revealeda role
for fucosyl oligosaccharides and their associated lectinsand
glycoproteins in the regulation of neurite growth andsynapse
formation. These findings may shed light onbehavioral and
electrophysiological studies implicatingFucR(1-2)Gal in long-term
memory storage. Alterations inneuronal morphology, such as dynamic
changes in dendriticspine number and shape, occur during memory
consolidationand LTP.196,197 One possibility is that the
interaction betweencertain FucR(1-2)Gal glycoproteins and lectins
may promotethe stabilization of synaptic connections that underlie
learningand memory. In addition, fucosylation may exert its
effectsindependently of lectins, by stabilizing fucosylated
glyco-proteins such as synapsin or modulating their functions.
Thecontinued development and application of chemical tools
hastremendous potential to expand our understanding of the rolesof
fucosylated lectins and glycoproteins in the brain and mayprovide
exciting opportunities to modulate neuronal com-munication
processes.
4. O-GlcNAc Glycosylation
4.1. Structure and Biological FunctionsO-GlcNAc glycosylation is
the covalent attachment of
-N-acetylglucosamine to serine and threonine residues ofproteins
(Figure 9). Unlike other forms of glycosylation,O-GlcNAc is a
dynamic, reversible modification found onlyon intracellular
proteins, rendering it akin to protein phos-phorylation. A wide
range of proteins are O-GlcNAc-modified, including transcription
factors, nuclear pore pro-teins, cytoskeletal proteins, and
synaptic proteins.8,12,198,199202
Figure 8. Chemical probe for imaging lectin receptors (top)
andstaining of hippocampal neurons in culture (bottom panels)
withthe probe demonstrating the presence of FucR(1-2)Gal lectins
alongthe cell body and neurites. Cells were treated with 3 mM of
theimaging probe (A) or biotin (B), labeled with a
streptavidin-dyeconjugate, and imaged by fluorescence microscopy.
Images courtesyof C. Gama.
1716 Chemical Reviews, 2008, Vol. 108, No. 5 Murrey and
Hsieh-Wilson
-
Several excellent reviews have described the functional rolesof
O-GlcNAc in transcription,203 apoptosis,204,205
signaltransduction,199 nutrient sensing,206,207 and
proteasomaldegradation.206 O-GlcNAc glycosylation has also
beenimplicated in the cellular stress response208,209 and is
inducedby oxidative, osmotic, metabolic, and chemical
stress.8,206Levels of O-GlcNAc glycosylation are altered in
diseasestates such as cancer, diabetes, and Alzheimers
dis-ease.201,204,207,210215 Moreover, one of the hallmarks
ofAlzheimers disease is the formation of neurofibrillary tanglesby
hyperphosphorylated tau protein,216 and several studiessuggest that
O-GlcNAc glycosylation negatively regulatesthe ability of tau to
become phosphorylated.217,218 Thus, theinvestigation of O-GlcNAc
function may provide insightsinto our understanding of critical
cellular processes anddiseases.
4.2. Neurobiological Functions of O-GlcNAcEmerging evidence
indicates an important role for O-
GlcNAc glycosylation in the nervous system. The enzymesthat
catalyze the addition and removal of O-GlcNAc,O-GlcNAc transferase
(OGT) and O-GlcNAcase (OGA), aremost highly expressed in the
brain219 and are enriched inboth pre- and postsynaptic nerve
terminals.220 OGT expres-sion is critical for cell survival,221 and
neuronal-specificdeletion of the OGT gene in mice leads to
abnormaldevelopment, defects in motor coordination, and
earlyneonatal death.222 Thus far, more than 50 neuronal
proteinshave been shown to be O-GlcNAc-modified, includingproteins
involved in transcription (e.g., CREB (cAMP-response element
binding-protein), Sox2 (SRY box-contain-ing gene 2), ATF-2
(activating transcription factor-2)),neuronal signaling (synGAP
(synaptic Ras GTPase activatingprotein)), bassoon, the guanine
nucleotide exchange factorPDZ-GEF, and synapsin I), synaptic
plasticity (synaptopodinand -catenin), and neurodegenerative
disease (tau and APP(-amyloid precursor protein)).8,202,217,223227
Finally, O-GlcNAc glycosylation levels are dynamically modulated
byexcitatory stimulation of the brain in ViVo and upon activationof
specific kinase pathways in cultured cerebellar neurons.223
Despite its importance, the functional roles of
O-GlcNAcglycosylation are only beginning to be understood in
thebrain. A major challenge has been the difficulty of detectingand
studying the modification in ViVo. Similar to phospho-rylation,
O-GlcNAc is often dynamic, substoichiometric,targeted to
subcellular compartments, and prevalent on lowabundance regulatory
proteins. The sugar is also bothenzymatically and chemically
labile. For example, massspectrometry analyses to identify
O-GlcNAc-modified pro-teins and map glycosylation sites are
challenged by loss ofthe modification upon collision-induced
dissociation (CID).The lack of a well-defined consensus sequence
for OGT hasprecluded the determination of in ViVo glycosylation
sitesbased on primary sequence alone. Furthermore, the complex-ity
of the nervous system and its unique technical challenges(e.g.,
postmitotic cells, multiple cell types, blood-brainbarrier, complex
organization) greatly complicates efforts to
study O-GlcNAc glycosylation and necessitates the develop-ment
of rapid, highly sensitive detection methods. Here, wedescribe
chemical approaches undertaken to overcome thesechallenges and
highlight how they have advanced ourunderstanding of the roles of
O-GlcNAc glycosylation inneuronal function and dysfunction.
4.3. Chemical Tools To Study O-GlcNAcGlycosylation4.3.1.
Chemoenzymatic Labeling of O-GlcNAc Proteins
4.3.1.1. Rapid, Sensitive Detection. Traditional methodsfor
detecting O-GlcNAc-modified proteins often suffer fromlimited
sensitivity and specificity. For instance, radiolabelingof the
proteins using UDP-[3H]-galactose and (1-4)-galactosyltransferase
(GalT), an enzyme that transfers [3H]-galactose onto terminal
GlcNAc groups of glycoproteins,228can require weeks for
visualization and lacks the sensitivityto detect certain
O-GlcNAc-modified proteins. Lectins228 andantibodies229,230 are
also effective methods, but they bindonly a subset of the
O-GlcNAc-modified proteins (usuallythose with multiple
glycosylation sites) and have limitedaffinity and specificity.
In response, a chemoenzymatic approach for taggingO-GlcNAc
proteins was developed by Hsieh-Wilson and co-workers that allows
for more rapid and sensitive detection.An unnatural substrate for
GalT was designed, in which abioorthogonal ketone moiety was
appended to the C-2position of galactose (UDP-ketogalactose probe,
Figure10A).231 Studies by Qasba and colleagues had demonstratedthat
a mutant form of GalT (Y289L) tolerates minorsubstitutions at this
position.232 Once transferred, the ketonemoiety can be reacted with
an aminooxy biotin derivative,thus permitting the sensitive
detection of O-GlcNAc-modifiedproteins by chemiluminescence.231
Notably, this methodenables the identification of
O-GlcNAc-glycosylated proteinsthat elude detection using other
methods. For example,detection of the glycoproteins R-crystallin
and CREB wasaccomplished within minutes, whereas lectins and
antibodiesfailed to detect the modification on these proteins and
tritiumlabeling required more than a week to develop.231 Thus,
thischemoenzymatic approach provides superior sensitivity rela-tive
to traditional methods and accelerates the identificationof new
O-GlcNAc-modified proteins.
4.3.1.2. Identification of O-GlcNAc-Glycosylated Pro-teins from
Cells. Selective biotinylation of proteins usingthe chemoenzymatic
approach also facilitates the parallelpurification of
O-GlcNAc-modified proteins from cell ortissue extracts by affinity
chromatography.233 Previousmethods have necessitated purification
of individual proteinsprior to analysis, a tedious and
time-consuming process.Using the chemoenzymatic approach, the
tagged O-GlcNAcproteins can be isolated in a single step by
streptavidinaffinity chromatography and interrogated for
modificationin parallel by Western blotting.233 This strategy was
used todemonstrate that the AP-1 transcription factors c-Fos
andc-Jun, as well as the activating transcription factor ATF-1,are
O-GlcNAc-modified in HeLa cells.233 In addition, theidentification
of O-GlcNAc on CREB-binding protein (CBP)reveals a new class of
O-GlcNAc-glycosylated proteins, thehistone acetyltransferases
(HAT). Thus, glycosylation canbe readily investigated across
structurally or functionallyrelated proteins, as well as novel
functional classes. Together,studies have revealed that a broad
number of transcriptional
Figure 9. O-GlcNAc glycosylation.
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1717
-
components are O-GlcNAc-glycosylated,202,223,233 and O-GlcNAc
may function as a general regulatory modificationfor the control of
transcription.239,240
4.3.1.3. Proteome-Wide Analyses. When used in con-junction with
high-throughput mass spectrometry, thechemoenzymatic approach can
be exploited for proteome-wide analyses of O-GlcNAc-modified
proteins.202 Proteinsfrom cell lysates are chemoenzymatically
labeled and pro-teolytically digested. The desired glycopeptides
are thencaptured by avidin affinity chromatography and analyzed
byHPLC in line with tandem mass spectrometry (LC-MS/MS).The
ketogalactose-biotin tag facilitates the isolation ofO-GlcNAc
glycopeptides from complex mixtures. Thisenrichment step is often
crucial for detecting low-abundancepost-translational
modifications. The tag also provides aunique signature on the mass
spectrometer, thus enablingunambiguous identification of
O-GlcNAc-modified peptidesand mapping of glycosylation sites to
specific functionaldomains within a protein. Using this approach,
Hsieh-Wilson,Peters, and colleagues reported the first
proteome-wideidentification of O-GlcNAc-modified proteins from
themammalian brain.202 Nearly 100 peptides were
identifiedcontaining the mass spectrometry signature, and 34 of
thesepeptides were successfully sequenced. The sequenced pep-tides
identified 25 different proteins from rat brain. Of theproteins
identified, only two proteins had been previouslyreported, and 23
were novel O-GlcNAc-glycosylated pro-teins, thus significantly
expanding the repertoire of proteinsknown to be modified.
This method demonstrates the power of chemical-taggingapproaches
to accelerate the high-throughput identificationof O-GlcNAc
glycoproteins. Notably, many of the proteinsidentified have
important functional roles in gene regulation,
cytoskeletal dynamics, neuronal signaling, and
synapticplasticity. For example, synaptopodin, synGap, and
shank2(SH3 and multiple ankyrin repeat domains protein 2)
arecritical for the regulation of dendritic spine
formation.234236Synaptopodin and -catenin have important roles in
learningand memory,234,237 and the guanine nucleotide
exchangefactor PDZ-GEF is involved in the assembly of
signaltransduction complexes at the synapse.238 Together,
thesestudies suggest that O-GlcNAc glycosylation may play a rolein
mediating neuronal communication and signaling net-works.
Consistent with this observation, Burlingame and co-workers
recently employed lectin weak-affinity chromatog-raphy in
conjunction with mass spectrometry to identify
18O-GlcNAc-glycosylated proteins from the postsynaptic den-sity
fraction of rat brain.224 The proteins represent multiplefunctional
classes, and several proteins involved in synapticvesicle cycling
were found to be extensively O-GlcNAc-glycosylated, such as
bassoon, piccolo, and synapsin I.224
While the chemoenzymatic approach has broad applicationto the
study of O-GlcNAc-glycosylated proteins from celland tissue
extracts, O-GlcNAc proteins cannot be labeled inanimals using this
method. In addition, the determination ofexact glycosylation sites
is still difficult, because theketogalactose-biotin moiety can be
lost upon CID in themass spectrometer. Instead, O-GlcNAc
modification sites aremapped to short amino acid sequences within
proteins, whichstill provides insight into the function of the
modification.Despite these limitations, the chemoenzymatic
labelingstrategy is so powerful for in Vitro analysis and
proteomicsthat a variation of this approach is now
commerciallyavailable for fluorescent labeling or biotinylation of
O-GlcNAc-glycosylated proteins using [3 + 2] cycloadditionchemistry
(Figure 10B).
Figure 10. (A) Chemoenzymatic approach for tagging O-GlcNAc
glycosylated proteins, (B) UDP-azidogalactose probe for [3 +
2]cycloaddition chemistry using the chemoenzymatic approach, and
(C) GlcNAz and biotin phosphine probe for metabolic labeling of
O-GlcNAc-modified protein using the Staudinger ligation.
1718 Chemical Reviews, 2008, Vol. 108, No. 5 Murrey and
Hsieh-Wilson
-
4.3.2. Metabolic Labeling of O-GlcNAc Proteins
4.3.2.1. IncorporationofGlcNAzintoO-GlcNAcProteins.A
complementary strategy that enables tagging of O-GlcNAcylated
proteins in living cells involves metabolicallylabeling the
proteins with unnatural GlcNAc derivatives.Bertozzi and colleagues
demonstrated that N-(2-azidoacetyl)-glucosamine (GlcNAz, Figure
10C) is processed by enzymesin the hexosamine salvage pathway,
resulting in incorporationof a bioorthogonal azide functionality
into O-GlcNAc-glycosylated proteins.241 The azido group can be
subse-quently labeled with triarylphosphines via the
Staudingerligation. Using this approach, the authors
demonstratedsuccessful incorporation of GlcNAz into both nuclear
andcytoplasmic proteins of cultured Jurkat T lymphocyte cells.In
particular, selective labeling and detection of the nuclearpore
protein p62, a known O-GlcNAc-modified protein with>10
glycosylation sites,242 was shown using a phosphine-FLAG probe.
Although incomplete labeling of O-GlcNAc-glycosylated proteins
limits the sensitivity of this approachrelative to the
chemoenzymatic strategy described above,metabolic labeling with
GlcNAz sugars can be performedin living cells and might allow for
the dynamic imaging ofO-GlcNAc-glycosylated proteins in ViVo.
4.3.2.2. Proteomic Analysis by Metabolic Labeling.Although
metabolic labeling has not yet been applied toneurons, it
represents another powerful chemical approachfor the
high-throughput identification of O-GlcNAc-modifiedproteins. Zhao
and colleagues labeled O-GlcNAc proteinsin the HeLa cervical cancer
cell line with GlcNAz and taggedthem with a biotin phosphine
reagent (Figure 10C).243,244Tryptic digestion of the
affinity-captured proteins, followedby LC-MS/MS analysis, led to
the identification of 199putative O-GlcNAc-modified proteins.
Because the presenceof the GlcNAc moiety was inferred rather than
detecteddirectly, independent confirmation of the modification
byimmunoblotting was required and demonstrated on 23 of the199
proteins.
While this method provides a powerful chemical tool forprofiling
O-GlcNAc-modified proteins, there are some limi-tations of this
procedure for in ViVo labeling in the brain.Most sugars do not
cross the blood-brain barrier,245 andthus in ViVo labeling with
these molecules would entailinvasive surgical procedures for
intracranial administrationrather than simple intraperitoneal
injection. In addition,metabolic labeling is not quantitative,
which may limit thesensitivity of detection as well as preclude the
ability tomonitor glycosylation dynamics. Despite these
limitations,the approach has been successfully employed to
investigatethe O-GlcNAc proteome in both mammalian and insect
celllines.243,244 In the future, metabolic labeling could prove
auseful tool for studying the O-GlcNAc proteome in
culturedneurons.
4.3.3. Methods for Mapping Exact Glycosylation Sites
4.3.3.1. The -Elimination Followed by Michael Addi-tion with
Dithiothreitol (BEMAD) Approach. The iden-tification of O-GlcNAc
modification sites within proteins iscritical for elucidating the
functions of O-GlcNAc in specificbiological contexts. Nonetheless,
the exact sites of glycosy-lation remain unknown for most proteins.
Mapping glyco-sylation sites has been challenging due to the low
abundanceof the modification and the lability of the glycosidic
linkageduring fragmentation on a mass spectrometer, which canresult
in the loss of direct amino acid identification. Hartand co-workers
showed that the labile GlcNAc moiety couldbe replaced with a more
stable sulfide adduct by alkaline-induced -elimination followed by
Michael addition withdithiothreitol (BEMAD, Figure 11).246 The
resulting sulfideadduct is not cleaved upon CID, thereby allowing
sites ofglycosylation to be more readily determined. However,
alimitation of this approach is that it is often destructive
toproteins,247,248 and selectivity controls must be performedto
distinguish among O-GlcNAc, O-phosphate, and otherO-linked
carbohydrates.246 When biotin pentylamine is usedin place of
dithiothreitol, O-GlcNAc-modified peptides canbe selectively
biotinylated, enriched by affinity chromatog-raphy, and identified
by LC-MS/MS analysis. This methodhas been successfully employed to
identify novel O-GlcNAcsites on purified glycoproteins such as
synapsin I and proteinsfrom a purified rat brain nuclear pore
complex.246 Furtherextension of BEMAD to complex mixtures for the
high-throughput mapping of O-GlcNAc sites is an important
futuregoal.
4.3.3.2. Electron Transfer Dissociation (ETD) and Elec-tron
Capture Dissociation (ECD) Coupled with LectinAffinity
Chromatography or Chemoenzymatic Labeling.Recently, the development
of novel fragmentation methodsfor mass spectrometry has facilitated
the identification ofO-GlcNAc modification sites. Electron transfer
dissociation(ETD) and electron capture dissociation (ECD) use
thermalelectrons to produce sequence specific-peptide
fragmentationwithout the loss of labile post-translational
modifications suchas O-GlcNAc and O-phosphate.249 ECD has recently
beenused by Burlingame and co-workers to identify
O-GlcNAcglycosylation sites following enrichment of the
modifiedpeptides by lectin weak-affinity chromatography.224
Theauthors were able to identify glycosylation sites on
severalneuronal proteins such as spectrin 2, shank2, bassoon,
andpiccolo.
While ECD requires the use of a Fourier transform
massspectrometer, ETD has the advantage of being performed
inappropriately modified ion trap mass spectrometers, renderingthe
technology powerful and more accessible. Hsieh-Wilson,Coon, and
colleagues have implemented ETD fragmentationto map glycosylation
sites on neuronal proteins following
Figure 11. BEMAD approach for mapping O-GlcNAc glycosylation
sites.
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1719
-
chemoenzymatic labeling and enrichment by avidin
affinitychromatography. The authors identified glycosylation
siteson multiple proteins such as the neuron-specific
transcrip-tional repressor BHC80, the transcriptional repressor
p66,the transcriptional coactivator SRC-1, and the zinc
fingerRNA-binding protein.223 With further methodological
refine-ments and advances in database search algorithms forfragment
ions, it is anticipated that ETD and ECD willbecome increasingly
powerful tools for the study of O-GlcNAc glycosylation.
4.3.4. Monitoring O-GlcNAc Dynamics
Unlike most forms of protein glycosylation,
O-GlcNAcglycosylation is reversible and dynamic. Several studies
haveshown that global O-GlcNAc levels in cells change withinminutes
of activation by specific extracellular stimuli.250,251O-GlcNAc
levels are also highly responsive to cellularglucose
concentrations, as approximately 2-5% of allglucose is metabolized
through the hexosamine biosynthesispathway to generate
UDP-GlcNAc.252254 Furthermore,studies have suggested a potential
interplay between O-GlcNAc glycosylation and phosphorylation in
neurons. Aninverse relationship between O-GlcNAc and O-phosphatewas
observed upon activation of protein kinase C (PKC) orcAMP-dependent
protein kinase (PKA) in the cytoskeletalprotein fraction of
cultured cerebellar neurons.255 As de-scribed below, recent
quantitative proteomics studies haveshown that O-GlcNAc
glycosylation is dynamically inducedby excitatory stimulation of
the mammalian brain in ViVo.223Finally, O-GlcNAc glycosylation is
known to be dysregulatedin multiple disease states and is believed
to contribute tothe etiology of certain diseases, such as diabetes,
Alzheimersdisease, and cancer.207,252,256,257
Despite considerable investigation, the specific
proteinsundergoing dynamic changes in glycosylation remain
largelyunknown. Moreover, the molecular mechanisms and signal-ing
pathways involved in the regulation of OGT and OGAare poorly
understood. As such, there is a great need todevelop chemical tools
to monitor changes in glycosylationon specific proteins and at
specific modification sites in bothnormal and disease states. We
describe below some of thechemical approaches that have been
developed to addressthese challenges.
4.3.4.1. FRET-Based Sensors. Mahal and colleaguesdeveloped a
fluorescence resonance energy transfer (FRET)-based sensor to
investigate O-GlcNAc glycosylation dynam-ics in living cells.258
Their approach uses two fluorophores,enhanced cyan and yellow
fluorescent protein, separated by
a known OGT substrate domain and the bacterial O-GlcNAclectin
GafD (Figure 12). Upon O-GlcNAc glycosylation ofthe substrate
domain, the GafD domain binds the carbohy-drate moiety, bringing
the fluorophores into close proximityand leading to a concomitant
increase in FRET. The authorsdetected a significant increase in
FRET from HeLa cellstransfected with the sensor construct upon
treatment withglucosamine or the OGA inhibitor PUGNAc
(O-(2-acet-amido-2-deoxy-D-glucopyranosylidene)amino-N-phenylcar-bamate,
Figure 14). 258 This biological sensor represents apromising tool
for the investigation of O-GlcNAc glycosy-lation dynamics in
response to a variety of cellular stimuli.
4.3.4.2. The Quantitative Isotopic and
ChemoenzymaticTagging(QUIC-Tag)ApproachforQuantitativeProteomics.Hsieh-Wilson,
Peters, and co-workers developed a methodto probe dynamic changes
in O-GlcNAc glycosylation usingquantitative mass spectrometry-based
proteomics.223 TheirQUIC-Tag approach (quantitative isotopic and
chemoenzy-matic tagging) involves chemoenzymatically labeling
pro-teins from two different cell states (e.g., normal
versusdiseased; stimulated versus unstimulated) with the
keto-galactose-biotin group as described above (Figure 13).223After
proteolytic digestion, the resulting peptides are isoto-pically
labeled with either heavy or light isotope tags usingreductive
amination chemistry to distinguish the two popula-tions. The
peptides are subsequently combined, and thebiotinylated O-GlcNAc
peptides are captured using avidinchromatography. MS analysis
reveals two ions for eachglycosylated peptide (corresponding to
each of the twoisotopically labeled forms), and calculation of the
peak areasmeasures the change in glycosylation level for each
peptide.Importantly, as the observed peptides are sequenced
usingCID or ETD MS, the method identifies specific
proteinsundergoing dynamic changes in glycosylation and can beused
to monitor changes at particular glycosylation siteswithin
proteins.
This approach has advantages over other methods ofO-GlcNAc
detection. For instance, lectins and O-GlcNAcantibodies are
typically used to detect only global changesin O-GlcNAc
glycosylation by Western blotting and do notmonitor individual
glycosylation sites. Metabolic labelingusing GlcNAz may alter the
kinetic efficiency of O-GlcNActransfer to protein substrates, as
well as influx through thehexosamine biosynthesis pathway, which
complicates effortsto quantify dynamic changes in response to
cellular stimuli.In contrast, the QUIC-Tag approach is performed
ondenatured protein lysates and thus preserves the
physiological
Figure 12. A fluorescence resonance energy transfer (FRET)-based
sensor to detect O-GlcNAc glycosylation levels.
1720 Chemical Reviews, 2008, Vol. 108, No. 5 Murrey and
Hsieh-Wilson
-
glycosylation state of the protein without perturbing
intra-cellular glycosylation pathways.
By this approach, O-GlcNAc glycosylation was found tobe
stimulated upon PUGNAc treatment of cortical neuronsor kainic
acid-induced excitatory stimulation of rodent brainsin ViVo.223
Robust changes in O-GlcNAc glycosylation wereobserved at specific
sites on several proteins, whereas othermodification sites remained
unchanged, suggesting thatO-GlcNAc is subject to complex regulation
in neurons. Forexample, glycosylation of early growth response-1
(EGR-1), a transcription factor involved in long-term
memoryformation and cell survival,259,260 increased greater than
10-fold after kainic acid stimulation. Because the dynamic
glycosylation site within EGR-1 lies within its
transactivationdomain, O-GlcNAc glycosylation may modulate the
tran-scriptional activity of EGR-1 and modulate gene
expression.Cumulatively, these studies indicate that O-GlcNAc
glyco-sylation is reversible, subject to complex regulation,
andinduced by neuronal activity, which supports the notion
thatO-GlcNAc represents an important regulatory modificationin the
brain.
4.3.4.3. Stable Isotope Labeling with Amino Acids
inCellCulture(SILAC)CoupledwithAffinityChromatography.Recently,
Hart and co-workers employed the SILAC (stableisotope labeling with
amino acids in cell culture) methodfor quantitative proteomics261
in conjunction with immu-noaffinity chromatography to investigate
the interplay be-tween O-GlcNAc and phosphorylation in COS-7
kidneyfibroblast cells.262 Cells from two different states
werelabeled with either heavy or light isotopes of arginine
andcombined. Proteins of interest were subsequently isolatedby
affinity chromatography using a general O-GlcNAcantibody, resolved
by SDS-PAGE, proteolytically digested,and analyzed by LC-MS/MS.
Using this approach, Hart and colleagues investigated theeffects
of lithium inhibition of glycogen synthase kinase-3(GSK-3) on
O-GlcNAc glycosylation levels. GSK-3 isinvolved in multiple
intracellular signaling cascades and isimplicated in the etiology
of Alzheimers disease, diabetes,and bipolar disorder, thus making
it a desirable therapeutictarget.263,264 The authors identified 10
proteins that wereenriched after LiCl treatment, suggesting that
they underwentincreases in O-GlcNAc glycosylation. The increases
inglycosylation were confirmed on four proteins by
immuno-precipitation. Interestingly, many proteins exhibited
nochange, and 19 proteins showed decreases in glycosylation.These
studies suggest that a complex interplay exists betweenO-phosphate
and O-GlcNAc within signaling networks.
Although this approach works well for dividing cells,SILAC is
not amenable to tissues and quiescent cells suchas neurons. In
addition, the method does not readily enabledirect detection of the
O-GlcNAc modification, and thusindependent confirmation by
immunoprecipitation is required.Nonetheless, this approach provides
another powerful strat-egy to investigate the cellular dynamics of
O-GlcNAc glyco-sylation.
Figure 13. QUIC-Tag approach for quantifying dynamic changes in
glycosylation.
Figure 14. Small-molecule OGA inhibitors.
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1721
-
4.3.4.4. Small-Molecule Inhibitors of OGT and OGA.Traditional
genetic approaches have revealed insights intothe functions of OGT
and OGA in ViVo. For example, geneticdeletion of the OGT gene in
mice has revealed that OGT iscritical for cell survival, and
neuron-specific deletion of OGTresults in defects in mouse
embryogenesis, loss of locomotorcontrol, and neonatal death.221,222
Although such studies haverevealed an important role for these
enzymes in neuraldevelopment, investigations into the functions of
O-GlcNAcremain challenging, particularly in adult animals.
Thedevelopment of small-molecule inhibitors for OGT and OGAhas been
actively pursued to enable direct temporal andspatial control over
OGT and OGA activity.
Well-known small-molecule inhibitors of OGT such asalloxan
(Figure 14) show multiple nonspecific effects suchas inhibition of
OGA and glucokinase,265,266 as well asformation of superoxide
radicals.267 To develop betterpharmacological agents, Walker and
co-workers screened alibrary using a high-throughput,
fluorescence-based assay andidentified several novel compounds that
inhibited OGTactivity in Vitro.268 Notably, the compounds
selectivelyinhibited OGT but not MurG, a related enzyme that also
usesUDP-GlcNAc as a substrate.
As PUGNAc, the most commonly used OGA inhibitor,suffers from
nonspecific activity toward -hexosaminidase,269several groups are
working to develop more selectiveinhibitors. The Vocadlo and
Hanover groups have extendedthe N-acyl substituent of PUGNAc to
generate inhibitors with10-fold selectivity for OGA over
-hexosaminidase.269,270van Aalten and colleagues developed a
nagstatin derivativebased the crystal structure of a bacterial OGA
(Figure 14).271This molecule contains an isobutanamido group at the
N8position that improves selectivity by fitting into a pocket ofthe
enzyme and a phenethyl group at the C2 position thatinteracts with
a solvent-exposed tryptophan from bacterialOGA. More recently, the
Hanover and Vocadlo groupsindependently developed novel OGA
inhibitors based on thenonspecific hexosaminidase inhibitor
GlcNAc-thiazaoline, byadding fluoro, azido, or alkyl substituents
(Figure 14). Theresultant inhibitors exhibited over 3000-fold
selectivity forOGA over -hexosaminidase.272,273
The development of such compounds may enable theselective
inhibition of OGT and OGA in cultured neurons,as well as in ViVo.
The ability to perturb O-GlcNAc enzymesand glycosylation levels
with small molecules should revealnew information about the
functional roles of O-GlcNAcglycosylation in the nervous system, as
well as facilitate theidentification of signaling pathways that
regulate OGT andOGA.
5. Glycosaminoglycans
5.1. Structure and DiversityGlycosaminoglycans (GAGs) are
sulfated, linear polysac-
charides that represent a central component of the
extracel-lular matrix (ECM) and are involved in a myriad of
biologicalfunctions, including blood coagulation,274,275
angiogene-sis,276278 tumor growth and metastasis,279281
neuriteoutgrowth,282285spinalcordinjury,286288anddevelopment.289291They
are composed of repeating disaccharide units containinga hexuronic
acid sugar linked to a hexosamine sugar.292,293There are several
classes of GAGs (Figure 15), each of whichare distinguished by
backbone composition, including heparinand heparan sulfate (HS),
chondroitin sulfate (CS), dermatan
sulfate (DS), keratan sulfate (KS), and hyaluronic acid
(HA).Heparin and HS contain D-glucosamine (GlcN) and
eitherD-glucuronic acid (GlcA) or L-iduronic acid (IdoA)
connectedby R(1-4) and (1-4) linkages. In contrast, CS
polymerscontain N-acetylgalactosamine (GalNAc) instead of GlcNAcin
alternating (1-3) and (1-4) linkages to GlcA, whereasDS polymers
have both GlcA and IdoA linked to GalNAc.Heparin/HS and CS/DS are
attached to proteins throughO-linkages to serine residues via a
GlcA(1-3)Gal-(1-3)Gal(1-4)Xyl (Xyl ) xylose) tetrasaccharide
linker,forming glycoconjugates known as proteoglycans.294296 KSis
attached to proteoglycans through either N- or
O-linkages.Hyaluronic acid is unique in that it is not
protein-bound andis reportedly synthesized in the plasma
membrane,296,297whereas proteoglycans are synthesized in the Golgi
appa-ratus.292,293
In addition to having different backbone compositions,GAGs
display remarkable structural variation through sul-fation of
various hydroxyl groups along the polysaccharidebackbone (Figure
15). The sulfation patterns of GAGs areincredibly diverse, owing to
the large number of potentialsulfation sites and possible
combinations of differentiallysulfated disaccharides linked in
tandem. For example, heparinand HS disaccharide units can be
sulfated at the C-2 positionof IdoA or the C-3 and C-6 positions of
GlcN. The C-2 amineof GlcN can also be acetylated, sulfated, or
unmodified.Similarly, CS can be sulfated at the C-4 and C-6
positionsof GalNAc, as well as the C-2 and C-3 positions of GlcA.A
simple HS disaccharide has 48 potential sulfated se-quences,
yielding tetrasaccharides with over 2300 possiblesulfation
sequences.
GAGs also vary in chain length from 10 to 200disaccharide units,
with clusters of low and high sulfationalong the polysaccharide
backbone.298 Structural studiessuggest that GAGs can adopt a
variety of helical conforma-tions, such as variance in helical
pitch that may depend onthe associated counterion.299,300 Further
structural diversityis obtained from the conformational flexibility
of the pyra-nose ring of IdoA, which exists in equilibrium between
thechair and skew-boat conformations when sulfated at the
C-2position.298 Thus, the combination of different sequences,charge
distributions, and conformations gives rise to tre-mendous chemical
and structural diversity within glycosami-noglycan chains.
Figure 15. Structures of GAG subclasses. Potential sulfation
sitesare indicated in red. R ) SO3 or H; R1 ) SO3, H, or Ac; n
)10-200.
1722 Chemical Reviews, 2008, Vol. 108, No. 5 Murrey and
Hsieh-Wilson
-
5.2. Neurobiological Functions5.2.1. Neuronal Development
Evidence from genetic and biochemical approaches sug-gests that
the sulfation patterns of GAGs are important formodulating their
biological activity and can exert profoundeffects on organismal
development. For instance, mutationof the
N-deacetylase-N-sulfotransferase gene (Ndst-1) in-volved in HS
biosynthesis inhibits growth factor signalingthat disrupts normal
embryonic development in Droso-phila.290 HS and CS have been shown
to interact withnumerous growth factors and axon guidance proteins
in asulfation-specific manner.283,301308 Moreover, the
sulfationpatterns of HS and CS change during the course of
braindevelopment,309,310 and specific CS sulfation patterns
aredifferentially expressed in certain brain regions.311,312
Thesulfation patterns of HS and CS are also organ- and
age-specific, as is the expression of different
sulfotransferases.309,310Thus, HS and CS sulfation patterns in the
brain are tightlyregulated with the exquisite spatial and temporal
controlrequired for neuronal development.
5.2.2. Axon Guidance
In the developing nervous system, neurons are presentedwith a
variety of molecular cues that guide axons to theirproper targets.
HS sulfation has been implicated in axontargeting through the
interaction of the HS proteoglycanglypican-1 with Slit, a secreted
protein important for axonguidance, axon branching, and neuronal
cell migration.313,314Slit repelsaxonalgrowthbybinding to
theRoboreceptor.314,315Removal of HS by heparinase treatment or
addition ofexogenous HS containing specific sulfation patterns
inhibitsSlit binding to Robo and abolishes the axonal
repulsionmediated by Slit.304,315 These results suggest that HS
andparticular HS sulfation patterns play important roles
inmediating the chemotropic actions of Slit. In other studies,HS
sulfation was shown to be critical for neuronal outgrowthand axon
guidance in Caenorhabditis elegans. Using geneticapproaches, Hobert
and colleagues demonstrated that certainneuronal subtypes require
the HS-modifying enzymes C5-epimerase, 2-O-sulfotransferase, and
6-O-sulfotransferase forproper axon guidance.316 Interestingly,
other subclasses ofneurons require only the C5-epimerase or
2-O-sulfotrans-ferase, and some neuronal subtypes do not require
any ofthe HS modifying enzymes. Cumulatively, these
studiesdemonstrate that HS sulfation patterns play important
rolesin neuronal development and may encode axon guidance cuesto
direct neurons to their proper targets in ViVo.
5.2.3. Spinal Cord Regeneration
Chondroitin sulfate proteoglycans (CSPGs) are crucialcomponents
of perineuronal nets, structures of ECM mol-ecules surrounding the
soma and proximal dendrites ofcertain neurons in the brain and
spinal cord.317,318 CSPGsand other ECM molecules are recruited to
sites of CNS injuryand form a portion of the glia scar, a structure
that inhibitsaxonal regeneration and contributes to permanent
paralysisin ViVo. Several groups have demonstrated the importanceof
CSPGs and their associated sugar chains in mediatingneuronal
inhibition after spinal cord injury. For instance,CSPGs have been
shown to inhibit the neurite outgrowth ofDRG and CGN neurons in
Vitro.319,320 Moreover, degradingCS chains with chondroitinase ABC
(ChABC), an enzyme
that cleaves CS into disaccharide units, reverses the
inhibitoryeffects of CSPGs on neurite outgrowth.321,322 Most
notably,Fawcett, McMahon, and colleagues discovered that
ChABCdigestion of CSPGs promotes spinal cord regeneration inViVo,
with concomitant partial recovery of proprioceptivebehaviors and
locomotor skills in mice.323,324 These and otherstudies indicate
that CSPGs exert a crucial inhibitory roleon neuronal regeneration
and represent valid targets fortherapeutic intervention. Such
studies also underscore theimportance of CS glycosaminoglycans in
this process andthe need to further understand the molecular
mechanismsand sulfation patterns involved in directing their
activity.
5.3. Challenges to the Study of GAGsWhile GAGs play a
fundamental role in many neurobio-
logical processes, a molecular level understanding of the
rolesof specific sulfation sequences in mediating GAG functionsis
largely unknown. GAG biosynthesis is not template drivenand lacks
the proofreading capabilities of DNA biosynthesis,which results in
greater chemical heterogeneity and structuraldiversity within GAG
chains. Thus, GAGs purified fromnatural sources are often mixtures
of compounds that containdifferent sulfation patterns and chain
lengths. Characteriza-tion of these structures is challenging and
is often describedsimply in terms of the percent composition of
distinct sulfateddisaccharide subunits. Little is known about the
precise linearsequences of GAG polysaccharides, although methods
tosequence short oligosaccharide sequences are
becomingavailable.325327 Given these challenges, the synthesis
ofhomogeneous oligosaccharides containing defined
sulfationsequences has the potential to significantly advance
ourunderstanding of the structure-activity relationships
ofglycosaminoglycans. Here, we will highlight chemical ap-proaches
that have helped to decipher the roles of GAGs inthe nervous system
and efforts to develop GAG-basedtherapeutics for neurodegenerative
diseases.
5.4. Synthetic Molecules for ProbingStructure-Activity
Relationships
As described above, the sulfation patterns of GAGs areimportant
for directing their neurobiological functions.Although genetic
approaches have revealed crucial roles forGAGs in neural
development, such experiments lead toglobal changes in sulfation
throughout the carbohydratechain, precluding the identification of
specific sulfation motifsresponsible for biological activity. The
use of chemicallydefined small-molecule GAGs has provided insight
into theirneurobiological roles and demonstrated the importance
ofspecific sulfation sequences in mediating GAG functions.
5.4.1. Synthesis of Glycosaminoglycans
Early work on glycosaminoglycans focused primarily onthe
synthesis of heparin oligosaccharides.328336 Heparin hasbeen used
since the 1940s as an antithrombic agent, and aunique heparin
pentasaccharide sequence was discovered in the1980s as a potent
factor Xa inhibitor.298 The first syntheses ofheparin
pentasaccharides required over 60 steps, producedheparin in
relatively low yield, and were impractical for thedevelopment of
synthetic drugs. Since then, the efforts ofmultiple laboratories
have contributed methods that allowfor efficient syntheses of
heparin, HS, and their ana-logues.337344
Chemical Neurobiology of Carbohydrates Chemical Reviews, 2008,
Vol. 108, No. 5 1723
-
GAGs are notoriously difficult to synthesize, requiring
theformation of stereospecific glycosidic linkages, uronic
aciddonors and acceptors with low chemical reactivity,
andsophisticated protecting group strategies to effect
regiose-lective sulfation. Heparin, HS, and DS oligosaccharides
alsonecessitate efficient syntheses of the challenging
L-idopyra-nosyl sugar. The synthesis of GAGs has been summarizedin
several excellent reviews (see refs 337, 344347) Recently,there has
been great interest in generating libraries of sulfatedcompounds to
probe the role of sulfation and identifybiologically active
sulfation motifs.2,285,339,340,342,348 Ingeneral, these approaches
implement modular, convergentsynthetic strategies that afford
multiple sulfated structuresfrom a common disaccharide synthon and
thus minimize thenumber of steps.
Other strategies have employed chemoenzymatic routesto generate
defined GAG oligosaccharides. Kobayashi andco-workers have
capitalized on the promiscuity of hyal-uronidase, an enzyme that
normally catalyzes the hydrolysisof chondroitin in ViVo, to effect
glycosidic bond formationand generate GAG polymers.349353 They were
able todemonstrate the efficient polymerization of
N-acetylhyalo-biuronate [GlcA(1-3)GlcNAc] and
N-acetylchondrosine[GlcA(1-3)GalNAc] derivatives to form HA and
nonsul-fated chondroitin, respectively, as well as unnatural
cho