Synthesis, biological activity, and preliminary pharmacokinetic evaluation of analogues of a phosphosulfomannan angiogenesis inhibitor (PI-88)

Article abstract of DOI:10.1021/jm050618p

The phosphosulfomannan 1 (PI-88) is a mixture of highly sulfated oligosaccharides that is currently undergoing clinical evaluation in cancer patients. As well as its anticancer properties, 1 displays a number of other interesting biological activities. A

Full text of DOI:10.1021/jm050618p

J. Med. Chem. 2005, 48, 8229-8236
8229
Synthesis, Biological Activity, and Preliminary Pharmacokinetic Evaluation of
Analogues of a Phosphosulfomannan Angiogenesis Inhibitor (PI-88)
Tomislav Karoli,† Ligong Liu,† Jon K. Fairweather,† Edward Hammond,† Cai Ping Li,† Siska Cochran,†
Kicki Bergefall,‡ Edward Trybala,‡ Russell S. Addison,§ and Vito Ferro*,†
Drug Design Group, Progen Industries Limited, Brisbane, Queensland, Australia, Department of Clinical Virology,
Go¨teborg University, Go¨teborg, Sweden, and Centre for Studies in Drug Disposition, Department of Medicine,
The University of Queensland, Royal Brisbane Hospital, Queensland, Australia
Received June 30, 2005
The phosphosulfomannan 1 (PI-88) is a mixture of highly sulfated oligosaccharides that is
currently undergoing clinical evaluation in cancer patients. As well as its anticancer properties,
1 displays a number of other interesting biological activities. A series of analogues of 1 were
synthesized with a single carbon (pentasaccharide) backbone to facilitate structural charac-
terization and interpretation of biological results. In a fashion similar to 1, all compounds were
able to inhibit heparanase and to bind tightly to the proangiogenic growth factors FGF-1, FGF-
2, and VEGF. The compounds also inhibited the infection of cells and cell-to-cell spread of
herpes simplex virus (HSV-1). Preliminary pharmacokinetic data indicated that the compounds
displayed different pharmacokinetic behavior compared with 1. Of particular note was the
n-octyl derivative, which was cleared 3 times less rapidly than 1 and may provide increased
systemic exposure.
Introduction
The antiangiogenic phosphosulfomannan PI-88 (1,
Figure 1)1,2 is a promising inhibitor of tumor growth and
metastasis1,3,4 and is currently under evaluation in
phase II clinical trials in cancer patients.5 Compound 1
exerts antiangiogenic effects by antagonizing the inter-
actions of proangiogenic growth factors (FGF-1, FGF-
2, and VEGF) and their receptors with heparan sulfate
(HS).1,6 Compound 1 is also a potent inhibitor of
heparanase, a glycosidase that cleaves the HS side
chains of proteoglycans and plays an important role in
metastasis and angiogenesis.7-9 In addition to its an-
ticancer effects, 1 inhibits the blood coagulation cas-
cade,10-12 blocks vascular smooth muscle cell prolifera-
Figure 1. Structures of compounds 1-3.
tion and intimal thickening,13 inhibits herpes simplex
indicate that the dominant pentasaccharide component
is also the most biologically active and that the terminal
6-O-phosphate group can be replaced by a sulfo group
without detriment to biological activity.6,14,17 The com-
pounds were further simplified by making them ano-
merically pure, although the final products were still
mixtures of highly but incompletely sulfated forms, as
has been well established for compounds of this
type.1,6,17-19 The synthetic modifications were specifi-
cally chosen with a view to altering the pharmacokinetic
properties of the analogues in a favorable manner
compared with the parent compound20,21 to possibly
allow for less frequent dosing while biological activity
is maintained.
virus (HSV) infection of cells and cell-to-cell spread of
HSV-1 and HSV-2,14 and reduces proteinuria in a model
of passive Heymann nephritis.15
Compound 1 is a mixture of highly sulfated, mono-
phosphorylated mannose oligosaccharides ranging in
size from di- to hexasaccharide.16,17 The major compo-
nents are penta- (∼60%) and tetrasaccharides (∼30%).
While 1 is a promising clinical candidate,5 the fact that
it is a mixture complicates its characterization and the
assessment of structure-activity relationships. The
aims of this study were therefore to synthesize several
simple, easier to characterize analogues of 1 and to
determine if this could be accomplished while maintain-
ing biological activity. Pentasaccharide analogues where
the terminal 6-O-phosphate of 1 has been replaced by
sulfate were chosen for synthesis because a number of
studies of derivatives of the individual components of 1
Results and Discussion
Synthesis. The pentasaccharide R-D-Man-(1f3)-R-
D-Man-(1f3)-R-D-Man-(1f3)-R-D-Man-(1f2)-D-Man (2)
was selected as the starting point for the synthesis of
analogues of 1 via modification of the reducing end. The
types of modifications chosen were based on the intro-
duction of various groups that might be expected to alter
* Corresponding author. Phone: (+617) 3273 9150. Fax: (+617)
3375 6746. E-mail: vito.ferro@progen.com.au.
† Progen Industries Limited.
‡ Go¨teborg University.
§ The University of Queensland, Royal Brisbane Hospital.
10.1021/jm050618p CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/17/2005

8230 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Karoli et al.
alcohol was initially scaled up, a small amount of
powdered 3-Å molecular sieves was added into the
reaction mixture as a precaution to trap any moisture
before the addition of the promoter (BF3 etherate).
However, it was found that the molecular sieves dra-
matically slowed the reaction (24 h vs 2 h), while the
product profiles and yields were nearly identical to the
procedure excluding them. Both benzyl and n-octyl
glycosides (4 and 5) were isolated in moderate to good
yields (45-66%). Deprotection under Zemple´n condi-
tions followed by sulfonation and size exclusion chro-
matography provided the corresponding products 6 and
7 in 44% and 77% yield, respectively.
Unlike for the glycosylation of simple alcohols (benzyl
and n-octyl), the peracetate 3 was not a powerful enough
donor for condensation of the monomethyl ether of
PEG5000 using the conditions developed above. To obtain
the glycoside, a more reactive donor was required. The
glycosyl imidate 8 was an obvious choice and it was
prepared in a straightforward manner by anomeric
deacetylation with benzylamine in THF followed by
treatment with trichloroacetonitrile and DBU as a base.
A TMSOTf-promoted glycosylation proceeded well and
the PEGylated glycoside 9 was obtained in good yield
(80%). Compound 9 was deprotected to the polyol 10
(89%) and sulfonated to give the sulfated PEG5000-OMe
mannopentaoside 11 in 42% yield. In a similar fashion,
the sulfated PEG2000-OMe mannopentaoside 14 was
prepared in similar overall yield.
For the synthesis of the nitrogen-linked analogues,
peracetate 3 was reacted with trimethylsilyl azide and
tin(IV) chloride as catalyst to give the R-azide 15 in good
yield (82%).25 Azide 15 was subjected to the Staudinger
reaction in the presence of phenoxyacetyl chloride to
produce amide 16 in moderate yield. This application
of the Staudinger reaction is known26 to produce ano-
merization under certain conditions via ring opening of
the glycosylphosphazene intermediate. Little anomer-
ization was observed, but the contaminating triphenyl
phosphine oxide could not be removed by flash chroma-
tography. Zemple´n deacetylation and sulfonation pro-
ceeded smoothly and the contaminating materials were
easily separated from the desired sulfated compound 17,
obtained in 34% yield over two steps.
The biotinylated target 19 was chosen to serve two
tasks: to provide a compound based on the above-
mentioned design requirements (i.e., enhanced lipophi-
licity) and to provide an analogue suitable for immobi-
lization onto streptavidin-coated surfaces, such as a
BIAcore sensor chip, for potential assay development.
Literature precedent27 exists for the in situ preparation
of amides via reduction of an azide by hydrogen and
palladium in the presence of an N-hydroxysuccinimyl
(NHS) ester. In the present case, no reduction of the
azide was observed under the conditions described in
the literature (1 atm of H2, 10% Pd/C). Neither increas-
ing the H2 pressure to 100 psi nor the use of catalytic
transfer hydrogenation (NH4‚HCOO) was successful. A
stepwise approach28 was thus attempted. Hydrogena-
tion of the azide required the use of Adam’s catalyst
(PtO2) for 18 h at 100 psi of H2. After 1 h, the azide was
observed to be completely consumed, but the product
appeared to be a dimerized intermediate.29 The amine
condensed with biotinamidocaproate NHS ester in py-
Figure 2. Structures of compounds 4-19.
the pharmacokinetic properties compared to 1 itself,
such as lipophilic groups (e.g., n-octyl) and poly(ethylene
glycol) (PEG) groups.22 It was considered that such
modifications could be obtained either by glycosylation
of an appropriate alcohol or by acylation of an anomeric
amino group (introduced via a glycosyl azide). While 2
can be synthesized in a stepwise manner from D-
mannose,23 for the purposes of this study the starting
material was sourced from the neutral oligosaccharide
fraction obtained by mild acid-catalyzed hydrolysis of
the extracellular phosphomannan of the yeast Pichia
(Hansenula) holstii NRRL Y-2448.24 The pure pentasac-
charide, previously isolated by size exclusion chroma-
tography of the above neutral oligosaccharide fraction,24
was acetylated to give the peracetate 3. Alternatively,
3 could be isolated by flash chromatography following
acetylation of the neutral oligosaccharide fraction.
Various promoters (e.g., tin(IV) chloride, zinc chloride,
TMSOTf) were investigated for direct glycosylation of
peracetate 3 initially with benzyl alcohol as the glycosyl
acceptor. The best conditions were found to be boron
trifluoride etherate (4 equiv) in 1,2-dichloroethane at
60 °C, which cleanly gave complete conversion to
product within 2 h. This protocol was used for the larger
scale preparation of both the benzyl and n-octyl glyco-
sides 4 and 5 (Figure 2). When the reaction with benzyl

Analogues of the Angiogenesis Inhibitor PI-88
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26 8231
Table 1. Growth Factor Binding, Inhibition of Heparanase, Inhibition of HSV-1 Infectivity, and Inhibition of Cell-to-Cell Spread of
HSV-1 by Compound 1 and Analogues
Kd
IC50, µM
HSV-1
cell-to-cell
spread
FGF-1
(pM)
FGF-2
(nM)
VEGF
(nM)
heparanase
inhibn
HSV-1
infectivity
compd
1
6
281-444a
120 ( 25
144 ( 8
82-140a
86 ( 7
1.9-6.9a
1.7 ( 0.2
1.7 ( 0.1
8.1 ( 0.6
3.5 ( 0.8
7.1 ( 0.6
7.2 ( 0.6
0.98 ( 0.11
1.83 ( 0.48
1.64 ( 0.41
6.03 ( 1.05
2.12 ( 0.15
2.02 ( 0.28
1.85 ( 0.31
2b
1b
1
0.4
11
7
2
7
68 ( 3
1
11
14
17
19
361 ( 28
88 ( 17
150 ( 9
114 ( 13
112 ( 9
84 ( 8
not tested
10
7
2
660 ( 40
390 ( 70
5
3
a Data from ref 6. b Data from ref 14.
ridine at 60 °C for 3 days to produce the desired amide
18 in 36% overall yield. Zemple´n deacetylation and
sulfonation then gave the sulfated oligosaccharide 19
in moderate yield (61%).
Antiviral Studies. In addition to its anticancer
activities, 1 is a potent inhibitor of cell infection by
HSV.14 Sulfated polysaccharides such as heparin or
chondroitin sulfate E are also known inhibitors of the
early step(s) of viral infection of cells, such as viral
attachment and entry into cells. Their antiviral activi-
ties usually increase with increasing molecular weight,
but this feature has been reported to adversely affect
the ability of the polymer to penetrate into deep layers
of stratified tissue,30 thus limiting their potential anti-
viral application. In contrast to sulfated polysaccharides,
1 also efficiently inhibits the cell-to-cell spread of the
virus, most likely because its relatively small size
permits access to the narrow intercellular space.14 The
effects of compounds 1, 6, 7, 11, 14, 17, and 19 on
infection of cells by HSV-1 and on the cell-to-cell spread
of this virus are shown in Figure 3, parts A and B,
respectively. The IC50 values for these compounds
interpolated from the data are presented in Table 1. The
n-octyl derivative 7 inhibited viral spread and viral
infectivity similarly to 1, while the benzyl derivative 6
was only slightly less active than 1. The higher molec-
ular weight PEGylated derivatives 11 and 14 were
considerably weaker inhibitors of viral cell-to-cell spread.
This observation is in line with our previous finding14
that compounds of relatively high molecular weight
were poor inhibitors of HSV-1 cell-to-cell spread, most
likely because of their inability to enter the narrow
intercellular space.
Pharmacokinetic Studies. Compounds 6, 7, 11, 14,
and 19 were selected for preliminary pharmacokinetic
analysis in male Sprague-Dawley rats. These com-
pounds together with the parent compound 1 were
radiolabeled with 35S by sulfonating their respective
polyol precursors with 35SO3 pyridine complex.31 The
compounds were obtained in good chemical and radio-
chemical purity (as determined by HPLC; see Support-
ing Information). The sample of [35S]11 contained ∼45%
salt (by HPLC), but this was not associated with any
radioactivity, so the sample was considered suitable for
use. Compounds 1, 6, and 14 were obtained with good
specific activities (∼30 µCi/mg) while 7, 11, and 19 were
somewhat less active (∼6 µCi/mg), but still suitable for
the study.32 Six groups of rats (n ) 4 in each group)
were dosed intravenously with the radiolabeled com-
pounds (ca. 0.5-10 µCi/animal), plus the corresponding
nonlabeled compound, to a final dose of 2.5 mg/kg. Blood
samples were collected for determination of plasma
radioactivity at timed intervals over a 48-h period after
dosing. Urine and feces were also collected for deter-
The sulfonation of oligosaccharides larger than dis-
accharides with reagents such as sulfur trioxide pyri-
dine or trimethylamine complex is known to not go to
completion.1,6,17-19 Following literature precedent,1,6,17
the sulfonation of the polyol precursors of the target
compounds did not go to completion, and thus com-
pounds 6, 7, 11, 14, 17, and 19 were all obtained as
reproducible mixtures of highly but incompletely sul-
fated forms. Compounds 11 and 14 are even more
heterogeneous due to PEGylation. The purity of the
compounds was therefore determined by a combination
of size exclusion HPLC, capillary electrophoresis, and
1H NMR spectroscopy (see Experimental Section and
Supporting Information). All were of satisfactory purity
(78-100%), although compound 11 contained approxi-
mately 55% salt.
Heparanase Inhibition Studies. Compounds 6, 7,
11, 14, 17, and 19 were tested for their ability to inhibit
human platelet heparanase, and the results are pre-
sented in Table 1. The assays were performed using a
Microcon ultrafiltration assay,23 which relies on the
principle of physically separating HS that has been
digested by heparanase from native HS to determine
heparanase activity. The assay uses ultrafiltration
devices (Microcon YM-10) to separate the smaller
heparanase-cleaved HS fragments from native HS.
Compound 1, used as a standard in this assay, has an
IC50 ) 0.98 µM. All compounds tested inhibited hepara-
nase to a similar extent as 1 with IC50 values of ∼1-2
µM. The exception was the PEG5000 derivative 11, which
was approximately six times less active than 1 (IC50
6 µM).
)
Growth Factor Binding Studies. The binding
affinities of compounds 6, 7, 11, 14, 17, and 19 for the
proangiogenic growth factors FGF-1, FGF-2, and VEGF
were determined with a BIAcore (surface plasmon
resonance) solution affinity assay.6 It is important to
note that ligand binding to the growth factors can only
be detected when the interaction involves the HS
binding site, thus eliminating the chance of evaluating
nonspecific binding to other sites on the protein. A 1:1
stoichiometry was assumed for all protein:ligand inter-
actions. The results indicate that the compounds gener-
ally retain their affinities for the growth factors com-
pared with the parent compound and in some cases even
show a modest improvement (Table 1).

8232 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Karoli et al.
Conclusions. Several analogues of the anticancer
agent 1 were prepared and their biological activities
were evaluated. The analogues were based on a single
pentasaccharide backbone for ease of synthesis and to
facilitate their structural characterization and the
interpretation of biological results. The compounds
displayed similar activities to 1 in terms of their ability
to inhibit heparanase, to bind tightly to proangiogenic
growth factors, and to inhibit cell infection and cell-to-
cell spread of HSV-1, although the higher molecular
weight PEGylated derivatives were only weak inhibitors
of viral cell-to-cell spread. Of particular note was the
demonstration that it is possible to alter the pharma-
cokinetic properties of the analogues relative to 1 by
appropriate modifications at the reducing end of the
carbohydrate chain. The n-octyl derivative 7 is particu-
larly promising in this regard with early indications that
it is cleared less rapidly than 1. Taken together, the
results indicate that it is possible to prepare simple
analogues of 1 with improved pharmacokinetic proper-
ties that retain or enhance biological activity and that
are easier to synthesize and to characterize. Efforts are
now under way to prepare analogues with optimized
properties.
Experimental Section
General. General experimental details have been given
previously.6 Heparanase was isolated from human platelets
as described by Freeman and Parish.33 PI-88 (1) was prepared
as previously described.17 35SO3‚pyridine was from Amersham
Biosciences, UK. The purity of polysulfated products (see
Supporting Information) was determined by size exclusion
HPLC and/or capillary electrophoresis (CE) as described
previously.6
General Procedure for Deacetylation. A solution of the
peracetate in anhydrous MeOH (0.1 M) was treated with a
solution of NaOMe in MeOH (1.35 M, 0.2-0.6 equiv). The
mixture was stirred at room temperature for 1-3 h (monitored
by TLC). Acidic resin AG-50W-X8 (H+ form) was added to
adjust the pH to 6-7, the mixture was filtered and the resin
was rinsed with MeOH. The combined filtrate and washings
were concentrated in vacuo and dried to give the polyol
product.
Figure 3. Effect of compounds 1, 6, 7, 11, 14, 17, and 19 on
HSV-1 infectivity (A) and HSV-1 cell-to-cell spread (B). For
explanations, see the Experimental Section. In panel A, the
results are expressed as a percentage of the number of viral
plaque forming units (pfu) formed in cells infected with the
compound-treated virions relative to mock-treated controls. In
panel B, the results are expressed as a percentage of the
average area of 20 viral plaques formed in the continuous
presence of compound relative to mock-treated control cells.
General Procedure for Sulfonation. A mixture of the
polyol and SO3‚trimethylamine or SO3‚pyridine complex (2
equiv per alcohol) in DMF was heated (60 °C, overnight). The
cooled (room temperature) reaction mixture was treated with
MeOH and then made basic (to pH >10) by the addition of
Na2CO3 (10% w/w). The mixture was filtered and the filtrate
evaporated and coevaporated (H2O). The crude polysulfated
material was dissolved in H2O and subjected to size exclusion
chromatography (see below) to yield the sulfated product. After
lyophilization the product was passed through an ion-exchange
column (AG-50W-X8, Na+ form, 1 × 4 cm, deionized H2O, 15
mL) in order to transfer the product uniformly into the sodium
salt form. The solution collected was evaporated and lyophi-
lized to give the final product as a colorless glass or white
power.
Size Exclusion Chromatography. Size exclusion chro-
matography (SEC) was performed over Bio-Gel P-2 in a 5 ×
100 cm column and a flow rate of 2.8 mL/min of 0.1 M NH4-
HCO3, collecting 2.8 min (7.8 mL) fractions. Fractions were
analyzed for carbohydrate content by spotting onto silica gel
plates and visualization by charring, and/or analyzed for
polysulfated species by the dimethyl methylene blue (DMB)
test.34 Finally, fractions were checked for purity by CE,17 and
those deemed to be free of salt were pooled and lyophilized.
In cases where the presence of under sulfated byproducts or
other salt contaminants was detected (normally only in small
amounts), a Sephedex LH20 column chromatography step (2
mination of the extent of excretion of radioactivity over
this period. The data obtained are summarized in Table
2.
The log plasma concentration-time profiles were
curvilinear over the period 0-12 h. A notionally linear
period was observed between 0.75 and 4.0 h for all
compounds, except 11, and was used to calculate indica-
tive pharmacokinetic parameters (see footnote a in
Table 2) for comparison between the groups. Radiolabel
was rapidly eliminated from plasma following iv dosing
with all compounds studied, with half-lives of ap-
proximately 1 h or less being observed over the notion-
ally linear 0.75-4.0 h postdose interval. Clearance of
the label was most rapid for 14 (404 ( 59.5 mL/h/kg)
and slowest for 7 (83.6 ( 9.1 mL/h/kg). Compound 7 also
displayed the highest results for C0 (35.6 µg equiv/mL).
The overall results for 7 indicate that it is cleared three
times less rapidly compared with 1 and may have an
increased systemic exposure. Urinary excretion of ra-
diolabel accounted for 40-80% of the radiolabeled doses,
while excretion in the feces was low (approximately 1%
of the dose).

Analogues of the Angiogenesis Inhibitor PI-88
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26 8233
Table 2. Pharmacokinetic Parameters Determined for 35S-Labeled Compounds Following Iv Administration to Male
Sprague-Dawley Rats (mean values ( SD, n ) 4)
1
6
7
11
14
19
C0 (µg equiv/mL)
AUC0-12h
17.7 ( 2.23
9.6 ( 1.9
20.5 ( 1.3
12.6 ( 1.2
35.6 ( 3.1
29.7 ( 3.4
30.5 ( 2.3
14.7 ( 1.2
17.1 ( 1.8
6.2 ( 1.0b
14.0 ( 0.84
6.5 ( 0.4c
(µg equiv‚h/mL)
t1/2a,d (h)
0.83 ( 0.09
0.844 ( 0.096
250 ( 27.6
43.1 ( 1.9
0.83 ( 0.02
0.836 ( 0.024
199 ( 13.2
38.4 ( 3.8
1.10 ( 0.09
0.633 ( 0.053
83.6 ( 9.1
22.9 ( 2.2
41.8 ( 1.5
2.81 ( 0.04
0.247 ( 0.003
172 ( 11.8
24.9 ( 2.9
0.59 ( 0.01
1.17 ( 0.024
404 ( 59.5
44.5 ( 4.5
79.1 ( 3.6
0.79 ( 0.03
0.879 ( 0.028
380 ( 24.3c
55.1 ( 2.6
ka (h-1
)
Cla (mL/h/kg)
Vda (mL)
urinary recovery
(0-48 h) (% dose)
59.1 ( 13.1
39.3 ( 5.5
66.5 ( 9.4
80.5 ( 3.9
a Apparent values. b Calculated over 0-8 h postdose interval only. c Calculated over 0-4 h postdose interval only. d Calculated over
the 0.75-4.0 h postdose interval for 1, 6, 7, 14 and 19; calculated over the 4.0-12 h postdose interval for 11.
× 95 cm, deionized water, 1.2 mL/min, 3.5 min/fraction) was
applied to remove them completely.
75.6, 75.5(3), 75.4(8), 74.4, 73.8, 73.1, 73.0, 72.8, 72.7, 71.8,
71.3, 70.7, 70.6, 70.4, 69.9, 69.8, 69.7, 68.0, 67.8, 67.5, 66.6,
66.3(7), 66.3(5).
Peracetate 3. (a) The pentasaccharide R-D-Man-(1f3)-R-
D-Man-(1f3)-R-D-Man-(1f3)-R-D-Man-(1f2)-D-Man (2) (1.03
g, 95% by HPLC), previously isolated by size exclusion chro-
matography of the neutral fraction obtained from the mild
acid-catalyzed hydrolysis of the extracellular phosphomannan
from P. holstii NRRL Y-2448 according to the literature
procedure,24 was acetylated using sodium acetate (1.2 g) in
acetic anhydride (50 mL) at 140 °C for 18 h. After normal
workup the crude product was purified by flash chromatog-
raphy (EtOAc-hexane, 4:1) to give the peracetylated pen-
tasaccharide 3 as a colorless gum (810 mg): 1H NMR (400
MHz, CDCl3) δ 6.14 (d, 0.84H, J ) 2.0, RH1I), 5.71 (d, 0.16H,
J ) 0.9, âH1I), 5.30-5.10 (m, 8H), 5.00-4.85 (m, 7H), 4.25-
3.70 (m, 19H), 2.20-1.90 (m, 51H); HRMS calcd for C64H87O43
[M + H]+ 1543.4623, found 1543.4599.
(b) Alternatively, the neutral fraction from the mild acid-
catalyzed hydrolysis of the extracellular phosphomannan from
P. holstii NRRL Y-2448 was directly acetylated (excess Ac2O/
pyridine/DMAP) at room temperature for 3 days. The per-
acetylated pentasaccharide 3 was then isolated by flash
chromatography as described in method a above.
Octyl Glycoside 5. The peracetate 3 was glycosylated with
octanol as described above for 4 to give the product (5) as a
colorless gum (207 mg, 66%): Rf ) 0.41, hexanes-EtOAc, 1:3;
1H NMR (CDCl3, 400 MHz) δ 5.23-5.09 (m, 8 H), 4.96-4.82
(m, 8 H), 4.23-3.71 (m, 19 H), 3.59 (dt, 1 H, J ) 9.4, 6.8 Hz,
OCH2R), 3.35 (dt, 1 H, J ) 9.4, 6.8, OCH2R), 2.11, 2.10(2), 2.09-
(8), 2.06, 2.05, 2.04(4), 2.04(1), 2.03(8), 2.03, 2.02, 2.01, 1.99-
(3), 1.98(8), 1.96, 1.94, 1.90 (16s, 48 H, 16 × Ac), 1.52 (quintet,
2 H, J ) 7.2 Hz, CH2), 1.27-1.18 (m, 10 H, (CH2)5), 0.80 (t, 3
H, J ) 7.2 Hz, CH3); 13C NMR (CDCl3, 100 MHz) δ 170.4(0)
(×2), 170.3(8) (×2), 170.3, 170.2, 170.1, 169.9 (×2), 169.8(2),
169.7(5), 169.6, 169.5, 169.4(4), 169.3(5), 169.3, 99.1 (×2), 98.8,
98.7, 98.0, 77.0, 75.0, 74.8(3), 74.7(5), 71.0, 70.8, 70.7, 70.1,
69.4(9), 69.4(7), 69.3(0), 69.2(7), 69.2, 68.3, 68.2(0), 68.1(6),
67.2, 66.6(4), 66.6(0), 66.1, 65.4, 62.4, 62.3, 61.8, 61.5, 31.5,
29.1, 29.0, 28.9, 25.9, 22.4, 20.7(3), 20.7(0), 20.6(7), 20.6, 20.5,
20.4(3), 20.4(0), 20.3(9), 20.3(7), 13.8.
Octyl Glycoside Polysulfate 7. Compound 5 was deacety-
lated (HRMS calcd for polyol C38H69O26 [M + H]+ 941.40778,
found 941.4060) and sulfonated according to the general
procedures to give the product (7) as a white powder (195 mg,
72%): 1H NMR (D2O, 400 MHz) δ 5.33 (s, 1 H), 5.29 (d, 1 H,
J ) 1.6 Hz), 5.24 (d, 1 H, J ) 1.6 Hz), 5.21 (d, 1 H, J ) 1.6
Hz), 5.03 (dd, 1 H, J ) 2.8, 2.0 Hz), 4.87 (d, 1 H, J ) 1.6 Hz),
4.86-4.83 (m, 2 H), 4.70-3.92 (m, 27 H), 3.59 (dt, 1 H, J )
9.6, 7.0 Hz), 3.44 (dt, 1 H, J ) 9.6, 7.0 Hz), 1.48-1.40 (m, 2
H), 1.21-1.08 (m, 10 H), 0.68 (t, 3 H, J ) 7.2 Hz); 13C NMR
(D2O, 100 MHz) δ 100.5, 100.4, 100.1, 100.0, 99.0, 98.4(1), 98.3-
(8), 98.3(6), 98.3(5), 76.8(5), 76.7(9), 76.7, 76.6, 76.5(2), 76.4-
(7), 76.0, 75.4(0), 75.3(5), 75.3, 75.2, 74.3, 73.0(5), 72.9(9), 72.7,
72.6, 71.7, 70.4, 70.2, 69.8(4), 69.7(5), 69.6, 69.1, 67.8(5), 67.7-
(7), 66.5, 66.2, 31.5, 30.0, 28.8, 25.8, 22.5, 14.0.
Trichloroacetimidate 8. (a) A mixture of the peracetate
3 (68 mg, 51 µmol) and BnNH2 (17 µL, 152 µmol) in THF (2
mL) was stirred (room temperature) for 2 d. The mixture was
diluted with CHCl3 (20 mL) and subjected to workup. The
organic phase was evaporated and coevaporated (2 × 10 mL
of MeCN) and used in the following reaction without further
purification. (b) DBU (10 µL, 6.7 µmol) was added to a solution
of the crude product from step a and trichloroacetonitrile
(1.0 mL, 10 mmol) in 1,2-DCE (4 mL), and the combined
mixture was stirred (0 °Cf12 °C, overnight). The mixture was
concentrated and the residue subjected to flash chromatogra-
phy (50f90% EtOAc/hexanes) to yield 8 as a pale yellow oil
(35 mg, 48%, two steps): 1H NMR (400 MHz, CDCl3) δ 8.70
(s, 1 H, NH), 6.32 (d, 1 H, J ) 2.0, H1I), 5.36-5.13 (m, 8 H),
5.00-4.90 (m, 6 H), 4.26-3.75 (m, 20 H), 2.15-1.94 (m, 48
H).
PEG5000 Polysulfate 11. (a) A mixture of the imidate 8 (33
mg, 20.2 µmol) and PEG5000-monomethyl ether (151 mg, 30.3
µmol) in 1,2-DCE (3 mL) was stirred in the presence of
molecular sieves (50 mg of 3-Å powder) under an atmosphere
of argon (10 min). The mixture was cooled (-20 °C) with
continuous stirring (10 min) prior to the addition of TMSOTf
(5 µL, 2.8 µmol). After 20 min, Et3N (10 µL) was introduced
and the mixture was filtered. The solvent was evaporated and
Benzyl Glycoside 4. To a solution of the peracetate 3 (225
mg, 0.146 mmol) in dry 1,2-DCE (7.3 mL, 0.02 M) was added
benzyl alcohol (45 µL, 0.438 mmol). Boron trifluoride etherate
(37 µL, 0.292 mmol) was added and the mixture was stirred
under an atmosphere of argon at 60 °C for 2 h. The mixture
was cooled and triethylamine (244 µL, 1.75 mmol) was added.
The mixture was diluted with dichloromethane, washed with
saturated aqueous sodium carbonate, and dried (anhydrous
MgSO4). The dried solution was filtered and the filter cake
washed with dichloromethane. The combined filtrate and
washings were concentrated, loaded onto silica gel, and
purified by flash chromatography (hexanes-EtOAc, 6:1 f 1:4)
to give the benzyl glycoside 4 as a colorless gum (108 mg,
1
46%): Rf ) 0.32, hexanes-EtOAc, 1:3; H NMR (CDCl3, 400
MHz) δ 7.35-7.27 (m, 5 H, Ph), 5.30-5.12 (m, 8 H), 5.00-
4.85 (m, 8 H), 4.68, 4.50 (AB q, J ) 11.8 Hz, PhCH2), 4.27-
3.74 (m, 19 H), 2.14(4), 2.13(5), 2.13, 2.10, 2.08(4), 2.07(9),
2.07(6), 2.06(9), 2.06(6), 2.06 (×2), 2.02, 2.00, 1.99, 1.97, 1.94
(15s, 48 H, 16 × Ac); 13C NMR (CDCl3, 100 MHz) δ 171.0,
170.5(3), 170.5(1), 170.5(0), 170.4, 170.3, 170.2, 170.0(4), 170.0-
(2), 169.8(9), 169.8(8), 169.7, 169.6, 169.5(6), 169.4(6), 169.3,
136.1, 128.5, 128.2, 127.9, 99.2 (×2), 98.9, 98.8, 97.3, 76.7, 75.1,
74.9(9), 74.9(7), 71.1, 70.9, 70.8, 70.2, 69.7, 69.5(9), 69.5(6),
69.4(2), 69.3(7), 69.2, 68.6, 68.3, 67.1, 66.7(3), 66.6(7), 66.1,
65.5, 62.4, 62.1, 61.9, 61.6, 60.2, 20.9, 20.8(2), 20.8(0), 20.7(8),
20.7, 20.6, 20.5(4), 20.5(1), 20.4(9), 20.4(6).
Benzyl Glycoside Polysulfate 6. Compound 4 was deacety-
lated (HRMS calcd for polyol C37H59O26 [M + H]+ 919.3296,
found 919.3279) and sulfonated according to the general
procedures to give the product (6) as a white powder (76.1 mg,
44%): 1H NMR (D2O, 400 MHz) δ 7.35-7.26 (m, 5 H, Ph), 5.32
(s, 1 H), 5.30 (d, 1 H, J ) 1.2 Hz), 5.26 (d, 1 H, J ) 2.0 Hz),
5.24 (d, 1 H, J ) 1.6 Hz), 5.05 (dd, 1 H, J ) 2.8, 2.0 Hz), 5.00
(d, 1 H, J ) 2.0 Hz), 4.87-4.85 (m, 2 H), 4.68-4.34 (m, 12 H),
4.32-3.86 (m, 17 H); 13C NMR (D2O, 100 MHz) δ 137.0, 129.5,
129.4, 129.1, 100.5(9), 100.5(6), 100.2, 97.9, 93.8, 76.9, 76.8,

8234 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Karoli et al.
the residue subjected to flash chromatography (0-7.5% MeOH/
CHCl3) to yield the glycoside 9 as a colorless glass (104 mg,
80%, based on average Mr 6483): 1H NMR (400 MHz, CDCl3)
δ 5.28-4.87 (m, 14 H), 4.43-3.42 (m, 829 H,), 3.34 (s, 3 H,
OMe), 2.15-1.94 (m, 48 H).
(b) Compound 9 (104 mg, 16 µmol) was deacetylated
according to the general procedure to yield the polyol 10 as a
colorless wax (82 mg, 89%, based on average Mr 5769). This
residue was used in the next reaction without further purifica-
tion or characterization.
heated to 60 °C for 3 days with stirring. The solution was
evaporated and subjected to flash chromatography (Et3N
washed silica gel, EtOAc-hexane, 4:1 f MeOH-EtOAc, 3:7)
to give 30.8 mg (36% over two steps) of amide 18: 1H NMR
(400 MHz, CDCl3) δ 7.41 (br d, 1 H, J ) 9.4, NH), 6.47, 6.17
(2 × br s, 2 × 1 H, imide NHs), 5.40 (br d, 1 H, J ) 9.4, H1I),
5.40-4.90 (m, 16 H), 4.52 (dd, 1 H, J ) 4.9, 7.5, biotin-H4),
4.36-3.72 (m, 20 H), 3.25-3.12 (m, 3 H), 2.91 (dd, 1 H, J )
5.0, 13.0, biotin-H5A), 2.75 (d, 1 H, J ) 12.9, biotin-H5B),
2.27-1.96 (m, 52 H), 1.82-1.29 (m, 12 H, alkyl chains).
(c) The polyol 10 (82 mg, 14 µmol) was sulfonated according
to the general procedure to yield compound 11 as a colorless
foam (45 mg, 42%, based on average Mr 7401): 1H NMR (400
MHz, D2O) δ 5.34-4.87 (m, 7 H), 4.71-3.97 (m, 20 H), 3.76-
3.35 (m, 432 H), 3.23 (s, 3 H, OMe).
PEG2000 Polysulfate 14. (a) A mixture of the imidate 8 (60
mg, 36.5 µmol) and PEG2000-OMe (110 mg, 55.0 µmol) was
treated with TMSOTf as described for PEG5000-OMe to yield
the glycoside 12 as a colorless glass (96 mg, 74%): 1H NMR
(400 MHz, CDCl3) δ 5.28-5.13, 5.00-4.87, 4.27-3.40 (3 m,
H1I-V,2I-V,3I-V,4I-V,5I-V,6aI-V,6bI-V,OCH2CH2O), 3.34 (s, 3 H,
OMe), 2.15-1.94 (16 s, 3 H each, Ac).
Biotinamidocaproamide Polysulfate 19. The peracetate
18 (30 mg, 16.3 µmol) was deacetylated and sulfonated
according to the general procedures to yield 28 mg (61% for
two steps) of 19 after lyophilization: 1H NMR (400 MHz, D2O,
solvent suppressed, affected by amide rotamers) δ 5.60-4.75
(m, 7 H), 4.68 (dd, 1 H, J ) 4.7, 7.2 Hz, biotin-H4), 4.60-3.60
(m, 26 H), 4.21 (dd, 1 H, J ) 4.4, 7.2 Hz, biotin-H3), 3.33-
3.16 (m, 1 H, biotin-H2), 3.07-2.97 (m, 3 H, biotin-H5A +
CH2N), 2.92 (dd, 1 H, J ) 4.9, 13.5 Hz, biotin-H5B), 2.33-
2.14 (m, 2 H, COCH2B), 2.09 (t, 2 H, J ) 7.4 Hz, COCH2A),
1.63-1.15 (m, 12 H, alkyl chains).
Radiolabeling of Compounds. The polyol precursors for
1, 6, 7, 11, 14, and 19 (2 mg of each) were desiccated under
vacuum over P2O5 for 3 days. Into each vial was syringed 50
µL of a stock solution of 35SO3‚pyridine complex (1.77 mg, 2.0
mCi) and SO3‚Me3N (2 mg) in anhydrous DMF (300 µL, redried
over freshly activated 3-Å molecular sieves). A further 600 µL
of anhydrous DMF was added to the SO3 vial and was
distributed to each sample vial. The samples were heated to
60 °C for 6 h. SO3‚Me3N (14 mg in 300 µL of anhydrous DMF)
was added to each vessel, and the resulting solutions were
heated to 60 °C overnight. The vials were cooled to room
temperature and each sample was quenched by addition of
Na2CO3 (saturated aqueous adjusted to pH 8-9), evaporated
to dryness, and subjected to SEC (Bio-Gel P-2, 2.6 × 90 cm,
flow rate 30 mL/h, 5 min/fraction). Fractions containing the
desired material were detected using a G-M counter and DMB
test followed by CE. As anticipated, apart from the PEGylated
compounds, all samples were detected clearly by CE in the
expected fractions, and the fractions deemed to be free of
contaminants were pooled and lyophilized. The PEGylated
samples were pooled, based on the DMB assay and radioactiv-
ity of the fractions, and lyophilized. The specific activities and
the chemical and radiochemical purities (determined by HPLC
as previously described31) are presented in the Supporting
Information.
Heparanase Inhibition Assays. The heparanase assays
were performed using a Microcon ultrafiltration assay. A
reaction was set up with a volume of 100 µL, 40 mM acetate
buffer (pH 5.0), 0.1 mg/mL BSA, human platelet heparanase
(100 ng), 5 µM [3H]-labeled HS,33 and various concentrations
of inhibitors. The reactions were set up with all components
except the [3H]-labeled HS and allowed to equilibrate for 10
min at 22 °C. The assays were then initiated by adding the
HS, and immediately 20 µL was taken and mixed with 80 µL
of 10 mM phosphate (pH 7.0), and the 100 µL was transferred
to a Microcon YM-10 concentrator which was then centrifuged
at approximately 14 000g for 5 min. The solution that passed
through the membrane (filtrate) was retained. This sample
was considered the time ) 0 sample. The assays (now 80 µL
in volume) were allowed to react at 22 °C for 4 h and then the
filtration step was repeated for three aliquots of 20 µL from
each assay. The time ) 0 filtrate and the three 4-h filtrate
samples were counted for 3H. The difference between the time
) 0 and the averaged 4 h samples gave the amount of
heparanase activity. All inhibition assays were run with a
heparanase standard assay that was identical to the assay
composition above, except no inhibitor was present; the
amount of heparanase inhibition in the other assays was
determined by comparison with this standard. The IC50 for 1
in this assay was 0.98 ( 0.11 µM. The results are presented
in Table 1.
(b) The product from step a was deacetylated according to
the general procedure to yield, presumably, the polyol 13 as a
colorless wax (63 mg, 81%). This residue was used in the next
reaction without further purification or characterization.
(c) The product from step b was sulfonated according to the
general procedure to yield compound 14 as a colorless powder
(47 mg, 68%): 1H NMR (400 MHz, D2O) δ 5.34-3.97 (m, 498
H), 3.80-3.35 (m, 81 H), 3.23 (s, 3 H, OMe).
Azide 15. A solution of peracetate 3 (270 mg, 175 µmol),
TMSN3 (60 mg, 525 µmol), and SnCl4 (200 µL of 1 M in CH2-
Cl2) in anhydrous CH2Cl2 (20 mL) was stirred overnight in
the dark. Additional quantities (3 equiv) of TMSN3 and SnCl4
were added, and stirring was continued in the dark overnight
again. Ice and NaHCO3 (saturated aqueous) were added, and
the mixture was extracted with EtOAc, washed with brine,
evaporated, and subjected to flash chromatography (EtOAc-
hexane, 1:1 f 4:1) to yield azide 15 as a colorless oil (218 mg,
82%): 1H NMR (400 MHz, CDCl3) δ 5.52 (d, 1 H, J ) 2.0, H1I),
5.29-5.12 (m, 8 H), 5.02-4.87 (m, 7 H), 4.29-3.76 (m, 19 H),
2.18-1.95 (m, 48 H); 13C NMR (100 MHz, CDCl3) δ 170.5(9),
170.5(7), 170.5(6), 170.4, 170.3, 170.2, 170.1, 169.9(9), 169.9-
(8), 169.9(5), 169.7(3), 169.6(9), 169.6(6), 169.6, 169.5, 169.3,
99.3(0), 99.2(7), 99.1, 99.0, 88.1, 75.2, 75.1, 74.8, 71.1, 70.9,
70.8, 70.6, 69.7, 69.5, 69.4, 69.2, 68.3, 67.3, 66.8, 66.7, 65.5(9),
65.5(8), 62.6, 62.2, 62.0, 61.7, 20.8(8), 20.8(6), 20.8, 20.7, 20.6-
(2), 20.5(8), 20.5(7), 20.5; HRMS calcd for C62H84N3O41 [M +
H]+ 1526.4583, found 1526.4557.
Phenoxyacetamide 16. A solution of 15 (32 mg, 21 µmol),
PPh3 (11 mg, 42.6 µmol), and phenoxyacetyl chloride (7.3 mg,
43 µmol) in anhydrous acetonitrile (5 mL) was stirred at 0 °C
for 4 h then at room temperature overnight. EtOAc and
NaHCO3 (saturated aqueous) were added, and the organic
layer was washed with brine, dried (MgSO4), and subjected to
flash chromatography (EtOAc-hexane, 3:2 f 9:1) to yield 11.4
mg (33%) of amide 16 with some remaining PPh3/PPh3O: 1H
NMR (400 MHz, CDCl3) δ 7.36-7.32 (m, 2 H), 7.18 (br d, 1 H,
J ) 8.1, NH), 7.00-6.90 (m, 3 H), 5.79 (dd, 1 H, J ) 3.8, 8.2,
H1I), 5.32-4.97 (m, 15 H), 4.60-3.76 (m, 21 H), 2.20-1.95 (m,
48 H, AcO); HRMS calcd for C70H92NO43 [M + H]+ 1634.5045,
found 1634.5002.
Phenoxyacetamide Polysulfate 17. The peracetate 16 (11
mg, 6.7 µmol) was deacetylated and sulfonated according to
the general procedures to yield 6 mg (34% for 2 steps) of 17
after lyophilization: 1H NMR (400 MHz, D2O, solvent sup-
pressed) δ 7.30-7.21 (m, 2 H, ArHm), 6.96-6.84 (m, 3 H,
ArHo,p), 5.56-3.59 (m, 30 H affected by suppression).
Biotinamidocaproamide 18. A mixture of 15 (70 mg, 46
µmol) and Adam’s catalyst (2 mg) in EtOAc-EtOH (2:1, 3 mL)
was stirred under H2 (100 psi) overnight, filtered, evaporated,
and coevaporated with anhydrous pyridine. Biotinamidoca-
proate N-hydroxysuccinimide ester (31 mg, 68 µmol) and
anhydrous pyridine (1 mL) were added, and the mixture was
Growth Factor Binding Assays. Binding affinities of
ligands for the growth factors FGF-1, FGF-2, and VEGF were

Analogues of the Angiogenesis Inhibitor PI-88
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26 8235
measured using a surface plasmon resonance solution affinity
assay performed on a BIAcore 3000 (BIAcore, Uppsala, Swe-
den) operated using the BIAcore Control Software, as previ-
ously described.6 Sensorgram data were analyzed using the
BIAevaluation software (BIAcore). Background sensorgrams
were subtracted from experimental sensorgrams to produce
curves of specific binding, and baselines were subsequently
adjusted to zero for all curves. Standard curves relating the
relative response value to the injected protein concentration
were linear, indicating that the binding response was propor-
tional to the protein concentration, and thus suggesting that
the binding experiments were conducted under mass transport
conditions.35 Therefore, the relative binding response for each
injection can be converted to free protein concentration using
the equation
Pharmacokinetic Studies. Male Sprague-Dawley rats
(250-350 g) were obtained from the Herston Medical Research
Centre at Royal Brisbane Hospital. The animals were allowed
free access to food and water before and during the experi-
ments, during which they were maintained unrestrained in
metabolism cages. Animal procedures were approved by the
University of Queensland Group 5 Animal Ethics Committee.
Rats were anaesthetized 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 full range of
movement. The animals were carefully monitored during
recovery (1-4 h).
Stock dosing solutions were prepared by mixing appropriate
amounts of unlabeled and radiolabeled drug (dissolved in
phosphate-buffered saline) to give a total drug concentration
of 1.25 mg/mL. All doses were administered as a bolus
intravenous injection of 2.5 mg/kg in a dose volume of 2 mL/
kg. The total amount of radioactivity administered to each rat
was 0.5-10 µCi. The dose level used in this study is 10-fold
lower than the no-effect dose previously established for acute
toxicity of PI-88. Blood samples (∼250 µL) were collected
predose and at 5, 15, 30, and 45 min and 1, 1.5, 2, 4, 8, 12, 24,
36, and 48 h after dosing. The blood samples were immediately
centrifuged, and the plasma was collected. At completion of
the experiments, the animals were killed by a lethal overdose
of iv pentobarbitone anaesthetic (Nembutal). Urine was col-
lected from each animal at intervals of 0-12 h, 12-24 h, and
24-48 h after dosing. Cage washings (∼15 mL of deionized
water) were also collected. At the end of the experiment,
bladder contents were aspirated from each animal and added
to the 24-48 h voidings. Feces were collected over the same
time intervals as the urine.
r
[P] ) [P]total
rm
where r is the relative binding response and rm is the maximal
binding response.
Binding equilibria established in solution prior to injection
were assumed to be of 1:1 stoichiometry. Therefore, for the
equilibrium
P + L h P‚L
where P corresponds to the growth factor protein, L is the
ligand, and P‚L is the protein:ligand complex, the equilibrium
equation is
[P][L]
Kd )
[P·L]
and the binding equation6 can be expressed as
Aliquots of plasma (100 µL) and urine and cage washings
(500 µL) were transferred directly to 6-mL polypropylene
scintillation vials for determination of radioactivity. Feces
collected during each time period (from one animal dosed with
each compound) were weighed and homogenized in 4 volumes
of deionized water using a mechanical homogenizer. Ap-
proximately 1 g (accurately weighed) of this slurry was
transferred to a 20-mL glass scintillation vial, 2 mL of tissue
solubilizer was added, and the vials were capped and incubated
at 60 °C for at least 24 h. Radioactivity was measured following
mixing of samples with Packard Ultima Gold liquid scintilla-
tion counting cocktail (2.0 mL for plasma and dose, 5.0 mL
for urine and cage washings, 10 mL for feces). Counting was
conducted on a Packard Tri-Carb liquid scintillation counter.
Any result less than 3 times the background was considered
less than the lower limit of quantification and was not used
in calculations. Plasma, urine, and cage washings were
counted in triplicate within 5 days of collection and were not
corrected for radiochemical decay. Feces were processed as a
batch at the completion of the study and the counts from these
samples were corrected for radiochemical decay. Plasma
pharmacokinetic parameters were calculated using PK Solu-
tions 2.0 software (Summit Research Services) and are pre-
sented in Table 2.
(Kd + [L]total + [P]total
)
[P] ) [P]total
-
+
2
2
(Kd + [L]total + [P]total
)
- [L]total[P]total
x
4
The Kd values given are the values fitted, using the binding
equation, to a plot of [P] versus [L]total. Where Kd values were
measured in duplicate, the values represent the average of the
duplicate measurements. The results are presented in Table
1.
Antiviral Assays. Monolayer cultures of African green
monkey kidney cells36 and herpes simplex virus (HSV-1)
KOS321 strain37 were used throughout. The antiviral assays
for compounds 1, 6, 7, 11, 14, 17, and 19 were performed as
described previously.14 Briefly, the effects of the compounds
on the infection of cells by exogenously added virus were tested
by mixing serial 5-fold dilutions of compound (at 0.032-20 µM)
with approximately 200 plaque forming units of the virus.
Following incubation of the virus and compound for 10 min at
room temperature, the mixture was added to the cells and left
on the cell monolayer for 2 h at 37 °C. Subsequently, the
inoculum was aspirated and replaced with an overlay medium
of 1% methylcellulose solution in Eagle’s minimum essential
medium. The viral plaques that developed after incubation of
cells for 3 days at 37 °C were stained with 1% crystal violet
solution and counted. The effects of the compounds on cell-to-
cell spread of HSV-1 were tested by adding serial 5-fold
dilutions of compound (at 0.032-20 µM) in the serum-free
overlay medium to cells after their infection with HSV-1. After
incubation of the compound with the cells for 3 days at 37 °C,
the images of 20 plaques were captured and subjected to area
determination using IM500 software (Leica). The results for
viral infection of cells and for viral cell-to-cell spread are shown
in Figure 3, parts A and B, respectively, while the derived IC50
values are presented in Table 1. Note that compound 11 was
tested only in the cell-to-cell spread assay because of lack of
compound.
Acknowledgment. We thank Drs. Brian Creese and
Ian Bytheway for useful discussions and critical evalu-
ation of the manuscript. We also thank Prof. Ron
Dickinson for useful discussions and Andrew Wright
and Sarika Prasad for technical assistance. This work
was partially funded by an AusIndustry START grant
to Progen Industries Ltd.
Supporting Information Available: 1H and 13C NMR
spectra for selected new compounds. Tables of HPLC and CE
purity data for compounds 6, 7, 11, 14, 17,and 19 and chemical
and radiochemical purity and specific activity of radiolabeled

8236 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Karoli et al.
activity of the phosphosulfomannan PI-88. Eur. J. Med. Chem.
2002, 37, 783-791.
(18) Wessel, H. P.; Iberg, N. The smooth muscle cell antiproliferative
activity of heparan sulfate model oligosaccharides. Bioorg. Med.
Chem. Lett. 1996, 6, 427-430.
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(19) Wessel, H. P.; Chucholowski, A.; Fingerle, J.; Iberg, N.; Ma¨rki,
H. P.; Mu¨ller, R.; Pech, M.; Pfister-Downar, M.; Rouge, M.;
Schmid, G.; Tschopp, T. From glycosaminoglycans to heparinoid
mimetics with antiproliferative activity. Carbohydrate mimics:
Concepts and methods; Wiley-VCH: Chichester, UK, 1998; pp
417-431.
(20) Creese, B.; Ferro, V.; King, A. R.; Mardon, K.; Dickinson, R. G.;
Punler, M. J.; Dodds, H. Pharmacokinetics of [35S]PI-88 in male
rats. Proc. Aust. Soc. Clin. Exp. Pharmacol. Toxicol. 2001, 9,
57.
(21) Creese, B.; Ribbons, K.; Harvey, W. D.; McBurney, A.; Douglas,
S.; Fareed, J. Use of activated partial thromboplastin time
(APTT) as a surrogate measure of plasma drug concentration
in toxicity studies of PI-88, a novel anticancer and anticoagulant
drug. Toxicology 2001, 164, 232.
(1) 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.
(2) Khachigian, L. M.; Parish, C. R. Phosphomannopentaose sulfate
(PI-88): Heparan sulfate mimetic with clinical potential in
multiple vascular pathologies. Cardiovasc. Drug Rev. 2004, 22,
1-6.
(3) Joyce, J. A.; Freeman, C.; Meyer-Morse, N.; Parish, C. R.;
Hanahan, D. A functional heparan sulfate mimetic implicates
both heparanase and heparan sulfate in tumor angiogenesis and
invasion in a mouse model of multistage cancer. Oncogene 2005,
24, 4037-4051; Corrigendum; Oncogene 2005, 24, 4163.
(4) Iversen, P. O.; Sorenson, D. R.; Benestad, H. B. Inhibitors of
angiogenesis selectively reduce the malignant cell load in rodent
models of human myeloid leukemias. Leukemia 2002, 16, 376-
381.
(22) Harris, J. M.; Martin, N. E.; Modi, M. Pegylation: A novel
process for modifying pharmacokinetics. Clin. Pharmacokinet.
2001, 40, 539-551.
(23) Fairweather, J. K.; Hammond, E.; Ferro, V. Manuscript in
preparation.
(5) Ferro, V.; Don, R. The development of the novel angiogenesis
inhibitor PI-88 as an anticancer drug. Australas. Biotechnol.
2003, 13, 38-39.
(24) Ferro, V.; Fewings, K.; Palermo, M. C.; Li, C. Large-scale
preparation of the oligosaccharide phosphate fraction of Pichia
holstii NRRL Y-2448 phosphomannan for use in the manufac-
ture of PI-88. Carbohydr. Res. 2001, 332, 183-189.
(25) The stereoselective nature of this reaction was confirmed by the
absence of signals in the 1H and 13C NMR spectra attributable
to the â-anomer of 15, itself prepared via the displacement of
the corresponding peracetylated R-anomeric bromide with so-
dium azide.
(6) 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.
(7) Parish, C. R.; Freeman, C.; Hulett, M. D. Heparanase: A key
enzyme involved in cell invasion. Biochim. Biophys. Acta 2001,
1471, M99-M108.
(8) Ferro, V.; Hammond, E.; Fairweather, J. K. The development
of inhibitors of heparanase, a key enzyme involved in tumour
metastasis, angiogenesis and inflammation. Mini-Rev. Med.
Chem. 2004, 4, 159-165.
(26) Mizuno, M.; Muramoto, I.; Kobayashi, K.; Yaginuma, H.; Inazu,
T.
A
simple method for the synthesis of Nâ-glycosylated-
asparagine and -glutamine derivatives. Synthesis 1999, 162-
165.
(9) Simizu, S.; Ishida, K.; Osada, H. Heparanase as a molecular
target of cancer chemotherapy. Cancer Sci. 2004, 95, 553-558.
(10) Wall, D.; Douglas, S.; Ferro, V.; Cowden, W.; Parish, C. Char-
(27) Olsen, J. A.; Severinsen, R.; Rasmussen, T. B.; Hentzer, M.;
Givskov, M.; Nielsen, J. Synthesis of new 3- and 4-substituted
analogues of acyl homoserine lactone quorum sensing autoin-
ducers. Bioorg. Med. Chem. Lett. 2002, 12, 325-328.
(28) Ying, L.; Gervay-Hague, J. Synthesis of N-(fluoren-9-ylmethoxy-
carbonyl)glycopyranosylamine uronic acids. Carbohydr. Res.
2004, 339, 367-375.
(29) Paul, B.; Korytnyk, W. Synthesis of 2-acetamido-3,4,6-tri-O-
acetyl-2-deoxy-â-D-glucopyranosylamine and dimer formation.
Carbohydr. Res. 1978, 67, 457-468.
(30) Betz, G.; Nowbakht, P.; Imboden, R.; Imanidis, G. Heparin
penetration into and permeation through human skin from
aqueous and liposomal formulations in vitro. Int. J. Pharm.
2001, 228, 147-159.
acterisation of the anticoagulant properties of
a range of
structurally diverse sulfated oligosaccharides. Thromb. Res.
2001, 103, 325-335.
(11) Demir, M.; Iqbal, O.; Hoppensteadt, D. A.; Piccolo, P.; Ahmad,
S.; Schultz, C. L.; Linhardt, R. J.; Fareed, J. Anticoagulant and
antiprotease profiles of a novel natural heparinomimetic man-
nopentaose phosphate sulfate (PI-88). Clin. Appl. Thromb.
Hemost. 2001, 7, 131-140.
(12) Piccolo, P.; Iqbal, O.; Demir, M.; Ma, Q.; Gerbutavicius, R.;
Fareed, J. Global anticoagulant effects of a novel sulfated
pentomannan oligosaccharide mixture. Clin. Appl. Thromb.
Hemost. 2001, 7, 149-152.
(13) Francis, D. J.; Parish, C. R.; McGarry, M.; Santiago, F. S.; Lowe,
H. C.; Brown, K. J.; Bingley, J. A.; Hayward, I. P.; Cowden, W.
B.; Campbell, J. H.; Campbell, G. R.; Chesterman, C. N.;
Khachigian, L. M. Blockade of vascular smooth muscle cell
proliferation and intimal thickening after balloon injury by the
sulfated oligosaccharide PI-88: Phosphomannopentaose sulfate
directly binds FGF-2, blocks cellular signaling, and inhibits
proliferation. Circ. Res. 2003, 92, e70-77.
(14) Nyberg, K.; Ekblad, M.; Bergstro¨m, T.; Freeman, C.; Parish, C.
R.; Ferro, V.; Trybala, E. The low molecular weight heparan
sulfate-mimetic, PI-88, inhibits cell-to-cell spread of herpes
simplex virus. Antiviral Res. 2004, 63, 15-24.
(15) Levidiotis, V.; Freeman, C.; Punler, M.; Martinello, P.; Creese,
B.; Ferro, V.; van der Vlag, J.; Berden, J. H. M.; Parish, C. R.;
Power, D. A. A synthetic heparanase inhibitor reduces pro-
teinuria in passive Heymann nephritis. J. Am. Soc. Nephrol.
2004, 15, 2882-2892.
(16) Ferro, V.; Li, C.; Fewings, K.; Palermo, M. C.; Linhardt, R. J.;
Toida, T. Determination of the composition of the oligosaccharide
phosphate fraction of Pichia (Hansenula) holstii NRRL Y-2448
phosphomannan by capillary electrophoresis and HPLC. Car-
bohydr. Res. 2002, 337, 139-146.
(17) 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
(31) Ferro, V.; Li, C.; Wang, B.; Fewings, K.; King, A. R.; Hammond,
E.; Creese, B. R. Synthesis of [14C]- and [35S]-labelled PI-88 for
pharmacokinetic and tissue distribution studies. J. Labelled
Compd. Radiopharm. 2002, 45, 747-754.
(32) All the samples were treated identically so it is unclear as to
why this difference arose.
(33) Freeman, C.; Parish, C. R. Human platelet heparanase: Puri-
fication, characterization and catalytic activity. Biochem. J.
1998, 330, 1341-1350.
(34) Farndale, R. W.; Buttle, D. J.; Barrett, A. J. Improved quanti-
tation and discrimination of sulphated glycosaminoglycans by
use of dimethylmethylene blue. Biochim. Biophys. Acta 1986,
883, 173-177.
(35) Karlsson, R.; Roos, H.; Fa¨gerstam, L.; Persson, B. Kinetic and
Concentration Analysis Using BIA Technology. Methods: Com-
panion Methods Enzymol. 1994, 6, 99-110.
(36) Gunalp, A. Growth and cytopathic effect of rubella virus in a
line of Green Monkey kidney cells. Proc. Soc. Exp. Biol. Med.
1965, 118, 185-190.
(37) Holland, T. C.; Homa, F. L.; Marlin, S. D.; Levine, M.; Glorioso,
J. Herpes simplex virus type 1 glycoprotein C-negative mutants
exhibit multiple phenotypes, including secretion of truncated
glycoproteins. J. Virol. 1984, 52, 566-574.
JM050618P