Assessment of chemoselective neoglycosylation methods using chlorambucil as a model

Article abstract of DOI:10.1021/jm101024j

To systematically assess the impact of glycosylation and the corresponding chemoselective linker upon the anticancer activity/selectivity of the drug chlorambucil, herein we report the synthesis and anticancer activities of a 63-member library of chloramb

Full text of DOI:10.1021/jm101024j

J. Med. Chem. 2010, 53, 8129–8139 8129
DOI: 10.1021/jm101024j
Assessment of Chemoselective Neoglycosylation Methods Using Chlorambucil as a Model
Randal D. Goff and Jon S. Thorson*
Wisconsin Center for Natural Products Research, University of Wisconsin National Cooperative Drug Discovery Group,
Pharmaceutical Sciences Division, School of Pharmacy, University of Wisconsin;Madison, 777 Highland Avenue,
Madison, Wisconsin 53705, United States
Received August 6, 2010
To systematically assess the impact of glycosylation and the corresponding chemoselective linker upon
the anticancer activity/selectivity of the drug chlorambucil, herein we report the synthesis and anticancer
activities of a 63-member library of chlorambucil-based neoglycosides. A comparison of N-alkoxyamine-,
N-acylhydrazine-, and N-hydroxyamine-based chemoselective glycosylation of chlorambucil revealed
sugar- and linker-dependent partitioning among open- and closed-ring neoglycosides and correspond-
ing sugar-dependent variant biological activity. Cumulatively, this study represents the first neoglyco-
randomization of a synthetic drug and expands our understanding of the impact of sugar structure upon
product distribution/equilibria in the context of N-alkoxyamino-, N-hydroxyamino-, and N-acylhydrazine-
based chemoselective glycosylation. This study also revealed several analogues with increased in vitro
anticancer activity, most notably D-threoside 60 (NSC 748747), which displayed much broader tumor
specificity and notably increased potency over the parent drug.
Introduction
To assess the impact of alternative chemoselective glycosyl-
ation methods in the context of neoglycorandomization, we
selected the synthetic anticancer agent chlorambucil (Figure 1, 1)
as a model. First synthesized over 5 decades ago,25 chlo-
rambucil remains a current treatment for chronic lymphocytic
leukemia (CLLa)26-30 and has served as the basis for newer
generation analogues such as the recently approved bend-
amustine (3).31 A nitrogen mustard, 1 leads to guanine alkyla-
tion and DNA cross-linking and ultimately prohibits DNA
replication and transcription.32,33 The primary cellular uptake
mechanism of 1 is passive diffusion,34 and like many cyto-
toxics, the lack of nitrogen mustard tumor-specificity con-
tributes to serious side effects.35,36 Thus, improvements have
focused upon (i) the development of tumor-activated pro-
drugs, exemplified by the hypoxia-activated N-oxide PX-478
(Figure 1, 2) currently in phase I37 or (ii) modifications to
engage tumor-specific transport, exemplified by the β-
D-glucosyl analogue of ifosfamide (glufosfamide, Figure 1, 4)
which is actively transported into tumor cells by the sodium/
D-glucose cotransporter SGLT3 (SAAT1).38 While the specific
glycosylation of 1 has presented analogues that display slight
improvements in a perceived therapeutic index (slightly
improved in vitro potency and subtle reductions of in vivo
peripheral toxicity), the analogues synthesized to date have been
restricted to the use of D-gluco- or D-galacto-based sugars.39-46
To systematically assess the impact of glycosylation and the
corresponding chemoselective linker upon the anticancer
activity/selectivity of 1, herein we report the synthesis and
anticancer activities of a 54-member library of 1 neoglyco-
sides. Several analogues with increased in vitro anticancer
activity were identified, most notably D-threoside 60 (NSC
748747) which displayed much broader tumor specificity and
increases in potency of up to 15-fold compared to 1, repre-
senting the most active chlorambucil glycoside reported to
date. Representative sugars identified as hits in the context of
The sugars attached to pharmaceutically important natural
products dictate the pharmacokinetics and/or pharmaco-
dynamics of the selected agent.1-3 Yet studies designed to
systematically understand and/or exploit the attachment of
carbohydrates in drug discovery remain limited by the avail-
ability of practical synthetic and/or biosynthetic tools.4-8
Neoglycosylation takes advantage of a chemoselective reac-
tion between free reducing sugars and N-methoxyamino-
substituted acceptors.9-19 This reaction has enabled the
process of “neoglycorandomization” wherein alkoxyamine-
appended natural product-based drugs are differentially glyco-
sylated with a wide array of natural and unnatural reducing
sugars.4-8,20-24 Neoglycorandomization has led to the dis-
covery of cardenolide neoglycosides with enhanced in vitro
and in vivo anticancer activity and lower in vivo toxicity,20,23
colchicine neoglycosides with a novel anticancer mechanism
and lower in vivo toxicity,21 vancomycin neoglycosides that
displayed improved in vitro potency against clinical isolates of
vancomycin-resistant Enterococci,22 and betulinic acid neo-
glycosides improved for either in vitro anticancer or antiviral
potency.24 While these examples clearly highlightthe potential
impact of differential glycosylation indrug lead discovery, this
work has not addressed the impact of the chemoselective
glycosylation “handles”17 employed.
*To whom correspondence should be addressed. Phone: (608) 262-
3829. Fax: (608) 262-5345. E-mail: jsthorson@pharmacy.wisc.edu.
a Abbreviations: BH3 Et3N, borane-triethylamine complex; CLL,
chronic lymphocytic leukemia; CNS, central nervous system; DIC,
N,N0-diisopropylcarbodiimide; DMAP, 4-(N0,N0-dimethylamino)pyridine;
EDAC, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide; GI50, growth
inhibitory concentration for 50% of the cell population under study;
GLUT1, glucose transporter 1; LAH, lithium aluminum hydride;
NMM, N-methylmorpholine; SAAT1, sodium/amino acid transporter 1;
SGLT3, sodium/glucose transporter 3; SPE, solid phase extraction;
THF, tetrahydrofuran.
3
r
2010 American Chemical Society
Published on Web 10/25/2010
pubs.acs.org/jmc

8130 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Goff and Thorson
N-methoxyamino-substituted 1 were subsequently conju-
gated via N-acylhydrazine- and N-hydroxyamine-based
strategies and the products characterized and evaluated for
anticancer activity. Analysis of these second generation neo-
glycoside analogues revealed sugar-dependent partitioning
among open- and closed-ring neoglycosides and correspond-
ing sugar-dependentvariant biologicalactivity. Cumulatively,
this study represents the first neoglycorandomization of a
synthetic drug and expands our understanding of the impact
of sugar structure upon product distribution/equilibria in
the context of N-alkoxyamino-, N-hydroxyamino-, and N-
acylhydrazine-based chemoselective glycosylation.
amide 6 using the EDAC coupling agent in excellent yield
(97%) and then selectively reduced to the corresponding
aldehyde with LAH. Compound 7 was subsequently con-
densed with methoxyamine HCl in the presence of organic
3
base, providing a mixture of E- and Z-oximes (8). Reduction
of the carbon-nitrogen double bond was accomplished with
BH3 Et3N in a manner similar to that for previously de-
3
scribed neoaglycons.20,21,23 This convenient strategy yielded
neoaglycon 9 in four steps with an overall yield of 37% and
required only a single chromatographic purification. While
standard neoglycosylation conditions (3/1 DMF/HOAc)9-22
provided a slightly greater yield (68%) of neo-D-riboside 53
in a pilot reaction, the use of MeOH with a small molar
excess of HOAc (i.e., 1.5 equiv) as an acidic proton source
was sufficient for the reaction to proceed in an equivalent
amount of time (54% yield, Table S1 in Supporting Infor-
mation) and also simplified subsequent solvent evacuation.
By use of these optimized conditions (90 μM 9, 2 equiv of
sugar, 40 °C, 1.5 equiv of HOAc in MeOH), a 54-member
library of neoglycosides (10-63; see Figure S1 in Supporting
Information) was synthesized with an average isolated yield
of 63% wherein high-throughput solid phase extraction
provided an average purity of 92.9% (see Table S3 in
Supporting Information). Product yields paralleled the
overall reactivity trend: tetroses (79 ( 10%) > deoxy sugars
(69 ( 6%) > pentoses (62 ( 7%) > hexoses (56 ( 5%),
likely reflecting the general reducing sugar solubility in the
selected solvent system. Unlike prior neoglycorandomized
libraries,9-24 a preference for β-anomer formation of chlor-
ambucil neoglycosides was generally less predominate.
Rather, a strong 1,2-trans relationship was typically observed
(Table S2 in Supporting Information).
Results and Discussion
Chlorambucil N-Alkoxyamino-Based Neoglycorandomiza-
tion (Scheme 1). Chlorambucil was converted to the Weinreb
Anticancer Activity of Chlorambucil N-Alkoxyamino-Based
Neoglycosides. The antiproliferative activity (i.e., GI50) of
neoglycosides 10-63 was evaluated using a 10-member panel
of human carcinomas from various lung, colorectal, liver,
breast, prostate, CNS, and ovarian cell lines with 1 and
neoaglycon 9 as comparators (see Figure 2). On the basis of
this study, aglycon 9 displayed slightly better GI50 values
(1.1-4.4 μM) compared to the parent drug 1 (5.0-11.0 μM)
and the vast majority of the library members had an average
GI50 also comparable to or lower than chlorambucil (see
Table S5 in Supporting Information). Out of the 54 analogues,
19 possessed GI50 values in the high nanomolar range in at least
one cell line, six of which displayed GI50 values in the high
nanomolar range in three or more lines. Of this latter set, two
neoglycosides (D-glucuronolactonide 39 and D-threoside 60;
Figure 1. Chlorambucil (1) and other nitrogen mustard compounds
including N-oxide 2 (PX-478), FDA-approved bendamustine (3),
and glycosylated variants glufosfamide (4) and 5. The latter was
found to inhibit the brain/erythrocyte D-glucose GLUT1 transporter
and thereby decrease glucose uptake (ref 42).
Scheme 1. Synthesis of the N-Alkoxyamino-Based Chlorambucil Librarya
a (a) MeON(H)Me, NMM, EDAC, 0 °C (97%); (b) LAH, THF, 0 °C (70%); (c) MeONH3Cl, Et3N, EtOH (84%); (d) BH3 Et3N, HCl, EtOH, 0 °C
3
(65%); (e) reducing sugar, MeOH, HOAc (1.5 equiv), 40 °C (63% av).

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8131
Figure 2. Summary of GI50 data from the high-throughput growth inhibition assay of 10-63 (reciprocal values displayed). Comparisons were
performed against the aglycon (9) and chlorambucil (1). GI50 data and error values are provided in Supporting Information Table 5.
Representative cancer cell lines tested include NCI-H460 (lung), A549 (lung), Du145 (prostate), SKOV3 (ovary), Hep3b (liver), SF268 (brain),
MCF7 (breast), HT29 (colorectal), HCT15 (colorectal), H1299 (lung).
see Figure 3) were identified as the two most potent neoglycosides.
In comparison to the parent 1, 39 displayed a 6-fold improve-
ment in average growth inhibition across the cell panel, with the
most sensitive cell line being SF268 glioblastoma (12-fold im-
proved). Likewise, D-threoside 60 presented an average elevated
potency of 8-fold over 1, with 12-, 13-, and 15-fold improved
activities toward HT29 (colorectal), H1299 (lung), and HCT15
(colorectal) cancers, respectively, over the parent drug.
The activity assessment described above revealed the follow-
ing general structure-activity relationships. First, sugars that
favor furanosyl-derived neoglycosides (e.g., threosides 60 and
61, glucuronolactonide 39, arabinosides 14 and 15, lyxosides 41
and 42, and xylosides 62 and 63) led to the greatest improve-
ments in anticancer activity. Second, relative configuration
within this furanoside-derived neoglycoside series influenced
potency. Specifically, those tetroses or pentoses with a 2,3-trans
dihydroxy orientation [e.g., D-threoside 60 (2S,3R), L-threoside
61 (2R,3S), and 39 (D-2S,3R)] were generally more potent
than those with a corresponding 2,3-cis configuration [e.g.,
D-erythroside 19 (2R,3R), D-riboside 53 (2R,3R), andL-riboside
54 (2S,3S)], while D- and L-saccharide enantiomers of this
furanoside-derived group were found to have similar values.
Third, among neohexosides, a 3R hydroxyl group (i.e., axial in
the chair conformation) enhances selectivity toward lung
(H1299) and colon (HCT-15) cancer cell lines (e.g., allosides
10 and 11, altrosides 12 and 13, and L-guloside 40). Finally,
most neoglycosides derived from sugars known to be GLUT
substrates and to mediate GLUT-dependent uptake of con-
jugates (e.g., D-glucoside 28 or 3-methoxy-D-glucoside 33)47-50
were not among the most active and/or selective hits identified.
Additionally, comparison of the anomeric composition of
either the most active neoglycosides or the library as a whole
to the inhibitory data does not reveal that a distinct correlation
exists between anomers and activity. To illustrate using the
Figure 3. Structures of the most antiproliferative chlorambucil
N-alkoxyamino-based neoglycosides against a 10-member carcinoma
panel.

8132 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Goff and Thorson
Scheme 2. Synthesis of the N-Acylhydrazine- and N-Hydroxyamino-Based Chlorambucil Librariesa
a (a) HONH3Cl, Et3N, EtOH (98%); (b) BH3 Et3N, HCl, EtOH, 0 °C (65%); (c) reducing sugar, MeOH, HOAc (1.5 equiv), 40 °C (47% av);
3
(d) (i) N-hydroxysuccinimide, DIC, THF, 40 °C, (ii) NH2NH2, pyridine, DMAP, 40 °C (85%); (e) reducing sugar, MeOH, HOAc (1.5 equiv), 40 °C (62% av).
antiproliferative neo-D-pentosides (the two with the most
extreme anomeric biases), D-arabinoside 14 (R/β 19/1) and
D-xyloside 62 (β only) have similar GI50 values in 9 of the
10 cell lines, not providing any distinction between the two
anomers and overall impact on growth inhibition.
Chlorambucil N-Hydroxyamino-Based Neoglycosylation
(Scheme 2). Given the clear impact of chlorambucil neoglyco-
sylation upon anticancer activity, we subsequently set out to
examine the specific contribution of the neoglycoside handle.
Similar to alkoxyamino-based chemoselective neoglycosyla-
tion,9-24 previous studies have revealed hydroxyamines51-53
and hydrazides54-59 to also provide the corresponding
closed-ring glycosides. Additionally, we envisioned that use
of a hydroxyamino handle may allow for additional modi-
fication at the hydroxyl group, providing a facile means for
further diversification.23 To create the two modified agly-
cons, only slight changes in the synthesis were required. In a
fashion similar to the procedure described in Scheme 1,
aldehyde 7 was combined with hydroxyamine HCl to form
a mixture of E- and Z-oximes (64) that were reduced with
BH3 Et3N complex. As with the synthesis described in
3
Scheme 1, only aglycon 65 required column chromatography
for purification in the four-step process from 1 (Scheme 2,
44% overall yield). Hydrazide formation (71) was achieved
by reacting the N-hydroxysuccinimidyl ester of 1 with
hydrazine in the presence of DMAP in pyridine. By use of
the previously described neoglycosylation conditions (see
Scheme 1), the focused hydroxyamine (66-70) and hydra-
zide (72-75) neoglycosyl sets were synthesized using sugars
identified as hits from the alkoxyamino-based series 10-63,
specifically D-glucuronolactone and D-threose, which had
superior performance throughout the panel; D-xylose, the
most active of the pentoses; and D-fucose, which displayed
the most pronounced selectivity toward one tumor line
(HCT-15 colorectal).
NMR analysis of the 66-70 neoglycosides revealed that
the desired closed-ring R- and β-anomeric forms were usually
in equilibrium with an open-chain imine isomer. Evidence
for this was based upon the observed 1H NMR chemical shift
of the H1 (doublets) and H2 (doublets of doublets) protons
at 7.2-7.5 and 5.1-5.0 ppm, respectively, indicating the
presence of an iminyl double bond. Such nitrone formation is
Figure 4. 1H NMR spectra of N-hydroxyaminochlorambucil glyco-
sides: (A) equilibrium between the cyclic neoglycoside and acyclic
nitrone of D-fucoside 66; (B) nitrone form of the chlorambucil
D-threoside 70. Both spectra were obtained at 500 MHz in CD3OD.
well-precedented for the condensation of aldehydes and
monosubstituted hydroxyamines.60 However, in the context
of glycoside formation, only a handful of such glycosides
have been reported, comprising protected saccharides and
small N-alkyl-N-hydroxyamines.51-53 From the current study,
the nature of the sugar influences the thermodynamic equi-
librium and thereby product distribution. Specifically, the
D-fucoside-derived 66 predominately adopted the closed-ring
isomer (67%), D-glucuronolactone (67) and D-ribose (68)
favored the open-chain nitrone (71% and 50%, respectively),

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8133
Figure 5. Interconversion of N-hydroxyaminochlorambucil-D-riboside
between the cyclic neoglycoside and acyclic nitrone 68 and per-
acetylated analogue 69. Both spectra were obtained at 500 MHz
with 68 in CD3OD and 69 in CDCl3.
while D-threose (70) led solely to nitrone (Figure 4, Table S3
in Supporting Information). These equilibria could be shifted
upon nitrone modification. For example, peracetylation of
neoriboside 68 using acetic anhydride and DMAP in THF
promoted ring closure based upon 1H NMR (Figure 5) and
ESI-MS analysis. Deacetylation of the peracetate 69 in
base reestablished the mixture of isomers, demonstrating
that changing the electronics of the nitrone oxygen will
promote ring-opening and closure, as will the type of ligated
sugar. Interestingly, the anomeric ratio also changed, mov-
ing from a 1/1 ratio as the nitrone to 2/1 as the tetraacetate,
contrasting from the 1/3 R/β ratio of methoxyamine
D-riboside (53).
Figure 6. 1H NMR spectra of N-hydrazidochlorambucil glyco-
sides: (A) equilibrium between the cyclic neoglycoside and acyclic
imine of D-glucurono-6,3-lactonide 73; (B) nitrone form of the chlo-
rambucil D-threoside 74. Both spectra were obtained at 500 MHz in
CD3OD.
Chlorambucil N-Acylhydrazine-Based Neoglycosylation
(Scheme 2). A conformational study of sugar acetylhydra-
zides by Bendiak indicated that natural hexose and pentose
(e.g., D-Glc, D-Gal, D-Xyl, etc.) analogues formed pyrano-
1
sides, as evidenced by H NMR.54 Similar findings were
reported for mono- and oligosaccharide chemoselective
ligations with biotinyl,55 long-chain acyl,56 peptidyl,57 and
adipyl hydrazides.59 In contrast, the current chlorambucil
study revealed a trend that mirrored the trend observed for
the 66-70 subset. Specifically, D-fucoside 72 was formed only
as the cyclic compound (67% yield), D-glucuronolactonide
73 and D-xyloside 75 were a mixture of the open and closed
conformers (46% and 47%, respectively), and D-threoside 74
adopted the open-chain imine (86%). These results indicate
that closure of the glycoside may be dependent not only on
the type of sugar but possibly on the nature of the aglycon as
well. It was also notable that only the β-anomer of the closed
rings was observed while the hydrazylimines were isolated as
mixtures of E- and Z-isomers (Figure 6).
Figure 7. Summary of GI50 data from the high-throughput growth
inhibition assay of 66-70 hydroxyamines and 72-75 hydrazides
(reciprocal values displayed). Comparisons were performed against
methoxyamines 39 and 60, aglycons 65 and 71, and chlorambucil
(1). GI50 data and error values are provided in Supporting Informa-
tion Table 5. Representative cancer cell lines tested include NCI-
H460 (lung), A549 (lung), Du145 (prostate), SKOV3 (ovary),
Hep3b (liver), SF268 (brain), MCF7 (breast), HT29 (colorectal),
HCT15 (colorectal), H1299 (lung).
Anticancer Activity of Chlorambucil N-Hydroxyamino- and
N-Acylhydrazine-Based Neoglycosides. Neoglycosides 66-70,
72-75, and their respective aglycons 65 and 71 were assayed for
antiproliferative activity using the same cell lines as for 10-63.
In general, the alternative handle glycosides did not perform as

8134 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Goff and Thorson
well as their methoxyamine analogues with the most notable
difference being between the 66-70 and 72-75 analogues.
While some of the hydroxyamine compounds produced com-
parable inhibitory responses to the methoxyamine group,
72-75 had universally diminished anticancer properties over
10-63 with average GI50 values 5- to 13-fold greater across the
10-member panel. 66-70 and 72-75 were also found to have
GI50 values up to 6-fold larger than the corresponding agly-
cons, though aglycons 65 and 71 were of similar activity to 9
(Table S5 in Supporting Information).
Two exceptions to this overall trend, D-riboside 68 and
its peracetylated variant 69, were 2- to 5-fold more active
than the alkoxyamino-based 53 in 6 of the 10 cell lines
(0.53-1.3 μM vs 1.8-5.9 μM, respectively). 68 and 69 are
also notable as being among the most active (i.e., mid- to
high-nanomolar range) neoglycosides against the SKOV3
ovarian cancer line (see Figure 7). These data suggest the type
of neoglycoside handle employed greatly influences desired
activity. When taken into account that aglycons 9, 65, and 71
are of similar potency, the variation in activity between the
methoxyamino-, hydroxyamino-, and acylhydrazine-based
neoglycosides implicates the nature of the glycosidic bond to
be the most significant contributor wherein a prevalence of
the acyclic nitrone or imine conjugate led to a reduction in
potency.
Experimental Section
Materials and General Methods. Mass spectrometric data
were obtained on either a Waters (Milford, MA) LCT time-
of-flight spectrometer for electrospray ionization (ESI) or a Varian
ProMALDI (Palo Alto, CA) Fourier transform ion cyclotron
resonance mass spectrometer (FTICR) equipped with a 7.0 T
actively shielded superconducting magnet and a Nd:YAG laser.
NMR spectra were obtained on a Varian UnityInova 500 MHz
instrument (Palo Alto, CA) using 99.8% CDCl3 with 0.05% v/v
TMS or 99.8% CD3OD in ampules. 1H and 13C chemical shifts
were referenced to TMS (for CDCl3) or nondeuterated solvent
(for CD3OD). Multiplicities are indicated by s (singlet), d
(doublet), t (triplet), q (quartet), qui (quintet), m (multiplet),
and br (broad). Chemical shift assignments for anomeric
mixtures, where possible, are noted as R or β with the atom
responsible for the shift. 1H NMR characterization was supple-
mented with gCOSY for all neoglycoside library members as
well as 13C and gHSQC for pilot reactions and alternative
handle compounds. Tetrahydrofuran was dried using a column
of activated alumina. All other solvents were used as provided
by the supplier. Reagents were obtained from Aldrich or Sigma
and were used as received. Flash chromatography was per-
formed using 40-63 μm particle size silica gel. Thin layer chro-
matography was performed on aluminum-backed, 254 nm
UV-active plates with a silica gel particle size of 60 μm. Library
purity was assessed by reverse phase HPLC using a Varian (Walnut
Creek, CA) ProStarunitwithaPhenomenex(Torrance, CA) Luna
C18 4.6 mm ꢀ 250 mm column running a H2O/MeCN 90/10 to
10/90 gradient over 13 m, followed by a 5 m isocratic flow, at a
rate of 1.0 mL/m, A254 detection. The purity of the neoglycosides
and aglycons was assessed to be greater than 95%, unless
specified otherwise (see Table S4 in Supporting Information).
N,O-Dimethyl-4-(4-N0,N0-bis(2-chloroethyl)amino)phenyl-
butanamide (6). Chlorambucil (1, 516 mg, 1.70 mmol) was
dissolved in CH2Cl2 (15 mL) and cooled to 0 °C before adding
N,O-dimethylhydroxylamine HCl (181 mg, 1.86 mmol) and
N-methylmorpholine (200 μL, 1.82 mmol). The coupling agent
EDAC (347 mg, 1.81 mmol) was then added to the reaction
slowly over 5 min to ensure dissolution. After 60 min, the
reaction was quenched with 5% HCl (20 mL) and the acidic
layer extracted with CH2Cl2 (20 mL). The organic layers were
combined, washed with saturated aqueous NaHCO3 (20 mL),
and the basic layer was extracted with CH2Cl2 (20 mL). After the
organic layers were combined, washed with brine (20 mL), and
dried with Na2SO4, solvent removal yielded a colorless oil
(571 mg, 97%, Rf = 0.23, EtOAc/Hex 1:2) that was used
Conclusions
This study revealed a facile four-step process, which could
be conducted on gram scale in less than 2 h of reaction time
and required only a single chromatographic separation to
modify the drug chlorambucil (1) for chemoselective gly-
cosylation. Whileprior synthesesof1 glycoconjugatesfocused
upon the use of typical metabolic sugars designed to enhance
sugar-mediated uptake and led to modest overall improve-
ments compared to 1,39,40 the current study revealed that
anticancer potency optimization was best accomplished via
conjugation with novel nonmetabolic sugars, culminating in
the discovery of D-threoside 60 as the most active chlorambu-
cil glycoside reported to date. The discovery of 60 opens the
door to a series of new questions relating to the precise
mechanism(s) of improvement, including among the many
possibilities: (i) modulation of uptake (via novel targeting of
known transporters and/or even raising the possibility of new
sugar transport/receptor mediated-processes); (ii) intracellular
stabilization of the alkylating reagent (basically extending the
intracellular T1/2); (iii) enhancement of the productive agent-
DNAinteractions(e.g., DNA affinityand/or specificity); and/
or even (iv) alternative targeting of the active species (e.g.,
RNA, proteins, and/or membrane targets). While the specific
mechanisms remain to be elucidated, it is important to note
that precedent does exists for 1 glucosylation to alter the
mechanism of cellular uptake.34,42
In addition to the lead discovery aspect of this project, it is
also important to note that the corresponding in-depth product
distribution analysis among N-alkoxyamino-, N-acylhydrazine-,
and N-hydroxyamino-based neoglycosylation reactions revealed
sugar-dependent partitioning among open- and closed-ring
neoglycosides in the last two cases. Thus, this cumulative study
also sheds new light (and a potential note of caution) on an
underappreciated chemical variability of glycoconjugates gen-
erated via the N-acylhydrazine- and N-hydroxyamine-based
chemoselective glycosylation methods more commonly applied
in glycobiology.18
1
without further purification. H NMR (CDCl3, 500 MHz) δ
7.09 (d, J = 8.8 Hz, 2 H), 6.62 (d, J = 8.8 Hz, 2 H), 3.71-3.65
(m, 4 H), 3.64 (s, 3 H), 3.62-3.59 (m, 4 H), 3.17 (s, 3 H), 2.58 (t,
J = 7.7 Hz, 2 H), 2.44 (t, J = 7.0 Hz, 2 H), 1.95-1.88 (m, 2 H);
13C NMR (CDCl3, 125 MHz) δ 144.37, 131.17, 129.80, 112.31,
61.30, 53.74, 40.69, 34.33, 32.29, 31.35, 26.44; HRMS (ESI) m/z
for C16H25Cl2N2O2 ([M þ H]þ) 347.1295, calcd 347.1288.
4-(4-N0,N0-Bis(2-chloroethyl)amino)phenylbutanal (7). Weinreb
amide 6 (571 mg, 1.64 mmol) was dissolved in anhydrous THF
(8 mL) under Ar and cooled to 0 °C. A suspension of lithium
aluminum hydride in anhydrous THF (1.0 M, 1 mL) was added in
one aliquot. After 5 min, the reaction was quenched with saturated
aqueous KHSO4 (5 mL) followed by deionized water (5 mL).
Extraction of the aldehyde was performed with Et2O (2 ꢀ 15 mL),
and the organic layer was dried with Na2SO4. Solvent evapora-
tion provided a yellowish oil (334 mg, 70%, Rf = 0.65, EtOAc/
Hex 1:2). 1H NMR (CDCl3, 500 MHz) δ 9.80 (s, 1 H), 7.13 (d, J =
8.8 Hz, 2 H), 6.70 (d, J = 8.8 Hz, 2 H), 3.77-3.74 (m, 4 H), 3.69-
3.66 (m, 4 H), 2.63 (t, J = 7.6 Hz, 2 H), 2.49 (td, J = 7.3, 1.5 Hz,
2 H), 2.00-1.94 (m, 2 H); 13C NMR (CDCl3, 125 MHz) δ 202.42,
144.46, 130.37, 129.69, 112.23, 53.56, 43.16, 40.60, 33.91, 23.93;
HRMS (MALDI) m/z for C14H20Cl2NO ([M þ H]þ) 288.090 91,
calcd 288.091 65.

Article
N-Methoxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutanimine
(8). Aldehyde 7 (334 mg, 1.16 mmol) was dissolved in absolute
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8135
4-(4-N0,N0-Bis(2-chloroethyl)amino)phenylbutanoic Hydrazide (71).
Chlorambucil (1, 319 mg, 1.05 mmol) was dissolved in THF (10 mL),
along with N-hydroxysuccinimide (132 mg, 1.15 mmol) and DIC
(180μL, 1.16 mmol). The mixture was warmed to 40 °C and stirred for
5.5 h. Hydrazine (40 μL, 1.27 mmol), pyridine (10 μL), and DMAP
(5 mg, 0.04 mmol) were then introduced, and the reaction proceeded
for another 20 min. The solvent was removed in vacuo, yielding a
yellowish solid, which was suspended in MeOH/CH2Cl2 3:97 and
purified by column chromatography (SiO2, MeOH/CH2Cl2 3:97).
The hydrazide product was collected as an opaque oil that solidified at
-20 °C to a white amorphous solid (285 mg, 85% Rf = 0.39, MeOH/
CH2Cl2 5:95). 1H NMR (CDCl3, 500 MHz) δ 7.31 (s br, 1 H), 7.05
(d, J = 8.7 Hz, 2 H), 6.61 (d, J = 8.7 Hz, 2 H), 3.90 (s br, 2 H),
3.71-3.66 (m, 4 H), 3.63-3.58 (m, 4 H), 2.54 (t, J = 7.6 Hz, 2 H),
2.17-2.13 (m, 2 H), 1.95-1.89 (m, 2 H); 13C NMR (CDCl3, 125
MHz) δ 173.80, 144.46, 130.47, 129.72, 112.26, 53.62, 40.66, 40.88,
34.10, 33.77, 27.19; HRMS (ESI) m/z for C14H22Cl2N3O ([M þ H]þ)
318.1140, calcd 318.1134.
EtOH (10 mL) followed by the addition of MeONH2 HCl (247 mg,
3
2.95 mmol) and Et3N(400μL, 2.87 mmol). After 35 min, the solvent
was removed in vacuo, providing the crude product as a white
crystalline solid. The material was suspended in EtOAc/Hex 1:7
(50 mL) and filtered through a silica gel plug, which was flushed with
solvent (400 mL). The purified imine was provided as a colorless oil
containing a mixture of E- and Z-isomers (308 mg, 84%, RfA
=
0.51, RfB = 0.43, EtOAc/Hex 1:7). 1H NMR (CDCl3, 500 MHz) δ
7.36 (t, J = 6.2 Hz, 1 H), 7.05 (d, J = 8.8 Hz, 2 H), 6.61 (d, J = 8.8
Hz, 2 H), 3.80 (s, 3 H), 3.69-3.65 (m, 4 H), 3.61-3.57 (m, 4 H), 2.55
(t, J = 7.7 Hz, 2 H), 2.18 (q, J = 7.7 Hz, 2 H), 1.75 (qui, J = 7.7 Hz,
2H);13CNMR(CDCl3, 125 MHz) δ150.58, 144.38, 130.87, 129.74,
112.28, 61.27, 53.67, 40.61, 34.14, 29.03, 28.72; HRMS (ESI) m/z for
C15H23Cl2N2O ([M þ H]þ) 317.1187, calcd 317.1182.
N-Methoxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutylamine
(9). Imine 8 (308 mg, 0.971 mmol) was dissolved in absolute EtOH
(15 mL), cooled to 0 °C, and the reducing agent BH3 Et3N (710 μL,
General Procedure for Neoglycoside Library Synthesis, Purifi-
cation, and Characterization. Aglycon 9, 65, or 71 (typically
0.12-0.16 mmol) was added to 1 dram vials along with stir fleas
and dissolved in MeOH such that the aglycon was at 90-100 mM.
Glacial acetic acid (1.5 equiv) was introduced, reducing sugars
(2 equiv) were added, the vials capped, and the vessels placed on
a heating block/stir plate to react at 40 °C for 3-48 h. The vial
caps were removed, and the solvent was evaporated by a Speedvac
apparatus (55 °C, 3 h). Crude neoglycosides were suspended
in CH2Cl2 (200 μL) using a vortex mixer and then loaded onto
2000 mg silica gel solid phase extraction (SPE) columns (Alltech,
Deerfield, IL) that were prewashed with MeOH/CH2Cl2 2/98.
The SPEs were eluted using a vacuum manifold, collecting
fractions with a volume of approximately 2 mL. After the initial
two fractions were obtained, eluting any unreacted aglycon or
relatively nonpolar material, the following step gradients were
used: MeOH/CH2Cl2 5/95 for tetroses, pentoses, and substit-
uted hexoses; MeOH/CH2Cl2 10/90 for hexoses; MeOH/CHCl3
15/85 for disaccharides; MeOH/CHCl3 20/80 for glycurono-
sides. For polyprotected saccharides, a gradient of EtOAc/
Hex 1/6 to 1/5 was used. Typically, all neoglycoside was eluted
by the 10th or 11th fraction, leaving unreacted sugar on the
column. The fractions containing pure product were identified
by TLC using UV light (254 nm) and p-anisaldehyde stain, then
combined and dried. Library members were characterized by
1H and gCOSY NMR as well as either high-resolution ESI or
MALDI mass spectrometry (see Tables S2 and S3 in Supporting
Information). Anomeric ratios were obtained by comparison of
anomeric proton integration (see Tables S2 and S3 in Support-
ing Information).
3
4.83 mmol) was added in one aliquot. Concentrated HCl/EtOH 1:1
(800 μL) was then dripped in slowly over 5 min. The reaction was
quenched with saturated aqueous NaHCO3 (5 mL) 5 min after the
acid solution was completely added, resulting in a white slurry. The
organic solvent was removed under reduced pressure. Then more
NaHCO3 (5 mL) solution was added. The aqueous material was
extracted with CH2Cl2 (3 ꢀ 20 mL) and dried with Na2SO4. The
crude oil was purified by flash chromatography (SiO2, EtOAc/
Hex 1:2), providing the aglycon as a colorless oil (200 mg, 65%,
Rf = 0.37, EtOAc/Hex 1:2). 1H NMR (CDCl3, 500 MHz) δ7.05 (d,
J = 8.6 Hz, 2 H), 6.60 (d, J = 8.7 Hz, 2 H), 5.50 (s br, 1 H),
3.69-3.65 (m, 4 H), 3.61-3.57 (m, 4 H), 3.50 (s, 3 H), 2.91 (t, J =
7.1 Hz, 2 H), 2.53 (t, J = 7.5 Hz, 2 H), 1.61-1.58 (m, 2 H),
1.55-1.49 (m, 2 H); 13C NMR (CDCl3, 125 MHz) δ 144.20, 131.63,
129.62, 112.22, 61.80, 53.65, 51.80, 40.61, 34.66, 29.32, 26.98; HRMS
(ESI) m/z for C15H25Cl2N2O ([M þ H]þ) 319.1334, calcd 319.1339.
N-Hydroxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutanimine
(64). Aldehyde 7 (442 mg, 1.53 mmol) was dissolved in absolute
EtOH (10 mL) followed by the addition of HONH2 HCl (269 mg,
3
3.87 mmol) and Et3N(550μL, 3.95 mmol). After 25 min, the solvent
was removed in vacuo, providing the crude product as a white flaky
solid. The material was purified by column chromatography (SiO2,
EtOAc/Hex 1:2), which yielded the imine as a colorless mixture of
E- and Z-isomers (455 mg, 98%, RfA = 0.53, RfB = 0.45, EtOAc/
Hex 1:2). 1H NMR (CDCl3, 500 MHz) δ 7.29 (t, J = 6.3 Hz, 1 H),
7.06 (d, J = 8.8 Hz, 2 H), 6.61 (d, J = 8.8 Hz, 2 H), 3.68-3.65 (m,
4 H), 3.61-3.58 (m, 4 H), 2.55 (t, J = 7.7 Hz, 2 H), 2.15 (q, J =
7.7 Hz, 2 H), 1.75 (qui, J = 7.7 Hz, 2 H); 13C NMR (CDCl3,
125 MHz) δ 147.98, 144.45, 131.01, 129.69, 112.22, 53.76, 40.59,
33.23, 29.10, 28.73; HRMS (ESI) m/z for C14H21Cl2N2O ([M þ
H]þ) 303.1019, calcd 303.1025.
N-Methoxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutyl-
amino-D-riboside (53). Aglycon 9 (40 mg, 0.13 mmol) was placed
into a 1 dram vial, dissolved in MeOH (1.34 mL), and mixed
with glacial acetic acid (10.7 μL). After addition of D-ribose
(57 mg, 0.38 mmol), the mixture was capped, warmed to 40 °C,
and allowed to stir for 3 h. Solvent was subsequently removed in
vacuo and the resulting crude solid suspended in MeOH/
CH2Cl2 2:98 (250 μL) by vortex mixer. The mixture was purified
by SPE (SiO2, MeOH/CH2Cl2 2:98 to 5:95), providing the white
solid neoglycoside as a mixture of anomers (31 mg, 54%, Rf =
0.34, MeOH/CH2Cl2 5:95; R/β1:3). 1HNMR(CD3OD, 500 MHz)
δ 7.08 (d, J = 8.7 Hz, 2 H), 6.68 (d, J = 8.7 Hz, 2 H), 4.56 (d, J =
3.4 Hz, 0.25 H, R-H10), 4.27 (d, J = 8.7 Hz, 0.75 H, β-H10), 4.15
(dd, J = 5.4, 3.5 Hz, 0.25 H, R-H20), 4.12 (s br, 0.75 H, β-H30),
3.99 (t, J = 5.6 Hz, 0.25 H, R0 -H30), 3.90-3.86 (m, 0.25, R-H40),
3.74-3.68 (m, 5 H, R-H5A þ β-H5A0), 3.68-3.60 (m, 6.5 H,
N-Hydroxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutyl-
amine (65). Imine 64 (455 mg, 1.50 mmol) was dissolved in
absolute EtOH (10 mL), cooled to 0 °C, and the reducing agent
BH3 Et3N (1.1 mL, 7.5 mmol) was added in one aliquot.
3
Concentrated HCl/EtOH 1:1 (1.26 μL) was then dripped in
slowly over 5 min. The reaction was quenched with saturated
aqueous NaHCO3 (5 mL) 5 min after the acid solution was
completely added, resulting in a white slurry. The organic sol-
vent was removed under reduced pressure. Then more NaHCO3
(5 mL) solution was added. The aqueous material was extracted
with CH2Cl2 (4 ꢀ 25 mL) and dried with Na2SO4. The crude oil
was purified by column chromatography (SiO2, MeOH/CH2Cl2
2:98 to 5:95), providing the aglycon as a colorless oil (302 mg,
66%, Rf = 0.36, MeOH/CH2Cl2 5:95). 1H NMR (CDCl3,
500 MHz) δ 7.05 (d, J = 8.7 Hz, 2 H), 6.61 (d, J = 8.7 Hz,
2 H), 5.23 (s br, 2 H), 3.70-3.64 (m, 4 H), 3.63-3.57 (m, 4 H),
3.40 (s, 2 H), 2.58-2.51 (m, 2 H), 1.96-1.84 (m, 2 H), 1.67-1.55
(m, 2 H); 13C NMR (CDCl3, 125 MHz) δ 144.60, 131.08, 129.84,
112.47, 53.81, 50.43, 40.88, 34.58, 29.00, 27.16; HRMS (ESI)
m/z for C14H23Cl2N2O ([M þ H]þ) 305.1175, calcd 305.1182.
0
0
R-H5B þ β-H20 þ β-H40 þ β-H5B ), 3.58 (s, 3 H), 3.05-2.98
(m, 0.75 H, β-H2A), 2.97-2.92 (m, 0.25 H, R-H2A), 2.86-2.77
(m, 1 H, H2B), 2.56 (t, J = 7.2 Hz, 2 H), 1.71-1.56 (m, 4 H); 13
C
NMR (CD3OD, 125 MHz) δ 145.86, 132.73, 130.66, 113.61,
100.76 (R-C10), 91.14 (β-C10), 84.50 (R-C40), 73.16 (R-C20), 72.52
(R-C30), 72.46 (β-C30), 68.94 (β-C20), 68.65 (β-C40), 65.94

8136 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Goff and Thorson
(β-C50), 64.21 (R-C50), 62.82, 54.76 (R-C2), 54.68, 54.05 (β-C2),
41.82, 35.76, 30.70, 28.12; HRMS (ESI) m/z for C20H32Cl2-
N2NaO5 ([M þ Na]þ) 473.1581, calcd 473.1587.
(nitrone-C40), 69.65 (β-C50), 66.13 (nitrone-C20), 65.02 (nitrone-
C2), 55.41 (β-C2), 53.42, 40.66, 34.55 (β-C5), 33.99 (nitrone-C5),
29.15 (β-C3), 28.17, 26.79 (nitrone-C3); HRMS (MALDI) m/z
for C20H28Cl2N2NaO6 ([M þ Na]þ) 485.122 54, calcd 485.121 66.
N-Hydroxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutylamino-
D-riboside (68). By use of aglycon 65 (50 mg, 0.16 mmol), the prod-
uct mixture was yielded as a white solid that was visualized as a
single spot by TLC (18 mg, 25%, Rf = 0.39 MeOH/CH2Cl2 10:90,
R/β/nitrone 1:1:2). 1H NMR (CD3OD, 500 MHz) δ 7.17 (d, J =
6.3 Hz, 0.5 H, nitrone-H10), 7.07 (d, J = 8.7 Hz, 2 H), 6.66 (d, J =
8.7 Hz, 2 H), 4.99 (dd, J = 6.3, 3.5 Hz, 0.5 H, nitrone-H20), 4.51
(d, J = 2.9 Hz, 0.25 H, R-H10), 4.18-4.17 (m, 0.25 H, R-H20), 4.16
(d, J = 8.7 Hz, 0.25 H, β-H10), 4.12-4.09 (m, 0.5 H, R/β-H40 þ
β-H20), 3.89-3.86 (m, 0.5 H, R-H30 þ R/β-H5A0), 3.86-3.85 (m,
N-Methoxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutylamino-
L-riboside (54). By use of the same procedure as 53, 9 (52 mg,
0.16 mmol) combined with L-ribose (48 mg, 0.32 mmol), yielded
the anomeric mixture as a white solid (40 mg, 54%, Rf = 0.34,
MeOH/CH2Cl2 5:95; R/β 1:3). 1H NMR (CD3OD, 500 MHz) δ
7.05 (d, J = 8.6 Hz, 2 H), 6.67 (d, J = 8.6 Hz, 2 H), 4.52 (d, J =
3.4 Hz, 0.25 H, R-H10), 4.25 (d, J = 8.7 Hz, 0.75 H, β-H10), 4.11
(dd, J = 5.4, 3.5 Hz, 0.25 H, R-H20), 4.08 (s br, 0.75 H, β-H30),
3.95 (t, J = 5.5 Hz, 0.25 H,0R-H30), 3.87-3.83 (m, 0.25, R-H40),
3.71-3.67 (m, 5 H, R-H5A þ β-H5A0), 3.65-3.61 (m, 5.75 H,
0
0
R-H5B þ β-H40 þ β-H5B ), 3.60-3.59 (m, 0.75 H, β-H20), 3.55
(s, 3 H), 3.00-2.95 (m, 0.75 H, β-H2A), 2.94-2.89 (m, 0.25 H,
R-H2A), 2.81-2.76 (m, 1 H, H2B), 2.53 (t, J = 7.5 Hz, 2 H),
1.66-1.55 (m, 4 H); 13C NMR (CD3OD, 125 MHz) δ 145.96,
132.83, 130.69, 113.70, 100.85 (R-C10), 91.20 (β-C10), 84.59
(R-C40), 73.24 (R-C20), 72.61 (R-C30), 72.56 (β-C30), 69.02
(β-C20), 68.74 (β-C40), 66.00 (β-C50), 64.28 (R-C50), 62.81,
54.78 (R-C2), 54.76, 54.07 (β-C2), 41.86, 35.80, 30.75, 28.15;
HRMS (ESI) m/z for C20H32Cl2N2NaO5 ([M þ Na]þ) 473.1579,
calcd 473.1581.
0
1 H, nitrone-H2), 3.84-3.81 (m, 0.25 H, R/β-H5A ), 3.78 (dd, J =
0
11.5, 3.5 Hz, 0.5 H, nitrone-H30), 3.74-3.67 (m, 5 H, nitrone-H5A
0
0
þ R/β-H40 þ R/β-H5B ), 3.65-3.58 (m, 5 H, nitrone-H5B
þ
R/β-H5B þ β-H30), 3.55-3.51 (m, 0.5 H, nitrone-H40), 3.08-
3.02 (m, 0.25 H, R-H2A), 2.98-2.93 (m, 0.25 H, R-H2B), 2.74-2.67
(m, 0.5 H, β-H2), 2.57 (t, J = 7.4 Hz, 2 H), 1.92-1.86 (m, 2 H),
1.68-1.60 (m, 2 H); 13C NMR (CD3OD, 125 MHz) δ 144.79,
144.63 (nitrone-C10), 130.80, 129.52, 112.45, 100.86 (R-C10), 91.12
(β-C10), 83.43 (R-C30), 74.19 (nitrone-C30), 72.98 (R-C20), 72.31
(nitrone-C40), 70.79 (β-C20), 70.71 (R/β-C40), 68.16 (nitrone-C20),
67.88 (β-C30), 67.55 (R/β-C40), 64.85 (R/β-C50), 64.82, 63.58
(nitrone-C50), 62.59 (R/β-C50), 53.54, 40.57, 34.09, 28.22, 26.78;
HRMS (ESI) m/z for C19H30Cl2N2NaO5 ([M þ Na]þ) 459.1441,
calcd 459.1424.
0
N-Hydroxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutylamino-
D-fucoside (66). By use of aglycon 65 (63 mg, 0.21 mmol), the
mixture of compounds was yielded as a white solid (54 mg, 58%,
Rf = 0.11, MeOH/CH2Cl2 5:95, R/β/nitrone 2.7:1:1.8). 1H NMR
(CD3OD, 500 MHz) δ 7.25 (d, J = 5.8 Hz, 0.33 H, nitrone-H10),
7.10-7.07 (m, 2 H), 6.70-6.67 (m, 2 H), 5.06 (dd, J = 5.8,
2.1 Hz, 0.33 H, nitrone-H20), 4.40 (d, J = 4.9 Hz, 0.18 H, R-H10),
4.30 (dd, J = 5.8, 4.9 Hz, 0.18 H, R-H20), 4.08 (dd, J = 6.6,
1.8 Hz, 0.33 H, nitrone-H50), 3.96 (dd, J = 7.5, 5.8 Hz, 0.18 H,
R-H30), 3.89 (dd, J = 8.7, 2.1 Hz, 0.33 H, nitrone-H30), 3.86-
3.79 (m, 1.33 H, nitrone-H2 þ R-H50 þ β-H10), 3.73-3.70
(m, 4.49 H, β-H20), 3.67-3.63 (m, 4.18 H, R-H40), 3.61-3.58
(m, 0.98 H, β-H40 þ β-H50), 3.53-3.50 (m, 0.49 H, β-H30), 3.48
(dd, J = 8.7, 1.8 Hz, 0.33 H, nitrone-H40), 3.11-3.06 (m, 0.49 H,
β-H2A), 3.05-3.01 (m, 0.18 H, R-H2A), 2.79-2.75 (m, 0.49 H,
β-H2B), 2.74-2.68 (m, 0.18 H, R-H2B), 2.60-2.54 (m, 2 H),
1.91-1.87 (m, 0.66 H, nitrone-H3), 1.68-1.63 (m, 3.34 H),
1.30-1.26 (m, 3 H, H60); 13C NMR (CD3OD, 125 MHz) δ
145.54 (nitrone-C10), 144.64, 131.68, 129.45, 112.41, 99.46 (R-
C10), 94.89 (β-C10), 86.70 (β-C20), 78.52 (R-C20), 77.32 (R-C30),
74.99 (β-C30), 73.68 (nitrone-C40), 72.48 (β-C50), 72.36 (β-C40),
71.32 (nitrone-C30), 67.76 (R-C50), 67.38 (R-C40), 67.26 (nitrone-
C20), 66.21 (nitrone-C50), 64.69 (nitrone-C2), 54.87 (β-C2),
53.72 (R-C2), 53.49, 40.64, 34.59, 29.36, 28.16 (R/β-C3), 26.97
(nitrone-C3), 18.95 (R/β-C60), 18.66 (nitrone-C60); HRMS
(MALDI) m/z for C20H32Cl2N2NaO5 ([M þ Na]þ) 473.159 11,
calcd 473.158 05.
N-Acetoxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutyl-
amino-D-riboside Peracetate (69). The D-riboside mixture desig-
nated 68 (23 mg, 0.053 mmol) was peracetylated by dissolving in
THF (5 mL) and adding acetic anhydride (0.5 mL, 5 mmol),
DMAP (1 mg, 0.008 mmol), and Et3N (0.5 mL, 4 mmol). After
20 min, the solvent was removed in vacuo and the residue was
purified by SPE chromatograpy (SiO2, EtOAc/Hex 2:3). The
product was collected as a mixture of anomers of a colorless oil
(32 mg, >99%, Rfβ = 0.38 RfR = 0.32, EtOAc/Hex 2:3, R/β =
2:1). 1H NMR (CDCl3, 500 MHz) δ 7.06-7.04 (m, 2 H), 6.61 (d,
J = 8.6 Hz, 2 H), 5.62 (s, 0.33 H, β-H30), 5.27-5.25 (m, 0.67 H,
R-H20), 5.21 (dd, J = 6.1, 5.4 Hz, 0.67 H, R-H30), 5.07 (dd, J =
9.2, 2.6 Hz, 0.33 H, β-H20), 5.04-5.01 (m, 0.33 H, β-H40), 4.89
(d, J = 2.9 Hz, 0.67 H, R-H10), 4.58 (d, J = 9.2 Hz, 0.33 H,
β-H10), 4.33-4.31 (m, 0.67 H, R-H5A0), 4.23-4.20 0(m, 0.67 H,
R-H40), 4.16 (dd, J = 11.7, 5.4 Hz, 0.67 H, R-H5B ), 3.97 (dd,
J = 10.9, 5.4 Hz, 0.33 H, β-H5A0), 3.71-3.67 (m, 4.33 H,
0
β-H5B ), 3.63-3.60 (m, 4 H), 3.20-3.15 (m, 0.33 H, β-H2A),
3.11-3.06 (m, 0.67 H, R-H2A), 2.99-2.91 (m, 1 H, R-H2B
þ
β-H2B), 2.57-2.47 (m, 2 H), 2.11 (s, 3 H), 2.10 (s, 3 H), 2.09 (s, 3
H), 2.05 (s, 3 H), 1.67-1.46 (m, 4 H); 13C NMR (CDCl3, 125
MHz) δ 170.72, 169.76, 169.51, 169.49, 144.35, 131.49, 129.74,
112.36, 96.56 (R-C10), 88.60 (β-C10), 77.85 (R-C40), 71.99 (R-
C20), 70.58 (R-C30), 68.93 (β-C30), 66.47 (β-C40), 65.72 (β-C20),
63.55 (R-C50), 63.21 (β-C50), 54.06, 53.79, 40.71, 34.55, 29.20,
26.53, 20.93, 20.88, 20.68, 20.65; HRMS (ESI) m/z for
C27H38Cl2N2NaO9 ([M þ Na]þ) 627.1859, calcd 627.1847.
N-Hydroxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutyl-
amino-D-threoside Nitrone (70). By use of aglycon 65 (35 mg,
0.11 mmol), the neoglycoside nitrone was yielded as a colorless
syrup (23 mg, 49%, Rf = 0.11, MeOH/CH2Cl2 5:95). 1H NMR
(CD3OD, 500 MHz) δ 7.22 (d, J = 5.8 Hz, 1 H, H10), 7.10 (d,
J = 8.6 Hz, 2 H), 6.70 (d, J = 8.6 Hz, 2 H), 4.84 (dd, J = 5.8,
4.2 Hz, 1 H, H20), 3.87-3.84 (m, 3 H, H30 þ H40), 3.75-3.72 (m,
4 H), 3.68-3.61 (m, 6 H), 2.59 (t, J = 7.4 Hz, 2 H), 1.93-1.87
(m, 2 H), 1.69-1.63 (m, 2 H); 13C NMR (CD3OD, 125 MHz) δ
146.11, 146.01 (C10), 132.10, 130.72, 113.71, 73.93 (C30), 68.77
(C20), 65.95 (C40), 64.09, 54.70, 41.85, 35.30, 29.41, 27.93;
HRMS (MALDI) m/z for C18H28Cl2N2NaO4 ([M þ Na]þ)
429.131 88, calcd 429.131 83.
N-Hydroxy-4-(4-N0,N0-bis(2-chloroethyl)amino)phenylbutyl-
amino-D-glucurono-6,3-lactonide (67). By use of aglycon 65
(58 mg, 0.19 mmol), the mixture of compounds was yielded as
a colorless syrup (49 mg, 56%, Rf = 0.28, MeOH/CH2Cl2 5:95,
1
R/β/nitrone 0:1:2.5). H NMR (CD3OD, 500 MHz) δ 7.25 (d,
J = 6.3 Hz, 0.71 H, nitrone-H10), 7.10 (d, J = 8.6 Hz, 2 H), 6.70
(d, J = 8.6 Hz, 2 H), 5.11-5.08 (m, 0.71 H, nitrone-H20),
4.85-4.83 (m, 0.29 H, β-H40), 4.79 (d, J = 4.4 Hz, 0.29 H,
β-H30), 4.57 (d, J = 2.3 Hz, 0.29 H, β-H10), 4.54 (d, J = 4.6 Hz,
0.71 H, nitrone-H50), 4.49 (d, J = 2.0 Hz, 0.29 H, β-H50), 4.47
(d, J = 2.8 Hz, 0.71 H, nitrone-H30), 4.46 (d, J = 2.3 Hz, 0.29 H,
β-H20), 4.33 (dd, J = 4.6, 2.8 Hz, 0.71 H, nitrone-H40), 3.91 (q,
J = 7.0 Hz, 1.42 H, nitrone-H2), 3.74-3.71 (m, 4 H), 3.67-3.65
(m, 4 H), 2.99-2.96 (m, 0.29 H, β-H2A), 2.69-2.64 (m, 0.29 H,
β-H2B), 2.59 (t, J = 7.5 Hz, 1.42 H, nitrone-H5), 2.69-2.64 (dd,
J = 7.2, 6.7 Hz, 0.58 H, β-H5), 1.94-1.87 (m, 1.42 H, nitrone-
H3), 1.68-1.62 (m, 2.58 H, β-H3 þ H4); 13C NMR (CD3OD,
125 MHz) δ 176.39 (C60), 144.87, 141.51 (nitrone-C10), 130.76,
129.48, 112.45, 104.94 (β-C10), 85.74 (β-C30), 80.64 (nitrone-
C30), 77.57 (β-C40), 77.24 (β-C20), 70.90 (nitrone-C50), 69.97
4(4-N0,N0-Bis(2-chloroethyl)amino)phenylbutylhydrazido-
β-D-fucoside (72). By use of aglycon 71 (47 mg, 0.15 mmol), the

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8137
product was yielded as a colorless syrup (46 mg, 67%, Rf = 0.28
MeOH/CH2Cl2 10:90). H NMR (CD3OD, 500 MHz) δ 7.09
13C NMR (CD3OD, 125 MHz) δ 174.67, 144.87, 130.53, 129.49,
112.43, 91.78 (β-C10), 87.51 (imine A-C20), 77.17 (β-C30), 72.44
(imine A/B-C40), 72.43 (imine B-C30), 71.57 (imine A/B-C40),
72.00 (imine B-C20), 71.34 (β-C20), 70.92 (imine A-C30), 70.05
(β-C40), 67.28 (β-C50), 62.95 (imine-C50), 53.43, 40.60, 33.99,
33.22, 27.67; HRMS (MALDI) m/z for C19H29Cl2N3NaO5
([M þ Na]þ) 472.137 12, calcd 472.137 65.
1
(d, J = 8.6 Hz, 2 H), 6.70 (d, J = 8.6 Hz, 2 H), 3.86 (d, J =
8.7 Hz, 1 H, H10), 3.74-3.72 (m, 4 H), 3.68-3.65 (m, 4 H),
3.64-3.62 (m, 2 H, H30 þ H50), 3.54 (dd, J = 9.5, 3.2 Hz, 1 H,
H40), 3.48 (d, J = 8.7 Hz, 1 H, H20), 2.56 (t, J = 7.5 Hz, 2 H),
2.21 (t, J = 7.5 Hz, 2 H), 1.90 (qui, J = 7.5 Hz, 2 H), 1.28 (d, J =
6.4 Hz 3 H); 13C NMR (CD3OD, 125 MHz) δ 174.72, 144.86,
130.58, 129.53, 112.44, 91.30 (C10), 74.29 (C40), 72.12 (C30),
72.03 (C50), 68.38 (C20), 53.45, 40.61, 34.02, 33.28, 27.65, 24.17,
15.93 (C60); HRMS (MALDI) m/z for C20H31Cl2N3NaO5 ([M þ
Na]þ) 486.152 33, calcd 486.153 30.
Cell Proliferation Assays. Testing was performed by the
Keck-UWCCC Small Molecule Screening Facility (Madison,
WI). General carcinoma cell line maintenance, compound
handling, and assay protocols have been previously reported.20
Briefly, cells were plated in 50 μL volumes at a density of 500
cells per well in 384-well clear-bottom tissue culture plates.
Serial dilutions of 30 mM DMSO compound stock solutions
were accomplished in 96-well plates using a BioTek Precision XS
liquid handler (Winooski, VT) to a concentration 100ꢀ greater
than that of the most dilute assay. Final dilutions were per-
formed in a 384-well plate in quadruplicate using a Beckman-
Coulter Biomek FX liquid handler with a 384-channel pipetting
head (Fullerton, CA) and were stored at -20 °C when not in use.
Compounds were then added to the culture plates by the Biomek
FX handler and were incubated at 37 °C for 7 days in an
atmosphere containing 5% CO2. The calcein AM reagent
(acetoxymethyl ester, 10 μM) and ETHD-1 (100 μM, 30 μL
total) were added, the treated cells incubated for 30 m at 37 °C,
and plates read for fluorescent emission using a Tecan Safire2
microplate reader (Duram, NC) at the appropriate wavelengths.
CellTiter-Glo reagent (15 μL; Promega Corp., Madison, WI)
was added, and the plates were incubated for 10 m at room tem-
perature with gentle agitation to lyse the cells. Each plate was
re-examined for luminescence to verify inhibition. GI50 values
for cytotoxicity were determined using XLfit 4.2 as previously
reported.20 For compound 60, the most active neoglycoside, the
NCI60 human tumor cancer cell line screen was performed by
the Developmental Therapeutics Program of the National
Cancer Institute (Rockville, MD) as previously described.61,62
4-(4-N0,N0-Bis(2-chloroethyl)amino)phenylbutylhydrazido-
D-glucurono-6,3-lactonide (73). By use of aglycon 71 (61 mg,
0.19 mmol), the mixture of compounds was yielded as a colorless
syrup (42 mg, 44%, RfA = 0.55, RfB = 0.47, MeOH/CH2Cl2
10:90, R/β/imine = 0:1:1). 1H NMR (CD3OD, 500 MHz) δ 7.57
(d, J = 4.3 Hz, 0.35 H, imine A-H10), 7.41 (d, J = 4.1 Hz, 0.15 H,
imine B-H10), 7.09 (d, J = 8.6 Hz, 2 H), 6.70 (d, J = 8.6 Hz, 2
H), 4.91 (dd, J = 6.4, 4.7 Hz, 0.5 H, β-H40), 4.85-4.79 (m, 0.5 H,
β-H30), 4.78 (d, J = 1.4 Hz, 0.5 H, β-H10), 4.68 (dd, J = 7.6, 4.3
Hz, 0.35 H, imine A-H20), 4.63 (dd, J = 7.8, 4.1 Hz, 0.15 H,
imine B-H20), 4.57-4.54 (m, 1 H, imine-H40/H50 þ β-H20),
4.53-4.50 (m, 1.35 H, imine A-H30 þ imine-H40/H50 þ β-
H50), 4.46-4.44 (m, 0.15 H, imine B-H30), 3.75-3.71 (m, 4 H),
3.68-3.65 (m, 4 H), 2.71-2.54 (m, 2 H), 2.27 (t, J = 7.5 Hz,
1 H), 2.20-2.15 (m, 1 H), 1.96-1.87 (m, 2 H); 13C NMR
(CD3OD, 125 MHz) δ 176.71 (imine-C60), 176.08 (β-C60),
171.83, 148.17 (imine A-C10), 144.93, 144.85 (imine B-C10),
130.37, 129.50, 112.43, 99.79 (β-C10), 84.14 (β-C30), 81.62
(imine B-C30), 81.53 (imine A-C30), 77.83 (β-C40), 70.96 (imine
B-C40/C50), 70.88 (β-C20), 70.11 (imine A-C40/C50), 69.97 (imine
B-C40/C50), 69.84 (imine A-C40/C50), 69.49 (β-C50), 69.20 (imine
A-C20), 69.59 (imine B-C20), 53.40, 40.62, 33.99, 33.58 (β-C2),
33.47 (imine-C2), 27.36; HRMS (MALDI) m/z for C20H27Cl2-
N3NaO6 ([M þ Na]þ) 498.116 31, calcd 498.116 91.
4-(4-N0,N0-Bis(2-chloroethyl)amino)phenylbutylhydrazido-
D-threosideimine (74). By use of aglycon 71 (36 mg, 0.11 mmol),
the neoglycoside imine was yielded as a colorless syrup (41 mg,
Acknowledgment. This work was supported by funding
from the NIH (Grants U19 CA113297 and AI52218) and the
Laura and Edward Kremers Chair in Natural Products
Chemistry (J.S.T.). Analytical support was provided by the
School of Pharmacy Analytical Instrumentation Center, and
cancer cell line cytotoxicity screening was provided by the
Keck UWCCC Small Molecule Screening Facility.
1
85%, Rf = 0.19, MeOH/CH2Cl2 10:90). H NMR (CD3OD,
500 MHz) δ 7.53 (d, J = 4.9 Hz, 0.6 H, imine A-H10), 7.37 (d,
J = 5.5 Hz, 0.4 H, imine B-H10), 7.08 (d, J = 8.6 Hz, 2 H), 6.70
(d, J = 8.6 Hz, 2 H), 4.32 (t, J = 4.4 Hz, 0.6 H, imine A-H20),
4.25 (m, 0.4 H, imine B-H20), 3.78-3.70 (m, 5 H, imine-H30),
3.69-3.60 (m, 6 H, imine-H40), 2.65 (t, J = 7.5 Hz, 0.8 H, imine
B-H2), 2.58 (t, J = 7.5 Hz, 1.2 H, imine A-H2), 2.26 (t, J =
7.5 Hz, 1.2 H, imine A-H4), 1.98-1.90 (m, 2.8 H, imine-H3 þ
imine B-H3); 13C NMR (CD3OD, 125 MHz) δ 171.67, 151.04
(imine A-C10), 147.62 (imine B-C10), 144.92, 130.40, 129.49,
112.44, 73.65 (imine B-C30), 73.46 (imine A-C30), 71.49 (imine
B-C20), 71.40 (imine A-C20), 62.60 (imine-C40), 53.42, 40.59,
34.19 (imine B-C2), 34.00 (imine A-C2), 33.56 (imine A-C4),
31.76 (imine B-C4), 27.41 (imine A-C3), 26.82 (imine B-C3);
HRMS (MALDI) m/z for C18H27Cl2N3NaO4 ([M þ Na]þ)
442.127 21, calcd 442.127 08.
Supporting Information Available: Experimental procedures
and characterization data for aglycons 9, 65, and 71; character-
ization, structures, and growth inhibition data for the neoglyco-
side library; NCI60 anticancer screen data for 60; and NMR
spectra. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
ꢀ
ꢀ
(1) Kren, V.; Rezanka, T. Sweet antibiotics. The role of glycosidic
residues in antibiotic and antitumor activity and their randomiza-
tion. FEMS Microbiol. Rev. 2008, 32, 858–889.
4-(4-N0,N0-Bis(2-chloroethyl)amino)phenylbutylhydrazido-
D-xyloside (75). By use of aglycon 71 (60 mg, 0.19 mmol), the
mixture of compounds was yielded as a colorless syrup (40 mg,
45%, RfA = 0.36, RfB = 0.29, MeOH/CH2Cl2 10:90, R/β/imine
= 0:2.3:1). 1H NMR (CD3OD, 500 MHz) δ 7.54 (d, J = 5.0 Hz,
0.15 H, imine A-H10), 7.38 (d, J = 5.5 Hz, 0.15 H, imine B-H10),
7.09 (d, J = 8.6 Hz, 2 H), 6.70 (d, J = 8.6 Hz, 2 H), 4.45 (d, J =
3.8 Hz, 0.15 H, imine A-H20), 4.40 (t, J = 4.9 Hz, 0.15 H, imine
B-H20), 3.89 (dd, J = 11.3, 5.4 Hz, 0.7 H, β-H5A0), 3.85 (d, J =
8.7 Hz, 0.7 H, β-H10), 3.75-3.72 (m, 4.3 H, imine B-H30 þ imine
A/B-H40), 3.68-3.65 (m, 4.75 H, imine A/B-H40 þ imine-H50),
3.59 (dd, J = 7.8, 4.0 Hz, 0.15 H, imine A-H30), 3.52-3.45 (m,
0.7 H, β-H40), 3.39-3.36 (m, 0.7 H, β-H30), 3.23-3.16 (m, 1.4 H,
(2) Thorson, J. S.; Hosted, T. J., Jr.; Jiang, J.; Biggins, J. B.; Ahlert, J.
Nature’s carbohydrate chemists: the enzymatic glycosylation of
bioactive bacterial metabolites. Curr. Org. Chem. 2001, 5, 139–167.
(3) Weymouth-Wilson, A. C. The role of carbohydrates in biologically
active natural products. Nat. Prod. Rep. 1997, 14, 99–110.
ꢁ
(4) Salas, J. A.; Mendez, C. Engineering the glycosylation of natural
products in actinomycetes. Trends Microbiol. 2007, 15, 119–232.
(5) Thibodeaux, C. J.; Melanc-on, C. E.; Liu, H.-W. Unusual sugar
biosynthesis and natural product glycodiversification. Nature
2007, 446, 1008–1016.
(6) Blanchard, S.; Thorson, J. S. Enzymatic tools for engineering natural
product glycosylation. Curr. Opin. Chem. Biol. 2006, 10, 263–271.
(7) Griffith, B. R.; Langenhan, J. M.; Thorson, J. S. “Sweetening”
natural products via glycorandomization. Curr. Opin. Biotechnol.
2005, 16, 622–630.
0
β-H20 þ β-H5B ), 2.56 (t, J = 7.5 Hz, 2 H), 2.27-2.24 (m, 0.6 H,
imine-H2), 2.21-2.17 (m, 1.4 H, β-H2), 1.96-1.87 (m, 2 H);

8138 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Goff and Thorson
(8) Langenhan, J. M.; Griffith, B. R.; Thorson, J. S. Neoglycorando-
mization and chemoenzymatic glycorandomization: two comple-
mentary tools for natural product diversification. J. Nat. Prod.
2005, 68, 1696–1711.
van der Jagt, R. H.; Cheson, B. D. Bendamustine is effective
therapy in patients with rituximab-refractory, indolent B-cell
non-hodgkin lymphoma: results from a multicenter study. Cancer
2010, 116, 106–114.
(32) Millard, J. T.; Raucher, S.; Hopkins, P. B. Mechlorethamine cross-
links deoxyguanosine residues at 50-GNC sequences in duplex
DNA fragments. J. Am. Chem. Soc. 1990, 112, 2459–2460.
(33) Denny, W. A. DNA minor groove alkylating agents. Curr. Med.
Chem. 2001, 8, 533–544.
(34) Bank, B. B.; Kanganis, D.; Liebes, L. F.; Silber, R. Chlorambucil
pharmacokinetics and DNA binding in chronic lymphocytic
leukemia lymphocytes. Cancer Res. 1989, 49, 554–559.
(35) Dorr, R. T.; Fritz, W. L. Cancer Chemotherapy Handbook; Elsevier
Science: New York, 1980.
(36) Sanderson, B. J. S.; Johnson, K. J.; Henner, W. D. Induction of
mutant lymphocytes in cyclophosphamide- and chlorambucil-
treated patients. Mutagenesis 2001, 16, 197–202.
(37) Palayoor, S. T.; Mitchell, J. B.; Cerna, D.; Degraff, W.
John-Aryankalayil, M.; Coleman, C. N. PX-478, an inhibitor of
hypoxia-inducible factor-1R, enhances radiosensitivity of prostate
carcinoma cells. Int. J. Cancer 2008, 123, 2430–2437.
(38) Veyhl, M.; Wagner, K.; Volk, C.; Gorboulev, V.; Baumgarten, K.;
Weber, W.-M.; Schaper, M.; Bertam, B.; Wiessler, M.; Koepsell,
H. Transport of the new chemotherapeutic agent β-D-glucosyliso-
phosphþoramide mustard (D-19575) into tumor cells is mediated by
the Na -D-glucose cotransporter SAAT1. Proc. Natl. Acad. Sci.
U.S.A. 1998, 95, 2914–2919.
(39) (a) Schein, P. S.; Green, D.; Hammer, C. F.; McPherson, E.; Talebian,
A. 6-Bis-(2-chloroethyl)amino-6-deoxy-D-galactopyranose hydro-
chloride: synthesis, chemical characterization, murine P388 anti-
tumor activity, and bone marrow toxicity. J. Pharm. Sci. 1989, 78,
918–921.
(40) Atassi, G.; Dumont, P.; Gosse, C.; Fournier, J. P.; Gouyette, A.;
Roger, P. Characterization of the antitumor activity against solid
tumors of a new nitrosoureido sugar: Cy 233. Cancer Chemother.
Pharmacol. 1989, 25, 205–209.
(41) Sosnovsky, G.; Rao, N. U. M. In the search for new anticancer
drugs. XXIII: exploration of a predictive design for anticancer
drugs of carbohydrates containing N-nitrosochloroethylamino,
N-nitrosomethyl, and N-nitrosoaminoxyl components. J. Pharm.
Sci. 1991, 80, 693–699.
(42) Halmos, T.; Santarromana, M.; Antonakis, K.; Scherman, D.
Synthesis of glucose-chlorambucil derivatives and their recognition
by the human GLUT1 glucose transporter. Eur. J. Pharmacol.
1996, 318, 477–484.
(43) Iglesias-Guerra, F.; Candela, J. I.; Bautista, J.; Alcudia, F.;
Vega-Perez, J. M. Potential anticancer drugs, part 2. Alkylating
agents from sugars. Alkyl hexopyranoside derivatives as carrier
systems for chlorambucil. Carbohydr. Res. 1999, 316, 71–84.
(44) Vega-Perez, J. M.; Candela, J. I.; Blanco, E.; Iglesias-Guerra, F.
Potential anticancer drugs. Part 3. Alkylating agents from sugars.
Stereoselective synthesis of 2,3-diaminoglucoses from 2-nitroalkenes,
as intermediates in the synthesis of carriers of chlorambucil.
Tetrahedron 1999, 55, 9641–9650.
(45) Iglesias-Guerra, F.; Candela, J. I.; Blanco, E.; Alcudia, F.;
Vega-Perez, J. M. Alkylating agents from sugars: synthesis of
chlorambucil derivatives carried by chiral glycosyl glycerols derived
from D-glucosamine. Chirality 2002, 14, 199–203.
(46) Reux, B.; Weber, V.; Galmier, M.-J.; Borel, M.; Madesclaire,
M.; Madelmont, J.-C.; Debiton, E.; Coudert, P. Synthesis and
cytotoxic properties of new fluorodeoxyglucose-coupled chloram-
bucil derivatives. Bioorg. Med. Chem. 2008, 16, 5004–5020.
(47) Thorens, B.; Mueckler, M. Glucose transporters in the 21st century.
Am. J. Physiol.: Endocrinol. Metab. 2010, 298, 141–145.
(48) Zhao, F.-Q.; Keating, A. F. Functional properties and genomics of
glucose transporters. Curr. Genomics 2007, 8, 113–128.
(49) Macheda, M. L.; Rogers, S.; Best, J. D. Molecular and cellular
regulation of glucose transporter (GLUT) proteins in cancer.
J. Cell. Physiol. 2005, 202, 654–662.
(9) Peri, F.; Dumy, P.; Mutter, M. Chemo- and stereoselective glyco-
sylation of hydroxylamino derivatives: a versatile approach to
glycoconjugates. Tetrahedron 1998, 54, 12269–12278.
(10) Peri, F.; Deutman, A.; La Ferla, B.; Nicotra, F. Solution and solid-
phase chemoselective synthesis of (1-6)-amino(methoxy) di- and
trisaccharide analogs. Chem. Commun. 2002, 1504–1505.
(11) Carrasco, M. R.; Nguyen, M. J.; Burnell, D. R.; MacLaren, M. D.;
Hengel, S. M. Synthesis of neoglycopeptides by chemoselec-
tive reaction of carbohydrates with peptides containing a novel
N0-methyl-aminooxy amino acid. Tetrahedron Lett. 2002, 43, 5727–
5729.
(12) Carrasco, M. R.; Brown, R. T. A versatile set of aminooxy amino
acids for the synthesis of neoglycopeptides. J. Org. Chem. 2003, 68,
8853–8858.
(13) Carrasco, M. R.; Brown, R. T.; Serafimova, I. M.; Silva, O.
Synthesis of N-Fmoc-O-(N0-Boc-N0-methyl)-aminohomoserine,
an amino acid for the facile preparation of neoglycopeptides.
J. Org. Chem. 2003, 68, 195–197.
(14) Peri, F. Extending chemoselective ligation to sugar chemistry:
convergent assembly of bioactive neoglycoconjugates. Mini-Rev.
Med. Chem. 2003, 3, 651–658.
ꢁ
ꢀ
(15) Peri, F.; Jimenez-Barbero, J.; Garcıa-Aparicio, V.; Tvaroska, I.;
´
Nicotra, F. Synthesis and conformational analysis of novel
N(OCH3)-linked disaccharide analogues. Chem.;Eur. J. 2004,
10, 1433–1444.
(16) Peri, F.; Nicotra, F. Chemoselective ligation in glycochemistry.
Chem. Commun. 2004, 623–627.
(17) Langenhan, J. M.; Thorson, J. S. Recent carbohydrate-based
chemoselective ligation applications. Curr. Org. Synth. 2005, 2,
59–81.
(18) Carrasco, M. R.; Brown, R. T.; Doan, V. H.; Kandel, S. M.; Lee,
F. C. 2-(N-Fmoc)-3-(N-Boc-N-methoxy)-diaminopropanoic acid,
an amino acid for the synthesis of mimics of O-linked glycopep-
tides. Biopolymers 2006, 84, 414–420.
(19) Nicotra, F.; Cipolla, L.; Peri, F.; La Ferla, B.; Redaelli, C.
Chemoselective neoglycosylation. Adv. Carbohydr. Chem. Biochem.
2007, 61, 353–398.
(20) Langenhan, J. M.; Peters, N. R.; Guzei, I. A.; Hoffman, M. A.;
Thorson, J. S. Enhancing the anticancer properties of cardiac
glycosides by neoglycorandomization. Proc. Nat. Acad. Sci. U.S.A.
2005, 102, 12305–12310.
(21) Ahmed, A.; Peters, N. R.; Fitzgerald, M. K.; Watson, J. A., Jr.;
Hoffmann, F. M.; Thorson, J. S. Colchicine glycorandomization
influences cytotoxicity and mechanism of action. J. Am. Chem. Soc.
2006, 128, 14224–14225.
(22) Griffith, B. R.; Krepel, C.; Fu., X.; Blanchard, S.; Ahmed, A.;
Edmiston, C. E.; Thorson, J. S. Model for antibiotic optimization
via neoglycosylation: synthesis of liponeoglycopeptides active
against VRE. J. Am. Chem. Soc. 2007, 129, 8150–8155.
(23) Langenhan, J. M.; Engle, J. M.; Slevin, L. K.; Fay, L. R.; Lucker,
R. W.; Smith, K. R.; Endo, M. M. Modifying the glycosidic linkage
in digitoxin analogs provides selective cytotoxins. Bioorg. Med.
Chem. Lett. 2008, 18, 670–673.
(24) Goff, R. D.; Thorson, J. S. Enhancing the divergent activities of
betulinic acid via neoglycosylation. Org. Lett. 2009, 11, 461–464.
(25) Everett, J. L.; Roberts, J. J.; Ross, W. C. J. Aryl-2-haloalkylamines.
XII. Some carboxylic derivatives of N,N-bis(2-chloroethyl)aniline.
J. Chem. Soc. 1953, 2386–2392.
(26) Jaksic, B.; Brugiatelli, M.; Krc, I.; Losonczi, H.; Holowiecki, J.;
Planinc-Peraica, A.; Kusec, R.; Morabito, F.; Iacopino, P.; Lutz,
D. High dose chlorambucil versus Binet’s modified cyclophos-
phamide, doxorubicin, vincristine, and prednisone regimen in the
treatment of patients with advanced B-cell chronic lymphocytic
leukemia. Results of an international multicenter randomized trial.
Cancer 1997, 79, 2107–2114.
(50) Joost, H.-G.; Thorens, B. The extended GLUT-family of sugar/
polyol transport facilitators: nomenclature, sequence characteris-
tics, and potential function of its novel members (review). Mol.
Membr. Biol. 2001, 18, 247–256.
(51) Merino, P.; Franco, S.; Merchan, F. L.; Tejero, T. A facile synthe-
sis of glycosyl hydroxylamines. Synth. Commun. 1997, 27, 3529–
3537.
(52) Dondoni, A.; Giovannini, P. P.; Perrone, D. New synthesis of
pyrrolidine homoazasugars via aminohomologation of furanoses
and their use for the stereoselective synthesis of aza-C-disacchar-
ides. J. Org. Chem. 2002, 67, 7203–7214.
(27) Robak, T.; Kasznicki, M. Alkylating agents and nucleoside ana-
logues in the treatment of B cell chronic lymphocytic leukemia.
Leukemia 2002, 16, 1015–1027.
(28) Nicolle, A.; Proctor, S. J.; Summerfield, G. P. High dose chlor-
ambucil in the treatment of lymphoid malignancies. Leuk. Lymphoma
2004, 45, 271–275.
(29) Robak, T. Recent progress in the management of chronic lympho-
cytic leukemia. Cancer Treat. Rev. 2007, 33, 710–728.
(30) Robak, T.; Jamroziak, K.; Robak, P. Current and emerging
treatments for chronic lymphocytic leukemia. Drugs 2009, 69,
2415–2449.
(31) Kahl, B. S.; Bartlett, N. L.; Leonard, J. P.; Chen, L.; Ganjoo, K.;
(53) Cicchi, S.; Marradi, M.; Corsi, M.; Faggi, C.; Goti, A. Chiral
nitrones from lactols, 1. Preparation of N-glycosylhydroxylamines
Williams, M. E.; Czuczman, M. S.; Robinson, K. S.; Joyce, R.;

Article
and their oxidation to nitrones for the enantioselective synthesis of
isoxazolidines. Eur. J. Org. Chem. 2003, 4152–4160.
(54) Bendiak, B. Preparation, conformation, and mild hydrolysis of
1-glycosyl-2-acetylhydrazines of the hexoses, pentoses, 2-acetamido-
2-deoxyhexoses, and fucose. Carbohydr. Res. 1997, 304, 85–90.
(55) Leteux, C.; Childs, R. A.; Chai, W.; Stoll, M. S.; Kogelberg, H.;
Feizi, T. Biotinyl-L-3-(2-naphthyl)-alanine hydrazide derivatives of
N-glycans: versatile solid-phase probes for carbohydrate-recognition
studies. Glycobiology 1998, 8, 227–236.
(56) Auge, J.; Lubin-Germain, N. Glycosylhydrazides, a new class
of sugar surfactants. Preparation and amphiphilic properties
of 1-glycosyl-2-acylhydrazines. J. Carbohydr. Chem. 2000, 19, 379–
392.
(57) Peluso, S.; Ufret, M. L.; O’Reilly, M. K.; Imperiali, B. Neoglyco-
peptides as inhibitors of oligosaccharyl transferase: insight into
negotiating product inhibition. Chem. Biol. 2002, 9, 1323–1328.
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8139
(58) Guillaumie, F.; Thomas, O. R. T.; Jensen, K. J. Immobilization of
pectin fragments on solid supports: novel coupling by thiazolidine
formation. Bioconjugate Chem. 2002, 13, 285–294.
(59) Flinn, N. S.; Quibell, M.; Monk, T. P.; Ramjee, M. K.; Urch, C. J.
A single-step method for the production of sugar hydrazides:
intermediates for the chemoselective preparation of glycoconjugates.
Bioconjugate Chem. 2005, 16, 722–728.
(60) Grigor’ev, I. A. Nitrones: Novel Strategies in Synthesis. In Nitrile
Oxides, Nitrones, and Nitronates in Organic Synthesis, 2nd ed.;
Feuer, H., Ed.; John Wiley & Sons, Inc.: Hoboken, NJ, 2008;
pp 129-434.
(61) Developmental Therapeutics Program NCI/NIH Home Page.
http://dtp.nci.nih.gov/branches/btb/ivclsp.html (accessed May 4,
2010).
(62) Shoemaker, R. H. The NCI60 human tumour cell line anticancer
drug screen. Nat. Rev. Cancer 2006, 6, 813–823.