Synthesis of a targeted library of heparan sulfate hexa- to dodecasaccharides as inhibitors of β-secretase: Potential therapeutics for Alzheimer's disease

Article abstract of DOI:10.1002/chem.201204519

Heparan sulfates (HS) are a class of sulfated polysaccharides that function as dynamic biological regulators of the functions of diverse proteins. The structural basis of these interactions, however, remains elusive, and chemical synthesis of defined structures represents a challenging but powerful approach for unravelling the structure-activity relationships of their complex sulfation patterns. HS has been shown to function as an inhibitor of the β-site cleaving enzyme β-secretase (BACE1), a protease responsible for generating the toxic Aβ peptides that accumulate in Alzheimer's disease (AD), with 6-O-sulfation identified as a key requirement. Here, we demonstrate a novel generic synthetic approach to HS oligosaccharides applied to production of a library of 16 hexa- to dodecasaccharides targeted at BACE1 inhibition. Screening of this library provided new insights into structure-activity relationships for optimal BACE1 inhibition, and yielded a number of potent non-anticoagulant BACE1 inhibitors with potential for development as leads for treatment of AD through lowering of Aβ peptide levels. Copyright

Full text of DOI:10.1002/chem.201204519

FULL PAPER
DOI: 10.1002/chem.201204519
Synthesis of a Targeted Library of Heparan Sulfate Hexa- to
Dodecasaccharides as Inhibitors of b-Secretase: Potential Therapeutics
for Alzheimerꢀs Disease
Ralf Schwçrer,[a] Olga V. Zubkova,[a] Jeremy E. Turnbull,*[b] and Peter C. Tyler*[a]
Abstract: Heparan sulfates (HS) are a
class of sulfated polysaccharides that
function as dynamic biological regula-
tors of the functions of diverse pro-
teins. The structural basis of these in-
teractions, however, remains elusive,
and chemical synthesis of defined
structures represents a challenging but
powerful approach for unravelling the
structure–activity relationships of their
complex sulfation patterns. HS has
been shown to function as an inhibitor
of the b-site cleaving enzyme b-secre-
tase (BACE1), a protease responsible
for generating the toxic Ab peptides
that accumulate in Alzheimerꢀs disease
(AD), with 6-O-sulfation identified as
a key requirement. Here, we demon-
strate a novel generic synthetic ap-
proach to HS oligosaccharides applied
to production of a library of 16 hexa-
to dodecasaccharides targeted at
BACE1 inhibition. Screening of this li-
brary provided new insights into struc-
ture–activity relationships for optimal
BACE1 inhibition, and yielded
a
number of potent non-anticoagulant
BACE1 inhibitors with potential for
development as leads for treatment of
AD through lowering of Ab peptide
levels.
Keywords: Alzheimerꢀs · carbohy-
drates · glycosylation · heparan sul-
fate · synthesis
Introduction
focus of a number of programs searching for therapeutics to
alleviate the symptoms of Alzheimerꢀs disease (AD).[7] This
neurodegenerative disease was first described by Alois Alz-
heimer in 1906. A key characteristic is the deposition of in-
soluble accumulations of the amyloid b-peptide in the brain.
The amyloid b-peptide is itself derived from step-wise ex-
tracellular enzymatic cleavage of the amyloid precursor pro-
tein by BACE1, and then intramembrane cleavage by g-sec-
retase. The accumulation of amyloid b-peptide is a critical
driving force for Alzheimerꢀs disease pathology according to
the amyloid cascade hypothesis.[8] There is also evidence
that amyloid b-peptide is involved in hyperphosphorylation
of the tau protein and subsequent formation of neurofibril-
lary tangles.[9]
Although the inhibition of BACE1 by HS and heparin is
interesting, heparin, which is a highly sulfated variant of HS
and a widely used anticoagulant drug, would have serious
blood-thinning side effects that would likely preclude its use
for the treatment of AD. However, the selective chemical
modification of heparin has given rise to some homogeneous
HS polymers with potent BACE1 inhibitory activity and no
significant anticoagulant properties,[6,10] whereas dermatan
sulfate had no effect.[6] In addition some discrete synthetic
HS tetrasaccharides have shown surprising (though still
moderate) potency as BACE1 inhibitors.[11] This suggests
that larger synthetic oligosaccharides might have applica-
tions as high potency BACE1 inhibitors with potential as
novel therapeutics for AD.
Heparan sulfate (HS) is a linear polysaccharide with highly
diverse functionality. Through attachment to core proteins it
is present as a glycoprotein (“proteoglycan”) on mammalian
cell surfaces and in the extracellular matrix. HS has a disac-
charide repeating unit of d-glucosamine (Glc) and d-glucur-
onic acid (GlcA) or l-iduronic acid (IdoA), which may be
variously O- or N-sulfated or N-acetylated, generating poly-
anionic structural diversity. HS regulates many important bi-
ological processes;[1] significant examples include growth
factors and their receptors,[2] cytokines,[3] stem cells,[4] Wnt
and bone morphogenetic proteins.[5] The diversity of cell sig-
nalling events influenced by HS is presumed to be a func-
tion of the microheterogeneity of its structure with specific
sequences and/or sulfation patterns required for particular
protein binding and biological activities.[1b]
In addition HS was identified as the first natural regulator
of the cleavage of the amyloid precursor protein by b-secre-
tase.[6] This cleavage is a critical step in the formation of
amyloid plaques present in the brains of Alzheimerꢀs pa-
tients and the enzyme b-secretase (BACE1) has been the
[a] Dr. R. Schwçrer, Dr. O. V. Zubkova, Dr. P. C. Tyler
Carbohydrate Chemistry, Industrial Research, Ltd.
P.O. Box 31310, Lower Hutt (New Zealand)
E-mail: p.tyler@irl.cri.nz
[b] Prof. J. E. Turnbull
Institute of Integrative Biology
University of Liverpool, Liverpool L69 3BX (UK)
E-mail: J.Turnbull@liverpool.ac.uk
The ability to synthesise discrete HS oligomers of octasac-
charide size and larger would offer potential therapeutics
for Alzheimerꢀs as well as other diseases in which HS–pro-
tein interactions play a significant role. Importantly the syn-
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201204519.
Chem. Eur. J. 2013, 19, 6817 – 6823
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6817

thesis of such oligosaccharides would also facilitate the
wider elucidation of specific HS–protein interactions—a key
bottleneck in this field. For these reasons we decided to de-
velop generic approaches for the synthesis of any desired
HS oligosaccharide. In the first instance we have focused on
the synthesis of a targeted library of potential BACE1 inhib-
itors.
Studies of chemically modified heparins have indicated
that having 6-O-sulfate substituents on the glucosamine resi-
dues is important for potency as BACE1 inhibition, whereas
N-acetate substituents are tolerated, and an additional 2-O-
sulfate on the uronic acids offers some improved activity.[6,10]
The synthesis and testing of discrete sulfated oligosacchar-
ides would offer confirmation of these results and potential-
ly provide lead single chemical entity drug compounds. We
decided to make targets both with and without the uronic
acid 2-O-sulfates, as reducing sulfation levels is likely to be
beneficial for bioavailability and reducing off-target effects.
In addition, we wished to explore the relative importance of
glucuronic and iduronic acid residues, which has not been
elucidated for this protein. Our initial synthetic targets then
became hexa-, octa-, deca- and dodecasaccharides contain-
ing GlcNAc6S-UA or GlcNAc6S-UA2S disaccharides with
either d-glucuronic or l-iduronic acids (Figure 1).
Figure 2. Possible disaccharide donors.
of the uronic acid. However, uronic acids make relatively
poor donors and can be unreliable when assembling large
oligosaccharide chains. Donor 2 has the advantage that the
uronic acid glycosidic linkage is established, but it relies on
achieving very good a-selectivity in glycosylations with this
donor. In our hands, and those of others,[12o] the use of such
2-azido-glycosyl donors cannot be guaranteed to result in
high anomeric selectivity. Subtle changes in protecting
groups will sometimes exert significant effects on anomeric
selectivity and we did not want to have to separate anomeric
mixtures in a large oligosaccharide chain. Consequently, we
have settled on the use of donors represented by 3, which
has the advantages of 1 but uses the much more reactive d-
glucose or l-idose moieties as donors, which can be oxidised
later to the uronic acids. Since we began this work others
have reported similar strategies although with significant
differences in donor activation and protecting group
choices.[12a,j,l,p]
Figure 1. Synthetic targets.
Results and Discussion
For the purposes of synthesising the BACE1 inhibitor tar-
gets outlined above, after considerable experimentation we
settled on the protecting group combination in the disac-
charides 4 and 5, which offered benefits in ease of synthesis
and the required selectivity. There are four orthogonal ester/
carbonate protecting groups that can be selectively removed
as required (vide infra). The building blocks 4 can serve as
acceptors by removal of the Fmoc groups or be converted
into donors by hydrolysis of the methoxyphenyl glycosides
and then formation of the corresponding trichloroacetimi-
dates.
Selection of building blocks for synthesis: A number of
groups have reported methods for the synthesis of HS oligo-
saccharides.[11,12] Some are elegant without any being com-
pletely general and/or suitable for scale-up. Recently, a
novel chemoenzymatic approach has used a number of HS
biosynthetic enzymes to assemble some defined low-molecu-
lar-weight (LMW) heparin oligosaccharides including a hep-
tasaccharide analogue of the heparin pentasaccharide drug,
Arixtra.[13] Although this strategy clearly has a lot of prom-
ise, controlling enzyme specificity to deliver pure, as well as
non-natural products has yet to be overcome. All fully syn-
thetic approaches so far have used disaccharide building
blocks as the basis for assembling their targets. The disac-
charide donors can be oriented with either the uronic acid
or the glucosamine moieties at the reducing end, represent-
ed by the general structures 1 and 2, respectively (Figure 2).
With donor 1, an advantage is that the sometimes difficult-
to-establish a-stereochemistry of the glucosamine is already
in place. Also, the a-l- or b-d-stereochemistry of the uronic
acid can be achieved by using a participating group at O-2
Synthesis of disaccharide building blocks: Commercially
available methyl 2-azido-2-deoxy-1-thio-b-d-glucopyranoside
(6) was converted into the benzylidene derivative 7,[14] fol-
lowed by standard protecting group manipulation to give 8
and the alternative 4-O-benzyl derivative 9 was also pre-
pared as shown (Scheme 1). The d-gluco- and l-ido- build-
ing blocks were both prepared from 1,2:5,6-di-O-isopropyli-
dene-a-d-glucofuranose (Scheme 2), in which the latter was
converted into the methoxyphenyl glycoside 10[15] and then
the acceptor 11. Alternatively, the l-ido tetrabenzoate 12[16]
6818
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 6817 – 6823

Synthetic Heparan Sulfates in Drug Discovery
FULL PAPER
Scheme 1. Preparation of azide-glycosyl donors. a) TrCl, py; b) BnBr,
NaH; c) aq. HOAc, 808C; d) Ac2O, py; e) HCl, MeOH; f) AcCl, py,
À788C to room temperature; g) FmocCl, py.
Scheme 3. Preparation of disaccharides. a) NIS, AgOTf, DCM/toluene;
b) CeACHTUNRTGNE(UNG NH4)2NO3, aq. CH3CN; c) Cl3CCN, NaH, DCM; d) Et3N, DCM.
Scheme 2. Preparation of glycosyl acceptors. a) ClAcCl, py, À788C to
room temperature; b) p-methoxyphenol, BF3·EtO2; c) NaOMe; d) Me2C-
ACHTUNGTRENNUNG(OMe)2, TsOH; e) BzCl, py; f) aq. HOAc, 808C.
coupled with acceptors 11 and 14 in high yields, and a-selec-
tivity to give 15–18 (Scheme 3). Compounds 15–17 are crys-
talline, facilitating large-scale synthesis of these building
blocks.
was prepared and converted via the 4,6-O-isopropylidene
derivative 13 into the acceptor 14. The donors 8 and 9 were
Scheme 4. Synthesis of fully protected homogenous oligosaccharides. a) TMSOTf; b) Et3N; c) TMSOTf, 19, 20, 28 or 29; d) TMSOTf, 20, 28 or 29;
e) TMSOTf, 29. The descriptors ido/gluco are used to describe the configuration of the carbohydrate residues representing uronic acids in the target mol-
ecule, ordered from non-reducing end to reducing end (all gluco for a GlcNAc-GlcA-GlcNAc-GlcA-GlcNAc-GlcA hexasaccharide).
Chem. Eur. J. 2013, 19, 6817 – 6823
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6819

J. E. Turnbull, P. C. Tyler et al.
Oligosaccharide assembly: Initially homogeneous oligosac-
charides were synthesised, with either glucuronic or iduronic
acid precursor components. The disaccharide 15 was con-
verted into donor 19 by ceric ammonium nitrate promoted
hydrolysis of the methoxyphenyl glycoside followed by for-
mation of the trichloroacetimidate, and was separately con-
verted into an acceptor 21 by selective removal of the Fmoc
group with mild base (Scheme 3). Coupling of 19 and 21 af-
forded only the tetrasaccharide 22 in 87% yield (Scheme 4).
Removal of the Fmoc gave the tetrasaccharide acceptor 23,
which was glycosylated with donor 19 to give hexasacchar-
ide 24, and the process was repeated to give 25 and then oc-
tasaccharides 26 and 27. The glycosylation yields were uni-
formly high (>80%) and no by-products of a-anomer or or-
thoester formation were detected. Methoxyphenyl glycoside
protection at the reducing end was retained as a useful
NMR marker unlikely to influence the biological activity.
This procedure was also successful in the idose series so that
disaccharide 17 afforded both the donor 28 and acceptor 30
(Scheme 3). Sequential glycosylations again afforded excel-
lent yields (>80% with no b-l anomers or orthoesters de-
tected). In this way the oligosaccharide acceptors 32, 34, 36
and 38 were prepared. Then glycosylation of acceptors 23,
25 and 27 with acetimidate 20 and of acceptors 32, 34, 36,
and 38 with acetimidate 29 afforded hexa-, octa-, deca- and
dodecasaccharides 39–45 with excellent overall efficiency.
Some oligosaccharides with mixed glucuronic and iduronic
acid residues were also synthesised with a view to exploring
the structure–activity relationship (SAR) of these com-
pounds. Starting with the disaccharide acceptors 21 and 30
and the tetrasaccharide acceptor 32 appropriate sequential
glycosylation reactions were performed as before to afford
the mixed octasaccharides 50, 53, 58 and 61 (Scheme 5). The
glycosylation efficiencies with the particular set of protect-
ing groups used here were uniformly very high and this al-
lowed oligosaccharide assembly on a multigram scale with
relative ease.
Scheme 5. Synthesis of mixed octasaccharides. a) TMSOTf; b) Et3N;
c) TMSOTf, 19 or 28; d) TMSOTf, 20 or 29. The descriptors ido/gluco
are used to describe the configuration of the carbohydrate residues repre-
senting uronic acids in the target molecule, ordered from non-reducing
end to reducing end (ido-gluco for a GlcNAc-IdoA-GlcNAc-GlcA tetra-
saccharide).
Conversion to sulfated oligosaccharide targets: The fully
protected hexa-, octa-, deca- and dodecasaccharides 39–45,
50, 53, 58 and 61 have three types of orthogonal ester pro-
tecting groups and the azide moieties that require processing
to generate the target sulfated oligosaccharides. Firstly, the
chloroacetate groups were selectively removed with mild
base to expose from 3 to 6 primary hydroxyl groups without
detectable loss of other ester protecting groups. These alco-
hols were then oxidised to the corresponding carboxylic
acids in good yield using TEMPO/BAIB conditions in aque-
ous acetonitrile, and protection as the methyl esters was ef-
fected with diazomethane (or trimethylsilyldiazomethane).
The overall yields for multiple oxidation and ester formation
were not optimised but if sufficient time was allowed for
complete oxidation the yields were approximately 75%. The
azide groups were then converted directly to the N-acetates
with thioacetic acid in pyridine affording 76–82 (Scheme 6)
and 129–132 (Scheme 7). There remained the primary ace-
tate and secondary benzoate groups, which needed to be dif-
ferentiated to allow selective sulfation on the primary cen-
tres. Solvolysis of the acetates under mild base conditions
was not sufficiently selective as some removal of benzoates
occurred, especially for the longer oligosaccharides. Howev-
er, acidic solvolysis gave high yields of products resulting
from loss of only the acetates and without any sign of
anomerization of the glycoside linkages. Subsequent sulfa-
tion of the exposed hydroxyl groups afforded 90–96 and
137–140. Then saponification of both the benzoate and
methyl esters was effected and the products were fully de-
protected by hydrogenolysis to the target compounds 104–
110 and 145–148. It was necessary to add methanolic or
aqueous ammonia to the hydrogenolysis solvent to prevent
some loss of sulfate groups. Alternatively, the exposed sec-
ondary hydroxy groups after saponification in 97–103 were
also sulfated, and then global deprotection of the products
by hydrogenolysis gave the alternatively sulfated oligosac-
6820
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 6817 – 6823

Synthetic Heparan Sulfates in Drug Discovery
FULL PAPER
number of the octasaccharides
were surprisingly potent, with
similar activities to decasacchar-
ides of the same basic structure
(Table 1). The low potency hex-
asaccharides (104, 107) were
similar in activity to the most
active tetrasaccharides descri-
bed in previous studies,[11] and
were approximately 15- to 20-
fold less potent than equivalent
octasaccharides (105, 108).
Regarding the relative contri-
bution of IdoA and GlcA resi-
dues, for the homogenous re-
peating unit structures in the
absence of 2-O-sulfation, there
were small differences in poten-
cy (maximum two- to three-
fold), suggesting that GlcA may
confer a marginal improvement
in activity over IdoA, at least
for the hexa- and octasacchar-
ides. The IC50 of the non-regu-
lar octasaccharide compounds
with mixed uronate residues
was similar (~100–300 nm) to
those containing solely GlcA
(260 nm), and only slightly
more active than those contain-
ing solely IdoA (~900 nm).
However, of much greater sig-
nificance was the evidence for
dramatically higher activity
conferred by the additional
presence of a 2-O-sulfate on
the UA residues, irrespective of
the GlcA or IdoA configura-
Scheme 6. Synthesis of final hexa-, octa-, deca- and dodecasaccharides. a) DABCO, MeCN/EtOH, 708C;
b) TEMPO, BAIB, aq. MeCN, then CH2N2; c) py, HSAc; d) HCl, MeOH/DCM; e) SO3·NMe3, DMF, 608C;
f) NaOH, aq. MeOH; g) H2, Pd(OH)2/C, aq. THF. The descriptors ido/gluco are used to describe the configu-
ration of the carbohydrate residues representing uronic acids in the target molecule (n=3, ido for a GlcNAc-
IdoA-GlcNAc-IdoA-GlcNAc-IdoA-GlcNAc-IdoA octasaccharide).
charide targets 116–120. We found the two-step sulfation
procedure to work more efficiently than removing both the
acetate and benzoate groups by saponification followed by
sulfation.
tion. The differences were approximately 20-fold for dodec-
asaccharides (120 vs. 110), about 70- to 90-fold for decasac-
charides (117 and 119 vs. 106 and 109) and approximately
55- to 230-fold for octasaccharides (116 and 118 vs. 105 and
108). Such large differences were unexpected since previous
data indicated only a threefold difference in activity for hep-
arin polysaccharides lacking 2-O-sulfates.[10]
Overall, these data support the view that octa- (116 and
118) and decasaccharides (117 and 119) containing either
GlcNAc6S-GlcA2S or GlcNAc6S-IdoA2S units represent
realistic lead compounds for the development of fully syn-
thetic BACE1 inhibitors, with a requisite balance of activity,
size and relative ease of synthesis. Only modest gains in po-
tency were evident for deca- and dodeca- compared to octa-
saccharides (~two- to threefold), which are unlikely to justi-
fy the additional complexity and cost of their synthesis.
To assess the potential off-target anticoagulant activity of
the saccharides, they were all tested in the dose range 0.004
to 50 mgmLÀ1 in a Factor Xa assay employing cleavage of a
Compounds potently inhibit BACE1 protease and lack anti-
coagulant activity: The ability of the target compounds to
inhibit the BACE1 protease was next investigated using flu-
orescence resonance energy transfer (FRET) peptide cleav-
age assays (see the Supporting Information). All of the com-
pounds inhibited BACE1, with IC50 values varying from low
potency (5–12 mm) for the hexasacharides to high potency
(1–5 nm) for a number of the octa-, deca- and dodecasac-
charides (Table 1). Their potency clearly increased with in-
creasing size, within classes of compounds with similar re-
peating backbones; this is in agreement with previous data
from heparin size fractions.[10b] However, the previous data
indicated a significant reduction in activity (~tenfold) below
the decasaccharide size, whereas here we found that a
Chem. Eur. J. 2013, 19, 6817 – 6823
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6821

J. E. Turnbull, P. C. Tyler et al.
Scheme 7. Synthesis of final mixed octasaccharides. a) DABCO, MeCN/EtOH, 708C; b) TEMPO, BAIB, aq. MeCN, then CH2N2; c) py, HSAc; d) HCl,
MeOH/DCM; e) SO3·NMe3, DMF, 608C; f) NaOH, aq. MeOH; g) H2, Pd(OH)2/C, aq. THF. The descriptors ido/gluco are used to describe the configura-
tion of the carbohydrate residues representing uronic acids in the target molecule (gluco-ido-ido-gluco for a GlcNAc-GlcA-GlcNAc-IdoA-GlcNAc-
IdoA-GlcNAc-GlcA octasaccharide).
peptide substrate (see the Supporting Information). None
displayed any measurable ability to accelerate antithrombin-
III mediated inactivation of Factor Xa, consistent with our
previous data which indicated that selectively desulfated
heparins with equivalent repeating structures had anticoagu-
lant activity >1000-fold lower than parental heparin.[10b]
Unlike heparin, such synthetic compounds would thus be
expected to have no significant side effects related to antico-
agulation. Furthermore, with regard to potential off-target
inhibition of other structurally related and physiologically
relevant aspartic proteases, we have previously shown that
similar compounds (NAcLMWH and analogues) have mark-
edly reduced potency for enzymes, such as renin and pepsin
(>100-fold) and cathepsin D (~10–100-fold).[10b] This selec-
tivity is encouraging with regard to the potential therapeutic
applications of these compounds as BACE1 inhibitors.
moieties, targeted at optimised inhibition of BACE1 with re-
duced off-target activities.[10] The majority were regular, ho-
mogenous repeat compounds, but the synthetic strategy also
afforded irregular structures with mixed sequences contain-
ing either GlcA or IdoA residues; altered patterns of other
O-sulfate or indeed N-sulfate groups could easily be intro-
duced with usage of appropriate disaccharide building
blocks. Thus, the approach has generic potential to produce
any desired HS structure at scale for future chemical biology
studies. Here, we have identified surprisingly potent, fully
synthetic inhibitors as small as octasaccharides; this finding
was unexpected based on previous studies with heterogene-
ous structures available by semisynthetic chemical desulfa-
tion of heparin.[10b] These compounds have significant poten-
tial as leads for the development of new drugs that inhibit
BACE1, particularly as LMW heparin oligosaccharides have
been demonstrated to cross the blood–brain barrier,[17] and
oral availability is a tractable route for such saccharides
using appropriate formulation and encapsulation strat-
egies.[18] Successful in vivo delivery and BACE1 inhibition
by these saccharides could lower brain Ab levels and poten-
tially provide a disease-modifying treatment for Alzheimerꢀs
disease.
Conclusion
We have described a successful synthetic approach that ena-
bles the general synthesis of specific HS oligosaccharides of
a range of sizes and structures. The building blocks we de-
veloped have allowed the efficient assembly of a range of
hexa- to dodecasaccharides. The targeted library of 16 sul-
fated oligosaccharides presented here is to our knowledge
the largest of its type reported to date. We have focused on
relatively low-sulfated saccharides containing NAc and 6S
6822
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 6817 – 6823

Synthetic Heparan Sulfates in Drug Discovery
FULL PAPER
Table 1. Inhibitory activities of synthetic compounds and heparin for cleavage of
FRET peptide by BACE1 enzyme.
[7] a) R. Silvestri, Med. Res. Rev. 2009, 29, 295–338; b) V.
John, J. P. Beck, M. J. Bienkowski, S. Sinha, R. L. Heinrik-
son, J. Med. Chem. 2003, 46, 4625–4630; c) S. L. Cole, R.
Vassar, J. Biol. Chem. 2008, 283, 29621–29625.
Compound
IC50
Structure[b]
MW
[nm][a]
[8] J. Hardy, D. J. Selkoe, Science 2002, 297, 353–356.
[9] L. M. Ittner, J. Gçtz, Nat. Rev. Neurosci. 2011, 12, 65.
[10] a) S. J. Patey, E. A. Edwards, E. A. Yates, J. E. Turnbull,
Neurodegener. Dis. 2008, 5, 197–199; b) S. J. Patey, E. A.
Edwards, E. A. Yates, J. E. Turnbull, J. Med. Chem. 2006,
49, 6129–6132.
[11] S. Arungundram, K. Al Mafraji, J. Asong, F. E. Leach,
I. J. Amster, A. Venot, J. E. Turnbull, G. J. Boons, J. Am.
Chem. Soc. 2009, 131, 17394–17405.
[c]
heparin
NAcLMWH
0.2 [IdoA2S-GlcNS6S]n
12000[d]
4000[e]
[c]
1.8 [IdoA2S-GlcNAc6S]n
hexasaccharides
104
107
5300
13000
GlcNAc6S-[GlcA-GlcNAc6S]2-GlcA
GlcNAc6S-[IdoA-GlcNAc6S]2-IdoA
1634
1634
octasaccharides
105
108
116
118
145
260
890
GlcNAc6S-[GlcA-GlcNAc6S]3-GlcA
GlcNAc6S-[IdoA-GlcNAc6S]3-IdoA
4.7 GlcNAc6S-[GlcA2S-GlcNAc6S]3-GlcA2S
3.9 GlcNAc6S-[IdoA2S-GlcNAc6S]3-IdoA2S
2138
2138
2546
2546
2138
[12] a) Y. P. Hu, S. Y. Lin, C. Y. Huang, M. M. Zulueta, J. Y.
Liu, W. Chang, S. C. Hung, Nat. Chem. 2011, 3, 557–563;
b) F. Baleux, L. Loureiro-Morais, Y. Hersant, P. Clayette,
F. Arenzana-Seisdedos, D. Bonnaffꢂ, H. A. Lortat-Jacob,
Nat. Chem. Biol. 2009, 5, 743–748; c) C. L. Cole, S. U.
Hansen, M. Barꢃth, G. Rushton, J. M. Gardiner, E. Avi-
zienyte, G. C. Jayson, PLoS ONE 2010, 5, e11644; d) C.
Noti, P. H. Seeberger, Chem. Biol. 2005, 12, 731–756;
e) M. Petitou, A. Imberty, P. Duchaussoy, P. A. Driguez,
M. L. Ceccato, F. Gourvenec, P. Sizun, J. P. Hꢂrault, S.
Pꢂrez, J. M. Herbert, Chem. Eur. J. 2001, 7, 858–873;
f) A. Lubineau, H. Lortat-Jacob, O. Gavard, S. Sarrazin,
D. Bonnaffꢂ, Chem. Eur. J. 2004, 10, 4265–4282; g) J. L.
de Paz, M. Martꢄn-Lomas, Eur. J. Org. Chem. 2005, 1849–
1858; h) Y. Zhou, F. Lin, J. Chen, B. Yu, Carbohydr. Res.
2006, 341, 1619–1629; i) C. Chen, B. Yu, Bioorg. Med.
Chem. Lett. 2009, 19, 3875–3879; j) J. C. Lee, X. A. Lu,
S. S. Kulkarni, Y. S. Wen, S. C. Hung, J. Am. Chem. Soc.
2004, 126, 476–477; k) T. Polat, C. H. Wong, J. Am.
Chem. Soc. 2007, 129, 12795–12800; l) L. D. Lu, C. R.
Shie, S. S. Kulkarni, G. R. Pan, X. A. Lu, S. C. Hung, Org.
Lett. 2006, 8, 5995–5998; m) S. Sarrazin, D. Bonnaffꢂ, A.
Lubineau, H. Lortat-Jacob, J. Biol. Chem. 2005, 280,
37558–37564; n) H. A. Orgueira, A. Bartolozzi, P. Schell,
R. E. J. N. Litjens, E. R. Palmacci, P. H. Seeberger, Chem.
Eur. J. 2003, 9, 140–169; o) G. J. S. Lohman, P. H. See-
berger, J. Org. Chem. 2004, 69, 4081–4093; p) S. C. Hung,
X. A. Lu, J. C. Lee, M. D.-T. Chang, S. I. Fang, T. C. Fan,
M. M. Zulueta, Y. Q. Zhong, Org. Biomol. Chem. 2012,
10, 760–772.
110
GlcNAc6S-GlcA-GlcNAc6S-IdoA-GlcNAc6S-
IdoA-GlcNAc6S-GlcA
GlcNAc6S-IdoA-GlcNAc6S-GlcA-GlcNAc6S-
IdoA-GlcNAc6S-GlcA
GlcNAc6S-GlcA-GlcNAc6S-IdoA-GlcNAc6S-
GlcA-GlcNAc6S-IdoA
GlcNAc6S-GlcA-GlcNAc6S-GlcA-GlcNAc6S-
IdoA-GlcNAc6S-IdoA
146
147
148
310
280
230
2138
2138
2138
decasaccharides
106
109
117
119
250
200
GlcNAc6S-[GlcA-GlcNAc6S]4-GlcA
GlcNAc6S-[IdoA-GlcNAc6S]4-IdoA
3.5 GlcNAc6S-[GlcA2S-GlcNAc6S]4-GlcA2S
2.2 GlcNAc6S-[IdoA2S-GlcNAc6S]4-IdoA2S
2639
2639
3149
3149
dodecasaccharides
110
120
32
GlcNAc6S-[IdoA-GlcNAc6S]5-IdoA
1.3 GlcNAc6S-[IdoA2S-GlcNAc6S]5-IdoA2S
3142
3754
[a] For comparative IC50 values expressed by weight [mgmLÀ1] and corresponding
standard deviation values, see the Supporting Information. [b] GlcNAc, N-acetylated
glucosamine; GlcA, glucuronic acid; IdoA, iduronic acid; 6S, 6-O-sulfate; 2S, 2-O-sul-
fate. NS, N-sulfate. The numbered synthetic compounds have a reducing end methoxy-
phenyl group (data not shown). [c] Predominant disaccharide repeating unit. [d] Hepa-
rin, average MW 12 kDa. [e] N-Acetylated low-molecular-weight heparin.
[13] a) Y. Xu, S. Masuko, M. Takieddin, H. Xu, R. Liu, J. Jing,
S. A. Mousa, R. J. Linhardt, J. Liu, Science 2011, 334, 498–501; b) R.
Liu, Y. Xu, M. Chen, M. Weꢅwer, X. Zhou, A. S. Bridges, P. L.
DeAngelis, Q. Zhang, R. J. Linhardt, J. Liu, J. Biol. Chem. 2010,
285, 34240–34249.
Experimental Section
Full experimental details including characterisation and NMR spectra of
new compounds are given in the Supporting Information.
[14] P. Rajaratnam, P. Gupta, P. Katavic, K. Kuipers, N. Huyh, S. Ryan,
T. Falzun, G. B. Tometzki, L. Bornaghi, G. Le Thanh, G. Abbenante,
L. Liu, W. Meutermans, N. Wimmer, M. L. West, Aust. J. Chem.
2010, 63, 693–699.
Acknowledgements
[15] N. Karst, J. C. Jacquinet, Eur. J. Org. Chem. 2002, 815–825.
[16] N. Barroca, J. C. Jacquinet, Carbohydr. Res. 2000, 329, 667–679.
[17] a) B. Leveugle, W. Ding, F. Laurence, M. P. Dehouck, A. Scanameo,
R. Cecchelli, H. Fillit, J. Neurochem. 2002, 70, 736–744; b) Q. Ma,
B. Dudas, M. Hejna, U. Cornelli, J. M. Lee, S. Lorens, R. Mervis, I.
Hanin, J. Fareed, Thromb. Res. 2002, 105, 447–453.
Drs. Herbert Wong and Yinrong Lu are thanked for an exceptional
NMR and mass spectroscopy service. J.E.T. gratefully acknowledges the
financial support of the Biotechnology and Biological Sciences Research
Council UK and the Medical Research Council UK, and Elizabeth Ed-
wards for technical assistance with the BACE inhibition assays.
[18] a) V. Hoffart, A. Lamprecht, P. Maincent, T. Lecompte, C. Vigner-
on, N. Ubrich, J. Controlled Release 2006, 113, 38–42; b) N. A. Mot-
lekar, B. C. Youan, J. Controlled Release 2006, 113, 91–101; c) S. S.
Lanke, S. G. Gayakwad, J. G. Strom, M. J. D’Souza, J. Microencapsu-
lation 2009, 26, 493–500; d) J. W. Park, O. C. Jeon, S. K. Kim, T. A.
Al-Hilal, K. M. Lim, H. T. Moon, C. Y. Kim, Y. Byun, Thromb. Hae-
mostasis 2011, 105, 1060–1071.
[1] a) J. R. Bishop, M. Schuksz, J. D. Esko, Nature 2007, 446, 1030–
1037; b) J. Turnbull, A. Powell, S. Guimond, Trends Cell. Biol. 2001,
11, 75–82.
[2] M. Mohammadi, S. K. Olsen, O. A. Ibrahimi, Cytokine Growth
Factor Rev. 2005, 16, 107–137.
[3] H. Lortat-Jacob, Curr. Opin. Struct. Biol. 2009, 19, 543–548.
[4] S. Piecewicz, S. Sengupta, Stem Cells Dev. 2011, 20, 749–758.
[5] J. Filmus, M. Capurro, J. Rast, Genome Biol. 2008, 9, 224.
[6] Z. Scholefield, E. A. Yates, G. Wayne, A. Amour, W. McDowell,
J. E. Turnbull, J. Cell Biol. 2003, 163, 97–107.
Received: December 19, 2012
Revised: February 22, 2013
Published online: April 3, 2013
Chem. Eur. J. 2013, 19, 6817 – 6823
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6823