Mannose-pepstatin conjugates as targeted inhibitors of antigen processing

Article abstract of DOI:10.1039/b600060f

The molecular details of antigen processing, including the identity of the enzymes involved, their intracellular location and their substrate specificity, are still incompletely understood. Selective inhibition of proteolytic antigen processing enzymes such as cathepsins D and E, using small molecular inhibitors such as pepstatin, has proven to be a valuable tool in investigating these pathways. However, pepstatin is poorly soluble in water and has limited access to the antigen processing compartment in antigen presenting cells. We have synthesised mannose-pepstatin conjugates, and neomannosylated BSA-pepstatin conjugates, as tools for the in vivo study of the antigen processing pathway. Conjugation to mannose and to neomannosylated BSA substantially improved the solubility of the conjugates relative to pepstatin. The mannose-pepstatin conjugates showed no reduction in inhibition of cathepsin E, whereas the neomannosylated BSA-pepstatin conjugates showed some loss of inhibition, probably due to steric factors. However, a neomannosylated BSA-pepstatin conjugate incorporating a cleavable disulfide linkage between the pepstatin and the BSA showed the best uptake to dendritic cells and the best inhibition of antigen processing. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b600060f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Mannose–pepstatin conjugates as targeted inhibitors of antigen processing†
Paul Free,a,b Christopher A. Hurley,a Takashi Kageyama,c Benjamin M. Chainb and Alethea B. Tabor*a
Received 4th January 2006, Accepted 8th March 2006
First published as an Advance Article on the web 4th April 2006
DOI: 10.1039/b600060f
The molecular details of antigen processing, including the identity of the enzymes involved, their
intracellular location and their substrate specificity, are still incompletely understood. Selective
inhibition of proteolytic antigen processing enzymes such as cathepsins D and E, using small molecular
inhibitors such as pepstatin, has proven to be a valuable tool in investigating these pathways. However,
pepstatin is poorly soluble in water and has limited access to the antigen processing compartment in
antigen presenting cells. We have synthesised mannose–pepstatin conjugates, and neomannosylated
BSA–pepstatin conjugates, as tools for the in vivo study of the antigen processing pathway. Conjugation
to mannose and to neomannosylated BSA substantially improved the solubility of the conjugates
relative to pepstatin. The mannose–pepstatin conjugates showed no reduction in inhibition of cathepsin
E, whereas the neomannosylated BSA–pepstatin conjugates showed some loss of inhibition, probably
due to steric factors. However, a neomannosylated BSA–pepstatin conjugate incorporating a cleavable
disulfide linkage between the pepstatin and the BSA showed the best uptake to dendritic cells and the
best inhibition of antigen processing.
internalises such ligands;4 mannosylation of peptide and protein
antigens has been shown to result in a significant increase in T-
Introduction
cell response.5 High affinity synthetic ligands for the mannose
receptor have recently been developed, based on multivalent
cluster mannosides which can simultaneously bind to the mul-
tiple carbohydrate recognition domains. In particular, lysine-
based cluster mannosides bearing up to six terminal mannose
groups, connected to the lysine residues by flexible spacers, have
been studied and shown to have subnanomolar affinity for the
mannose receptor.6 These and similar cluster mannosides have
been used to block mannose receptors7 and for the efficient, cell-
specific delivery of antigens,8 fluorophores9 and PNA.10 Mannose-
targeted liposomes have also been synthesised for gene11,12,13 and
antigen14,15 delivery to dendritic cells, and other scaffolds, such
as mannosylated cyclodextrins16 and mannosylated bovine serum
albumin17 have also been used to target the mannose receptor
and deliver therapeutic agents to macrophages. In spite of the
advantages of such an approach, targeted delivery of pepstatin
has seldom been explored,18,19 and receptor-mediated uptake to
macrophages and dendritic cells via the mannose receptor has not
previously been disclosed.
Uptake, proteolytic degradation, and display of foreign antigens
by professional antigen-presenting cells such as dendritic cells
are essential steps in the immune response to antigens. Under-
standing how foreign antigens, once internalised, are proteolysed
is therefore key to understanding both the normal function of
the immune system1 and the abnormal antigen processing which
occurs in autoimmune diseases.2 One approach to the study of
the role of proteolysis in antigen processing and presentation
is to use proteolytic inhibitors, such as pepstatin, to block the
processing enzymes, such as cathepsins D and E.3 However, there
are several major problems with this approach. Aspartic protease
inhibitors such as pepstatin are only sparingly soluble in water,
and are inefficiently transported across membranes. Moreover, in
the absence of any tissue or cell targeting specificity, it is also
difficult to carry out reliable in vivo studies. We therefore sought to
develop a system for the solubilisation, cell-specific delivery and
receptor-mediated uptake of pepstatin.
Antigen uptake by dendritic cells and macrophages is mediated
by the mannose receptor, a membrane-associated protein bearing
eight carbohydrate recognition domains. The mannose receptor
binds to mannosylated polysaccharides and glycoproteins and
In this paper we report the design, synthesis and biological
characterisation of four mannose–pepstatin bioconjugates, de-
signed for receptor-mediated uptake of pepstatin to dendritic
cells. We have studied the enzyme inhibition properties of
these bioconjugates and their uptake into dendritic cells, and
report their utility as tools for studying the antigen processing
pathway.
aUniversity College London, Department of Chemistry, Christopher In-
gold Laboratories, 20 Gordon Street, London, WC1H 0AJ, UK. E-mail:
a.b.tabor@ucl.ac.uk; Fax: 020 7679 7463; Tel: 020 7679 4637
bUniversity College London, Department of Immunology, The Windeyer
Building, 46 Cleveland Street, London, W1T 4JF, UK
cCenter for Human Evolution Modeling Research, Primate Research Insti-
tute, Kyoto University, Inuyama, 484-8506, Japan
† Electronic supplementary information (ESI) available: NOESY spec-
trum of 7; 1H and 13C assignments for the acetylated compound, 13;
MALDI-TOF mass spectra of BSA, mannose–BSA conjugate 17, and
4; details of the analysis of the number of sugars, and of free amino
groups, on mannose–BSA conjugate 17; 1H NMR spectra of 7–9, 11,
13, 1, 14–16, 2, N-(t-butyloxycarbonyl), Nꢀ-(pepstatinyl)cystamine, N-
(pepstatinyl)cystamine, and 19. See DOI: 10.1039/b600060f
Results and discussion
Design of mannose–pepstatin conjugates
We considered several factors for the design of the mannose–
pepstatin conjugates. Although many studies have indicated the
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1817–1830 | 1817
©

View Article Online
importance of polyvalent mannose clusters for successful receptor
binding and uptake,20 it is clear that for some receptors a single
glycosidic unit is sufficient,18 and we therefore sought to compare
the delivery of pepstatin bearing a single mannose unit to pepstatin
conjugated to a polymannose cluster. The optimal spacer length
between individual mannoses in mannose clusters, and between
the cluster mannoside and the antigen or inhibitor being delivered,
has not been determined for all systems: however, studies of the
interaction of mannosylated liposomes with human phagocytes
indicated that longer spacer lengths such as PEG 6, PEG 8 led
to good uptake by the mannose receptor.11 It had previously
been reported that the conjugation of pepstatin to PEG, whilst
increasing the solubility, seriously decreased the aspartic protease
inhibition, probably due to steric hinderance between the inhibitor
and the conjugated PEG.21 In order to reduce the possibility
of steric hinderance between the polymannose cluster and the
proteolytic enzymes under study, we also designed probes where
the pepstatin and mannose units were conjugated via a disulfide
bond. Recent studies have shown that specific redox enzymes,
as well as reducing agents, are prevalent in endosomes and
lysosomes, and that therefore targeted drug delivery strategies can
be enhanced by the use of cleavable disulfide linkages between the
receptor targeting moiety and the drug22 allowing the drug to be
released from the targeting moiety by reductive cleavage in the
endosome. We hypothesised that mannose–pepstatin conjugates
with a disulfide bond in the linker could also be internalised via
receptor-mediated endocytosis and would subsequently be cleaved
in the endosome to release the pepstatin.
We therefore designed mannose–pepstatin conjugates 1–4
(Fig. 1). Conjugates 1 and 2 bear a single mannose unit, in contrast
to 3 and 4, in which the pepstatin is attached to neomannosylated
bovine serum albumin (BSA). In addition, conjugates 2 and 4 have
a cleavable disulfide bond between the pepstatin and the targeting
moiety, whereas 1 and 3 are not cleavable. We envisaged that these
four conjugates would enable us to explore the effects of polyvalent
and cleavable mannose targeting moieties on pepstatin delivery.
Synthesis of mannose–pepstatin conjugates 1 and 2
A convergent route to the monomannosylated conjugates 1 and 2
was used that would allow incorporation of either a cleavable or
a non-cleavable linker at a late stage. The mannose was attached
to the linker via a flexible spacer group using standard methods,
as follows. Mannose was converted to the known 2,3,4,6-tetra-O-
acetyl-a-D-mannopyranosyl trichloroimidate 523 and was reacted
with benzyl 4-hydroxybutanoate 624 to afford 7 (Scheme 1). The
benzyl ester was then deprotected to give 8 and converted to the
N-hydroxysuccinimide 9.
For the synthesis of mannose–pepstatin conjugate 1, 9 was
then reacted with the non-cleavable linker N-(t-butyloxycarbonyl)
propane-1,3-diamine18 10 to give 11 (Scheme 2). Deprotection with
TFA was followed by coupling to pepstatin-N-hydroxysuccinimide
Fig. 1 Structure of the mannose–pepstatin conjugates 1 (non-cleavable) and 2 (cleavable) and of the neomannosylated BSA–pepstatin conjugates 3
(non-cleavable) and 4 (cleavable).
1818 | Org. Biomol. Chem., 2006, 4, 1817–1830
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
to give thiourea-bridged glycodendrimers has previously been
reported,26 and we therefore elected to use this approach in
our synthesis. Thus, 4-isothiocyanatophenyl a-D-mannopyranose
was reacted with BSA in 0.15 M NaCl27 to give mannose–BSA
conjugates 17 (Scheme 3). Three methods were used to determine
the average amount of conjugation of isothiocyanomannosyl
sugar derivatives: mass spectroscopy (MALDI-TOF); biochemical
analysis for total mannose content;28 and analysis of free amino
groups.29 These methods indicated that the mannose–BSA con-
jugates 17 had, on average, between 22 and 24 mannosyl units
attached.
In order to chemoselectively attach a pepstatin–non-cleavable
linker moiety, pepstatin-N-hydroxysuccinimide 12 was reacted
with N-(t-butyloxycarbonyl) propane-1,3-diamine 10, the Boc
group removed and the resulting amine conjugated to iodoacetic
anhydride to give the iodoacetamide 18. Similarly, pepstatin-N-
hydroxysuccinimide 12 was reacted with N-(t-butyloxycarbonyl)
cystamine 14, the Boc group removed and the resulting amine
conjugated to iodoacetic anhydride to give the iodoacetamide
19. The inherent insolubility of pepstatin in many solvents
ensured that 18 and 19, and the intermediates in their synthesis,
could be readily purified prior to coupling with the mannose–
BSA conjugate. Thus, reaction of 18 with 17 gave the non-
cleavable neomannosylated BSA–pepstatin conjugate 3 directly,
and reaction of 19 with 17 afforded the cleavable neomannosylated
BSA–pepstatin conjugate 4.
Scheme
1 Synthesis of mannose-linker 9: a) Ac2O, Et3N (99%);
b) H2NNH2.AcOH, DMF (77%); c) K2CO3, Cl3CCN, CH2Cl2 (96%); d)
6, BF3OEt2 (55%); e) H2, Pd/C (85%); f) EDCI, NHS (quant.).
12 to give 13. Removal of the acetate groups then afforded 1. Simi-
larly, 9 was reacted with the cleavable linker N-(t-butoxycarbonyl)
cystamine 14 to give 15; deprotection with TFA and coupling to
pepstatin-N-hydroxysuccinimide 12 gave 16, which was likewise
deprotected to afford 2.
Synthesis of neomannosylated BSA–pepstatin conjugates 3 and 4
BSA is an inexpensive and readily available protein, soluble in
many different aqueous buffers,25 with a large number of Lys
residues and a single Cys residue that is not involved in a disulfide
bridge. We planned to use this inherent orthogonality of functional
groups to form the desired conjugates 3 and 4. It was envisaged
that the mannose moieties could be selectively conjugated to the
Lys-NH2 groups, leaving the Cys-SH available for chemoselective
coupling to a suitable linker–pepstatin moiety. As we wished to
avoid a final deacetylation step in the synthesis of the neomanno-
sylated BSA–pepstatin conjugates, it was also necessary to alter
the strategy for mannosylation of BSA. Conjugation of oligovalent
amines with unprotected NCS-functionalised mannose derivatives
Biochemical evaluation of the mannose–pepstatin and
neomannosylated BSA–pepstatin conjugates
It was evident during synthesis that both monomannosylated
conjugates 1 and 2 and the neomannosylated BSA–pepstatin
conjugates 3 and 4 had much improved solubility compared to
pepstatin A alone. The monomannosylated conjugates 1 and 2 had
excellent solubility in organic solvents, especially methanol, and
Scheme 2 Synthesis of mannose–pepstatin conjugates 1 and 2; a) TFA; b) pepstatin N-hydroxysuccinimide 12, Et3N (45% over 2 steps); c) LiOH, MeOH
(70%); d) TFA; e) pepstatin N-hydroxysuccinimide 12, Et3N (55% over 2 steps); f) LiOH, MeOH (66%).
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1817–1830 | 1819
©

View Article Online
Scheme 3 Synthesis of neomannosylated BSA–pepstatin conjugates 3 and 4; a) BSA, 0.15 M NaCl (74%); b) 10 or 14, DMF; c) TFA, Et3SiH, thioanisole,
phenol; d) iodoacetic anhydride (18: 40% over 3 steps: 19: 42% over 3 steps); e) PBS, DMF (3: 73%: 4: 70%).
their solubility in water was much higher than that of pepstatin.
At high concentrations in water (>0.5 mM) a slight cloudiness was
visible, but at the lower concentrations used in biological assays
(<54 lM, see below) the compounds appeared to be completely
water soluble. The neomannosylated BSA–pepstatin conjugates
were highly soluble in water (greater than 1 mM).
The inhibitory activity of the conjugates was tested against
cathepsin E using a fluorogenic assay.30 Fig. 2 shows representative
plots of this data for monomannosylated conjugate 2 and neoman-
nosylated BSA–pepstatin conjugate 4. Similar data were obtained
for conjugates 1 and 3 (not shown). The best fit curve for the
data allowed estimates for IC50 values. IC50 for monomannosylated
Effect of the mannose–pepstatin and neomannosylated
BSA–pepstatin conjugates on antigen processing in A20 cells and
dendritic cells
In order to test the activity of the pepstatin conjugates on antigen
processing, we made use of a well-established assay31,32 in which
ovalbumin is processed and presented to an antigen-specific T cell
hybridoma, DO11-10, which recognizes ovalbumin amino acids
323–339. Co-culture of antigen presenting cells, antigen and the
T cell hybridoma DO11-10 induces the release of the cytokine
interleukin 2 (IL-2). We explored the activity of the inhibitors
using two types of antigen presenting cell types, A20 a B cell
lymphoma which does not express the mannose receptor33 but is
known to use the aspartic protease cathepsin E for processing,
or mouse bone marrow-derived dendritic cells which do express
mannose receptor(s).33,34
Pepstatin inhibited processing of ovalbumin by A20 cells as
described previously,31 and did not inhibit presentation of the
“preprocessed” ovalbumin peptide 323–339 (Fig. 3, lower panels),
demonstrating that inhibition was at the level of processing and not
presentation, or T cell activation. High concentrations of inhibitor
were required (around 105 higher than the IC50 measured in
enzymatic assays), and some toxicity (reflected in the inhibition of
ovalbumin peptide presentation (lower right panel) was observed
at the highest concentration.
conjugates 1 and 2 are in the picomolar range (conjugate 1 IC50
133
=
23 pM, n = 3; conjugate 2 IC50 = 86
47 pM, n = 4)
similar to that of pepstatin (not shown). IC50 for neomannosylated
BSA–pepstatin conjugate 3 and 4 are in the low nanomolar range
(conjugate 3 IC50 = 1.3 0.4 nM, n = 3; conjugate 4 IC50 = 3.5
0.4 nM, n = 3. In contrast, the neomannosylated BSA conjugate
17 without the addition of pepstatin had little inhibitory activity
(IC50 > 1 lM).
Attachment of a small linker and sugar unit does not therefore
cause any significant reduction in the inhibitory effects of pepstatin
A, while attachment of the larger neomannosylated BSA causes
a reduction in IC50 of approximately 10 fold, perhaps because the
large carrier reduces the accessibility of the pepstatin.
1820 | Org. Biomol. Chem., 2006, 4, 1817–1830
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
tions used, they precipitated out of solution in the cell culture
media required for these assays. Addition of DMSO to the stock
1
3
solution (to a final concentration of that required for pepstatin
A) restored solubility. The monomannosylated conjugates 1 and 2
also inhibited processing of ovalbumin, showing a dose response
rather similar to pepstatin. No toxicity or inhibition of peptide
presentation was observed at any concentration tested, perhaps
1
3
because the solutions contained only
the concentration of
DMSO as solvent. These conjugates therefore both retain the
inhibitory activity of pepstatin, albeit with improved solubility
properties.
In contrast, neither neomannosylated conjugates 3 or 4 showed
consistent inhibition of ovalbumin processing/presentation by
A20 cells (Fig. 4). Although some slight inhibition was observed
with conjugate 4 this did not exceed that shown by the mannose–
BSA conjugate 17 which did not carry pepstatin, and was pre-
sumably due to non-specific effects. For reasons which remained
Fig. 2 Inhibitory activity of mannose–pepstatin and neomannosylated
BSA–pepstatin conjugates against cathepsin E. Inhibitors were incubated
at the concentrations shown in the presence of fluorogenic peptide
substrate (5 lM) and purified rat cathepsin E (50 ng ml−1) as described
in Experimental procedures. Results for each inhibitor are representative
of at least three experiments and expressed as % activity relative to the
activity in the absence of inhibitor.
Fig. 4 Antigen processing by A20 cells in the presence of neomannosylated
BSA–pepstatin conjugates 3 and 4, or the mannose–BSA conjugate 17. The
results are expressed as % inhibition of the IL-2 responses in the presence
of 54 lM inhibitor. The figure shows the average of four independent
experiments. Other experimental details as for Fig 3.
Preliminary experiments showed that although the monoman-
nosylated conjugates were soluble in water at the lM concentra-
Fig. 3 Antigen processing by A20 cells in the presence of mannose–pepstatin conjugates 1 and 2, or pepstatin. A20 cells and DO11-10 T cells were co-cultured
in the presence of ovalbumin (left panels, 3 mg ml−1) or ovalbumin peptide 323–339 (right panels, 0.1 lg ml−1) and inhibitor or an equivalent concentration
of DMSO in culture medium. IL-2 was collected and measured by ELISA as described in Experimental procedures and results are expressed as % of
IL-2 in the presence of antigen but no inhibitor (corresponds to approximately 1 ng ml−1 IL-2). Dotted line shows IL-2 levels in absence of antigen or
inhibitor.
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1817–1830 | 1821
©

View Article Online
Fig. 5 Antigen processing by dendritic cells in the presence of mannose–pepstatin conjugates 1 and 2. Dendritic cells and DO11-10 T cells were co-cultured
in the presence of ovalbumin (left panel, 3 mg ml−1) or ovalbumin peptide 323–339 (right panel, 0.3 lg ml−1) and inhibitor or an equivalent concentration
of DMSO in culture medium. IL-2 was collected and measured by CTLL bioassay as described in Experimental procedures and results are expressed as
% of IL-2 in the presence of antigen but no inhibitor (corresponds to approximately 1 ng ml−1 IL-2). IL-2 levels in absence of antigen were less than 10%.
The result shows one representative of four experiments.
Fig. 6 Antigen processing by dendritic cells in the presence of neomannosylated BSA–pepstatin conjugates 4, or the mannose–BSA conjugate 17. Dendritic
cells and DO11-10 T cells were co-cultured in the presence of ovalbumin (left panel, 0.5 mg ml−1) or ovalbumin peptide 323–339 (right panel, 0.05 lg ml−1
)
and inhibitor in culture medium. IL-2 was collected and measured by CTLL bioassay as described in Experimental procedures, and results are expressed
as % of IL-2 in the presence of antigen but no inhibitor (corresponds to approximately 1 ng ml−1 IL-2). IL-2 levels in absence of antigen were less than
10%. The result shows one representative of eight experiments.
unclear, conjugate 3 actually appeared to enhance IL-2 release, but
this phenomenon was not explored further. The lack of inhibition
by the mannosylated BSA–pepstatin conjugates may reflect either
the increased IC50 of these compounds demonstrated in Fig 2,
and/or a decreased ability of these much larger molecules to enter
the A20 cells, and reach the target enzyme.
A20 cells, although efficient at processing via fluid phase
uptake,35 do not carry the mannose receptor, and cannot therefore
take up the mannosylated conjugates via receptor mediated
uptake. We therefore used dendritic cells, which show extremely
efficient receptor mediated uptake via the mannose receptor to
present ovalbumin to the DO11-10 T cells. Neither pepstatin (not
shown) nor the monomannosylated conjugates 1 and 2 inhibited
ovalbumin presentation above the level of the DMSO solvent, up
to concentrations of 18 lM (Fig. 5). Higher concentrations were
toxic to the dendritic cells.
Initial experiments using the neomannosylated-BSA conjugates
were carried out at the same concentration ranges. Although
conjugates 3 and 4 inhibited processing of ovalbumin at these
concentrations, there was no difference between those conjugates
carrying pepstatin (i.e. 3 and 4) and the mannose–BSA conjugate
17. Inhibition was presumed to be due to non-specific effects
of the high concentrations of mannose BSA conjugates, and
these experiments were therefore not pursued further. However,
conjugate 4 showed consistent inhibition of ovalbumin, but not
ovalbumin peptide presentation, at concentrations of around 1 ×
10−5 lM. A representative experiment is shown in Fig. 6, and
an average inhibition of 71%
8 was obtained at 2 lM in
8 different dendritic cell preparations. Preliminary experiments
suggested that the neomannosylated conjugate 3 (without the
cleavable disulfide bond linker) was much less active, suggesting
the presence of a cleavable disulfide bond within the linker was an
essential feature of the design.
Conclusion
Our objective in synthesising the mannose derivatives of pepstatin
was to increase solubility, and at the same time target this
inhibitor to the endocytic pathway of antigen presenting cells thus
enhancing bio-availability. All the conjugates synthesised showed
the expected increase in solubility in aqueous solution, although
for the monomannosylated derivatives, this increase was limited,
and the compounds still required significant concentrations of
DMSO to remain in solution in physiological media, which
contain salts and serum proteins. Nevertheless, synthesis of these
compounds was useful, since it demonstrated that the synthetic
strategy of linking a pepstatin to the mannose via a linker did not
1822 | Org. Biomol. Chem., 2006, 4, 1817–1830
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
interfere with the inhibitory activity of the pepstatin, measured
either enzymatically or in the A20 antigen processing bioassay.
The monomannosylated derivatives and pepstatin itself were
not able to inhibit processing by dendritic cells. Although the
molecular reasons for this are not known, it is likely to reflect
the greatly enhanced efficiency of processing/presentation by
dendritic cells, which provides a much more stringent test of
inhibitors of processing enzymes. Indeed it is notable that few
previous studies have published data showing inhibition of antigen
processing in dendritic cells by small molecular weight protease
inhibitors.36
However, inhibition of processing by dendritic cells could be
achieved by the use of the neomannosylated BSA–pepstatin
conjugate 4. Since the inhibitory activity of this conjugate against
purified enzyme was actually less than free pepstatin (see Fig. 2),
this increased bioactivity presumably reflects the much higher
avidity of binding and uptake via the mannose receptor by
the conjugate carrying multiple mannose residues, over that of
a monomannosylate.6,17,20 Once inside the cell, the intracellular
environment is likely to favour reduction of the disulfide linker,
releasing pepstatin to interact efficiently with its target.
In summary, we have demonstrated that conjugation of pep-
statin to single, or multiple, mannose units constitutes a pow-
erful method for solubilising such small molecule inhibitors, for
targeting them to the desired cells and facilitating their uptake.
Such conjugates constitute important new tools for investigating
the antigen processing pathway, in vitro and in vivo. Inhibition
of processing by pepstatin conjugates clearly implicates aspartic
proteinases in the antigen processing pathway of dendritic cells.
At least two aspartic proteinases, cathepsins D and E, are known
to be expressed in the immune system, and we have extended
this work, using neomannosylated conjugate 4 in conjunction
with genetic approaches, to further dissect which enzyme is
actually functionally the most important.37 Further studies, using
fluoresceinated derivatives of the conjugate to track uptake and
intracellular distribution, are planned.
carried out using a Shimadzu FTIR-8700 Fourier transform infra-
red spectrometer and analysed with Shimadzu FT IRS software.
Analytical HPLC (flow rate 1 ml min−1) was carried out using
a Millipore Waters 600E system controller and pump system, Pye
Unicam PU4025 UV detector, Shimadzu C-R6A chromatopac
and a Vydac 218TP54 protein & peptide C18 column, 4.6 mm ×
25 cm. Preparative HPLC (flow rate 15 ml min−1) was carried out
using a Millipore Waters 600E system controller and pump system,
Gilson Holochrome UV detector, Millipore Waters 745B data
module and a Vydac 218TP1022 protein & peptide C18 column,
21.4 mm × 25 cm. tlc was carried out on pre-coated 0.25 mm thick
Merck 60 F254 silica plates. Reverse phase silica plates used were
Merck RP-8 F254S. Silica for chromatography was obtained from
BDH.
Dialysis was performed using standard 12 000 molecular weight
cut-off (MWCO) dialysis tubing. Before use, the tubing was
prepared by first boiling for 10 min in a large volume of 2% (w/v)
sodium bicarbonate and 1 mM EDTA (pH 8.0), then rinsing the
tubing thoroughly in 1 mM EDTA (pH 8.0), before boiling again
for 10 minutes in 1 mM EDTA (pH 8.0). The tubing was allowed
to cool before use, and was stored in solution (water or buffer)
◦
at 4 C. Centrifugal filtration devices used were either Millipore
centriplus YM-30 (Millipore (UK) Ltd) with a maximum 15 ml
capacity and used within a Beckman J2-21 swing bucket centrifuge,
or Millipore ultrafree-4 filtration unit (5 ml maximum capacity)
using a fixed 35◦ angle centrifuge (Jouan MR 1812, Jouan S. A.,
France).
Tetrahydrofuran (THF) was dried by distillation from a sus-
pension of THF with sodium and benzophenone. Dimethyl
formamide (DMF) was dried by distillation from a suspension of
DMF and calcium hydride. CH2Cl2, Et3N and EtOH were dried
over CaH powder and stored over molecular sieves when needed.
With all reactions that used dry solvents, an inert atmosphere of
N2 or Ar was used.
(3’-Benzyloxycarbonylpropyl)-2,3,4,6-tetra-O-acetyl-a-D-
mannopyranoside 7
Benzyl 4-hydroxybutanoate 624 (2.68 g, 5.4 mmol) and
2,3,4,6-tetra-O-acetyl-a-D-mannopyranosyl trichloroacetimidate
523 (2.11 g, 10.8 mmol, 2 eq.) were added to a flask. Water
was azeotropically removed from the mixture with anhydrous
toluene (4 × 10 ml). Anhydrous CH2Cl2 (30 ml) was added and
the mixture stirred vigorously and cooled to −40 ◦C using an
MeCN–CO2 (S) bath. BF3·Et2O (1.4 ml, 10.9 mmol, 2 eq.) was
added slowly to this reaction mixture over 10 min during which
time the reaction mixture turned yellow. The disappearance of the
acetimidate 5 was monitored by tlc with maximum conversion to 7
by 4 h. The reaction mixture was washed with ice-cold 50% sodium
bicarbonate solution (3 × 60 ml) and then with ice-cold 50% brine
solution (3 × 60 ml). The organic layer was dried over MgSO4
and filtered. Removal of the CH2Cl2 in vacuo gave a yellow oil,
which was purified by reverse phase chromatography (gradient:
100% water to 80% MeOH–water: Rf 0.35 (3 : 1 MeOH–water;
reverse-phase tlc plates)). The column fractions containing 7 were
pooled and the MeOH removed in vacuo. The remaining aqueous
solution was extracted with CH2Cl2 (50 ml). The CH2Cl2 layer was
dried (MgSO4) and the solvent removed in vacuo to give the titled
Experimental
General experimental procedures
Unless specified, all reagents were purchased from commercial
suppliers and were used without further purification. Pepstatin
was obtained from Calbiochem and Bachem. BSA was purchased
from Sigma-Aldrich. Isothiocyanato-phenyl a-D-mannopyranose
is available from Sigma-Aldrich, or was synthesised from 4-
nitrophenyl-a-D-mannopyranose (Acros) via hydrogenation of the
nitro group followed by reaction with CSCl2.27 1H and 13C spectra
were recorded on Bruker AC 300, 400 and 500 instruments. The
chemical shift data for each signal is given in units of d relative to
tetramethylsilane (TMS) where d (TMS) = 0. Coupling constants
(J) are quoted in Hz. CI and FAB mass spectra were recorded
using a VG ZAB SE mass spectrometer. APCI and ES mass spectra
were recorded on a Micromass Quattro mass spectrometer. Optical
rotation was recorded using an AA-10 automatic polarimeter
(Optical Activity Ltd, UK) and a 5 cm length, 1.5 ml cell (d = 0.5).
Melting points were measured using a Gallenkamp melting point
apparatus and are uncorrected. Infra-red (IR) spectroscopy was
product as a yellow wax (1.57 g, 3.0 mmol, 55%); [a]D + 38.0◦
21
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1817–1830 | 1823
©

View Article Online
(c 1, CHCl3); mmax (CHCl3) 3024–3014 (w), 1751–1740 (s), 1454–
1429 (w), 1369 (m), 1240 (s) cm−1; 1H NMR dH (500 MHz, CDCl3)
1.94 (2H, m, CH2CH2CH2O), 1.96 (3H, s), 2.01 (3H, s), 2.07 (3H,
s), 2.13 (3H, s), 2.45 (2H, t, J 8.5, CH2CH2CH2O), 3.37 (1H, ddd, J
5.7, 6.6 and 9.8, CH2CH2CH2O), 3.72 (1H, ddd, J 5.7, 6.9 and 9.8,
CH2CH2CH2O), 4.05 (1H, ddd, J 2.4, 5.3 and 9.7, H-5), 4.08 (1H,
dd, J 2.4 and 12.3, H-6), 4.25 (1H, dd, J 5.3 and 12.3, H-6), 4.76
(1H, d, J 1.8, H-1), 5.12 (2H, s, PhCH2), 5.21 (1H, dd, J 1.8, and
3.4, H-2), 5.24 (1H, t, J 9.7, H-4), 5.29 (1H, dd, J 3.4 and 9.7, H-
3), 7.34 (5H, s, Ph); 13C NMR dH (100 MHz, CDCl3) 20.58, 20.59,
20.62, 20.8, 24.5, 30.8, 62.3, 66.0, 66.3, 67.1, 68.4, 69.0, 69.4, 97.3,
128.2, 128.5, 135.8, 169.6, 169.8, 169.9, 170.5, 172.7; m/z (FAB)
547.1796 (M + Na+. C25H32O12 requires 547.1792), m/z (APCI+)
169 (100%), 331 (29%) and 547 (16%). NOESY experiments were
used to determine which anomer had been produced during the
coupling reaction: NOE correlations were observed between H-3
and H-5, but not H-1, implying that H-1 is equatorial and that
the a-anomer had been produced. The relevant portion of the
spectrum is shown in the ESI.†
(w), 1229 (s) and 1203 (s) cm−1; 1H NMR dH (300 MHz, CDCl3)
2.01–2.12 (2H, m, CH2CH2CH2O) 1.97 (3H, s), 2.02 (3H, s), 2.08
(3H, s), 2.13 (3H, s), 2.73 (2H, t, J 7.0, CH2CH2CH2O), 2.80–2.82
(4H, m, NHS), 3.52 (1H, dt, J 5.9 and 10.0, CH2CH2CH2O), 3.81
(1H, dt, J 6.1 and 10.0, CH2CH2CH2O), 3.92–3.96 (1H, m, H-5),
4.07 (1H, dd, J 2.4 and 12.3, H-6), 4.27 (1H, dd, J 5.1 and 12.3, H-
6), 4.81 (1H, d, J 1.4, H-1), 5.22–5.31 (3H, m, H-2, H-3 and H-4);
13C NMR dH (100 MHz, CDCl3) 13.9, 13.94, 20.7, 20.9, 24.5, 25.4,
30.8, 62.5, 66.0, 67.3, 68.6, 69.3, 69.4, 97.5, 169.8, 170.2, 170.3,
170.9, 172.2, 177.1; m/z (ES+) 331 (8%) and 554 (100%).
N-(t-Butyloxycarbonyl)propane-1,3-diamine 10
Anhydrous THF (20 ml) was added to a flask containing diamino-
propane (4 ml, 47.4 mmol, 4 eq.) and stirred vigorously. Di-t-
butyl dicarbonate (2.56 g, 11.8 mmol) was dissolved in anhydrous
THF (2 ml) and slowly added to the flask. The mixture was
stirred vigorously for 3 h, by which time the starting di-t-butyl
dicarbonate had completely reacted (tlc 20% MeOH–CHCl3 +
1% Et3N, di-t-butyl dicarbonate Rf 0.9, 10 Rf 0.2). The THF was
removed in vacuo and the crude product redissolved in 100 ml
ether. The ethereal layer was washed with water (1 × 100 ml),
water–1% brine–1% 0.2 M sodium acetate buffer (2 × 100 ml) and
then with brine (1 × 100 ml). The aqueous layers were combined,
and brine (40 ml) and saturated sodium bicarbonate (40 ml) added.
This was then extracted with CHCl3 (200 ml, then 3 × 50 ml). The
combined CHCl3 layers were dried over MgSO4 and the solvent
removed in vacuo to give 10 (1.175 g, 6.73 mmol, 57%) as a yellow
oil identical by NMR to the literature.18
(3’-Oxycarbonylpropyl)-2,3,4,6-tetra-O-acetyl-a-D-
mannopyranoside 8
To a flask containing 3-benzyloxycarbonylpropyl-2,3,4,6-tetra-
O-acetyl-a-D-mannopyranoside 7 (1.16 g, 2.2 mmol) was added
anhydrous EtOH (30 ml) and 10% Pd/C (0.1 g). This was
hydrogenated at room temperature and 1 atm with continuous
stirring in the presence of an excess of H2 over 4 h, monitoring
by tlc. Filtration through Celite and removal of the ethanol in
vacuo yielded the title compound as an opaque viscous yellow oil
21
(0.82 g, 1.89 mmol, 85%); Rf 0.6 (1 : 4 ethanol–CH2Cl2); [a]D
+
36.0◦ (c 1, CHCl3); mmax (CHCl3) 3024–3014 (w), 1757–1730 (s),
1429 (w), 1369 (m), 1252 (s) cm−1; 1H NMR dH (300 MHz, CDCl3)
1.94–2.02 (2H, m, CH2CH2CH2O) 1.98 (3H, s), 2.03 (3H, s), 2.08
(3H, s), 2.13 (3H, s), 2.44 (2H, t, J 6.9, CH2CH2CH2O), 3.49 (1H,
dt, J 5.9 and 9.6, CH2CH2CH2O), 3.68 (1H, dt, J 5.9 and 9.6,
CH2CH2CH2O), 3.93–3.99 (1H, m, H-5), 4.07 (1H, dd, J 2.3 and
12.3, H-6), 4.25 (1H, dd, J 5.3 and 12.3, H-6), 4.79 (1H, d, J 1.6,
H-1), 5.21 (1H, dd, J 1.6 and 3.0, H-2), 5.25–5.31 (2H, m, H-3 and
H-4), 7.98 (1H, br s, –COOH); 13C NMR dH (100 MHz, CDCl3)
20.6, 20.7, 20.8, 24.4, 30.7, 62.4, 66.0, 67.0, 67.2, 68.5, 69.1, 69.4,
97.5, 169.8, 170.1, 170.8, 176.7, 177.9; m/z (FAB) 557.1316 (M +
Na+. C18H26O12 requires 557.1322), m/z (ES+) 169 (8%), 331 (5%)
and 457 (100%)
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)-
propylcarbonyl)-Nꢀ-(t-butyloxycarbonyl)propane-1,3-diamine 11
3-Succinimidoxycarbonylpropyl-2,3,4,6-tetra-O-acetyl-a-D-man-
nopyranoside 9 (1 g, 1.9 mmol, 1.2 eq.) was dissolved in
anhydrous CH2Cl2 (10 ml) in a round-bottomed flask. N-(t-
Butyloxycarbonyl)propane-1,3-diamine 10 (279 mg, 1.6 mmol)
was dissolved in anhydrous CH2Cl2 (2 ml) and added to the
reaction at room temperature with stirring. A further 8 ml of
anhydrous CH2Cl2 was added to the reaction along with three
drops of anhydrous Et3N. The reaction was stirred overnight, then
washed with 20% brine (3 × 50 ml) and the organic layer dried over
MgSO4. Removal of the solvent in vacuo gave an oil, which was
purified by flash chromatography (EtOAc: Rf 0.35 (1 : 19 EtOH–
CH2Cl2)) to give 11 as an oil (0.78 g, 1.31 mmol, 82%) that partially
3-Succinimidoxycarbonylpropyl-2,3,4,6-tetra-O-acetyl-a-D-
mannopyranoside 9
◦
crystallised overnight at 4 C; [a]D + 23.2◦ (c 2.5, CHCl3); mmax
21
(CHCl3) 3024–3014 (w), 1745 (s), 1700, 1662 (m), 1512 (m), 1440–
1
3-Oxycarbonylpropyl-2,3,4,6-tetra-O-acetyl-a-D-mannopyrano-
side 8 (1.4 g, 2.7 mmol) was dissolved in anhydrous CH2Cl2
(30 ml), and the flask cooled in an ice–water bath. To the reac-
tion was added 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride (EDCI, 1 g, 5.4 mmol, 2 eq.) and N-hydroxy-
succinimide (NHS) (0.62 g, 5.4 mmol, 2 eq.). The reaction was
stirred for 18 h and allowed to reach room temperature. The
reactionmixturewas washed with 20%brine(3 × 50 ml), dried over
Na2SO4, filtered and the solvent removed in vacuo to obtain 9 as a
pale yellow oil (1.55 g, 2.9 mmol) which was used without further
purification; mmax (CHCl3) 3024–3014 (w), 1741 (s), 1431 (w), 1369
1433 (br w), 1367 (w), 1230 (s) cm−1; H NMR dH (300 MHz,
CDCl3) 1.33 (9H, s), 1.52 (2H, quin, J 5.9, NCH2CH2CH2N),
1.79–1.92 (2H, m, CH2CH2CH2O), 1.89 (3H, s), 1.94 (3H, s), 2.0
(3H, s), 2.05 (3H, s), 2.20 (2H, t, J 7.3, CH2CH2CH2O), 3.05 (2H,
q, J 6.1 Hz, NCH2CH2CH2N or NCH2CH2CH2N), 3.18 (2H, m,
NCH2CH2CH2N or NCH2CH2CH2N), 3.39 (1H, dt, J 6.2 and
9.8, CH2CH2CH2O), 3.66 (1H, dt, J 6.0 and 9.8, CH2CH2CH2O),
3.86–3.91 (1H, m, H-5), 3.97–4.02 (2H, m, H-6 and NH), 4.18 (1H,
dd, J 5.2 and 12.2, H-6), 4.70 (1H, d, J 1.4, H-1), 5.11–5.22 (3H,
m, H-2, H-3 and H-4), 6.44 (1H, m, NH); 13C NMR dH (100 MHz,
CDCl3) 20.7, 20.71, 20.8, 25.1, 28.3, 30.1, 32.8, 62.4, 66.1, 67.5,
1824 | Org. Biomol. Chem., 2006, 4, 1817–1830
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
68.5, 69.1, 69.5, 97.5, 169.7, 170.0, 170.1, 170.7, 172.5; m/z (ES+)
331 (24%), 629 (100%, [M + Na]+).
0.84–0.98 (30H, m, CH3), 1.30–1.36 (2H, m), 1.37 (3H, d, J 7.2),
1.54–1.65 (4H, m), 1.69 (2H, quin, J 6.8), 1.93–1.95 (2H, m), 1.95
(3H, s), 2.03 (3H, s), 2.05 (3H, s), 2.13 (3H, s), 2.02–2.10 (3H,
m), 2.11–2.14 (2H, m), 2.27 (2H, d, J 6.7), 2.31 (2H, t, J 7.4),
2.34–2.36 (2H, m), 3.17–3.26 (4H, m), 3.51 (1H, ddd, J 5.9, 6.4,
9.8), 3.75 (1H, ddd, J 6.0, 6.6, 9.8), 3.90–3.96 (2H, m), 3.98–4.03
(3H, m), 4.09 (1H, dd, J 2.5, 12.2), 4.13 (1H, d, J 7.8 Hz), 4.15
(1H, d, J 8.1), 4.23–4.28 (2H, m), 4.82 (1H, d, J 1.4), 4.88 (1H, s,
OH), 5.20–5.24 (3H, m); 13C NMR dH (CD3OD, 125.7 MHz) 18.1
(C-16ꢀ), 18.9, 19.1, 19.9, 20.0, 20.6, 20.64, 20.7, 22.39, 22.4, 22.77,
22.8, 23.7, 23.8, 25.8 & 25.9, 26.6, 27.5, 30.1, 31.4 & 31.5, 33.7,
37.8 (2C), 41.3, 41.4, 41.6 & 42.0 (4C), 46.0, 51.4, 52.3 & 52.8, 60.7
& 60.9, 63.6, 67.8, 68.6, 69.8, 70.7, 70.8, 71.1, 71.5, 98.9, 171.5,
171.52, 171.6, 172.4, 173.8, 173.9, 174.1, 174.2, 175.4, 175.5,
174.8; m/z (ES+) 1180.6886 (M + Na+. C55H95N7O19 requires
1180.6580), m/z (APCI+) 169.0 (100%), 331.2 (14%), 810.7 (5%,
Pepstatin N-hydroxysuccinimide 12
Pepstatin A (176 mg, 260 lmol), EDCI (493 mg, 2.6 mmol, 10 eq.)
and NHS (296 mg, 2.6 mmol, 10 eq.) were placed in a flask. To
this was added anhydrous DMF (5 ml) and the reaction stirred
overnight. The DMF was removed in high vacuo. The solid was
loosened from the vessel walls, washed with water (20 ml) and
collected by filtration. The solid was further washed with water
(180 ml) and then with diethyl ether (50 ml). The wet solid was
then dried for 24 h under vacuum over anhydrous P2O5. This gave
187 mg (239 ◦lmol, 93%) of the titled product as a white solid;
mp 227–229 C; mmax (KBr) 3500–3200 (s), 2959–2871 (m), 1818
(w), 1785 (w), 1740 (s), 1650–1617 (s), 1534 (s) cm−1; 1H NMR dH
(d6-DMSO, 300 MHz) 0.50–0.70 (30H, m, CH3), 0.98 (3H, d, J
6.81), 0.95–1.25 (4H, m), 1.25–1.40 (2H, m), 1.70–1.85 (3H, m),
1.85–1.95 (2H, m), 2.25–2.33 (4H, m), 2.60 (4H, s), 3.55–3.77 (4H,
m), 3.90–4.10 (3H, m), 4.58 (1H, d, J 4.4, OH), 5.04 (1H, d, J 9.6,
OH), 7.42 (2H, m, NH), 7.73 (1H, d, J 9.6, NH), 7.78 (1H, d, J 9.1,
NH), 7.87 (1H, d, J 6.7, NH); 13C NMR dH (d6-DMSO, 75.4 MHz)
18.1, 18.2, 18.3, 18.38, 18.4, 18.41, 19.28, 19.33, 21.6, 21.8, 22.3,
23.3, 23.5, 24.2, 24.8, 25.5, 25.7, 30.1, 30.2, 30.3, 44.4, 48.3, 50.5,
57.8, 169.2, 169.3, 169.7, 170.2, 170.5, 170.6, 171.6, 172.4; m/z
(ES) 805.4194 (M + Na+. C38H66N6O11 requires 805.4687), m/z
(APCI+) 215.4 (83%), 283.4 (33%), 511.6 (67%), 668.7 (100%),
783.7 (33%).
[M − mannopyranose]+), 1158.9 (7% [M + H]+). Full 1H and 13
C
NMR analysis was carried out on 13 (COSY, TOCSY) and the
chemical shifts fully assigned: these assignments were then used
to evaluate other pepstatin conjugates. Details of the 1H and 13
assignments are given in the ESI.†
C
N-(3-(a-D-Mannopyranosyloxy)propylcarbonyl)-Nꢀ-
(pepstatinyl)propane-1,3-diamine 1
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)propylcar-
bonyl)-Nꢀ-(pepstatinyl)propane-1,3-diamine 13 (20 mg, 17.3
lmol), was dissolved in MeOH (5 ml) in a 25 ml round-bottomed
flask. LiOH (1 M, aq) solution was added dropwise with stirring
until pH 11 was reached. Stirring was continued for 6 h before
removal of the solvents in vacuo. The crude product was purified
by preparative HPLC (gradient, 20% MeCN–water to 60%
MeCN–water over 25 min, RT = 13 min). Removal of the MeCN
in vacuo and lyophilisation gave the titled product as a white
powder (12 mg, 70%, 12 lmol); mp 180 ◦C (dec.); mmax (KBr)
3400–3200 (br s), 2960–2880 (s), 2500–2380 (w), 1685–1632 (s),
1560–1542 (m), 1468–1437 (m), 1206–1187 (s), 1138 (s) cm−1;
1H NMR dH (CD3OD, 400 MHz) 0.86–0.99 (30H, m, CH3),
1.28–1.34 (2H, m), 1.38 (3H, d, J 7.2), 1.53–1.64 (4H, m), 1.68
(2H, m), 1.85–1.92 (2H, m), 2.02–2.10 (3H, m), 2.12 (2H, d, J
6.5), 2.26–2.36 (6H, m), 3.20–3.23 (4H, m), 3.42–3.52 (2H, m),
3.59 (1H, t, J 9.5), 3.66–3.75 (3H, m), 3.78 (1H, dd, J 1.7 and
3.3), 3.82 (1H, dd, J 2.2 and 11.7), 3.93–4.03 (4H, m), 4.11–4.15
(2H, m), 4.26 (1H, q, J 7.1), 4.72 (1H, d, J 1.5), 4.85 (1H, s,
OH), 7.47 (1H, d, J 9.4 NH), 7.70 (1H, d, J 9.3, NH), 7.99 (2H,
m, NH); 13C NMR dH (CD3OD, 125.7 MHz) 18.1, 18.9, 19.2,
20.0, 20.1, 22.4, 22.5, 22.8, 22.9, 23.8, 23.9, 25.9, 26.0, 26.9, 27.5,
30.1, 31.5, 31.6, 34.0, 37.9 (2C), 41.3, 41.5, 41.7, 42.0, 46.0, 51.4,
52.4, 52.8, 60.8, 60.9, 63.0, 67.8, 68.7, 71.2, 71.5, 72.2, 72.7, 74.7,
101.7, 173.8, 174.0, 174.1, 174.2, 175.5, 175.7, 175.8; m/z (ES+)
1012.6140 (M + Na+. C47H87N7O15Na requires 1012.6157), m/z
(ES+) 828.6 (32%, [M − mannopyranose]+), 990.7 (100%, [M +
H]+), 1012.8 (44%, [M + Na]+).
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)-
propylcarbonyl)-Nꢀ-(pepstatinyl)propane-1,3-diamine 13
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)propylcar-
bonyl)-Nꢀ-(t-butyloxycarbonyl)propane-1,3-diamine 11 (650 mg,
1.1 mmol) was added to a flask and dissolved in anhydrous CH2Cl2
(20 ml). TFA (2 ml) was added slowly to the flask while stirring.
After 30 min the reaction was stopped and 10 ml of CH2Cl2
was added. The reaction mixture turned greenish. The solvent
and TFA were removed in vacuo during which time the colour
gradually disappeared to obtain N-(3-(2,3,4,6-tetra-O-acetyl-
a -D -mannopyranosyloxy)propylcarbonyl)propane-1,3-diamine
trifluoroacetate as a viscous oil (0.7 g) which was used without
further purification.
Pepstatin N-hydroxysuccinimide 12 (30 mg, 38 lmol) and
N-(3-(2,3,4,6-tetra-O-acetyl-a-D-mannopyranosyloxy)propylcar-
bonyl)propane-1,3-diamine trifluoroacetate (83 mg, 137 lmol,
5.6 eq.) were both dissolved in anhydrous DMF (5 ml) in a flask.
Anhydrous Et3N (1 ml) was added with stirring and the reaction
was stirred at room temperature for 18 h. The DMF and Et3N
were removed in vacuo, and the crude product was dissolved in
MeOH and filtered through cotton wool to remove pepstatin
dimethylamine (insoluble in MeOH). Purification by preparative
HPLC (gradient, 60% MeOH–water to 95% MeOH–water
over 25 min), and removal of the MeOH in vacuo followed by
lyophilisation, gave the titled product (RT 11 min, Rf 0.2 (1 : 9
MeOH–CH2Cl2)) (20 mg, 17.3 lmol, 45%) as a white powder,
mp 180 ◦C (dec.) mmax (CHCl3) 3400–3200 (w, amide), 3024–2855
(m), 2401 (w), 1747 (s), 1647 (s), 1537 (m), 1466–1435 (br m),
N-(t-Butyloxycarbonyl)cystamine 14
To a flask containing water (5 ml) and dioxane (4 ml) were added
cystamine dihydrochloride (1.29 g, 5.7 mmol, 4 eq.) and sodium
bicarbonate (0.48 g, 5.7 mmol, 4 eq.) with vigorous stirring.
1
1371 (w), 1227–1213 (s) cm−1; H NMR dH (CD3OD, 500 MHz)
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1817–1830 | 1825
©

View Article Online
Di-t-butyl dicarbonate (313 mg, 1.43 mmol) was dissolved in dio-
xane (1 ml) and slowly added to the flask. The mixture was stirred
vigorously. By 2 h a residue had formed which was re-dissolved by
the addition of MeCN (5 ml). After 4 h the organic solvents were
removed in vacuo to give an aqueous suspension. A few drops of
acetate buffer (0.2 M) were added, resulting in the precipitation
of a solid; this was filtered and washed several times with 10%
brine and 10% bicarbonate solution (50 ml total). The filtrate was
extracted with CH2Cl2 (4 × 50 ml) and the combined organic layers
dried over MgSO4. Removal of the solvent in vacuo gave 11 as a
viscous yellow oil (154 mg, 0.61 mmol, 43%); mmax (CHCl3) 3454
The reaction was monitored by tlc. After 4 h the solvent was
removed in vacuo and purified by flash chromatography (4 cm,
isocratic, 10% EtOH–CH2Cl2) to obtain N-(3-(2,3,4,6-tetra-
O-acetyl-a-D-mannopyranosyloxy)propylcarbonylcystamine
trifluoroacetate as a yellow/orange oil (141 mg, 202 lmol, 75%;
Rf 0.3 (1 : 9 EtOH–CH2Cl2)).
Pepstatin N-hydroxysuccinimide 12 (36 mg, 46 lmol) and
N-(3-(2,3,4,6-tetra-O-acetyl-a-D-mannopyranosyloxy)propylcar-
bonylcystamine trifluoroacetate (33 mg, 48 lmol, > 1 eq.) were
dissolved in anhydrous DMF (4 ml) in a round-bottomed flask.
Anhydrous Et3N (1 ml) was added with stirring and the reaction
was stirred for 18 h at room temperature. The DMF and Et3N
were removed in vacuo, and the crude product dissolved into
MeOH and filtered through cotton wool to remove pepstatin
dimethylamine (insoluble in MeOH). Purification by preparative
HPLC (gradient, 60% MeOH–water to 95% MeOH–water
over 25 min) gave the titled product (RT 11 min, Rf 0.2 (1 : 9
MeOH–CH2Cl2)). Removal of the MeOH in vacuo, and removal
of the water by lyophilisation gave the titled product as a white
powder (28.5 mg, 23 lmol, 55%); mp 180 ◦C (dec.); mmax (CHCl3)
3400–3200 (m), 2962–2850 (m), 1746 (s), 1652 (s), 1540 (m),
1
(w), 3000–2900 (w), 1720–1680 (s), 1506 (s) cm−1; H NMR dH
(CDCl3, 400 MHz) 1.39 (9H, s, t-Bu), 1.78 (2H, s, NH2), 2.72 (2H,
t, J 6.2, NCH2CH2S), 2.73 (2H, t, J 6.3, NCH2CH2S), 2.96 (2H, t,
J 6.1, BocNHCH2CH2S), 3.38–3.44 (2H, m, SCH2CH2NH2), 5.03
(1H, br s, NH); 13C NMR dH (CDCl3, 100.6 MHz) 28.3, 38.4, 39.3,
40.4, 42.5, 155.7; m/z (FAB) 253.1010 (M + H+. C9H20N2O2S2
requires 253.1044), m/z (APCI+) 252.9 (56%).
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)-
propylcarbonyl-Nꢀ-(tert-butyloxycarbonyl)cystamine 15
1
1466–1438 (br m), 1370 (w), 1231–1225 (s) cm−1; H NMR dH
3-Succinimidoxycarbonylpropyl - 2,3,4,6 - tetra - O - acetyl - a - D-
mannopyranoside 9 (500 mg, 0.94 mmol, 1.7 eq.) was dissolved
in anhydrous CH2Cl2 (10 ml) in a round-bottomed flask. N-
(t-Butyloxycarbonyl)cystamine 14 (140 mg, 0.55 mmol) was
dissolved in anhydrous CH2Cl2 (2 ml) and added to the reaction
while stirring at room temperature. A further 8 ml of anhydrous
CH2Cl2 was added to the reaction along with three drops of
anhydrous Et3N. The reaction was stirred overnight, then washed
with 20% brine (3 × 50 ml) and the organic layer dried over MgSO4.
Removal of the solvent in vacuo gave an oil, which was purified by
flash chromatography (4 cm, gradient, 50% hexane–ethyl acetate
to 10% hexane–ethyl acetate, change in 10% hexane every 50 ml;
Rf 0.3 (1 : 9 hexane–ethyl◦acetate)) to give 15 as an oil (244 mg, 363
(CD3OD, 500 MHz) 0.87–0.98 (30H, m, CH3), 1.28–1.35 (2H,
m), 1.38 (3H, d, J 7.2), 1.54–1.64 (4H, m), 1.92–1.97 (2H, m),
1.95 (3H, s), 2.03 (3H, s), 2.06 (3H, s), 2.13 (3H, s), 2.02–2.12
(3H, m), 2.12–2.14 (2H, m), 2.30–2.37 (6H, m), 2.82 (2H, t, J
6.8), 2.83 (2H, t, J 6.6), 3.40–3.55 (5H, m), 3.73–3.78 (1H, m),
3.90–4.03 (5H, m), 4.10 (1H, dd, J 2.6 and 12.3), 4.12–4.16 (2H,
m), 4.23–4.29 (2H, m), 4.82 (1H, d, J 1.0), 4.89 (1H, s, OH),
5.20–5.26 (3H, m), 7.45 (1H, d, J 9.4, NH), 7.67 (1H, d, J 9.1,
NH), 7.96 (1H, d, J 8.4, NH), 8.01 (1H, d, J 7.9, NH), 8.06 (1H,
t, J 5.6, NH), 8.17 (1H, t, J 5.6, NH), 8.20 (1H, d, J 5.7, NH); 13
C
NMR dH (CD3OD, 125.7 MHz) 18.1, 18.9, 19.1, 20.0, 20.1, 20.6,
20.64, 20.66, 20.7, 22.4 (2 C), 22.7, 22.8, 23.7, 23.8, 25.85, 25.9,
26.5, 27.5, 31.4, 31.5, 33.6, 38.3, 38.6, 39.6, 39.7, 41.3, 41.4, 41.6,
41.9, 46.0, 51.4, 52.3, 52.7, 60.8, 60.9, 63.6, 67.3, 68.5, 69.8, 70.7,
70.8, 71.1, 71.5, 98.9, 171.5, 171.55, 171.6, 172.4, 173.7, 173.9,
174.1, 174.2, 175.4, 175.6, 175.8; m/z (ES+) 1258.6237 (M + Na+.
C56H97N7O19S2Na requires 1258.6178), m/z (ES+) 1236.6 (30%,
[M + H]+), 1258.6 (100%, [M − H + Na]+).
21
lmol, 66%); [a]D + 22.4 (c 2.5, CHCl3); mmax (film) 3400–3100 (s),
2980–2900 (s), 1760–1650 (s), 1550–1504 (m), 1435 (w), 1367 (w),
1280–1220 (m) cm−1; 1H NMR dH (400 MHz, CDCl3) 1.38 (9H, s),
1.88–1.94 (2H, m, CH2CH2CH2O) 1.94 (3H, s), 2.0 (3H, s), 2.05
(3H, s), 2.10 (3H, s), 2.27 (2H, t, J 7.2, CH2CH2CH2O), 2.73 (2H,
t, J 6.7, NHCH2CH2S), 2.79 (2H, t, J 5.9, NHCH2CH2S), 3.38
(2H, q, J 6.4, NHCH2CH2S), 3.41–3.47 (1H, m, CH2CH2CH2O),
3.51 (2H, q, J 6.0, NHCH2CH2S), 3.70 (1H, dt, J 5.9 and 9.8,
CH2CH2CH2O), 3.90–3.95 (1H, m, H-5), 4.02–4.07 (1H, m, H-6),
4.27 (1H, dd, J 5.2 and 12.2, H-6), 4.75 (1H, d, J 1.5, H-1), 5.11
(1H, br d, NH), 5.16–5.27 (3H, m, H-2, H-3 and H-4), 6.53 (1H,
br d, NH); 13C NMR dH (100 MHz, CDCl3) 20.6, 20.7, 20.8, 25.0,
28.3, 32.5, 37.7, 38.1, 38.2, 39.4, 62.4, 66.0, 67.4, 68.4, 69.0, 69.4,
97.4, 155.9, 169.6, 169.9, 170.1, 170.6, 172.4; m/z (FAB) 669.2371
(M+. C27H44NO13S2 requires 669.2363), m/z (FAB+) 169 (80%),
331 (100%), 669 (50%, [M + H]+).
N-(3-(a-D-Mannopyranosyloxy)propylcarbonyl)-Nꢀ-
(pepstatinyl)cystamine 2
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)propylcar-
bonyl)-Nꢀ-(pepstatinyl)cystamine 16 (25 mg, 20.2 lmol), was
dissolved in MeOH (5 ml) in a 25 ml round-bottomed flask
and stirred. LiOH (1 M, aq) solution was added dropwise with
stirring until pH 11 was reached. Stirring was continued for
6 h before removal of the solvents in vacuo. The crude product
was purified by preparative HPLC (gradient, 20% MeCN–water
to 60% MeCN–water over 25 min, RT = 12.8 min). Removal
of the MeCN in vacuo and lyophilisation gave 2 as a white
powder (14.3 mg, 13.4 lmol, 66%); mp 180 ◦C (dec.); mmax (KBr)
3600–3200 (br s), 2965–2930 (w), 1690–1630 (s), 1565–1534 (m),
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)-
propylcarbonyl)-Nꢀ-(pepstatinyl)cystamine 16
1
N-(3-(2,3,4,6-Tetra-O-acetyl-a-D-mannopyranosyloxy)propylcar-
bonyl-Nꢀ-(tert-butyloxycarbonyl)cystamine 15 (183 mg, 270 lmol)
was dissolved in anhydrous CH2Cl2 (5 ml) in a round-bottomed
flask. TFA (0.5 ml) was added slowly to the flask while stirring.
1455–1430 (m), 1208 (s), 1146 (s) cm−1; H NMR dH (CD3OD,
500 MHz) 0.87–0.99 (30H, m, CH3), 1.28–1.36 (2H, m), 1.39
(3H, d, J 8.0), 1.58–1.63 (4H, m), 1.85–1.92 (2H, m), 2.04–2.11
(3H, m), 2.14 (2H, d, J 6.7), 2.29–2.38 (6H, m), 2.83 (4H, t,
1826 | Org. Biomol. Chem., 2006, 4, 1817–1830
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
J 6.7), 3.47–3.53 (6H, m), 3.61 (1H, t, J 9.6), 3.64–3.72 (3H,
m), 3.79 (1H, dd, J 1.7 and 3.3), 3.88 (1H, dd, J 2.5 and 11.5),
3.94–4.02 (4H, m), 4.12–4.17 (2H, m), 4.27 (1H, m), 4.74 (1H,
d, J 1.7), 4.92 (1H, s, OH); 13C NMR dH (CD3OD, 125.7 MHz)
18.1, 19.0, 19.2, 20.0, 20.1, 22.4, 22.5, 22.8, 22.9, 23.8, 23.9, 25.9,
26.0, 26.9, 27.5, 31.5, 31.6, 33.9, 38.4, 38.6, 39.7, 39.74, 41.3, 41.4,
41.6, 42.0, 46.0, 51.4, 52.4, 52.8, 60.9, 61.0, 63.0, 67.8, 68.7, 71.1,
71.5, 72.2, 72.7, 74.7, 101.6, 173.9, 174.0, 174.1, 174.2, 175.5,
175.9, 180.9; m/z (ES+) 1090.5787 (M + Na+. C48H89N7O15S2Na
requires 1090.5755), m/z (ES+) 283.2 (38%), 906.5 (22%, [M −
mannopyranose]+), 1068.4 (20%, [M + H]+).
(1H, dd, J 7.2 and 8.9), 4.17 (1H, dd, J 7.4 and 8.7), 4.24 (1H,
quin, J 7.2), 4.82 (1H, d, J 4.8, OH), 4.83 (1H, d, J 4.9, OH), 7.30
(1H, d, J 9.2, NH), 7.43 (1H, d, J 8.8, NH), 7.75 (1H, d, J 8.9,
NH), 7.79 (1H, d, J 8.8, NH), 7.89 (1H, d, J 7.4, NH); 13C NMR
dH (DMSO, 125.7 MHz) 18.1, 18.2, 19.2, 19.3, 21.6, 21.9, 22.2,
23.2, 23.4, 24.1, 25.6, 28.2, 30.0, 30.3, 36.1, 37.5, 38.6, 39.6, 39.8,
39.9, 40.1, 40.2, 44.4, 48.3, 50.4, 50.7, 54.8, 57.8, 58.0, 69.0, 69.1,
77.4, 155.5, 170.6, 170.7, 170.78, 170.8, 171.1, 171.5, 172.1; m/z
(ES+) 864.5785 (M + Na+. C42H79N7O10Na requires 864.5786),
m/z (ES+) 842.6 (100%, [M + H]+), 864.6 (27%, [M + Na]+).
N-(Pepstatinyl)propane-1,3-diamine18
Mannosylated BSA 17
To a flask containing N-(t-butyloxycarbonyl)-Nꢀ-(pepstatinyl)-
propane-1,3-diamine (75 mg, 89 lmol) was added TFA (4 ml),
water (125 ll, 6.9 lmol), triethylsilane (125 ll, 0.77 lmol),
thioanisole (250 ll, 2.1 lmol) and phenol (250 ll, 2.8 lmol).
The reaction was stirred vigorously at room temperature for 4 h.
The TFA was removed in vacuo, and the resulting solid purified
by preparative HPLC (50% MeOH–water to 100% MeOH over
25 min, RT = 14–15 min; Rf 0.2 (1 : 9 MeOH–CHCl3)) to obtain
the title compound as a white powder (60 mg, 81 lmol, 91%); mp
214–216 ◦C (lit18 207–210 ◦C); mmax (KBr) 3500–3200 (s), 3079 (w),
2960–2872 (s), 2416 (w), 1700–1600 (s), 1537 (s), 1467–1417 (m),
1204–1138 (m) cm−1; 1H NMR dH (CD3OD, 500 MHz) 0.86–0.98
(30H, m, CH3), 1.27–1.35 (2H, m), 1.39 (3H, d, J 7.3), 1.55–1.64
(4H, m), 1.84 (2H, quin, J 7.0), 2.02–2.10 (3H, m), 2.14 (2H, d, J
6.8), 2.30–2.38 (4H, m), 2.98 (2H, t, J 7.1), 3.20–3.25 & 3.32–3.36
(2H, m), 3.92–3.97 (2H, m), 3.99–4.03 (2H, m), 4.10–4.14 (2H,
m), 4.21 (1H, q, J 7.2), 4.97 (1H, s, OH); 13C NMR dH (CD3OD,
125.7 MHz) 18.0, 18.9, 19.2, 19.9, 20.0, 22.3, 22.4, 22.7, 22.8,
23.7 & 23.8 (10 C), 25.8, 25.9, 27.4, 28.7, 31.3, 31.4, 36.8, 38.2,
41.2, 41.3, 41.6 (2C), 46.0, 51.6, 51.9, 52.7, 61.0 (2 C), 71.0, 71.3,
173.9, 174.1, 174.3, 174.5, 175.5, 175.9; m/z (ES+) 742.5440 (M +
H+. C37H71N7O8 requires 742.5442), m/z (ES+) 742.6 (100%, [M +
H]+).
4-Isothiocyanatophenyl-a-D-mannopyranose (43 mg, 0.14 mmol,
approx. 60 eq.) and bovine serum albumin (153 mg, ∼2.3 lmol)
were dissolved in NaCl solution (0.15 M, 10 ml) in a 50 ml round-
bottomed flask and stirred vigorously. The pH was adjusted to
pH 9 by the dropwise addition of 0.1 M NaOH, and the reaction
was stirred continuously for 6 h, maintaining the same pH. At the
end of this time the reaction mixture was stored overnight at 4 ◦C
in the fridge. The pH was then adjusted to pH 7 and the reaction
mixture dialysed with 0.15 M NaCl solution (1L) for three days,
changing the solution twice a day and stirring the dialysis solution
continuously within a fridge. The dialysis tubing contents were
further purified and de-salted using centrifugal filtration devices
(Millipore) spun within a fixed 35◦ angle centrifuge at a relative
centrifugal force (rcf) of 3000 for 90 min, 21 ◦C (maximum
4 ml solution per device) before re-suspending the gelatinous
product at the bottom of the device with water (4 ml). This
was repeated three times to maximise purification. Lyophilisation
gave 17 as a fine white powder (126 mg, 74%); mp dec. above
260 ◦C; MS (MALDI-TOF): MS peak range = 67,000–79,000;
top of peak for BSA ∼66,430, 17 ∼74,032, difference = 7602,
implying approximately 24.3 sugar units conjugated to each
BSA molecule. Biochemical analysis: Analysis of free TNBS
reactive amino groups28 showed an approximate conjugation of
23 sugar units. Analysis of attached sugar units27 showed an
approximate conjugation of 22.5 sugar units. The mass spectra,
and details of the biochemical analysis, are shown in the ESI.†
N-(Iodoacetyl)-Nꢀ-(pepstatinyl)propane-1,3-diamine 18
N-(Pepstatinyl)propane-1,3-diamine (10 mg, 13.5 lmol) was
placed in a flask and any water present removed by azeotroping
with anhydrous MeCN (3 × 1 ml). The amine was dissolved
in anhydrous DMF (3 ml) and iodoacetic anhydride (5.3 mg,
15 lmol, 1.1 eq.) was added dropwise with stirring. The flask
was covered with aluminium foil to exclude all light and the
reaction stirred continuously for 2 h. The DMF was removed
in vacuo and the crude product stored overnight in a fridge. The
flask contents were then washed with water containing 1% v/v
saturated NaHCO3 solution and 10% v/v saturated KI solution.
The product precipitated and was filtered through cotton wool.
The solid product was washed again with the aqueous KI–
NaHCO3 solution (40 ml) and once with water. The precipitate
was then dissolved in EtOH and transferred to a round-bottomed
flask. Removal of the EtOH in vacuo gave 18 as a white powder
(6 mg, 6.6 lmol, 49%), which was stored in a fridge under Ar
and used in the next reaction as soon as possible; mp 189–191 ◦C;
mmax (KBr) 3500–3150 (s), 3079 (w), 2959–2871 (m), 1700–1635 (s),
N-(t-Butyloxycarbonyl)-Nꢀ-(pepstatinyl)propane-1,3-diamine18
N-(t-Butyloxycarbonyl)propane-1,3-diamine 10 (111 mg,
640 lmol, 10 eq.), pepstatin N-hydroxysuccinimide 12 (50 mg,
64 lmol) and anhydrous DMF (6 ml) were added to a round-
bottomed flask and stirred continuously for 2 days. The solvent
was removed in vacuo and the crude product purified by flash
chromatography (1 cm3, isocratic, 10% MeOH–CH2Cl2: Rf 0.5 (1 :
9 MeOH–CHCl3)). The solvent was removed in vacuo to give an
off-white solid. This was re-crystallised from CHCl3–cyclohexane
to obtain the title c◦ompound as a white powder (48 mg, 57 lmol,
89%); mp 219–221 C (lit18 215–217 ◦C); mmax (KBr) 3400–3200 (s,
amide), 2960–2872 (s), 2360–2331 (m), 1652–1634 (s), 1533 (m),
1465–1437 (m), 1171 (m) cm−1; 1H NMR dH (DMSO, 500 MHz)
0.77–0.85 (30H, m, CH3), 1.18 (3H, d, J 7.1), 1.21–1.26 (2H, m),
1.33–1.37 (2H, m), 1.36 (9H, s), 1.46–1.52 (2H, m), 1.92–1.98 (3H,
m), 2.00–2.05 (4H, m), 2.11 (2H, d, J 6.2), 2.90 (2H, q, J 6.4),
2.97–3.05 (2H, m), 3.27–3.34 (2H, m), 3.75–3.83 (4H, m), 4.11
1
1550–1507 (s) cm−1; H NMR dH (CD3OD, 500 MHz) 0.86–0.98
(30H, m, CH3), 1.32–1.39 (2H, m), 1.38 (3H, d, J 7.2), 1.54–1.65
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1817–1830 | 1827
©

View Article Online
(4H, m), 1.70 (2H, quin, J 6.8), 2.02–2.10 (3H, m), 2.13 (2H, d,
J 6.7), 2.27 (2H, d, J 6.7), 2.34–2.37 (2H, m), 3.20–3.34 (4H, m),
3.69 (2H, s), 3.93–4.01 (4H, m), 4.13 (1H, d, J 7.9), 4.15 (1H,
d, J 8.1), 4.26 (1H, q, J 7.3), 4.89 (1H, s, OH), 7.32 (1H, d, J
9.3, NH), 7.45 (1H, s, NH), 7.68 (1H, d, J 9.0, NH), 7.89 (1H, s,
NH); 13C NMR dH (CD3OD, 125.7 MHz) −2.0 (CH2-I), 18.1, 18.9,
19.1, 20.0, 20.1, 22.4 (2H), 22.7, 22.8, 23.7, 23.8, 25.8, 25.9, 27.5,
29.8, 31.4, 31.5, 37.8, 38.4, 41.3, 41.4, 41.7, 42.0, 46.0, 51.4, 52.3,
52.8, 60.7, 60.9, 71.1, 71.4, 171.4, 173.8, 173.9, 174.1, 174.2, 175.4,
175.8; m/z (ES+) 932.4311 (M + Na+. C39H72IN7O9Na requires
932.4334), m/z (ES+) 932.3 (83%, [M + Na]+).
942.5384), m/z (ES+) 920.6 (100%, [M + H]+), 942.6 (53%, [M–
H + Na]+).
N-(Pepstatinyl)cystamine
To a flask containing N-(t-butyloxycarbonyl)-Nꢀ-(pepstatinyl)-
cystamine (65 mg, 70 lmol) was added TFA (4 ml), water (125 ll,
6.9 lmol), triethylsilane (125 ll, 0.77 lmol), thioanisole (250 ll,
2.1 lmol) and phenol (250 ll, 2.8 lmol). The reaction was stirred
vigorously for 4 h. The TFA was removed in vacuo, and the mixture
purified by preparative HPLC (40% MeCN–water to 80% MeCN
over 30 min; RT = 5 min; Rf 0.2 (1 : 9 MeOH–CHCl3)) to obtain
the title compound as a white powder (37 mg, 45 lmol, 64%); mp
219–221 ◦C; mmax (KBr) 3500–3200 (s), 2960–2872 (m), 1700–1600
Non-cleavable neomannosylated BSA–pepstatin conjugate 3
Mannosylated BSA 17 (30 mg, ∼3.8 × 10−7 moles) was placed in
a 25 ml round-bottomed flask and dissolved in 2 ml of phosphate
buffer saline (PBS, 1 M, pH 8.0) with stirring. To this was added N-
(iodoacetyl)-Nꢀ-(pepstatinyl)propane-1,3-diamine 18 (6 mg ml−1
stock in anhydrous DMF, 350 lg, 3.8 × 10−7 moles) and the
reaction was stirred overnight at room temperature. The reaction
mixture was dialysed once against 1 M PBS solution (1L, pH 7.4)
for one day, changing the solution once and stirring the dialysis
solution continuously at 4 ◦C. Further purification and de-salting
was achieved by using Millipore ultrafree-4 centrifugal filtration
devices spun within a fixed 35◦ angle centrifuge at a relative
centrifugal force (rcf) of 3000 for 90 min at 21 ◦C (crude product
made up to 4 ml with water and total volume of 10% DMSO). The
gelatinous product at the bottom of the device was re-suspended in
water (4 ml) and re-centrifuged in a new filtration device (3000 rcf,
90 min) and repeated within the same device by re-suspending
in water (4 ml). Lyophilisation gave 3 as a fine white powder
(22 mg, 73%); mp dec. above 260 ◦C; MS (MALDI-TOF): MS
peak range = 67 500–80 000.
(s), 1540 (s), 1470–1419 (w), 1203–1138 (m) cm−1; H NMR dH
1
(CD3OD, 400 MHz) 0.86–0.98 (30H, m, CH3), 1.27–1.35 (2H, m),
1.38 (3H, d, J 7.3), 1.55–1.63 (4H, m), 2.02–2.10 (3H, m), 2.13
(2H, d, J 6.5), 2.28–2.38 (4H, m), 2.87 (2H, t, J 6.7), 2.97 (2H, t,
J 6.8), 3.24–3.32 (2H, m), 3.42–3.62 (2H, m), 3.91–4.04 (4H, m),
4.10–4.17 (2H, m), 4.23 (1H, q, J 7.3), 4.93 (1H, s, OH), 7.47 (1H,
d, J 9.4, NH), 7.63 (1H, d, J 9.7, NH), 7.67 (1H, d, J 9.1, NH),
7.97 (1H, d, J 8.3, NH), 8.05 (1H, d, J 7.63, NH), 7.63 (1H, s,
NH); 13C NMR dH (CD3OD, 100.6 MHz) 18.0, 18.9, 19.2, 20.0,
20.1, 22.4 (2C), 22.7, 22.8, 23.7, 23.8, 25.8, 25.9, 27.5, 31.3, 31.5,
35.6, 38.3, 39.4, 39.5, 41.2, 41.3, 41.6, 41.9, 46.0, 51.5, 52.2, 52.7,
60.9 (2C), 71.0, 71.5, 173.8, 173.9, 174.2, 174.3, 175.5, 175.9; m/z
(ES+) 820.5039 (M + H+. C38H73N7O8S2 requires 820.5040), m/z
(ES+) 820.5 (100%, [M + H]+).
N-(Iodoacetyl)-Nꢀ-(pepstatinyl)cystamine 19
N-(Pepstatinyl)cystamine (7 mg, 8.3 lmol) was placed in a round-
bottomed flask and any water present removed by azeotroping
with anhydrous MeCN (3 × 1 ml). The amine was then dissolved
in anhydrous DMF (3 ml) and iodoacetic anhydride (3.3 mg,
9.4 lmol, 1.1 eq.) added dropwise with stirring. The flask was
covered with aluminium foil to exclude all light and the reaction
was stirred continuously for 2 h. The DMF was removed in vacuo
and the crude product stored overnight at 4 ◦C. The next day
the flask contents were washed with water containing 1% v/v
saturated bicarbonate solution and 10% v/v saturated potassium
iodide (KI) solution. The product precipitated and was filtered
through cotton wool. The solid product was washed again with
the aqueous KI/bicarbonate solution (40 ml) and once with water.
The precipitate was then dissolved in EtOH and transferred to a
round-bottomed flask. Removal of the EtOH in vacuo gave the
title compound as a white powder (6 mg, 6.1 lmol, 83%), which
was stored in a fridge under Ar and used in the next reaction as
soon as possible; mp 198–201 ◦C; mmax (KBr) 3600–3100 (m), 2959–
N-(t-Butyloxycarbonyl)-Nꢀ-(pepstatinyl)cystamine
N-(t-Butyloxycarbonyl)cystamine 14 (161 mg, 640 lmol, 10 eq.),
pepstatin N-hydroxysuccinimide 12 (50 mg, 64 lmol) and
anhydrous DMF (6 ml) were added to a round-bottomed flask
and stirred continuously for 2 days at room temperature. The
solvent was then removed in vacuo and the mixture purified by
flash chromatography (1 cm3, isocratic, 10% MeOH–CH2Cl2, Rf
0.55). This gave a crude white powder, which was re-crystallised
from hot EtOH to obtain the title compound as a white powder
(47 mg, 64 lmol, 80%); mp 244–246 ◦C; mmax (KBr) 3600–3100 (s),
2959–2871 (s), 1635 (s), 1534 (s), 1466–1437 (m), 1171 (m) cm−1; 1H
NMR dH (DMSO, 500 MHz) 0.77–0.85 (30H, m, CH3), 1.18 (3H,
d, J 7.1), 1.20–1.26 (2H, m), 1.33–1.38 (2H, m), 1.36 (9H, s), 1.48–
1.52 (2H, m), 1.90–1.99 (3H, m), 2.00–2.08 (4H, m), 2.11 (2H, d, J
6.2), 2.72 (2H, t, J 6.5), 2.73 (2H, t, J 6.5), 3.16–3.20 (2H, m), 3.25–
3.31 (2H, m), 3.74–3.81 (4H, m), 4.11 (1H, d, J 7.2), 4.11 (1H, d, J
7.3), 4.23 (1H, q, J 7.0), 4.82 (1H, d, J 5.3, OH), 4.83 (1H, d, J 5.2,
OH), 7.32 (1H, d, J 9.2, NH), 7.45 (1H, d, J 8.8, NH), 7.77 (1H,
1
2871 (m), 1700–1617 (m), 1540 (w), 1067 (s) cm−1; H NMR dH
(CD3OD, 500 MHz) 0.87–0.99 (30H, m, CH3), 1.28–1.36 (2H, m),
1.38 (3H, d, J 7.3), 1.57–1.62 (4H, m), 2.03–2.11 (3H, m), 2.13 (2H,
d, J 6.7), 2.28 (2H, d, J 6.7), 2.34–2.37 (2H, m), 2.83 (4H, t, J 6.6),
3.43–3.61 (4H, m), 3.71 (2H, s), 3.91–4.04 (4H, m), 4.12 (1H, d, J
7.8), 4.15 (1H, d, J 8.0), 4.26 (1H, q, J 9.3), 4.89 (1H, s, OH), 7.43
(1H, s, NH), 7.45 (1H, s, NH), 7.68 (1H, d, J 9.0, NH); 13C NMR
dH (CD3OD, 125.7 MHz) −2.1 (CH2-I), 18.1, 18.9, 19.1, 20.0, 20.1,
22.4 (2C), 22.7, 22.8, 23.7, 23.8, 25.8, 25.9, 27.5, 31.4, 31.6, 38.2,
38.3, 39.7, 40.1, 41.3, 41.4, 41.7, 42.0, 46.0, 51.4, 52.3, 52.8, 60.8,
d, J 9.0, NH), 7.81 (1H, d, J 8.8, NH), 7.91 (1H, d, J 7.4, NH); 13
C
NMR dH (DMSO, 125.7 MHz) 18.2, 18.3, 19.2, 19.3, 21.6, 21.9,
22.2, 23.3, 23.5, 24.2, 25.7, 28.2, 30.1, 30.3, 37.1, 37.6, 37.9, 38.0,
38.6, 39.7, 39.8, 40.0, 40.1, 40.2, 44.3, 48.3, 50.4, 50.7, 57.7, 57.9,
69.0, 69.1, 77.8, 155.5, 170.7, 170.8, 170.9, 171.0, 171.02, 171.6,
172.2; m/z (ES+) 942.5378 (M + Na+. C43H81N7O10S2Na requires
1828 | Org. Biomol. Chem., 2006, 4, 1817–1830
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
60.9, 71.1, 71.4, 173.8, 173.9, 174.1, 174.2, 175.4, 175.5, 175.8; m/z
(ES+) 988.2 (56%, [M + H]+), 1010.3 (100%, [M − H + Na]+).
Eagles medium (Cancer Research UK) containing 10% foetal
calf serum (FCS; PAA Laboratories, Linz, Austria); 100 U ml−1
Penicillin, 100 lg ml−1 Streptomycin (Cancer Research UK) and
50 lM 2-mercaptoethanol (2-ME; Gibco, Paisley, UK). Cells were
maintained at a density of between 2 × 105 cells/ml and 5 ×
105cells/ml. RPMI 1640 with L-Glutamine (Gibco). CTLL-2 (a
kind gift from Dr B. Stockinger, National Institute for Medical
Research, London, UK) is an IL-2-dependent human T-cell line.
Cells were cultured in complete RPMI 1640 medium containing
10% FCS.
Cleavable neomannosylated BSA–pepstatin conjugate 4
Mannosylated BSA 17 (30 mg, ∼3.8 × 10−7 moles) was placed in
a 25 ml round-bottomed flask and dissolved in 2 ml of phosphate
buffer saline (PBS, 1 M, pH 8.0) with stirring. To this was added N-
(iodoacetyl)-Nꢀ-(pepstatinyl)cystamine 19 (taken from 6 mg ml−1
stock in anhydrous DMF, 380 lg, 3.8 × 10−7 moles) and the
reaction was continually stirred overnight. The reaction mixture
was dialysed once against 1 M PBS solution (1L, pH 7.4) for
one day, changing the solution once and stirring the dialysis
solution continuously within a fridge. Further purification and
de-salting was achieved by using Millipore ultrafree-4 centrifugal
filtration devices spun within a fixed 35◦ angle centrifuge at a
relative centrifugal force (rcf) of 3000 for 90 min, 21 ◦C (crude
product made up to 4 ml with water and total volume of 10%
DMSO). The gelatinous product at the bottom of the device was
re-suspended in water (4 ml) and re-centrifuged in a new filtration
device (3000 rcf, 90 min) and repeated within the same device by
re-suspending in water (4 ml). Lyophilisation gave 4 as a fine white
powder (21 mg, 70%); dec. above 260 ◦C; MS (MALDI-TOF):
MS peak range = 69 000–81 500. In order to remove any last
traces of solvent or other small molecular weight impurities, 5 mg
material was dissolved in 0.5 ml water and passed over a prepacked
column of Sephadex G10 chromatography medium (Pharmacia),
preequilibrated in tissue culture medium containing 10% foetal
calf serum. The conjugate was eluted following the manufacturers
instructions, and stored in aliquots at −20 ◦C until required.
Dendritic cells were obtained from bone marrow of Balb/c mice
as described elsewhere.37
The assay for antigen processing/presentation has been de-
scribed previously.31 Briefly DO11-10 cells (5 × 104) and either
A20 cells (5 × 104) or dendritic cells (2.5 × 104) were co-cultured
in wells of a tissue culture plate (Nunc) in a total volume of 200 ll.
Cells were stimulated with ovalbumin (Sigma Aldrich, Poole,
Dorset), or a synthetic peptide (Cancer Research UK) coding for
the ovalbumin sequence 323–339 which is the epitope recognised
by the DO11-10 T cells. The cells are cultured for 24 hours
at 37 ◦C, 5% CO2, and supernatants collected and stored at
−20 ◦C until assayed for IL-2. IL-2 levels were measured either by
ELISA (eBioscience, London, UK) according to manufacturer’s
instructions (Fig. 3), or by bioassay using the IL-2 indicator cell
line CTLL, and measuring incorporation of tritiated thymidine
(Figs 4–6) as described.31 Results are expressed as % IL-2, relative
to IL-2 levels in the presence of ovalbumin or ovalbumin synthetic
peptide, but no inhibitor. 100% levels correspond to 1–10 ng ml−1
IL-2 in the culture supernatants.
Acknowledgements
Enzymatic studies of inhibitors
Rat Cathepsin E was prepared as described.30 Stocks were stored
in 0.05 M sodium acetate buffer, 50% glycerol, 0.25 M NaCl, at
−20 ◦C.
We thank UCL for the award of a Graduate School Interdisci-
plinary Scholarship (to P. F. F.) the Japanese Science and Edu-
cation Ministry/British Council for a MONBUSHO Fellowship
(to P. F. F.) and the Nuffield Foundation for an Undergraduate
Research Bursary (to C. A. H.).
The fluorogenic peptide substrate MOCAc-Gly-Lys-Pro-Ile-
Leu-Phe-Phe-Arg-Leu-Lys(Dnp)c-NH2 (5 ll of 100 lM in H2O;
BAChem, Meyerside, UK) known to be cleaved by cathepsin D
or E30 was added to 80 ll 0.1 M sodium acetate buffer (pH 4.0).
To this was added 5 ll of inhibitor solution at a series of two-fold
dilutions from various concentrations (dependent on the inhibitor
used). Samples were incubated at 37 ◦C for several minutes in
a water bath. 10 ll of enzyme solution (0.01 M sodium acetate
buffer, pH 5.5) at 0.5 lg ml−1 was added to each sample, and
References
1 (a) J. A. Villadangos, R. A. Bryant, J. Deussing, C. Driessen, A. M.
Lennon-Dumenil, R. J. Riese, W. Roth, P. Saftig, G. P. Shi, H. A.
Chapman, C. Peters and H. L. Ploegh, Immunol. Rev., 1999, 172, 109–
120; (b) P. G. Medd and B. M. Chain, Semin. Cell Dev. Biol., 2000, 11,
203–210.
2 H. Yang, M. Kala, B. G. Scott, E. Goluszko, H. A. Chapman and P.
Christadoss, J. Immunol., 2005, 174, 1729–1737.
◦
the reaction was allowed to continue at 37 C for 30 min before
adding 200 ll of 5% trichloroacetic acid to precipitate the enzymes.
Negative/positive controls were included as a sample without
enzyme, or contained no inhibitor, respectively. Samples were
centrifuged at 13 000 rpm for a few seconds before reading OD
(excitation 328 nm, emission 393 nm) using a Hitachi fluorescence
spectrophotometer (Hitachi, Japan).
Data were analysed by the Dixon method38 plotted as fluores-
cence (converted to % activation) vs. inhibitor concentration to
obtain the IC50 value for the appropriate inhibitor.
3 N. Katunuma, Y. Matsunaga, K. Hiumeno and Y. Hayashi, Biol. Chem.,
2003, 384, 883–890.
4 P. D. Stahl and R. A. B. Ezekowitz, Curr. Opin. Immunol., 1998, 10,
50–55.
5 M. C. A. A. Tan, A. M. Mommaas, J. W. Drtijfhout, R. Jordens,
J. J. M. Onderwater, D. Verwoerd, A. A. Mulder, A. N. van der Heiden,
D. Scheidegger, L. C. J. M. Oomen, T. H. M. Ottenhoff, A. Tulp, J. J.
Neefjes and F. Koning, Eur. J. Immunol., 1997, 27, 2426.
6 E. A. L. Biessen, F. Noorman, M. E. van Teijlingen, J. Kuiper, M.
Barrett-Bergshoeff, M. K. Bijsterbosch, D. C. Rijken and T. J. C. van
Berkel, J. Biol. Chem., 1996, 271, 28024–28030.
7 F. Noorman, M. Barrett-Bergshoeff, E. A. L. Biessen, E. van de Bilt,
T. J. C. van Berkel and D. C. Rijken, Hepatology, 1997, 26, 1303–1310.
8 C. Grandjean, C. Rommens, H. Gras-Masse and O. Melnyk, Angew.
Chem. Int. Ed., 2000, 39, 1068–1072.
9 C. Grandjean, C. Rommens, H. Gras-Masse and O. Melnyk, Tetrahe-
dron Lett., 1999, 40, 7235–7238.
Assays for antigen processing/presentation
A20 cells (B lymphoma) and DO11-10 cells (T cell hybridoma
specific for ovalbumin) were cultured in Dulbecco’s modified
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1817–1830 | 1829
©

View Article Online
10 O. Kinsel, D. Fattori, P. Ingallinella, E. Blanchi and A. Pessi, J. Pept.
Sci., 2003, 9, 375–385.
11 A. Engel, S. K. Chatterjee, A. Al-arifi, D. Reimann, J. Langner and P.
Nuhn, Pharm. Res., 2003, 20, 51–57.
12 A. Du¨ffels, L. G. Green, S. V. Ley and A. D. Miller, Chem.–Eur. J.,
2000, 6, 1416–1430.
13 S. Kawakami, A. Sato, M. Nishikawa, F. Yamashita and M. Hashida,
Gene Ther., 2000, 7, 292–299.
14 S. Espuelas, P. Haller, F. Schuber and B. Frisch, Bioorg. Med. Chem.
Lett., 2003, 13, 2557–2560.
15 M. J. Copland, M. A. Baird, T. Rades, J. L. McKenzie, B. Becker, F.
Reck, P. C. Tyler and N. M. Davies, Vaccine, 2003, 21, 883–890.
16 C. Ortiz Mellet, J. Defaye and J. M. Garc´ıa Ferna´ndez, Chem.–Eur. J.,
2002, 8, 1982–1990.
24 A. E. Weber, T. A. Halgren, J. J. Doyle, R. J. Lynch, P. K. S. Siegl, W. H.
Parsons, W. J. Greenlee and A. A. Patchett, J. Med. Chem., 1991, 34,
2692–2701.
25 G. T. Hermanson, Bioconjugate Techniques, Academic Press, San
Diego, 1996, pp. 423–426.
26 C. Kieburg and T. K. Lindhorst, Tetrahedron Lett., 1997, 38, 3885–
3888.
27 C. R. McBroom, C. H. Samanen and I. J. Goldstein, Methods Enzymol.,
1972, 28, 212–219.
28 M. Dubois, K. A. Gilles, J. K. Hamilton, P. A. Rebers and F. Smith,
Anal. Chem., 1956, 28, 350–356.
29 (a) R. Fields, Methods Enzymol., 1972, 25, 464–469; (b) A. C. C.
Spadaro, W. Draghetta, S. Nassif del Lama, A. C. M. Camargo and
L. J. Greene, Anal. Biochem., 1979, 96, 317–321.
17 M. M. Ponpipom, R. L. Bugianesi, J. C. Robbins, T. W. Doebber and
T. Y. Shen, J. Med. Chem., 1981, 24, 1388–1395.
18 Pepstatin-mannose-6-phosphate bioconjugates, targeted to specific
mannose-6-phosphate receptors: B. Hamdaoui, G. Dewynter, F.
Capony, J.-L. Montero, C. Toiron, M. Hnach and H. Rochefort, Bull.
Soc. Chim. Fr., 1994, 131, 854–864.
19 Pepstatin-asialofetuin bioconjugates for receptor-mediated uptake into
hepatocytes; K. Furuno, N. Miwa and K. Kato, J. Biochem., 1983, 93,
249–256.
20 (a) M. E. Taylor, K. Bezouska and K. Drickamer, J. Biol. Chem., 1992,
267, 1719–1726; (b) M. E. Taylor and K. Drickamer, J. Biol. Chem.,
1993, 268, 399–404; (c) N. P. Mullin, K. T. Hall and M. E. Taylor,
J. Biol. Chem., 1994, 269, 28405–28413.
30 Y. Yasuda, T. Kageyama, A. Akamine, M. Shibata, E. Kominami, Y.
Uchiyama and K. Yamamoto, J. Biochem. (Tokyo), 1999, 125, 1137–
1143.
31 K. Bennett, T. Levine, J. S. Ellis, R. J. Peanasky, J. Kay and B. M. Chain,
Eur. J. Immunol., 1992, 22, 1519–1524.
32 R. Shimonkevitz, J. Kappler, T. Marrack and H. Grey, J. Exp. Med.,
1992, 158, 303.
33 T. P. Levine and B. M. Chain, Adv. Exp. Med. Biol., 1993, 329, 11–15.
34 K. Inaba, M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S.
Muramatsu and R. M. Steinman, J. Exp. Med., 1992, 176, 1693–1702.
35 T. P. Levine and B. M. Chain, Proc. Natl. Acad. Sci. U. S. A., 1992, 89,
8342–8346.
36 R. L. Thurmond, S. Sun, C. A. Sehon, S. M. Baker, H. Cai, Y. Gu,
W. Jiang, J. P. Riley, K. N. Williams, J. P. Edwards and L. Karlsson,
J. Pharmacol. Exp. Ther., 2004, 308, 268–276.
37 B. M. Chain, P. Free, P. Medd, C. Swetman, A. B. Tabor and N.
Terrazzini, J. Immunol., 2005, 174, 1791–1800.
38 M. Dixon, Biochem. J., 1972, 129, 197–202.
21 J. Brygier, J. Vincentelli, M. Nijs, C. Guermant, C. Paul, D. Baeyens-
Volant and Y. Looze, Appl. Biochem. Biotechnol., 1994, 47, 1–10.
22 G. Saito, J. A. Swanson and K.-D. Lee, Adv. Drug Deliv. Rev., 2003, 55,
199–215.
23 M. Mori, Y. Ito and T. Ogawa, Carbohydr. Res., 1990, 195, 199–224.
1830 | Org. Biomol. Chem., 2006, 4, 1817–1830
This journal is
The Royal Society of Chemistry 2006
©