Synthesis, characterization and anti-diabetic effect of bis[(1-R-imidazolinyl)phenolato]oxovanadium(IV) complexes

Article abstract of DOI:10.1016/j.ica.2010.03.028

The five-coordinate oxovanadium(IV) complexes; [VO(pimin)₂] (1a), [VO(Etpimin)₂] (2) and [VO(EtOHpimin)₂] (3), were prepared by reacting the ligands; 2-(2′-hydroxyphenyl)-1H-imidazoline (piminH), 2-(2′-hydroxyphenyl)-1-eth

Full text of DOI:10.1016/j.ica.2010.03.028

Inorganica Chimica Acta 363 (2010) 2215–2221
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Synthesis, characterization and anti-diabetic effect
of bis[(1-R-imidazolinyl)phenolato]oxovanadium(IV) complexes
b
c
Ryan S. Walmsley a, Zenixole R. Tshentu a, , Manuel A. Fernandes , Carminita L. Frost
*
a Department of Chemistry, Rhodes University, P.O. Box 94, Grahamstown 6140, South Africa
b Molecular Sciences Institute, School of Chemistry, University of Witwatersrand, P.O. Wits 2050, Johannesburg, South Africa
c Department of Biochemistry and Microbiology, Nelson Mandela Metropolitan University, P.O. Box 77000, Port Elizabeth 6031, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
The five-coordinate oxovanadium(IV) complexes; [VO(pimin)2] (1a), [VO(Etpimin)2] (2) and [VO(EtOHpi-
min)2] (3), were prepared by reacting the ligands; 2-(20-hydroxyphenyl)-1H-imidazoline (piminH),
2-(20-hydroxyphenyl)-1-ethylimidazoline (EtpiminH) and 2-(20-hydroxyphenyl)-1-ethanolimidazoline
(EtOHpiminH), with VOSO4. The complexes were characterized by elemental analysis, IR, UV–Vis and cyc-
lic voltammetry. All complexes show V@O stretching vibrations between 932 and 987 cmꢀ1. The pres-
ence of three d–d transition occurring between 400 and 625 nm and the irreversible oxidation
(VIV ? VV) between 400 and 490 mV confirm the d1 electronic configuration of the complexes. The solid
state structures of [VO(pimin)2] (1a) and its autoxidation hydrolysis product [VO2(pimin)(piminH0)] (1b)
were determined by single crystal X-ray diffraction. The geometry of [VO(pimin)2] was found to be inter-
mediate between trigonal bipyramidal and square pyramidal and sits on a crystallographic twofold axis,
while the geometry of [VO2(pimin)(piminH0)] was distorted trigonal bipyramidal. Potentiometric
titrations were used to determine the protonation and stability constants for the ligands and oxovana-
dium(IV) complexes, respectively. The species existing over a biological pH range were also investigated.
The in vitro studies indicated that the oxovanadium(IV) complexes were effective in enhancing glucose
uptake in the 3T3-L1 adipocytes, C2C12 muscle cells and Chang liver cell lines. In these cell lines, the
anti-hyperglycemic effect was equivalent to or surpassed the effect of metformin.
Received 2 December 2009
Received in revised form 8 March 2010
Accepted 11 March 2010
Available online 16 March 2010
Keywords:
Oxovanadium
Imidazoline
Glucose uptake
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
stream, where the vanadium complexes may experience ligand
substitution as well as undergo redox reactions (VIV ꢀ VV)
Diabetes mellitus has become a growing health concern with
approximately 180 million people currently affected by the disease
worldwide [1]. Almost 80% of the deaths associated with diabetes
occur in low and middle-income countries [2], and the total num-
ber of those afflicted is set to double by 2030 [3]. The ever increas-
ing prevalence of diabetes has driven the development of
pharmaceuticals that do not exhibit the undesirable side-effects
of current therapies, and vanadium has shown promise in this re-
gard [4,5]. Inorganic salts of vanadium have shown significant
insulin-enhancing effects [6,7] but the potential toxicity of vana-
date and low absorption rate of vanadyl sulfate [8,9] has pushed
the focus towards designing new organovanadium compounds.
Tailoring of the properties of these vanadium complexes can thus
be achieved by simple modification of the organic ligand.
[10,11]. Vanadium may then be shuttled by bio-ligands to the cell
which it enters via diffusion, endocytosis or ion-channels (in the
form of vanadate) [12].
Vanadate is structurally similar to phosphate and is thus able to
bind and inhibit protein tyrosine phosphatase (PTP), an enzyme
which counteracts the autophosphorylation effects of insulin. As
a result, the insulin-induced signal transduction pathway remains
intact and the metabolism of glucose is improved [12]. The most
illustrious of these glucose-lowering organovanadium compounds
include bis(picolinato)oxovanadium(IV) [13], bis(maltolato)oxova-
nadium(IV) (BMOV), and bis(ethylmaltolato)oxovanadium(IV)
(BEOV), with the latter having undergone Phase IIa clinical trials
[14].
In this paper, we present the synthesis and characterization of
three novel oxovanadium(IV) compounds with 2-(20-hydroxy-
phenyl)-1R-imidazoline ligands. The stability and pH speciation
of the complexes, as well as their in vitro glucose-lowering effect,
was investigated. Certain imidazoline compounds have been
shown to have anti-diabetic effect [15,16], hence we have chosen
this group as a binding moiety to oxovanadium(IV).
The ligands stabilize vanadium under the pH conditions of the
digestive tract and assist absorption of vanadium into the blood
* Corresponding author. Tel.: +27 46603 8846; fax: +27 46622 5109.
E-mail address: z.tshentu@ru.ac.za (Z.R. Tshentu).
0020-1693/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2010.03.028

2216
R.S. Walmsley et al. / Inorganica Chimica Acta 363 (2010) 2215–2221
lowed to proceed for a further 2 h. The precipitate was collected,
washed with water, then methanol and dried at 100 °C. Yield:
56.9%. IR (cmꢀ1, KBr disk): 931,
(V@O); 3260, (N–H); 1609,
(C@N). Anal. Calc. for C18H18N4O3V: C, 55.53; H, 4.66; N, 14.39.
Found: C, 55.43; H, 4.74; N, 14.16%. UV–Vis (DMSO) kmax
ꢀ1 cmꢀ1): 617 (55), 542 (44), 403sh (196).
Upon standing the synthetic mother liquor at room tempera-
2. Experimental
m
m
2.1. Materials, methods and instrumentation
m
(e,
All chemicals were purchased from commercial sources (Sig-
ma–Aldrich, Merck) and used without further purification. Sol-
vents were of reagent grade. All oxovanadium(IV) complexes
were synthesized under an argon atmosphere.
M
ture and under aerobic conditions, yellow crystals of
[VO2(pimin)(piminH0)] (1b) were obtained. Yield wrt V: 12.3%. IR
The infrared spectra were recorded on a Perkin Elmer 2000 FTIR
spectrometer in the mid-IR range (4000–400 cmꢀ1) as KBr pellets.
1H and 13C NMR spectra of all ligands were recorded on a Bruker
AMX 400 NMR MHz spectrometer and reported relative to tetra-
methylsilane (d 0.00). Electronic spectra were recorded on a Varian
Cary 500 Scan UV–Vis spectrophotometer using 1 cm quartz cells
and dimethylsulfoxide as the solvent. Microanalysis was carried
out using a Vario Elementar Microcube ELIII. Cyclic voltammetry
was performed using a BAS CV 100 Cyclic Voltammogram. Potenti-
ometric studies were performed using a Metrohm 794 Titrino
equipped with a Metrohm LL Ecotrode. A Bio-Tek KC4 powerwave
XS microtiter plate reader was used to measure absorbance for 3-
(cmꢀ1, KBr disk): 931, 886,
m(V@O); 3191, m(N–H); 1617, m(C@N).
Anal. Calc. for C18H19N4O4V: C, 53.21; H, 4.71; N, 13.79. Found: C,
53.14; H, 4.71; N, 13.82%.
2.2.5. [VO(Etpimin)2] (2)
This was prepared in a similar manner as above except that
EtpiminH was used. Yield: 47.6%. IR (cmꢀ1, KBr disk): 987,
m
(V@O); 1602,
5.88; N, 12.58. Found: C, 59.04; H, 5.88; N, 12.27%. UV–Vis (DMSO)
kmax
, Mꢀ1 cmꢀ1): 621 (120), 547 (99), 401sh (300).
m(C@N). Anal. Calc. for C22H26N4O3V: C, 59.32; H,
(e
(4,5-dimethylthiazol-2-yl)-2,5-diphenyl
(MTT) and Glucose assays.
tetrazolium
bromide
2.2.6. [VO(EtOHpimin)2] (3)
This was prepared in a similar manner as for 1a except that
EtOHpiminH was used. Yield: 52%. IR (cmꢀ1, KBr disk): 966,
m
(V@O); 1603,
5.49; N, 11.74. Found: C, 55.30; H, 5.63; N, 11.66%. UV–Vis (DMSO)
kmax
, Mꢀ1 cmꢀ1): 625 (84), 548 (66), 404sh (269).
m(C@N). Anal. Calc. for C22H26N4O5V: C, 55.35; H,
2.2. Preparative work
2.2.1. 2-(20-Hydroxyphenyl)-1H-imidazoline (piminH)
(e
Methyl salicylate (1.84 g, 0.012 mol) was added to an excess of
ethylenediamine (3.2 g, 0.053 mol) in a conical flask and heated in
a microwave for 7 min at 180 W. Excess ethylenediamine was dis-
tilled off under reduced pressure until a solid material was left be-
hind. This solid was then digested overnight in chloroform, filtered
and then washed with cold chloroform to afford a cream solid.
Yield: 78.1%. 1H NMR (400 MHz, DMSO-d6): d 3.71 (s, 4H, Im–
CH2), 6.69 (t, 1H, Ar–H) 6.77 (d, 1H, Ar–H), 7.27 (t, 1H, Ar–H),
7.56 (d, 1H, Ar–H). 13C NMR (400 MHz, DMSO-d6): d 46.5, 110.2,
115.5, 118.4, 127.2, 132.7, 163.5, 166.1; IR (cmꢀ1, KBr disk):
2.3. Cyclic voltammetry
Cyclic voltammograms of 1a, 2 and 3 were recorded using a
glassy carbon electrode as the working electrode, platinum wire
as the counter electrode and silver chloride-coated silver wire as
the reference electrode. For all complexes, DMSO was used as the
solvent and tetrabutylammonium perchlorate was used as the sup-
porting electrolyte. Argon was bubbled through the solutions for
5 min before each run. A scan rate of 100 mV sꢀ1 was used.
3217, m(N–H); 1618, m(C@N). Anal. Calc. for C9H10N2O: C, 66.65;
H, 6.21; N, 17.27. Found: C, 66.68; H, 6.20; N, 16.98%.
2.4. Potentiometric studies
2.2.2. 2-(20-Hydroxyphenyl)-1-ethylimidazoline (EtpiminH)
The protonation and stability constants for the ligands and oxo-
vanadium(IV) complexes were determined by potentiometric titra-
tion of approximately 25 ml samples. All solutions were prepared
using freshly boiled and degassed deionized milli-Q water to en-
sure the removal of dissolved oxygen and carbon dioxide. The li-
gand concentration was 1 mM and metal-to-ligand ratios of 1:1,
1:5 and 1:10 were used. Titrations were performed over the pH
range of 2–11 under a continuous flow of purified nitrogen using
HCl and tetramethylammonium hydroxide (TMAOH). The vana-
dium stock solution containing 0.10 M HCl was standardized by
titration with permanganate. The ionic strength of the titration
solutions was kept constant at 0.10 M tetramethylammonium
chloride (TMACl). Titrations were controlled using Tiamo software.
The glass electrode was calibrated for a strong acid–base reaction
by the Gran-method [17] using the program GLEE [18], to deter-
mine the standard potential Eo. The ionic product of water (pKw)
of 13.83(1) at 25.0 0.1 °C in 0.10 M TMACl was used in all
calculations [19]. The hydrolysis model of a vanadyl system was
included in the model; [VO(OH)]+ (log b10ꢀ1 = ꢀ5.94) and
This was prepared in a similar manner as above except that N-
ethylethylenediamine was used. Yield: 50.3%. 1H NMR (d, 400 MHz,
CDCl3): 1.26 (t, 3H, N–CH2CH3), 3.42 (q, 2H, N–CH2), 3.50 (t, 2H,
Im–CH2), 3.91 (t, 2H, Im–CH2), 6.76 (t, 1H, Ar–H), 6.98 (d, 1H,
Ar–H), 7.27 (t, 1H, Ar–H), 7.36 (t, 1H, Ar–H). 13C NMR (d,
400 MHz, CDCl3): 14.50, 45.52, 50.47, 50.92, 112.31, 117.11,
118.66, 127.31, 132.16, 162.19, 167.09. IR (cmꢀ1, KBr disk): 1608,
m(C@N); Anal. Calc. for C11H14N2O: C, 69.45; H, 7.42; N, 14.73.
Found: C, 69.54; H, 7.67; N, 14.44%.
2.2.3. 2-(20-Hydroxyphenyl)-1-ethanolimidazoline (EtOHpiminH)
This was prepared in a similar manner as for piminH above ex-
cept that N-ethanolethylenediamine was used. Yield: 54.1%. 1H
NMR (d, 400 MHz, D2O): 3.57 (t, 2H, N–CH2), 3.79 (t, 2H, N–
CH2CH2), 4.11 (m, 4H, Im–CH2), 6.70 (t, 1H, Ar–H), 6.76 (d, 1H,
Ar–H), 7.27 (d, 1H, Ar–H), 7.43 (t, 1H, Ar–H). 13C NMR (d,
400 MHz, D2O): 43.13, 49.15, 49.19, 58.18, 112.21, 114.18,
121.81, 130.19, 134.93, 166.02, 169.49. IR (cmꢀ1, KBr disk): 3138
[(VO)2(OH)2]2+ (log b20ꢀ2 = ꢀ6.95), while [VO(OH)3]ꢀ (log b10ꢀ3
=
m(O–H); 1612, m(C@N). Anal. Calc. for C11H14N2O2: C, 64.06; H,
ꢀ18.0) and [(VO)2(OH)5]ꢀ (log b20ꢀ5 = ꢀ22.0) [20] did not fit. The
concentration stability constants bpqr = [MpLqHr]/[M]p[L]q[H]r were
calculated by using the computer program HYPERQUAD [21]. The final
values of the constants were obtained from an average of six inde-
pendent titrations using an average of 400 data points in total for
each refinement.
6.84; N, 13.58. Found: C, 64.31; H, 6.55; N, 13.44%.
2.2.4. [VO(pimin)2] (1a)
To a solution of Hpimin (0.25 g, 1.5 mmol) in methanol (5 ml)
was added vanadyl sulfate (0.152 g, 0.70 mmol) in water (5 ml).
A blue–green precipitate formed immediately. The reaction was al-

R.S. Walmsley et al. / Inorganica Chimica Acta 363 (2010) 2215–2221
2217
2.5. X-ray crystal structure resolution and refinement
Intensity data for 1a and 1b were collected on a Bruker APEX II
CCD area detector diffractometer with graphite monochromated
Mo K
a
radiation (50 kV, 30 mA) using the APEX 2 data collection
-scans of width
software [22]. The collection method involved
x
0.5° and 512 ꢁ 512 bit data frames. Data reduction was carried
out using the program SAINT+ [23] and absorption corrections were
made using the program SADABS [23].
Scheme 1. Synthesis of bis[(imidazolinzyl)phenolato]oxovanadium(IV) complexes.
The crystal structures were solved by direct methods using SHEL-
XTL [24]. Non-hydrogen atoms were first refined isotropically fol-
lowed by anisotropic refinement by full matrix least-squares
calculations based on F2 using SHELXTL. Hydrogen atoms were first
located in the difference map then positioned geometrically and al-
lowed to ride on their respective parent atoms. Diagrams and pub-
lication material were generated using SHELXTL, PLATON [25] and ORTEP-
3 [26].
2.6. In vitro studies
1a
2
2.6.1. Maintenance of cells lines
3T3-L1 preadipocytes, were maintained in Dulbecco’s modified
eagle medium (DMEM) supplemented with 10% fetal bovine ser-
um (FBS), while C2C12 mouse skeletal myoblasts and Chang liver
cells were maintained in RPMI-1640 medium (Sigma) supple-
mented with 10% FBS. These were incubated at 37 °C in a humid-
ified incubator with 5% CO2. Cells were subcultured at 70%
confluence and seeded at a density of 35 000 cells/ml (for 3T3-
L1) and 25 000 cells/ml (for Chang and C2C12) in 24-well culture
plates.
3
0.1
0.3
0.5
0.7
Potential (V)
Fig. 1. Cyclic voltammograms of complexes [VO(pimin)2] (1a), (Epa = 490 mV);
[VO(Etpimin)2] (2), (Epa = 440 mV); and [VO(EtOHpimin)2] (3), (Epa = 400 mV); in
DMSO.
tion to a methanolic solution of the corresponding ligand in a
1:2 M ratio (Scheme 1). The complexes which precipitated out of
solution were easily collected by filtration.
2.6.2. MTT cytotoxicity studies
Cells were seeded in 24-well plates (Nunc) at densities of
25 000 for C2C12 and Chang liver cells, and 35 000 cells/mL for
3T3-L1. After overnight attachment, the culture medium was re-
placed with medium containing the test compounds at a range of
concentrations (0.5–100 lM). Cells were incubated for 48 h at
37 °C, after which the MTT assay was performed [27].
Complexes 1a, 2 and 3 were soluble in alcohols, DMF and
DMSO. Complexes 2 and 3 were also soluble in dichloromethane,
acetonitrile and slightly soluble in water. The hydrolysis of these
complexes accompanied by autoxidation occurred when the
mother liquor was left to stand at room temperature. This was
noted by the change in color of the solution from green (VIV) to yel-
low–green (a mixture of VIV and VV). From the mother liquor of 1a,
yellow crystals deposited from the yellow–green solution. This oxi-
dation product was identified as the dioxovanadium(V) species, 1b.
The low oxidation potential of the complexes, determined by
cyclic voltammetry, corresponds well with the facile aerobic oxida-
tion observed in solution. All complexes display irreversible VIV–VV
oxidation peaks between 400 and 490 mV (Fig. 1).
2.6.3. Glucose uptake studies
The glucose uptake assay was performed using the GLUCOSE
(Glu-cinet) kit (BAUER) [28]. Control cells (Con) were represented
by untreated differentiated fat (3T3-L1), liver (Chang) and muscle
(C2C12) cells incubated in culture media. Positive control cells
(Met) were represented by untreated differentiated fat, liver and
muscle cells exposed to metformin. Cells were then exposed for
48 h to the test compounds. Thereafter glucose uptake was deter-
mined and the cell number was normalized using the MTT assay, as
laid out by Mossman [27].
3.2. Spectroscopic characterization
All of the ligands display infrared stretching frequencies be-
tween 1618 and 1608 cmꢀ1 which can be assigned to the azome-
thine (C@N) stretch [29]. Coordination to oxovanadium(IV) was
evident from a ꢂ9 cmꢀ1 shift of this stretch to lower frequencies
2.6.4. Statistical analysis
Error bars indicate the standard error of the mean (SEM) unless
specified otherwise (n = 3). The two-tail paired test was used to
determine significance of results (p < 0.05) and (p < 0.01).
[31]. The uncoordinated ligands show phenolic m(C–O) stretches
between 1287 and 1269 cmꢀ1 which upon coordination shift to a
3. Results and discussion
higher wavelength of 1313–1319 cmꢀ1 [32]. The
m(V@O) bands ap-
pear at 932, 987 and 965 cmꢀ1 for 1a, 2 and 3, respectively, and are
within the reported range for V@O stretches of 930–1030 cmꢀ1
3.1. Synthesis and general considerations
[33]. For the dioxovanadium(V) complex 1b, two m(V@O) bands ap-
The ligands were synthesized according to a method by Parík et
al. [29] with modifications by us. The ligand 2-(20-hydroxyphenyl)-
1H-imidazoline was substituted with the R-groups (R = Et, EtOH)
pear at 931 and 886 cmꢀ1. The higher energy band is assigned to
the mas stretch and the lower energy band is assigned to the ms
+
stretch of the cis-VO2 moiety [34].
with the goal of balancing the lipophilicity/hydrophilicity,
requirement of these metallopharmaceuticals [30]. The complexes
were synthesized by addition of an aqueous vanadyl sulfate solu-
a
In general, square pyramidal oxovanadium(IV) complexes dis-
play three low intensity d–d transitions in the range of 330–
1000 nm [35]. The high energy transition, b2 ? a1, occurs between

2218
R.S. Walmsley et al. / Inorganica Chimica Acta 363 (2010) 2215–2221
0.5
0.4
0.3
0.2
0.1
0
3
2
1a
360
415
470
525
580
635
690
Wavelength (nm)
Fig. 2. Electronic spectra of [VO(pimin)2] (1a), [VO(Etpimin)2] (2) and [VO(EtOHpi-
min)2] (3).
401 and 404 cmꢀ1 as a shoulder to a charge transfer band, hence
the higher intensity. The b2?b1 and b2 ? e transitions fall in the
range 538–545 cmꢀ1 and 617–625 cmꢀ1, respectively (Fig. 2).
Fig. 4. ORTEP diagram of 1b, showing 50% thermal probability ellipsoids.
3.3. X-ray crystallography
The blue–green crystals of 1a suitable for X-ray crystallography
were obtained upon recrystallization of the compound in acetoni-
trile at ꢀ20 °C. The yellow crystals of 1b were obtained from stand-
ing a methanolic mother liquor of 1a at room temperature for
4 days. The ORTEP diagrams for the structures of 1a and 1b are rep-
resented by Figs. 3 and 4, respectively. The selected crystallo-
graphic data is presented in Table 1 and selected bond lengths
and angles in Table 2. For 1a, the V@O bond length of 1.619(2) Å
is slightly longer than those observed for similar compounds,
which fall in the range of 1.591–1.605 Å [36,37]. The geometry of
1a is intermediate between trigonal bipyramidal and square pyra-
Table 1
Selected crystallographic data for [VO(pimin)2] (1a) and [VO2(pimin)(piminH0)] (1b).
Compound
[VO(pimin)2] (1a)
[VO2(pimin)(piminH0)] (1b)
Empirical formula
Formula weight
Crystal color
Crystal system
Space group
Temperature (K)
a (Å)
C18H18N4O3V
389.30
blue–green
orthorhombic
Fdd2
173(2)
12.4698(7)
36.810(2)
7.3123(4)
90
C18H19N4O4V
406.32
yellow
monoclinic
P21/c
173(2)
8.9544(8)
22.6383(19)
8.6648(7)
90
b (Å)
c (Å)
midal with
s = 0.49 [38] and sits on a crystallographic twofold axis.
a
(°)
b (°)
90
90
3356.5(3)
8
90.736(2)
90
1756.3(3)
4
Vanadium is situated 0.550 Å above the plane defined by the N2O2
ligand donor atoms and the N(1)–V(1)–O(2) and O(1)–V(1)–O(2)
angles are 99.12(6)° and 113.92(6)°, respectively. The N(1)–V(1)–
N(1) and O(1)–V(1)–O(1) angles in the equatorial plane are
161.8(1)° and 132.2(1)°, respectively, showing marked deviations
from an ideal of 180° expected for a square pyramidal geometry.
The bidentate ligand coordinates through the neutral imidazo-
line nitrogen and phenolate oxygen resulting in the formation of
the neutral bis[(imidazolinyl)phenolato]oxovanadium(IV) com-
plex. The average bite angle of the ligands is 86.32(8)°. The V(1)–
O(1) and V(1)–N(1) bond lengths have averages of 1.911(1) and
2.048(2) Å, respectively. The imidazoline and phenyl rings are
not co-planar, and have an average dihedral angle of 20.11° be-
c
(°)
V (Å3)
Z
qcalc (g/cm3)
Wavelength (Å)
Total reflections
Unique reflections
R
1.541
0.71073
4875
1827
0.0384
0.0735
1.537
0.71073
12 519
4234
0.0501
0.0962
Rw
Table 2
Selected bond lengths (Å) and angles (o) for [VO(pimin)2] (1a) and
[VO2(pimin)(piminH0)] (1b).
[VO(pimin)2] (1a)
[VO2(pimin)(piminH0)] (1b)
V(1)–O(2)
V(1)–O(1)
V(1)–N(1)
O(1)–V(1)–O(1)
N(1)–V(1)–N(1)
O(1)–V(1)–O(2)
N(1)–V(1)–O(2)
N(1)–V(1)–O(1)
N(1)–V(1)–O(1)i
1.619(2)
1.911(1)
2.048(2)
132.2(1)
161.8(1)
113.91(6)
99.12(6)
86.26(8)
86.37(8)
V(1)–O(3)
V(1)–O(4)
V(1)–O(1)
V(1)–O(2)
1.617(1)
1.648(1)
1.949(1)
1.932(1)
2.062(2)
108.3(1)
128.08(9)
123.40(9)
159.07(8)
V(1)–N(1)
O(3)–V(1)–O(4)
O(1)–V(1)–O(3)
O(1)–V(1)–O(4)
N(1)–V(1)–O(2)
tween them. As expected, the imidazoline ring is not planar with
the torsion angles C(9)–C(8)–N(1)–C(7) = 13.4(3)° and C(8)–C(9)–
N(2)–C(7) = 16.9(3)°.
Fig. 3. ORTEP diagram of 1a, showing 50% thermal probability ellipsoids.

R.S. Walmsley et al. / Inorganica Chimica Acta 363 (2010) 2215–2221
2219
The five-coordinate dioxovanadium(V) complex, [VO2(pimin)-
(piminH0)] (1b), adopts a highly distorted trigonal bipyramidal
geometry ( = 0.59) with O(1), O(3) and O(4) forming the trigonal
s
plane. The deviations in this plane are given by O(1)–V(1)–
O(3) = 128.05(9)°, O(1)–V(1)–O(4) = 123.40(9)° and O(3)–V(1)–
O(4) = 108.3(1)°. The N(1)–V(1)–O(2) at 159.02(8)° is also far from
the ideal of 180°.
The V(1)–O(3) and V(1)–O(4) at 1.617(1) and 1.648(1) Å,
respectively, are within the lengths expected for a five-coordinate
dioxovanadium(V) systems [34]. However, the V(1)–O(4) length is
slightly longer due to strong hydrogen bonding with N(2)–H of the
neighboring molecule [N(2)–Hꢃ ꢃ ꢃO(4) = 2.00(2) Å]. One of the li-
gands attaches to the dioxovanadium(V) center in a monodentate
fashion through the phenolate oxygen while the other ligand at-
taches in a bidentate manner through the imidazoline nitrogen
(N1) and phenolate oxygen (O1) with a bite angle of 81.56(8)°.
The imidazoline nitrogen (N3) of the mono-coordinated ligand is
protonated to afford a positively charged species which neutralizes
the extra negative charge resulting from the two phenolate oxy-
gens and the two dioxo ligands, and the vanadium(V) center.
Fig. 5. Species distribution diagram for the complexation of VO(IV) with EtpiminH
(LH) [CVO = 0.001 mM and CL = 0.004 mM].
3.4. pH speciation
This hydrolysis could be prevented by encapsulating the com-
plexes within a suitable drug carrier capsule.
The protonation constants of EtpiminH and EtOHpiminH were
determined using aqueous potentiometric titrations. Solution
studies with the ligand piminH were precluded due to the low
water solubility of this compound. The stability constants for the
VO(IV)–EtpiminH and EtOHpiminH systems were determined
and are summarized in Table 3.
3.5. In vitro studies
The vanadium compounds (VOSO4, 1a, 2 and 3) showed no
The first binding constants could be calculated with a 1:1 me-
tal-to-ligand ratio and precipitation due to the formation of hydro-
lysis products was observed for titrations with a 1:2 metal-to-
ligand ratio. The best fitting experimental curves were obtained
using a high metal-to-ligand ratio (10-fold), allowing for the deter-
mination of the second binding constants. The species distribution
diagram was generated for the VO-EtpiminH system, using the
program HySS [40], by inclusion of protonation, stability and
hydrolysis constants (Fig. 5).
Both EtpiminH and EtOHpiminH are basic ligands with pK val-
ues of 7.52 and 7.33, and 11.26 and 11.09 for the imidazoline nitro-
gen and phenolate oxygen, respectively. As a consequence of this,
hydrolysis of vanadyl occurs at a low pH (3 < pH < 7). As the pH in-
creases, however, complex formation prevents the hydrolysis.
Experimental data points up to pH 9 fitted well for the complexa-
tion studies, and the binary hydroxyl species [(VO)2(OH)5]ꢀ and
[VO(OH)3]ꢀ did not fit within the model for highly basic solutions.
This is probably due to the high ligand excess and the high thermo-
dynamic stabilization of the vanadium complexes in this pH range
due to the basic nature of the ligands. The overall stability con-
stants for these VO-ligand systems are high at 17.13(2) and
16.66(3) for VO-EtpiminH and VO-EtOHpiminH systems, respec-
tively. Both show good overall stability, but are however, subject
to hydrolysis at acidic pH range due to the basic nature of the li-
gands which tend to interact with vanadyl at relatively high pH.
cytotoxicity between 0.5 and 10
C2C12 cell lines tested. At concentrations above 10
the vanadium compounds proved to be cytotoxic. Thus, the glucose
uptake ability of the vanadium compounds, at non-cytotoxic con-
l
M in the 3T3-L1, Chang and
l
M, however,
centrations of 0.5, 1 and 10
lM, was screened. The results for the
1
lM concentrations are presented in Fig. 5, while the results for
all concentrations tested can be found in Table 4.
In the 3T3-L1 adipocytes (Fig. 6), the oxovanadium(IV) com-
pounds (VOSO4, 1a, 2 and 3) enhanced glucose uptake significantly,
either equaling or surpassing the effects observed for metformin, a
drug commonly used in the treatment of type II diabetes [41]. The
uptake of glucose in C2C12 muscle cells was not as significant as
observed in the other cell lines (Fig. 6). Nevertheless, vanadyl sul-
fate as well as compounds 1a and 2 improved glucose uptake rel-
ative to the control (Con). The Chang liver cells exhibited the
greatest response to the tested compounds compared to the other
cell lines tested.
Vanadyl sulfate proved to be the most effective of the oxovana-
dium(IV) compounds in stimulating the uptake of glucose. This can
be attributed to the fact that an in vitro model does not take into
account that uncomplexed vanadyl (VOSO4) may be simply oxi-
dized to vanadate, the active species for PTP-inhibition, while com-
pounds 1a–3 must first undergo oxidation and ligand dissociation.
This extra ligand dissociation step may be the reason for the de-
creased in vitro activity of the synthesized compounds [12]. How-
ever, the increased stability of compounds 1a, 2 and 3, imparted by
the organic ligands, may allow vanadium to survive the conditions
of the digestive system unlike VOSO4 which undergoes facile
hydrolysis and subsequent excretion. It has also been shown that
the intestinal cell permeability of complexed vanadium com-
pounds far exceeds that of the inorganic salt (VOSO4) [30]. The ef-
fect of the R-substituents (R = H, Et, EtOH) in compounds 1a–3
makes little difference to the in vitro activity, but would perhaps
become important in an in vivo model when considering the lipo-
philicity/hydrophilicity of the complexes. Although the in vitro
model used may have certain shortcomings, in that it does not
Table 3
Protonation (log K) and complex formation constants (log b) for VO(IV)-ligand
systems at 25 0.1 °C and I = 0.10 M (TMACl).
Reaction
Ligand
EtpiminH EtOHpiminH Maltol [39]
+
pK1
pK2
LH2 ꢀ H+ + LH
7.52(1)
11.26(1)
10.73(2)
7.33(1)
LH ꢀ H+ + Lꢀ
11.09(1)
10.53(2)
16.66(3)
8.44(2)
8.80(2)
16.29(2)
Log b110 VO2+ + Lꢀ ꢀ [VO(L)]+
Log b120 VO2+ + 2Lꢀ ꢀ [VO(L)2] 17.13(2)

2220
R.S. Walmsley et al. / Inorganica Chimica Acta 363 (2010) 2215–2221
Table 4
The effects of metformin (Met), VOSO4, [VO(pimin)2] (1a), [VO(Etpimin)2] (2) and [VO(EtOHpimin)2] (3) on 3T3-L1, Chang and C2C12 cells at 0.5, 1.0 and 10 lM concentrations.
Basal glucose uptake is represented as 100%.
3T3-L1 adipocytes
Chang hepatocytes
C2C12 myoblasts
Concentration (
Compound
Metformin
VOSO4
[VO(pimin)2] (1a)
[VO(Etpimin)2] (2)
[VO(EtOHpimin)2] (3)
lM)
0.5
1
10
0.5
1
10
0.5
1
10
110 11.1
126 28.7
120 3.2
117 0.9
111 11.6
163 17.2
154 24.7
141 27.0
138 32.7
128 31.0
112 5.3
111 3.2
107 4.8
113 31.0
72 3.9
123 10.2
122 3.2
115 11.1
124 4.7
129 4.9
121 9.7
122 20.2
117 3.7
131 12.4
109 7.2
126 26.1
163 7.2
122 20.4
116 22.3
117 31.3
124 31.9
235 15.9
88 45.2
87 4.8
50 19.5
87 2.8
96 8.8
24 3.5
114 8.42
Microbiology) for assisting with the biological assays. For financial
support, we thank the South African Medical Research Council
(MRC) and National Research Foundation (NRF).
200
**
3T3-L1
180
**
Chang
160
C2C12
140
120
100
80
Appendix A. Supplementary material
*
*
CCDC 756221 and 756222 contain the supplementary crystallo-
graphic data for 1a and 1b. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. Supplementary data associ-
ated with this article can be found, in the online version, at
doi:10.1016/j.ica.2010.03.028.
60
40
20
0
References
Con
Met
VOSO4
1a
2
3
Treatment (1 μM)
[1] D.O. Abegunde, C.D. Mathers, T. Adam, M. Ortegon, K. Strong, Lancet 370
(2007) 1929.
[2] P. Zimmet, K.G.M.M. Alberti, J. Shaw, Nature 414 (2001) 782.
[3] S. Wild, G. Roglic, A. Green, R. Sicree, H. King, Diabetes Care 27 (2004) 1047.
[4] S. Del Prato, N. Pulizzi, Metabolism 55 (2006) 20.
[5] M. Stumvoll, B.J. Goldstein, T.W. van Haeften, Lancet 365 (2005) 1333.
[6] M.C. Cam, R.A. Pederson, R.W. Brownsey, J.H. McNeill, Diabetologia 36 (1993)
218.
Fig. 6. The effects of metformin (Met), vanadyl sulfate (VOSO4), [VO(pimin)2] (1a),
[VO(Etpimin)2] (2) and [VO(EtOHpimin)2] (3) (1 M) on 3T3-L1, Chang and C2C12
l
glucose uptake. The basal glucose uptake, is represented as 100% (Con). Error bars
indicate SEM (n = 3), *p < 0.05 and **p < 0.01 relative to the (Con).
[7] E.H. Clayton, A.G. Tahiliani, J.H. McNeill, Science 227 (1985) 1474.
[8] M.D. Cohen, Toxicol. Ecotoxicol. News 3 (1996) 132.
[9] I.A. Setyawati, K.H. Thompson, V.G. Yuen, Y. Sun, M. Battell, D.M. Lyster, C. Vo,
T.J. Ruth, S. Zeisler, J.H. McNeill, C. Orvig, J. Appl. Physiol. 84 (1998) 569.
[10] T. Jakusch, D. Hollender, E.A. Enyedy, C.S. Gonzalez, M. Montes-Bayon, A. Sanz-
Medel, J.C. Pessoa, I. Tomaz, T. Kiss, Dalton Trans. (2009) 2428.
[11] D. Sanna, E. Garribba, G. Micera, J. Inorg. Biochem. 103 (2009) 648.
[12] K.G. Peters, M.G. Davis, B.W. Howard, M. Pokross, V. Rastogi, C. Diven, K.D.
Greis, E. Eby-Wilkens, M. Maier, A. Evdokimov, S. Soper, F. Genbauffe, J. Inorg.
Biochem. 96 (2003) 321.
[13] H. Sakurai, K. Fujii, H. Watanabe, H. Tamura, Biochem. Biophys. Res. Commun.
214 (1995) 1095.
[14] K.H. Thompson, J. Lichter, C. LeBel, M.C. Scaife, J.H. McNeill, C. Orvig, J. Inorg.
Biochem. 103 (2009) 554.
[15] T.L. Berridge, J.C. Doxey, A.G. Roach, Eur. J. Pharmacol. 213 (1992) 213.
[16] P. Proks, I. Treinies, H.–J. Mest, S. Trapp, Eur. J. Pharmacol. 452 (2002) 11.
[17] G. Gran, Analyst 77 (1952) 661.
[18] P. Gans, B. O’Sullivan, Talanta 51 (2000) 33.
[19] C. Bazzicalupi, A. Bencini, A. Bianchi, A. Danesi, C. Giorgi, B. Valtancoli, Inorg.
Chem. 48 (2009) 2391.
[20] R.P. Henry, P.C.H. Mitchell, J.E.J. Prue, J. Am. Chem. Soc., Dalton Trans. (1973)
1156.
[21] P. Gans, A. Sabatini, A. Vacca, Talanta 43 (1996) 1739.
[22] Bruker, APEX2. Version 2.0-1, Bruker AXS Inc., Madison, Wisconsin, USA, 2005.
[23] Bruker, SAINT-NT. Version 6.0. (includes XPREP and SADABS), Bruker AXS Inc.,
Madison, Wisconsin, USA, 2005.
[24] Bruker, SHELXTL. Version 5.1. (includes XS, X., XP, XSHELL), Bruker AXS Inc.,
Madison, Wisconsin, USA, 1999.
[25] A.L. Spek, J. Appl. Crystallogr. 36 (2003) 7.
[26] L.J. Farrugia, J. Appl. Crystallogr. 30 (1997) 565.
[27] T. Mossman, J. Immunol. Met. 65 (1983) 55.
exactly replicate in vivo conditions [42], it does provide a simple
and ethical platform for the rapid assessment of new complexes.
4. Conclusions
A series of bis[1-R-(imidazolinyl)phenolato]oxovanadium(IV)
compounds were successfully synthesized and characterized. The
compounds 1a–3 undergo aerobic oxidation in solution to form
the dioxovanadium(V) analogs, one such product (1b) was isolated
and identified by X-ray crystallography. The acidity constants of
EtpiminH and EtOHpiminH, and the stability constants of their cor-
responding VO(IV) complexes were determined by potentiometric
acid–base titrations. These compounds showed encouraging stabil-
ity and resistance to hydrolysis at relatively high pH, however,
modification of the ligands to improve the stability at low pH is
currently in progress. The glucose uptake effect of the oxovana-
dium(IV) compounds was investigated using 3T3-L1 adipocytes,
C2C12 muscle cells and Chang liver cells. In the 3T3-L1 adipocytes
and Chang liver cells, all of the oxovanadium(IV) compounds (1a, 2
and 3) enhanced glucose uptake, while in the C2C12 muscle cells,
all oxovanadium compounds besides 3 showed anti-hyperglycemic
activity.
[28] A.S. Keston, in: 129th Meeting American Chemistry Society, 1956, p. 31c.
[29] P. Parík, S. Senauerová, V. Lisková, K. Handlír, M. Ludwig, J. Heterocycl. Chem.
43 (2006) 835.
Acknowledgements
[30] K.H. Thompson, B.D. Liboiron, Y. Sun, K.D.D. Bellman, I.A. Setyawati, B.O.
Patrick, V. Karunaratne, G. Rawji, J. Wheeler, K. Sutton, S. Bhanot, C. Cassidy,
J.H. McNeill, V.G. Yuen, C. Orvig, J. Biol. Inorg. Chem. 8 (2003) 66.
[31] A.D. Westland, M.T.H. Tarafder, Inorg. Chem. 20 (1981) 3992.
[32] S. Bhattacharya, T. Ghosh, Trans. Met. Chem. 27 (2002) 89.
[33] G. Wilkinson, R.D. Gillard, J.A. McCleverty, Comprehensive Coordination
Chemistry, Elsevier, Pergamon Press, New York, 1987.
The authors thank Dr. E. Antunes from DST/Mintek-NIC (Rhodes
University) for the microanalysis results and Mr. T. Mugadza
(Rhodes University, Chemistry Department) for assisting with elec-
trochemistry. We acknowledge Miss C. Davison from Nelson Man-
dela Metropolitan University (Department of Biochemistry and

R.S. Walmsley et al. / Inorganica Chimica Acta 363 (2010) 2215–2221
2221
[34] A.G.J. Ligtenbarg, A.L. Spek, R. Hage, B.L. Feringa, J. Am. Chem. Soc., Dalton
Trans. (1999) 659.
[39] K. Saatchi, K.H. Thompson, B.O. Patrick, M. Pink, V.G. Yuen, J.H. McNeill, C.
Orvig, Inorg. Chem. 44 (2005) 2689.
[35] J. Selbin, Chem. Rev. 65 (1965) 153.
[40] L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini, A. Vacca, Coord. Chem. Rev.
184 (1999) 311.
[41] M.B. Davidson, A.L. Peters, Am. J. Med. 102 (1997) 99.
[42] A.A. Nejo, G.A. Kolawole, A.R. Opuka, J. Wolowska, P. O’Brien, Inorg. Chim. Acta
362 (2009) 3993.
[36] C. Bolm, T.K.K. Luongb, K. Harms, Chem. Ber. 130 (1997) 887.
[37] M. Melchior, K.H. Thompson, J.M. Jong, S.J. Rettig, E. Shuter, V.G. Yuen, Y. Zhou,
J.H. McNeill, C. Orvig, Inorg. Chem. 38 (1999) 2288.
[38] A.W. Addison, T.N. Rao, J. Chem. Soc., Dalton Trans. (1984) 1349.