Maqbool A Siddiqui1, Shailesh Ambre1, Susan K Keay2,3,4, Jeffrey M Rhyne4, Chen-Ou Zhang4, Joseph J Barchi1. 1. Chemical Biology Laboratory, Center for Cancer Research, National Cancer Institute at Frederick, Frederick, Maryland 21702, United States. 2. Medical Service, Veterans Administration Maryland Health Care System, Baltimore, Maryland 21201, United States. 3. Baltimore Research and Education Foundation, Baltimore, Maryland 21201, United States. 4. Department of Medicine and Department of Pathology, University of Maryland School of Medicine, Baltimore, Maryland 21201, United States.
Abstract
Glycoamino acid analogues of the Thomsen-Friedenreich antigen disaccharide, where the 4' and 4″ hydroxyl groups were substituted with fluorine or hydrogen, were synthesized and incorporated into the asialylated antiproliferative factor (as-APF), a biologically active form of APF, a glycopeptide found in the urine of patients with interstitial cystitis. Various strategies were employed to incorporate the fluorine atom at the 4-positions of either the galactose or N-acetylgalactosamine unit of the disaccharide antigen, based on stereochemistry and reactivity. These glycopeptides were evaluated in antiproliferative assays on both primary normal bladder epithelial cells and T24 bladder carcinoma cells. Unlike many previously published substitutions to APF, mono-4'-fluorination of the GalNAc residue did not affect the activity, whereas fluoro-derivatives of the galactose 4″-position or both 4' and 4″ hydroxyls showed a reduced potency relative to the monosubstituted GalNAc derivative. A fourth compound where the 4″ position of galactose was deoxygenated showed a lower potency than the parent and monosubstituted compounds. These results suggest that specific substitutions in the sugar moieties in the APF can be tolerated, and the glycomimetic design of APF analogues can include fluorine in the GalNAc sugar of the disaccharide.
Glycoamino acid analogues of the Thomsen-Friedenreich antigen disaccharide, where the 4' and 4″ hydroxyl groups were substituted with fluorine or hydrogen, were synthesized and incorporated into the asialylated antiproliferative factor (as-APF), a biologically active form of APF, a glycopeptide found in the urine of patients with interstitial cystitis. Various strategies were employed to incorporate the fluorine atom at the 4-positions of either the galactose or N-acetylgalactosamine unit of the disaccharide antigen, based on stereochemistry and reactivity. These glycopeptides were evaluated in antiproliferative assays on both primary normal bladder epithelial cells and T24 bladder carcinoma cells. Unlike many previously published substitutions to APF, mono-4'-fluorination of the GalNAc residue did not affect the activity, whereas fluoro-derivatives of the galactose 4″-position or both 4' and 4″ hydroxyls showed a reduced potency relative to the monosubstituted GalNAc derivative. A fourth compound where the 4″ position of galactose was deoxygenated showed a lower potency than the parent and monosubstituted compounds. These results suggest that specific substitutions in the sugar moieties in the APF can be tolerated, and the glycomimetic design of APF analogues can include fluorine in the GalNAc sugar of the disaccharide.
All mammalian cell
surfaces are coated with oligosaccharide chains
(termed “glycans”) that are presented in various guises
through covalent attachment to proteins and lipids.[1] Proteins primarily display two types of glycosylation:
N-linked (glycan with a reducing end anomeric “β”
N-linkage to the carboxamide of an asparagine residue) and O-linked
(glycan with a reducing end anomeric “α” O-linkage
to the hydroxyl group of a serine or threonine residue); and these
sugars contribute to a variety of protein functions (e.g., folding,
infection, cell–cell communication, and immune responses).[2] Many of these glycan chains are highly modified
during various disease states, such as cancer,[3] with many of these aberrant glycans becoming targets of the immune
system and are hence known as tumor-associated carbohydrate antigens
(TACAs). Two prevalent O-linked TACAs expressed on many solid tumors
are Tn[4] (Thomsen nouvelle, GalNAcα-O-Ser/Thr, Tnag) and its extended core-1 structure,
TF (Thomsen–Friedenreich, Galβ1-3GalNAcα-O-Ser/Thr, TFag) antigens.[5] Truncation of larger normal cell oligosaccharide chains
to these smaller glycan structures serves to expose underlying peptide
sequences within the extracellular domains of cell-surface proteins
(e.g., mucins). As a result, these novel epitopes can be recognized
by the immune system, and thus various strategies to design immunotherapies
against TACAs have been evaluated in recent years.[6−14] In addition, TFag has been directly tied to the metastasis
of breast,[15] prostate,[16−18] and pancreatic[19] cancerthrough its interaction with tumor-derived
galectin-3.Although modified TACAs such as TFag have
been used
in immunological/antitumor studies outlined above, TFag is also found on selected natural products, which have biological
activity that is dependent on the presence of the disaccharide. The
conotoxins,[20] cod glycopeptide,[16] and the antiproliferative factor (APF) from
interstitial cystitis (IC) patients[21] all
contain TFag, which contributes directly to their biological
function. Hence, modified versions of TFag and their use
in synthetic congeners of these active glycopeptides may serve to
influence either the mode of action or the potency of these agents.
Our group has extensively studied APF, a highly potent sialylated
TFag-containing glycopeptide, discovered in the urine of
patients with IC/painful bladder syndrome (PBS), a debilitating disease
of the bladder. The APF treatment of normal bladder (NB) cells causes
cellular changes that are reminiscent of the disease.[22,23] In addition, APF is not only highly potent against NB epithelial
cells but also a powerful antiproliferative agent against a variety
of tumor cell lines in vitro.[24] Our lab
has prepared several analogues of this glycopeptide and has developed
a solid structure–activity relation profile for the antiproliferative
activity of APF.[25,26] Although we previously documented
that a TFagdisaccharide moiety and proper stereochemistry
are critical for its activity, we had not prepared analogues where
selected functional groups of the sugars were substituted. In past
reports on immunological studies of TFag-containing mucinglycopeptides, specific functional groups of TFag were
substituted with a fluorine atom, which is often used as a surrogate
for both hydroxyl groups and hydrogen atoms.[9,27,28] However, the axial 4′-position is
more difficult to replace with fluorine, and hence only one study
of a TF antigen with a 4″-fluoro modification on the galactose
residue was recently reported.[6] Herein,
we describe our own synthesis of a modified version of the previously
reported 4″-fluoro-galactoseglycoamino acids (GAAs) along
with three other analogues (two fluorinated and one 4″-deoxy)
of the TFag and their incorporation into GAAs (Figure A). We used these
new GAAs as building blocks in Fmoc-based solid-phase peptide synthesis
(SPPS) to prepare novel analogues of APF (Figure B) and studied the effect that these substitutions
had on the antiproliferative activity.
Figure 1
(A) Target glycoamino
acids and (B) glycopeptides used in this
study. APF and asialylated APF (as-APF) are labeled
accordingly.
(A) Target glycoamino
acids and (B) glycopeptides used in this
study. APF and asialylated APF (as-APF) are labeled
accordingly.
Results and Discussion
4″-F-Galβ1-3GalNAc
TF GAA (1)
The synthesis of both monomeric 4′-fluoro-galactose
and 4′-fluoro-galactosamine
derivatives has been described in the literature (vide infra); however,
reports for the synthesis of GAAs, with these congeners as precursors
for the glycopeptide synthesis, are sparse. A very recent report by
Hoffmann-Röder and co-workers outlined the first synthesis
of MUC1-based glycopeptides using a TF-antigen derivative with a 4″-fluoro-galactose
moiety.[6] The 4″-fluoro group imparted
resistance to hydrolysis by β-galactosidase, and the authors
were able to prepare MUC1glycopeptide conjugates of this analogue
with the tetanus toxoid protein1 and generate an immune response in
mice.[6] We have prepared a similar 4″-F-Gal-GalNAc-Thr
GAA building block via an alternative process outlined in Scheme . The known galactosamine
precursor A (Figure ) used for our synthesis was initially derived from
our previous work on an improved TFag synthesis,[29,30] but arrived via a slightly shorter procedure, which we found as
more efficient and avoids an azidonitration step. Starting from galactosamine
hydrochloride, access to acetylated 2-azido galactose peracetate derivative 9 proceeded smoothly via the diazo transfer reaction of Wong,[31] using the modified imidazole-based transfer
reagent of Goddard-Borger and Stick.[32] Adapting
from our previous work, this was processed in six high-yielding steps
to the known[30] acceptor A.
The known 4′-fluoro-galactosyl donor 10 was simply
prepared in five steps by the method of Koch and Chambers[33] through the selective acetylation of the 2′,
3′, and 6′ positions of methyl-d-glucose via
a stannyl intermediate and subsequent diethylaminosulfur trifluoride
(DAST) fluorination of the sole free 4′ hydroxyl group to afford
the 2,3,6-triacetyl-4-F-methyl galactoside. Acetolysis to the anomeric
acetate, followed by its selective removal with hydrazine acetate
and trichloroimidate formation afforded 10(34) in 66% overall yield. Glycosylation with acceptor A gave β1-3 TFag-disaccharide derivative 11 (67%). Transformation of 11 into a glycoamino
acid building block involved tert-butyldimethylsilyl removal and trichloroimidate
formation to 12, glycosylation to 13 with
the protected threonine derivative B (Figure ), and concomitant azide reduction/carboxylate
deprotection, which afforded α-glycoamino acid reagent 1 in 32% overall yield.
Scheme 1
Reagents and conditions:
(i) A, TMS-OTf, dichloromethane (DCM), 4 Å molecular
sieves
(MS), 67% (ii) Et3N·3HF, tetra-n-butylammonium
fluoride (TBAF), tetrahydrofuran (THF), 76%; (iii) CCl3CN, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), DCM, 76%; (iv) B, TMS-OTf, DCM/THF 3:1, 4 Å MS, 74.4%; and (v) Zn, THF/AcOH/Ac2O (6:3:1) 71.4%.
Figure 2
Various acceptors used for the synthesis
of (A) compound 11, (B) compounds 13, 29, and 30, and (C) compound 32.
Various acceptors used for the synthesis
of (A) compound 11, (B) compounds 13, 29, and 30, and (C) compound 32.Reagents and conditions:
(i) A, TMS-OTf, dichloromethane (DCM), 4 Å molecular
sieves
(MS), 67% (ii) Et3N·3HF, tetra-n-butylammonium
fluoride (TBAF), tetrahydrofuran (THF), 76%; (iii) CCl3CN, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), DCM, 76%; (iv) B, TMS-OTf, DCM/THF 3:1, 4 Å MS, 74.4%; and (v) Zn, THF/AcOH/Ac2O (6:3:1) 71.4%.
4″-F- and 4′,4″-Difluoro
Galβ1-3GalNAc
TF GAAs (2 and 3)
N-Acetyl galactosamine derivatives bearing an axial 4′ fluorine
are known in the literature; but to prepare compound 2, the TFag-glycoamino acid with 4′-F GalNAc (as
well as 3 with both axial hydroxyls substituted with
fluorine), our strategy necessitated some modified anomeric chemistry
to obtain an efficient fluorination at the 4′ and 4″
positions (Scheme ). Similar to galactosamine, glucosamine hydrochloride was transformed
into the peracetylated 2-azido derivative. Selective removal of anomeric
acetate, donor formation to the known trichloroimidate 14,[35] and glycosylation with 2-naphthalenemethanol
afforded a 9:1 inseparable α/β mixture of naphthyl glycosides 15 in 76% overall yield (three steps). The naphthyl system
was installed because previous reports suggested that fluorination
by the displacement of a hydroxy derivative at C4′ in a gluco-configuration
would be facilitated by bulky anomeric appendages that are in the
α position.[36] Hence, acetate hydrolysis
followed by 4′,6′ benzylidene formation resulted in
a now separable mixture of alcohols 16a and 16b. Pure α anomer 16a was 3′-benzoylated
to yield 17, and the 4′-position was unlocked
by the reductive opening of the 4′,6′-benzylidene ring,
allowing smooth conversion of compound 18 with DAST in
DCM to 4′-fluorinated galactosamine analogue 19 in 84% yield. This is one of the high-yielding 4′-axial fluorinations
of a 2-azido sugar reported to date.[37−40] Removal of the 3′-benzoate
ester afforded acceptor 20.
Scheme 2
Reagents
and conditions: (i)–(iii)
ref (35); (iv) 2-napthylmethanol,
TMS-OTf, DCM/ether (2:1) 87.3%; (v) 25% NaOMe/MeOH, 82%; (vi) benzaldehyde
dimethylacetal, camphor-10-sulfonic acid (CSA), dimethylformamide
(DMF), 16a, 73.6%, 16b, 8.4%; (vii) benzoyl
chloride/pyridine, 91.6%; (viii) NaBH3CN, 4 Å MS,
THF; then 2 M HCl in ether, 85.7%; (ix) DAST, DCM, 84%; and (x) 25%
NaOMe/MeOH, 94.3%.
Reagents
and conditions: (i)–(iii)
ref (35); (iv) 2-napthylmethanol,
TMS-OTf, DCM/ether (2:1) 87.3%; (v) 25% NaOMe/MeOH, 82%; (vi) benzaldehydedimethylacetal, camphor-10-sulfonic acid (CSA), dimethylformamide
(DMF), 16a, 73.6%, 16b, 8.4%; (vii) benzoyl
chloride/pyridine, 91.6%; (viii) NaBH3CN, 4 Å MS,
THF; then 2 M HCl in ether, 85.7%; (ix) DAST, DCM, 84%; and (x) 25%
NaOMe/MeOH, 94.3%.The synthesis of glycoamino
acids 2 and 3 (Scheme ) converged
at this point using 20 as a common acceptor. Disaccharide
formation between 20 and a galactosedonor however proved
problematic. We attributed this to the possible “disarming”
of the C3′ hydroxyl group reactivity through an inductive or
conformational effect imparted by the axial fluorine atom. Several
methods were attempted to enhance the reactivity of the acceptor hydroxyl
group, but none proved expedient or high-yielding. Optimum yields
were obtained by the “classical” method of reacting
peracetylated galactosyl bromide 21 or fluorinated derivative 22(33) with the acceptor under the
auspices of silver triflate catalysis, with yields of pure β-disaccharides 23 or 24 of 60 and 55%, respectively. We were
pleased at this moderate but an acceptable yield and the stereoselectivity
in these novel reactants; although this seemingly simple 1,3 glycosylation
using an assisting acetate at O2′ of the galacto-donor should
result in near-exclusive beta products, several have reported mixtures
and hydrolysis products that hamper the efficiency of this coupling.[27,41,42] Another bright spot in this method
was that approximately 15% of the precious fluorinated acceptor could
be recovered during purification. Elaboration to the appropriate GAAs 2 and 3 involved the following straightforward
sequence. Reductive removal of both naphthyl and benzyl groups followed
by acetylation afforded compounds 25 and 26. Formation of anomeric bromides 27 and 28 and glycosylation with amino acid B and silver perchlorate
afforded α-linkedGAAs 29 and 30.
Simultaneous azide reduction, N-acetylation, and phenacyl ester removal,
as above, afforded GAAs 2 and 3 in a high
yield. The only downside to this highly efficient sequence is the
amino acid glycosylation step: although this proceeds in a nearly
quantitative yield for both fluorinated derivatives, the anomeric
ratio only slightly favored the α product (reagents and conditions, Scheme ), even if the anomers
were easily separable. The stereoselectivity of the glycosylation
undoubtedly suffered because of the lack of any steric or electronic
bias in the 4-fluorodonor. The timing of this step, however, was
difficult to avoid because it was necessary to fluorinate the sugar
building blocks before the introduction of the amino acid portion
of the molecule. Hence, any structural features that may bias the
glycosylation toward specific anomers are removed before this step.
Glycoamino acid formation at the monomer stage of the 4′-fluorogalactosamine derivatives did not enhance the anomeric ratio in a
favorable way.
Scheme 3
Reagents and conditions: (i)
AgOTf, DCM, 4 Å MS, 55% (23) 60% (24); (ii) 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ), DCM/H2O (20:1), then Ac2O, pyridine, 53% (25) 47%
(26); (iii) TMSBr, BiBr3, DCM, 98% (27) 87% (28); (iv) B, AgClO4, DCM, 4 Å MS; 97% (29, α/β;
56:41) 95% (30, α/β; 51:44); and (v) Zn,
THF/AcOH/AC2O (6:3:1) 84% (2) 93% (3).
Reagents and conditions: (i)
AgOTf, DCM, 4 Å MS, 55% (23) 60% (24); (ii) 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ), DCM/H2O (20:1), then Ac2O, pyridine, 53% (25) 47%
(26); (iii) TMSBr, BiBr3, DCM, 98% (27) 87% (28); (iv) B, AgClO4, DCM, 4 Å MS; 97% (29, α/β;
56:41) 95% (30, α/β; 51:44); and (v) Zn,
THF/AcOH/AC2O (6:3:1) 84% (2) 93% (3).
4″-Deoxy-Galβ1-3GalNAc TF GAA
(4)
We also prepared the TF antigen GAA (4) deoxygenated
at the galactose 4′ position. Commercially available methyl(2,3,6-tri-O-benzoyl)galactopyranoside (Scheme ) was processed through the 4′-thiocarbonylimidazole
derivative with tri-N-butyl tin hydride and azobisisobutyronitrile
in refluxing toluene to afford the known methyl-(2,3,6-tri-O-benzoyl)-4′-deoxyglucoside.[34] The benzoyl groups were replaced with acetates, and the
anomeric center was processed similar to the preparation of 10 above to afford 4′-deoxy gluco-donor 31(34) in nearly 50% yield over seven steps.
Glycosylation with the known Fmoc/t-butyl protected
acceptor C(43) (Figure ) proceeded to give β-disaccharide 32 in 70% yield. Simple, high-yield processing of 32 through 33 (benzylidene removal and acetylation) and 34 (azide reduction/acetylation) afforded, after the t-butyl group hydrolysis, final GAA 4 (65%
over four steps).
Scheme 4
Reagents and conditions: (i) C, TMS-OTf, 1,2-dichloroethane, 4 Å MS, 70%; (ii) iodine,
MeOH/H2O, reflux; (iii) acetic anhydride, pyridine, 73.4%,
two steps; (iv) chloroform/pyridine/thioacetic acid (1:1:1), 88.7%;
and (v) 95% aq trifluoroacetic acid (TFA), 94.9%.
Reagents and conditions: (i) C, TMS-OTf, 1,2-dichloroethane, 4 Å MS, 70%; (ii) iodine,
MeOH/H2O, reflux; (iii) acetic anhydride, pyridine, 73.4%,
two steps; (iv) chloroform/pyridine/thioacetic acid (1:1:1), 88.7%;
and (v) 95% aq trifluoroacetic acid (TFA), 94.9%.
Synthesis and Characterization of APF Peptides
With
the GAAs in hand, four new APF peptide analogues were synthesized
by standard Fmoc-based SPPS methods. Procedures for amino acid coupling
were similar to those in our previous reports,[26,44] with minor variations. The peptides were synthesized under microwave
radiation on a CEM Liberty peptide synthesizer using 2-chlorotrityl
resin. Amino acids were double-coupled to the resin to ensure the
complete reaction of each amino acid and to avoid the production of
deletion sequences. Coupling of the glycoamino acid was performed
manually under the conditions we had developed previously.[45] The final peptides were deprotected on resin
with hydrazine hydrate, cleaved from the resin, and high-performance
liquid chromatography (HPLC) purified. All glycopeptides were >95%
pure by the HPLC analysis (see the Supporting Information). All purified glycopeptides were characterized
by high-resolution mass spectrometry and multiple nuclear magnetic
resonance (NMR) experiments. Both 1H and 13C
spectra are shown for compounds 5–8 in the Supporting Information along with the assignment
tables for fluorinated derivatives 5–7. The 19F spectra for the fluorinated final APF derivatives are shown
in Figure . Inspection
shows that the coupling constants for each monofluoro derivative are
very similar, indicating that substitution does not affect the ring
pucker to any substantial degree. We have performed comprehensive
modeling and NMR studies on fluorinated as-APF analogues 5–7, and the results from these studies will be reported in a future
manuscript.
Figure 3
Proton-coupled fluorine 19 spectra of compounds 5, 6, and 7 (left to right). Chemical shifts are
referenced to external TFA at −76.55 ppm.
Proton-coupled fluorine 19 spectra of compounds 5, 6, and 7 (left to right). Chemical shifts are
referenced to external TFA at −76.55 ppm.
Antiproliferative Activity of APF Analogues
The four
glycopeptide analogues were tested for antiproliferative activity
against NB epithelial cells and the T24 bladder carcinoma cell line
(Figures and 5), as performed in the previous studies of APF analogues.[21,22] Determination of cell proliferation was performed with the WST-1
assay.[46,47] This is similar to other chromogenic assays
but produces a formazan product that is water-soluble (MTT product
is insoluble) and more stable than other (XTT) products (see e.g.,
refs (48) and (49)). Replacement of the 4-OH
group of galactosamine with fluorine (compound 6) resulted
in a compound whose potency was very similar to that of the parent as-APF, whereas remaining analogues 5, 7, and 8 were more than 3-fold less potent than 6 (Figure B). We also tested our new analogues against T24 bladder cancer cells
as our previous studies have shown that these tumor cells are nearly
as sensitive to the activity of APF as NB cells.[21,24] The results were similar to the data for NB cells: compound 6 was the most potent with 5, 7,
and 8 being ∼3 to 5-fold less potent (Figure ).
Figure 4
Antiproliferative WST-1
assay on derivatives 5–8 relative to as-APF against NB epithelial cells.
The error bars are standard deviations based on a minimum of three
separate experiments.
Figure 5
Antiproliferative WST-1 assay on derivatives 5–8 relative to as-APF against T24 bladder carcinoma
cells. The error bars are standard deviations based on two separate
experiments.
Antiproliferative WST-1
assay on derivatives 5–8 relative to as-APF against NB epithelial cells.
The error bars are standard deviations based on a minimum of three
separate experiments.Antiproliferative WST-1 assay on derivatives 5–8 relative to as-APF against T24 bladder carcinoma
cells. The error bars are standard deviations based on two separate
experiments.These results shed additional
light on the requirements for the
antiproliferative activity of various functional aspects of the sugar
portion of the molecule. We had previously shown that both disaccharide
and α linkage to the peptide were absolutely necessary for the
maintenance of cellular activity.[44] In
addition, one of our original analogues that retained full activity
contained a lactosamine unit (Galβ1-4GlcNAc, unpublished results)
instead of TFag, strongly suggesting that the reducing
sugar can tolerate changes and still remain active, whereas the galactose
moiety at the nonreducing end remained important for activity. The
results presented here also showed that modification of the 4″-position
of the galactose ring is detrimental to activity. Coupled with our
previous results, indicating that the removal of the galactose moiety
rendered the as-APF inactive,[44] our current data therefore confirm the importance of a
galacto-type stereochemistry/functionality in the nonreducing sugar
for optimal as-APF activity. Replacement of either
hydrogens or hydroxyl groups with fluorine is a common substitution
in drug discovery. The fluorine atom is highly compact and electronegative,
making it a suitable isostere for either a small H atom or a hydrogen
bonding/accepting OH group. Fluorine atoms can only accept hydrogen
bonds, and this property can be used as an advantage in various medicinal
chemistry campaigns. The strong electronegativity of fluorine can
also have dramatic effects on the conformation and/or binding properties
of small molecules.
Conclusions
We prepared several
threonine glycoamino acid analogues of the
TFagdisaccharide and used them to prepare analogues of
a unique antiproliferative agent, APF. The synthesis of the GAAs entailed
exploring a variety of glycosylation conditions that were amenable
to the modified reactivity of the intermediates along the synthetic
routes. All molecules were efficiently prepared, and all GAAs were
incorporated into the APF peptide sequence by microwave-assisted SPSS.
The additional structure–activity relationships that were gleaned
from this study will help us design novel glycomimetics of APF for
the future design of inhibitors for IC/PBS as well as therapeutic
leads against various malignancies.
Materials and Methods
General
All solvents were purchased from either VWR
(Radnor, PA) or Thermo Fisher Scientific (Waltham, MA) and dried using
a glass contour room temperature drying system; the reagents/chemicals
were purchased from Sigma-Aldrich (St Louis, MO), and amino acids
and peptide synthesis reagents were purchased from Bachem (Torrance,
CA). Thin-layer chromatography analyses were performed on Analtech
(Newark, DE) glass-backed plates for ultraviolet detection at 245
nm. Purification of intermediates was performed either with manually
packed silica gel columns or on an ISCO CombiFlash+ system. Final
glycoaminoacid and glycopeptide products were purified on a Waters
2545 Prep HPLC system interfaced to a 2767 sample manager. Liquid
chromatography/mass spectrometric analysis was performed with either
a Shimadzu 2020 LCMS system or an Agilent 1100 LCMS system.Routine 1H and 13C NMR data were collected
on a Bruker NanoBay 400 MHz spectrometer, and 19F data
were collected at 376 MHz on the same instrument with a multinuclear
SMART Probe at 25 °C and processed with MestreNova software (Santiago
de Compostela, Spain). The spectra were referenced to the residual
protonated solvent peak. Final glycopeptides were analyzed at 500
MHz on a Bruker AVANCE III spectrometer with a triple resonance cryoprobe
(TCI). Experiments for resonance assignments included 1-D proton and
carbon, along with 2-dimensional COSY, TOCSY, HSQC, and HMBC data.
Data were collected in both 9:1 H2O/D2O and
100% D2O for assignment purposes.
To a solution of 11 (0.30 g; 0.42
mmol) in anhydrous THF (15 mL) was added Et3N·3HF
(1.5 mL; 9.10 mmol). The reaction was stirred at room temperature
for 18 h and then treated with a solution of TBAF (1 M, 1.0 mL, 1.0
mmol). After 2 h of stirring, the mixture was diluted with ethyl acetate
(60 mL), washed with water (1 × 25 mL), and dried. After filtration
and concentration, the residual intermediate was used directly in
the next step. This intermediate was dissolved in dry DCM (15 mL)
and cooled in an ice bath. Trichloroacetonitrile (0.46 mL; 4.58 mmol)
and DBU (0.05 mL; 0.344 mmol) were added. The ice bath was removed,
and the mixture was stirred for 20 h. The solvent was removed in vacuo
and purified by flash chromatography using ethyl acetate/hexanes (1:4
to 2:1) to give 179 mg of product 12 (75.8%). H NMR (400 MHz, chloroform-d): δ 8.68 (s, 1H, Cl3CC(N)), 7.27–7.48 (m, 5H, ), 6.49 (d, J = 2.7 Hz, 1H, ), 5.50 (s, 1H, PhC), 5.31
(dd, J = 8.1, 10.3 Hz, 1H, ), 4.92 (ddd, J = 2.7, 10.3, 27.7 Hz, 1H, ), 4.79 (dd, J = 50.3, 2.7 Hz, 1H, ),
4.41 (dist d, 1H, ), 4.33 (ddd, J = 11.2, 6.2, 0.8 Hz, 1H, ), 4.23 (dd, J = 12.5, 1.5 Hz, 1H, ), (ddd, J = 11.2, 6.7
Hz, 1H, ), 4.14–4.06
(m, apparent AB of ABX, 2H, ), 3.97 (dd, J = 12.8, 2.7 Hz, 1H, ), 3.82 (br s, 1H, ), 3.80 (br dt, J = 26.0, 6.7 Hz, ), 2.043 (s, 3H, ), 2.015 (s, 3H, ), 1.989
(s, 3H, ). C NMR (101 MHz, chloroform-d): δ
169.4, 169.3, 168.3, 136.4, 128.0, 127.2, 125.1, 100.8, 99.7, 94.8,
84.6 (JC4,F = 187 Hz), 74.4 (JC3,F = 20 Hz), 70.5, 70.4, 70.3, 70.1, 67.7 (JC5,F = 15.4 Hz), 64.4, 60.44 (JC6,F = 5.3 Hz), 19.8, 19.7 (2× C). HRMS: compound is
unstable in the mass spectrometer and used as is.
To
a solution of 2-azido-2-deoxy-3,4,6-tri-O-acetyl-d-glucopyranosyl trichloroacetimidate (14;[35] 9.13 g; 19.19 mmol) in DCM/diethylether (2:1,
150 mL) were added 2-naphthylmethanol (2.50 g; 15.80 mmol) and activated
4 Å MS (5 g). The mixture was cooled to −40 °C, and
TMSOTf (0.67 mL; 3.79 mmol) was added dropwise and stirred for 30
min at the same temperature. The reaction was allowed to warm to room
temperature for over 30 min and then cooled to 0 °C before being
quenched with trimethylamine (0.70 mL). The mixture was filtered through
a celite pad and concentrated under vacuum. The crude product was
purified by flash chromatography (toluene/acetone, 10:1) to give 15 as a mixture of isomers (7.90 g; 87.30%) (note:
the α-isomer is assigned; the β-isomer H1 is assigned). H NMR (400 MHz, chloroform-d): δ 7.91–7.80
(m, 4H, Ar(C)), 7.55–7.47 (m,
3H, Ar(C)), 5.55 (dd, J = 10.6, 9.2 Hz, 1H, ), 5.10 (d, J = 3.8 Hz, 1H, ), 5.07 (d, J = 10.6 Hz, 1H, ), 4.92 (d, J = 12.0 Hz, 1H, Nap-CH), 4.80 (d, J = 12.0 Hz, 1H, Nap-CH), 4.49 (d, J = 8.1 Hz,
0.1H, ),
4.27 (dd, J = 12.2, 4.2 Hz, 1H ), 4.08 (ddd, J =
10.1, 4.2, 2.4 Hz, 1H, ), 4.03 (dd, J = 12.2, 2.3 Hz, 1H, ), 3.35 (dd, J = 10.6, 3.6 Hz, 1H, ), 2.09 (s, 3H, OAc(CH3)), 2.09 (s, 3H,
OAc(C)), 2.03 (s, 3H, OAc(C)). C NMR (101 MHz, CDCl3): δ 170.71, 170.12,
169.81, 133.49, 133.32, 133.26, 128.69, 128.11, 127.88, 127.53, 126.51,
126.46, 126.03, 100.47, 96.77, 70.61, 70.36, 68.68, 67.98, 61.94,
61.06, 20.86. HRMS: C23H25N3O8 ESI M + NH4 calcd: 489.1980; observed,
489.1967.
(2-Napthylmethyl) 2-Azido-2-deoxy-4,6-O-benzylidine-α-d-glucopyranoside (16a) and 1-(2-Napthylmethyl)
2-Azido-2-deoxy-4,6-O-benzylidine-β-d-glucopyranoside (16b)
To a solution of (2-naphthylmethyl)
2-azido-2-deoxy-3,4,6-tri-O-acetyl-d-glucopyranoside
(15; 7.90 g; 16.77 mmol) in methanol (70 mL) was added
a solution of 25% sodium methoxide in methanol (4 mL). The reaction
mixture was stirred at room temperature for 3 h and then cooled in
an ice bath before being neutralized with 2 M HCl/diethylether solution.
The reaction mixture was concentrated under vacuum and purified by
flash chromatography (DCM/MeOH 15:1) to give (2-naphthylmethyl)-2-azido-2-deoxy-glucopyranoside
(4.75 g; 82.0%). This material (13.75 mmol) was dissolved in dry DMF
(60 mL) and cooled in an ice bath to 0 °C. To this solution were
added benzaldehydedimethylacetal (6.5 mL; 43.10 mmol) and (+) CSA
(0.65 g; 2.79 mmol). The mixture was stirred at room temperature for
4 h. The mixture was diluted with EtOAc (150 mL) and washed with saturated
aqueous ammonium chloride (2 × 70 mL). The organic layer was
dried, concentrated in vacuo, and purified by flash chromatography
(EtOAc/hexanes 1:4) to give desired α-isomer 16a (4.39 g; 73.6%) and β-isomer 16b (0.50 g; 8.4%). 16a—H NMR (400
MHz, chloroform-d): δ 7.90–7.81 (m,
4H, Ar-C), 7.55–7.45 (m, 5H,
Ar(C)), 7.43–7.32 (m, 3H, Ar(C)), 5.55 (s, 1H, PhC), 5.05 (d, J = 3.7 Hz, 1H, α), 4.95
(d, J = 12.2 Hz, 1H, Nap-CH), 4.78 (d, J = 12.1 Hz, 1H, Nap-CH), 4.33 (t, J = 10.1,
9.6 Hz, 1H, ), 4.28 (dd, J = 10.2, 5.2 Hz, 1H, ), 3.98 (td, J = 9.9, 4.9 Hz, 1H, ),
3.76 (t, J = 10.3 Hz, 1H, ), 3.56 (t, J = 9.3 Hz, 1H, ), 3.32 (dd, J = 10.0, 3.7 Hz,
1H, ). C NMR (101 MHz, CDCl3): δ 136.98, 133.94,
133.33, 133.29, 129.55, 128.62, 128.55, 128.10, 127.90, 127.22, 126.45,
126.41, 126.35, 125.90, 102.29, 97.53, 82.02, 70.14, 69.05, 68.98,
63.28, 62.80. HRMS: C24H23N3O5 ESI M + H calcd: 434.1710; observed, 434.1702.16b—H NMR (400 MHz, chloroform-d): δ 7.93–7.81
(m, 4H, Ar-C), 7.59–7.43 (m,
5H, Ar-C), 7.43–7.32 (m, 3H,
Ar-C), 5.56 (s, 1H, PhC), 5.11 (d, J = 11.8 Hz, 1H, Nap-CH), 4.87 (d, J = 11.8 Hz,
1H, Nap-CH), 4.56 (d, J = 7.9 Hz, 1H, ), 4.39 (dd, J = 10.5, 5.0 Hz, 1H, ), 3.83 (t, J = 10.3
Hz, 1H, ), 3.66 (t, J = 9.2 Hz, 1H, ), 3.59 (t, J = 9.1 Hz, 1H, ), 3.51
(t, J = 9.3, 8.0 Hz, 1H, ), 3.43 (td, J = 9.6, 5.0 Hz, 1H, ). C NMR (101 MHz, CDCl3): δ 136.91, 133.89, 133.35, 133.32,
129.54, 128.59, 128.54, 128.13, 127.89, 127.27, 126.42, 126.39, 126.36,
125.94, 102.18, 101.19, 80.77, 72.21, 71.70, 68.70, 66.69, 66.39. HRMS: C24H23N3O5 ESI M + H calcd: 434.1710; observed, 434.1702.
A solution of 34 (0.235 g;
0.24 mmol) and 95:5 TFA/H2O (5 mL) was stirred at room
temperature for 60 min. The mixture was concentrated under vacuum
at 15 °C and then coevaporated with toluene (3 × 5 mL) at
20 °C. The crude product was purified by silica gel chromatography
using a gradient of ethyl acetate/acetic acid (50:1 to 50:3) to give
compound 4 (0.208 g; 94.9%). H NMR (400 MHz, methanol-d4): δ 7.82 (d, J = 7.6 Hz, 2H, Ar(C)), 7.69 (dd, J = 7.6, 3.2 Hz,
2H, Ar(C)), 7.41 (t, J = 7.4 Hz, 2H, Ar(C)), 7.33 (td, J = 7.3, 2.4 Hz, 2H, Ar(C)), 5.38 (d, J = 3.3 Hz, 1H, ), 4.95 (ddd, J = 11.4, 9.5, 5.4 Hz, 1H, ), 4.88 (d, J =
4.1 Hz, 1H, ), 4.66 (dd, J = 9.6, 7.8 Hz, 1H, ), 4.61–4.46
(m, 3H, , Fmoc(C)), 4.36
(q, J = 6.7 Hz, 1H, Thr(C), 4.32–4.24 (m, 2H, , Fmoc(C)), 4.24–4.16
(m, 3H, , Thr(C), ), 4.13
(dd, J = 11.4, 5.0 Hz, 2H, ), 3.97 (dd, J = 11.4, 7.6 Hz,
1H, ), 3.89 (dd, J = 11.1, 3.3 Hz, 1H, ), 3.77 (dd, J = 11.0, 5.2 Hz, 1H, ), 2.14–2.05 (m, 7H, (C), 2× OAc(C)), 2.04 (s, 3H,
OAc(C)), 2.00 (s, 3H, OAc(C)), 1.99 (s, 3H, OAc(C)), 1.98
(s, 3H, NHAc(C)), 1.59 (q, J = 12.0 Hz, 1H, (C)), 1.23 (d, J = 6.4 Hz, 2H, Thr(C)). C NMR (101 MHz, methanol-d4): δ 172.49, 172.32, 171.96, 171.79,
171.34, 159.05, 145.37, 145.17, 142.86–142.48 (m), 128.84,
128.21, 126.13, 126.01, 121.01 (d, J = 3.0 Hz), 102.42,
100.76, 74.46, 73.70, 71.83, 71.41, 70.72, 68.81, 67.62, 66.14, 64.09,
33.34, 23.29, 20.74, 19.20. HRMS: C43H52N2O19 ESI M + H calcd: 901.3237; observed,
901.3193.
Loading of the C-Terminal Amino Acid on Chlorotrityl
Chloride
Resin
2-Chlorotrityl chloride resin (1 g; loading capacity
1.5 mmol/g) was placed in a 10 mL polypropylene fritted column, suspended
in dry DCM for 15 min, and further washed with dry DCM (2 × 7
mL) and then with dry DMF (3 × 7 mL). Fmoc-Ala-OH (0.467 g, 1.5
mmol) was dissolved in dry DMF (5 mL) and added to the resin, followed
by diisopropylethylamine (DIPEA, 520 μL, 3.0 mmol). The resin
was shaken with this mixture at room temperature for 45 min, after
which the reagents were drained and the resin was washed with dry
DMF (3 × 7 mL) followed by dry DCM (5 × 7 mL). The coupling
and washing steps were repeated, and the resin was dried under vacuum
overnight. The alanine loading was calculated, using a method reported
previously by Gude et al., to be 0.4 mmol/g.[50]
Synthesis of Octapeptide (H-VPAAVVVA-CTC-Resin)
SPPS was performed on a CEM Liberty microwave peptide
synthesizer with a CEM Discover microwave generator. Fmoc-Ala-CTC
resin (250 mg, 0.1 mmol) was swelled at room temperature for 30 min
and then loaded into the peptide synthesizer, and the following seven
amino acids in the sequence (10 equiv each) were double-coupled under
microwave catalysis using 0.5 M (1-[Bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate)
(HATU) and 2 M DIPEA (20 W; 50 °C
for 10 min, 2×). Fmoc-deprotection was affected by microwaves
in the presence of 20% piperidine in N-methyl-2-pyrrolidone
with 0.1% hydroxybenzotriazole (35 W; 50 °C for 2 min; 2×).
The resin-attached octapeptide (Fmoc-VPAAVVVA) was retrieved from
the synthesizer, washed with dry DCM (3 × 5 mL), and dried overnight.
The Fmoc-number following the synthesis of the octapeptide was determined
to be 0.3 mmol/g. The terminal Fmoc-group was removed manually by
shaking the resin with 20% piperidine in DMF (7 mL, 30 min; 2×).
The deprotected resin was washed with dry DMF (3 × 7 mL) followed
by dry DCM (3 × 7 mL).
General Method for Coupling GAAs 1-4 to H-VPAAVVVA-CTC-Resin
Fmoc-GlycoAA-OH (0.9 equiv) and HATU
(0.9 equiv) were dissolved
in DMF (2 mL) and added to the resin (0.5 μmol) at room temperature.
Finally, 2,4,6-trimethyl pyridine (1.5 equiv) was added to the mixture,
and the resin was shaken at room temperature for 3 h. The reagents
were drained, and the resin was washed with DMF (3 × 5 mL). The
Fmoc-group was removed using 20% piperidine in DMF (2 × 2 mL).
The resin was washed with DMF (3 × 5 mL) and then treated with
10% hydrazine hydrate in DMF for 4 h. The resin was filtered and washed
again with DMF (3 × 5 mL) followed by DCM (5 × 5 mL) and
then dried. The glycopeptide was cleaved off the resin by treatment
with TFA/H2O (95:5; 2 mL) for 2 h. The TFA solution was
collected, and the resin was washed with TFA (2 × 1 mL). The
TFA solutions were combined, and the acid was removed under vacuum
at 20 °C. The residue was redissolved in Milli-Q water, lyophilized
to a white foam, and purified by preparative HPLC (see the Supporting Information).
WST-1 Cell Proliferation
Assay
Primary NB epithelial
cells were grown from bladder tissue biopsies, as previously described.[21] For this study, the cells were obtained from
either the bladder biopsies of patients who were at least 18 years
old or from the cadaveric tissue, in accordance with the guidelines
of the Institutional Review Board of the University of Maryland School
of Medicine. None of the cell donors had any history of a functional
bladder disorder. T24 (HTB-4) cells were obtained from ATCC and cultured,
as previously described.[24]Bladder
epithelial and carcinoma cell proliferations were measured using a
WST-1 assay (Roche) according to the manufacturer’s instructions.
For this assay, NB cells were plated onto 96-well tissue culture plates
(VWR) at a density of 1.5 × 104 cells/well, cultured
in minimum essential medium (MEM) containing 10% heat-inactivated
fetal bovineserum (FBS), 1% antibiotic/antimycotic solution, and
1% l-glutamine (all from Sigma) at 37 °C in a 5% CO2 atmosphere overnight; T24 cells were plated at a density
of 1.5 × 103 cells/well, cultured in McCoy’s
5A medium (Invitrogen) containing 10% heat inactivated FBS, 1% antibiotic/antimycotic
solution, 1% l-glutamine, and 2.2 g/L sodium bicarbonate
(all from Sigma). The next day, the medium was changed to serum-free
MEM medium (NB cells) or serum-free McCoy’s medium (T24 cells),
and the cells were cultured for an additional 24 h. On the third day,
APF derivatives were diluted in acetonitrile/H2O (double-distilled)
(1:1) and added to the serum-free cell medium; cell controls received
acetonitrile/H2O (double distilled) (1:1) alone. The cells
were then incubated at 37 °C in a 5% CO2 atmosphere
for an additional 48 h, after which the WST-1 reagent was applied
to the cells and incorporation was measured at 450 nm with the reference
wavelength at 690 nm, using a Molecular Devices microplate reader.
Authors: Kristopher R Koch; Chen-Ou Zhang; Piotr Kaczmarek; Joseph Barchi; Li Guo; Hanief M Shahjee; Susan Keay Journal: Invest New Drugs Date: 2011-09-20 Impact factor: 3.850
Authors: Shantibhusan Senapati; Pallavi Chaturvedi; William G Chaney; Subhankar Chakraborty; Vinayaga S Gnanapragassam; Aaron R Sasson; Surinder K Batra Journal: Clin Cancer Res Date: 2010-11-08 Impact factor: 12.531
Authors: Piotr Kaczmarek; Susan K Keay; Gillian M Tocci; Kristopher R Koch; Chen-Ou Zhang; Joseph J Barchi; David Grkovic; Li Guo; Christopher J Michejda Journal: J Med Chem Date: 2008-09-13 Impact factor: 7.446
Authors: Susan K Keay; Zoltan Szekely; Thomas P Conrads; Timothy D Veenstra; Joseph J Barchi; Chen-Ou Zhang; Kristopher R Koch; Christopher J Michejda Journal: Proc Natl Acad Sci U S A Date: 2004-07-28 Impact factor: 11.205
Authors: Ernest G Nolen; Ezra S Hornik; Kendra B Jeans; Weiyu Zhong; Danielle M LaPaglia Journal: Tetrahedron Lett Date: 2021-04-23 Impact factor: 2.032