Convenient microwave-assisted synthesis of lipophilic sulfenamide prodrugs of metformin

Article abstract of DOI:10.1016/j.ejps.2013.05.023

A convenient microwave-assisted synthesis of lipophilic sulfenamide prodrugs of antidiabetic agent, metformin, is reported in this study. These acyclic prodrugs were synthesized directly from selected disulfides with basic metformin and silver nitrate by a one-pot reaction under microwave irradiation. The prepared prodrugs had significantly increased lipophilicity, which resulted in excellent permeability of the octylthio prodrug of metformin across a Caco-2 cell monolayer. According to our preliminary in vivo studies, the octylthio prodrug was also absorbed mostly intact after oral administration in rats. In conclusion, this study shows that these types of more lipophilic sulfenamide prodrugs can be promising candidates to improve permeability and passive absorption of highly water-soluble metformin.

Full text of DOI:10.1016/j.ejps.2013.05.023

European Journal of Pharmaceutical Sciences 49 (2013) 624–628
Contents lists available at SciVerse ScienceDirect
European Journal of Pharmaceutical Sciences
journal homepage: www.elsevier.com/locate/ejps
Convenient microwave-assisted synthesis of lipophilic sulfenamide
prodrugs of metformin
⇑
Kristiina M. Huttunen , Jukka Leppänen, Krista Laine, Jouko Vepsäläinen, Jarkko Rautio
School of Pharmacy, Faculty of Health Sciences, University of Eastern Finland, P.O. Box 1627, FI-70211 Kuopio, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
A convenient microwave-assisted synthesis of lipophilic sulfenamide prodrugs of antidiabetic agent, met-
formin, is reported in this study. These acyclic prodrugs were synthesized directly from selected disul-
fides with basic metformin and silver nitrate by a one-pot reaction under microwave irradiation. The
prepared prodrugs had significantly increased lipophilicity, which resulted in excellent permeability of
the octylthio prodrug of metformin across a Caco-2 cell monolayer. According to our preliminary
in vivo studies, the octylthio prodrug was also absorbed mostly intact after oral administration in rats.
In conclusion, this study shows that these types of more lipophilic sulfenamide prodrugs can be promis-
ing candidates to improve permeability and passive absorption of highly water-soluble metformin.
Ó 2013 Elsevier B.V. All rights reserved.
Received 24 March 2013
Received in revised form 21 May 2013
Accepted 26 May 2013
Available online 1 June 2013
Keywords:
Lipophilic
Metformin
Microwave-assisted synthesis
Prodrug
Sulfenamide
1. Introduction
and cysteine, and free thiol-containing proteins (Nti-Addae et al.,
2011; Nti-Addae and Stella, 2011). Sulfenamides are also versatile
Metformin is an oral anti-diabetic agent used as a first-line drug
of choice for the treatment of type 2 diabetes (Holman, 2007; Kir-
pichnikov et al., 2002; Strack, 2008). It improves hyperglycemia by
increasing insulin sensitivity, suppressing hepatic glucose produc-
tion and enhancing peripheral glucose uptake. The cellular mecha-
nisms behind these effects have been proposed to occur via
AMP-activated kinase (AMPK) (Andujar-Plata et al., 2012; Boyle
et al., 2010). Metformin, like all the other bisguanides, is a strong
base (pKa ca. 11–12) and highly water soluble. Therefore, it is ab-
sorbed slowly and incompletely from the upper intestine after oral
administration (Strack, 2008). With its effective doses (0.5–2.0 g
per day) metformin frequently causes uncomfortable gastrointesti-
nal adverse effects, such as abdominal pain, diarrhea, nausea, and
vomiting.
prodrugs, since the nature of the promoiety can be altered to meet
the requirements of the prodrug; i.e., aqueous solubility can be in-
creased by adding polar groups and lipophilicity improved by
lengthening or branching the alkyl chain. In the present study,
we introduce an efficient microwave-assisted synthesis method
to prepare lipophilic sulfenamide prodrugs of metformin. We also
report physicochemical characterization of the prepared prodrugs
and discuss the factors affecting bioconversion of sulfenamide
prodrugs.
2. Materials and methods
2.1. General synthetic methods
As there is a great demand for an improved metformin therapy,
we have studied more lipophilic prodrugs of metformin with the
aim of increasing permeability and passive diffusion across the cell
membranes. This is expected to result in better oral absorption of
metformin and eventually improve the clinical outcome. Sulfena-
mides have proved to be promising candidates as more lipophilic
prodrugs of metformin, since the powerful cyclization behavior
can be avoided by this prodrug approach (Huttunen et al., 2012,
2009). Generally, sulfenamide prodrugs are rapidly bioconverted
to their parent drugs by endogenous free thiols, like glutathione
All the reactions were performed with reagents of commercial
high purity quality without further purification. Reactions were
monitored by thin-layer chromatography using aluminum sheets
coated with silica gel 60 F245 (0.24 mm) with suitable visualization.
Microwave irradiation experiments were carried out in a Biotage
Initiator microwave reactor (Biotage, Uppsala, Sweden) in a pres-
sure-rated glass tubes. Purifications of the final compounds were
performed either by a preparative high performance liquid chro-
matography (HPLC) system, which consisted of an Schimadzu
SCL-10AVP System Controller (Schimadzu Corporation, Kyoto,
Japan), LC-8A Preparative Pump, SPD-10AVP UV–VIS Detector
(operated at 220 and 254 nm), and FRC-10A Fraction Collector on
⇑
Corresponding author. Tel.: +358 40 3553 684.
E-mail address: Kristiina.Huttunen@uef.fi (K.M. Huttunen).
a Kromasil 100 C8 column (150 mm  20 mm, 5
lm) (Higgins
0928-0987/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ejps.2013.05.023

K.M. Huttunen et al. / European Journal of Pharmaceutical Sciences 49 (2013) 624–628
625
Analytical, Inc., Southborough, MA, USA) or by CombiFlash Com-
2.3. HPLC analyzes
panion Flash Chromatography Instrument (Teledyne Isco, Inc., Lin-
coln, NE, USA) on a RediSep Rf Normal Phase Silica Flash Column.
1H and 13C nuclear magnetic resonance (NMR) spectra were re-
corded on a Bruker Avance 500 spectrometer (Bruker Biospin,
Fällanden, Switzerland) operating at 500.13 MHz and 125.75,
respectively, using tetramethylsilane as an internal standard. pH-
dependent NH-protons of the compounds were not observed. The
final products were also characterized by mass spectroscopy with
a Finnigan LCQ quadrupole ion trap mass spectrometer (Finnigan
MAT, San Jose, CA, USA) equipped with an electrospray ionization
source. The purities of the final compounds were determined to
be >95% by an analytical HPLC (Agilent Technologies Inc., Wilming-
ton, DE, USA) on a Supelco Supelcosil LC-Si analytical column
The analyzes were performed on the HPLC system, which con-
sisted of a Agilent 1100 binary pump (Agilent Technologies Inc.,
Wilmington, DE, USA), a 1100 micro vacuum degasser, a HP 1050
Autosampler, a HP 1050 variable wavelength detector (operated
at 235 nm). The chromatographic separations were achieved on a
Supelco Supelcosil LC-Si analytical column (4.6 mm  250 mm,
5
l
m) (Supelco Inc., Bellefonte, PA, USA) by using isocratic elution
of acetonitrile and 10 mM ammonium acetate buffer (pH 5.0) with
ratio of 60:40 (v/v) at the flow rate 1.0 mL/min at room
a
temperature.
2.4. Distribution coefficients of metformin and the prodrugs 1–4 in
(4.6 mm  250 mm, 5
lm) (Supelco Inc., Bellefonte, PA, USA).
octanol/water
The distribution coefficients (logD) of metformin and metfor-
min prodrugs 1–4 were determined at room temperature from
the distribution of the compounds between a mixture of 1-octanol
and 50 mM acetate buffer (pH 4.0) or phosphate buffer (pH 6.5 or
2.2. General procedure for the synthesis of sulfenamide prodrugs 1–4
Metformin (N,N-dimethyl imidodicarbonimidic diamide hydro-
chloride) (1.0 eq.) in 1 M NaOH (1.5 eq.) was stirred at room tem-
perature for 30 min. Water was evaporated in vacuo and the
residue was dissolved in MeOH. The solvent was evaporated and
the residue was redissolved in cold anhydrous MeOH. NaCl was fil-
tered out of the solution and the filtrate was evaporated to yield
basic metformin as a white solid (99%).
7.4). Metformin or prodrug solution (50 lg/mL) in 50 mM buffer
was added to 1-octanol saturated with desired buffer (the ratio
varied with the compounds). The mixtures were shaken for 2 h
and the phases were then separated. The concentration of metfor-
min or the prodrugs 1–4 in the buffers before and after the shaking
were analyzed by HPLC.
Basic metformin (2.0 eq.), AgNO3 (1.0 eq.) and disulfide (com-
mercial or prepared as previously described (Kirihara et al., 2007)
(1.0 eq.) were dissolved in anhydrous MeOH in a sealed pressure-
rated glass tube and irradiated at 80 °C in a microwave reactor
for 30 min. The reaction mixture was filtered and the filtrate was
treated by AcOH (2.2 eq.) or 1 M HCl (1.1 eq.). The solvent was re-
moved under reduced pressure and the residue was purified either
by preparative HPLC on Kromasil 100 C8 column eluting with 0.1%
AcOH solution and acetonitrile (15:85, v/v) or by CombiFlash Com-
panion Flash Chromatography Instrument eluting with MeOH/
DCM (1:5) solution to obtain the compounds 1–4.
2.5. Permeation studies of metformin and the prodrug 3 across Caco-2
monolayer
The bidirectional Caco-2 permeability experiments were con-
ducted at Cyprotex Discovery Ltd., (Cheshire, UK). Caco-2 cells ob-
tained from the American Type Culture Collection (ATCC) were
used between passage numbers 40–60. Cells were seeded onto
Millipore Multiscreen Caco-2 plates at 1 Â 105 cells/cm2. They
were cultured for 20 days in Dubecco’s Modified Eagle’s Medium
(DMEM) and media was changed every 2 or 3 days. On day 20
the permeability study was performed. Hanks Balanced Salt Solu-
tion (HBSS) pH 7.4 buffered with 25 mM HEPES and 4.45 mM glu-
cose at 37 °C was used as the medium in the permeability studies.
Incubations were carried out in an atmosphere of 5% CO2 with a
relative humidity of 95% at 37 °C. The solutions were made by
diluting 10 mM metformin or the prodrug 3 in DMSO with HBSS
2.2.1. N1,N1-dimethyl-N4-(butylthio)-bisguanidine (1)
1H NMR (CD3OD): d ppm 3.04 (s, 6H), 2.75 (t, 3JHH = 7.4 Hz, 2H),
3
1.68–1.60 (m, 2H), 1.51–1.43 (m, 2H), 0.95 (t, JHH = 7.4 Hz, 3H);
13C NMR (CD3OD): d ppm 161.45, 160.62, 40.02, 38.06 (2C),
30.94, 22.72, 14.02. MS (ESI+) for C8H20N5S (M)+: Calcd 218.33,
Found 218.13.
to give a final concentration of 10 lM (final DMSO concentration
1%). The fluorescent integrity marker lucifer yellow was also in-
cluded in the dosing solution. Test and control compounds (ateno-
lol, propranolol, and talinolol) were quantified by LC–MS/MS. The
permeability coefficients (Papp) for metformin and the prodrug 3
in the apical to basolateral (A–B) and basolateral to apical (B–A)
direction across Caco-2 cells were calculated.
2.2.2. N1,N1-dimethyl-N4-(hexylthio)-bisguanidine (2)
3
1H NMR (CD3OD): d ppm 3.11 (s, 6H), 2.95 (t, JHH = 7.6, 2H),
1.70–1.62 (m, 2H), 1.49–1.41 (m, 2H), 1.36–1.30 (m, 4H), 0.94–
0.89 (m, 3H); 13C NMR ((CD3)2SO): d ppm 159.72, 157.62, 51.99,
35.92 (2C), 31.08, 27.71, 22.11, 22.02, 12.93. MS (ESIÀ) for
C
10H22N5S (MÀH)À: Calcd 244.39, Found 244.18.
2.6. In vitro bioactivation of the prodrugs 1–4
2.2.3. N1,N1-dimethyl-N4-(octylthio)-bisguanidine (3)
The rates of bioactivation of the prodrugs 1–4 in 1 mM reduced
glutathione (GSH) or cysteine (CYS) solutions were determined at
37 °C. The incubation mixtures were prepared by mixing 100 mM
GSH or CYS stock solution in water with preheated 50 mM phos-
phate buffer (pH 7.4). 10 mM prodrug stock solution in 80% aceto-
3
1H NMR (CDCl3): d ppm 3.15 (s, 6H), 3.06 (t, JHH = 7.6 Hz, 2H),
1.74–1.66 (m, 2H), 1.50–1.42 (m, 2H), 1.38–1.26 (m, 8H), 0.93–
0.88 (m, 3H); 13C NMR (CD3OD): d ppm 159.64, 157.57, 53.96,
37.48 (2C), 32.87, 30.21, 30.08, 29.37, 23.65, 23.37, 14.41. MS
(ESIÀ) for C12H26N5S (MÀH)À: Calcd 272.44, Found 272.21.
nitrile (the final concentration of prodrugs was 100 lM and the
acetonitrile concentration 1%) was added and the mixture was
incubated from 10 min to 6 h at 37 °C. The samples were with-
drawn at appropriate intervals and the reactions were terminated
by the addition of ice-cold stopping solution, containing 2.5 mM N-
methylmaleimide in acetonitrile. The samples were centrifuged for
15 min at 12,000g at room temperature and kept on ice until the
supernatants were analyzed by the HPLC. The pseudo-first-order
2.2.4. N1,N1-dimethyl-N4-((2-ethylhexyl)thio)-bisguanidine (4)
1H NMR (CD3OD): d ppm 3.12 (s, 6H), 2.97–2.92 (m, 2H), 1.80–
1.73 (m, 1H), 1.53–1.38 (m, 4H), 1.37–1.26 (m, 4H), 0.96–0.89 (m,
6H); 13C NMR ((CD3)2SO): d ppm 159.52, 157.55, 56.26, 36.04 (2C),
34.55, 32.28, 28.30, 25.63, 22.41, 13.01, 9.60. MS (ESIÀ) for
C
12H26N5S (MÀH)À: Calcd 272.44, Found 272.22.

626
K.M. Huttunen et al. / European Journal of Pharmaceutical Sciences 49 (2013) 624–628
half-lives (t1/2) for the rates of bioconversion of the prodrugs were
calculated from the slope of the linear portion of the plotted loga-
rithm of remaining prodrug concentration versus time.
Metformin, being nearly completely in ionized form under studied
conditions, had very low logD values (Fig. 1), while the prodrugs,
as expected, were much more lipophilic. The shortest alkyl group
in the promoiety (butylthio derivative 1) had only a modest
improvement on lipophilicity, although lengthening and branching
the alkyl chain increased lipophilicity gradually. The longer alkyl
groups seemed to increase the amount of the prodrugs in octanol
phase, especially at pH values between 6.5 and 7.4 (Fig. 1). There-
fore, these prodrugs have a great potential to improve passive
intestinal absorption of metformin, since the pH of intestine varies
typically between 4.5 and 8.0.
3. Results and discussion
3.1. Synthesis
Previously sulfenamide prodrugs have been prepared by con-
verting appropriate disulfides to sulfenyl chlorides or utilizing thi-
ophthalimides as
a sulfur transfer agent to attach various
The bidirectional permeability studies were carried out across
the Caco-2 (human epithelial colorectal adenocarcinoma) cell
monolayer, which is a well-accepted in vitro model of human intes-
tinal absorption. However, we have noted that due to the differ-
ences in used Caco-2 cell lines, this model does not always truly
correlate with the in vivo intestinal absorption of metformin
(Natoli et al., 2011). For example, in our earlier studies metformin
did not cross the monolayer at all (Huttunen et al., 2012), although
other studies have shown that metformin is transported across the
Caco-2 cell monolayer both paracellularly and via specific trans-
porters, like organic cation transporters (OCTs) (Horie et al.,
2011; Proctor et al., 2008). Despite this fact, we used this model
in the present study to evaluate intestinal permeability of the most
lipophilic prodrug 3 with the octylthio promoiety. This sulfena-
mide prodrug had good permeability in both directions (Table 1).
Compared to our previous results, the octylthio derivative 3 had
1000-times better permeability in both directions across the
Caco-2 cell monolayer than the cyclohexylthio derivative 5 (Ta-
ble 1), which indicates that sulfenamide prodrugs with longer acy-
clic alkyl chains could significantly increase passive absorption of
metformin. Furthermore, in our earlier studies, only small amount
of the prodrug 5 was able to cross the Caco-2 cell monolayer intact
and a large amount of it was prematurely converted to metformin
on the apical side, which in turn decreased the overall permeability
(Huttunen et al., 2012). This same phenomenon was also reported
with sulfenamide prodrug of NH-acidic linezolid (Guarino et al.,
2011; Nti-Addae et al., 2012). In the present study with the octyl-
thio derivative 3, no metformin was found on either side of the
absorption membrane, and therefore it was evident that the pro-
drug 3 was not converted prematurely to metformin. In that sense,
longer acyclic alkyl chains seem to be more useful as sulfenamide
prodrug structures for basic compounds, like metformin, as they
somehow can stabilize the prodrug structure.
promoieties to the nitrogen of the parent drug (Guarino et al.,
2007; Hemenway et al., 2007). However, sulfenyl chlorides are
known to be unstable and undergo side reactions with several
functional groups in the parent drugs (Davis et al., 1977). Despite
numerous attempts, we could not obtain even a trace of sulfena-
mide prodrugs of metformin by the sulfenyl chloride method. Syn-
theses via the thiophthalimide route were more successful, but the
overall yields were quite poor (5–20%). Therefore, we wanted to
test if the sulfenamide prodrugs of metformin could be prepared
by a microwave-assisted one-pot reaction (Larhed and Hallberg,
2001; Wathey et al., 2002). Metal ion-assisted synthesis has been
reported to be a convenient method to prepare sulfenamides di-
rectly from disulfides and amines, since the complexation of the
metal ion renders the disulfide bond to the nucleophilic attack of
the amine (Bentley et al., 1971; Davis et al., 1977). The use of silver
nitrate as a metal ion source, the excess use of the amine and long
reaction times at room temperature have guaranteed good yields
of sulfenamides. We followed this procedure and irradiated the
reaction mixtures at 80 °C in a microwave reactor for 30 min to ob-
tain the desired sulfenamide prodrugs of metformin with good
yields (Scheme 1). However, we had difficulties to get enough pure
compounds, since these sulfenamides seemed to be somewhat
unstable in their basic form. Therefore, the reaction mixtures were
quickly filtrated and acidified by AcOH or dilute HCl, which stabi-
lized the compounds. The sulfenamide salts were then purified
chromatographically to remove the excess of metformin and to ob-
tain over 95% pure metformin sulfenamide prodrugs with moder-
ate yields (10–30%). Due to the instability of the final compounds
before their salt formation, it seemed to be quite difficult to im-
prove the overall yields by changing the reaction conditions. How-
ever, the proposed microwave-assisted synthesis shortens the
reaction and work up times significantly when comparing to the
other methods. In this study, we prepared a set of metformin sulf-
enamides with acyclic promoieties. This includes three promoi-
eties with straight alkyl chain; butylthio (1), hexylthio (2), and
octylthio (3), and one promoiety with branched alkyl group; 2-eth-
ylhexylthio (4) (Scheme 1).
3.3. In vitro bioconversion
The rates of bioconversion of the sulfenamide prodrugs 1–4
were evaluated in 1 mM reduced glutathione (GSH) and cysteine
(CYS) solutions at 37 °C (pH 7.4). Unexpectedly, the compound 1
with butylthio promoiety was the only sulfenamide prodrug that
degraded and released metformin quantitatively (Table 2). As com-
pared to our previous results of sulfenamide prodrug of metformin
with cyclohexylthio promoiety (5) (Huttunen et al., 2012), the
3.2. Lipophilicity and permeability
Lipophilicity of the sulfenamide prodrugs 1–4 and metformin
were determined by octanol–water distribution coefficients (logD)
in buffered solutions at pH 4.0, 6.5, and 7.4 at room temperature.
Scheme 1. Microwave-assisted synthesis method of sulfenamide prodrugs of metformin.

K.M. Huttunen et al. / European Journal of Pharmaceutical Sciences 49 (2013) 624–628
627
Table 2
Half-lives of the metformin sulfenamide prodrugs 1–5 in 1 mM reduced glutathione
solution (GSH) and 1 mM cysteine solution (CYS) at 37 °C (mean SD; n = 3).
Compounds
Half-live (t1/2) with GSH
Half-live (t1/2) with CYS
1
2–4
5
14.52 1.11 s
9.48 1.39 s
a
a
–
–
4.68 0.07 minb
n.d.c
a
b
c
No degradation was observed during the 6 h incubation.
Huttunen et al. (2012).
Not determined.
Fig. 1. Octanol–water distribution coefficients (logD) of metformin and the
sulfenamide prodrugs 1–4 at pH 4.0, 6.5 and 7.4 at room temperature (mean SD,
n = 3).
However, the steric hindrance does not explain why the prodrugs
2–4 were not bioconverted by free thiols. One factor that affects
SN2 reactions is the solvent. When comparing the log D values of
the prodrugs 1 and 2 (Fig. 1), there is over 200-fold difference at
pH 4.0 and over 3500-fold difference at pH 7.4 in their hydrophilic-
ities, the compound 1 being more hydrophilic. In fact, the com-
pounds 2–4 are so lipophilic that they most probably stack
together due to the van der Waals interactions of hydrocarbon tails
and form micelle-type of structures in buffer solutions. That would
explain why the free thiols are unable to react with these prodrugs,
since nitrogen–sulfur bonds of the prodrugs could be hided inside
the micelles. Thus, increasing the lipophilicity of highly water-sol-
uble parent drugs by lengthening the hydrocarbon tail of the pro-
moiety may generate surfactant-type prodrugs, which may have
impact on the bioconversion and liberation rate of the parent drug.
Therefore, physicochemical and stereoelectronic requirements
need to be considered carefully when designing more lipophilic
sulfenamide prodrugs of highly water-soluble parent drugs.
As expected on the basis of the in vitro results, the preliminary
in vivo studies of the prodrug 1 in rats showed that this compound
was most probably predisposed to premature degradation on the
brush border of the enterocytes. After the oral administration of
the prodrug 1, the AUCs calculated from liberated metformin was
almost same as after the oral administration of metformin itself
and no changes in metformin absorption profile was seen with this
prodrug. It has been proposed that the premature bioconversion of
sulfenamide prodrugs may occur by free thiol rich proteins on the
apical side of the enterocytes (Guarino et al., 2011; Nti-Addae et al.,
2012) and in that sense the butylthio prodrug 1 is not a good can-
didate for further pharmacokinetic studies. The preliminary in vivo
studies of the octylthio prodrug 3 were more encouraging, since
this prodrug seemed to be absorbed mainly intact after its oral
administration. However, more experiments are needed to reveal
the real pharmacokinetic profile of the prodrug 3, as it was found
to be distributed to deep compartment, most probably to red blood
cells, but possibly to other tissues as well.
bioconversion rate of the acyclic sulfenamide 1 in GSH solution
was almost 20-times faster than the one of the cyclic sulfenamide
prodrug 5. In the presence of smaller thiol, cysteine, the bioconver-
sion rate of the prodrug 1 was even faster (Table 2). Interestingly,
the sulfenamide prodrugs of NH-acidic compounds behaved oppo-
sitely, showing faster bioconversion rate in GSH than CYS solution
(Nti-Addae and Stella, 2011). As the bioconversion is a non-enzy-
matic reaction, there are several factors that can affect the rate of
sulfur-nitrogen bond cleavage of acid and basic sulfenamide
prodrugs.
It has been proposed that the bioconversion reaction of
NH-acidic sulfenamides occurs by a bimolecular nucleophilic sub-
stitution (SN2) mechanism, where a nucleophilic thiolate anion at-
tacks on the electropositive sulfur atom that has donated its lone
pair of electrons to the nitrogen atom of the sulfenamide bond
(Nti-Addae et al., 2011). Therefore, the nucleophilicity of the react-
ing thiol and the leaving group capacity of the amine affect greatly
the sulfur-nitrogen bond breakage and the disulfide bond forma-
tion. However, in our case, the bioconversion reaction is more com-
plex, since metformin itself has a lot of lone pair electrons allowing
complicate equilibrium between different resonance forms, as
shown in Scheme 2. In the bioconversion reaction, the sulfur atom
of the free thiol is attached to the sulfur atom of the prodrug. At the
same time, the cleavage of the S–H bond of the free thiol is assisted
by the lone pair electrons of the nearest tertiary nitrogen, leading
to formation of free metformin and the disulfide.
Like all SN2 reactions, the bioconversion of sulfenamide pro-
drugs of metformin seemed to be slightly sensitive to increasing
steric hindrance, as the nucleophile cannot approach the backside
of the electrophilic sulfur atom. The bulkier promoiety of the cyclo-
hexylthio derivative 5 decreased the bioconversion rate when
compared to the one of the butylthio derivative 1 (Table 2).
Table 1
Caco-2 cell permeability of metformin and the prodrugs 3 and 5 from apical to basolateral side (AB ? BL) and from basolateral to apical side (BL ? AP) (n = 2).
Papp (Â10À6 cm/s)
AB ? BL
BL ? AP
a
a
Metformin
Prodrug 3
Prodrug 5
–
–
NH NH2
N
N
N
N
NH2
44.80 7.44
35.00 2.97
NH NH2
S
S
N
N
H
0.039 0.004b
0.068 0.027b
NH NH2
N
N
H
a
Not detected in the receiver compartment.
Huttunen et al. (2012).
b

628
K.M. Huttunen et al. / European Journal of Pharmaceutical Sciences 49 (2013) 624–628
Scheme 2. Proposed bioconversion reaction mechanism of sulfonamide prodrug of metformin.
Boyle, J.G., Salt, I.P., McKay, G.A., 2010. Metformin action on AMP-activated protein
4. Conclusion
kinase: a translational research approach to understanding a potential new
therapeutic target. Diabet. Med. 27, 1097–1106.
In the present study, we propose a fast and convenient micro-
wave-assisted one-pot synthesis to prepare more lipophilic sulfe-
namide prodrugs of metformin. Lipophilicity of the prodrugs was
increased by lengthening or branching the alkyl group of the pro-
moiety, which in turn improved bidirectional permeability of the
octylthio prodrug 3 across the Caco-2 monolayer. Interestingly,
the promoieties that had alkyl chain length exceeding four carbon
atoms were not cleaved by a free thiol-mediated chemical reaction
in vitro in aqueous media. The reason may be that these surfactant-
type compounds stack together in aqueous media, which prevents
free thiols to react with the hided sulfenamide prodrug bonds.
Therefore, the physicochemical and stereoelectronic properties of
the promoieties need to take into a careful consideration when
designing novel more lipophilic sulfenamide prodrugs of highly
water-soluble parent drugs. However, the preliminary in vivo stud-
ies in rats indicated that the octylthio prodrug 3 is a promising can-
didate to improve permeability and passive absorption of
metformin, as it was absorbed mainly intact after its oral
administration.
Davis, F.A., Friedman, A.J., Kluger, E.W., et al., 1977. Chemistry of the sulfur–nitrogen
bond. 12. Metal-assisted synthesis of sulfenamide derivatives from aliphatic
and aromatic disulfides. J. Org. Chem. 42, 967–972.
Guarino, V.R., Karunaratne, V., Stella, V.J., 2007. Sulfenamides as prodrugs of NH-
acidic compounds: a new prodrug option for the amide bond. Bioorg. Med.
Chem. Lett. 17, 4910–4913.
Guarino, V.R., Nti-Addae, K., Stella, V.J., 2011. MDCK cell permeability
characteristics of a sulfenamide prodrug: strategic implications in considering
sulfenamide prodrugs for oral delivery of NH-acids. Bioorg. Med. Chem. Lett. 21,
172–175.
Hemenway, J.N., Nti-Addae, K., Guarino, V.R., et al., 2007. Preparation,
characterization and in vivo conversion of new water-soluble sulfenamide
prodrugs of carbamazepine. Bioorg. Med. Chem. Lett. 17, 6629–6632.
Holman, R., 2007. Metformin as first choice in oral diabetes treatment: the UKPDS
experience. J. Ann. Diabetol. Hotel Dieu., 13–20.
Horie, A., Sakata, J., Nishimura, M., et al., 2011. Mechanisms for membrane transport
of metformin in human intestinal epithelial Caco-2 cells. Biopharm. Drug
Dispos. 32, 253–260.
Huttunen, K.M., Mannila, A., Laine, K., et al., 2009. The first bioreversible prodrug of
metformin with improved lipophilicity and enhanced intestinal absorption. J.
Med. Chem. 52, 4142–4148.
Huttunen, K.M., Leppanen, J., Vepsalainen, J., et al., 2012. In vitro and in vivo
evaluation of a sulfenamide prodrug of basic metformin. J. Pharm. Sci. 101,
2854–2860.
Kirihara, M., Asai, Y., Ogawa, S., et al., 2007. A mild and environmentally benign
oxidation of thiols to disulfides. Synthesis, 3286–3289.
Kirpichnikov, D., McFarlane, S.I., Sowers, J.R., 2002. Metformin: an update. Ann. Int.
Med. 137, 25–33.
Acknowledgments
Larhed, M., Hallberg, A., 2001. Microwave-assisted high-speed chemistry: a new
technique in drug discovery. Drug Discov. Today 6, 406–416.
Natoli, M., Leoni, B.D., D’Agnano, I., et al., 2011. Cell growing density affects the
structural and functional properties of Caco-2 differentiated monolayer. J. Cell
Physiol. 226, 1531–1543.
Nti-Addae, K.W., Stella, V.J., 2011. In vitro conversion of model sulfenamide
prodrugs in the presence of small molecule thiols. J. Pharm. Sci. 100, 1001–
1008.
Nti-Addae, K.W., Laurence, J.S., Skinner, A.L., et al., 2011. Reversion of sulfenamide
prodrugs in the presence of free thiol-containing proteins. J. Pharm. Sci. 100,
3023–3037.
Nti-Addae, K.W., Guarino, V.R., Dalwadi, G., et al., 2012. Determination of the
permeability characteristics of two sulfenamide prodrugs of linezolid across
Caco-2 cells. J. Pharm. Sci. 101, 3134–3141.
We would like to thank Mrs. Miia Reponen for skillful technical
assistance with prodrug synthesis, Mrs. Helly Rissanen for the
skilled technical assistant with the physicochemical characteriza-
tion and bioconversion studies and M.Sc. Lauri Peura and M.Sc. Sini
Mustalahti for performing the preliminary in vivo studies. Tim
Smith from Cyprotex Discovery Ltd., (Cheshire, UK) is also
acknowledged for conducting Caco-2 studies. The work was finan-
cially supported by the Finnish Funding Agency for Technology and
Innovation (Tekes), the Academy of Finland (Grant #256837), and
Emil Aaltonen Foundation.
Proctor, W.R., Bourdet, D.L., Thakker, D.R., 2008. Mechanisms underlying saturable
intestinal absorption of metformin. Drug Metab. Dispos. 36, 1650–1658.
Strack, T., 2008. Metformin: a review. Drugs Today (Barc) 44, 303–314.
Wathey, B., Tierney, J., Lidstrom, P., et al., 2002. The impact of microwave-assisted
organic chemistry on drug discovery. Drug Discov. Today 7, 373–380.
References
Andujar-Plata, P., Pi-Sunyer, X., Laferrere, B., 2012. Metformin effects revisited.
Diabetes Res. Clin. Pract. 95, 1–9.
Bentley, M.D., Douglass, I.B., Lacadie, J.A., et al., 1971. New one-step synthesis of
sulfenamides from alkyl and aryl disulfides. J. Chem. Soc. [Section] D: Chem.
Commun., 1625–1626.