Synthesis and characterization of new and potent α-lipoic acid derivatives

Article abstract of DOI:10.1016/j.bmc.2003.11.025

α-Lipoic acid [5-[1,2]-dithiolan-3-yl-pentanoic acid (LA)] is a natural antioxidant and cofactor of several enzymes. It increases the glucose transport activity in skeletal muscles and adipocytes in a non-insulin dependent manner. Therefore, LA is widely used in Type 2 diabetic patients as an oral auxiliary drug. However, large doses of LA (0.8-1.8 gr/day po) are required due to its unfavorable pharmacokinetic parameters. In order to improve these parameters, we synthesized ester and amide LA derivates. Two of these newly synthesized compounds, 5-[1,2]-dithiolan-3-yl-pentanoic acid 3-(5-[1,2]dithiolan-3yl-pentanoylamino)-propyl]-amide (AN-7) and 5-[1,2]-dithiolan-3-yl-pentanoic acid 3-(5-[1,2]-dithiolan-3yl-pentanoyloxy)- propyl ester (AN-8) augmented the rate glucose transport in myotubes in culture in the absence or presence of insulin. Their potency was 12-fold higher than that of the parent compound; their maximal stimulatory effect was 1.5-fold higher than that of LA. When tested in vivo in streptozotocin-diabetic C57/ Black mice, AN-7 (10 mg/kg/day for 2 weeks, sc) reduced blood glucose level by 39% while a higher dose of LA (50 mg/kg/day for 2 weeks, sc) lowered it by 30%. These results indicate that AN-7 is more potent than LA in augmenting glucose transport in skeletal muscles and reducing blood glucose in diabetic animals.

Full text of DOI:10.1016/j.bmc.2003.11.025

Bioorganic & Medicinal Chemistry 12 (2004) 1183–1190
Synthesis and characterization of new and potent ꢀ-lipoic
acid derivatives
a
b
b,
c
Arie Gruzman, Adel Hidmi, Jehoshua Katzhendler, * Abdalla Haj-Yehie
a
and Shlomo Sasson
aDepartment of Pharmacology, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem POBox 12272,
Jerusalem, Israel
Department of Medicinal Chemistry, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem POBox 12272,
Jerusalem, Israel
Department of Pharmaceutics, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem POBox 12272,
Jerusalem, Israel
b
c
Received 21 August 2003; accepted 19 November 2003
Abstract—a-Lipoic acid [5-[1,2]-dithiolan-3-yl-pentanoic acid (LA)] is a natural antioxidant and cofactor of several enzymes. It
increases the glucose transport activity in skeletal muscles and adipocytes in a non-insulin dependent manner. Therefore, LA is
widely used in Type 2 diabetic patients as an oral auxiliary drug. However, large doses of LA (0.8–1.8 gr/day po) are required due
to its unfavorable pharmacokinetic parameters. In order to improve these parameters, we synthesized ester and amide LA derivates.
Two of these newly synthesized compounds, 5-[1,2]-dithiolan-3-yl-pentanoic acid 3-(5-[1,2]dithiolan-3yl-pentanoylamino)-propyl]-
amide (AN-7) and 5-[1,2]-dithiolan-3-yl-pentanoic acid 3-(5-[1,2]-dithiolan-3yl-pentanoyloxy)-propyl ester (AN-8) augmented the
rate glucose transport in myotubes in culture in the absence or presence of insulin. Their potency was 12-fold higher than that of the
parent compound; their maximal stimulatory effect was 1.5-fold higher than that of LA. When tested in vivo in streptozotocin-
diabetic C57/Black mice, AN-7 (10 mg/kg/day for 2 weeks, sc) reduced blood glucose level by 39% while a higher dose of LA (50
mg/kg/day for 2 weeks, sc) lowered it by 30%. These results indicate that AN-7 is more potent than LA in augmenting glucose
transport in skeletal muscles and reducingblood gl ucose in diabetic animals.
#
2003 Elsevier Ltd. All rights reserved.
1. Introduction
plications7 and an impairment of peripheral insulin
sensitivity.
8
,9
Lipoic acid (thiocitic acid) (LA) is a cofactor of mito-
chondrial decarboxylacion enzymatic reactions and is
essential for an adequate ATP production from glucose
LA is an oral food additive and is sold over the coun-
1
0
ter. Jacob et al. has reported that an acute iv infusion
of LA improves insulin-stimulated metabolic clearance
rate of glucose as well as insulin sensitivity in type 2
1
via the citric acid cycle. LA is also a multifunctional
2
antioxidant. Its antioxidant properties are divided into
four categories: (I) scavenger of reactive oxygen species,
1
1
diabetes patients. When tested in vitro, LA increases
the rate of glucose transport in cultured rat L6 myo-
tubes and 3T3-L1 adipocytes and in isolated rat dia-
(
II) regeneration and de novo syntheses of endogenous
antioxidants (i.e., glutathione and vitamin E), (III)
metal ion chelatingactivity, and (IV) repairingoxidative
1
2ꢀ14
phragm and cardiomyocytes.
This effect is blocked
damage in macromolecules.3 Oxidative stress is asso-
ꢀ5
by wortmannin, an inhibitor of phosphoinositol-3-kinase
(PI3-kinase), a key enzyme in the insulin transduction
pathway. However, the concentrations of LA required
to induce this direct stimulatory effects in vitro are con-
siderably high (2.0–2.5 mM). It was calculated that a
single oral dose of 600 mg in man elevates plasma LA
concentration to only 10–25 mM without any consider-
ciated with various diseases leadingto multicellular and
6
organ damages. For instance, in diabetes mellitus oxi-
dative stress is associated with an early onset of com-
Keywords: Antihyperglycemic drugs; Glucose transport; a-Lipoic acid,
Prodrug.
1
5
able antihyperglycemic activity. There are two possi-
*
Correspondingauthor. Tel.: +972-2-675-8693; fax: +972-2-675-
076; e-mail: katzhe@cc.huji.ac.il
7
ble reasons for the marginal effect of an oral dose LA
0
968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.11.025

1184
A. Gruzman et al. / Bioorg. Med. Chem. 12 (2004) 1183–1190
on blood glucose levels and peripheral insulin sensitivity:
a short plasma half-life (30 min) and low bioavailability
None of the PEG-derivatives (AN-5, AN-6 and AN-11)
or the ester derivatives of hexane (AN-30), octane (AN-
31) and decane (AN-32) improved the rate of hexose
transport in L6 myotubes (data not shown). However,
AN-7 and AN-8 were active in vitro. These two com-
pound as well as LA increased the rate of hexose trans-
port in L6 myotubes in a dose- and a time-dependent
manner, as shown in Figures 1–4. In these experiments
the myotubes were pre-conditioned at 5.0 mM glucose
or 23.0 mM glucose for 24 h. Pre-incubation of the
myotubes at 23.0 mM glucose induced down-regulation
1
5
(
30%). Indeed, repeated oral administration of LA in
1
6
man does not elevate its plasma level above 25 mM,
while an iv infusion of LA results in higher plasma
concentration of the drugand improved effects.
2
We have undertaken to synthesize more potent and
effective LA derivatives. Several compounds were syn-
thesized and tested both in vivo and in vitro. Among
those 5-[1,2]-dithiolan-3-yl-pentanoic acid 3-(5-[1,2]dithio-
lan-3-yl-pentanoylamino)-propyl]-amide (AN-7) was
found to be more potent than LA in vitro and in vivo.
2
0
of the hexose transport system, as was shown before.
Maximal stimulatory effects of LA (2.5 mM) on the rate
of hexose transport were 39ꢁ9% (meanꢁSEM, n=3)
and 51ꢁ6% in L6 myotubes maintained at 5.0 and 23.0
mM glucose, respectively (Fig. 1). Higher concentra-
tions of LA were toxic to the myotubes. AN-7 and AN-
8 increased the rate of hexose transport by 92ꢁ11%
and 82ꢁ6.8% in L6 myotubes at 5.0 mM glucose and
by 120ꢁ8.9% and 93ꢁ11.2% at 23 mM. These
maximal effects were obtained with 200 mM of each
compound (Fig. 2). Maximal stimulatory effect of LA
on myotubes exposed to 5.0 and 23.0 mM glucose was
observed 2 h after its addition and thereafter it rapidly
declined (Fig. 3 and 4). In contrast, half maximal and
maximal effects of AN-7 and AN-8 were observed at 27
and 36 h, respectively, after their addition to the myo-
tubes cultures exposed to 5.0 mM (Fig. 3). The corre-
2
. Results and discussion
LA-N-hydroxy soccinimide (NHS) was synthesized
1
7
accordingto Nefkens and Tesser
(Scheme 1). The
bond between the alkyl chain and the LA was either an
ester or amide. Amide derivatives (AN-11), (AN-5) and
(
PEG-diamine, PEG-monoamine or propane-1,3-dia-
AN-7) were synthesized by bindingof LA-NHS to
1
8
mine, respectively (Scheme 1). Ester derivatives were
synthesized by couplingof LA-NHS to dihydroxy-PEG,
propane-1,3-diol, hexane-1,6-diol, octane-1,8-diol and
decane1,10-diol to produce (AN-6), (AN-8), (AN-30),
(
AN-31) and (AN-32), respectively19 (Scheme 1).
Scheme 1.

A. Gruzman et al. / Bioorg. Med. Chem. 12 (2004) 1183–1190
1185
spondingvalues in myotubes exposed to 23.0 mM gl u-
cose were 22 and 36 h (Fig. 4). These data suggest a
slow release of active LA from the prodrugmolecules
followingaccumulation in myotubes. The combined
stimulatory effects of insulin and LA, AN-7 or AN-8
were additive both in L6 myotubes under the normo- or
hyperglycemic conditions (Fig. 5). AN-7 and AN-8 were
not toxic to the myotubes cultures.
on blood glucose levels in streptozotocin-diabetic male
C57/Black mice. AN-7 and LA reduced blood glucose
levels in mildly diabetic mice by 30.2ꢁ5 (meanꢁSEM,
n=8) and 40ꢁ13%, respectively. Lower doses of AN-7
(1 or 5 mg/kg BW/day) had no noticeable effect on
blood glucose level of mildly or severly diabetic mice. A
higher dose of AN-7 (20 mg/kg BW/day) produced
similar glucose lowering effect in mildly diabetic mice as
10 mg/kg/day (data not shown). AN-8 failed to change
Table 1 shows the effects of a 2-week treatment with LA
50 mg/kg BW/day), AN-7 or AN-8 (10 mg/kg BW/day)
blood glucose levels in these mice. Neither LA nor AN-
7 or AN-8 had a significant effect on blood glucose
(
Figure 1. Dose–response of LA effect of on hexose transport in L6 myotubes. L6 myotubes were maintained for 24 h at 5.0 (open symbols) or 23.0
mM glucose (closed symbols). The cells then received increasing concentrations of LA (*) or 10 mL of the vehicle DMSO (&). The standard
3
[
H]dGlc uptake assay was performed 2 h after the addition of the LA to the myocyte cultures as described under. 100% value was taken for control
cc oe lnl t mr o al i. ntained at 5.0 mM glucose (1.21ꢁ0.08 nmol dGlc/mgprotein/min). Mean ꢁSEM, are shown: n=3. *p<0.05, compared to the respective
Figure 2. Dose–response of AN-7 and AN-8 effects on hexose transport in L6 myotubes. L6 myotubes were maintained for 24 h at 5.0 (open sym-
bols) or 23.0 mM glucose (closed symbols). The cells then received increasing concentrations of ~,~: AN-7; &,&: AN-8; and *,*: 10 mL of
vehicle DMSO. After 36 h incubation, the myocyte cultures were taken for the standard [ H]dGlc uptake assay as described. 100% value was taken
3
rf oe sr p ce oc nt i tvr eo lc oc en l tl r mo l a. intained at 5.0 mM glucose (1.68ꢁ0.10 nmol dGlc/mgprotein/min). Mean ꢁSEM, are shown: n=3. *p<0.05, compared to the

1186
A. Gruzman et al. / Bioorg. Med. Chem. 12 (2004) 1183–1190
in man.21 Therefore, Evan and Goldfine have suggested
that effort should be made to maintain a therapeutically
effective levels of LA in the plasma for an extended
period of time that may result in an increased insulin
sensitivity and eventually improvement of glycemic con-
trol in type 2 diabetic patients. To meet these require-
ments we have synthesized prodrugs with improved
2
levels in severely diabetic (Table 1) or in non-diabetic
control mice (data not shown). One week after dis-
continuation of LA and AN-7 treatments, blood glu-
cose levels returned to the same levels of control mice.
Clinical studies suggest that, due to the limited bio-
availability of LA, large oral doses of the compound are
2
1
4
22,23
required for an oral therapy. Some studies indicate
that LA augments insulin sensitivity in Type 2 diabetics.
However, this effect seems limited due to the reduced
bioavailability of LA and its low plasma concentration
lipid solubility that can release LA intracellularly.
Amongall the LA derivatives we synthesized only the
amidopropane derivative (AN-7) was active in both in
Figure 3. Time–course of a-Lipoic acid, AN-7 and AN-8-dependent stimulation of hexose transport in L6 myotubes at 5.0 mM glucose. L6 myo-
tubes were preincubated at 5.0 mM glucose for 24 h, The culture medium was changed to fresh medium with the same glucose concentration in the
presence 10 mL of vehicle DMSO (*), 2.5 mM LA (&), 200 mM AN-7 (^), or 200 mM of AN-8 (~). Followingincubations at the indicated times
3
the myotubes were taken for the standard [ H]dGlc uptake assay as described. 100% value was taken for control cell maintained at 5.0 mM glucose
(1.11ꢁ0.07 nmol dGlc/mgprotein/min). Mean ꢁSEM, are shown: n=3. *p<0.05, compared to the respective control.
Figure 4. Time–course curves of a-lipoic acid, AN-7 and AN-8-dependent stimulation of hexose transport in L6 myotubes at 23.0 mM concen-
tration glucose. L6 myotubes were preincubated at 23.0 mM glucose for 24 h. The culture medium was changed to fresh medium with the same
glucose concentration in the presence DMSO (*), 2.5 mM LA (&), 200 mM AN-7 (^), or 200 mM of AN-8 (~). Followingincubations at the
3
indicated time the myotubes were taken for the standard [ H]dGlc uptake assay as described. 100% value was taken for control cell maintained at
23.0 mM glucose. (0.69ꢁ0.08 nmol dGlc/mgprotein/min). Mean ꢁSEM, are shown: n=3. *p<0.05, compared to the respective control.

A. Gruzman et al. / Bioorg. Med. Chem. 12 (2004) 1183–1190
1187
betic mice is attributed to the impact of the disease on
the entire glucose homeostatic system.
The ester derivative AN-8, which was active in vitro,
failed to decrease blood glucose level in diabetic mice.
This is perhaps due to a rapid hydrolysis of the com-
2
4
pound by non-specific esterases in the plasma, which
precludes a substantial accumulation of this molecule in
cells. We suggest that the lack of in vitro effect of all the
other ester and amide derivatives that we have synthesized
may be due to steric hindrances related to the presence
of PEG or an alkyl linker containingmore than six
methylene groups that produce compounds resistant to
estrases or amidases.
3. Conclusions
Figure 5. Effect of insulin a-lipoic acid, AN-7 and AN-8 on hexose
transport in L6 myotubes. L6 myotubes were preincubated at 5.0 or at
2
3.0 mM glucose for 24 h before experiment. Medium was changed to
fresh medium with the same glucose concentration in the presence of
0 mL of vehicle DMSO, 200 mM AN-7, or 200 mM of AN-8. Seven h
1
. AN-7 and AN-8 are more potent and effective
thanLA inaugmentingtherateof hexosetransport
in L6 myotubes in culture. Their effects were addi-
tive to the stimulatory effect of insulin, indicating
that they operate an intracellular mechanism
distinguishable from that employed by insulin.
. The longer time period required to achieve
maximal effect of AN-7 and AN-8 in myotubes
seems to depend on the release rate of LA from
the produrgmolecules intracellularly.
1
before the uptake, assay medium was changed to a serum-free supple-
mented medium with the 0.5% (w/v) BSA and same additions. Insulin
(200 nM) was added duringthe last 1 h of incubation. LA (2.5 mM)
3
was added 2 h prior to the uptake assay. [ H]dGlc uptake assay was
performed 2 h after LA addition and 36 h after AN-7 and AN-8
additions. 100% value was taken for control cell maintained at 5.0
mM glucose (1.23ꢁ0.09 nmol dGlc/mgprotein/min). Mean ꢁSEM,
are shown: n=3. *p<0.05, compared to the respective control.
2
3
. AN-7 is more potent in vivo then the parent
compound LA; its blood glucose lowering effect
was similar to that of LA, but was obtained with
a 5-fold lower dose.
vitro and in vivo experimental systems. This compound
was 12-fold more potent in vitro than LA in increasing
the rate of hexose transport in L6 myotubes with or
without insulin. It was 5-fold more potent than LA in
reducingblood gl ucose levels in mildly diabetic mice.
Both AN-7 and AN-8 were ꢂ1.5-fold more effective
then LA in augmenting the rate of hexose transport in
L6 myotubes. Another strikingdifference between LA
and these compounds is the longer period required to
exert their stimulatory effects. Yet, the effect of AN-7 in
vivo was similar to that of LA, but a 5-fold lower doses.
The lack of effect of both LA and AN-7 in severe dia-
4. AN-7 and similar amide derivatives may serve as
prototype drugs to develop potent LA prodrugs
as an auxiliary drugtherapy for the treatment of
diabetes.
4. Experimental
1-(5-[1,2] Dithiolan-3-yl-pentanoyl)-pyrrolidine-2,5-dione
2
5ꢀ27
(LA-NHS) and PEG diamine (H N-PEG-NH )
2
2
were synthesized and characterized in our laboratory.
dl-a-lipoic acid (LA), N-N -dicyclohexyl carbodiimide
(
propane diol, 1,3-propane diamine, 1,6-hexane diol, 1,8-
0
DCC), N-hydroxy succinimide (NHS), PEG 2000, 1,3-
Table 1. Effects of LA and AN-7 on blood glucose level in diabetic
mice
ocatane diol, 1,10-decane diol, 2-deoxy-d-glucose
Test compound
Blood glucose (mg/dl)
Severe diabetes Mild diabetes
(
dGlc), streptozotocine, dimethyl sulfoxide (DMSO),
sodium citrate, were dotained from Aldrich, Sigma
Chemical Co (St. Louis, MO, USA) and used as
received. a-Minimal essential medium (a-MEM), fetal
calf serum (FCS), trypsin, glutamine and antibiotics
were purchased from Biological Industries (Kibbutz
Before
treatment
After
treatment
Before
treatment
After
treatment
DMSO
LA
AN-7
AN-8
571ꢁ17
514ꢁ38
556ꢁ23
582ꢁ20
517ꢁ86
547ꢁ62
502ꢁ71
530ꢁ79
219ꢁ32
235ꢁ34
246ꢁ33
194ꢁ21
198ꢁ15
163ꢁ33*
148ꢁ29*
209ꢁ69
1
Beth-Haemek, Israel). Insulin Actrapid HM was from
Novo Nordisk, Denmark. American Radiolabeled Che-
3
micals (St. Louis, MO, USA) supplied 2-[1,2- H(N)]
Six-week old male C57 black mice (BW 27–32 g) were made diabetic
by a single STZ injection as described above. Two weeks later they
were divided into two groups (mild and severe diabetes) and received
daily sc injections of DMSO (20 mL), LA (50 mg/kg BW, daily), AN-7
or AN-8, (10 mg/kg BW daily, for 2 weeks). Non-fasting blood glu-
cose levels were determined at 9 AM by Glucometr Elite. MeanSD
deoxy-d-glucose (60 Ci/mmol), d-glucose was pur-
chased from Merck (Darmstadt, Germany). Bradford
protein assay kit was from Bio-Rad Laboratories GmbH
(Munich, Germany). Meltingpoints were determined
with a meltingpoint apparatus (MeltingPoint SMP.
Stuart Scientific). Infrared (IR) spectra were recorded
an on VECTOR 22, Bruker spectrophotometer.
(
n=8).
p<0.05, significantly different from the respective to control,
Mann–Whitney test.
*

1188
A. Gruzman et al. / Bioorg. Med. Chem. 12 (2004) 1183–1190
4
.1. NMR
Hebrew University of Jerusalem. Male C57 black mice
were purchased from Harlan Laboratories (Jerusalem,
Israel) and housed (eight mice per cage) at the animal
The proton NMR spectra were obtained on a Varian
VXR-300 (300 MHz) spectrometer equipped with a 5-
mm switchable probe, and data were processed usinga
VNMR software. Chloroform-d and dimethyl sulfox-
ide-d6 were used as solvents; usingtetramethylsilane
ꢃ
facility with a 12-h light–dark cycle at 23 C. Diabetes
was induced in 6-week-old mice by a single ip injec-
tion of streptozotocin (150 mg/kg body weight) as
3
1
described. The diabetic mice were divided into two
groups: The mildly diabetic group that had non-fasting
blood glucose less than 250 mg/dL and a severe diabetic
group in which blood glucose ranged between 250 and
600 mg/dL. Two weeks following the induction of
diabetes both diabetic and control animals received
daily sc injections of LA (50 mg/kg body weight), AN-7
or AN-8 (10 mg/kg body weight) or the vehicle, DMSO,
daily, for 2 weeks. The volume of each injection was
20 mL.
Me Si (d 0.00 ppm) as an internal standard. The split-
4
tingpattern abbreviations are as follows: s, sin gl et; d,
doublet; t, triplet; q, quartet; m, unresolved multiplet
due to the field strength of the instrument; and dd,
doublet of doublet.
4
.2. Electrospray ionization mass spectrometry (ESI-MS)
Electrospray ionization mass spectrometry was mea-
sured on a Thermo Quest Finnigan LCQ-Duo in the
positive ion mode. In most cases, elution was in 20:79:1
water/methanol/acetic acid at a flow rate of 15 mL/min.
4.7. Glucose determination
Data were processed usingThermoQuest Finni ag n’s
Glucose concentration in culture medium samples and
in blood (taken from the tip of tails) was determined
TM
Xcalibur
software.
Biomass Calculation and Deconvolution
TM
with Glucometer Elite
(Bayer, Puteaux, France).
and blood glucose test strips
4
.3. Purification
Flash column chromatography on Merck silica gel 60
particle size 230–400 mesh) was used for purification.
CH Cl and CH OH were used in different ratio as elu-
4.8. Statistical analysis
(
Statistical analysis was done usingMann–Whitney
test.
2
2
3
ent. Purification on HPLC (Gilson) was performed on
C18, preparative and semi preparative columns, using
acetonitrile and double distilled water in different ratios
as eluent solvent. Analytical TLC was performed on
silica gel 60 F254-precoated plates purchased from
Merck (Darmstadt, Germany). The compounds were
visualized by usingUV li gh t or I 2 vapor.
4.9. Preparation of 1-(5-[1,2]dithiolan-3-yl-pentanoyl)-
pyrrolidine-2,5-dione (LA-NHS)
NHS (3.1g, 27.0 mmol), and N,N dicyclohexyl-carbdii-
mide (DCC) (5.57 g, 27.0 mmol) were dissolved in dry
dioxane (50 mL). A solution of LA (5 g, 26.0 mmol) in
dry dioxane (20 mL) was added slowly and the stirred
solution was left for 48 h at room temperature. The
dicyclohexyl urea (DCU) precipitate was filtered and the
solvent was evaporated. The powder was dissolved in 5
4
.4. Cell cultures
A subclone of rat skeletal muscle L6 cells (curtsey of
Dr. Nava Bashan, Ben-Gurion University, Beer-
Sheva. Israel), selected for high fusion potential was
used. Cells were grown in a-MEM supplemented with
ꢃ
mL of 2-propanol and left overnight at 4 C. The white
precipitate formed was filtered and dried.
1
1
0% (v/v) FCS, 100 U/mL penicillin G and 100 mg/mL
The yield was 57%. H NMR (DMSO): d 2.78 (t, 4H,
streptomycin, in a 95%:5% air/CO humidified incu-
bator at 37 C. Differentiation of myocytes into
[OC–CH –CH –CO]), 2.42–2,51 (m, 3H, [S–CH –CH –
CH-S]), 2.29 (t, 2H, [CH –CO–O–N]), 1.84 (m, 2H, [S–
2
2
2
2
2
ꢃ
myotubes was induced with 2% (v/v) FCS, as descri-
2
CH –CH –CH–S]), 1.37–1.62 (m, 4H, [CH –CH –CH –
2
2
2
2
2
2
8
bed. The culture medium was changed at least every
8 h.
CH –CO]), 1.25 (m, 2H, [CH –CH –CH –CH –CO]),
2 2 2 2 2
1
3
4
C NMR (DMSO): d 176.0, 166.4, 53.7, 42.9, 37.3,
6.8, 32.6, 25.9, 24.1, 21.3. Anal. calcd for
3
C H NO S : C, 47.50; H, 5.65; N, 4.62; S, 21.14.
Found: C, 47.48; H, 5.71; N 4.57; S, 21.10.
1
2
17
4 2
4
.5. Hexose uptake assay
3
The [ H] dGlc uptake assay was performed as descri-
2
9
bed. Carrier-mediated dGlc uptake was calculated on
the basis of protein quantity, determined by Bradford’s
4
.10. General procedure for the synthesis of (5-[1,2]di-
thiolan-3yl-pentanoylamino)-O-methoxy-PEG]-amide
AN-5) and di-(5-[1,2]dithiolan-3yl-pentanoylamino)-
3
0
method. Test compounds were added to cell cultures
from stocks solution in DMSO by a 1000-dilution (final
concentration 1% v/v). DMSO reduced the rate of hex-
ose transport by less than 5%.
(
PEG]-amide (AN-11)
4.10.1. Preparation of di-(5-[1,2]dithiolan-3yl-pentanoy-
lamino)-PEG]-amide, AN-11. PEG diamine M =2000
r
g/mol (5 g, 2.5 mmol) was dissolved in 1:1 ratio of 15%
NaHCO in water and dioxane (50 mL). A solution of
4
.6. Animals
3
All experiments were performed accordingto gu idelines
of the Care and Use of Laboratory Animals of the
LA-NHS (3.03 g, 10 m mol) in dioxane (50 mL) was
added and the stirred solution was left for 48 h at room

A. Gruzman et al. / Bioorg. Med. Chem. 12 (2004) 1183–1190
1189
temperature. The solution was filtered and the solvent
water and dioxane was removed. The powder left was
dissolved in 30 mL of hot EtOH/CH Cl and was fil-
were removed by a Dean–Stark apparatus. The solution
was evaporated under reduced pressure, and the resi-
due was dissolved in CH Cl . DCC (2.06 g, 10 mmol)
2
2
2
2
tered again and evaporated. The solid residue was dis-
solved in hot EtOH/ether 1:1 solution (10 mL), and left
overnight at 4 C. The white precipitate formed was fil-
tered and dried. The product was purified on silica-gel
by flash chromatography.
and LA (2.06 g, 10 mmol) in CH Cl was added and the
2 2
mixture was stirred for 48 h at room temperature.
Consequently, the precipitate was filtered and the
solvent was evaporated to yield white solid. The solid
was dissolved in a minimum volume of hot EtOH
ꢃ
ꢃ
and the solution was left at 4 C for 24 h. The
The yield was 72%. IR Vmax (KBr): 3447.13, 2887.35,
ꢀ
white precipitate was filtered, washed by ether and
purified by silica-gel through flash chromatography
technique.
1
1
1
653.27, 1280.22, 1113.27, 685.85 cm
H NMR
(
DMSO): d 3.70 (t, 4H, –NH–CH2–CH –O), 3.62 (m,
2
4
1
H, –NH–CH –CH –O, 1H, H–S–S–), 3.4–3.60 (m,
2
2
68H, –[CH –CH –O]n), 3.20 (m, 2H, –S–S–2H), 2.42
2
The yield was: 85%. IR Vmax (KBr): 2885.62, 1734.47,
ꢀ
2
1 1
(
m, 1H, CH –CH–S–), 2.38 (t, 2H, –CH2–CO–), 1.90–
2
1241.94, 1111.80, 685.38 cm . H NMR (DMSO, d)
1
3
1
.45 (m, 7H, –(CH2)3, CH2–CH–S–).
C NMR
4.10 (t, 4H, 2* –(OC–O–CH2–CH )), 3.70 (t, 1H, 1H,
H–S–S–) 3.60 (m, 4H, –O–OC–CH –CH –O), 3.40–3.58
2
(
3
DMSO): d 175.9, 75.4, 73.6, 60.0, 47.5, 42.6, 39.0, 38.0,
2.2, 28.9. Anal. calcd for C105H206N O S : C, 53.41;
H, 8.79; N, 1.19; S, 5.43. Found: C, 54.68; H, 8.94; N
2
2
(m, 168H, –[CH2–CH2–O]n), 3.16 (m, 2H, –S–S–2H),
2.40 (m, 1H, CH –CH–S–), 2.30 (t, 2H, –CH2–
2
46 4
2
1
.02; S, 5.00.
CO–), 1.9–1.45 (m, 7H, –[CH ] , CH –CH–S–).
3
2
2
1
3
C NMR (DMSO): d 176.7, 76.3, 73.8, 64.2, 60.0, 42.6,
42.1, 38.0, 37.2, 32.1, 28.2. Anal. calcd for C106H207O S :
4.10.2. (5-[1,2]Dithiolan-3yl-pentanoylamino)-O-meth-
oxy-PEG]-amide (AN-5). Yield 78%. IR Vmax (KBr):
3
cm
48
4
C, 53.37; H, 8.7; S, 5.43. Found: C, 54.21; H, 8.22; S,
4.62.
447.69, 2886.80, 1653.99, 1280.35, 1113.75, 685.70
ꢀ1 1
H NMR (DMSO): d 3.70 (t, 2H, –NH–CH2–
CH –O), 3.62 (m, 4H, –NH–CH –CH –O, 1H, H–S–S–),
2
2
2
3
.4–3.60 (m, 168H, –[CH –CH –O]n), 3.25 (s,
2
4.11. General procedure for the synthesis of 5-[1,2]dithio-
lan-3-yl-pentanoic acid 3-(5-[1,2]dithiolan-3yl-pentanoy-
loxy)-propyl ester (AN-8), 5-[1,2]dithiolan-3-yl-pentanoic
acid 6-(5-[1,2]dithiolan-3yl-pentanoyloxy)-hexyl ester
(AN-30); 5-[1,2]dithiolan-3-yl-pentanoic acid 8-(5-[1,2]di-
thiolan-3yl-pentanoyloxy)-octyl ester (AN-31); and 5-
[1,2]dithiolan-3-yl-pentanoic acid 10-(5-[1,2]dithiolan-3yl-
pentanoyloxy)-decyl ester (AN-32) 5-[1,2]dithiolan-3-yl-
pentanoic acid 3-(5-[1,2]dithiolan-3yl-pentanoyloxy)-pro-
pyl ester (AN-8)
2
3
CH–S–), 2.38 (t, 2H, –CH2–CO–), 1.90–1.45 (m, 7H,
H,OCH ), 3.20 (m, 2H, –S–S–2H), 2.42 (m, 1H, CH –
3
2
1
3
–
7
3
5
8
(CH )3, CH –CH–S–). C NMR (DMSO): d 175.9,
2
2
5.2, 74.5, 73.7, 73.5, 72.1, 62.0, 60.1, 47.5, 42.6, 39.1,
8.0, 32.1, 29.0. Anal. calcd for C H NO S : C,
9
7
197
45 2
4.01; H, 9.02; N, 0.64; S, 2.91. Found: C, 53.31; H,
.17 ; S, 2.97.
4.10.3. 3-(5-[1,2]dithiolan-3yl-pentanoylamino)-propyl]-
amide (AN-7). 1,3 Diamino propane (0.25 mL, 3 mmol)
was dissolved in a solution of 1:1 15% NaHCO in
3
To propane-1,3-diol (0.25 mL, 3.45 mmol) in CH Cl
2
2
water and dioxane. A solution of LA-NHS (3.63 g, 12
mmol) in dioxane was added and left stirred for 48 h at
room temperature. The solution was then filtered, and
the solvent was evaporated. The left residue was dis-
solved in 30 mL of a hot solution of EtOH/CH Cl (1:1)
(100 mL), DCC (2.84 g, 13.8 mmol) and LA (2.8.4 g,
13.8 mmol) were dissolved in 100 mL CH Cl . The mix-
ture was stirred for 48 h at room temperature. The
DCU precipitate was filtered and the solvent was evapo-
rated. The remaingpowder was dissolved in a solution
2
2
2
2
and filtered again. After evaporation of the solvent the
remaining solid was purified by silica-gel through flash
chromatography and preparative HPLC and the pure
product, AN-7, was collected.
of 15% (w/v) NaHCO in water, and the precipitate
3
was, filtered and purified on a silica-gel column via flash
chromatography, followed by preparative HPLC to
obtain pure compound of AN-8.
ꢃ
ꢃ
2930.15, 1709.33, 1234.59, 663.14 cm
The yield was: 75%; mp=79–81 C. IR V
3
(KBr):
The yield was: 80%, mp=99–100 C. IR V
ꢀ1
(KBr):
max
H NMR
max
ꢀ
1
1
1
303.18, 2930.69, 1732.75, 1205.83, 649.64 cm
NMR (DMSO, d) 3.62 (m, 1H, H–S–S–), 3.20–3.10 (dt,
H, –NH–CH2–), 3.07 (m,2H, –HN–CH –CH2–) 3.05
(m, 2H, –S–S–2H), 2.4(m, 1H, CH –CH–S–), 2.03 (t, 2H, –
CH2–CO–NH), 1.9–1.45 (m, 7H, –(CH ) , CH –CH–S–).
.
H
.
(DMSO, d) 4.10 (t, 1H, 1H, H-S-S-), 3.90 (t, 4H, 2* –
(OC–O–CH2–CH )), 3.60 (m, 2H, –OC–O–CH –CH2–),
3.16 (m, 2H, –S–S–2H), 2.40 (m, 1H, CH –CH–S–), 2.30
4
2
2
2
2
2
(t, 2H, –CH2–CO–), 1.9–1.45 (m, 7H, –[CH2] , CH2–
3
2
3
2
1
3
13
C NMR (DMSO): d 175.8, 75.4, 73.6, 59.9, 44.0, 42.0,
9.2, 38.1,37.8, 31.6, 29.0. Anal. calcd for
CH–S–). C NMR (DMSO): d 173.3, 64.5, 60.0, 42.0,
3
C H N O S : C, 50.63; H, 7.60; N, 6.21; S, 28.46.
38.1, 38.0, 37.3, 32.1, 28.3. Anal. calcd for C H O S :
C, 50.52; H, 7.36; N, 3.10; S, 28.39. Found: C, 50.50; H,
+
7.31; N, 2.98; S, 28.31. ESMS m/z: 452.4 (MH ).
1
9
32
4 4
1
9
34
2
2 4
Found: C, 50.53; H, 7.59; S, 28.41. ESMS m/z:
+
4
51.3(MH ).
4.11.1. 5-[1,2]Dithiolan-3-yl-pentanoic acid 6-(5-[1,2]di-
thiolan-3yl-pentanoyloxy)-hexyl ester, (AN-30). Yield:
4
.10.4. Di-(5-[1,2]dithiolan-3yl-pentanoyloxy)-PEG ester
AN-6). HO-PEG-OH M =2000 Da (5 g, 2.5 mmol)
ꢃ
(
was dissolved in toluene (100 mL) and traces of water
82%, mp=205–207 C. IR
V
(KBr): 2930.60,
r
max
ꢀ
1 1
1709.71, 1234.91, 664.06 cm . H NMR (DMSO, d)

1190
A. Gruzman et al. / Bioorg. Med. Chem. 12 (2004) 1183–1190
4
.10 (t, 1H, 1H, H–S–S–), 3.90 (t, 4H, 2* –(OC–O– 2. Evans, J. L.; Goldfine, I. D. Diabetes Technol. Ther. 2000,
CH2–CH )), 3.60 (m, 8H, –OC–O–CH –(CH2Þ –), 3.16
2, 401.
. Biewenga, G. P.; Haenen, G. R.; Bast, A. A. Gen. Phar-
macol. 1997, 29, 315.
4. Packer, L.; Witt, E. H.; Tritschler, H. J. Free Rad. Biol.
Med. 1995, 19, 227.
. Han, D.; Handleman, G.; Marcocci, L.; Sen, C. K.; Roy,
S.; Kobuchi, H.; Tritschler, H. J.; Flohe, L.; Packer, L.
Bio Factors 1997, 6, 321.
2
2
4
3
(
m, 2H, –S–S–2H), 2.40 (m, 1H, CH –CH–S–), 2.30 (t,
2
2
H, –CH –CO–), 1.9–1.45 (m, 7H, –[CH ] , CH2–CH–
2 2
3
1
3
S–). C NMR (DMSO): d 174.3, 60.0, 56.5, 42.0, 38.1,
7.9, 35.7, 34.3, 32.2, 28.6, 28.4. Anal. calcd for
C H O S : C, 53.40; H, 7.74; S, 25.92. Found: C,
3
5
2
2
38
4 4
+
5
3.31; H, 7.69; S, 26.10. ESMS m/z: 495.3(MH ).
6
. Betteridge, D. J. Metabolism 2000, 49, 3.
4.11.2. 5-[1,2]Dithiolan-3-yl-pentanoic acid 8-(5-[1,2]di-
thiolan-3yl-pentanoyloxy)-octyl ester, (AN-31). Yield:
7. Baynes, J. W.; Thorpe, S. R. Diabetes 1999, 48, 1.
8. Paolissio, G.; Giugliano, D. Diabetologia 1996, 39, 357.
9. Wollheim, C. Diabetologia 2000, 43, 265.
ꢃ
7
1
4
9%, mp=187–190 C. IR
709.33, 1234.85, 664.66 cm . H NMR (DMSO, d)
.10 (t, 1H, 1H, H–S–S–), 3.90 (t, 4H, 2* –(OC–O–
V
(KBr): 2929.81,
max
ꢀ
1
1
10. Biewenga, G.; Haenen, G. R.; Bast, A. Drug Metab. Rev.
1
997, 29, 1025.
1
1. Jacob, S.; Henriksen, E. J.; Schiemann, A. L.; Simon, I.;
Clancy, D. F.; Tritschler, H. J.; Jung, W. I.; Aufustine,
H. J.; Dietze, G. J. Arzeneimittel-Forschund 1995, 45, 872.
2. Haaguard, N.; Haaguard, E. Biochim. Biophys. Acta 1970,
CH2–CH )), 3.60 (m, 12H, –OC–O–CH –(CH2Þ –), 3.16
2
2
6
(
m, 2H, –S–S–2H), 2.40 (m, 1H, CH –CH–S–), 2.30 (t,
2
2
H, –CH –CO–), 1.9–1.45 (m, 7H, –[CH ] , CH –CH–
2 2
3 2
1
1
1
3
S–). C NMR (DMSO): d 173.4, 67.6, 59.9, 42.0, 38.0,
5.6, 34.3, 32.1, 32.0, 29.0, 28.6, 28.3 . Anal. calcd for
C H O S : C, 55.13; H, 8.10; S, 24.53. Found: C,
2
3. Estrada, D. E.; Ewart, H. S.; Tsakiridis, T.; Volchuk, A.;
22, 583.
3
2
4
42
4 4
Ramlal, T.; Tritshler, H. J.; Klip, A. Diabetes 1996, 45, 1798.
14. Ramrath, S.; Tritshler, H. J.; Eckel, J. Horm. Metab. Res.
999, 31, 632.
+
5
5.12; H, 8.16; S, 24.63. ESMS m/z: 523.2 (MH ).
1
1
5. Breithaupt-Grogler, K.; Neibch, G.; Schneider, E.; Erb,
K.; Hermann, R.; Blume, H. H.; Schug, B. S.; Belz, G. C.
Eur. J. Pharm. Sci. 1999, 8, 57.
4.11.3. 5-[1,2]Dithiolan-3-yl-pentanoic acid 10-(5-[1,2]di-
thiolan-3yl-pentanoyloxy)-decyl ester (AN-32). Yield:
ꢃ
8
1
4
7%, mp=175–176 C. IR
709.11, 1234.53, 662.27 cm . H NMR (DMSO, d)
.10 (t, 1H, 1H, H–S–S–), 3.90 (t, 4H, 2* –(OC–
V
(KBr): 2918.94,
max
1
6. Marangon, K.; Devaraj, S.; Tirosh, O.; Packer, L.; Jialal,
I. Free. Rad. Biol. Med. 1999, 27, 1114.
7. Nefkens, G.; Tesser, G. J. Am. Chem. Soc. 1961, 83, 1263.
ꢀ
1 1
1
OCꢀH2–CH )), 3.60 (m, 12H, –OC–O–CH –(CH2) –),
2
2
6
18. Anderson, G. W.; Zimmerman, J. E.; Callahan, F. M.
J. Am. Chem. Soc. 1964, 86, 1839.
3
.16 (m, 2H, –S–S–2H), 2.40 (m, 1H, CH –CH–S–),
2
2
.30 (t, 2H, –CH –CO–), 1.9–1.45 (m, 7H, –[CH ] ,
2
3
19. Spirada, P. K. Chem. Phys. Lipids 2000, 48, 157.
20. Walker, P.; Ramla, T.; Donovan, J.; Doering, T.; Sandra,
A.; Clip, A.; Pessin, E. J. Biol. Chem. 1989, 264, 6587.
2
1
3
CH –CH–S–). C NMR (DMSO): d 173.3, 63.8, 59.9,
42.0, 37.9, 35.6, 34.3, 32.2, 32.0, 29.4, 28.9, 27.9 . Anal.
calcd for C H O S : C, 56.68; H, 8.42; S, 23.48. Found:
2
2
1. Teishert, J.; Preiss, R. Int. J. Clin. Pharm. Ther. Toxicol.
1992, 30, 511.
2
6
46
4 4
+
C, 56.64; H, 8.32; S, 23.39. ESMS m/z: 552.5 (MH ).
2
2
2. Junge, W.; Krisch, K. Crit. Rev. Toxicol 1975, 3, 371.
3. Mark, L.; Kayden, H.; Steele, J.; Cooper, J.; Berlin, I.;
Rovenstein, E.; Brodie, B. J. Pharmacol. Exp. Ther. 1951,
Acknowledgements
1
02, 5.
2
4. Heymann,E. In The Organic Chemistry of Drug Design
And Drug Actions; Academic Press; New-York, 1980, Vol.
2, p. 291
This work was supported by grants from Found of Alex
Grass Center for DrugDesi ng and Synthesis at the
Hebrew University and The Yedidut Foundation Mex-
ico. S. S., A.H.Y. and J.K. are a members of the David
R. Bloom Center for Pharmacy at the Hebrew Uni-
versity of Jerusalem. A.G. received a fellowship award
from the Diabetes Research Center of the Hebrew Uni-
versity of Jerusalem.
25. Anderson, G. W.; Zimmerman, J. E.; Callahan, F. M. J.
Am. Chem. Soc. 1963, 85, 3039.
2
2
2
6. Luo, Y.; Prestwich, G. D. Biocojugate Chem. 2001, 12, 1085.
7. Zalipsky, S. US Patent 5,122,614, 1992
8. Kozlovsky, N.; Rudich, A.; Potashnik, R.; Bashan, N.
Free Rad. Biol. Med. 1997, 23, 859.
2
9. Kaiser, N.; Sasson, S.; Feener, E.; Boukobza-Vardi, N.;
Higashi, S.; Moller, D. E.; Davidheiser, S.; Przybylski,
R. J.; King, G. L. Diabetes 1993, 42, 80.
References and notes
3
3
0. Bradford, M. Anal. Biochem. 1972, 72, 248.
1. Haluzic, M.; Nedvedkova, J. Physiol. Res. 2000, 49 (Suppl 1),
S37.
1
. Randle, P. J. Diabetes Metab. Rev. 1998, 14, 263.