A visualizable chain-terminating inhibitor of glycosaminoglycan biosynthesis in developing zebrafish

Article abstract of DOI:10.1002/anie.201310569

Heparan sulfate (HS) and chondroitin sulfate (CS) glycosaminoglycans (GAG) are proteoglycan-associated polysaccharides with essential functions in animals. They have been studied extensively by genetic manipulation of biosynthetic enzymes, but chemical to

Full text of DOI:10.1002/anie.201310569

Angewandte
Chemie
DOI: 10.1002/anie.201310569
Glycosylation Inhibitor
AVisualizable Chain-Terminating Inhibitor of Glycosaminoglycan
Biosynthesis in Developing Zebrafish**
Brendan J. Beahm, Karen W. Dehnert, Nicolas L. Derr, Joachim Kuhn, Johann K. Eberhart,
Dorothe Spillmann, Sharon L. Amacher, and Carolyn R. Bertozzi*
Abstract: Heparan sulfate (HS) and chondroitin sulfate (CS)
glycosaminoglycans (GAG) are proteoglycan-associated poly-
saccharides with essential functions in animals. They have been
studied extensively by genetic manipulation of biosynthetic
enzymes, but chemical tools for probing GAG function are
limited. HS and CS possess a conserved xylose residue that
links the polysaccharide chain to a protein backbone. Here we
report that, in zebrafish embryos, the peptide-proximal xylose
residue can be metabolically replaced with a chain-terminating
4-azido-4-deoxyxylose (4-XylAz) residue by administration of
UDP-4-azido-4-deoxyxylose (UDP-4-XylAz). UDP-4-XylAz
disrupted both HS and CS biosynthesis and caused devel-
opmental abnormalities reminiscent of GAG biosynthesis and
laminin mutants. The azide substituent of protein-bound 4-
XylAz allowed for rapid visualization of the organismal sites
of chain termination in vivo through bioorthogonal reaction
with fluorescent cyclooctyne probes. UDP-4-XylAz therefore
complements genetic tools for studies of GAG function in
zebrafish embryogenesis.
of animal cell surfaces and the extracellular matrix. The most
widely studied GAGs are heparan sulfate (HS) and chon-
droitin sulfate (CS), which have distinct disaccharide repeats
but share a common core tetrasaccharide that links them to an
underlying protein scaffold (Figure 1A). Variable sulfation
along a GAG chain can generate an enormous array of
structures, allowing them to interact with diverse biomole-
cules. For example, GAG binding to signaling molecules,
including members of the Hedgehog, TGFb and FGF
families, and to extracellular matrix molecules and basement
membrane components such as laminin, is essential for proper
animal development.[1,2] Accordingly, mutations in the GAG
biosynthetic machinery can lead to dramatic phenotypes.[3]
For example, mice lacking Ext1 or Ext2, the enzymes
responsible for polymerizing HS, die in early development
G
lycosaminoglycans (GAGs), linear polysaccharides com-
posed of repeating disaccharides, are important components
[*] B. J. Beahm, Dr. K. W. Dehnert, Prof. C. R. Bertozzi
Department of Chemistry and Molecular and Cell Biology
Howard Hughes Medical Institute, University of California
Berkeley, CA 94720 (USA)
E-mail: crb@berkeley.edu
N. L. Derr, Prof. S. L. Amacher
Departments of Molecular Genetics and Molecular and Cellular
Biochemistry; Center for RNA Biology, Ohio State University
Columbus, OH 43210 (USA)
Dr. J. Kuhn
Institut fꢀr Laboratoriums- und Transfusionmedizin
Herz- und Diabeteszentrum Nordrhein-Westfalen
Georgstraße 11, 32545 Bad Oeynhausen (Germany)
Prof. J. K. Eberhart
Department of Molecular Biosciences, University of Texas
Austin, TX 78713 (USA)
Dr. D. Spillmann
Department of Medical Biochemistry and Microbiology
The Biomedical Center, Uppsala University
Husargatan 3, PO Box 582, 75123 Uppsala (Sweden)
Figure 1. Xylose initiates diverse glycosaminoglycan structures and can
therefore be targeted for metabolic replacement with a visualizable
chain-truncating analog. A) Heparan sulfate and chondroitin sulfate
structures. R1 =H or SO32À; R2 =Ac or SO3ÀÀ. GlcA, glucuronic acid;
IdoA, iduronic acid; GlcNAc, N-acetylglucosamine; GalNAc, N-acetyl-
galactosamine; Gal, Galactose; Xyl, Xylose; asterisk, position at which
epimerization can occur. B) With an azide in place of the hydroxy
group that is elaborated during GAG enogation, 4-XylAz both inhibits
GAG biosynthesis and enables visualization using a cyclooctyne probe
conjugated to a fluorophore. Green star, fluorophore.
[**] We thank J. Jewett, E. Sletten, B. Belardi, J. Hudak, T. Gallagher, S.
Laughlin and B. Swarts for helpful discussions and for critical
reading of the manuscript. This work was supported by NIH grants
to C.R.B. (GM58867) and S.L.A. (GM61952). The lamb1ab1166 allele
was isolated in a screen supported by NIH grant HD22486 (to
Charles B. Kimmel).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201310569.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
from defective gastrulation.[4–6] GAG biosynthesis mutations
in zebrafish embryos have been shown to cause craniofacial
developmental defects.[7]
While nucleotide sugars cannot be effectively adminis-
tered to cultured cells, the situation is quite different for
zebrafish embryos. Our lab as well as Wu and co-workers have
recently shown that nucleotide sugars can be delivered to all
the cells of a zebrafish embryo by microinjection into the yolk
sack at the 1–4 cell stage.[26–28] This observation suggested
a route to inhibiting GAG biosynthesis with UDP-xylose
analogs. In addition to providing a convenient route for
delivery of nucleotide sugars, zebrafish have become a power-
ful, well-established animal model for vertebrate develop-
ment, a biological setting in which GAGs are known to have
numerous functions.[3]
Here we report that UDP-4-azido-4-deoxyxylose (UDP-
4-XylAz) acts as chain-terminating metabolic inhibitor of
GAG biosynthesis in developing zebrafish. After injection
into zebrafish embryos, the compound is recognized as
a substrate by one or more protein XylTs, resulting in the
addition of 4-XylAz to sites where GAGs would normally be
elaborated. Replacement of the C-4 hydroxy group with an
azide both inhibits GAG elaboration and enables visual-
ization of sites within the embryo where inhibition has
occurred by reaction with cyclooctyne-functionalized imaging
probes (Figure 1B).[26,27,29] This dual-function inhibitor adds
to the much-needed chemical toolkit for studying GAGs
in vivo.
Studies of GAG biology typically involve genetic manip-
ulation of biosynthetic enzymes (e.g., by gene disruption,
antisense knockdown or RNAi) followed by phenotypic
analysis. However, the enzymes may have functions beyond
their roles in catalysis. For example, several proteins involved
in HS biosynthesis are known to exist in multi-enzyme
complexes in the Golgi compartment.[8–10] As a consequence,
diminished expression of one enzyme can indirectly affect the
activity, stability and subcellular localization of others.[8–10] In
other areas of biology, such complexities have motivated the
development of complementary chemical tools that target
proteins of interest in a fundamentally different manner.[11]
Chemical tools for modulating the biosynthesis of GAGs
in cells and organisms are presently quite limited. The
majority of work has focused on aryl xylosides as “primers”
of GAG biosynthesis.[12,13] As the first sugar of the conserved
core tetrasaccharide (Figure 1A), xylose is attached through
a b-linkage to serine residues by the action of one or more
protein xylosyltransferases (XylTs).[14] The xylose residue is
then elaborated with three more conserved residues before
commitment to a specific GAG chain. b-Xylosides with
hydrophobic aglycones can mimic xylosylated proteins and
compete for the elaborating enzymes. These compounds do
not technically inhibit GAG biosynthesis; rather, they pro-
vide an alternative substrate for GAG elaboration and lead to
the production of proteoglycans without GAGs. Further, the
GAG-modified b-xylosides are secreted into the extracellular
environment where they can acquire new activities, con-
founding interpretation of the observed biological effects.[15]
Recently, this approach was modified to generate true
inhibitors of GAG biosynthesis by replacing the C-4 hydroxy
group of b-xylosides, the position from which the GAG chain
is polymerized, with a fluorine atom.[16,17] These fluorinated
xylose analogs effectively inhibit GAG biosynthesis in
cultured cells.
At the outset of this work we prepared the panel of azido
UDP-xylose analogs shown in Figure 2, wherein each hydroxy
group of xylose was individually replaced with an azide group.
To prepare UDP-4-XylAz, we began by selectively benzoy-
lating the 1-, 2- and 3-hydroxy groups of l-arabinose, the C-4
Chain-terminating metabolic inhibitors comprise a second
though less well-developed class of GAG biosynthesis
disruptors.[18–21] These compounds are simple deoxy or
fluorinated sugars that lack the key hydroxy group needed
for polysaccharide elongation. Their metabolic conversion to
nucleotide-sugar glycosyl donors and subsequent enzymatic
incorporation into growing GAG chains leads to truncation
and, therefore, incomplete GAG biosynthesis. A complica-
tion inherent to this approach is that most GAG monosac-
charide constituents are widely distributed among other
glycan types. The corresponding deoxy and fluoro analogs
therefore perturb numerous glycan structures in addition to
GAGs.[19,22] The exception is xylose, which is almost exclu-
sively found in GAGs.[23,24] But chain-terminating xylose-
based GAG inhibitors have not been explored, perhaps due to
the absence of a salvage pathway that would convert simple
xylose analogs to the UDP-activated form used by protein
xylosyltransferases.[25] This problem would be obviated by the
direct use of UDP-xylose analogs as chain-terminating
metabolic inhibitors, but the anionic nature of nucleotide
sugars renders them cell membrane impermeant.
Figure 2. UDP-XylAz analogs. UDP-4-XylAz (1), UDP-2-XylAz (2), and
UDP-3-XylAz (3).
epimer of xylose, using a known procedure.[30] The free
hydroxy group of 1,2,3-tri-O-benzoyl-l-arabinose (4) was
then converted to an azide, giving 5 as shown in Scheme 1.
The benzoyl groups were exchanged with acetyl groups to
facilitate production of the desired a anomer during phos-
phorylation of the C-1 hydroxy group in a later step. Selective
deprotection of the anomeric position with hydrazine acetate
produced 6, which was reacted with diallyl-N,N-diisopropyl-
phosphoramidite to give the globally protected xylose-1-
phosphate analog 7.[31] After deprotection of the phosphate
group and immediate reaction with UMP-N-methylimidazo-
lide, UDP-4-XylAz was obtained.[31] Similar syntheses for
UDP-2-XylAz (2) and UDP-3-XylAz (3) are detailed in the
Supporting Information (Scheme S1 and S2).
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Scheme 1. UDP-4-XylAz synthesis. Reagents/conditions: a) 1. Tf2O,
pyridine, DCM, À208C; 2. LiN3, DMF, RT, 85% (2 steps); b) 1.
NaOMe, MeOH, RT; 2. Ac2O, pyridine, 08C to RT, 80% (2 steps);
c) hydrazine acetate, DMF, RT, 54%; d) 1. diallyl N,N-diisopropylphos-
phoramidite, 1H-tetrazole, DCM, RT; 2. mCPBA, DCM, À408C, 31% (2
steps); e) 1. [Pd(PPh3)4], sodium p-toluenesulfinate, THF, MeOH, RT;
2. UMP-N-methylimidazolide, MeCN, 08C; 3. NEt3, MeOH, water, RT,
22% (3 steps). DCM, dichloromethane; DMF, dimethylformamide;
mCPBA, meta-chloroperbenzoic acid.
Each azido UDP-xylose derivative, 1–3, was tested for its
ability to deliver azidoxylose into zebrafish proteoglycans. In
these experiments, 50–100 pmol of 1, 2 or 3 was injected into
zebrafish embryos at the 1–4 cell stage. After developing to
24 h post fertilization (hpf), embryos were incubated with
difluorocyclooctyne-AlexaFluor 488 (DIFO-488, 200 mm),
mounted in agarose, and imaged by confocal microsco-
py.[26,27,29,32] While no labeling was detectable for embryos
injected with 2 or 3 (Figure S3), azide-dependent labeling was
observed for embryos injected with UDP-4-XylAz (1)
(Figure 3). Flow cytometry, which provides more sensitive
detection than fluorescence microscopy, confirmed that
azide-dependent labeling was only observed after treatment
with 1 (Figure S4).
Figure 4. Effect of UDP-4-XylAz on A) HS expression, B) CS expression,
C) HS total sulfation, D) CS total sulfation, E) HS disaccharide compo-
sition, and F) CS disaccharide composition. Gray bars, untreated;
black bars, UDP-4-XylAz. Error bars denote the standard deviation
from three replicate experiments. *p<0.05, **p<0.01, ***p<0.001.
As human XylT1 is highly homologous to its zebrafish
counterpart, sharing 75% identity, this result supports the
claim that XylT1 is able to transfer UDP-4-XylAz onto
proteoglycans in zebrafish.
Figure 3. UDP-4-XylAz enables in vivo imaging of the site of inhibition
in zebrafish embryos. Scale bars: 10ꢁ, 200 mm; 20ꢁ, 50 mm.
Having confirmed that XylT1 can transfer UDP-4-XylAz
to an acceptor peptide in vitro, we disrupted translation of
XylT1 in zebrafish using a splice-blocking morpholino
oligonucleotide (MO) and analyzed the effect on 4-XylAz
labeling.[33] We used a previously-reported XylT1 MO and
verified its effect on the XylT1 transcript by reverse tran-
scription polymerase chain reaction (RT-PCR) (Figure S6 and
S7).[33] Then, embryos were injected with 50 pmol of UDP-4-
XylAz alone or 50 pmol of UDP-4-XylAz and 6 ng of XylT1
MO, and were allowed to develop to 24 hpf. Enveloping layer
To establish that UDP-4-XylAz delivers 4-XylAz to sites
of GAG glycosylation, we performed biochemical assays with
human xylosyltransferases 1 and 2 (XylT1 and XylT2), which
are known to initiate GAG glycosylation.[33] An acceptor
peptide was incubated with UDP-4-XylAz and either XylT1
or XylT2 and the glycopeptide product was analyzed by mass
spectrometry (Figure S5). While XylT1 transferred 4-XylAz
to the acceptor peptide, XylT2 was inactive on the substrate.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Comparison of embryos
injected with the XylT1
MO
and
untreated
embryos confirmed that
the MO had no effect on
fucosylation, and, by infer-
ence, global glycosylation.
After confirming that
replacement of the C-4 hy-
droxy group of xylose with
an azide prevents elabora-
tion by b4-galatosyltrans-
ferase (b4GalT7)[34] in vi-
tro (Figure S10), we inves-
tigated the effects of UDP-
4-XylAz treatment on the
total amount of HS and CS
per embryo, which we
quantified using a previ-
ously described method.[7]
GAGs were isolated from
UDP-4-XylAz-treated or
untreated embryos, enzy-
matically degraded into
disaccharides, and ana-
lyzed by reverse phase ion
pairing HPLC. Treatment
with UDP-4-XylAz caused
Figure 5. UDP-4-XylAz injection causes embryological defects. Compared to 24 hpf controls (A), injected
embryos (B) have a shortened axis, disorganized brain with enlarged hindbrain ventricle, small eyes, and
abnormally shaped myotomes (see inset). By 48 hpf, notochord defects and tail blistering are prominent in
treated versus control embryos (C and D). At 6 dpf, treated embryos have neurocranium defects, including
a narrow ethmoid plate (arrow) and less cohesive midline cartilagenous cells (arrowhead), and the absence of
a differentiated notochord (asterisk) (E and F), phenotypes strikingly similar to that of lamininb1ab1166 mutant
embryos (G and H). Another craniofacial defect in treated embryos is the narrowing of the lower jaw (I and J);
visceral skeletal elements (K and L), including the pharyngeal arches (not shown) are largely normal. Alcian-
stained cartilage elements are blue; Alizarin-stained bone elements are red. Although the alizarin staining in
the lamb1ab1166 mutant panel had faded significantly prior to imaging, additional preparations confirm the
notochord defect described for laminin mutants.[37]
a
reduction in HS and
CS levels by 47% and
77%, respectively (Fig-
ure 4A,B). The unequal
effect of the inhibitor on
HS and CS levels is consis-
tent with observations
in zebrafish mutants with
genetic knockdowns of
the GAG biosynthesis
enzymes responsible for
constructing the core tetra-
cells were labeled with NHS-647 and the embryos were
dissociated into single cells, incubated with DIFO-biotin
followed by avidin-AlexaFluor-488, and analyzed by flow
cytometry. Knockdown of XylT1 activity by MO treatment
reduced UDP-4-XylAz-dependent fluorescence labeling by
30% (Figure S8). This partial decrease in labeling was
expected because embryos have maternally provided XylT
activity that cannot be affected by the XylT1 MO.[33]
We next sought to ensure that the decrease in labeling
upon treatment with the MO was not due to a non-specific,
global perturbation of glycan biosynthesis. Fucose is a mono-
saccharide not found in zebrafish GAGs; its incorporation
into cell-surface glycans should therefore be unaffected by the
XylT1 MO. To determine the effects of the XylT1 MO on cell-
surface fucosylation, a proxy for non-GAG glycosylation, we
dissociated MO-treated embryos into single cells, incubated
them with a biotinylated Aleuria aurantia lectin that recog-
nizes fucose, incubated with avidin-AlexaFluor-488, and
measured for fluorescence by flow cytometry (Figure S9).[27]
saccharide.[7] We further analyzed the extent of sulfation of
HS and CS in the presence and absence of the metabolic
inhibitor. In line with the reduction in GAG levels, HS and CS
sulfation were decreased by 53% and 83%, respectively
(Figure 4C,D). In parallel, we analyzed the disaccharide
compositions of HS and CS in the presence and absence of
UDP-4-XylAz. In agreement with its effects on total sulfa-
tion, the inhibitor did not appreciably alter the HS or CS
disaccharide composition (Figure 4E,F).
Finally, we examined the morphological effects of GAG
disruption by UDP-4-XylAz injection. Zebrafish embryos
treated with the metabolic inhibitor displayed an observable
phenotype that is subtle at early stages (18 hpf, data not
shown) and becomes more prominent as the embryo develops
(Figure 5). UDP-4-XylAz-treated fish had a shortened axis,
disorganized brain, small eyes, notochord defects, and abnor-
mally shaped myotomes by 24 hpf, as well as prominent tail
blistering by 48 hpf (Figure 5A–D). These phenotypes are
reminiscent of those observed for zebrafish with GAG or
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
proteoglycan defects induced by genetic methods,[7,35] and of
zebrafish laminin (lam) mutants, particularly lama1/bashful,
lamb1a/grumpy, and lamc1/sleepy (sly).[36–39] By 6 dpf, cranio-
facial abnormalities were apparent in live UDP-4-XylAz-
treated fish. At this stage, the craniofacial skeleton is
comprised of a simple pattern of cartilage and bone elements;
thus, we utilized stains for cartilage and bone (alcian and
alizarin, respectfully) to better understand the impact of
UDP-4-XylAz on these structures. In inhibitor-treated
embryos, alcian staining of cartilaginous elements revealed
defects in the neurocranium, including a narrow ethmoid
plate and less cohesive midline cartilage cells, while alizarin
staining of developing bone tissues revealed a lack of
notochord differentiation (Figure 5E,F). A similar ethmoid
plate phenotype is observed in embryos with deficiencies in
Hedgehog signaling, defects in cytokine signaling, or excess
retinoic acid signaling,[40–42] yet the alcian stains are most
strikingly similar to those of lamb1ab1166 mutants (Figure 5,
compare E and F with G and H), especially in light of other
phenotypic similarities between inhibitor-treated and laminin
mutant embryos. Another craniofacial element defect in
UDP-4-XylAz-injected embryos includes a narrowing of the
lower jaw elements (Figure 5I,J); otherwise visceral skeletal
elements, including the pharyngeal arches (not shown) are
largely normal (Figure 5K,L). During early development,
Laminin and proteoglycan interactions are critical for base-
ment membrane integrity and interactions with growth
factors; the phenotypic similarities of UDP-4-XylAz-injected
and lamb1ab1166 mutant embryos suggest that the inhibitor
may interfere with these interactions.[43]
[3] H. E. Bꢀlow, O. Hobert, Annu. Rev. Cell Dev. Biol. 2006, 22,
375 – 407.
[4] X. Lin, G. Wei, Z. Shi, L. Dryer, J. D. Esko, D. E. Wells, M. M.
Matzuk, Dev. Biol. 2000, 224, 299 – 311.
[5] K. J. Mitchell, K. I. Pinson, O. G. Kelly, J. Brennan, L. Zupicich,
P. Scherz, P. A. Leighton, L. V. Goodrich, X. W. Lu, B. J. Avery,
P. Tate, K. Dill, E. Pangilinan, P. Wakenight, M. Tessier-Lavigne,
W. C. Skarnes, Nat. Genet. 2001, 28, 241 – 249.
[6] D. Stickens, B. M. Zak, N. Rougier, J. D. Esko, Z. Werb,
Development 2005, 132, 5055 – 5068.
[7] K. Holmborn, J. Habicher, Z. Kasza, A. S. Eriksson, B. Filipek-
Gorniok, S. Gopal, J. R. Couchman, P. E. Ahlberg, M. Wiweger,
D. Spillmann, J. Kreuger, J. Ledin, J. Biol. Chem. 2012, 287,
33905 – 33916.
[8] C. McCormick, G. Duncan, K. T. Goutsos, F. Tufaro, Proc. Natl.
Acad. Sci. USA 2000, 97, 668 – 673.
[9] C. Senay, T. Lind, K. Muguruma, Y. Tone, H. Kitagawa, K.
Sugahara, K. Lidholt, U. Lindahl, M. Kusche-Gullberg, EMBO
Rep. 2000, 1, 282 – 286.
[10] M. A. Pinhal, B. Smith, S. Olson, J. Aikawa, K. Kimata, J. D.
Esko, Proc. Natl. Acad. Sci. USA 2001, 98, 12984 – 12989.
[11] A. C. Dar, K. M. Shokat, Annu. Rev. Biochem. 2011, 80, 769 –
795.
[12] M. Okayama, K. Kimata, S. Suzuki, J. Biochem. 1973, 74, 1069 –
1073.
[13] N. B. Schwartz, L. Galligani, P. L. Ho, A. Dorfman, Proc. Natl.
Acad. Sci. USA 1974, 71, 4047 – 4051.
[14] I. B. H. Wilson, Cell. Mol. Life Sci. 2004, 61, 794 – 809.
[15] R. K. Margolis, B. Goossen, H. Tekotte, L. Hilgenberg, R. U.
Margolis, J. Cell Sci. 1991, 99, 237 – 246.
[16] D. R. Garud, V. M. Tran, X. V. Victor, M. Koketsu, B. Kuberan,
J. Biol. Chem. 2008, 283, 28881 – 28887.
[17] Y. Tsuzuki, T. K. N. Nguyen, D. R. Garud, B. Kuberan, M.
Koketsu, Bioorg. Med. Chem. Lett. 2010, 20, 7269 – 7273.
[18] S. S. Thomas, J. Plenkiewicz, E. R. Ison, M. Bols, W. Zou, W. A.
Szarek, R. Kisilevsky, Biochim. Biophys. Acta Mol. Basis Dis.
1995, 1272, 37 – 48.
[19] R. L. Hull, S. Zraika, J. Udayasankar, R. Kisilevsky, W. A.
Szarek, T. N. Wight, S. E. Kahn, Am. J. Physiol. Cell Physiol.
2007, 293, C1586 – 1593.
[20] J. Nigro, A. Wang, D. Mukhopadhyay, M. Lauer, R. J. Midura, R.
Sackstein, V. C. Hascall, J. Biol. Chem. 2009, 284, 16832 – 16839.
[21] X. M. Van Wijk, V. L. Thijssen, R. Lawrence, S. A. van den
Broek, M. Dona, N. Naidu, A. Oosterhof, E. M. van de Westerlo,
L. J. Kusters, Y. Khaled, T. A. Jokela, P. Nowak-Sliwinska, H.
Kremer, S. E. Stringer, A. W. Griffioen, E. van Wijk, F. L.
van Delft, T. H. van Kuppevelt, ACS Chem. Biol. 2013, 8,
2331 – 2338.
In conclusion, UDP-4-XylAz is an effective chain-termi-
nating metabolic inhibitor of GAG biosynthesis in zebrafish.
The compound has a unique attribute in that its organismal
sites of inhibition can be visualized in vivo through bio-
orthogonal reaction with fluorescent cyclooctynes. Compar-
ison of GAGs from UDP-4-XylAz and untreated embryos
revealed differences in GAG abundance that likely cause the
specific embryological defects observed. This metabolic
inhibitor should aid in elucidating novel roles of GAG
chains during vertebrate development.
Experimental Section
[22] C. J. Dimitroff, R. J. Bernacki, R. Sackstein, Blood 2003, 101,
602 – 610.
[23] H. Nishimura, S. Kawabata, W. Kisiel, S. Hase, T. Ikenaka, T.
Takao, Y. Shimonishi, S. Iwanaga, J. Biol. Chem. 1989, 264,
20320 – 20325.
General procedure for visualizing GAG inhibition in vivo: At the 1–4
cell stage, zebrafish embryos were injected with 50 pmol of UDP-4-
XylAz. After developing to 24 hpf, zebrafish were incubated in
a 200 mm solution of DIFO-488 for 1 h at 28.58C, washed, and imaged
by confocal microscopy.
[24] K.-i. Inamori, T. Yoshida-Moriguchi, Y. Hara, M. E. Anderson,
L. Yu, K. P. Campbell, Science 2012, 335, 93 – 96.
[25] H. Bakker, T. Oka, A. Ashikov, A. Yadav, M. Berger, N. A.
Rana, X. Bai, Y. Jigami, R. S. Haltiwanger, J. D. Esko, R.
Gerardy-Schahn, J. Biol. Chem. 2009, 284, 2576 – 2583.
[26] J. M. Baskin, K. W. Dehnert, S. T. Laughlin, S. L. Amacher, C. R.
Bertozzi, Proc. Natl. Acad. Sci. USA 2010, 107, 10360 – 10365.
[27] K. W. Dehnert, B. J. Beahm, T. T. Huynh, J. M. Baskin, S. T.
Laughlin, W. Wang, P. Wu, S. L. Amacher, C. R. Bertozzi, ACS
Chem. Biol. 2011, 6, 547 – 552.
Received: December 5, 2013
Published online: && &&, &&&&
Keywords: biosynthesis · click chemistry · glycosaminoglycans ·
.
inhibitors · zebrafish
[1] K. Nybakken, N. Perrimon, Biochim. Biophys. Acta Gen. Subj.
[28] D. Soriano del Amo, W. Wang, H. Jiang, C. Besanceney, A. C.
Yan, M. Levy, Y. Liu, F. L. Marlow, P. Wu, J. Am. Chem. Soc.
2010, 132, 16893 – 16899.
2002, 1573, 280 – 291.
[2] E. Hohenester, P. D. Yurchenco, Cell Adhes. Migr. 2013, 7, 56 –
63.
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
[29] S. T. Laughlin, J. M. Baskin, S. L. Amacher, C. R. Bertozzi,
Science 2008, 320, 664 – 667.
D. L. Stemple, Z. Rangini, H. Yang, W. Driever, Development
1996, 123, 165 – 178.
[30] J. F. Batey, C. Bullock, E. OꢁBrien, J. M. Williams, Carbohydr.
Res. 1975, 43, 43 – 50.
[31] H. C. Hang, C. Yu, M. R. Pratt, C. R. Bertozzi, J. Am. Chem.
Soc. 2004, 126, 6 – 7.
[32] J. M. Baskin, J. A. Prescher, S. T. Laughlin, N. J. Agard, P. V.
Chang, I. A. Miller, A. Lo, J. A. Codelli, C. R. Bertozzi, Proc.
Natl. Acad. Sci. USA 2007, 104, 16793 – 16797.
[33] M. Mueller, Ph.D. thesis, The University of Chicago, 2009.
[34] Y. Nakamura, N. Haines, J. Chen, T. Okajima, K. Furukawa, T.
Urano, P. Stanley, K. D. Irvine, K. Furukawa, J. Biol. Chem.
2002, 277, 46280 – 46288.
[37] M. J. Parsons, S. M. Pollard, L. Safflde, B. Feldman, P. Coutinho,
E. M. A. Hirst, D. L. Stemple, Development 2002, 129, 3137 –
3146.
[38] J. D. Paulus, M. C. Halloran, Dev. Dyn. 2006, 235, 213 – 224.
[39] M. Dolez, J. F. Nicolas, E. Hirsinger, Development 2011, 138, 97 –
106.
[40] N. Wada, Y. Javidan, S. Nelson, T. J. Carney, R. N. Kelsh, T. F.
Schilling, Development 2005, 132, 3977 – 3988.
[41] K. Laue, M. Jänicke, N. Plaster, C. Sonntag, M. Hammerschmidt,
Development 2008, 135, 3775 – 3787.
[42] E. C. O. Killian, D. A. Birkholz, K. B. Artinger, Dev. Biol. 2009,
333, 161 – 172.
[35] J. Topczewski, D. S. Sepich, D. C. Myers, C. Walker, A. Amores,
Z. Lele, M. Hammerschmidt, J. Postlethwait, L. Solnica-Krezel,
Dev. Cell 2001, 1, 251 – 264.
[43] G. E. Davis, F. G. Klier, E. Engvall, C. Cornbrooks, S. Varon, M.
Manthorpe, Neurochem. Res. 1987, 12, 909 – 921.
[36] A. F. Schier, S. C. F. Neuhauss, M. Harvey, J. Malicki, L.
SolnicaKrezel, D. Y. R. Stainier, F. Zwartkruis, S. Abdelilah,
6
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
These are not the final page numbers!

Angewandte
Chemie
Communications
Glycosylation Inhibitor
Visualizing inhibition: Metabolic incor-
poration of an azide modified xylose
residue inhibits elaboration of the glycan.
Further, the azide group enables rapid
visualization of the sites of inhibition in
vivo during zebrafish development using
Cu-free click chemistry.
B. J. Beahm, K. W. Dehnert, N. L. Derr,
J. Kuhn, J. K. Eberhart, D. Spillmann,
S. L. Amacher,
C. R. Bertozzi*
&&&& — &&&&
A Visualizable Chain-Terminating
Inhibitor of Glycosaminoglycan
Biosynthesis in Developing Zebrafish
Angew. Chem. Int. Ed. 2014, 53, 1 – 7
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7
These are not the final page numbers!