Chromium(V) glutathione complexes are among the likely reactive intermediates in Cr(VI)-induced genotoxicity and carcinogenicity. The first definitive structure of one such complex, [Cr(V)O(LH(2))(2)](3)(-) (I; LH(5) = glutathione = GSH), isolated from the reaction of Cr(VI) with excess GSH at pH 7.0 (O'Brien, P.; Pratt, J.; Swanson, F. J.; Thornton, P.; Wang, G. Inorg. Chim. Acta 1990, 169, 265-269), has been determined by a combination of electrospray mass spectrometry (ESMS), X-ray absorption spectroscopy (XAS), EPR spectroscopy, and analytical techniques. In addition, Cr(V) complexes of GSH ethyl ester (gamma-Glu-Cys-GlyOEt) have been isolated and characterized by ESMS, and Cr(III) products of the Cr(VI) + GSH reaction have been isolated and characterized by ESMS and XAS. The thiolato and amido groups of the Cys residue in GSH are responsible for the Cr(V) binding in I. The Cr-ligand bond lengths, determined from multiple-scattering XAFS analysis, are as follows: 1.61 A for the oxo donor; 1.99 A for the amido donors; and 2.31 A for the thiolato donors. A significant electron withdrawal from the thiolato groups to Cr(V) in I was evident from the XANES spectra. Rapid decomposition of I in aqueous solutions (pH = 1-13) occurs predominantly by ligand oxidation with the formation of Cr(III) complexes of GSH and GSSG. Maximal half-lives of the Cr(V) species (40-50 s at [Cr] = 1.0 mM and 25 degrees C) are observed at pH 7.5-8.0. The experimental data are in conflict with a recent communication (Gaggelli, E.; Berti, F.; Gaggelli, N.; Maccotta, A.; Valensin, G. J. Am. Chem. Soc. 2001, 123, 8858-8859) on the formation of a Cr(V) dimer as a major product of the Cr(VI) + GSH reaction, which may have resulted from misinterpretation of the ESMS and NMR spectroscopic data.

We aim to divide the problem space of fine-grained recognition into some specific regions. To achieve this, we develop a unified framework based on a mixture of experts. Due to limited data available for the fine-grained recognition problem, it is not feasible to learn diverse experts by using a data division strategy. To tackle the problem, we promote diversity among experts by combing an expert gradually-enhanced learning strategy and a Kullback-Leibler divergence based constraint. The strategy learns new experts on the dataset with the prior knowledge from former experts and adds them to the model sequentially, while the introduced constraint forces the experts to produce diverse prediction distribution. These drive the experts to learn the task from different aspects, making them specialized in different subspace problems. Experiments show that the resulting model improves the classification performance and achieves the state-of-the-art performance on several fine-grained benchmark datasets.

A new Cr(V) complex, K[CrVO(qaH3)2].H2O (Ia; qaH3 = quinato = (1R,3R,4R,5R)-1,3,4,5-tetrahydroxycyclohexanecarboxylato(2-)), synthesized by the reaction of K2Cr2O7 with excess qaH5 in MeOH (Codd, R.; Lay, P. A. J. Am. Chem. Soc. 1999, 121, 7864-7876), has been characterized by microanalyses, electrospray mass spectra, and UV-visible, CD, IR, EPR, and X-ray absorption spectroscopies. This complex is of interest because of its ability to act as both a structural and a biomimetic model for a range of Cr(V) species believed to be generated in vivo during the intracellular reduction of carcinogenic Cr(VI). The Na+ analogue of Ia (Ib) has also been isolated and characterized by microanalyses and IR and X-ray absorption spectroscopies. The reaction of Cr(VI) with MeOH in the presence of qaH5 that leads to I proceeds via a Cr(IV) intermediate (observed by UV-visible spectroscopy), and a mechanism for the formation of I has been proposed. DMF or DMSO solutions of I are stable for several days at 25 degrees C, while I in aqueous solution (pH = 4) disproportionates to Cr(VI) and Cr(III) in minutes. The likely structures in the solid state for Ia (14 K) and Ib (approximately 293 K) have been determined using both single-scattering (Ia,b) and multiple-scattering (Ia) analyses of XAFS data. These analyses have shown the following: (i) In agreement with the results from the other spectroscopic techniques, the quinato ligands are bound to Cr(V) by 2-hydroxycarboxylato moieties, with Cr-O bond lengths of 1.55, 1.82, and 1.94 A for the oxo, alcoholato, and carboxylato O atoms, respectively. (ii) The position of an oxo O atom is somewhat disordered. This is consistent with molecular mechanics modeling of the likely structures. The XAFS, EPR, and IR spectroscopic evidence points to the existence of hydrogen bonds between the oxo ligand and the 3,4,5-OH groups of the quinato ligands in the solid state of I.

The Cr(V) products of the reaction of Cr(VI) with H2O2 were studied by EPR spectroscopy. In addition to the well-characterized tetrakis(η2-peroxo)chromate(V) complex, [Cr(O2)4]3-, with g iso = 1.9723 (A iso = 18.4 × 10-4 cm-1), three new species were observed with isotropic EPR parameters, g iso = 1.9820, g iso = 1.9798 (A iso = 16.3 × 10-4 cm-1), and g iso = 1.9764 (A iso = 18.1 × 10-4 cm-1). While [Cr(O2)4]3- is stable at high concentrations of H2O2 and in alkaline solution, the species with a signal at g iso = 1.9798 is stabilized at low relative concentrations of H2O2 and in neutral solution. The signal at g iso = 1.9764 is most prominent in weakly acidic (pH = 4−7) solutions and low relative concentrations of H2O2. Finally, the signal at 1.9820 is only minor and is apparent at low pH values and low [H2O2]. From the pH and [H2O2] dependences, and by analogy with the V(V) chemistry, the species giving rise to the signals at g iso = 1.9820, g iso = 1.9798, and g iso = 1.9764 are assigned as the oxo(η2-peroxo)chromium(V), [Cr(O)(O2)(OH2) n ]+, aquaoxobis(η2-peroxo)chromate(V), [Cr(O)(O2)2(OH2)]-, and the hydroxotris(η2-peroxo)chromate(V), [Cr(O2)3(OH)]2-, complexes, respectively. The implications of these Cr(V) peroxo species for understanding the in vitro DNA damage caused by Cr(VI) and H2O2 and the genotoxicity of carcinogenic Cr(VI) complexes are discussed.

The nature of the long-lived EPR-active Cr(V) species observed in cells and biological fluids exposed to carcinogenic Cr(VI) has been definitively assigned from detailed kinetic and spectroscopic analyses of a model reaction of Cr(VI) with p-bromobenzenethiol (RSH) in the presence or absence of cyclic 1,2-diols (LH(2)) in aprotic or mixed solvents. The first definitive structures for Cr(V) complexes with a monodentate thiolato ligand, [Cr(V)O(SR)(4)](-) (g(iso) = 1.9960, A(iso) = 14.7 x 10(-4) cm(-1)), [Cr(V)OL(SR)(2)](-) (g(iso) = 1.9854, A(iso) = (15.8-16.2) x 10(-4) cm(-1)) and [Cr(V)(O)(2)(SR)(2)](-) (g(iso) = 1.9828, A(iso) = 6.8 x 10(-4) cm(-1)) were assigned by EPR spectroscopy and electrospray mass spectrometry. The unusually low A(iso) ((53)Cr) value for the latter species is consistent with its rare four-coordinate, bis-oxido structure. The [Cr(V)OL(SR)(2)](-) species are responsible for the transient g(iso) approximately 1.986 EPR signals observed in living cells and animals treated with Cr(VI) (where RSH and LH(2) are biological thiols and 1,2-diols, respectively). For the first time, concentrations of Cr(V) intermediates formed during the reduction of Cr(VI) were determined by quantitative EPR spectroscopy, and a detailed reaction mechanism was proposed on the basis of stochastic simulations of the kinetic curves for Cr(V) species. A key feature of the proposed mechanism is the regeneration of Cr(V) species in the presence of Cr(VI) through the formation of organic free radicals, followed by the rapid reactions of the formed radicals with Cr(VI). The concentration of Cr(V) grows rapidly at the beginning of the reaction, reaches a steady-state level, and then drops sharply once Cr(VI) is spent. Similar mechanisms are likely to operate during the reduction of Cr(VI) in biological environment rich in reactive C-H bonds, including the oxidative DNA damage by Cr(V) intermediates.