Determination of amines using 2-(11H-benzo[a]carbazol-11-yl) ethyl chloroformate (BCEC-Cl) as labeling reagent by HPLC with fluorescence detection and identification with APCI/MS

A pre-column derivatization method for the sensitive determination of amines using a labeling reagent 2-(11H-benzo[a]-carbazol-11-yl) ethyl chloroformate (BCEC-Cl) followed by high-performance, liquid chromatography with fluorescence detection has been developed. Identification of derivatives was carried out by LC/APCI/MS in positive-ion mode. The chromophore of 1,2-benzo-3,4-dihydrocarbazole-9-ethyl chloroformate (BCEOC-Cl) reagent was replaced by 2-(11H-benzo[a]-carbazol-11-yl) ethyl functional group, which resulted in a sensitive fluorescence derivatizing reagent BCEC-Cl. BCEC-Cl could easily and quickly label amines. Derivatives were stable enough to be efficiently analyzed by HPLC and showed an intense protonated molecular ion corresponding m/z [M+ H](+) under APCI/MS in positive-ion mode. The collision-induced dissociation of the protonated molecular ion formed characteristic fragment ions at m/z 261.8 and m/z 243.8 corresponding to the cleavages of CH2O-CO and CH2-OCO bonds. Studies on derivatization demonstrated excellent derivative yields over the pH 9.0-10.0. Maximal yields close to 100% were observed with three- to four-fold molar reagent excess. In addition, the detection responses for BCEC-derivatives were compared to those obtained using 1,2-benzo-3,4-dihydrocarbazole-9-ethyl chloroformate (BCEOC-Cl) and 9-fluorenyl methylchloroformate, (FMOC-Cl) as labeling reagents. The ratios I-BCEC/I-BCEOC = 1.94-2.17 and I-BCEC/I-FMOC = 1.04-2.19 for fluorescent (FL) responses (here, I was relative fluorescence intensity). Separation of the derivatized amines had been optimized on reversed-phase Eclipse XDB-C-8 column. Detection limits calculated from 0.50 pmol injection, at a signal-to-noise ratio of 3, were 1.77-14.4 fmol. The relative standard deviations for within-day determination (n = 11) were 1.84-2.89% for the tested amines. The mean intra- and inter-assay precision for all amines levels were < 3.64% and 2.52%, respectively. The mean recoveries ranged from 96.6% to 107.1% with their standard deviations in the range of 0.8-2.7. Excellent linear responses were observed with coefficients of > 0.9996. (C) 2006 Elsevier B.V. All rights reserved.

Evaluation of HER2, p53, bcl-2, topoisomerase II-alpha, heat shock proteins 27 and 70 in primary breast cancer and metastatic ipsilateral axillary lymph nodes.

BACKGROUND: Breast cancer is a heterogeneous disease. Predictive biological markers (BM) of responsiveness to therapy need to be identified. Evaluation of BM is mainly done at the primary site. However, in the adjuvant therapy of breast cancer, the main goal is control of micrometastases. It is still unknown whether heterogeneity in the expression of BM between the primary site and its micrometastases exists. OBJECTIVE: To evaluate the expression of some BM with potential predictive value from the primary breast cancer site and metastatic ipsilateral axillary lymph nodes. PATIENTS AND METHODS: Focality (percentage of positive cells) and intensity staining scores were evaluated for each marker. Freshly cut sections (4 microm) from embedded blocks of breast cancer fixed in formalin or bouin were put onto superfrost slides (Menzel-Gläser). Protein expression was evaluated immunohistochemically (IHC) using monoclonal antibodies against: topo II-alpha (clone KiS1, 1 microg/ml, Roche) with a trypsine pre-treatment (P); HSP27 (clone G3.1, 1/60, Biogenex), HSP70 (clone BRM.22, 1/80, Biogenex) and HER2 (clone CB11, 1/40, Novocastra; without P); p53 (clone D07, 1/750, Dako) and bcl-2 (clone 124, 1/60, Dako) with citrate buffer as P. RESULTS: Overall, the percentage of discordant marker status in the primary tumour and its metastatic lymph nodes was 2% for HER2, 6% for p53, 15% for bcl-2, 19% for topoisomerase II-alpha, 24% for HSP27 and 30% for HSP70. For the subgroup of patients with positive BM in the primary tumour, the percentage of discordance was 6% for HER2, 7% for p53, 14% for bcl-2, 19% for HSP70, 21% for topoisomerase II-alpha and 36% for HSP27. For the subgroup of patients with positive BM in the lymph nodes, the percentage of discordance was 9% for bcl-2, 15% for HER2 and p53, 21% for topoisomerase II-alpha, 22% for HSP27 and 25% for HSP70. CONCLUSIONS: 1) No biological marker had 100% concordant results. 2) Although some discordant cases might be explained by the limitations of the IHC technique, future studies aiming to evaluate the predictive value of BM in the adjuvant therapy of breast cancer should take into account a possible difference in BM expression between the primary and the metastatic sites.

A genome-wide association study identifies susceptibility variants for type 2 diabetes in Han Chinese.

To investigate the underlying mechanisms of T2D pathogenesis, we looked for diabetes susceptibility genes that increase the risk of type 2 diabetes (T2D) in a Han Chinese population. A two-stage genome-wide association (GWA) study was conducted, in which 995 patients and 894 controls were genotyped using the Illumina HumanHap550-Duo BeadChip for the first genome scan stage. This was further replicated in 1,803 patients and 1,473 controls in stage 2. We found two loci not previously associated with diabetes susceptibility in and around the genes protein tyrosine phosphatase receptor type D (PTPRD) (P = 8.54x10(-10); odds ratio [OR] = 1.57; 95% confidence interval [CI] = 1.36-1.82), and serine racemase (SRR) (P = 3.06x10(-9); OR = 1.28; 95% CI = 1.18-1.39). We also confirmed that variants in KCNQ1 were associated with T2D risk, with the strongest signal at rs2237895 (P = 9.65x10(-10); OR = 1.29, 95% CI = 1.19-1.40). By identifying two novel genetic susceptibility loci in a Han Chinese population and confirming the involvement of KCNQ1, which was previously reported to be associated with T2D in Japanese and European descent populations, our results may lead to a better understanding of differences in the molecular pathogenesis of T2D among various populations.

Structure and denaturation of 4-chlorobenzoyl coenzyme A dehalogenase from Arthrobacter sp. strain TM-1

The secondary structure of the trimeric protein 4-chlorobenzoyl coenzyme A dehalogenase from Arthrobacter sp. strain TM-1, the second of three enzymes involved in the dechlorination of 4-chlorobenzoate to form 4-hydroxybenzoate, has been examined. E(mM) for the enzyme was 12.59. Analysis by circular dichroism spectrometry in the far uv indicated that 4-chlorobenzoyl coenzyme A dehalogenase was composed mostly of alpha-helix (56%) with lesser amounts of random coil (21%), beta-turn (13%) and beta-sheet (9%). These data are in close agreement with a computational prediction of secondary structure from the primary amino acid sequence, which indicated 55.8% alpha-helix, 33.7% random coil and 10.5% beta-sheet; the enzyme is, therefore, similar to the 4-chlorobenzoyl coenzyme A dehalogenase from Pseudomonas sp. CBS-3. The three-dimensional structure, including that of the presumed active site, predicted by computational analysis, is also closely similar to that of the Pseudomonas dehalogenase. Study of the stability and physicochemical properties revealed that at room temperature, the enzyme was stable for 24 h but was completely inactivated by heating to 60 degrees C for 5 min; thereafter by cooling at 1 degrees C min(-1) to 45 degrees C, 20.6% of the activity could be recovered. Mildly acidic (pH 5.2) or alkaline (pH 10.1) conditions caused complete inactivation, but activity was fully recovered on returning the enzyme to pH 7.4. Circular dichroism studies also indicated that secondary structure was little altered by heating to 60 degrees C, or by changing the pH from 7.4 to 6.0 or 9.2. Complete, irreversible destruction of, and maximal decrease in the fluorescence yield of the protein at 330-350 nm were brought about by 4.5 M urea or 1.1 M guanidinium chloride. Evidence was obtained to support the hypothetical three-dimensional model, that residues W140 and W167 are buried in a non-polar environment, whereas W182 appears at or close to the surface of the protein. At least one of the enzymes of the dehalogenase system (the combined 4-chlorobenzoate:CoA ligase, the dehalogenase and 4-hydroxybenzoyl coenzyme A thioesterase) appears to be capable of association with the cell membrane.

Neuropeptide Y Y(1) receptor regulates protein turnover and constitutive gene expression in hypertrophying cardiomyocytes.

Increased levels of neuropeptide Y correlate with severity of left ventricular hypertrophy in vivo. At cardiomyocyte level, hypertrophy is characterised by increased mass and altered phenotype. The aims were to determine the contributions of increased synthesis and reduced degradation of protein to neuropeptide Y-mediated increase in mass, assess effects on gene expression, and characterise neuropeptide Y Y receptor subtype involvement. Neuropeptide Y (10 nM) increased protein mass of adult rat ventricular cardiomyocytes maintained in culture (24 h) (16%>basal) and de novo protein synthesis (incorporation of [14C]phenylalanine) (18%>basal). Neuropeptide Y (100 nM) prevented degradation of existing protein at 8 h. Actinomycin D (5 µM) attenuated increases in protein mass to neuropeptide Y (=1 nM) but not to neuropeptide Y (10 nM). [Leu31, Pro34]neuropeptide Y (10 nM), an agonist at neuropeptide Y Y1 receptors, increased protein mass (25%>basal) but did not stimulate protein synthesis. Neuropeptide Y-(3–36) (10 nM), an agonist at neuropeptide Y Y2 receptors, increased protein mass (29%>basal) and increased protein synthesis (13%>basal), respectively. Actinomycin D (5 µM) abolished the increase in protein mass elicited by neuropeptide Y-(3–36) but not that by [Leu31, Pro34]neuropeptide Y. BIBP3226 [(R)-N2-(diphenylacetyl)-N-(4-hydroxyphenylmethyl)-d-arginine amide] (1 µM), a neuropeptide Y Y1 receptor subtype-selective antagonist, and T4 [neuropeptide Y-(33–36)]4, a neuropeptide Y Y2 receptor subtype-selective antagonist, attenuated the increase in protein mass to 100 nM neuropeptide Y by 68% and 59%, respectively. Neuropeptide Y increased expression of the constitutive gene, myosin light chain-2 (MLC-2), maximally at 12 h (4.7-fold>basal) but did not induce (t=36 h) expression of foetal genes (atrial natriuretic peptide (ANP), skeletal-a-actin and myosin heavy chain-ß). This increase was attenuated by 86% and 51%, respectively, by BIBP3226 (1 µM) and T4 [neuropeptide Y-(33–36)]4 (100 nM). [Leu31, Pro34]neuropeptide Y (100 nM) (2.4-fold>basal) and peptide YY-(3–36) (100 nM) (2.3 fold>basal) increased expression of MLC-2 mRNA at 12 h. In conclusion, initiation of cardiomyocyte hypertrophy by neuropeptide Y requires activation of both neuropeptide Y Y1 and neuropeptide Y Y2 receptors and is associated with enhanced synthesis and attenuated degradation of protein together with increased expression of constitutive genes but not reinduction of foetal genes.

Australia Telescope Compact Array 1.2cm Observations of the Massive Star Forming Region G305.2+0.2

We report on Australia Telescope Compact Array observations of the massive star-forming region G305.2+0.2 at 1.2 cm. We detected emission in five molecules towards G305A, confirming its hot core nature. We determined a rotational temperature of 26 K for methanol. A non-local thermodynamic equilibrium excitation calculation suggests a kinematic temperature of the order of 200 K. A time-dependent chemical model is also used to model the gas-phase chemistry of the hot core associated with G305A. A comparison with the observations suggest an age of between 2 × 104 and 1.5 × 105 yr. We also report on a feature to the south-east of G305A which may show weak Class I methanol maser emission in the line at 24.933 GHz. The more evolved source G305B does not show emission in any of the line tracers, but strong Class I methanol maser emission at 24.933 GHz is found 3 arcsec to the east. Radio continuum emission at 18.496 GHz is detected towards two H ii regions. The implications of the non-detection of radio continuum emission towards G305A and G305B are also discussed.

Voltammetry of oxygen in the room-temperature ionic liquids 1-ethyl-3-methylimidazolium bis((trifluoromethyl)sulfonyl)imide and hexyltriethylammonium bis((trifluoromethyl)sulfonyl)imide: One-electron reduction to form superoxide. Steady-state and transient behavior in the same cyclic voltammogram resulting from widely different diffusion coefficients of oxygen and superoxide

The electrochemical reduction of oxygen in two different room-temperature ionic liquids, 1-ethyl-3-methylimidazolium bis((trifluoromethyl)sulfonyl)imide ([EMIM][N(Tf)(2)]) and hexyltriethylammonium bis((trifluoromethyl)sulfonyl)imide ([N-6222][N(Tf)(2)]) was investigated by cyclic voltammetry at a gold microdisk electrode. Chronoamperometric measurements were made to determine the diffusion coefficient, D, and concentration, c, of the electroactive oxygen dissolved in the ionic liquid by fitting experimental transients to the Aoki model. [Aoki, K.; et al. J. Electroanal. Chem. 1981, 122, 19]. A theory and simulation designed for cyclic voltammetry at microdisk electrodes was then employed to determine the diffusion coefficient of the electrogenerated superoxide species, O-2(.-), as well as compute theoretical voltammograms to confirm the values of D and c for neutral oxygen obtained from the transients. As expected, the diffusion coefficient of the superoxide species was found to be smaller than that of the oxygen in both ionic liquids. The diffusion coefficients of O-2 and O-2(.-) in [N-6222][N(Tf)(2)], however, differ by more than a factor of 30 (D-O2 = 1.48 x 10(-10) m(2) s(-1), DO2.- = 4.66 x 10(-12) m(2) s(-1)), whereas they fall within the same order of magnitude in [EMIM][N(Tf)(2)] (D-O2 = 7.3 x 10(-10) m(2) s(-1), DO2.- = 2.7 x 10(-10) m(2) s(-1)). This difference in [N-6222][N(Tf)(2)] causes pronounced asymmetry in the concentration distributions of oxygen and superoxide, resulting in significant differences in the heights of the forward and back peaks in the cyclic voltammograms for the reduction of oxygen. This observation is most likely a result of the higher viscosity of [N-6222][N(Tf)(2)] in comparison to [EMIM][N(Tf)(2)], due to the structural differences in cationic component.

Geometric bonding effects in the X2A1, A2∑+u, and B2IIg states of Li2F

Published ab-initio and pseudopotential calculations for the dialkali halide systems suggest that the preferred co-linear geometry is for the metal to approach the metal end of the alkali halide. Here, ab-initio calculations on the Li2F system reveal that the well depth on the halide side in this radical is much deeper and is a local saddle-point associated with the ionic non-linear global minima. Although many features of the pseudopotential surfaces are confirmed, significant differences are apparent including the existence of a linear excited state instead of a triangular one, a considerably deeper global minimum some 50% lower in energy and a close approach between the X2A1 and the states, with the minimum 87 kJ mol-1 below the ground state asymptote. All the results can be rationalised as the avoided crossings between a long range, covalent potential dominant within the LiLiF geometry and an ionic state that forms the global minimum. Calculations on the 3rd 2A' potential indicate that even for Li + LiF collisions at ultracold temperatures the collision dynamics could involve as many as three electronic states.

The electrochemical oxidation of catechol and dopamine on platinum in 1-Ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([C(2)mim][NTf2]) and 1-Butyl-3-methylimidazolium tetrafluoroborate ([C(4)mim][BF4]): Adsorption effects in ionic liquid voltammetry

The electrochemical oxidation of catechol and dopamine has been studied at a platinum micro-electrode (10 pm diameter) in two room temperature ionic liquids (RTILs): 1-Ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([C(2)mim][NTf2]) and 1-Butyl-3-methylimidazolium tetrafluoroborate ([C(4)mim][BE4]). For catechol in [C(2)mim][NTf2], an electrochemically quasi-reversible oxidation peak was observed at 1.1 V vs. Pt with a back peak at 0.4 V vs. Pt. This is assigned to the two-electron oxidation of catechol to doubly protonated o-benzoquinone. Double-step chronoamperometry gave a diffusion coefficient for the catechol and the oxidised species which is 3.8 x 10(-11) m(2) s(-1) for both. For catechol in [C(4)mim][BF4], a two-electron oxidation wave was observed at 1.0 V vs. Pt with no back peak. Another peak at less positive potential was also observed at 0.6 V vs. Pt in [C(4)mim][BF4] but not in [C(2)mim][NTf2] which is assigned to the adsorption of electrochemically formed neutral o-benzoquinone on the platinum electrode. The oxidised protonated o-benzoquinone is suggested to be deprotonated by the [BF4](-) anion, but not by the [NTf2](-) anion: hence adsorption of the neutral species at the platinum electrode, not the charged species. For dopamine in both RTILs, two chemically irreversible oxidation peaks were observed at 0.75 V and 1.1 V vs. Pt, and assigned to the oxidation of dopamine to the corresponding semi-quinone and the quinone. Potential-step chronoamperometry was carried out on the oxidation waves of dopamine in [C(2)mim][NTf2] and the diffusion coefficient of species in solution was calculated to be 6.85 x 10(-12) m(2) s(-1) and confirmed that the waves corresponded to one and two electron processes. A third wave was observed at 1.8 V vs. Pt which is attributed to the oxidation of the amine group to a radical cation with likely subsequent follow up chemistry. In [C(4)mim][BF4] a peak at less positive potential was observed for dopamine, similar to catechol which is assigned to the adsorption of the neutral quinone species on the platinum electrode formed by the reaction of the removal of protons from the oxidised dopamine with the [BF4](-) anion. (C) 2009 Elsevier B.V. All rights reserved.

Lipoprotein composition in HNF1A-MODY: Differentiating between HNF1A-MODY and Type 2 diabetes.

INTRODUCTION:

The young-onset diabetes seen in HNF1A-MODY is often misdiagnosed as Type 2 diabetes. Type 2 diabetes, unlike HNF1A-MODY, is associated with insulin resistance and a characteristic dyslipidaemia. We aimed to compare the lipid profiles in HNF1A-MODY, Type 2 diabetes and control subjects and to determine if lipids can be used to aid the differential diagnosis of diabetes sub-type.
METHODS:

1) 14 subjects in each group (HNF1A-MODY, Type 2 diabetes and controls) were matched for gender and BMI. Fasting lipid profiles and HDL lipid constituents were compared in the 3 groups. 2) HDL-cholesterol was assessed in a further 267 patients with HNF1A-MODY and 297 patients with a diagnosis of Type 2 diabetes to determine its discriminative value.

RESULTS:

1) In HNF1A-MODY subjects, plasma-triglycerides were lower (1.36 vs. 1.93 mmol/l, p = 0.07) and plasma-HDL-cholesterol was higher than in subjects with Type 2 diabetes (1.47 vs. 1.15 mmol/l, p = 0.0008), but was similar to controls. Furthermore, in the isolated HDL; HDL-phospholipid and HDL-cholesterol ester content were higher in HNF1A-MODY, than in Type 2 diabetes (1.59 vs. 1.33 mmol/L, p = 0.04 and 1.10 vs. 0.83 mmol/L, p = 0.019, respectively), but were similar to controls (1.59 vs. 1.45 mmol/L, p = 0.35 and 1.10 vs. 1.21 mmol/L, p = 0.19, respectively). 2) A plasma-HDL-cholesterol > 1.12 mmol/L was 75% sensitive and 64% specific (ROC AUC = 0.76) at discriminating HNF1A-MODY from Type 2 diabetes.

CONCLUSION:

The plasma-lipid profiles of HNF1A-MODY and the lipid constituents of HDL are similar to non-diabetic controls. However, HDL-cholesterol was higher in HNF1A-MODY than in Type 2 diabetes and could be used as a biomarker to aid in the identification of patients with HNF1A-MODY.

Increased serum kallistatin levels in type 1 diabetes patients with vascular complications

Kallistatin, a serpin widely produced throughout the body, has vasodilatory, anti-angiogenic, anti-oxidant, and anti-inflammatory effects. Effects of diabetes and its vascular complications on serum kallistatin levels are unknown.

Toluene dioxygenase-catalysed oxidation route to angular cis-monohydrodiols and other bioproducts from bacterial metabolism of 1,2-dihydrobenzocyclobutene and derivatives

A mutant strain (UV4) of the soil bacterium Pseudomonas putida, containing toluene dioxygenase, has been used in the metabolic oxidation of 1,2-dihydrobenzocyclobutene 12 dagger and the related substrates 1,2-dihydrobenzocyclobuten-1-ol 13 and biphenylene 33. Stable angular cis-monohydrodiol metabolites (1R,2S)-bicyclo[4.2.0]octa-3,5-diene-1,2 7, (1S,2S,8S)-bicyclo[4.2.0]octa-3,5-diene-1,2,8-triol 8 and biphenylene-cis-1,8b-diol 9, isolated from each of these substrates, have been structurally and stereochemically assigned. The structure, enantiopurity and absolute configuration of the other cis-diol metabolites, (2R,3S)-bicyclo[4.2.0]octa-1(6),4-diene-2,3-diol 14 and cis-1,2-dihydroxy-1,2-dihydrobenzocyclobutene 16, and the benzylic oxidation bioproducts, 1,2-dihydrobenzocyclobuten-1-ol 13, 1,2-dihydrobenzocyclobuten-1-one 15 and 2-hydroxy-1,2-dihydrobenzocyclobuten-1-one 17, obtained from 1,2-dihydrobenzocyclobutene and 1,2-dihydrobenzocyclobuten-1-ol, have been determined with the aid of chiral stationary-phase HPLC, NMR and CD spectroscopy, and stereochemical correlation. X-Ray crystallographic methods have been used in the determination of absolute configuration of the di-camphanates 27 (from diol 7) and 32 (from diol 9), and the di-MTPA ester 29 (from diol 14) of the corresponding cis-diol metabolites. The metabolic sequence involved in the formation of bioproducts derived from 1,2-dihydrobenzocyclobutene 12 has been investigated.

Investigation into the effect of molybdenum-site substitution on the performance of Sr2Fe1.5Mo0.5O6-δ for intermediate temperature solid oxide fuel cells

In this paper, niobium doping is evaluated as a means of enhancing the electrochemical performance of a Sr₂Fe_1.5Mo_0.5O_6-δ (SFM) perovskite structure cathode material for intermediate temperature solid oxide fuel cells (IT-SOFCs) applications. As the radius of Nb approximates that of Mo and exhibits +4/+5 mixed valences, its substitution is expected to improve material performance. A series of Sr₂Fe_1.5Mo_0.5-xNb_xO_6-δ (x = 0.05, 0.10, 0.15, 0.20) cathode materials are prepared and the phase structure, chemical compatibility, microstructure, electrical conductivity, polarization resistance and power generation are systematically characterized. Among the series of samples, Sr₂Fe_1.5Mo_0.4Nb_0.10O_6-δ (SFMNb_0.10) exhibits the highest conductivity value of 30 S cm^-1 at 550°C, and the lowest area specific resistance of 0.068 Ω cm² at 800°C. Furthermore, an anode-supported single cell incorporating a SFMNb_0.10 cathode presents a maximum power density of 1102 mW cm^-2 at 800°C. Furthermore no obvious performance degradation is observed over 15 h at 750°C with wet H₂(3% H₂O) as fuel and ambient air as the oxidant. These results demonstrate that SFMNb shows great promise as a novel cathode material for IT-SOFCs.

Investigation into the effect of Fe-site substitution on the performance of Sr2Fe1.5Mo0.5O6-δ anodes for SOFCs

Ni-substituted Sr₂Fe_1.5-xNi_xMo_0.5O_6-δ (SFNM) materials have been investigated as anode catalysts for intermediate temperature solid oxide fuel cells. Reduced samples (x = 0.05 and 0.1) maintained the initial perovskite structure after reduction in H₂, while metallic nickel particles were detected on the grain surface for x = 0.2 and 0.3 using transmission electron microscopy. Temperature programmed reduction results indicate that the stable temperature for SFNM samples under reduction conditions decreases with Ni content. In addition, X-ray photoelectron spectroscopy analysis suggests that the incorporation of Ni affects the conductivity of SFNM through changing the ratios of Fe³⁺/Fe²⁺ and Mo⁶⁺/Mo⁵⁺. Sr₂Fe_1.4Ni_0.1Mo_0.5O_6-δ shows the highest electrical conductivity of 20.6 S cm^-1 at 800 °C in H₂. The performance of this anode was further tested with electrolyte-supported cells, giving 380 mW cm^-2 at 750 °C in H₂, hence demonstrating that Ni doping in the B-site is beneficial for Sr₂Fe_1.5Mo_0.5O_6-δ anode performance.