Synthetic 1-deoxynojirimycin N-substituted peptides offer prolonged disruption to N-linked glycan processing

Article abstract of DOI:10.1002/cmdc.201402186

A panel of 1-deoxynojirimycin (DNJ) N-linked peptides were synthesized. Their IC₅₀ values were measured in vitro against a-glucosidases I and II and were found to be in the micromolar range for both isozymes, and better than that of the iminosugar NB-DNJ (miglustat, 3) against a-glucosidase II. Cell-based studies revealed that although the free iminosugar 3 is most effective at disrupting N-linked glycan processing for shortterm incubations (one day), when the cell-based studies were extended to three days, the DNJ N-linked tetrapeptide KDEL, which is an endoplasmic reticulum (ER)-retaining sequence, performed far better than 3. In low inhibitor washout studies, NB-DNJ inhibition was decreased to zero after 24 h, but DNJ- KDEL retained 13% activity. This method offers a general approach for targeting drugs to the ER and prolonging their activity. Moreover, it is modular, so as new iminosugars of increased potency are discovered, they can be added to this template for targeting.

Full text of DOI:10.1002/cmdc.201402186

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201402186
Synthetic 1-Deoxynojirimycin N-Substituted Peptides Offer
Prolonged Disruption to N-Linked Glycan Processing
Aimꢀ Lꢁpez Aguilar,[a] Jaime Escribano,[a] Paul Wentworth, Jr.,*[a, b] and Terry D. Butters*[a]
A panel of 1-deoxynojirimycin (DNJ) N-linked peptides were
synthesized. Their IC50 values were measured in vitro against
a-glucosidases I and II and were found to be in the micromolar
range for both isozymes, and better than that of the iminosu-
gar NB-DNJ (miglustat, 3) against a-glucosidase II. Cell-based
studies revealed that although the free iminosugar 3 is most
effective at disrupting N-linked glycan processing for short-
term incubations (one day), when the cell-based studies were
extended to three days, the DNJ N-linked tetrapeptide KDEL,
which is an endoplasmic reticulum (ER)-retaining sequence,
performed far better than 3. In low inhibitor washout studies,
NB-DNJ inhibition was decreased to zero after 24 h, but DNJ–
KDEL retained 13% activity. This method offers a general ap-
proach for targeting drugs to the ER and prolonging their ac-
tivity. Moreover, it is modular, so as new iminosugars of in-
creased potency are discovered, they can be added to this
template for targeting.
Introduction
Iminosugars are polyhydroxylated alkaloid small molecules and
are structural mimics of monosaccharides or their hydrolysis
transition state, wherein the endocyclic oxygen atom is substi-
tuted by nitrogen.[1] 1-Deoxynojirimycin (DNJ, 1a) is a piperi-
dine 1-deoxyglucose mimic, naturally occurring in mulberry
plants, Streptomyces and Bacillus spp.,[2] and was the starting
point for the development of this class of compounds as thera-
peutic agents (Figure 1). There are two marketed first-genera-
disease[4] and Niemann–Pick type C (NPC) disease,[5] both lyso-
somal storage diseases (LSDs). Its proposed mechanism of
action is through the inhibition of glucosylceramide synthase,
a glycosyl transferase responsible for the synthesis of many
glycosphingolipids. Patients of type I Gaucher disease have
a dysfunctional glucocerebrosidase enzyme, leading to the ac-
cumulation of intracellular glucosylceramide. Decreasing the
activity of glucosylceramide synthase by 3 lowers the buildup
of glucosylceramide (a process called substrate reduction ther-
apy).[6]
The first-generation iminosugars have also shown potential
as antivirals. Most antiviral drugs target a gene product unique
to a specific pathogen, but this strategy is ultimately vulnera-
ble to the generation of drug-resistant strains. An approach to
circumvent the problem of resistance is to target a cellular pro-
cess of the host; one pathway that has been used is N-linked
glycan processing.[7] Disruption of N-linked glycan processing
in the endoplasmic reticulum (ER) by iminosugar-mediated glu-
cosidase inhibition interrupts virus-sensitive protein-folding
mechanisms, leading to decreased viral infectivity.[8] a-Glucosi-
dase inhibition has been shown to be a viable route for reduc-
ing the titer of HIV in vitro.[9] The antiviral effects of DNJ deriva-
tives have also been studied against hepatitis viruses. For ex-
ample, the use of NN-DNJ (N-nonyl-DNJ) against woodchuck
hepatitis virus showed a dose-dependent decrease in secreted
virus.[10] NB-DNJ (3) also showed antiviral effects against bovine
viral diarrhea virus, the hepatitis C model virus.[11]
Figure 1. Marketed drugs derived from 1-deoxynojirimycin 1a.
tion iminosugar drugs based on the DNJ scaffold: N-2HE-DNJ
(N-2-hydroxethyl-DNJ, miglitol, Glyset, 1b) and NB-DNJ (N-
butyl-DNJ, miglustat, Zavesca, 3).[3] Glyset has been licensed for
use in the treatment of type II diabetes since 1996, with its
proposed mode of action being to decrease glucose absorp-
tion from the gut through inhibition of intestinal a-1,4-glucosi-
dase. Zavesca is licensed for the treatment of type I Gaucher
[a] Dr. A. L. Aguilar, Dr. J. Escribano, Prof. P. Wentworth, Jr., Dr. T. D. Butters
Oxford Glycobiology Institute, Department of Biochemistry
University of Oxford, South Parks Road, Oxford OX1 3QU (UK)
E-mail: terry.butters@bt.internet.com
NB-DNJ is a known inhibitor of both a-glucosidase I (GluI)
and II (GluII),[12] and inhibition of these ER-localized enzymes
prevents newly formed glycoproteins from entering the calnex-
in/calreticulin cycle of protein folding by blocking the degluco-
sylation of glycoproteins.[13] It is this property of NB-DNJ that
has prompted its study as an antiviral agent.[14]
[b] Prof. P. Wentworth, Jr.
Department of Chemistry, The Scripps Research Institute
3030 Science Park Road, La Jolla, CA 92037 (USA)
E-mail: paulw@scripps.edu
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201402186.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 6
&
1
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
Iminosugars are thought to cross the plasma mem-
brane easily, such that their cytoplasmic concentra-
tion is equivalent to that in the extracellular matrix.[15]
In the cytosol, iminosugars interact directly with the
ceramide-specific glucosyltransferase on the cytosolic
side of the cis-Golgi, inhibiting glycolipid biosynthe-
sis. However, to inhibit GluI and/or GluII, the iminosu-
gars must gain access to the ER lumen. This is a clear
and unresolved problem for N-alkyl-DNJ iminosugars,
because studies have shown that the concentration
of these compounds required to inhibit ER GluI is
103–104-fold higher than the measured IC50 values of
purified enzyme.[10,16] It is this problem of how to im-
prove the efficacy of DNJ-like iminosugars in their
mode of action of perturbing N-linked glycan proc-
essing in the ER that is the focus of this study. The
process of delivering drugs to their target cells and
to the specific organelle where they are required has
been a long-standing challenge.[17] Cell-penetrating
peptides (CPPs) are seen as a possible solution to the
challenge of crossing the cell membrane.[18] However, Scheme 1. A) Synthesis of NCP-DNJ (4). Reagents and conditions: a) NaOH (0.5m), RT,
12 h, 84%; b) TBS-Cl, imidazole, DMF, RT, 30 min, quant.; c) K2CO3, benzyl bromide, ace-
targeting subcellular organelles, as is required for in-
tone, reflux, 18 h, 70%; d) TBAF; THF, O8C!RT, 3 h, 89%; e) Dess–Martin periodinane,
hibiting GluI and GluII, is even more challenging.
CH2Cl2, RT, 3 h, 45%; f) DNJ (1a), MeOH/AcOH (100:1), NaCNBH3 on solid support, RT, 3 h,
However, it has been known for some time that
57%; g) Pd/C, H2, MeOH, RT, 18 h, quant. B) Synthesis of 2a–d. Reagents and conditions:
newly synthesized proteins are tagged with short a) 1. DIC (4.4 equiv), HOBt (4.4 equiv), RT, 1.5 h, 2. TFA/TIS/H2O (97:2:1 v/v/v).
peptide sequences that direct them to a specific loca-
tion within the cell.[19] This tagging has recently been
applied to facilitate drug delivery to specific cellular organelles,
in particular the nucleus.[20]
catalytic hydrogenation (10% Pd/C) yielded NCP-DNJ (4),
which was then coupled with the requisite peptides that had
been prepared on solid-phase peptide synthesis (SPPS) resin
(Scheme 1B).
We synthesized and studied a number of structures that
comprise the iminosugar DNJ, N-linked through a common 6-
carbon spacer to peptide head groups that possess well-char-
acterized targeting functions. These head groups are either the
ER-retaining tetrapeptides (KDEL and KKAA,[21] in conjugates
2a,b), or the CPPs MAP and TAT[22] (in 2c,d; Scheme 1).
All peptides for this study were synthesized using 9-fluore-
nylmethyloxycarbonyl (Fmoc) SPPS on Wang resin. In brief,
after Fmoc deprotection with piperidine, each coupling was
achieved with the corresponding amino acid (AA, 4 equiv), dii-
sopropyl carbodiimide (DIC, 4.4 equiv) and 1-hydroxybenzotria-
zole (HOBt, 4.4 equiv) with shaking at room temperature for
1.5 h, followed by washing. Iminosugar-containing carboxylate
4 was coupled directly onto resins 19a–d under the same con-
ditions used for AA couplings. Cleavage with a TFA/TIS/H2O
cocktail (97:2:1% v/v/v) followed by cold diethyl ether precipi-
tation and purification by semi-preparative RP-HPLC gave pure
DNJ-KDEL (2a), DNJ-KKAA (2b), DNJ-MAP (2c), and DNJ-TAT
(2d) on a multi-milligram scale.
Results and Discussion
Synthesis of DNJ (1a) and DNJ–peptide conjugates 2a–d
DNJ (1a) was prepared on the gram scale by using a modifica-
tion of the synthesis reported by Fleet and co-workers[23] from
d-glucoronolactone (5) (discussed in the Supporting Informa-
tion). A facile synthesis of NCP-DNJ (4), the common iminosu-
gar-linker precursor to 2a–d, started with a base-catalyzed hy-
drolysis of e-caprolactone (12, Scheme 1A), followed by a step-
wise tert-butyldimethylsilyl (TBDMS) protection of the C6 hy-
droxy group and benzyl ester protection of the C1 carboxylate
to generate the stable intermediate 15 (70% overall yield). De-
silylation of 15, followed by Dess–Martin periodinane oxida-
tion, generated an unstable aldehyde 17, which was used im-
mediately after silica gel purification in a reductive amination
reaction with 1a. A number of conditions were studied to opti-
mize the yield of 18 (Supporting Information); the best results
were obtained with sodium cyanoborohydride on solid sup-
port with 3 ꢃ molecular sieves in a methanol/acetic acid mix-
ture at room temperature for 3 h. De-esterification of 18 by
Enzyme inhibition and cell-based studies
Primary enzyme studies involved inhibition of GluI-mediated
hydrolysis of Glc3Man7GlcNAc2 and GluII-mediated hydrolysis of
Glc1Man7GlcNAc2. The substrate glycans were labeled with 2-
anthranilic acid (2-AA) for UV-based detection as described pre-
viously by our research group.[24] Thus, the purified enzymes
were incubated with their substrate and a range of concentra-
tions of 2a–d and NCP-DNJ (4). The reaction products were
separated by normal phase (NP)-HPLC, and IC50 values were
determined. These studies revealed DNJ-MAP (2c) to be a poor
inhibitor of both enzymes (IC50 >100 mm), and so 2c was not
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 6
&
2
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
studied further. However, the remaining DNJ N-substituted
peptides 2a, (GluI IC50 17.5ꢀ1.3 mm, GluII IC50 23.5ꢀ1.2 mm),
2b (GluI IC50 34.6ꢀ1.5 mm, GluII IC50 8.2ꢀ1.4 mm), 2d (GluI IC50
23.3ꢀ1.1 mm, GluII IC50 42.1ꢀ1.1 mm), and NCP-DNJ 4 (GluI IC50
23.9ꢀ1.2 mm, GluII IC50 15.2ꢀ1.1 mm), although found to be
poorer inhibitors than the iminosugar NB-DNJ (3; GluI IC50
0.59 mm), were all similar inhibitors of GluII (IC50 53 mm) (data
for 3 taken from Ref. [24]).
Under normal conditions, there is little accumulation of glu-
cosylated free oligosaccharides (FOS) in the cytosol.[24] Howev-
er, following the addition of an a-glucosidase inhibitor, the
levels of glucosylated species increase significantly because
the misfolded glycoproteins in the ER are prevented from en-
tering the calnexin/calreticulin cycle, and instead are targeted
for ER-associated degradation (ERAD) involving glycan removal
and protein degradation.[13,15] FOS can therefore be used to
monitor inhibition of ER-localized GluI and GluII, the biological
functions of which are to direct newly synthesized proteins to
the ER/Golgi system.[19]
To assess intracellular glucosidase inhibition, CHO-K1 cells
were incubated with an iminosugar or iminosugar–peptide
construct, either 3, 4, or 2a,b,d at concentrations of either 10
or 100 mm for 24 h. These concentrations and incubation time
were selected based on our experience of the effective con-
centrations of NB-DNJ required to impact FOS, as reported pre-
viously.[24] Cell lysates were then analyzed for cytosolic FOS de-
rived from N-glycosylated glycoproteins with the 2-AA labeling
protocol described previously.[24] The appearance of mono- and
triglucosylated species by NP-HPLC correlates with inhibition
of GluI and GluII, respectively, in the ER lumen. For the N-sub-
stituted peptides 2a, 2b, 2d, and NCP-DNJ 4, no triglucosylat-
ed species were detected under these conditions, but mono-
glucosylated glycans did accumulate. For this study
Glc1Man5GlcNAc1 and Glc1Man4GlcNAc1 were selected to track
glucosidase inhibition, as their changes were easily observed,
and HPLC resolves their peaks with baseline separation (tR ~
26.8 and 31.2 min, respectively;[24] Figure 2A). Man3GlcNAc1
(tR ~21.0 min) was used as an internal standard (IS) because it
is a lysosomal degradation product that remains unaffected by
treatment with a-glucosidase inhibitors, as we previously re-
ported.[24,25]
Figure 2. FOS analysis of CHO-K1 cells. A) Representative NP-HPLC trace for
cell extracts in the a) absence or b) presence of NB-DNJ (3), c) DNJ-KKAA
(2b), or d) DNJ-KDEL (2a) (all at 500 mm) after 24 h incubation. B) Analysis of
the FOS Glc1Man5GlcNAc1, tR ~26.8 min vs. iminosugar inhibitor. C) Analysis
of the FOS Glc1Man4GlcNAc1, tR ~31.2 min vs. iminosugar inhibitor. Values are
normalized to Man3GlcNAc1 tR ~21.0 min.
PAR at 100 mm: 1.278 and 1.083). Thus, whereas DNJ-KDEL (2a)
is a marginally better inhibitor than NB-DNJ (3) against purified
GluII, the net effect on FOS in cells for short-term incubation
(24 h) showed no enhancement. Importantly, the most effec-
tive of the DNJ N-substituted peptides at elevating the Man4
glycan was DNJ-TAT (2d), suggesting that the problem with
the ER-targeted peptides 2a and 2b may be related in part to
problems of cellular uptake.
While FOS analysis is a useful tool for studying N-linked
glycan processing, its analysis is complicated owing to the net-
work of simultaneous enzymatic reactions that impact glycan
processing. However, what is clear in the inhibitor concentra-
tion range studied is that the N-substituted peptides 2a,b and
d are selective inhibitors of ER GluII, as no triglucosylated spe-
cies arising from inhibition of GluI were detected. In addition,
there appears to be no clear benefit imparted by the targeting
peptides KDEL (in 2a) KKAA (in 2b) and TAT (in 2d) in the 24 h
incubation (Figure 2B,C). The IC50 values of 2a (23.5 mm) and
2d (42.1 mm) are slightly lower than that published for NB-DNJ
(59 mm) against purified GluII.[24] However, the amounts of
Glc1Man5GlcNAc1 and Glc1Man4GlcNAc1 generated by 2a (mean
PAR at 100 mm: 0.467 and 0.445) and 2d (mean PAR at 100 mm:
0.339 and 0.937) during the 24 h incubation with these N-sub-
stituted peptide iminosugars are lower than those of 3 (mean
The design of ER-retaining peptides was intended to both
increase the effective concentration of DNJ in the ER and to
allow it to remain there upon extended treatment. Therefore,
three-day exposures of CHO-K1 cells to DNJ-KDEL (2a), NCP-
DNJ (4), and NB-DNJ (3) were undertaken, and the levels of
glucosylated FOS as markers for glucosidase inhibition were
measured (see above; Figure 3). A small but not statistically
significant increase in Glc1Man4GlcNAc1 production was mea-
sured in the presence of KDEL (conjugate 2a) relative to NB-
DNJ (3) from a one-day to three-day incubation (mean ratio
2a/3 48% day 1; mean ratio 2a/3 65% day 3, at 5 mm inhibitor
concentration; Figure 3A). However, if the inhibitor concentra-
tion was raised to 100 mm 2a and 3, the increase in
Glc1Man4GlcNAc1 production by 2a relative to 3 became signif-
icant (p<0.05, Figure 3A). In parallel, we studied the relative
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 6
&
3
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
At the low inhibitor concentration (5 mm), the level of glycans
produced by 2a and 3 are similarly lower amounts for both in-
hibitors, thus allowing a direct comparison.
We next studied the ability of 2a to persist within the ER.
CHO-K1 cells were incubated for 24 h with 2 and 3 at 5 and
100 mm, after which the media was replaced with drug-free
media, and the cells were incubated for a further 24 h. FOS
analysis was performed at the time of drug removal and at the
end of the study, and values are reported as a percent of the
amount of Glc1Man4GlcNAc1 at the end of the study relative to
that at the time of drug removal (Figure 3D). It is clear that
after 24 h recovery NB-DNJ (3) inhibition is not present at the
low concentration (5 mm) and is <5% at 100 mm. In contrast,
DNJ-KDEL (2a) retains 13% of its original activity at low con-
centration (5 mm) and 20% activity at 100 mm. This is a signifi-
cant benefit of the N-substituted peptide iminosugar DNJ-
KDEL tied to its head group having ER-retaining properties.
Conclusions
Iminosugars are amongst the most attractive class of carbohy-
drate mimics reported to date. In this current era of empty
drug pipelines, they occupy a new and largely unexplored
chemical space much prized by pharmaceutical companies.
The first-generation iminosugars have already demonstrated
their potential to provide drug molecules, with two licensed
compounds and several more in development. However, like
any first-generation class of compounds, there are teething
problems. The first iminosugars, while offering broad research
application, suffer from lack of selectivity and inadequate po-
tency which has lead to significant side effects in the clinic and
suspension of clinical trials. Second-generation iminosugars are
currently being developed, both in the academic research en-
vironment and pharma. Improved chemical access to these
complex structures, coupled with increasing understanding of
glycobiology and the role it can play in revealing new drugga-
ble targets, means that iminosugars are strategically placed to
impact health care across a range of disease classes for the
foreseeable future.
Figure 3. FOS analysis for long-term treatment of CHO-K1 cells with ER-re-
&
taining peptide constructs DNJ-KDEL (2a) after one-day ( ) and three-day
&
( ) treatment. A) Ratio of Glc1Man4GlcNAc1 generated by 2a and 3 as a per-
centage at either 5 or 100 mm. B) Ratio of Glc1Man4GlcNAc1 generated by 2a
and 4 as a percentage at 5 mm. C) Ratio of Glc1Man5GlcNAc1 generated by
2a and 3 as a percentage at either 5 or 100 mm. D) Percentage 2-AA labeled
Glc1Man4GlcNAc1 after 24 h recovery in drug-free media from values ob-
tained after 24 h drug treatment. Significance was determined by Student’s
t-test: ns p>0.05; *p<0.05, **p<0.01, ***p<0.001; n=3 using GraphPad
Prism 4.0.
production of Glc1Man4GlcNAc1 by 2a and 4 over time (Fig-
ure 3B). This analysis revealed a large and significant (p<
0.001) increase in the glycan generated by 2a to 4 throughout
the study (mean difference changed from 75% on day 1 to
250% on day 3; Figure 3B). In contrast to DNJ-KDEL (2a),
longer treatment with NCP-DNJ (4) does not induce higher in-
hibition, pointing to a mechanism more in line with most com-
pounds that cannot be retained in the ER lumen and that are
more readily metabolized and/or cleared by the cell.
The work described herein focuses on just one problem for
imunosugars: getting to a sub-cellular site of action to disrupt
N-linked glycan processing, the ER. Inclusion of ER-retaining
peptide tags and CPPs at flexible loci N-linked and known to
not impact GluI and GluII activity has generated compounds
with GluII selectivity and with IC50 values similar to or better
than that of NB-DNJ for this enzyme. In the cell-based studies,
there was little apparent benefit to the use of ER tags on N-
linked glycan processing. However, if longer incubation periods
were used and drug washout experiments were performed,
the benefit of ER retention to prolonging the N-linked glycan
disruption in the cell relative to NB-DNJ became clear. Thus,
the most promising effect of these N-linked peptide iminosu-
gars is the maintenance of activity following drug removal,
which hints strongly at a retention mechanism that is not ob-
served with non-peptide iminosugar conjugates. One of the
main benefits of this approach is its modularity. Thus, as
second-generation iminosugars are found that are increasingly
A similar long-term comparative study of the production of
the FOS Glc1Man5GlcNAc1 by 2a and 3 was then performed
(Figure 3C). At low inhibitor concentration (5 mm) there was
a large and significant (p<0.001) increase in the ratio of this
Man5 glycan generated by 2a relative to 3 (mean difference
changed from 30% on day 1 to 80% on day 3). Interestingly,
treatment of CHO-K1 cells with 2a and 3 at high inhibitor con-
centration (100 mm) causes an apparent lowering in the ratio
of Man5 glycan generated by 2a relative to 3 through the
study (mean difference changed from 55% on day 1 to 35%
on day 3; Figure 3C). This apparent fall in accumulation can be
explained by considering that Glc1Man5GlcNAc1 is a precursor
to Glc1Man4GlcNAc1, and DNJ-KDEL (2a) produces less net
amounts of total glycans than 3, and it is therefore likely that
most of the Glc1Man5GlcNAc1 generated by 2a at the higher
inhibitor concentration is transformed into Glc1Man4GlcNAc1.
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 6
&
4
&
ÞÞ
These are not the final page numbers!

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
more potent and selective for their target, they can be ap-
pended to these ER-retaining tags to direct them and retain
them at their site of action.
gꢀa (CONACYT Mꢁxico), and J.E. acknowledges the European
Commission for a Marie Curie Intra-European Fellowship (IEF;
PIEG-2009-255019). The Oxford Glycobiology Institute and Prof.
Raymond A. Dwek are thanked for additional support.
Experimental Section
Keywords: cell targeting
· glycosidases · iminosugars ·
CHO-K1 cells were cultured in Dulbecco’s modified Eagle’s medium
(DMEM; with l-glutamate) containing 10% fetal calf serum and
a 1% stock solution of penicillin/streptomycin (Invitrogen). CHO-K1
cells were seeded in six-well plates (2.5ꢄ105 cells per mL in 2 mL)
and allowed to recover overnight. Fresh medium containing DNJ
constructs (in water) or water as control was added after washing
the cells with phosphate-buffered saline (PBS; 2ꢄ1 mL), and the
cell cultures were incubated for 24 h. After this time, the medium
was removed, and the cells were washed with PBS (2ꢄ1 mL)
before being scraped into PBS (2 mL). The cells were then centri-
fuged (2000 rpm for 5 min), the supernatant PBS was decanted,
and the cell pellet was re-suspended in PBS (500 mL). The cells
were manually homogenized with a 2 mL glass Dounce homoge-
nizer, and an aliquot (30 mL) of each sample was separated for pro-
tein concentration determination. The homogenate was desalted
and de-proteinated by passage through a mixed-bed ion-exchange
column [380 mL packed resin of AG50W-X12 (H+ 100–200 mesh)
over 500 mL packed resin of AG3-X4 (OHꢁ, 100–200 mesh), elution
with water (4 mL)] pre-equilibrated with water (4 mL). The eluent
was concentrated in vacuo and re-suspended in water (30 mL). The
FOS were labeled with 2-AA by incubating at 808C for 1 h with la-
beling buffer (45 mg of anthranilic acid and 67.5 mg NaCHBH3 in
1.5 mL of a 4% (w/v) solution of sodium acetate containing 2% (w/
v) boric acid in methanol; 80 mL). The resulting mixture was diluted
with 1 mL 97:3 acetonitrile/water, and the solution was purified by
passing it through a Speed-Amide 2 column, pre-equilibrated with
water and acetonitrile (1 mL each). The column was washed with
acetonitrile/water (95:5, 2 mL) and eluted with water (2ꢄ750 mL).
Tris·HCl buffer (0.5m, pH 7.2, 175 mL) was added to each sample.
The labeled FOS were finally purified by using a concanavalin A
(ConA)–Sepharose 4B column (100 mL packed resin). The column
was pre-equilibrated with water (2ꢄ1 mL) followed by sequential
washes of 1 mm MgCl2, 1 mm CaCl2, 1 mm MnCl2 (1 mL), and
50 mm Tris·HCl buffer, pH 7.2 (2ꢄ1 mL). The sample was added
and allowed to pass through the column before washing with
50 mm Tris·HCl buffer, pH 7.2 (1 mL) and water (1 mL). The ConA-
bound FOS were then eluted with hot (708C) 0.5m methyl a-d-
mannopyranoside in 50 mm Tris·HCl buffer, pH 7.2 (2ꢄ1 mL). Ali-
quots (50 mL) of each 2-AA-labeled FOS sample were diluted with
acetonitrile (50 mL) and analyzed by HPLC (Program A; see Ref. [24]
for details). Glucose units were determined by comparison with
a 2-AA-labeled glucose oligomer ladder (derived from a partial hy-
drolysate of dextran) external standard using Peak Time software
(developed in-house). The peak area of each 2-AA-labeled species
was measured with Waters Empower 2 software (version 6.20).
inhibitors · peptide conjugates
[1] a) “Isofagomine and Noeuromycin Other 1-Azasugars, Iminosugar-Relat-
ed Glycosidase Inhibitors”, O. Lopez, M. Bols in Iminosugars: From Syn-
thesis to Therapeutic Applications (Eds.: P. Compain, O. R. Martin), Wiley,
Chichester, 2007, pp. 130–151; b) B. Winchester, G. W. J. Fleet, Glycobiol-
ogy 1992, 2, 199–210.
[2] N. Asano, R. J. Nash, R. J. Molyneux, G. W. J. Fleet, Tetrahedron: Asymme-
try 2000, 11, 1645–1680.
[3] a) G. Horne, F. X. Wilson, J. Tinsley, D. H. Williams, R. Storer, Drug Discov-
ery Today 2011, 16, 107–118; b) G. Horne, F. X. Wilson, Prog. Med. Chem.
2011, 50, 135–170.
[4] C. Dulsat, N. Mealy, Drugs Future 2009, 34, 147–149.
[5] J. E. Wraith, J. Imrie, Ther. Clin. Risk Manage. 2009, 5, 877–887.
[6] F. M. Platt, T. D. Butters in Lysosomal Storage Diseases (Eds.: J. A. Bar-
ranger, M. A. Cabrera-Salazar), Springer, London, 2007, pp. 153–168.
[7] R. Kornfeld, S. Kornfeld, Annu. Rev. Biochem. 1985, 54, 631–644.
[8] T. D. Butters, Tetrahedron: Asymmetry 2000, 11, 113–124.
[9] R. A. Gruters, J. Neefjes, M. Tersmette, R. E. Y. Goede, A. Tulp, H. G. Huis-
man, F. Miedema, H. L. Ploegh, Nature 1987, 330, 74–77.
[10] T. M. Block, X. Lu, A. S. Mehta, B. Blumberg, B. Tennant, M. Ebling, B.
Korba, D. M. Lansky, G. S. Jacob, R. A. Dwek, Nat. Med. 1998, 4, 610–
614.
[11] N. Zitzmann, A. S. Mehta, S. Carrouꢀe, T. D. Butters, F. M. Platt, J. McCau-
ley, B. Blumberg, R. A. Dwek, T. M. Block, Proc. Natl. Acad. Sci. USA 1999,
96, 11878–11882.
[12] A. B. Hughes, A. J. Rudge, Nat. Prod. Rep. 1994, 11, 135.
[13] C. Durrant, S. E. H. Moore, Biochem. J. 2002, 365, 239–247.
[14] R. A. Dwek, T. D. Butters, F. M. Platt, N. Zitzmann, Nat. Rev. Drug Discov-
ery 2002, 1, 65–75.
[15] H. R. Mellor, D. C. A. Neville, D. J. Harvey, F. M. Platt, R. A. Dwek, T. D. But-
ters, Biochem. J. 2004, 381, 861–866.
[16] B. Saunier, R. D. Kilker, J. S. Tkacz, A. Quaroni, A. Herscovics, J. Biol.
Chem. 1982, 257, 14155–14161.
[17] a) G. Orive, R. M. Hernꢅndez, A. R. Gascꢁn, A. Domꢆnguez-Gil, J. L.
Pedraz, Curr. Opin. Biotechnol. 2003, 14, 659–664; b) T. M. Allen, P. R.
Cullis, Science 2004, 303, 1818–1822.
[18] E. L. Snyder, S. F. Dowdy, Pharm. Res. 2004, 21, 389–393.
[19] S. D. Emr, J. Cell Biol. 1980, 86, 701–711.
[20] S. Misaghi, M. E. Pacold, D. Blom, H. L. Ploegh, G. A. Korbel, Chem. Biol.
2004, 11, 1677–1687.
[21] a) S. Munro, H. R. B. Pelham, Cell 1987, 48, 899–907; b) M. R. Jackson, T.
Nilsson, P. A. Peterson, J. Cell Biol. 1993, 121, 317–333.
[22] a) A. D. Frankel, C. O. Pabo, Cell 1988, 55, 1189–1193; b) L. N. Patel, J. L.
Zaro, W. C. Shen, Pharm. Res. 2007, 24, 1977–1992.
[23] D. Best, C. Wang, A. C. Weymouth-Wilson, R. A. Clarkson, F. X. Wilson,
R. J. Nash, S. Miyauchi, A. Kato, G. W. J. Fleet, Tetrahedron: Asymmetry
2010, 21, 311–319.
[24] D. S. Alonzi, D. C. A. Neville, R. H. Lachmann, R. A. Dwek, T. D. Butters,
Biochem. J. 2008, 409, 571–580.
[25] A. J. Rawlings, H. Lomas, A. W. Pilling, M. J. Lee, D. S. Alonzi, J. S. Roun-
tree, S. F. Jenkinson, G. W. J. Fleet, R. A. Dwek, J. H. Jones, T. D. Butters,
ChemBioChem 2009, 10, 1101–1105.
Acknowledgements
This work was supported by a grant from The Scripps Research
Institute and The Skaggs Institute for Chemical Biology (P.W.);
A.L.A. was funded by the Consejo Nacional de Ciencia y Tecnolo-
Received: May 9, 2014
Published online on && &&, 0000
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 6
&
5
&
ÞÞ
These are not the final page numbers!

FULL PAPERS
A. L. Aguilar, J. Escribano,
A trip to the ER: Peptide iminosugar
conjugates were synthesized, and the
inhibition of endoplasmic reticulum (ER)
a-glucosidase-mediated N-glycan proc-
essing was studied. A deoxynojirimycin
conjugate comprising the ER-retaining
tetrapeptide tag KDEL has significantly
prolonged ER activity relative to the
parent iminosugar. An exciting aspect of
this approach is its modularity; as more
potent drugs are discovered they can
be added to this template.
P. Wentworth, Jr.,* T. D. Butters*
&& – &&
Synthetic 1-Deoxynojirimycin N-
Substituted Peptides Offer Prolonged
Disruption to N-Linked Glycan
Processing
ꢂ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 0000, 00, 1 – 6
&
6
&
ÞÞ
These are not the final page numbers!