State-of-the art data normalization methods improve NMR-based metabolomic analysis

Extracting biomedical information from large metabolomic datasets by multivariate data analysis is of considerable complexity. Common challenges include among others screening for differentially produced metabolites, estimation of fold changes, and sample classification. Prior to these analysis steps, it is important to minimize contributions from unwanted biases and experimental variance. This is the goal of data preprocessing. In this work, different data normalization methods were compared systematically employing two different datasets generated by means of nuclear magnetic resonance (NMR) spectroscopy. To this end, two different types of normalization methods were used, one aiming to remove unwanted sample-to-sample variation while the other adjusts the variance of the different metabolites by variable scaling and variance stabilization methods. The impact of all methods tested on sample classification was evaluated on urinary NMR fingerprints obtained from healthy volunteers and patients suffering from autosomal polycystic kidney disease (ADPKD). Performance in terms of screening for differentially produced metabolites was investigated on a dataset following a Latin-square design, where varied amounts of 8 different metabolites were spiked into a human urine matrix while keeping the total spike-in amount constant. In addition, specific tests were conducted to systematically investigate the influence of the different preprocessing methods on the structure of the analyzed data. In conclusion, preprocessing methods originally developed for DNA microarray analysis, in particular, Quantile and Cubic-Spline Normalization, performed best in reducing bias, accurately detecting fold changes, and classifying samples.

     

Time course of pump inhibition by ouabain in amphibian epithelia

Inhibition by ouabain of rheogenic Na⁺ transport across the basolateral membranes of frog skin is found to be manifest within 3–4 min. This rate of pump inhibition is not different from the rate of diffusion through extracellular tissue layers between the serosal bath and the actual site of action, i.e., the epithelial cell layers. It is concluded that the well-known slow time course of decrease in transepithelial current flow is due ionic redistribution and conductance changes of the epithelial membranes secondary to pump inhibition.     

Effect of whole-body vibration with different frequencies and intensities on auditory evoked potentials and heart rate in man

Auditory evoked brain potentials (AEP) and electrocardiogram (ECG) were recorded from 9 healthy male subjects during sinusoidal whole-body vibration exposure (WBV) in the longitudinal (+/- az) direction with four frequencies (1 Hz, 2 Hz, 4 Hz, and 8 Hz) and two intensities as well as under non-WBV conditions. The sequences of the different experimental conditions were arranged according to a 9 X 9 Latin Square design. The sound of the electrohydraulic vibrator was masked by a constant noise level. A subtraction technique was used to eliminate vibration-synchronous activity contaminating the electroencephalogram. The AEP amplitude N1-P2 revealed systematic effects of different WBV frequencies and intensities. The amplitude decreased along with an increase in intensity (16 dB) by about 10 per cent. It diminished increasingly with a monotonic trend in the order non-WBV, WBV 8 Hz, WBV 4 Hz, WBV 2 Hz, and WBV 1 Hz. The interbeat-interval histograms computed from the ECG exhibited the highest mean values at MBV of 1 Hz, high intensity, and the lowest ones at WBV of 4 Hz, high intensity. The AEPs are reaffirmed as an informative measure for studying the WBV effect on central nervous information processing, although the modes of action are not yet fully known. Efferent influences on the acoustic input, cross-modality interaction, sensory mismatch, and changes of central nervous activation level are discussed as potential mechanisms.     

Dosimetry of low-energy neutrons using low-pressure proportional counters

Measurements in nearly monoenergetic beams of 144, 24.5, and 2 keV neutrons and of thermal neutrons have been performed with low-pressure proportional counters. The suitability of a tissue-equivalent proportional counter (TEPC) for dosimetry of low-energy neutrons has been investigated. In contrast to higher neutron energies, the modification of the primary radiation field by the detector wall and the contribution of secondaries produced in the gas are significant. These effects have been investigated by additional measurements with a carbon-walled proportional counter. The various physical processes of neutron interaction with wall and gas of the TEPC have been analyzed, and absorbed dose, kerma, and kerma contributions from the various processes are presented. In addition, dose contributions from contaminating neutrons and photons have been obtained for the calibration fields used. The results have been related to neutron fluence. The comparison with tabulated kerma factors shows excellent agreement, indicating the suitability of the TEPC method for dosimetry of low-energy neutrons.     

Resolution of d,l-menthol by interesterification with triacetin using the free and immobilized lipase of Candida cylindracea

Summary Interesterification in isooctane with triacetin as an acyl donor was found to be a new and effective method of racemic resolution of d,l-menthol, when using the free and immobilized lipase of Candida cylindracea. No water was produced by this highly stereoselective type of reaction in contrast to ester synthesis with acetic acid as an acyl donor. Even with diacetin no possible back reaction occurred and the enzyme was easily separated from the reaction solution as opposed to ester hydrolysis in aqueous systems. Inhibition of interesterification was caused by increasing concentrations of the acyl donor triacetin by more than 10 mmol·l^-1 on the one hand, and especially by diacetin on the other hand. The reaction product menthyl acetate had no influence. By adding water the interesterification activity of the lipase was reduced significantly. An alteration of the acyl donor triacetin to longerchained triglycerides caused changes in higher specific activities but poor enantioselectivities of the products, as in the case of ester synthesis starting from longer-chained organic acids.Dedicated to Prof. Dr. Fritz Wagner on the occasion of his 60th birthday     

Changes in auditory evoked brain potentials during ultra-low frequency whole-body vibration of man or of his visual surround

Auditory evoked brain potentials (AEP) were recorded from nine healthy male subjects during three types of condition: A - subject and visual field stationary; B - subject vibrated (z-axis, 0.6 Hz, 1.85 ms-2 rms), visual field stationary; C - subject stationary, visual field vibrated (as for B). The visual surround was confined to a checkerboard pattern in front of the subject. Auditory stimuli (1000 Hz, 86 dB, interstimulus interval 7 s) were delivered via headphones to evoke AEP. Vibration-synchronous activity in the EEG was eliminated by a subtraction technique. In comparison with condition A, conditions B and C caused an attenuation of P2 and N1P2 components of AEP together with an increased latency of N1. Effects of conditions B and C did not differ. Direct vestibular stimulation and mechanisms specific for whole-body vibration were rejected as modes of action. The AEP-changes and the subjective evaluation of experimental conditions, arousal and performance, as well as symptoms of kinetosis (motion sickness) suggest a sensory mismatch, leading to a "latent kinetosis" with de-arousal, as the dominating mechanism by which the processing of information was affected. This suggestion was supported by an additional pilot study. Under real working conditions a similar effect can be expected during relative motion between the driver and his visual surround, i.e. even with perfect vibro-isolation of the driver's seat.     

Genetics of the quantitative Lp(a) lipoprotein trait

The Lp(a) lipoprotein is a complex particle composed of a low density lipoprotein (LDL)-like lipoprotein and the disulfide bonded Lp(a) glycoprotein. The complex represents a quantitative genetic trait. SDS gel electrophoresis under reducing conditions of sera followed by immunoblotting with affinity-purified polyclonal anti-Lp(a) demonstrated inter- and intra-individual size heterogeneity of the glycoprotein with apparent Mr in the range 400-700kDa. According to their relative mobilities compared to apo B-100 the Lp(a) patterns were categorized into phenotypes F, B, S1, S2, S3 und S4 and into the respective double-band phenotypes. This size heterogeneity seems to be controlled by multiple alleles designated LpF, LpB, LpS1, LpS2, LpS3, LpS4 and a null allele (LpO) at a single locus. Phenotype frequencies observed in 441 unrelated subjects were in good agreement with those expected from the genetic hypothesis. Comparison of Lp(a) lipoprotein concentrations in the different phenotypes revealed a highly significant association of phenotypes B, S1 and S2 with high, and phenotypes S3 und S4 with intermediate Lp(a) concentrations. A third mode is represented by the null phenotype were no Lp(a) band is detected upon immunoblotting and Lp(a) lipoprotein is low or absent. We conclude that the same gene locus is involved in determining Lp(a) glycoprotein phenotype and Lp(a) lipoprotein concentrations in plasma. This major gene seems to be the Lp(a) glycoprotein structural gene locus.     

Genetics of the quantitative Lp(a) lipoprotein trait

Summary Lp(a) glycoprotein exhibits an apparent size polymorphism that is associated with genetically controlled Lp(a) lipoprotein concentrations in plasma (Utermann et al. 1988). We have tested the hypothesis that this polymorphism is genetically controlled by studying 15 matings with a total of 44 offspring. This confirmed our conclusion that Lp(a) types are controlled by a series of codominant alleles Lp^F, Lp^B, Lp^S1, Lp^S2, Lp^S3 and Lp^S4 and by a null allele Lp^o. Together with the data from the accompanying paper this indicates that the structural gene for the Lp(a) protein is the major gene locus determining Lp(a) lipoprotein concentrations in plasma.