Discovery of PG545: A highly potent and simultaneous inhibitor of angiogenesis, tumor growth, and metastasis

Article abstract of DOI:10.1021/jm201708h

Increasing the aglycone lipophilicity of a series of polysulfated oligosaccharide glycoside heparan sulfate (HS) mimetics via attachment of a steroid or long chain alkyl group resulted in compounds with significantly improved in vitro and ex vivo antiangiogenic activity. The compounds potently inhibited heparanase and HS-binding angiogenic growth factors and displayed improved antitumor and antimetastatic activity in vivo compared with the earlier series. Preliminary pharmacokinetic analyses also revealed significant increases in half-life following iv dosing, ultimately supporting less frequent dosing regimens in preclinical tumor models compared with other HS mimetics. The compounds also displayed only mild anticoagulant activity, a common side effect usually associated with HS mimetics. These efforts led to the identification of 3β-cholestanyl 2,3,4,6-tetra-O-sulfo-α-d-glucopyranosyl-(1→4)- 2,3,6-tri-O-sulfo-α-d-glucopyranosyl-(1→4)-2,3,6-tri-O-sulfo-α- d-glucopyranosyl-(1→4)-2,3,6-tri-O-sulfo-β-d-glucopyranoside, tridecasodium salt (PG545, 18) as a clinical candidate. Compound 18 was recently evaluated in a phase I clinical trial in cancer patients.

Full text of DOI:10.1021/jm201708h

Article
pubs.acs.org/jmc
Discovery of PG545: A Highly Potent and Simultaneous Inhibitor of
Angiogenesis, Tumor Growth, and Metastasis
Vito Ferro,*,†,§ Ligong Liu,†,∥ Ken D. Johnstone,†,⊥ Norbert Wimmer,†,# Tomislav Karoli,†,∥
Paul Handley,† Jessica Rowley,† Keith Dredge,† Cai Ping Li,†,∞ Edward Hammond,† Kat Davis,†
Laura Sarimaa,†,¶ Job Harenberg,‡ and Ian Bytheway†
†Drug Design Group, Progen Pharmaceuticals Limited, Brisbane, Queensland 4076, Australia
‡Clinical Pharmacology, Faculty of Medicine Mannheim, Ruprecht-Karls University of Heidelberg, Mannheim, Germany
S
* Supporting Information
ABSTRACT: Increasing the aglycone lipophilicity of a series
of polysulfated oligosaccharide glycoside heparan sulfate (HS)
mimetics via attachment of a steroid or long chain alkyl group
resulted in compounds with significantly improved in vitro and
ex vivo antiangiogenic activity. The compounds potently
inhibited heparanase and HS-binding angiogenic growth
factors and displayed improved antitumor and antimetastatic
activity in vivo compared with the earlier series. Preliminary
pharmacokinetic analyses also revealed significant increases in half-life following iv dosing, ultimately supporting less frequent
dosing regimens in preclinical tumor models compared with other HS mimetics. The compounds also displayed only mild
anticoagulant activity, a common side effect usually associated with HS mimetics. These efforts led to the identification of 3β-
cholestanyl 2,3,4,6-tetra-O-sulfo-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-sulfo-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-sulfo-α-D-
glucopyranosyl-(1→4)-2,3,6-tri-O-sulfo-β-D-glucopyranoside, tridecasodium salt (PG545, 18) as a clinical candidate. Compound
18 was recently evaluated in a phase I clinical trial in cancer patients.
that is currently in phase III clinical trials for post-resection
hepatocellular carcinoma.14,15 Recently we described16 the
synthesis and evaluation as angiogenesis inhibitors of a series of
polysulfated penta- and tetrasaccharide glycosides containing
the α(1→3)/α(1→2)-linked mannose sequence found in 1.
Compared with other HS mimetics, these compounds are
stable, fully sulfated, single chemical entities. These compounds
strongly inhibited heparanase and displayed potent activity in
cell-based and ex vivo angiogenesis assays, with tetramanno-
sides showing activity comparable to that of pentamannosides.
Selected compounds (e.g., 2 and 3) also showed good
antitumor activity in vivo in a mouse melanoma (solid
tumor) model resistant to 1. Furthermore, the lipophilic
modifications resulted in reduced anticoagulant activity, a
common side effect of HS mimetics,17 and conferred a
reasonable pharmacokinetic profile in the rat, thus building
on earlier results.18,19 These data supported the further
investigation of this class of compound as potential
antiangiogenic, anticancer therapeutics. Herein we describe
the extension of these studies to compounds with larger, more
lipophilic aglycones and with alternative saccharide backbones.
We show that these modifications endow the compounds with
more potent antiangiogenic activity and improved pharmaco-
kinetics, which have subsequently translated into significantly
INTRODUCTION
■
Angiogenesis, the growth of new blood vessels from the pre-
existing vasculature, is a critical process in tumor development,
and its inhibition is now a well-established strategy for the
treatment of cancer.1,2 Recently approved antiangiogenic
therapies include the anti-VEGF antibody bevacizumab and
various tyrosine kinase inhibitors (TKIs).3 While clinical
antitumor effects are evident, overall survival benefits to
patients have been relatively modest and a relapse to
progressive disease often ensues,4 possibly because the cancers
utilize alternative angiogenic mechanisms.5 More disturbingly,
recent studies in mice have shown that treatment with some
TKIs may result in a more aggressive phenotype and may
promote metastases in some circumstances.6−8 Additionally,
many of the currently approved antiangiogenic therapies are
associated with cardiotoxic side effects.9 The development of
new angiogenesis inhibitors with fewer side effects that target
multiple proangiogenic factors and also inhibit metastasis is
therefore a worthwhile strategy to provide improved survival
outcomes in cancer patients.
An example of the above approach is the use of heparan
sulfate (HS) mimetics to simultaneously block HS-mediated
growth factor signaling leading to angiogenesis10 and to inhibit
heparanase,11,12 an endo-β-glucuronidase that cleaves HS and
facilitates metastasis. Of particular note is muparfostat (also
known as PI-88, 1, Chart 1),13 a complex mixture of sulfated
oligosaccharides with antiangiogenic and antimetastatic activity
Received: December 19, 2011
Published: March 29, 2012
© 2012 American Chemical Society
3804
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
a
Chart 1. Structures of Oligomannoside HS Mimetics
a
R = SO3Na.
a
Scheme 1
a
Reagents and conditions: (a) cholesterol or cholestanol, TMSOTf, DCM, 0 °C; (b) (i) NaOMe/MeOH; (ii) AG-50W-X8 (H+ form); (c) (i)
SO3·Py, DMF, 60 °C; (ii) Na2CO3 (aq). For structures of aglycones A and B, see Chart 1.
greater potency in in vivo models and ultimately resulted in the
selection of a clinical candidate.
moderate yield (42−44%). Sulfonation of 16 gave a mixture of
products by CE,20 presumably due to the relative instability of
the cholesteryl group to the sulfonation conditions, requiring
purification by size exclusion chromatography. However,
sulfonation of cholestanyl glycoside 17 proceeded smoothly
to give 5 in >95% purity by CE following simple purification by
dialysis. Cholestanol was thus used as the steroidal glycosyl
acceptor of choice for subsequent compounds.
Compound 5 bound tightly to the angiogenic growth factors
FGF-1, FGF-2, and VEGF (Ki in the low nanomolar to
picomolar range, Table 1) as determined by a surface plasmon
resonance-based solution affinity assay21,22 and potently
inhibited heparanase activity23 (Ki = 5.8 nM, Table 2),24
indicating that the larger aglycone did not adversely affect
activity. When tested in growth factor-induced endothelial cell
proliferation assays25 using HUVECs (Table 2), compound 5
displayed significantly improved potency compared with the
RESULTS AND DISCUSSION
■
We reasoned that significantly increasing the lipophilicity of the
aglycone in this class of compound could result in increased
membrane and protein binding leading to increases in plasma
residence time and potency. In the earlier study16 the dodecyl
group gave no clear benefit over the shorter octyl group. The
first compounds targeted for synthesis were therefore the
significantly more lipophilic steroidal tetramannosides 4 and 5,
which were prepared according to Scheme 1. Glycosylation of
cholesterol or cholestanol with trichloracetimidate 1316 using
TMSOTf as the promoter gave exclusively the α-linked
glycosides 14 and 15 in moderate yield. Deprotection
(NaOMe in MeOH) followed by exhaustive sulfonation (excess
SO3·Py, DMF, 60 °C) gave the target compounds 4 and 5 in
3805
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
a
α(1→3)/α(1→2)-linked mannose oligosaccharides. Polysul-
fated maltohexaose, for example, displays similar potency to 1
in in vitro angiogenesis and heparanase inhibition assays.13 In
addition, maltosaccharides are commercially available and
inexpensive, being important products in the starch industry.
The corresponding polysulfated oligomaltosides should thus be
cheaper to manufacture than the oligomannosides where the
starting oligosaccharide must first be synthesized from
monosaccharide building blocks.26,29 Compounds 18−29
(Chart 2) were prepared similarly according to Scheme 1 via
TMSOTf-catalyzed reaction of peracetylated or perbenzoylated
trichloroacetimidate glycosyl donors with cholestanol or with 3-
azidopropanol followed by reduction and coupling with stearyl
chloride. The use of participating protecting groups (acetyl or
benzoyl) gave the expected β-linked glycosides as the sole
glycosylation products. In this series, cholestanol conjugates
with a triazole linker were synthesized via the click reaction
between 3β-O-(prop-2-ynyl)cholestanol and the corresponding
protected β-glycosyl azides, which were themselves readily
prepared from the corresponding α-glycosyl bromides by
nucleophilic displacement with azide ion (see Supporting
Information). The maltotetraosyl azide also provided access to
conjugate 21 in which the steroid (deoxycholic acid) was
attached with the opposite orientation, via reduction to the
glycosylamine (H2, 10% Pd/C) and HOBt-mediated coupling
with diacetyldeoxycholic acid.30
Table 1. Growth Factor Binding of Test Compounds
Kd (nM)
compd
FGF-1
FGF-2
9
VEGF
b
1
0.24 0.04
0.118 0.009
0.60 0.50
0.44 0.04
0.24 0.10
270 30
61
2.2 2.3
b
2
160 40
160 40
108 11
18
22
5
6
5
6
40 17
7
39
1570 150
631
6
1.04 0.19
1300 300
319 19
460 30
90 30
8
9
18.5 1.8
22.6 1.0
4
10
11
12
18
19
20
21
22
23
24
25
26
27
28
29
480 40
490 40
474.0 1.4
390 80
390 90
610 60
16
7
14.3 3.0
260 130
28.9 2.3
8
4
9.7 0.8
50
95
8
8
790 230
7.95 0.07
190 60
1.25 0.07
24.8 2.3
4.94 0.08
32.2 2.3
560 30
50
4
547 50
311 25
530 50
2200 400
3400 300
>3000
14
4
380 110
1200 400
1930 230
3000 400
1350 70
1600 300
840 130
3600 600
870 60
2000 400
1800 300
610 210
a
Binding affinities for growth factors FGF-1, FGF-2, and VEGF were
All test compounds were of ≥95% purity by CE following
purification by dialysis, and 1H and 13C NMR data were
consistent with full sulfonation, as noted for the earlier series.16
The compounds were evaluated for their ability to inhibit
heparanase, bind to angiogenic growth factors, and inhibit
growth factor-induced endothelial cell proliferation (Tables 1
and 2). The anticoagulant activity of the compounds was also
determined via the APTT and Heptest assays (Table 2), and
selected compounds were evaluated in the Matrigel micro-
tubule formation assay and the ex vivo rat aortic angiogenesis
assay (Figure 1).
determined using a surface plasmon resonance-based solution affinity
assay, as previously described.21,22 The Kd values represent the average
b
of at least duplicate measurements
the standard deviation. Data
from Johnstone et al.16
earlier series (e.g., the octyl tetramannoside 2) for all growth
factors tested (FGF-1, FGF-2, and VEGF). The IC50 values
were approximately 10-fold (for FGF-2) or 6-fold (for FGF-1
and VEGF) lower compared with the corresponding values for
2. A series of analogues of 5 were then synthesized, initially
based on the same oligomannose scaffold in 2 while varying the
length and linkage of the oligosaccharide. We were particularly
interested to see if the increased lipophilicity could improve the
potency of shorter oligosaccharides, as similar compounds had
previously possessed only modest HS-mimetic activity
compared with longer congeners.13,26 The introduction of a
spacer between the saccharide and the cholestanol was also
investigated, as was the use of an alternative highly lipophilic
aglycone, 3-stearamidopropyl. Compounds 6−12 were pre-
pared in a manner similar to that shown in Scheme 1 via
glycosylation, deprotection, and sulfonation. The compounds
with a triazole spacer (10, 12) were prepared via the Cu(I)-
catalyzed Huisgen 1,3-dipolar cycloaddition (“click”) reac-
tion27,28 between the protected azidopropyl glycoside and 3β-
O-(prop-2-ynyl)cholestanol. Reduction of the azide group
(Ph3P) followed by coupling with stearyl chloride furnished
the stearamidopropyl derivative 6.
Previous studies of sulfated oligosaccharides as HS mimetics
have shown that activity is dependent on monosaccharide
composition as well as on the length and linkage of the
oligosaccharide.13,17,26 Therefore, we also investigated HS
mimetics with alternative monosaccharides to D-mannose.
Particular attention was focused toward α(1→4)-linked
oligosaccharides of D-glucose (i.e., maltosaccharides) because
when exhaustively sulfonated, these compounds show potent
antiangiogenic and anti-heparanase activity similar to that of the
As seen with the earlier series,16 the presence of a lipophilic
aglycone significantly attenuated the anticoagulant activity of
the compounds compared with nonglycosylated 1. The data
also indicate that compounds with either a stearamidopropyl or
cholestanyl group, with or without a spacer, had strong affinities
for the angiogenic growth factors FGF-1, FGF-2, and VEGF,
with tetrasaccharides displaying Kd values similar to those of
their less lipophilic congeners from the earlier series.16 It was
also evident that affinity (particularly for VEGF and FGF-1)
was, not unexpectedly, dependent on oligosaccharide chain
length; e.g., compared with maltotetraoside 18, the affinity of
maltotrioside 22 for VEGF and FGF-1 was 6-fold and 3-fold
lower, whereas for maltoside 25 the affinity was 40-fold and 70-
fold lower, respectively. Interestingly, however, heparanase
inhibitory activity was significantly improved. Sulfated di- and
trisaccharides with small or no aglycones display only partial
competitive inhibition of heparanase26 and do not completely
inhibit the enzyme even at very high concentrations,
presumably because substrate access to the enzyme active site
is not completely blocked. For example, a sulfated mannobiose
only inhibited heparanase up to 58.6% with an IC50 of 1.57
0.31 μM (Kic = 2.23 0.20 μM; Kiu = 68.3 22.4 μM).26 In
contrast, all the compounds in this series completely inhibited
heparanase activity with the maltotrioside 22 showing activity
comparable to that of the maltotetraoside 18 (Ki of 9 and 6 nM,
3806
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
a
b
Table 2. Inhibition of Growth Factor-Induced HUVEC Proliferation, Inhibition of Heparanase, and Anticoagulant Activity
(at 0.1 mg/mL) of Test Compounds
c
IC50 (μM)
anticoagulant activity
Ki (nM)
heparanase
compd
FGF-1
FGF-2
VEGF
APTT (s)
Heptest (s)
d
e
1
42 14
10
9
5
20
9
6
5
7.9 0.6
>500
96.7
94.0
68.1
63.3
40.6
50.7
73.3
58.5
71.4
71.9
104.0
61.2
41.0
51.6
65.8
51.1
38.6
40.3
36.1
38.8
39.6
>500
79.5
28.0
24.5
29.6
25.5
27.3
33.1
24.2
26.4
28.6
45.5
30.2
26.8
29.2
32.2
24.9
23.9
26.7
23.9
25.3
25.3
d
2
9
2
2
59
7
5
1.29 0.25
1.54 0.16
0.99 0.10
>10
0.80 0.30
0.70 0.30
0.39 0.07
2.22 0.22
2.27 0.23
1.7 0.4
1.8 1.0
1.2 0.4
0.70 0.30
0.64 0.27
1.2 0.3
>50.0
1.2 1.0
0.60 0.60
0.22 0.16
1.8 1.1
1.0 0.5
2.1 1.9
1.7 1.2
1.4 1.1
0.5 0.3
2.2 0.9
1.8 0.6
>50.0
5.8 1.5
6.0 2.1
5.5 2.6
22.3 1.6
6.4 2.5
8.50 0.14
10.5 1.7
8.4 2.6
6.1 2.5
3.7 0.8
6
7
8
9
4.4 0.5
3.8 0.6
2.5 0.7
2.61 0.14
1.21 0.17
2.3 1.1
2.1 0.6
>50.0
10
11
12
18
19
20
21
22
23
24
25
26
27
28
29
16
20
4
5
2.5 1.4
0.47 0.23
1.7 0.8
4.0 1.5
4.1 0.9
2.2 0.4
2.3 1.2
0.80 0.50
1.7 1.3
>10
1.6 0.6
0.24 0.16
1.1 0.6
5.2 2.4
1.13 0.19
2.24 0.22
5.6 1.5
2.62 0.17
9.1 2.5
9.1 0.3
11.3 0.4
f
ND
f
4.17 0.08
2.9 0.4
5.6 0.5
2.43 0.12
ND
111 28
f
6
3
ND
2.7 0.4
30
7
a
The IC50 data represent the mean and SD where compounds were tested in two to eight independent experiments. Data without SD are from single
experiments. Ki the standard error determined from inhibition response curves as previously described.23 APTT and Heptest values are the
b
c
mean values of two determinations. The normal ranges for coagulation values in pooled human plasma are 26−36 s for APTT and 13−20 s for
d
e
f
heptest. Data from Johnstone et al.16 Data from Hammond et al.23 Not determined.
a
and conformational requirements for binding to HS-binding
Chart 2. Structures of HS Mimetics
proteins beyond simply chain length and charge density.31
The increased potency of the new compounds compared
with 1 and the earlier series became apparent in cell-based
assays used to determine antiangiogenic activity. We attribute
this increased potency to improved bioavailability, possibly
because of better localization to the cell surface membrane by
the lipophilic aglycone. Our results indicate that the
stearamidopropyl and cholestanyl aglycones have similar effects
in this regard. The first of these assays focuses on the
compounds’ ability to inhibit endothelial cell proliferation in
response to angiogenic growth factors (Table 2). With one
exception, all compounds potently inhibited FGF-1-, FGF-2-,
and VEGF-induced proliferation of HUVECs (Table 2), with
IC50 values up to 28-fold (for FGF-2) or 90-fold (for FGF-1
and VEGF) lower compared with 1 and 25-fold (for FGF-2),
19-fold (for FGF-1), or 40-fold (for VEGF) compared with 2.
The exception was compound 21, with a sulfo group at the
aglycone (deoxycholic acid) terminus, which had no inhibitory
activity up to 50 μM, presumably because the sulfo group
reduced the overall lipophilicity of the aglycone. The potency in
cell proliferation assays produced similar trends with respect to
oligosaccharide chain length, as seen in growth factor binding
assays, although the magnitude of the differences was smaller
and dependent on the growth factor. For example, in the
homologous series 25, 22, and 18 for VEGF the IC50 increases
10-fold (0.5−5 μM), for FGF-2 the increase is >10-fold (0.7 to
>10 μM), while for FGF-1 the increase is only ∼4-fold (1.2−4
μM). Similar results and trends for all compounds were also
a
R = SO3Na. For structures of aglycones B−D, see Chart 1.
respectively) and with even the mannobioside 8 and mannoside
27 showing good activity (Ki of 22 and 111 nM, respectively).
The nature of the oligosaccharide backbone had an
interesting effect on growth factor affinity. For example, while
the mannotetraoside 5 and maltotetraoside 18 had the same
inhibitory potency against heparanase and a similar affinity for
VEGF, 5 had a 13-fold greater affinity for FGF-1 and a 2.5-fold
greater affinity for FGF-2 than 18. These results reinforce the
view that different sulfated oligosaccharides have unique spatial
3807
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
Figure 1. Inhibition of in vitro and ex vivo angiogenesis by test compounds at 10 μM. (A) Inhibition of HUVEC tube formation. Tube formation
was assessed using low magnification (×40) and counting the total number of branching nodes connecting three or more tubules. Data are presented
as the mean number of branching nodes SD compared with controls. (B) Inhibition of microvessel outgrowths in the rat aortic assay, determined
using a scoring system outlined in the Experimental Section.
Figure 2. Dose related inhibition of in vitro and ex vivo angiogenesis by 18. (A) Dose related effect of 18 on capillary tube formation of dHMVECs
after 18 h of incubation on Matrigel basement membrane. Values are shown as mean number of branching nodes SD. Representative images of
control and 18 (10 μM) are also shown below. (B) Dose related antiangiogenic effect of increasing concentrations of 18 in Matrigel cultures of rat
aortic ring explants maintained in the absence of exogenous growth factors. Values are the mean
SD of microvessel outgrowths after daily
treatment for 6 days. Representative images of control and 18 (10 μM) are also shown below.
obtained using dermal human microvascular endothelial cells
(dHMVECs) in place of HUVECs (data not shown).
because the endothelial cells have not been preselected by
passaging and are thus not in a proliferative state at the time of
explantation.32 This may account for the fact that some
compounds were more potent in one assay over the other.
Figure 2 shows representative dose response data for 18
(ultimately selected as a clinical candidate; see below), which
indicates that it is a potent inhibitor of in vitro and ex vivo
angiogenesis with IC50 values of 2.0 0.5 μM and 1.1 0.2
μM, respectively, for the tube formation and rat aorta assays,
respectively. It is important to note that the compounds were
not cytotoxic at the inhibition values (IC50 and Kd/Ki) in the
above assays, e.g., the LC50 for 18 (78 μM)25 is approximately
80-fold higher than its IC50 in the cell proliferation assay. In
Endothelial cell tube formation on Matrigel using HUVECs
(or dHMVECs) was also strongly inhibited, with some
compounds (i.e., 5, 6, 18, and 23) showing >80% inhibition
at 10 μM (Figure 1A), a significant improvement compared
with 1 and 2.16 Selected compounds were also evaluated in the
ex vivo rat aortic angiogenesis assay (Figure 1B) with
compounds 5, 18, and 19 emerging as the most potent, all
inhibiting angiogenesis by over 80% when administered daily at
10 μM for 6−8 days.25 The rat aorta assay is considered to
come closer to simulating the in vivo situation not only because
it includes the surrounding nonendothelial cells but also
3808
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
experiments to assess tissue viability following 6 days of
treatment with the test compounds, aortic segments receiving
fresh medium containing VEGF (10 ng/mL) frequently
developed new microvessels within 72 h (data not shown),
confirming the lack of cytotoxicity.
mimetics. Changing the oligosaccharide backbone from
oligomannose to the cheaper and more readily accessible
maltotetraose also resulted in compounds with improved
manufacturability. These efforts led to the discovery of the
clinical candidate 18 (PG545), which was recently evaluated in
a phase I clinical trial in cancer patients. The antimetastatic
activity of 18, due to its potent inhibition of heparanase,
differentiates it from other antiangiogenic agents such as the
TKIs. In light of recent preclinical findings regarding anti-
VEGF therapies, this attribute may be of critical importance as
18 progresses through the clinic.
Several compounds including 18−20 and 22−24 were then
evaluated in syngeneic murine models of tumor growth and
metastasis, the B16 melanoma model and the B16 experimental
metastasis model. When administered by subcutaneous
injection as solutions in PBS for 12 consecutive days at a
dose of 10 mg/kg, compounds 18−20 potently inhibited the
development of metastatic nodules.25 In the B16 solid tumor
model, a model that is resistant to treatment with compound 1,
daily administration of compounds 18−20 at 15 mg/kg for 12
days, 3 days after tumor challenge, resulted in potent inhibition
of tumor growth with no evidence of palpable tumors until
cessation of drug treatment.25 This compares favorably with the
earlier series (e.g., 2, 3) where palpable tumors were evident
after several days of treatment at 30 mg/kg (b.i.d.). Compound
18 was also shown to significantly inhibit tumor growth in the
HT29 colon xenograft model when administered daily at 5 mg/
kg.25
Preliminary screening pharmacokinetic evaluation of com-
pounds 18−20 and 22−24 indicated that the more lipophilic
aglycones were indeed significantly affecting their pharmacoki-
netic properties, with apparent half-lives in the rat at least >3-
fold longer than for the octyl tetrasaccharide 2 (see Supporting
Information), presumably as a result of increased binding to
lipophilic cell membranes and plasma proteins. Subsequent
detailed pharmacokinetic analyses of 18 revealed that it has a
relatively long half-life. For example, the mean elimination half-
life of [3H]-labeled 18 in tumor-bearing mice was >50 h
following up to three sc bolus doses at 96 h intervals.33 The
long half-life of 18 enabled less frequent dosing regimens (e.g.,
weekly or twice weekly) in preclinical models while maintaining
substantial efficacy.33
The above studies, in combination with considerations such
as manufacturability and pharmacokinetic properties (see
below), ultimately resulted in the selection of compound 18
as the clinical candidate. Compound 18 was subsequently
subjected to extensive pharmacological evaluation in multiple
preclinical models, as detailed elsewhere.33 In summary, 18 was
shown to potently inhibit angiogenesis in vivo and solid tumor
growth in murine models of breast, prostate, liver, lung, colon,
head and neck cancers and melanoma. Compound 18 also
displayed potent antimetastatic activity in murine models of
experimental and spontaneous metastasis and showed
enhanced antitumor activity in a liver cancer model when
used in combination with the TKI sorafenib.33 It is important
to note that sorafenib, an approved drug for kidney and liver
cancer, showed no antimetastatic activity in these models.
In conclusion, further optimization of a series of polysulfated
oligosaccharide glycosides via increasing the lipophilicity of the
aglycone resulted in HS mimetics with potent and significantly
improved in vitro and ex vivo antiangiogenic activity and in vivo
antitumor and antimetastatic activity. Preliminary pharmacoki-
netic analyses also revealed significant increases in half-life
following iv dosing, ultimately supporting less frequent dosing
regimens compared with other HS mimetics.33 The improved
potency and pharmacokinetics may be because of better
localization to the cell surface membrane by the lipophilic
aglycone. The compounds displayed only mild anticoagulant
activity, a common side effect usually associated with HS
EXPERIMENTAL SECTION
■
General Information. Unless otherwise stated, nuclear magnetic
resonance (NMR) spectra were recorded at 400 MHz for 1H and 100
MHz for 13C, in either deuteriochloroform (CDCl3) with residual
CHCl3 (1H, δ 7.26) or deuterium oxide (D2O), employing residual
HOD (1H, δ 4.78) as internal standard, at ambient temperatures (298
1
K). Where appropriate, analyses of H NMR spectra were aided by
gCOSY experiments. Flash chromatography was performed on silica
gel (40−63 μm) under a positive pressure with specified eluants. All
solvents used were of analytical grade. The progress of reactions was
monitored by thin layer chromatography (TLC) using commercially
prepared silica gel 60 F254 aluminum-backed plates. Compounds were
visualized by charring with 5% sulfuric acid in methanol and/or under
ultraviolet light. Analytical HPLC was performed on a Waters Alliance
2690 separations module equipped with a Waters 2410 refractive index
detector, using a Phenomenex PolySep-GFC-P2000 (300 mm × 7.8
mm) column and 0.1 M NaNO3 as the mobile phase. The flow rate
was 0.8 mL/min, and the column temperature was set to 35 °C.
Capillary electrophoresis (CE) was performed in reverse polarity
mode on an Agilent CE system using 10 mM 5-sulfosalicylic acid, pH
3, as the background electrolyte and detection by indirect UV
absorbance at 214 nm, as previously described.20 Compound
1
homogeneity was determined by H and/or 13C NMR spectroscopy
and for the final products by CE and/or HPLC. Unless indicated
otherwise, the purity of the final products was ≥95%.
2,3,4,6-Tetra-O-acetyl-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-
acetyl-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-acetyl-α-D-gluco-
pyranosyl-(1→4)-1,2,3,6-tetra-O-acetyl-D-glucopyranose (30).
Dry maltotetraose (502 mg, 0.753 mmol) and DMAP (cat.) were
dissolved in dry pyridine (10 mL) and cooled at 0 °C. A solution of
Ac2O (2.8 g, 27.4 mmol, 2.6 equiv per hydroxy group) in pyridine (5
mL) was added dropwise at 0 °C. The mixture was stirred at 0 °C for 4
h and left at −20 °C for 48 h. The reaction was not completed;
therefore, additional Ac2O (1.0 g, 9.8 mmol) was added at 0 °C. The
mixture was stirred at room temperature for 16 h, cooled to 0 °C, and
treated with dry MeOH (10 mL). After being stirred at room
temperature for 2 h, the solution was concentrated and the residue
coevaporated with toluene (3 × 30 mL) to give the peracetate 30 as a
1
white solid (920 mg, 97%). H and 13C NMR spectra were in accord
with the literature.34
2,3,4,6-Tetra-O-acetyl-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-
acetyl-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-acetyl-α-D-gluco-
pyranosyl-(1→4)-2,3,6-tri-O-acetyl-α-D-glucopyranosyl Tri-
chloroacetimidate (31). To a precooled (0 °C) solution of
ethylenediamine (1.66 mmol, 0.11 mL) in dry THF (15 mL) was
added dropwise glacial acetic acid (0.90 mmol, 0.053 mL). A
precipitate was formed immediately. The peracetate 30 (900 mg,
0.717 mmol) was added at 0 °C and the mixture stirred at room
temperature for 2 h until TLC (toluene/EtOAc, 1:2) indicated the
absence of the starting material and the presence of a slower moving
product. The solution was neutralized using acetic acid (0.15 mL, to
pH 6). The solvent was blown out with a stream of air. The residue
was dissolved in EtOAc (100 mL), washed with saturated NaHCO3
solution (3 × 50 mL), water (3 × 10 mL), brine (30 mL), and dried
(Na2SO4). The solution was filtered and concentrated in vacuo to give
the hemiacetal as a yellow foam (830 mg, 0.684 mmol), which was
dissolved in dry DCM (5 mL). The solution was cooled to 0 °C, and
3809
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
K2CO3 (1.20 g, 8.60 mmol) and trichloroacetonitrile (0.849 mL, 8.40
mmol) were added. The mixture was stirred at room temperature for 2
h. The mixture was purified by flash chromatography (silica gel, 20 cm
× 1.5 cm, toluene−EtOAc, 1:2 → EtOAc, containing 0.2% (v/v)
Et3N) to give the desired trichloroacetimidate34 31 as a white fluffy
powder (795 mg, 86%). The compound was dried over P2O5
overnight and stored at −20 °C.
solution of NaOH (5 M, 5 drops) and dialyzed against water (4 L)
using a Slide-A-Lyzer cassette (2000 MWCO, 4−12 mL) for 16 h at
room temperature. The dialysis against water (4 L) was continued at 0
°C for 3 days, whereby the water was changed after each 24 h, as well
as an aqueous solution NH4HCO3 (3 M, 0.6 mL) was added to the
water to set the pH to ∼6.0−6.5. The desalted solution was then
lyophilized to afford the polysulfate 18 as a white fluffy powder (97
mg, 89%, 99% pure by CE20). 1H NMR (400 MHz, D2O): δ 5.63 (d, 1
H, J1,2 = 3.3, H1), 5.60 (d, 1 H, J1,2 = 3.6, H1), 5.50 (d, 1H, J1,2 = 3.6,
H1), 5.01 (d, 1H, J1,2 = 4.8, H1I), 4.10−4.93 (m, 23H, 4 × H2, 4 ×
H3, 4 × H-4, 3 × H5, 8 × H6), 4.05 (m, 1H, H5I), 3.76 (m, 1 H, H-3
chol), 0.54−1.97 (m, 33 H, 12 × CH2, 9 × CH), 0.86 (d, 3H, J = 6.5,
cholestanyl-CH3), 0.795 (d, 3H, J = 6.6, cholestanyl-CH3), 0.792 (d,
3H, J = 6.6, cholestanyl-CH3), 0.76 (s, 3H, cholestanyl-CH3), 0.61 (s,
3H, cholestanyl-CH3). Polysulfate 18 was also analyzed by 1D and 2D
NMR spectroscopy in D2O at 600 MHz, and the 1H and 13C chemical
shift data are presented in Tables S3 and S4. Chemical shifts were fully
assigned using gHSQC, gCOSY, gHMBC, TOCSY, and ROESY
techniques.
Biological Assays. Growth Factor Binding Assays. The
binding affinities of the compounds for the growth factors FGF-1,
FGF-2, and VEGF were determined on a BIAcore 3000 (BIAcore,
Uppsala, Sweden) using a surface plasmon resonance-based solution
affinity assay, as previously described.21,22 Heparin-coated CM4 sensor
chips were prepared via reductive amination of the reducing end
aldehyde with 1,4-diaminobutane, as previously described.22 Ligands
binding to FGF-1 and VEGF were measured in HBS-EP buffer (10
mM HEPES, pH 7.4, 150 mM NaCl, 3.0 mM EDTA, and 0.005% (v/
v) polysorbate 20), while binding to FGF-2 was measured in HBS-EP
buffer containing 300 mM NaCl. Sensorgram data were analyzed using
the BIAevaluation software, version 4.1 (BIAcore), and Kd values were
determined as previously described.22 The Kd values represent the
3β-Cholestanyl 2,3,4,6-Tetra-O-acetyl-α-D-glucopyranosyl-
(1→4)-2,3,6-tri-O-acetyl-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-
acetyl-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-acetyl-β-D-gluco-
pyranoside (32). The trichloroacetimidate 31 (300 mg, 0.221
mmol), cholestanol (2 equiv, 172 mg, 0.442 mmol), and 3 Å molecular
sieves (100 mg) were stirred in dry DCM (1.5 mL) for 0.5 h. A
solution of TMSOTf in dry DCM (0.4 M, 0.275 mL, 0.11 mmol, 0.5
equiv) was added dropwise at 0 °C. After 30 min at room temperature,
another portion of TMSOTf in dry DCM (0.4 M, 0.2 mL, 0.08 mmol,
0.36 equiv) was added and stirring continued for 30 min at room
temperature. The reaction was quenched by adding Et3N (0.025 mL)
at 0 °C. After being stirred for 10 min, the mixture was filtered through
a plug of Celite (0.5 cm) and washed with DCM (5 × 25 mL) and
EtOAc (3 × 25 mL). Both organic phases were washed separately with
saturated NaHCO3 solution (3 × 25 mL) and brine (25 mL). Aqueous
extracts were combined and re-extracted with EtOAc (3 × 30 mL).
The EtOAc extracts were combined and washed with brine (30 mL),
combined with the other organic extracts and dried (Na2SO4). The
solution was filtered and concentrated in vacuo to afford a yellow foam
(480 mg). Flash chromatographic purification (silica gel, 30 cm × 5
cm, toluene/EtOAc, 3:2 → 1:1 → 1:2 → EtOAc, containing 0.2%
Et3N (v/v)) afforded two fractions. Fraction A gave the desired β-
linked glycoside 32 as a white foam (81 mg, 23%), and fraction B
contained 77% partially deacetylated α-linked glycoside and 23%
1
partially deacetylated β-linked glycoside (118 mg). H NMR (CDCl3,
400 MHz): δ 5.24−5.45 (m, 7H, 3 × H1, 4 × H3), 5.08 (t, 1H, J3,4
=
average of at least duplicate measurements
SD. The data are
J4,5 = 9.80 Hz, H4IV) 4.86 (dd, 1H, J1,2 = 4.1, J2,3 = 10.4, H2IV), 4.70−
4.80 (m, 3H, 3 × H2), 4.63 (d, 1H, J1,2 = 7.7, H1I), 4.33−4.54 (m, 4H,
4 × H6), 3.86−4.31 (m, 10 H, 3 × H4, 3 × H5, 4 × H6), 3.70 (ddd,
1H, H5I), 3.56 (m, 1H, cholestanyl-H3), 2.20, 2.19, 2.16, 2.11, 2.07,
2.04, 2.03, 2.02, 2.015, 2.010, 2.00, 1.99 (s, 39H, 13 × Ac), 0.55−2.00
(m, 33H, 12 × CH2, 9 × CH), 0.90 (d, 3H, J = 6.6, cholestanyl-CH3),
0.871 (d, 3H, J = 6.6, cholestanyl-CH3), 0.867 (d, 3H, J = 6.6,
cholestanyl-CH3), 0.78 (s, 3H, cholestanyl-CH3), 0.65 (s, 3H,
cholestanyl-CH3).
presented in Table 1.
Heparanase Inhibition Assay. The inhibitory potency of the test
compounds against recombinant human heparanase was determined
by the fondaparinux assay as previously described.23 Briefly, assay
solutions (100 μL) comprised 40 mM sodium acetate buffer, pH 5.0,
100 mM fondaparinux (GlaxoSmithKline) with or without inhibitor.
Heparanase was added to a final concentration of 140 pM to start the
assay. The plates were sealed with adhesive tape and incubated at 37
°C for 2−24 h before the assays were stopped by the addition of 100
μL of a solution containing 1.69 mM 4-[3-(4-iodophenyl)-2-(4-
nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate (WST-1, Aus-
pep, Melbourne, Australia) in 0.1 M NaOH. The plates were resealed
with adhesive tape and developed at 60 °C for 60 min, and the
absorbance was measured at 584 nm (Fluostar, BMG Labtech). In
each plate, a standard curve constructed with D-galactose as the
reducing sugar standard was prepared in the same buffer and volume
over the range of 2−100 μM. The IC50 value was determined and
converted to a Ki using the method of Cheng and Prusoff.35 All curve
fitting was done using BIAevaluation software, version 4.1 (Biacore).
Growth Factor Induced Endothelial Cell Proliferation Assay.
Endothelial cell culture using HUVEC cells was carried out as
previously described.16 Test compounds were weighed out from
powder stocks and diluted in PBS to 10 mM stock solutions and
stored at −80 °C. For experiments, compounds were subsequently
diluted in EBM-2 medium (supplemented with 2% FBS and
gentamicin) to various working concentrations as required. The
proliferation assays using various concentrations of the growth factor
VEGF, FGF-1, or FGF-2 were carried out as previously described.16
Briefly, 100 μL of cells was added to each well at concentrations
between 1 × 103 and 3 × 103 per well. Growth factors and test
compounds were then added in 50 μL volumes at specified
concentrations (typically seven ranging from 1−50 μM in triplicate)
to obtain a final volume of 200 μL. Following incubation for 70 h, 20
μL of the CellTiter 96 Aqueous One solution cell proliferation assay
(Promega) was added for 2 h prior to reading the absorbance at 490
nm to obtain OD values. Once these values were subtracted from the
blanks (media only), the data were imported into BIAevaluation
3β-Cholestanyl α-D-Glucopyranosyl-(1→4)-α-D-glucopyrano-
syl-(1→4)-α-D-glucopyranosyl-(1→4)-β-D-glucopyranoside
(33). The peracetate 32 (75 mg, 0.047 mmol) was dissolved in
anhydrous MeOH (0.1 M) and treated with a solution of NaOMe in
MeOH (1.35 M, 0.2−0.6 equiv). The mixture was stirred at room
temperature for 3 h (monitored by TLC). Acidic resin AG-50W-X8
(H+ form) was added to adjust to pH 6−7. The mixture was filtered,
and the resin was rinsed with MeOH. The combined filtrate and
washings were concentrated in vacuo and thoroughly dried to give the
polyol 33 as a white solid (48 mg, 98%). ESMS m/z 1059.6 (M +
Na+); HR ESMS m/z 1059.5667 (M + Na+, calcd for C51H88O21Na+
1059.5710). Polyol 33 was analyzed by 1D and 2D NMR spectroscopy
1
in DMSO-d6 at 600 MHz and the H and 13C chemical shift data are
presented in Tables S1 and S2. Chemical shifts were fully assigned
using gHSQC, gCOSY, gHMBC, TOCSY, and ROESY techniques.
3β-Cholestanyl 2,3,4,6-Tetra-O-sulfo-α-D-glucopyranosyl-
(1→4)-2,3,6-tri-O-sulfo-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-
sulfo-α-D-glucopyranosyl-(1→4)-2,3,6-tri-O-sulfo-β-D-glucopyr-
anoside, Tridecasodium Salt (18, PG545). The polyol 33 (48 mg,
0.046 mmol) was dissolved in dry DMF (2.3 mL, 0.02 M) and freshly
washed and dried SO3−pyridine complex (285 mg, 3 equiv per OH
group, 1.79 mmol) added, and the mixture was stirred at 60 °C for 16
h. The reaction mixture was cooled to 0 °C for 10 min, then
neutralized by adding ice-cold aqueous NaOH solution (5 M, 2.1
equiv/SO3, 0.752 mL, 3.76 mmol) at 0 °C in one portion (to pH 12).
The suspension was stirred at 0 °C for 15 min, diluted with water (10
mL), and concentrated in vacuo at 40 °C. A pale yellow powder was
afforded, which was dissolved in water (10 mL), obtaining a solution
with pH 11.5. The solution was set to pH 12.5 by adding an aqueous
3810
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
software to determine IC50 values for each curve. The IC50 data
represent the mean and standard deviation to indicate the
interexperimental error where compounds were tested in two to
eight independent experiments. Data without SD are from single
experiments.
Preliminary Pharmacokinetic Analyses for Compounds 18,
23, 24. Male Sprague−Dawley rats (250−350 g) were obtained from
the Herston Medical Research Centre at Royal Brisbane Hospital,
Australia. The animals were allowed free access to food and water
before and during the experiment, during which they were maintained
unrestrained in metabolism cages. Animal procedures were approved
by The University of Queensland Animal Ethics Committee. Rats were
anesthetized with isoflurane (Forthane). A catheter was inserted in the
external jugular vein via an incision in the neck and was passed under
the skin to a second incision in the skin of the back (midline vicinity of
the scapulae). This was then exteriorized with the protection of a light
metal spring. The incision was closed and the spring fixed to the skin
with Michel sutures so that the rats had a full range of movement. The
animals were carefully monitored during recovery (1 h). The dose was
prepared in phosphate-buffered saline to give a total drug
concentration of 5.00 mg/mL. All doses were administered as a
bolus of 10 mg/kg in a dose volume of approximately 2 mL/kg. Blood
samples (about 250 μL) were collected predose and at 5, 10, 20, 35,
65, 125, 185, 305, and 455 min after dosing. The blood samples were
immediately centrifuged, and an aliquot of the plasma was transferred
to a vial for CE analysis. At completion of the experiments, the animals
were sacrificed by a lethal overdose of iv pentobarbitone.
Sample Preparation. Rat plasma samples (100 μL) were mixed
with 25 μL of 1% SDS and 75 μL of 5% CHAPS. Proteinase K solution
(50 μL of 40 mg/mL in 5 mM CaCl2) was added, and the mixture was
incubated at 37 °C for 1.5 h and then sonicated at 40 °C for 60 min.
An additional 230 μL of water was added to the sample solution, and
this was transferred onto a preconditioned (MeOH × 1 mL, water × 1
mL, 0.1 M NH4Ac × 1 mL) Waters Oasis WAX cartridge (30 μm, 1
mL/30 mg). The WAX cartridge was washed with 0.1 M NH4OAc (1
mL) followed by water (1 mL), 2% acetic acid (1 mL), water (1 mL),
MeOH (3 mL), and finally water (1 mL). The product was eluted off
the cartridge using ammonia solution [prepared by mixing 28% NH3
solution (18 mL) with 1:1 MeOH−water (32 mL)], and the solution
was freeze-dried and redissolved into 200 μL of water for CE analysis.
CE Analysis. Samples were analyzed by CE in reverse polarity mode
on an Agilent HP3D CE system equipped with a DAD detector.
Beckman eCap fused silica capillaries (75 μm i.d., 375 μm o.d., 66 cm
total length, 57.5 cm to detector) were used in these analyses. Indirect
UV detection was used relying on 10 mM 5-sulfsosalicylic acid, pH 3.0
(adjusted with aqueous NaOH), as the background electrolyte with
detection at 214 nM. The capillary temperature was 20 °C, and the
operating voltage was −25 kV. Samples were injected using
electrokinetic supercharging (−10 kV, 200 s) for online preconcentra-
tion. The limit of detection for 18 was 20 μg/mL.
Matrigel Microtubule Formation Assay. The tube formation
assay was performed as described previously.25 Briefly, HUVECs in the
fourth or fifth passage at 70−80% confluence were harvested and
resuspended in Lonza endothelial growth medium (EGM2) containing
all supplements as directed by the manufacturer, except heparin, at a
cell density of 4 × 105 cells per mL. For each set of triplicate wells, an
amount of 200 μL of cells (4 × 105/mL) was treated with an equal
volume of compound to obtain a final concentration of 10, 50, or 100
μM (thus ensuring 1 × 300 μL is available for each condition). A 100
μL aliquot of cells was then plated onto 96-well plates precoated with
growth factor reduced Matrigel (50 μL for 30 min followed by a
further 30 μL for 1 h) and incubated for 18 h. Tube formation was
examined by phase-contrast microscopy, and images were collected
using an Olympus C5050 digital camera. Tube formation inhibition
was quantified manually from images by recording the total number of
branching nodes connecting three or more tubules. Results are
expressed as percentage inhibition compared to control (untreated
HUVECs were used as a control for normal cell growth and tube
formation in Matrigel).
Ex Vivo Rat Aortic Angiogenesis Assay. The ex vivo rat aortic
angiogenesis assay was carried out as previously described.16,29 Briefly,
rat aortas were harvested from 6 to 10 week old Sprague−Dawley rats
by Tetra-Q, Brisbane, Australia, and trimmed of remaining fat and
connective tissue before cutting 1 mm ring sections which were
subsequently bisected and transferred to complete EBM-2 medium
(Lonza, Basel, Switzerland) containing 2% FCS and all SingleQuot
reagents except for heparin. Matrigel (BD Biosciences, San Jose, CA,
U.S.) was allowed to cool on ice, and once in a liquid form, 180 μL was
pipetted into 48-well tissue culture plates (Nunc, Rochester, NY, U.S.).
The plates were incubated at 37 °C for 30 min to allow Matrigel to
solidify. Aortic segments were then carefully placed on top on the
Matrigel in the center of each well, and 60 μL of Matrigel was pipetted
to cover the aortic segment. The plate was then returned to the
incubator for a further 20 min before each well was supplemented with
1.0 mL of medium in the absence (control) or presence of test
compounds usually at two concentrations within the range of 1−50
μM (test compound stock solutions (10 mM) were prepared using
sterile phosphate buffered saline, pH 7.2), depending on the particular
compound/experiment. Cultures were replenished as appropriate
every 48 h, and scoring of microvessels (from 0 [least] to 5 [most
positive]) was carried out at various time points up to 8 days by two
investigators in an independent fashion as previously described.36
Sprouting vessels were photographed using the 4× objective with an
Olympus C-7070 camera and an adaptor for the eyepiece. In some
instances, to determine the potential toxicity of compounds in this
assay, the viability of the tissue was assessed by withdrawing the
compound/medium from the culture on day 6 or 7 and adding
complete medium with VEGF (typically 10 ng/mL) for up to an
additional 7 days. In the absence of toxicity, the viable tissue should
sprout microvessels in response to the exogenous growth factor.
Anticoagulant Activity Assays. The anticoagulant activity of the
test compounds was determined by measuring the effect of various
concentrations of compound (0−100 μg/mL in PBS) on the elevation
of the APTT of pooled normal human plasma. Clotting assays were
performed with a ball coagulameter KC10 (Amelung, Lemgo,
Germany) using standard protocols according to the manufacturer’s
instructions. Unfractionated heparin (UFH) was used as the control.
The normal range of APTT for pooled normal human plasma is 26−
36 s. The coefficient of variation for APTT coagulation values of 36 s is
2%, and for coagulation values of 90 s and higher it is 5−7%. The
compounds were also evaluated in the Heptest as previously
described.16 Heptest values in normal plasma samples ranged from
Data Analysis. Plasma concentration versus time curves were
plotted (see Supporting Information), and apparent pharmacokinetic
parameters were estimated from the notionally linear portion of the
curve. The apparent terminal elimination half-lives (t1/2) for 18, 23,
and 24 were estimated to be 7.19, 5.86, and 5.89 h, respectively.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures and characterization data for selected
compounds; tables of 1H and 13C NMR chemical shift
assignments for compounds 33 and 18; plots of mean plasma
1
concentration versus time for compounds 18, 23, and 24; H
NMR spectra for compounds 30, 32, 33, and 18; capillary
electropherograms for sulfonated test compounds. This ma-
terial is available free of charge via the Internet at http://pubs.
acs.org.
AUTHOR INFORMATION
■
Corresponding Author
13 to 20 s (mean, 17.1
2.1 s). The coefficient of variation for
*Phone, +617-3346 9598; fax, +617-3365 4299; e-mail, v.ferro@
uq.edu.au.
Heptest coagulation values is 1.8% for values of 20−90 s and 3−4% for
values above 90 s.
3811
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
(12) McKenzie, E. A. Heparanase: a target for drug discovery in
cancer and inflammation. Br. J. Pharmacol. 2007, 151, 1−14.
(13) Parish, C. R.; Freeman, C.; Brown, K. J.; Francis, D. J.; Cowden,
W. B. Identification of sulfated oligosaccharide-based inhibitors of
tumor growth and metastasis using novel in vitro assays for
angiogenesis and heparanase activity. Cancer Res. 1999, 59, 3433−
3441.
(14) Liu, C. J.; Lee, P. H.; Lin, D. Y.; Wu, C. C.; Jeng, L. B.; Lin, P.
W.; Mok, K. T.; Lee, W. C.; Yeh, H. Z.; Ho, M. C.; Yang, S. S.; Lee, C.
C.; Yu, M. C.; Hu, R. H.; Peng, C. Y.; Lai, K. L.; Chang, S. S.; Chen, P.
J. Heparanase inhibitor PI-88 as adjuvant therapy for hepatocellular
carcinoma after curative resection: a randomized phase II trial for
safety and optimal dosage. J. Hepatol. 2009, 50, 958−968.
(15) Recent review: Kudchadkar, R.; Gonzalez, R.; Lewis, K. D. PI-
88: a novel inhibitor of angiogenesis. Expert Opin. Invest. Drugs 2008,
17, 1769−1776.
Present Addresses
§School of Chemistry and Molecular Biosciences, The University of
Queensland, Brisbane, Queensland 4072, Australia.
∥Institute for Molecular Bioscience, The University of Queens-
land, Brisbane, Qld 4072, Australia.
⊥National Food Technology Research Centre, Kanye, Botswana.
#Alchemia Ltd., Eight Mile Plains, Queensland 4113, Australia.
∞QLD Health Forensic and Scientific Services, Coopers Plains,
Queensland 4108, Australia.
¶Eskitis Institute for Cell and Molecular Therapies, Griffith
University, Nathan Qld 4111, Australia.
Notes
The authors declare no competing financial interest.
(16) Johnstone, K. D.; Karoli, T.; Liu, L.; Dredge, K.; Copeman, E.;
Li, C. P.; Davis, K.; Hammond, E.; Bytheway, I.; Kostewicz, E.; Chiu,
F. C.; Shackleford, D. M.; Charman, S. A.; Charman, W. N.;
Harenberg, J.; Gonda, T. J.; Ferro, V. Synthesis and biological
evaluation of polysulfated oligosaccharide glycosides as inhibitors of
angiogenesis and tumor growth. J. Med. Chem. 2010, 53, 1686−1699.
(17) Wall, D.; Douglas, S.; Ferro, V.; Cowden, W.; Parish, C.
Characterisation of the anticoagulant properties of a range of
structurally diverse sulfated oligosaccharides. Thromb. Res. 2001, 103,
325−335.
(18) Karoli, T.; Liu, L.; Fairweather, J. K.; Hammond, E.; Li, C. P.;
Cochran, S.; Bergefall, K.; Trybala, E.; Addison, R. S.; Ferro, V.
Synthesis, biological activity, and preliminary pharmacokinetic
evaluation of analogues of a phosphosulfomannan angiogenesis
inhibitor (PI-88). J. Med. Chem. 2005, 48, 8229−8236.
(19) Ferro, V.; Dredge, K.; Liu, L.; Hammond, E.; Bytheway, I.; Li,
C.; Johnstone, K.; Karoli, T.; Davis, K.; Copeman, E.; Gautam, A. PI-
88 and novel heparan sulfate mimetics inhibit angiogenesis. Semin.
Thromb. Hemostasis 2007, 33, 557−568.
(20) Yu, G.; Gunay, N. S.; Linhardt, R. J.; Toida, T.; Fareed, J.;
Hoppensteadt, D. A.; Shadid, H.; Ferro, V.; Li, C.; Fewings, K.;
Palermo, M. C.; Podger, D. Preparation and anticoagulant activity of
the phosphosulfomannan PI-88. Eur. J. Med. Chem. 2002, 37, 783−
791.
(21) Cochran, S.; Li, C.; Fairweather, J. K.; Kett, W. C.; Coombe, D.
R.; Ferro, V. Probing the interactions of phosphosulfomannans with
angiogenic growth factors by surface plasmon resonance. J. Med. Chem.
2003, 46, 4601−4608.
(22) Cochran, S.; Li, C. P.; Ferro, V. A surface plasmon resonance-
based solution affinity assay for heparan sulfate-binding proteins.
Glycoconjugate J. 2009, 26, 577−587.
(23) Hammond, E.; Li, C. P.; Ferro, V. Development of a
colorimetric assay for heparanase activity suitable for kinetic analysis
and inhibitor screening. Anal. Biochem. 2010, 396, 112−116.
(24) Note that this assay uses fondaparinux as substrate rather than
HS and no BSA carrier to stabilize the enzyme. The resultant Ki values
are approximately 20-fold lower than with HS-based assays which
require BSA and result in compound loss due to binding to BSA.
(25) Dredge, K.; Hammond, E.; Davis, K.; Li, C. P.; Liu, L.;
Johnstone, K.; Handley, P.; Wimmer, N.; Gonda, T. J.; Gautam, A.;
Ferro, V.; Bytheway, I. The PG500 series: novel heparan sulfate
mimetics as potent angiogenesis and heparanase inhibitors for cancer
therapy. Invest. New Drugs 2010, 28, 276−283.
ACKNOWLEDGMENTS
■
We thank Tony Carroll for high field NMR analysis of 18 and
33, Karine Mardon, Russell Addison, Edmund Kostewicz,
William Charman, and Tom Gonda for assistance with
pharmacokinetic and in vivo studies, and David Hume, Keith
Watson, Ron Dickinson, and Jeremy Turnbull for useful
discussions.
ABBREVIATIONS USED
■
HS, heparan sulfate; VEGF, vascular endothelial growth factor;
FGF, fibroblast growth factor; HUVEC, human umbilical vein
endothelial cell; dHMVEC, dermal human microvascular
endothelial cell; APTT, activated partial thromboplastin time;
PBS, phosphate buffered saline; TKI, tyrosine kinase inhibitor
REFERENCES
■
(1) Ferrara, N.; Kerbel, R. S. Angiogenesis as a therapeutic target.
Nature 2005, 438, 967−974.
(2) Folkman, J. Angiogenesis: an organizing principle for drug
discovery? Nat. Rev. Drug Discovery 2007, 6, 273−286.
(3) Herbst, R. S. Therapeutic options to target angiogenesis in
human malignancies. Expert Opin. Emerging Drugs 2006, 11, 635−650.
(4) Bergers, G.; Hanahan, D. Modes of resistance to anti-angiogenic
therapy. Nat. Rev. Cancer 2008, 8, 592−603.
(5) Casanovas, O.; Hicklin, D. J.; Bergers, G.; Hanahan, D. Drug
resistance by evasion of antiangiogenic targeting of VEGF signaling in
late-stage pancreatic islet tumors. Cancer Cell 2005, 8, 299−309.
(6) Paez-Ribes, M.; Allen, E.; Hudock, J.; Takeda, T.; Okuyama, H.;
Vinals, F.; Inoue, M.; Bergers, G.; Hanahan, D.; Casanovas, O.
Antiangiogenic therapy elicits malignant progression of tumors to
increased local invasion and distant metastasis. Cancer Cell 2009, 15,
220−231.
(7) Ebos, J. M.; Lee, C. R.; Cruz-Munoz, W.; Bjarnason, G. A.;
Christensen, J. G.; Kerbel, R. S. Accelerated metastasis after short-term
treatment with a potent inhibitor of tumor angiogenesis. Cancer Cell
2009, 15, 232−239.
(8) Loges, S.; Mazzone, M.; Hohensinner, P.; Carmeliet, P. Silencing
or fueling metastasis with VEGF inhibitors: antiangiogenesis revisited.
Cancer Cell 2009, 15, 167−170.
(9) Force, T.; Kolaja, K. L. Cardiotoxicity of kinase inhibitors: the
prediction and translation of preclinical models to clinical outcomes.
Nat. Rev. Drug Discovery 2011, 10, 111−126.
(10) Casu, B.; Vlodavsky, I.; Sanderson, R. D. Non-anticoagulant
heparins and inhibition of cancer. Pathophysiol. Haemostasis Thromb.
2008, 36, 195−203.
(26) Fairweather, J. K.; Hammond, E.; Johnstone, K. D.; Ferro, V.
Synthesis and heparanase inhibitory activity of sulfated mannooligo-
saccharides related to the antiangiogenic agent PI-88. Bioorg. Med.
Chem. 2008, 16, 699−709.
(27) Tornøe, C. W.; Christensen, C.; Meldal, M. Peptidotriazoles on
solid phase: [1,2,3]-triazoles by regiospecific copper(I)-catalyzed 1,3-
dipolar cycloadditions of terminal alkynes to azides. J. Org. Chem.
2002, 67, 3057−3064.
(28) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. A step-
wise Huisgen cycloaddition process: copper(I)-catalyzed regioselective
(11) Hammond, E.; Bytheway, I.; Ferro, V. Heparanase as a Target
for Anticancer Therapeutics: New Developments and Future
Prospects. In New Developments in Therapeutic Glycomics; Delehedde,
M., Lortat-Jacob, H., Eds.; Research Signpost: Trivandrum, India,
2006; pp 251−282.
3812
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813

Journal of Medicinal Chemistry
Article
“ligation” of azides and terminal alkynes. Angew. Chem., Int. Ed. 2002,
41, 2596−2599.
(29) Liu, L.; Johnstone, K.; Fairweather, J. K.; Dredge, K.; Ferro, V.
An improved synthetic route to the potent angiogenesis inhibitor
benzyl manα(1→3)-manα(1→3)-manα(1→3)-manα(1→2)-man hex-
adecasulfate. Aust. J. Chem. 2009, 62, 546−552.
(30) Subsequent treatment with NaOMe in MeOH to remove acyl
groups prior to sulfonation failed to remove the C-12 acetyl group
from deoxycholic acid.
́
(31) Reviews: (a) Imberty, A.; Lortat-Jacob, H.; Perez, S. Structural
view of glycosaminoglycan−protein interactions. Carbohydr. Res. 2007,
342, 430−439. (b) Coombe, D. R.; Kett, W. C. Heparan sulfate−
protein interactions: therapeutic potential through structure−function
insights. Cell. Mol. Life Sci. 2005, 62, 410−424. (c) Raman, R.;
Sasisekharan, V.; Sasisekharan, R. Structural insights into biological
roles of protein−glycosaminoglycan interactions. Chem. Biol. 2005, 12,
267−277. (d) Capila, I.; Linhardt, R. J. Heparin−protein interactions.
Angew. Chem., Int. Ed. 2002, 41, 391−412.
(32) Auerbach, R.; Lewis, R.; Shinners, B.; Kubai, L.; Akhtar, N.
Angiogenesis assays: a critical overview. Clin. Chem. 2003, 49, 32−40.
(33) Dredge, K.; Hammond, E.; Handley, P.; Gonda, T. J.; Smith, M.
T.; Vincent, C.; Brandt, R.; Ferro, V.; Bytheway, I. PG545, a dual
heparanase and angiogenesis inhibitor, induces potent anti-tumour and
anti-metastatic efficacy in preclinical models. Br. J. Cancer 2011, 104,
635−642.
(34) Narumi, A.; Miura, Y.; Otsuka, I.; Yamane, S.; Kitajyo, Y.; Satoh,
T.; Hirao, A.; Kaneko, N.; Kaga, H.; Kakuchi, T. End-functionalization
of polystyrene by malto-oligosaccharide generating aggregation-
tunable polymeric reverse micelle. J. Polym. Sci., Part A: Polym. Chem
2006, 44, 4864−4879.
(35) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition
constant (K1) and the concentration of inhibitor which causes 50 per
cent inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol.
1973, 22, 3099−3108.
(36) Min, J. K.; Han, K. Y.; Kim, E. C.; Kim, Y. M.; Lee, S. W.; Kim,
O. H.; Kim, K. W.; Gho, Y. S.; Kwon, Y. G. Capsaicin inhibits in vitro
and in vivo angiogenesis. Cancer Res. 2004, 64, 644−651.
3813
dx.doi.org/10.1021/jm201708h | J. Med. Chem. 2012, 55, 3804−3813