The Nonexistence of Procedures with Bounded Performance Characteristics in Certain Parametric Inference Problems

This paper gives a condition which implies the nonexistence of parametric statistical procedures with bounded risk or error performance characteristics. Many examples for which such a condition is satisfied are considered.     

Aurilide, a cytotoxic depsipeptide from the sea hare Dolabella auricularia: isolation, structure determination, synthesis, and biological activity

The bioassay-guided fractionation of the cytotoxic constituents of the Japanese sea hare Dollabella auricularia led to the isolation of aurilide (1), a 26-membered cyclodepsipeptide. The gross structure of 1 was established by spectroscopic analysis including 2D NMR techniques. The absolute stereostructure was determined by chiral HPLC analysis of acid hydrolysates of 1 and by the enantioselective synthesis of a degradation product arising from a dihydroxylated fatty acid portion. The enantioselective synthesis of 1 was achieved in 12% overall yield (16 steps) and confirmed the absolute stereostructure of 1. The cytotoxicity of 1 was evaluated using a synthetic sample, which was found to exhibit potent cytotoxicity against HeLa S₃ cells with an IC₅₀ of 0.011 μg/mL. Further biological and pharmacological studies of 1 have been carried out by using synthetic 1.     

The core structure of ukonan A, a phagocytosis-activating polysaccharide from the rhizome of Curcuma longa, and immunological activities of degradation products.

The controlled Smith degradation of ukonan A, a phagocytosis-activating polysaccharide isolated from the rhizome of Curcuma longa L., was performed. The reticuloendothelial system-potentiating, anti-complementary and alkaline phosphatase-inducing activities of ukonan A and its degradation products were investigated. Methylation analyses of both the primary and the secondary Smith degradation products indicated that the core structural features of ukonan A include a backbone chain mainly composed of beta-1,3-linked D-galactose, beta-1,4-linked D-xylose and alpha-1,2-linked L-rhamnose residues. All of the galactose units in the backbone carry side chains composed of alpha-L-arabino-beta-D-galactosyl or beta-D-galactosyl residues at position 6. Ukonan A has a remarkable effect on each of the three kinds of immunological activities. Periodate oxidation caused pronounced decrease or disappearance of the activities, but the controlled Smith degradation product having the core structure of polysaccharide showed considerable restoration of these activities.     

A study of GABA and its analogues using the molecular orbital method

GABA is a major neurotransmitter in the central nervous system. Data on GABA and its analogues calculated by using the ab initio and the MNDO method were compared with data obtained experimentally. The structures of GABA analogues calculated by the ab initio method agree well with the experimental data. This finding suggests the high reliability of this method. However, the structures of GABA analogues calculated by the MNDO method reflect only some aspects of the experimental data. Therefore the MNDO method should be used only for carefully selected chemical compounds.

The amino group in GABA and its analogues was proved to be the major active site. The electrostatic potential around the amino group in these compounds seems to be related to their biological activity. The difference in the electrostatic potential between the receptor binding molecules and the neuronal uptake molecules suggests that the structure of post-synaptic receptors might differ from that of uptake receptors. This finding suggests that there are at least two GABA-A receptors. GABA molecules seem to have a high potential for binding to the two receptors because they are highly flexible and can readily change their conformation. These results indicate a high reliability of the data calculated by the molecular orbital method and suggest that this method provides us with useful information that cannot be obtained experimentally.     

The S2 emission characteristics of several organic sulfur compounds obtained by molecular emission cavity analysis and pyrolysis

Emission profiles of several organic sulfur compounds are investigated by modified molecular emission cavity analysis (MECA). Thiourea, 1,3-diethylthiourea, S-methyl- cysteine and taurine are pyrolyzed in a hydrogen stream and the pyrolytic products are determined by gas chromatography. The S₂ emission mechanism is discussed on the basis of emission profiles and the composition of the pyrolytic products. Although some compounds give multipeaked responses, the splitting disappears when a worn surface cavity is used or oxalic acid is added to the sulfur compound in the cavity. When the emission profile from thiourea is compared with that from 1,3-diethylthiourea, it is clear that the multipeaked response is due to quenching by degradation products of the latter compound. The main product of pyrolysis is hydrogen sulfide. The emission intensity is related to the yield of hydrogen sulfide in pyrolysis. As methylmercaptan was not detected in the pyrolysis products, it is suggested that the quenching by the organic fragments results from their hydrogen consumption rather than their reaction with sulfur species. The S₂ emission from sulfur-containing compounds is rapidly complete in the presence of oxalic acid, and it is suggested that such compounds are subject to reductive breakdown in the cavity.     

Accurate and rapid estimation on spectral interferences in routine analysis by ICP-AES: use of mutual interference coefficients

A practical method to estimate spectral interferences and to select optimum analytical lines in ICP-AES is suggested. Depending on the matrix composition and the amounts of the analyte, the analytical lines suffering from little interferences and the limit of determination can be determined from calculation using spectral interference coefficients. For this calculation, the spectral interference coefficients, which are defined as apparent mass of the analyte equivalent to the spectral interference from unit mass of the interferent, are obtained experimentally for 639 emission lines of 68 elements. There is a good correlation between the coefficients obtained on two spectrometers having different resolutions.