Large-scale preparation of the oligosaccharide phosphate fraction of Pichia holstii NRRL Y-2448 phosphomannan for use in the manufacture of PI-88

Article abstract of DOI:10.1016/S0008-6215(01)00061-1

Mild acid-catalysed hydrolysis of the extracellular phosphomannan of the yeast Pichia holstii NRRL Y-2448 produces a high-molecular-weight phosphomannan core, a low-molecular-weight oligosaccharide phosphate fraction, and a neutral oligosaccharide fraction. A method was developed for the large-scale preparation of the oligosaccharide phosphate fraction, consisting predominantly of the pentasaccharide phosphate, 6-O-PO₃H₂-α-D-Man-(1 → 3)-α-D-man-(1 → 3)-α-D-man-(1 → 3)-α-D-man-(1 → 2)-D-man, for use in the manufacture of the promising new anti-cancer agent, PI-88. Further insights were also gained into the structure of the phosphomannan by HPLC analysis of the time course of the hydrolysis reaction.

Full text of DOI:10.1016/S0008-6215(01)00061-1

www.elsevier.nl/locate/carres
Carbohydrate Research 332 (2001) 183–189
Large-scale preparation of the oligosaccharide phosphate
fraction of Pichia holstii NRRL Y-2448 phosphomannan
for use in the manufacture of PI-88
Vito Ferro,* Kym Fewings, Maria C. Palermo, Caiping Li
Department of Research and De6elopment, Progen Industries Ltd, PO Box 28, Richlands BC, Qld 4077, Australia
Received 6 December 2000; accepted 19 February 2001
Abstract
Mild acid-catalysed hydrolysis of the extracellular phosphomannan of the yeast Pichia holstii NRRL Y-2448
produces a high-molecular-weight phosphomannan core, a low-molecular-weight oligosaccharide phosphate fraction,
and a neutral oligosaccharide fraction. A method was developed for the large-scale preparation of the oligosaccharide
phosphate fraction, consisting predominantly of the pentasaccharide phosphate, 6-O-PO3H2-a-
D
-Man-(1“3)-a-D-
Man-(1“3)-a- -Man-(1“3)-a- -Man-(1“2)- -Man, for use in the manufacture of the promising new anti-cancer
D
D
D
agent, PI-88. Further insights were also gained into the structure of the phosphomannan by HPLC analysis of the
time course of the hydrolysis reaction. © 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Pichia holstii; Phosphomannan; Oligosaccharide phosphate fraction; PI-88
1. Introduction
(1“6) to the terminal phosphate of the PC
and make up approximately 90% of the PS.
These side chains, which are capped by un-
phosphorylated residues,9,11 are susceptible to
mild acid hydrolysis of the phosphate diester
linkages thus giving ready access to both the
PC and 1. Indeed, the PC and low-molecular-
weight oligosaccharide preparations, previ-
ously assumed to consist solely of 1, have been
used extensively as tools and probes for study-
ing phosphomannosyl receptors.13–19
Recently, some light has been shed on the
nature of the phosphorylated side chains.12
The available data suggests that the majority
of these side chains are composed of up to at
least ten repeating pentasaccharide phosphate
residues. In the remaining side chains, some
pentasaccharide phosphate residues are re-
placed by tetrasaccharide phosphate and to a
much lesser extent by hexa-, tri- and disacchar-
The yeast Pichia (Hansenula) holstii NRRL
Y-2448, when grown aerobically in a nitrogen-
limited medium that uses
D
-glucose as the
carbon source and contains an excess of or-
thophosphate, produces an extracellular phos-
phomannan that has been the subject of
numerous studies.1–12 The phosphomannan
(PS) is composed of a highly branched, high-
molecular-weight (5–39×106) phosphoman-
nan core (PC) whose structure has only
recently been elucidated.11 Oligosaccharide
chains composed principally of the repeating
pentasaccharide phosphate 1, are linked a-
* Corresponding author. Tel.: +61-7-32739100; fax: +61-
7-33751168.
E-mail address: vito.ferro@progen.com.au (V. Ferro).
0008-6215/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(01)00061-1

184
V. Ferro et al. / Carbohydrate Research 332 (2001) 183–189
ide phosphates. There is also evidence that the
unphosphorylated units capping the side
chains are predominantly pentasaccharide and
disaccharide residues. In addition, it has been
shown that these low-molecular-weight
oligosaccharide preparations assumed to con-
sist solely of 1 also contain as much as 20–
25% of the tetrasaccharide phosphate 3, as
well as minor amounts of hexa-, tri- and disac-
charide phosphates. The presence of the phos-
phate group in these oligosaccharides prevents
their successful separation by size-exclusion
chromatography.12 Such a low-molecular-
weight oligosaccharide preparation (free of
neutral oligosaccharides) shall be referred to
as the oligosaccharide phosphate fraction or
OPF (Scheme 1).
The product of exhaustive sulfation of the
OPF, known as PI-88, has recently been iden-
tified as a promising inhibitor of tumour
growth and metastasis and is currently in
Phase I clinical trials.20 PI-88 exerts its an-
timetastatic effects by inhibiting hep-
aranase,21,22 an endoglucuronidase which
cleaves the heparan sulfate side chains of hep-
aran sulfate proteoglycans in the extracellular
matrix (ECM) surrounding tumour cells. The
degradation of the ECM facilitates the spread
of tumour cells by enabling them to both enter
into and escape from blood vessels and
lymphatics. Heparanase is also implicated in
angiogenesis, i.e., the growth of new blood
vessels from pre-existing blood vessels sur-
rounding tumours. Heparanase can act di-
rectly, by promoting invasion of endothelial
cells (vascular sprouting), and indirectly, by
releasing heparan sulfate-bound growth fac-
tors (e.g., bFGF) and generating heparan sul-
fate degradation fragments that promote
bFGF activity.21,22 PI-88 is thus also an in-
hibitor of angiogenesis by virtue of its inhibi-
tion of heparanase and by its ability to block
heparan sulfate binding of angiogenesis-induc-
ing growth factors.20 PI-88 also shows promise
as a potential anticoagulant/antithrombotic
agent with a novel mode of action.23 It en-
hances the ability of heparin cofactor II to
inhibit thrombin but does not interact with
AT-III.
Large quantities of clinical grade PI-88 are
required for ongoing pre-clinical studies (in-
cluding animal toxicology) and human clinical
trials, necessitating a reliable supply of the
OPF, preferably to good manufacturing prac-
tice (GMP) standards. No commercial pro-
duction of the OPF exists, and to date its
preparation has only been described on a lab-
oratory scale. These preparations usually in-
volve the isolation and purification by
differential ethanol precipitation of its barium
salt and are thus not easily scaleable, and the
product obtained is of questionable purity. A
Scheme 1.

V. Ferro et al. / Carbohydrate Research 332 (2001) 183–189
185
reliable method was thus sought for its prepa-
ration in multigram quantities that avoided
the use of toxic barium salts, was amenable to
scale up to pilot plant scale, and was of suffi-
cient purity for the GMP production of PI-88
for clinical trials.
mentation has been successfully scaled up into
a 500-L fermenter to provide up to 20 kg of
product per run (wet weight; dry weight ꢀ8
kg).
The hydrolysis of the PS was carried out for
6–10 h at 100 °C at a concentration of ap-
proximately 40–50 g/L, pH 2.2–2.5 using 1 M
HCl as the catalyst and in the presence of
KCl. The pH was closely monitored and was
found to rise slightly (up to pH 3) over the
first hour of reaction as the PS went into
solution. Therefore, it was readjusted back to
pH 2.2–2.5 by further additions of 1 M HCl.
It has been noted9,11 that under these hydroly-
sis conditions the side chains attached to the
2. Results and discussion
Previously described methods for the prepa-
ration of the OPF involve mild acid-catalysed
hydrolysis of the PS.8,9 After neutralisation,
the phosphorylated products, the OPF and the
PC, are converted into their barium salts by
the addition of barium acetate. The PC is then
precipitated with 0.1 volumes of ethanol and
removed by centrifugation or filtration. Addi-
tion of a further two volumes of ethanol to
the mother liquors then precipitates the OPF.
Alternatively, the OPF is isolated by size-ex-
clusion chromatography of the hydrolysate.10
The precipitation method has been found to
be less than optimal. The PC is incompletely
removed during the first precipitation step,
resulting in impure OPF which must be fur-
ther purified by size-exclusion chromatogra-
phy.9,17 The product may also be
contaminated with incompletely hydrolysed
oligosaccharide phosphate chains (predomi-
nantly made up of the decasaccharide diphos-
phate P-Man5-P-Man5) as well as smaller,
neutral oligosaccharides.9,17 The very large
volumes of ethanol required are also impracti-
cal for scale up. Moreover, the use of toxic
barium acetate needs to be avoided. The size-
exclusion method is also not well suited for
scaling up due to the very large quantities of
expensive size-exclusion media that are re-
quired, as well as the need to concentrate the
large volumes of hydrolysate that are obtained
prior to chromatography.
1
PC are incompletely removed. The H NMR
1
spectra of these products usually showed H
signals at l ꢀ5.7 (3JH,P ꢀ8 Hz), indicative of
the presence of phosphate diester linkages.
However, this was not found to be the case in
1
this instance, as shown by the H NMR spec-
trum of the isolated PC, which displayed no
such signals. These incomplete hydrolyses ob-
served previously may have been due to the
(unchecked) rise in the pH during the early
stages of the reaction.
The hydrolysis was routinely performed on
2.5 kg (wet weight) of PS. The reaction was
monitored by HPLC using a size-exclusion
column (Tosohaas TSK-Gel G2500PWXL)
with refractive index detection. Fig. 1(c) shows
a typical chromatogram at the completion of
the reaction (6 h). All of the components of
the fractions were isolated as described in
Section 3 and were identified by comparison
of their NMR and/or mass spectral data with
those in the literature.11,12 The fractions were
found to contain: the PC (A), P-Man5–P-
Man5 (B), OPF (C), a mixture of phosphate
and the unphosphorylated oligosaccharides 2,
4 and 6 (D), and the unphosphorylated disac-
charide 8 (E). Fractions F–H correspond to
salts (NaCl or KCl) and ethanol.
An analysis of the time course of the reac-
tion (Fig. 1(a–c)) indicates that the initial
products formed are the PC, poly-P-Man5
chains, the OPF and the disaccharide 8. Over
the course of the reaction, any remaining
phosphorylated side chains attached to the PC
are cleaved, the bulk of the poly-P-Man5
chains are hydrolysed to OPF, and some neu-
The synthesis began with the PS, which was
prepared by fermentation of P. holstii NRRL
Y-2448.2,9 The product was obtained by etha-
nol precipitation as a sticky gum with a high
water–ethanol content (ꢀ60%) that could be
1
used directly in the hydrolysis step. The H
NMR (300 MHz) spectrum of an analytical
sample of the PS recorded in D2O at 45 °C
was in accord with the literature.11 This fer-

186
V. Ferro et al. / Carbohydrate Research 332 (2001) 183–189
Fig. 1. HPLC analysis of the phosphomannan hydrolysate at time=0, 2 and 6 h.
phates 3 and 5, which are incorporated in
smaller amounts into the side chains.
tral oligosaccharides (2, 4 and 6) and phos-
phate are produced. Interestingly, the amount
of disaccharide 8 formed increases only
slightly over time. Attempts to convert the
remaining traces of fraction B into the OPF
by prolonging the reaction time did not yield
more OPF, but rather more unphosphorylated
products and phosphate (fraction D). This
indicated that the phosphate group of the
OPF was being cleaved. The optimal reaction
time was therefore determined to be 6–10 h.
The early appearance of 8 in the reaction
mixture, in a relatively unchanging amount
over time, confirms that in the PS the disac-
charide 8 is present mostly as an unphospho-
rylated capping unit on the side chains and
undergoes rapid hydrolysis. The lack of pen-
tasaccharide 2 in the initial stages of the reac-
tion also seems to indicate that this
pentasaccharide may not be a major unphos-
phorylated capping unit as has been sug-
gested.12 Pentasaccharide 2 probably arises to
a large extent from dephosphorylation of 1.
Similarly, the oligosaccharides 4 and 6 arise in
a similar manner from their respective phos-
After neutralisation of the hydrolysate and
adjustment of the pH to 9–9.5 with 1 M
NaOH, the OPF and unphosphorylated
oligosaccharides were separated from the PC
by ultrafiltration through a 10,000 nominal
molecular weight cut off (NMWCO) mem-
brane. The OPF, unphosphorylated oligosac-
charides and salts permeated through the
membrane whilst the high-molecular-weight
PC was retained. The separation was achieved
in diafiltration mode (against eight diavolumes
of purified water) with a tangential or
crossflow system. Such a system is readily
scaled up by increasing the available mem-
brane area and the flow rate. Membranes of
NMWCOs ranging from 3,000 to 100,000
were also successfully trialed, however, the
10,000 NMWCO membrane was the most
efficient.
Purification of the permeate was achieved
by ion-exchange chromatography on DEAE-
Spherilose resin equilibrated with 0.01 M
NH4HCO3. The OPF and traces of fraction B

V. Ferro et al. / Carbohydrate Research 332 (2001) 183–189
187
quantities to GMP standard, which is
amenable to scale up to pilot plant scale for
use in the manufacture of PI-88 for clinical
trials. We have also shed further light on the
structure of the phosphomannan.
bound to the resin whilst the unphosphory-
lated products were washed through. The
OPF was then eluted off with 0.25 M
NH4HCO3. The pooled fractions containing
the OPF were concentrated and desalted by
reverse osmosis prior to lyophilisation. The
reverse osmosis step was critical in rapidly
reducing the large volumes obtained (50–60 L
per hydrolysis) to manageable levels for
lyophilisation and is a key step in any further
scale up. Moreover, the reverse osmosis was
also able to remove the bulk of the inorganic
salts in the fractions, resulting in a very pure
final product containing only a few per cent of
inorganic salts. The OPF was obtained as a
hygroscopic powder that was approximately
93% pure by HPLC (single carbohydrate
peak) and could be used without further
purification for the production of PI-88. Ana-
lytical samples could be obtained by further
desalting on columns of Bio-Gel P-2, followed
by lyophilisation. This procedure has success-
fully been used to provide 500–600 g of OPF
per hydrolysis on a routine basis, and there is
potential for further scale up to pilot plant
scale because all the process steps are directly
scaleable.
3. Experimental
General.—NMR spectra were recorded us-
ing a Varian Unity-400, XL-300 or -200 spec-
trometer for solutions in D2O and are
1
referenced to internal acetone (l 2.225 for H
and l 31.0 for 13C) or to external H3PO4 (l
0.00 for 31P). Mass spectra were obtained by
electrospray-ionization (ESIMS) on a Fisons
VG Quattro II mass spectrometer. Analytical
HPLC was performed on a Waters Alliance
2690 separations module using a Tosohaas
TSK-Gel G2500PWXL 6 mm (300×7.8 mm)
size-exclusion column with a flow rate of 1.0
mL/min and 0.05 M NaCl as the mobile
phase. Detection was with a Waters 2410 re-
fractive index detector. Reverse osmosis was
performed on a Millipore Proscale system (op-
erating temperature 15–22 °C, feed pressure
1200 kPa, inlet flow rate 8 L/min) equipped
with two Helicon RO4 cartridges with
Nanomax 50 membrane (membrane area 0.37
m2/cartridge). Thin-layer chromatography
(TLC) was performed on E. Merck Kieselgel
60 F254 aluminium-backed sheets with spe-
cified eluents. Compounds were detected by
charring with 10% aq H2SO4. All reagents
used were analytical grade and were used
without further purification. All water used
was purified in house to USP purified water
standard. DEAE-Spherilose and Bio-Gel P-2
were from Isco and BioRad, respectively.
Phosphomannan from P. holstii NRRL Y-
2448 (PS).—The extracellular polysaccharide
from P. holstii NRRL Y-2448 was prepared
by fermentation according to the literature
procedure2,9 and was isolated by EtOH precip-
itation as a gum which was used without
further purification. A typical yield from a
400-L fermentation was 20 kg. The moisture–
EtOH content was determined to be approxi-
mately 60% by loss on drying. An analytical
sample was prepared by dissolving lyophilised
1
The H and 13C NMR spectra of the OPF
were in accord with the published spectra12
and displayed the expected major resonances
for 1 as well as a number of minor resonances
due to the presence of the homologous
oligosaccharide phosphates 3, 5 and 7. The 31P
NMR spectrum gave a single peak at l 3.44,
indicative of phosphate monoester.24 In addi-
tion, the mass spectrum, obtained in the nega-
tive-ion mode by the electrospray ionisation
technique (ESIMS), gave peaks for the ex-
pected major [M−H]− ions at m/z 907.3 and
745.3, respectively, for 1 and 3, as well as
peaks at m/z 583.3 and 421.3 for 5 and 7,
respectively.12 Interestingly, the reported peak
for hexasaccharide phosphate was not clearly
discernible, presumably due to the very small
amount present. Its presence was detected,
however, in the mass spectrum of fraction B
(m/z 979.3, [M−2H]2−, P2-Man11).12
In conclusion, we have developed a proce-
dure for preparing the oligosaccharide phos-
phate
fraction
of
the
extracellular
phosphomannan of P. holstii in multigram

188
V. Ferro et al. / Carbohydrate Research 332 (2001) 183–189
20 L by reverse osmosis. Reverse osmosis was
continued in diafiltration mode against
purified water until the conductivity of the
permeate was 50.2 mS/cm, and the solution
was then concentrated to a final volume of 6
L. Lyophilisation then gave the OPF as a
white, hygroscopic powder (566 g). The
product was of sufficient purity (ꢀ93% by
HPLC) for PI-88 synthesis. An analytical sam-
ple was prepared by desalting on a column of
Bio-Gel P-2 (5×100 cm) using 0.1 M
NH4HCO3 as eluent, followed by lyophilisa-
tion. ESIMS: m/z 907.3 ([M−H]−, 1), 745.3
([M−H]−, 3), 583.3 ([M−H]−, 5), 421.3
([M−H]−, 7). 31P NMR (162 MHz): l 3.44
PS (244 mg) in deionised water (100 mL) and
dialysing against two changes of deionised
water (4 L) followed by lyophilisation to give
1
a white amorphous solid (218 mg). H NMR
(300 MHz, 45 °C): l 5.68 (bd, 1 H, 3JH,P ꢀ7.5
Hz), 5.12 (bs, 3 H), 5.06 (bs, 1 H).
Hydrolysis of the phosphomannan.—A solu-
tion of KCl (600 g) in water (60 L), acidified
to pH 2.4 with 1 M HCl, was prepared in a 75
L Chemap stainless steel reaction vessel. The
PS (2.49 kg wet weight) was added in portions
to the solution via the addition portal, and the
mixture was heated at 100 °C for 7 h with
vigorous stirring. The pH was monitored
hourly and was readjusted to pH 2.3 after 1 h
by further addition of 1 M HCl. The progress
of the reaction was monitored by HPLC. At
the completion of the reaction, the hy-
drolysate was cooled to rt and then the pH
was adjusted to pH 9.5 by the addition of 1 M
NaOH.
1
(s). H and 13C NMR data were in accord
with the published spectra.12 1H NMR (300
MHz): l 5.38 (s, 0.9 H, a anomer of reducing
Man residue), 5.13 (s, 2 H), 5.09 (s, 1 H), 5.04
(s, 1 H), 4.93 (s, 0.1 H, b anomer of reducing
Man residue).
Isolation of oligosaccharides from fractions
D and E.—A sample of the unbound effluent
from the ion-exchange chromatography step
(2 L) was concentrated to dryness under vac-
uum. The residue was purified by size-exclu-
sion chromatography on a column of Bio-Gel
P-2 (5×100 cm) using 0.1 M NH4HCO3 as
eluent. The fractions were monitored by TLC
(3:2:1 EtOAc–MeOH–water). The following
pure oligosaccharides were isolated as white
amorphous solids after lyophilisation: 2 (Rf =
0.19, 240 mg), 4 (Rf=0.25, 303 mg), 6 (Rf=
0.33, 229 mg) and 8 (Rf=0.42, 310 mg).
HPLC analysis showed that fraction E was
made up of 8 (Rt=9.2 min) whilst fraction D
contained 6 (Rt =8.4 min), 4 (Rt =8.1 min)
Diafiltration of the hydrolysate.—The hy-
drolysate was diluted to 70 L and transferred
to a 150-L stainless steel tank. It was then
diafiltered against eight diavolumes of purified
water using an ultrafiltration system fitted
with two Sartorius Hydrosart 10K (NMWCO
10,000) filter cassettes (membrane area 0.6
m2/cassette) configured in series. Operating
parameters were: inlet pressure 200 kPa, outlet
pressure 150 kPa, retentate cross flow rate
15.6 L/min, permeate flux rate 66–87 L/h/m2
(16–22 °C). The progress of the separation
was monitored by HPLC. The permeate (630
L, conductivity=1.1 mS/cm) was split into six
batches for processing by ion-exchange
chromatography.
Ion-exchange chromatography of the perme-
ate ( fraction C).—A column of DEAE-Spher-
ilose (30 L) was equilibrated with 0.01 M
NH4HCO3 at a flow rate of 1.5 L/min. The
permeate (ꢀ100 L per run, six runs) was
loaded onto the column, and the neutral frac-
tion was washed from the column with 0.01 M
NH4HCO3 until the conductivity of the
effluent was within 0.2 mS/cm of the baseline
(ꢀ100 L). The product was then eluted off
using 0.25 M NH4HCO3. One-litre fractions
were collected and monitored by HPLC. The
appropriate fractions from each chromatogra-
phy run were combined and concentrated to
1
and 2 (Rt=7.9 min). The H and 13C NMR
spectra of the oligosaccharides were in accord
with the published spectra.12
High-molecular-weight core polysaccharide
( fraction A, PC).—A sample of the retentate
from the ultrafiltration step (50 mL) was di-
afiltered against six diavolumes (300 mL) of
water using an Amicon miniplate-3 biocon-
centrator (NMWCO 3,000). HPLC analysis
indicated that the retentate contained only a
single peak (Rt=5.6 min). The retentate was
filtered and lyophilised to give the PC as a
1
white, amorphous solid (137 mg). H NMR
(300 MHz): l 5.26 (bs, 1 H), 5.08 (bs, 3 H),
5.01 (bs, 2 H).

V. Ferro et al. / Carbohydrate Research 332 (2001) 183–189
189
[2] Jeanes, A.; Pittsley, J. E.; Watson, P. R.; Dimler, R. J.
Arch. Biochem. Biophys. 1961, 92, 343–350.
[3] Slodki, M. E.; Wickerham, L. J.; Cadmus, M. C. J.
Bacteriol. 1961, 82, 269–274.
[4] Jeanes, A.; Watson, P. R. Can. J. Chem. 1962, 40,
1318–1325.
Isolation of fraction B.—A sample of the
permeate from the ultrafiltration step was con-
centrated under vacuum and was then purified
by size-exclusion chromatography on
a
column of Bio-Gel P-2 (10×100 cm) using 0.1
M NH4HCO3 as eluent. The fractions were
monitored by TLC (7:3 EtOH–1 M ammo-
nium acetate), Rf=0.1 (vs 0.3 for OPF) and
HPLC (Rt =6.7 min). Fractions containing
pure product (single peak by HPLC) were
pooled and lyophilised to give fraction B (pre-
dominantly P-Man5 –P-Man5) as a white,
amorphous powder. 31P NMR (162 MHz) l
[5] Jeanes, A.; Pittsley, J. E.; Watson, P. R.; Sloneker, J. H.
Can. J. Chem. 1962, 40, 2256–2260.
[6] Slodki, M. E. Biochim. Biophys. Acta 1962, 57, 525–533.
[7] Dintzis, F. R.; Babcock, G. E.; Tobin, R. Solution Prop-
erties of Natural Polymers; Chemical Society: London,
1968; pp. 195–206 special publication 23.
[8] Slodki, M. E.; Ward, R. M.; Boundy, J. A. Biochim.
Biophys. Acta 1973, 304, 449–456.
[9] Bretthauer, R. K.; Kaczorowski, G. J.; Weise, M. J.
Biochemistry 1973, 12, 1251–1256.
[10] Slodki, M. E. In Industrial Polysaccharides: Genetic Engi-
neering: Structure/Property Relations and Applications;
Yalpani, M., Ed.; Elsevier: Amsterdam, 1987; pp. 109–
119.
[11] Parolis, L. A. S.; Duus, J.Ø.; Parolis, H.; Meldal, M.;
Bock, K. Carbohydr. Res. 1996, 293, 101–117.
[12] Parolis, L. A. S.; Parolis, H.; Kenne, L.; Meldal, M.;
Bock, K. Carbohydr. Res. 1998, 309, 77–87.
[13] Sahagian, G. C.; Distler, J. J.; Jourdian, G. W. Methods
Enzymol. 1982, 83, 392–396.
1
−1.34 (s), 4.60 (s). H NMR (300 MHz) l
3
5.66 (d, JH,P 7.8 Hz), 5.37 (s, 1 H), 5.10–4.86
(8 H). ESIMS: m/z 979.3 ([M−2H]2−, P2-
Man11), 898.3 ([M−2H]2−, P2-Man10), 817.4
([M−2H]2−, P2-Man9), 736.4 ([M−2H]2−
P2-Man8).
,
[14] Steiner, A. W.; Rome, L. H. Arch. Biochem. Biophys.
1982, 214, 681–687.
[15] Brown, W. J.; Farquhar, M. G. Cell 1984, 36, 295–307.
[16] Baba, T.; Watanabe, K.; Arai, Y. Carbohydr. Res. 1988,
177, 153–161.
Acknowledgements
We thank the Centre for Instrumental and
Developmental Chemistry at the Queensland
University of Technology (Brisbane), and the
Griffith University Magnetic Resonance Facil-
ity (Nathan) for the use of their NMR facili-
ties. We thank Dr David Armitt (ANU) for
the ESIMS determinations. We also thank Dr
Bin Wang and Leslie Tillack for the fermenta-
tion, Jeff Rood and Jason Ryan for technical
assistance and Dr Robert Don for useful
discussions.
[17] Murray, G. J.; Neville, Jr., D. M. J. Biol. Chem. 1980,
255, 11942–11948.
[18] Baba, T.; Watanabe, K.; Yonezawa, N.; Hiroto, M.;
Arai, Y. Carbohydr. Res. 1988, 177, 163–172.
[19] Yoshida, T.; Lee, Y. C. Carbohydr. Res. 1994, 251,
175–186.
[20] Parish, C. R.; Freeman, C.; Brown, K. J.; Francis, D. J.;
Cowden, W. B. Cancer Res. 1999, 59, 3433–3441.
[21] Hulett, M. D.; Freeman, C.; Hamdorf, B. J.; Baker, R.
T.; Harris, M. J.; Parish, C. R. Nat. Med. 1999, 5,
803–809.
[22] Vlodavsky, I.; Friedmann, Y.; Elkin, M.; Aingorn, H.;
Atzmon, R.; Ishai-Michaeli, R.; Bitan, M.; Pappo, O.;
Peretz, T.; Michal, I.; Spector, L.; Pecker, I. Nat. Med.
1999, 5, 793–802.
[23] Podger; D. M. AATD-Ad6ances in Anticoagulant, An-
tithrombotic and Thrombolytic Drugs; IBC 11th Annual
International Symposium; Waltham MA, October 2000.
[24] Gorenstein, D. G.; Shah, D. O. In Phosphorus-31 NMR
Principles and Applications; Gorenstein, D. G., Ed.; Aca-
demic: Orlando, 1984; pp. 549–591.
References
[1] Anderson, R. F.; Cadmus, M. C.; Benedict, R. G.;
Slodki, M. E. Arch. Biochem. Biophys. 1960, 89, 289–292.
.