Vanadium complexes with mixed O,S anionic ligands derived from maltol: Synthesis, characterization, and biological studies

Article abstract of DOI:10.1021/ic0486926

Four mixed O,S binding bidentate ligand precursors derived from maltol (3-hydroxy-2-methyl-4-pyrone) have been chelated to vanadium to yield new bis(ligand)oxovanadium(IV) and tris(ligand)vanadium(III) complexes. The four ligand precursors include two pyranthiones, 3-hydroxy-2-methyl-4-pyranthione, commonly known as thiomaltol (Htma), and 2-ethyl-3-hydroxy-4-pyranthione, commonly known as ethylthiomaltol (Hetma), as well as two pyridinethiones, 3-hydroxy-2-methyl-4(H)-pyridinethione (Hmppt) and 3-hydroxy-1,2-dimethyl-4- pyridinethione (Hdppt). Vanadium complex formation was confirmed by elemental analysis, mass spectrometry, and IR and EPR (where possible) spectroscopies. The X-ray structure of oxobis(thiomaltolato)vanadium(IV),VO(tma)₂, was also determined; both cis and trans isomers were isolated in the same asymmetric unit. In both isomers, the two thiomaltolato ligands are arranged around the base of the square pyramid with the V=O linkage perpendicular; the vanadium atom is slightly displaced from the basal plane [V(1) = 0.656(3) A, V(2) = 0.664(2) A]. All of the new complexes were screened for insulin-enhancing effectiveness in streptozotocin-induced diabetes in rats, and VO(tma) ₂ was profiled metabolically for urinary vanadium and ligand clearance by GFAAS and ESIMS, respectively. The new vanadium complexes did not lower blood glucose levels acutely, possibly because of rapid dissociation and excretion.

Full text of DOI:10.1021/ic0486926

Inorg. Chem. 2005, 44, 2678−2688
Vanadium Complexes with Mixed O,S Anionic Ligands Derived from
Maltol: Synthesis, Characterization, and Biological Studies
Vishakha Monga,† Katherine H. Thompson,† Violet G. Yuen,‡ Vijay Sharma,‡ Brian O. Patrick,†
John H. McNeill,‡ and Chris Orvig*,†
Medicinal Inorganic Chemistry Group, Department of Chemistry, and Faculty of Pharmaceutical
Sciences, UniVersity of British Columbia, 2036 Main Mall,
VancouVer, British Columbia V6T 1Z1, Canada
Received September 17, 2004
Four mixed O,S binding bidentate ligand precursors derived from maltol (3-hydroxy-2-methyl-4-pyrone) have been
chelated to vanadium to yield new bis(ligand)oxovanadium(IV) and tris(ligand)vanadium(III) complexes. The four
ligand precursors include two pyranthiones, 3-hydroxy-2-methyl-4-pyranthione, commonly known as thiomaltol (Htma),
and 2-ethyl-3-hydroxy-4-pyranthione, commonly known as ethylthiomaltol (Hetma), as well as two pyridinethiones,
3-hydroxy-2-methyl-4(H)-pyridinethione (Hmppt) and 3-hydroxy-1,2-dimethyl-4-pyridinethione (Hdppt). Vanadium
complex formation was confirmed by elemental analysis, mass spectrometry, and IR and EPR (where possible)
spectroscopies. The X-ray structure of oxobis(thiomaltolato)vanadium(IV),VO(tma)2, was also determined; both cis
and trans isomers were isolated in the same asymmetric unit. In both isomers, the two thiomaltolato ligands are
arranged around the base of the square pyramid with the V
dO linkage perpendicular; the vanadium atom is
slightly displaced from the basal plane [V(1) 0.656(3) Å, V(2)
)
)
0.664(2) Å]. All of the new complexes were
screened for insulin-enhancing effectiveness in streptozotocin-induced diabetes in rats, and VO(tma)2 was profiled
metabolically for urinary vanadium and ligand clearance by GFAAS and ESIMS, respectively. The new vanadium
complexes did not lower blood glucose levels acutely, possibly because of rapid dissociation and excretion.
Introduction
the well-known tyrosine phosphatase inhibition through
binding to cysteine at the putative active site.6 A third area
of interest, vanadium complexes with ligands that supply both
sulfur and oxygen donors for candidate therapeutic agents,7
is one that dovetailed nicely with our previous explorations
in the more general area of vanadium-containing insulin
mimics.8
The coordination chemistry of vanadium with sulfur-
containing ligands is an emerging field of interest with
relevance to several disparate biological systems.1-4 The
presence of vanadium-sulfur bonding in the active site of
certain nitrogenase enzymes has been well established,5 and
vanadium-sulfur coordination also appears to be pivotal to
The insulin-enhancing properties of vanadium complexes
have been thoroughly investigated, both in vitro and in
vivo.7-10 The absolute or relative lack of insulin, a hallmark
of all types of diabetes mellitus, can be partially overcome
* To whom correspondence should be addressed. Tel.: 604-822-4449.
Fax: 604-822-2847. E-mail: orvig@chem.ubc.ca
† Medicinal Inorganic Chemistry Group, Department of Chemistry.
‡ Faculty of Pharmaceutical Sciences.
(1) Thompson, K. H.; Orvig, C. Coord. Chem. ReV. 2001, 219-221, 1033
and references therein.
(6) Zhang, M.; Zhou, M.; Van Etten, R. L.; Stauffacher, C. V. Biochemistry
1997, 36, 15.
(2) Rehder, D. Coord. Chem. ReV. 1999, 182, 297 and references therein.
(3) (a) Vanadium in Biological Systems; Chasteen, N. D., Ed.; Kluwer
Academic Publishers: Dordrecht, The Netherlands, 1990. (b) Butler;
A.; Walker, J. V. Chem. ReV. 1993, 93, 1937. (c) Money, J. K.;
Huffman, J. C.; Christou, G. Inorg. Chem. 1988, 27, 507 and references
therein.
(7) (a) Sakurai, H.; Sano, H.; Takino, T.; Yasui, H. J. Inorg. Biochem.
2000, 80, 99. (b) Nekola, H.; Wang, D.; Gru¨ning, C.; Gatjens, J.;
Behrens, A.; Rehder, D. Inorg. Chem. 2002, 41, 2379.
(8) Thompson, K. H.; Orvig, C. Met. Ions Biol. Syst. 2004, 41, 221.
(9) Thompson, K. H.; McNeill, J. H.; Orvig, C. Chem. ReV. 1999, 99,
2561.
(10) Rehder, D.; Pessoa, J. C.; Geraldes, C. F. G. C.; Margarida, M.; Castro,
C. A.; Kabanos, T.; Kiss, T.; Meier, B.; Micera, G.; Pettersson, L.;
Rangel, M.; Salifoglou, A.; Turel, I.; Wang, D. J. Biol. Inorg. Chem.
2002, 7, 384.
(4) Crans, D. C.; Smee, J. J.; Gaidamauskas, E.; Yang, L. Chem. ReV.
2004, 104, 849 and references therein.
(5) (a) George, G. N.; Coyle, C. L.; Hales, B. J.; Cramer, S. P. J. Am.
Chem. Soc. 1988, 110, 4057. (b) Butler, A.; Carrano, C. J. Coord.
Chem. ReV. 1991, 109, 61.
2678 Inorganic Chemistry, Vol. 44, No. 8, 2005
10.1021/ic0486926 CCC: $30.25
© 2005 American Chemical Society
Published on Web 03/24/2005

Vanadium Complexes with Mixed O,S Anionic Ligands
by oral administration of either vanadium salts (e.g., VOSO4
or sodium vanadate) or organically chelated V(III),11 V(IV),12
or V(V)13 complexes. Numerous studies in experimental
animals have shown that tolerable levels of vanadium-
containing compounds can normalize plasma glucose, lipids,
and the thyroid hormone and at least partially counteract the
increased oxidative stress that accompanies diabetes mellitus
and contributes to secondary complications. Nonetheless, an
obviously limited “window of optimal utility”14,15 has spurred
continuing efforts to further tailor the ligands used for
vanadium complexation such that greater potency and
efficacy could be achieved.
In previous studies in our laboratories, both vanadium-
(III) and oxovanadium(IV) complexes with, primarily,
oxygen donor ligands have been shown effective, with one
compound, bis(ethylmaltolato)oxovanadium(IV), having com-
pleted phase I clinical trials.16
oxygen in maltol and ethylmaltol was replaced by sulfur,
yielding thiomaltol (Htma) and ethylthiomaltol (Hetma),
respectively. These compounds should bind to vanadium
through the thiocarbonyl sulfur and deprotonated hydroxyl
oxygen atoms, thus comprising mixed O,S donor ligands.
Two other ligands with mixed O,S donor atoms with the
N-R group substituted for the ring oxygen were also
synthesized: 3-hydroxy-2-methyl-4-pyridinethione (Hmppt)
and 3-hydroxy-1,2-dimethyl-4-pyridinethione (Hdppt). The
corresponding oxygen analogues of these compounds (i.e.,
CdO instead of CdS) are ligands that have been previously
synthesized and tested as vanadyl complexes for insulin-
like properties.10,16 The substitution of N-R in the ring
provides versatility in terms of varying the R group on the
nitrogen and, hence, changing chemical properties such as
lipophilicity, solubility, and stability of the vanadium com-
plex.
Inorganic vanadium compounds, although they are effec-
tive, have poor gastrointestinal absorption and require
relatively high doses for therapeutic efficacy.14 As shown
previously by our group,12,16 the vanadyl complexes with
maltol and ethylmaltol (approved food additives), namely,
bis(maltolato)oxovanadium(IV) (BMOV) and bis(ethylmal-
tolato)oxovanadium(IV) (BEOV), are several times more
potent than vanadyl sulfate.
The syntheses and characterizations of four mixed O,S
donor ligands and their corresponding vanadyl(IV) and
vanadium(III) complexes are described herein. Furthermore,
the results of the biological tests of the vanadium complexes
as insulin-enhancing agents as compared to BMOV are
presented.
Experimental Section
Among the many different complexes of vanadium that
have been synthesized, characterized, and biologically tested
by us8,9 and by others,10,17 there have been vanadium
complexes with mixed O,S donors and mixed O,N donors,
but so far none have proven to be substantially more potent
than BMOV or BEOV in vivo. In this work, we revisit the
maltol motif with small but significant modifications. The
aim of this study was to test the insulin-enhancing properties
of several new vanadium complexes with mixed O,S donor
ligands. The syntheses and characterization of four mixed
O,S donor ligands and their corresponding vanadyl(IV) and
vanadium(III) complexes are described here. The ketonic
Materials and Methods. The preparation and determination of
the acidity constants of the ligand precursors have been fully
described in an accompanying article.18 All solvents were reagent-
grade and were obtained from Fisher Scientific. All other chemicals
[VOSO4‚H2O, VCl3, NEt3, P2S5, NaCl, 1 M NaOH, 40% aqueous
MeNH2, NH4OH, HCl, and DTT (dithiothreitol)]were obtained from
Sigma-Aldrich or Fisher Scientific and were used as received
without further purification. Potassium biphthalate (KHP) was
obtained from Anachemia, and silica gel 230-400 mesh was
obtained from Silicycle. Water was deionized (Barnstead D8902
and D8904 Cartridges) and distilled (Hytrex II GX50-9-7/8 and
GX100-9-7/8 Cartridge filters) before use. Atomic absorption
standard (AAS) vanadyl solution for potentiometric and spectro-
photometric titrations was obtained from Sigma-Aldrich. Instru-
ment-quality (all metals <1 ppb) concentrated nitric acid was from
Seastar Chemicals Inc., Sidney, BC, Canada.
Instrumentation. Infrared spectra were recorded as KBr disks
in the range of 4000-500 cm-1 on a Galaxy Series 5000 FTIR
spectrometer and were standardized to polystyrene. Mass spectra
were obtained with a Kratos MS 50 (electron-impact ionization mass
spectrometry, EIMS), a Kratos Concept II H32Q (Cs+ liquid
secondary ion mass spectrometry, LSIMS), or a Micromass LCT
or Bruker Esquire ion trap (electrospray ionization, ESIMS)
spectrometer. Room-temperature (293 K) magnetic susceptibilities
(11) Melchior, M.; Rettig, S. J.; Liboiron, B. D.; Thompson, K. H.; Yuen,
V. G.; McNeill, J. H.; Orvig, C. Inorg. Chem. 2001, 40, 4686.
(12) Caravan, P.; Gelmini, L.; Glover, N.; Herring, F. G.; Li, H.; McNeill,
J. H.; Rettig, S. J.; Setyawati, I. A.; Shuter, E.; Sun, Y.; Tracey, A.
S.; Yuen, V. G.; Orvig, C. J. Am. Chem. Soc. 1995, 117, 12759.
(13) Crans, D. C.; Yang, L.; Jakusch, T.; Kiss, T. Inorg. Chem. 2000, 39,
4409.
(14) Thompson, K. H.; Battell, M.; McNeill, J. H. In Vanadium in the
EnVironment; Nriagu, J. O., Ed.; John Wiley: New York, 1998; Part
2, p 21.
(15) Thompson, K. H.; Orvig, C. Science 2003, 300, 936.
(16) Thompson, K. H.; Liboiron, B. D.; Sun, Y.; Bellman, K. D. D.;
Setyawati, I. A.; Patrick, B. O.; Karunaratne, V.; Rawji, G.; Wheeler,
J.; Sutton, K.; Bhanot, S.; Cassidy, C.; McNeill, J. H.; Yuen, V. G.;
Orvig, C. J. Biol. Inorg. Chem. 2003, 8, 66.
(17) Sakurai, H.; Tamura, A.; Fugono, J.; Yasui, H.; Kiss, T. Coord. Chem.
ReV. 2003, 245, 31.
(18) Monga, V.; Patrick, B. O.; Orvig, C. Inorg. Chem. 2005, 44, 2666-
2677.
Inorganic Chemistry, Vol. 44, No. 8, 2005 2679

Monga et al.
were measured on a Johnson Matthey magnetic susceptibility
balance; diamagnetic corrections were based on Pascal’s constants.
Mr. Peter Borda or Mr. Minaz Lakha in the Department of
Chemistry, University of British Columbia, performed the elemental
analyses of C, H, N, and S. Room-temperature and frozen-solution
X-band EPR spectra were recorded on a Bruker ECS-106 X-band
spectrometer. Temperature (125 or 298 K) was maintained by liquid
nitrogen flowing through a cryostat in conjunction with a Eurotherm
B-VT-2000 variable-temperature controller. The microwave fre-
quency and magnetic field were calibrated with an EIP 625A
microwave-frequency counter and a Varian E500 gaussmeter,
respectively. Computer simulations of the isotropic EPR spectra
were performed using Bruker’s WINEPR/SIMFONIA package. The
solution (298 K) spectrum of VO(dppt)2 (in MeOH) was recorded
in a 20-µL quartz capillary. Quartz tubes (4 mm i.d.) were used to
record the solution and frozen-solution spectra of VO(tma)2 and
bis(ethylthiomaltolato)oxovanadium(IV), VO(etma)2, in DMF and
bis(2-methyl-3-oxy-4-pyridinethionato)oxovanadium(IV), VO(mppt)2,
and bis(1,2-dimethyl-3-oxy-4-pyridinethionato)oxovanadium(IV),
VO(dppt)2, in MeOH/DMF (1:1). A Varian model SpectrAA 300
graphite furnace atomic absorption spectrometer (GFAAS) with
Zeeman background correction equipped with an autosampler was
used for the determination of vanadium concentrations in urine
samples of treated diabetic rats.
to the same method as employed for VO(tma)2, replacing Htma
with Hdppt. A pine-green precipitate was collected by vacuum
filtration, washed with 5 mL of cold water, and dried overnight in
vacuo to yield 42% based on V.
Method B. VOSO4‚3H2O (4.30 g, 0.020 mol) was added to a
solution of Htma (5.62 g, 0.040 mol) dissolved in 250 mL of hot
water (78 °C). The N-CH3 pyridinethione was prepared in situ by
slowly (over ∼2 h) adding 40% methylamine (17 mL, 0.20 mol).
Lowering the pH of the solution from ∼11 to ∼9 by addition of
concentrated H2SO4, dropwise, resulted in a solid precipitating from
the solution. The reaction mixture along with the precipitate was
refluxed overnight, cooled to room temperature, and vacuum filtered
to obtain a pine-green compound. The precipitate was washed twice
with 10 mL of cold water and dried overnight in vacuo. Analysis
verified the product to be identical to VO(dppt)2 synthesized by
method A, yield 5.48 g, 73% based on V. Anal. Calcd (found) for
C14H16N2O3S2V: C, 44.80 (44.87); H, 4.30 (4.42); N, 7.46 (7.51);
S, 17.08 (17.22). MS m/z (+LSIMS): 376 ([HVOL2]+), 359
([VL2]+). IR (cm-1, (4 cm-1): 962 (νVdO). Magnetic moment:
µeff ) 1.66 µB.
(v) Tris(thiomaltolato)vanadium(III), V(tma)3. VCl3 (1.42 g,
0.009 mol) was added to a solution of Htma (3.84 g, 0.027 mol) in
300 mL of degassed hot water and 50 mL of methanol at 80 °C.
The black reaction mixture was refluxed for 3 h and then cooled
to room temperature. The black, flaky precipitate formed was
collected by vacuum filtration, washed twice with 5 mL of cold
water, and dried overnight in vacuo to yield 2.77 g, 65% based on
V. Anal. Calcd (found) for C18H15O6S3V: C, 45.57 (45.67); H, 3.19
(3.11); S, 20.27 (20.21). MS m/z (EIMS): 474 ([VL3]+), 333
([VL2]+). IR (cm-1, (4 cm-1): 1580, 1500 (ring vibrations), 673
(νV-O). Magnetic moment: µeff ) 2.53 µB.
(vi) Tris(ethylthiomaltolato)vanadium(III), V(etma)3. VCl3
(0.165 g, 1.1 mmol) was added to a solution of Hetma (0.51 g, 3.3
mmol) in 10 mL of degassed water and 3 mL of MeOH at 83 °C
with Ar sparging. Triethylamine, N(Et)3 (0.45 mL, 3.2 mmol), was
added, and the solution was refluxed for 3 h and then cooled to
room temperature. A black crystalline, air-sensitive precipitate
[decomposes to VO(etma)2 in 5-6 days] was filtered under Ar,
washed with 5 mL of cold, degassed water, and dried overnight in
vacuo to yield 0.43 g, 79% based on V. Anal. Calcd (found) for
C21H21O6S3V: C, 48.83 (48.50); H, 4.10 (4.06). MS m/z (ESI-
MS): 516 ([VL3]+), 361 ([VL2]+). IR (cm-1, (4 cm-1): 1576,
1499 (ring vibrations), 673 (νV-O). Magnetic moment: µeff ) 2.52
µB.
Preparation of the Complexes. All complexation reactions and
the filtering of the vanadium(III) products were carried out under
Ar. Yields are presented for the analytically pure compounds and
were calculated on the basis of the respective vanadium starting
material for the vanadium complexes.
(i) Oxobis(thiomaltolato)vanadium(IV), VO(tma)2. VOSO4‚
3H2O (5.55 g, 0.026 mol) was added to a solution of Htma (7.27
g, 0.051 mol) in 100 mL of degassed water (78 °C). As the pH of
the solution was raised slowly (over 10 min) with 1 M NaOH to
∼5, a solid precipitated from the solution. The resulting mixture
along with the precipitate was refluxed for 3 h, cooled to room
temperature, and vacuum filtered to yield a dark olive-green
compound. The precipitate was washed with 5 mL of cold water
and dried overnight in vacuo to yield 7.6 g, 84% based on V. Anal.
Calcd (found) for C12H10O5S2V: C, 41.27 (40.99); H, 2.89 (2.85);
S, 18.36 (18.20). MS m/z (+LSIMS): 350 ([HVOL2]+), 333
([VL2]+). IR (cm-1, (4 cm-1): 974 (νVdO). Magnetic moment:
µeff ) 1.65 µB.
(ii) Bis(ethylthiomaltolato)oxovanadium(IV), VO(etma)2.
VO(etma)2 was synthesized according to the same method as
employed for VO(tma)2, replacing Htma with Hetma. A brown
precipitate was collected by vacuum filtration, washed with 5 mL
of cold water, and dried overnight in vacuo to yield 74% based on
V. Anal. Calcd (found) for C14H14O5S2V: C, 44.56 (44.16); H, 3.74
(3.71); S, 16.99 (16.75). MS m/z (+LSIMS): 378 ([HVOL2]+),
361 ([VL2]+). IR (cm-1, (4 cm-1): 972 (νVdO). Magnetic mo-
ment: µeff ) 1.71 µB.
(iii) Bis(2-methyl-3-oxy-4-pyridinethionato)oxovanadium(IV)
Trihydrate, VO(mppt)2‚3H2O. VO(mppt)2 was synthesized ac-
cording to the same method as employed for VO(tma)2, replacing
Htma with Hmppt. A grayish-brown precipitate was collected by
vacuum filtration, washed with 5 mL of cold water, and dried
overnight in vacuo to yield 78% based on V. Anal. Calcd (found)
for C12H12N2O3S2V‚3H2O: C, 35.91 (35.99); H, 4.52 (4.57); N,
6.98 (6.91); S, 15.96 (15.82). MS m/z (+LSIMS): 348 ([HVOL2]+),
331 ([VL2]+). IR (cm-1, (4 cm-1): 968 (νVdO). Magnetic mo-
ment: µeff ) 1.62 µB.
Spectrophotometric Titrations. The stepwise equilibrium con-
stants of the vanadyl complexes were determined by performing
spectrophotometric and potentiometric titrations. For the spectro-
photometric titrations, 30.00 mL of an aqueous solution containing
0.16 M NaCl, 16.2 mM [VO2+
]
[prepared by diluting a V(IV)
stock
AAS solution], and ∼0.9 mM ligand precursors was titrated with
∼0.1 M NaOH. The analytical solution was kept under constant
Ar flow in a titration vessel maintained at 25 ( 0.1 °C using a
Julabo UC circulating bath equipped with a Metrohm 6.0123.100
pH glass electrode, a Metrohm 6.0726.100 silver chloride reference
electrode, a model 665 Metrohm Dosimat autoburet, and a Metrohm
713 pH meter. An aliquot from the solution was transferred into a
cuvette (path length ) 1 cm), and the UV absorption was measured.
After measuring the absorbance, we carefully returned the aliquot
to the original solution, added titrant in 0.1-mL increments with
constant stirring, and recorded the next measurement after the pH
stabilized ((0.1 mV) . Absorbance data were corrected for dilution
factors and were fitted by nonlinear regression analysis using an
iterative procedure in the SigmaPlot 2001 program.
(iv) Bis(1,2-dimethyl-3-oxy-4-pyridinethionato)oxovanadium-
(IV), VO(dppt)2. Method A. VO(dppt)2 was synthesized according
2680 Inorganic Chemistry, Vol. 44, No. 8, 2005

Vanadium Complexes with Mixed O,S Anionic Ligands
Potentiometric Measurements. The general procedure for the
potentiometric titrations has been described in detail in an
accompanying article.18 All solutions were prepared with freshly
boiled deionized, distilled water that was purged with Ar during
boiling and cooling to ensure removal of CO2. The ionic strength
(I) of the titration solutions was kept constant with 0.16 M NaCl.
The value of pKw used at I ) 0.16 M NaCl and T ) 25 °C was
13.76.19 The titrations were controlled by a locally written Qbasic
program. The acidity constants of the ligand precursors were
calculated using the data within the range of ∼2.5 e pH e 10.5
on an IBM-compatible computer containing a Pentium II processor
using a curve-fit procedure (a Newton-Gauss nonlinear least-
squares program). The final values of the acidity constants for all
of the compounds18 were obtained from the average of at least 10
independent titrations. Potentiometric measurements of the ligand
precursors (HL) were conducted in the presence and absence of
the vanadyl ion. For the metal titrations, the ligand-to-metal ratio
was at least 3:1 to prevent hydrolysis. Stability constants of the
vanadyl complexes were determined under the same conditions as
employed for the free ligand and were calculated using the data
within the range of ∼4.5 e pH e 10.0 on an IBM-compatible
computer containing a Pentium II processor using a curve-fit
procedure (a Newton-Gauss nonlinear least-squares program). The
first equilibrium constant K110 was determined spectrophotometri-
cally and was maintained as a constant in potentiometric calculations
of K120 and K12-1. The hydrolysis model of a vanadyl system was
also included in the model.20,21 The final values of the equilibrium
constants of all of the complexes were obtained from the average
of eight independent titrations.
X-ray Crystallographic Analysis of VO(tma)2. Crystals were
mounted on a glass fiber, and measurements were made on a Bruker
X8 diffractometer with graphite monochromated Mo KR radiation.
Data were collected and integrated using the Bruker SAINT
software package.22 All data were processed and corrected for
Lorentz and polarization effects and absorption using the multiscan
technique (SADABS).23 The structure was solved by direct
methods.24 The material crystallizes with both the cis and trans
isomers in the asymmetric unit. Complete details of the X-ray
crystallographic analysis of VO(tma)2 can be found in the Sup-
porting Information.
Biological Studies. Male Wistar rats weighing 190-220 g were
cared for in accordance with the principles and guidelines of the
Canadian Council on Animal Care. The animals (two per cage)
were housed on a 12-h light/dark schedule. The cages were
maintained at 21 ( 1 °C, and the humidity was maintained at 54
( 2%. The animals were allowed ad libitum access to food (Purina
rat chow no. 5001) and tap water. Several experimental comparisons
were run sequentially to facilitate the care of the large number of
animals. VO(tma)2, V(tma)3, VO(dppt)2, and V(etma)3 were
administered by oral gavage; VO(etma)2 and VO(mppt)2 were
administered by ip (intraperitoneal) injections. In each case, after
a 5-day acclimatization period, animals were randomly assigned
to one of the following groups: control (C, nondiabetic), control
plus BMOV (CB), diabetic (D), diabetic treated with BMOV (DB),
or diabetic treated with a new thio-vanadium complex (DT); in the
last three screening assays [for V(etma)3, VO(mppt)2, and
VO(etma)2], animals were assigned to D, DB, and DT groups only.
The number of rats in each group for each study is given in the
insets of the individual graphs in the Supporting Information.
For animals in the D, DB, and DT groups, diabetes was induced
by tail vein injection of 60 mg/kg of streptozotocin (STZ, Sigma,
St. Louis, MO) in 0.9% NaCl under light halothane anaesthesia.
Diabetic status was confirmed 3 days after the STZ injection by a
glucometer (Ames Glucometer II, Miles Inc., Elkhart, Ind., and
Glucostix, Miles Canada Inc., Etobicoke, ON, Canada); blood
glucose levels greater than 13 mM were considered diabetic. For
the duration of the experiments (72 h), the plasma glucose levels
were monitored by using the Beckman Glucose Analyzer 2. Plasma
glucose levels less than 9 mM were considered normoglycemic.
Suspensions of the vanadium compounds to be tested were
prepared in a uniform volume (5 mL) of 1% carboxymethylcellulose
(CMC, 0.12 M) prepared immediately before administration. The
established ED50 dose for BMOV or BEOV was used for each
screening assay:8,16 0.6 mmol kg-1 po (per os) and 0.1 mmol kg-1
ip in a total volume of 5 mL kg-1. The control and diabetic groups
received the equivalent volume of 1% CMC alone.
The study was commenced 7 days after the STZ injection. Plasma
glucose levels were determined by collecting 100 µL of whole blood
prior to the start of treatment (time 0) and at set intervals up to 72
h following compound administration. Results were analyzed by
GLM ANOVA or two-way ANOVA followed by the Newman-
Keuls test (p < 0.05 accepted as significantly different) using the
Number Cruncher Statistical System (NCSS) and are presented as
means plus/minus the standard error of the mean (SEM) in the
graphs (see Supporting Information).
Determination of Vanadium and Vanadyl Complex Urinary
Excretion. Three STZ-diabetic male Wistar rats were administered
VO(tma)2, 0.1 mmoL kg-1 (1 mg V kg-1), ip. Immediately
following injection, the animals were placed in metabolic cages
equipped with urinary collection containers. These collection
containers were removed at the end of 2, 4, 8, 12, and 24 h; the
total volume of each urine sample was recorded; and the samples
were analyzed by positive ion mode electrospray ionization mass
spectrometry (+ESIMS) to identify the presence of VO(tma)2 or
its metabolites in the urine. The urine samples were also evaluated
for the total vanadium excreted by each animal up to 24 h after
injection. All glassware used in sample preparation was rinsed with
20% HNO3. Instrument-quality concentrated nitric acid (Seastar
Chemicals Inc., Sidney, BC, Canada; distilled in quartz, all metals
<1 ppb) and distilled, deionized water were used to prepare the
solutions and serial dilutions. The vanadium stock solution (100
ppb) was prepared by adding 10 µL of commercially available
standard vanadyl atomic absorption solution, 1020 µg mL-1 as
purchased from Sigma-Aldrich, to 990 µL of 5% HNO3 solution.
The urine samples were wet digested overnight with 50% HNO3
and then diluted as required for vanadium concentration measure-
ment by GFAAS.
Results and Discussion
(19) Martell, A. E.; Motekaitis, R. J. Determination and Use of Stability
Constants; VCH: New York, 1988; p 4.
Four neutral oxovanadium(IV) complexes, VO(tma)2,
VO(etma)2, VO(mppt)2, and VO(dppt)2, analogous to their
non-thio-containing pyrone and pyridinone analogues, have
been synthesized and characterized both physicochemically
and biologically. In each case, the ligand precursor had an
O,S binding moiety and was derived from one of two ligand
classes: 3-hydroxy-4-pyranthiones or 3-hydroxy-4-pyridine-
(20) Baes , C. F., Jr.; Mesmer, R. E. The Hydrolysis of Cations; John Wiley
& Sons: New York, 1986; p 200
(21) Chasteen, N. D. Struct. Bonding (Berlin) 1983, 53, 105.
(22) SAINT, version 6.02; Bruker AXS Inc.: Madison, WI, 1999.
(23) SADABS: Bruker Nonius Area Detector Scaling and Absorption
Correction, version 2.05; Bruker AXS Inc.: Madison, WI, 2002.
(24) SIR97: Altomare, A.; Burla, M. C.; Cammali, G.; Cascarano, M.;
Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.;
Spagna, A. J. Appl. Crystallogr. 1999, 32, 115.
Inorganic Chemistry, Vol. 44, No. 8, 2005 2681

Monga et al.
Scheme 1. Syntheses of the Oxovanadium(IV) Complexes (1:2
VO2+/HL) in H2O, ∼80 °C, pH ∼5
donating group, and the anionic hydroxyl oxygen. Metal
complexation to the pyranthione anion tma- has an electron-
withdrawing effect on the ring. The electropositive metal ion
pulls electron density away from oxygen and the ring, aiding
the nucleophilic attack by the amine on the ring. This one-
pot synthesis avoids the extra steps involved in isolating
Hdppt and yields VO(dppt)2 directly from Htma. The
procedure was easier and more efficient than the metathesis
method and gave a significantly higher overall yield (73%
vs 42%).
New V(III) complexes with pyranthiones, V(tma)3 and
V(etma)3, were synthesized in good yields (65-79%) from
VCl3 and the respective ligand precursors (Scheme 3).
Because of unavoidable partial hydrolysis/methanolysis of
VCl3, the complexation reaction did not yield the products
in 100% yield. For V(tma)3, the black solid precipitated
almost immediately upon addition of VCl3 to the solution,
but the reaction mixture was refluxed for 3 h and cooled
before filtering to maximize the yield. The synthesis of
V(etma)3 was similar, apart from the addition of a base (NEt3)
to precipitate the pure complex. Synthesis of the vanadium-
(III) complexes of the pyridinethiones, V(mppt)3 and V(d-
ppt)3, was also attempted in a fashion analogous to that for
V(etma)3 (Scheme 3). It was found that, despite various
conditions employed (i.e., temperatures, solvents, and bases),
the solid product that precipitated was the oxidized vanadyl
complex (VOL2). The vanadium(III) complex remained
dissolved in the solvent. Attempts were made to isolate the
pure complex by extracting the filtrant with CH2Cl2. Black
solids containing V(mppt)3 or V(dppt)3 could be isolated,
but in poor yields (11-21%) with some inseparable impuri-
ties and suspect elemental analyses. Any further attempts to
recrystallize the complexes resulted in oxidation of the
oxovanadium(IV) complexes. Similarly to their oxygen
counterparts, the thio-V(III) complexes were also found to
be air-sensitive to varying extents. V(tma)3 was found to be
the most stable (months), whereas V(etma)3 oxidized in 2-4
days and V(mppt)3 and V(dppt)3 decomposed within 24 h.
Previously, vanadium(III) complexes of the corresponding
pyrones and pyridinones have been prepared by in situ
reduction of the vanadyl complexes with sodium dithion-
ite.11,28 For the thio-vanadyl complexes, in situ reduction of
vanadyl to vanadium(III) could not be achieved despite
several attempts with different reducing agents and solvents.
Because the vanadyl complexes could be neither oxidized
nor reduced easily under various conditions, it appears that
these sulfur-substituted complexes are very stable under
ambient conditions.
thiones. The synthesis and complete characterization, includ-
ing acidity constants and X-ray crystal structures, of the
ligand precursors have been reported in an accompanying
article.18
Unlike their oxygen counterparts, all of the thio-vanadyl
complexes synthesized in this work were found to be stable
to oxidation both in the solid state and in solution in various
solvents (i.e., water, alcohols, CH3CN, CH2Cl2, toluene, and
THF). Of the four complexes, VO(dppt)2 was found to have
the lowest solubility.
Synthesis of the vanadyl complexes was accomplished
readily by adding vanadyl sulfate to aqueous solutions of
the corresponding ligand precursor in a 1:2 ratio, adjusting
the pH to ∼5 (with NaOH), and refluxing the resulting
solution for 3 h (Scheme 1). The reactions were carried out
at high temperatures (∼80 °C) because of the low solubilities
of the ligand precursors in water. The product that precipi-
tated from each reaction mixture was collected by vacuum
filtration, washed with cold water, and dried overnight in
vacuo. VO(tma)2, VO(etma)2, VO(mppt)2, and VO(dppt)2
were obtained in high yields (42-84%) as olive-green, dark
brownish-green, brown, and pine-green solids, respectively.
Interestingly, for VO(dppt)2, sometimes a light-gray hydrated
form of the complex was isolated [VO(dppt)2‚H2O].
An additional one-pot method was also used to synthesize
VO(dppt)2 (Scheme 2). The original method was similar to
that of the other VOL2 complexes, reacting Hdppt with
VOSO4 (Scheme 1), and while this work was ongoing, there
was a report of the synthesis of VO(dppt)2 from VOSO4 and
Hdppt.25 The one-pot method (vide infra) is more efficient
(Scheme 2); it was adapted from the previously reported
syntheses by our group26,27 wherein various metal ions (VO2+,
Al3+, Ga3+, and In3+) were complexed with Hdpp (3-
hydroxy-1,2-dimethyl-4-pyridinone), the oxygen analogue of
Hdppt. In this method, VOSO4 is added to a solution of
Htma, the precursor of Hdppt. The metal-pyranthione
complex is formed in situ, and addition of MeNH2 to this
solution converts the bound pyranthione to the N-CH3
pyridinethione. The conversion of a pyranthione to a pyridi-
nethione involves nucleophilic attack by the primary amine
at C-2 followed by the elimination of water and ring closure.
The nucleophilic attack is hindered by increased electron
density at C-2 because of the CH3 substituent, an electron-
All of the complexes were characterized by elemental
analysis, electrospray ionization (ESI) or liquid secondary
ion (LSI) mass spectrometry, magnetic moments, and IR and
EPR [for V(IV)] spectroscopies. Elemental analyses of all
of the complexes correlated well with the calculated values.
For VO(mppt)2, the complex was always isolated as a
trihydrate, indicating the increased hydrogen-bonding ability
of the N-H protons in the ring. For the N-CH3 pyridine-
(25) Katoh, A.; Tsukahara, T.; Saito, R.; Ghosh, K. K.; Yoshikawa, Y.;
Kojima, Y.; Tamura, A.; Sakurai, H. Chem. Lett. 2002, 1, 114.
(26) Zhou, Y. M.Sc. Thesis, University of British Columbia, Vancouver,
BC, Canada, 1993; p 8.
(27) Zhang, Z.; Hui, T. L. H.; Orvig, C. Can. J. Chem. 1989, 67, 1708.
(28) Dilli, S.; Patsalides, E. Aust. J. Chem. 1976, 29, 2389.
2682 Inorganic Chemistry, Vol. 44, No. 8, 2005

Vanadium Complexes with Mixed O,S Anionic Ligands
Scheme 2. One-Pot Synthesis of VO(dppt)2 from VOSO4, Htma, and 40% MeNH2
Scheme 3. Syntheses of Tris(ligand)vanadium(III) Complexes
The experimental magnetic moments of the four vanadyl
complexes were found to be in the range of 1.62-1.71 µB
at room temperature, which is consistent with the value
predicted from the spin-only formula (1.73 µB)31 of a
vanadium(IV) system. These values are slightly lower than
those of their corresponding oxygen analogues (∼1.72-1.77
µB)12,16,26 but are consistent with the presence of a single
unpaired electron and are within the range of the values of
other similar VO(O2S2) complexes (∼1.69 µB).30,32 The spin-
only magnetic moment that is expected for vanadium(III)
complexes is 2.83 µB.31 The room-temperature magnetic
moments of the pyranthione-V(III) complexes were found
to be 2.52-2.53 µB. These values are typical of octahedral
vanadium(III) complexes33,34 with two unpaired electrons and
are in the same range as the corresponding oxygen analogues
(2.5-2.7 µB). Because of the small amounts of products
obtained, the magnetic moments of the pyridinethione-
vanadium(III) complexes could not be obtained.
thione, the oxovanadium complex was isolated as a pine-
green solid for VO(dppt)2 and as a light-gray complex for
the monohydrate VO(dppt)2‚H2O. The monohydrate could
not be converted to the anhydrate even after extensive
heating.
Positive ion detection mode mass spectrometry was used
to analyze all of the complexes, and diagnostic mass spectra
were obtained with the expected isotopic patterns. The most
intense peak for the vanadyl complexes corresponded to
[VL2]+; other major peaks in the spectra were the protonated
parent-ion peak, [HVOL2]+, and its fragment, [VOL]+. A
characteristic small peak corresponding to [VL3]+ was also
usually observed for vanadyl complexes. The mass spectra
of the vanadium(III) complexes were dominated by the
parent-ion peak, [VL3]+, and the characteristic [VL2]+ peak.
The absence of the [HVOL2]+ peak from the spectra of V(III)
complexes confirms that the complex has not been oxidized
to the V(IV) species, even in the instrument. A small peak
corresponding to the [V2L5]+ ion, characteristic of tris-
(bidentate ligand) trivalent metal ion complexes, was also
observed for some of the complexes.11
Diagnostic shifts in the ligand stretches of the IR spectra
of both the oxovanadium(IV) and vanadium(III) complexes
were observed. In the oxovanadium(IV) and vanadium(III)
complexes, the pyranthione vibrations undergo a significant
bathochromic shift of ∼40-50 cm-1, whereas the pyridi-
nethione peaks shift by only ∼10-23 cm-1 to lower energy.
This bathochromic shift was also seen for the oxygen
analogues.11,26 The coordination of the ligand to vanadium
was confirmed by the disappearance of the strong νCdS band
and the appearance of a new peak around ∼673-681 cm-1
for V-O stretches. The most characteristic difference
between the IR spectra of the vanadyl(IV) and vanadium-
(III) complexes is the vanadyl stretch νVdO at 985 (50 cm-1
in the vanadyl(IV) complexes. The absence of a νVdO stretch
in the IR spectra of the vanadium(III) complexes confirmed
the lack of oxidation [of V(III) to VO2+]. For these thio-
vanadyl complexes, the νVdO stretch was observed between
962 and 974 cm-1, which is consistent with other similar
vanadyl complexes containing an O2S2 coordination sphere
(∼951-965 cm-1).29,30 Compared to the oxygen analogues,
these complexes had lower νVdO stretching frequencies, as
expected because of the replacement of oxygen with the less-
electronegative and heavier sulfur atom.
Room-temperature EPR spectra of the vanadyl complexes
show the expected eight-line pattern characteristic of a d1
V(IV) system. The isotropic EPR spectrum (298 K) of VO-
(tma)2 was obtained in toluene (giso ) 1.974, Aiso ) 85.0 ×
10-4 cm-1), and the parameters are remarkably similar to
the values reported previously for VO(tma)2 in CHCl3 (giso
) 1.976, Aiso ) 83.0 × 10-4 cm-1).35 The frozen-solution
X-band EPR spectra of all four thio-vanadyl complexes were
recorded. Figure 1 shows the anisotropic EPR spectrum (125
K) of VO(mppt)2 in MeOH/DMF (1:1). DMF was used
because of the limited solubility of the complex in MeOH
alone. The g and Az values of all four compounds (Table 1)
are within the ranges typical for an O2S2 coordination sphere
around a vanadyl ion.36,37 As expected, the g values of the
thio compounds are slightly higher than those of their oxygen
counterparts (giso ≈ 1.96-1.98), but the A values are lower
[Aiso ≈ (90-96) × 10-4 cm-1].16,38 Evidence of cis/trans
isomers was seen in the anisotropic spectra, but it was
unnecessary to include both isomers to obtain an acceptable
simulation fit.
(29) Challen, P. R.; McConville, D. B.; Youngs, W. J. Acta Crystallogr.
C:Cryst. Struct. Commun. 2000, 56, 310.
(30) Hodge, A.; Nordquest, K.; Blinn, E. L. Inorg. Chim. Acta 1972, 6,
491.
(31) Rodgers, G. E. Introduction to Coordination, Solid State and Descrip-
tiVe Inorganic Chemistry; McGraw-Hill: New York, 1994; p 76.
(32) Wen, T. B.; Shi, J. C.; Huang, X.; Chen, Z. N.; Liu, Q. T.; Kang, B.
S. Polyhedron 1998, 17, 331.
(33) Figgis, B. N.; Lewis, J.; Mabbs, F. E. J. Chem. Soc. 1960, 2480.
(34) Machin, D. J.; Murray, K. S. J. Chem. Soc. 1967, 1498.
(35) Bechmann, W.; Uhlemann, E. Z. Anorg. Allg. Chem. 1987, 544, 215.
(36) Tasiopoulos, A. J.; Troganis, A. N.; Deligiannakis, Y.; Evangelou,
A.; Kabanos, T. A.; Woollins, J. D.; Slawin, A. J. Inorg. Biochem.
2000, 79, 159.
(37) Chasteen, N. D. In Biological Magnetic Resonance; Berliner, L.,
Rueben, J., Eds.; Plenum Press: New York, 1981; Vol. 3, p 53.
(38) Rangel, M. Transition Met. Chem. 2001, 26, 219.
Inorganic Chemistry, Vol. 44, No. 8, 2005 2683

Monga et al.
Figure 1. Frozen-solution EPR spectrum of VO(mppt)2 in MeOH/DMF, experimental (dark gray) and simulated (light gray). (A and B) Expansions of the
vanadium hyperfine resonances associated with the principal z axis in the spectrum (T ) 125 K, ν ) 9.560 GHz).
Table 1. Spin Hamiltonian Parameters of the Vanadyl Complexes (DMF, T ) 125 K, g ( 0.001, A ( 0.1 × 10-4 cm-1
complex
gx
gy
gz
gavg
Ax
Ay
Az
Aav
ν (GHz)
VO(tma)2
VO(etma)2
VO(mppt)2
1.982
1.983
1.985
1.986
1.982
1.983
1.981
1.979
1.954
1.950
1.958
1.957
1.973
1.972
1.975
1.974
-52.0
-50.0
-42.0
-45.2
-51.5
-52.0
-52.0
-51.3
-153.0
-150.2
-151.0
-151.0
-85.5
-84.1
-81.7
-82.5
9.568
9.565
9.560
9.567
a
a
VO(dppt)2
a MeOH/DMF (1:1), gav ) (gx + gy + gz)/3, Aav ) (Ax + Ay + Az)/3.
The solution equilibria of vanadyl (VO2+) with each of
the four thio ligands were obtained by potentiometric and
spectrophotometric titrations. The acidity constants of the
ligands have been described in an accompanying article.18
The stepwise stability constants of the 1:1 and 1:2 metal-
to-ligand species (K110 and K120) and the overall stability
constant (â120) were determined as defined in eqs 1-5.
(M/L ≈ 12:1) were used, and the absorbance changes were
monitored between 290 and 310 nm as the pH was varied
from ∼2.4 to 3.8 (at pH > 4, metal hydrolysis can occur).
The high metal-to-ligand ratios were used to ascertain that
the absorbance changes are due to only the [(VO)L]+ species
(Figure 2).
As the ligand is deprotonated and chelated to the vanadyl
ion, the absorbance at ∼355 nm decreases, and the absor-
bance at ∼300 nm increases. The experimental spectropho-
tometric data were fitted by nonlinear regression analysis.
By using mass balance equations, the pH of the solution,
and the original analyte concentrations and constants (i.e.,
Ka and Kh), a theoretical data set could be generated. The
resultant mass balance equation in terms of K110 and
[(VO)L]+ is shown in eq 6 (charges have been omitted for
clarity).
Ka
HL y\
z H+ + L-
(1)
(2)
(3)
K110
VO2+ + L- y z [(VO)L]+
K120
[(VO)L]+ + L- y z (VO)L2
â120
VO2+ + 2L- y z (VO)L2
(4)
(5)
K12-1
(VO)L2 + OH- y z [(VO)L2(OH)]-
Kh
KaKh +
+ Ka[H] + 1
The 1:1 vanadyl/pyranthione (Htma and Hetma) species
form significantly even at pH less than 2; therefore, variable-
pH spectrophotometry was used to determine the 1:1 stability
constant (log K110) accurately. High metal-to-ligand ratios
(
)
[H]
K110
)
(6)
[VO]tLt
[(VO)L]
- [VO]t - Lt + [(VO)L]
2684 Inorganic Chemistry, Vol. 44, No. 8, 2005

Vanadium Complexes with Mixed O,S Anionic Ligands
Table 2. Stability Constants of the Species Formed between V(IV) (as
VO2+) and the Various (Pro)Ligands HLa
Htma
Hetma
Hmppt
Hdppt
b
pKa
8.12(1)
8.04(2)c
5.86(2)
13.9(2)
6.39(4)
8.20(1)
8.07(2)c
6.53(1)
14.6(1)
5.9(1)
8.92(1)
9.2(1)
6.9(1)
16.1(1)
5.53(5)
9.44(1)
7.95(2)
5.58(8)
13.53(8)
5.03(4)
log K110
log K120
log â120
log K12-1
a Potentiometric titrations, I ) 0.16 M NaCl, 25 °C. Values in parentheses
refer to the error (3σ) in the last digit. b Reference 18. c Spectrophotometric
titrations.
fitting of the theoretical absorbance values to the experi-
mental data. Five wavelengths with 14 data points in the
range of 2.4 e pH e 3.8 were fitted with calculated curves,
and the averages of at least three separate trials were taken
to obtain the final log K110 values of VO(tma)+ and
VO(etma)+. Figure 2 (bottom) shows the experimental data
(symbols) and calculated fit (solid lines) for one of the trials
of the vanadyl thiomaltolato system.
For the pyranthiones, log K110 values were fixed for
working up the potentiometric titration data to obtain K120
and K12-1. For the pyridinethiones, Hmppt and Hdppt, both
log K110 and log K120 were obtained from potentiometric
titrations. For the VO2+/Hmppt system, most of [(VO)L]+
did form at pH less than 2, but spectrophotometric titrations
could not be done with this system because of problems with
ligand dimerization. The pyridinethiones have a tendency
to tautomerize to the thiol form; in the presence of water
and molecular oxygen, two adjacent molecules in their thiol
forms can oxidatively dimerize to form a disulfide bond.
Because there is a higher concentration of the ligand
precursor in spectrophotometric titrations (as compared to
potentiometric titrations), the UV-vis spectrum indicated
dimerization before the ligand precursor could be deproto-
nated and complexed to oxovanadium(IV). Therefore, the
values given here for the VO2+/Hmppt system are the best
possible estimates from potentiometric titrations.
Figure 2. (Top) Variable-pH UV spectrophotometric titration of VO2+
/
Htma ([VO2+] ) 1.08 mM, [Htma] ) 0.09 mM) at 25 °C and I ) 0.16 M
NaCl. (Bottom) Experimental (data points) and calculated (lines) absorbance
values vs pH for the determination of log K110 at different wavelengths.
Ka is the acidity constant of the HL ligand, [VO]t is the total
concentration of vanadyl, Lt is the total concentration of the
ligand, [H] is the acid concentration derived from the pH of
the solution, [(VO)L] is the concentration of the 1:1 vanadyl/
pyranthione species at a given pH, and Kh ) 2.14 × 10-6 is
the first hydrolysis constant of VO2+ (eq 7).
[(VO)OH][H]
Kh )
(7)
[VO]
Rewriting eq 6, the equation can be solved for [(VO)L]
in terms of K110 using the quadratic formula (eq 8).
The titrations were carried out at 3 < pH < 11 (25 °C).
Vanadyl hydrolysis was included in the model using
constants from the literature.20 As seen in the oxygen
analogues,12,16 a hydroxo complex, [(VO)L2(OH)]- [incor-
porating a deprotonated water molecule coordinated to
(VO)L2], was also formed at higher pH in addition to the
[(VO)L]+ and (VO)L2 species. The stepwise formation
constants, log K110, log K120, and log K12-1, as well as the
overall stability constants, log â120, of all four thio-vanadyl
systems are listed in Table 2.
-B ( B2 - 4(K110)(K110LtMt)
x
[(VO)L] )
(8)
2K110
where
B ) K110Lt + K110Mt + KaKh +
Kh
+ Ka[H] + 1
(
)
[H]
The concentration of (VO)L can be calculated from the
Beer-Lambert law: [(VO)L]+ ) abs/ꢀ. Therefore, iterating
K110 and ꢀ as a function of pH results in the least-squares
Figure 3 shows the speciation diagrams of the VO2+/Hetma
and VO2+/Hdppt systems in the 2 < pH < 10 range. These
diagrams were calculated for a VO2+/L ratio of 1:3 and were
Inorganic Chemistry, Vol. 44, No. 8, 2005 2685

Monga et al.
Figure 4. ORTEP diagrams of cis- and trans-VO(tma)2 (50% thermal
ellipsoids).
Table 4. Selected Bond Lengths (Å) in VO(tma)2
cis-VO(tma)2
V(1)-O(5)
trans-VO(tma)2
V(2)-O(10)
1.614(5)
1.958(5)
1.936(5)
2.409(3)
2.403(3)
1.698(8)
1.712(8)
1.331(9)
1.315(9)
1.403(10)
1.419(11)
1.610(5)
1.949(5)
1.954(5)
2.432(3)
2.429(3)
1.686(8)
1.715(7)
1.328(9)
1.325(9)
1.429(11)
1.410(10)
V(1)-O(1)
V(1)-O(3)
V(1)-S(1)
V(1)-S(2)
C(1)-S(1)
C(7)-S(2)
C(2)-O(1)
C(8)-O(3)
C(1)-C(2)
C(7)-C(8)
V(2)-O(6)
V(2)-O(8)
V(2)-S(3)
V(2)-S(4)
C(13)-S(3)
C(19)-S(4)
C(14)-O(6)
C(20)-O(8)
C(13)-C(14)
C(19)-C(20)
Table 5. Selected Bond Angles (deg) in VO(tma)2
cis-VO(tma)2 trans-VO(tma)2
Figure 3. Speciation diagrams of VO2+/etma (top) and VO2+/dppt (bottom)
as a function of percent VO2+ versus pH (I ) 0.16 M NaCl, T ) 25 °C,
ligand/metal ratio g 3:1).
S(2)-V(1)-O(5)
106.8(2)
S(4)-V(2)-O(10)
S(4)-V(2)-O(8)
S(4)-V(2)-O(6)
S(3)-V(2)-O(10)
S(3)-V(2)-O(8)
S(3)-V(2)-O(6)
S(4)-V(2)-S(3)
105.9(2)
S(2)-V(1)-O(3)
S(2)-V(1)-O(1)
S(1)-V(1)-O(5)
S(1)-V(1)-O(3)
S(1)-V(1)-O(1)
S(2)-V(1)-S(1)
83.58(17)
144.48(17)
106.3(2)
144.43(18)
82.61(17)
86.73(9)
82.15(17)
87.88(18)
107.4(2)
85.90(17)
82.78(18)
146.70(10)
Table 3. Comparison of the Overall Stability Constants of the
Thio-Vanadyl Complexes and Their Oxygen Analogues
log â120
ref
VO(tma)2
VO(ma)2
13.9(2)
16.31(3)
14.6(1)
16.41(3)
16.1(1)
13.53(8)
22.25(8)
this work
12
this work
16
this work
this work
38
O(5)-V(1)-O(3) 109.3(3)
O(5)-V(1)-O(1) 108.8(3)
O(10)-V(2)-O(6) 109.9(2)
O(10)-V(2)-O(8) 107.9(2)
VO(etma)2
VO(ema)2
VO(mppt)2
VO(dppt)2
VO(dpp)2
O(3)-V(1)-O(1)
V(1)-S(1)-C(1)
V(1)-S(2)-C(7)
85.7(2)
95.7(3)
95.8(3)
O(6)-V(2)-O(8)
V(2)-S(3)-C(13)
V(2)-S(4)-C(19)
142.2(2)
96.4(3)
96.7(3)
V(1)-O(1)-C(2) 119.7(5)
V(1)-O(3)-C(8) 122.1(5)
V(2)-O(6)-C(14) 121.3(5)
V(2)-O(8)-C(20) 122.0(5)
increasing drug potency. This can be proven only after a
detailed study of the mechanism of action of these complexes
in vivo.
based on the stability constants listed in Table 2. The
diagrams clearly indicate that the dominant species at
physiological pH ≈ 7 are (VO)L2 and [(VO)L2(OH)]-. For
the biological activity of a complex, it is desired that the bis
complex species stays intact at high pH, a result seen in the
speciation diagrams. The anionic hydroxo complex could,
however, hinder the cellular absorption of the complex.
An X-ray crystal structure of VO(tma)2 was also obtained.
Green needle-shaped crystals of VO(tma)2 suitable for X-ray
diffraction were grown by mixing warm solutions of
VO(acac)2 in MeOH and Htma in THF (VO2+/Htma in a
1:2 ratio) followed by slow evaporation of the resulting
solution. The material crystallizes with both the cis and trans
(1:1) isomers in the asymmetric unit (Figure 4). To our
knowledge, this is the first example of a trans isomer of a
square-pyramidal vanadyl complex with an O2S2 coordination
sphere. All of the other similar complexes have been found
to form only the cis geometry in the solid state.29
Compared to those of their oxygen analogues, the overall
stability constants of the thio-vanadyl complexes are lower
(Table 3). This was expected because vanadium is a hard
acid and sulfur is a soft donor. Although the mixed O,S
bidentate ligands are strong chelators (based on log â120
values), they are weaker than the corresponding O,O biden-
tate ligands. Because free vanadyl might be responsible for
the insulin-enhancing effect, a slightly weaker coordinated
ligand might release the metal ion at the appropriate site more
easily than the more strongly chelating ligand, hence
Selected bond lengths and angles of both isomers are listed
in Tables 4 and 5. In both isomers, the two thiomaltolato
ligands are arranged around the base of the square pyramid
with the VdO linkage perpendicular. The vanadium atom
2686 Inorganic Chemistry, Vol. 44, No. 8, 2005

Vanadium Complexes with Mixed O,S Anionic Ligands
(tma)+ were observed, suggesting that the complex was not
excreted intact but was absorbed and rapidly decomposed
before excretion.
Figure S4 shows the plot of amount of V (in milligrams)
per sample and percent administered dose (%AD) versus time
(in hours) for rat 2 as an example with a clear pattern of
excretion (total %AD excreted by the other two rats was
<5% each at 24 h). For rat 2, the amount of vanadium
excreted (∼30%) with respect to the dose administered was
comparable to that reported in other studies in which
vanadium excretion was estimated.39-41 Sabbioni and Ma-
rafante studied the metabolic patterns of vanadium in rats
for 21 days.41 Measuring the excretion of vanadium in urine
(and feces), they found that ∼32% of the iv (intravenous)
administered dose of vanadium was excreted in 2-3 days.
Heinemann and co-workers studied the pharmacokinetics of
vanadium in five healthy volunteers and found that 52% of
the (iv) administered dose of vanadium was recovered in
the urine after 12 days.42 The data for these studies suggests
that the peak excretion occurs only after ∼24 h. Low
vanadium recovery precluded any definitive pharmacokinetic
profiling; however, the appearance of ligand-precursor peaks,
with no apparent intact complex, suggests that rapid dis-
sociation of the complex following injection had occurred.
While the present work was in progress, Sakurai and co-
workers25 reported positive glucose-lowering results of an
in vivo trial of VO(dppt)2 administered ip (10 mg V d-1) to
STZ-diabetic rats for 21 d, suggesting that, given a long
enough treatment period, this compound might yield thera-
peutically relevant relief of diabetic symptomatology.
Figure 5. Plasma glucose levels in two treated groups of STZ rats
compared to a diabetic group. (For all other DT groups, please see the
Supporting Information.)
is slightly displaced from the basal plane [V(1) ) 0.656(3)
Å, V(2) ) 0.664(2) Å]. The VdO bond distance is the same
in both isomers (cis, trans 1.61 Å), and the value for the cis
isomer of VO(tma)2 matches the VdO distances that were
reported for other cis-VO(O2S2) structures {e.g., VdO is
1.691(2) Å for [VO(mmp)2]2-, where mmpH2 ) 2-mercapto-
4-methylphenol}.29 Compared to the structure of its oxygen
analogue, VO(ma)2, the VdO bond length in VO(tma)2 is
longer [1.597(7) Å) in VO(ma)2], as suggested by the IR
data (vide supra).
Biological Studies
None of the treated rats showed any overt signs of toxicity
(noted principally as gastrointestinal distress in other studies),
nor were there significant changes in body weight over the
course of the 72-h screening, either by oral gavage or ip.
The only significant glucose lowering observed was in
BMOV-treated rats; all of the others showed little or no effect
of treatment up to the 24-h time point (Figure 5). No further
changes were noted for the duration of the screening assay.
As compared to the diabetic rats and diabetic rats treated
with BMOV, these thio-vanadium complexes did not produce
significant glucose lowering (p < 0.05) under these condi-
tions (Figures S1-S3).
During the acute screening trials, however, it was noted
that, for VO(tma)2, the rats excreted unusually bright yellow
urine 4 h postadministration. Because the ligand precursor,
thiomaltol, is a bright-yellow compound, it was predicted
that exceedingly fast decomposition of the complex and faster
excretion of vanadium (as compared to that from BMOV)
in the urine might be the reason that no significant glucose
lowering was observed for the vanadium complexes with
thio-substituted ligands. To test the hypothesis that thio
substitution promoted rapid excretion of both vanadium and
the ligand precursor, we measured urinary vanadium excre-
tion (GFAAS) over a 72-h period in three treated rats and
also analyzed urine samples by mass spectrometry (ESIMS).
Electrospray mass spectrometry was used to determine the
molecular weight of the major components of the urine
samples. The major peak in all of the spectra corresponded
to thiomaltol (m/z ) 143, [M + 1]+) with the expected
isotope pattern. No peaks corresponding to VO(tma)2 or VO-
Conclusions
Four vanadium(IV) and vanadium(III) complexes were
synthesized with each of four mixed O,S donor ligands, two
pyranthiones and two pyridinethiones. Although the synthesis
of the vanadyl complexes was uncomplicated, the vanadium-
(III) complexes were air-sensitive, and purification of the
product was difficult in some cases. At physiological pH,
the dominant forms of all four vanadyl complexes are
(VO)L2 and (VO)L2(OH), as expected (and desired). A
crystal structure determination of VO(tma)2 elucidated both
the cis and trans isomers in the same asymmetric unit, which
was not heretofore observed for vanadyl complexes with
mixed O,S ligation. No significant glucose lowering by thio-
vanadyl or V(III) complexes was apparent. Hence, for this
model, sulfur substitution did not increase the efficacy of
the vanadium complexes as insulin-enhancing agents as
compared to BMOV.
Acknowledgment. The authors thank the Canadian
Institutes of Health Research for financial support and Dr.
Bert Mueller, in the Department of Earth and Ocean
Sciences, UBC, for his expert assistance in the operation and
(39) McNeill, J. H.; Yuen, V. G.; Hoveyda, H. R.; Orvig, C. J. Med. Chem.
1992, 35, 1489.
(40) Cremer, K. D.; Cornelis, R.; Strijckmans, K.; Dams, R.; Lameire, N.;
Vanholder, R. J. Inorg. Biochem. 2002, 90, 71.
(41) Sabbioni, E.; Marafante, E. Bioinorg. Chem. 1978, 9, 389.
(42) Heinemann, G.; Fichtl, B.; Vogt, W. J. Clin. Pharmacol. 2003, 55,
241.
Inorganic Chemistry, Vol. 44, No. 8, 2005 2687

Monga et al.
Supporting Information Available: Crystallographic data for
VO(tma)2, bond lengths of cis- and trans-VO(tma)2, bond angles
of VO(tma)2, and plasma glucose levels of the DT groups. This
material is available free of charge via the Internet at http://
pubs.acs.org.
maintenance of the graphite furnace atomic absorption
spectrometer. The authors also acknowledge the following
people for their help with this work: Dr. Barry Liboiron for
spectrophotometric titrations and EPR spectroscopy, Dr. Song
Bin for potentiometric titrations, Tim Storr for EPR spec-
troscopy, and Robert An for crystal growing.
IC0486926
2688 Inorganic Chemistry, Vol. 44, No. 8, 2005