Analytica Chimica Acta (v.921, #C)

Application-Dedicated Selection of Filters (ADSF) using covariance maximization and orthogonal projection by Xavier Hadoux; Dinesh Kant Kumar; Marc G. Sarossy; Jean-Michel Roger; Nathalie Gorretta (1-12).
Visible and near-infrared (Vis-NIR) spectra are generated by the combination of numerous low resolution features. Spectral variables are thus highly correlated, which can cause problems for selecting the most appropriate ones for a given application. Some decomposition bases such as Fourier or wavelet generally help highlighting spectral features that are important, but are by nature constraint to have both positive and negative components. Thus, in addition to complicating the selected features interpretability, it impedes their use for application-dedicated sensors.In this paper we have proposed a new method for feature selection: Application-Dedicated Selection of Filters (ADSF). This method relaxes the shape constraint by enabling the selection of any type of user defined custom features. By considering only relevant features, based on the underlying nature of the data, high regularization of the final model can be obtained, even in the small sample size context often encountered in spectroscopic applications. For larger scale deployment of application-dedicated sensors, these predefined feature constraints can lead to application specific optical filters, e.g., lowpass, highpass, bandpass or bandstop filters with positive only coefficients.In a similar fashion to Partial Least Squares, ADSF successively selects features using covariance maximization and deflates their influences using orthogonal projection in order to optimally tune the selection to the data with limited redundancy. ADSF is well suited for spectroscopic data as it can deal with large numbers of highly correlated variables in supervised learning, even with many correlated responses.Display Omitted
Keywords: Filter selection; Variable selection; Dimension reduction; Orthogonal projection; Optical filters; Feature selection;

Glycan characterization of biopharmaceuticals: Updates and perspectives by Ana Planinc; Jonathan Bones; Bieke Dejaegher; Pierre Van Antwerpen; Cédric Delporte (13-27).
Therapeutic proteins are rapidly becoming the most promising class of pharmaceuticals on the market due to their successful treatment of a vast array of serious diseases, such as cancers and immune disorders. Therapeutic proteins are produced using recombinant DNA technology. More than 60% of therapeutic proteins are posttranslationally modified following biosynthesis by the addition of N- or O-linked glycans. Glycosylation is the most common posttranslational modifications of proteins. However, it is also the most demanding and complex posttranslational modification from the analytical point of view. Moreover, research has shown that glycosylation significantly impacts stability, half-life, mechanism of action and safety of a therapeutic protein. Considering the exponential growth of biotherapeutics, this present review of the literature (2009–2015) focuses on the characterization of protein glycosylation, which has witnessed an improvement in methodology. Furthermore, it discusses current issues in the fields of production and characterization of therapeutic proteins. This review also highlights the problem of non-standard requirements for the approval of biosimilars with regard to their glycosylation and discusses recent developments and perspectives for improved glycan characterization.Display Omitted
Keywords: Biopharmaceuticals; Biosimilars; N-glycosylation; Glycan separation; Glycan detection; Bioinformatics; Glycan analysis guidelines;

New multivariate calibration methods and other processes are being developed that require selection of multiple tuning parameter (penalty) values to form the final model. With one or more tuning parameters, using only one measure of model quality to select final tuning parameter values is not sufficient. Optimization of several model quality measures is challenging. Thus, three fusion ranking methods are investigated for simultaneous assessment of multiple measures of model quality for selecting tuning parameter values. One is a supervised learning fusion rule named sum of ranking differences (SRD). The other two are non-supervised learning processes based on the sum and median operations. The effect of the number of models evaluated on the three fusion rules are also evaluated using three procedures. One procedure uses all models from all possible combinations of the tuning parameters. To reduce the number of models evaluated, an iterative process (only applicable to SRD) is applied and thresholding a model quality measure before applying the fusion rules is also used. A near infrared pharmaceutical data set requiring model updating is used to evaluate the three fusion rules. In this case, calibration of the primary conditions is for the active pharmaceutical ingredient (API) of tablets produced in a laboratory. The secondary conditions for calibration updating is for tablets produced in the full batch setting. Two model updating processes requiring selection of two unique tuning parameter values are studied. One is based on Tikhonov regularization (TR) and the other is a variation of partial least squares (PLS). The three fusion methods are shown to provide equivalent and acceptable results allowing automatic selection of the tuning parameter values. Best tuning parameter values are selected when model quality measures used with the fusion rules are for the small secondary sample set used to form the updated models. In this model updating situation, evaluation of all possible models, thresholding, and iterative SRD performed equivalently for the three fusion rules with TR and PLS performed worse. While the application is model updating, the fusion processes are applicable to other situations requiring selection of multiple tuning parameter values.Display Omitted
Keywords: Sum of ranking differences; Multivariate calibration; Partial least squares; Tikhonov regularization; Calibration maintenance; Model updating; Fusion; Multicriteria optimization; Penalty regression;

PEBBLE (probe encapsulated by biologically localized embedding) nanosensor encapsulating an intensity-based fluorescence indicator and an inert reference fluorescence dye inside the pores of stable matrix can be used as a generalized wavelength-ratiometric probe. However, the lack of an efficient quantitative model render the choices of inert reference dyes and intensity-based fluorescence indicators used in PEBBLEs based generalized wavelength-ratiometric probes rather limited. In this contribution, an extended quantitative fluorescence model was derived specifically for generalized wavelength-ratiometric probes based on PEBBLE technique (QFMGRP) with a view to simplify the design of PEBBLEs and hence further extend their application potentials. The effectiveness of QFMGRP has been tested on the quantitative determination of free Ca2+ in both simulated and real turbid media using a Ca2+ sensitive PEBBLE nanosensor encapsulating Rhod-2 and eosin B inside the micropores of stable polyacrylamide matrix. Experimental results demonstrated that QFMGRP could realize precise and accurate quantification of free Ca2+ in turbid samples, even though there is serious overlapping between the fluorescence excitation peaks of eosin B and Ca2+ bound Rhod-2. The average relative predictive error value of QFMGRP for the test simulated turbid samples was 5.9%, about 2–4 times lower than the corresponding values of partial least squares calibration model and the empirical ratiometric model based on the ratio of fluorescence intensities at the excitation peaks of Ca2+ bound Rhod-2 and eosin B. The recovery rates of QFMGRP for the real and spiked turbid samples varied from 93.1% to 101%, comparable to the corresponding results of atomic absorption spectrometry.Display Omitted
Keywords: Probe encapsulated by biologically localized embedding; Generalized ratiometric probes; Quantitative fluorescence spectroscopy; Turbid media; Calcium;

Simultaneous quantitative profiling of 20 isoprostanoids from omega-3 and omega-6 polyunsaturated fatty acids by LC–MS/MS in various biological samples by Aude Dupuy; Pauline Le Faouder; Claire Vigor; Camille Oger; Jean-Marie Galano; Cédric Dray; Jetty Chung-Yung Lee; Philippe Valet; Cécile Gladine; Thierry Durand; Justine Bertrand-Michel (46-58).
Isoprostanoids are a group of non-enzymatic oxygenated metabolites of polyunsaturated fatty acids. It belongs to oxylipins group, which are important lipid mediators in biological processes, such as tissue repair, blood clotting, blood vessel permeability, inflammation and immunity regulation. Recently, isoprostanoids from eicosapentaenoic, docosahexaenoic, adrenic and α-linolenic namely F3-isoprostanes, F4-neuroprostanes, F2-dihomo-isoprostanes and F1-phytoprostanes, respectively have attracted attention because of their putative contribution to health. Since isoprostanoids are derived from different substrate of PUFAs and can have similar or opposing biological consequences, a total isoprostanoids profile is essential to understand the overall effect in the testing model. However, the concentration of most isoprostanoids range from picogram to nanogram, therefore a sensitive method to quantify 20 isoprostanoids simultaneously was formulated and measured by liquid chromatography-tandem mass spectrometry (LC–MS/MS). The lipid portion from various biological samples was extracted prior to LC–MS/MS evaluation. For all the isoprostanoids LOD and LOQ, and the method was validated on plasma samples for matrix effect, yield of extraction and reproducibility were determined. The methodology was further tested for the isoprostanoids profiles in brain and liver of LDLR−/− mice with and without docosahexaenoic acid (DHA) supplementation. Our analysis showed similar levels of total F2-isoprostanes and F4-neuroprostanes in the liver and brain of non-supplemented LDLR−/− mice. The distribution of different F2-isoprostane isomers varied between tissues but not for F4-neuroprostanes which were predominated by the 4(RS)-4-F4t-neuroprostane isomer. DHA supplementation to LDLR−/− mice concomitantly increased total F4-neuroprostanes levels compared to F2-isoprostanes but this effect was more pronounced in the liver than brain.Display Omitted
Keywords: Isoprostanes; Dihomo-isoprostanes; Neuroprostanes; Phytoprostanes; Mass spectrometry; Quantification; Oxidative stress; ROS;

As two members of the carbon materials family, carbon nanodots (CNDs) and graphene oxide (GO) possess many excellent optical properties resulting in a wide range of applications. In this work, the fluorescence of resultant dual-emission carbon nanodots (DECNDs) could be quenched by GO. In the presence of hemoglobin (Hb), the fluorescence would recover resulting from two interactions: one was the direct stacking effect of Hb on GO; the other one was that Hb could cover the surfaces of DECNDs; both of them would prevent the fluorescence quenching of DECNDs by GO. In the light of this mechanism, a novel fluorescent turn-off/on method has been developed for the detection of Hb based on DECNDs-GO system. By virtue of the dual emissions of these CNDs, it is noteworthy that both a single emission and ratiometric of dual emissions can be used to establish linear relationships of Hb: 0.05–300 nM (λem = 386 nm), 5–500 nM (λem = 530 nm), and 50–500 nM (I530/I410), with the corresponding limit of detection (LOD) as low as 20 pM, 2 nM and 20 nM, respectively. This present system is highly selective toward Hb over other proteins and this reliable method has been successfully applied for the detection of Hb in whole blood samples.Display Omitted
Keywords: Carbon nanodots; Dual emissions; Graphene oxide; Hemoglobin;

A new platform of a paper-based analytical device (PAD) for simultaneous forward and reverse ABO blood group typing has been reported. This platform can overcome the discrepancy results as influenced by the individual haematocrit. The test and the control of non-haemagglutination on each channel were performed in parallel. The PAD was fabricated by printing six parallel channels with wax onto Whatman No. 4 filter paper. An LF1 blood separation membrane was used for the separation of plasma from whole blood for reverse grouping. The blood group was identified by haemagglutination of the corresponding antigen–antibody. For forward grouping, Anti-A, -B and –A,B were treated on the test line of PAD, and inactivated Anti-A, -B and –A,B were immobilized on the control line. For reverse grouping, 30% standard A-cells, B- and O- were added to the test channel after plasma separation, and O-cells were used as a control. Then, 0.9% normal saline (NSS) containing 1% Tween-20 was bi-functionally used for dilution of the blood sample and elution of the non-agglutinated RBCs within the channels. The distance of agglutinated RBCs in each test line was compared with the distance of non-agglutinated RBCs in the parallel control line. The forward and reverse patterns of blood groups A, B, AB and O were a barcode-like chart in which the results can be visually analysed. The PAD has excellent reproducibility when 10 replications of the A, B, AB or O blood groups were performed. The results of both forward and reverse grouping were highly correlated with conventional methods compared with the slide method and tube method, respectively (n = 76). Thus, this ABO typing PAD holds great potential for future applications in blood typing point-of-care testing.Display Omitted
Keywords: ABO blood group; Forward and reverse typing; Haemagglutination; Paper-based analytical devices (PADs); Wax printing;

We have developed a ratiometric fluorescent probe BRT based on boron dipyrromethene (BODIPY) and rhodamine-thiohydrazide Förster resonance energy transfer (FRET) platform for sensing hypochlorous acid (HOCl) with high selectivity and sensitivity. The probe can detect HOCl in 15 s with the detection limit of 38 nM. Upon mixing with HOCl the fluorescence colour of probe BRT changed from green to orange. Moreover, probe BRT was applied to successfully monitor HOCl in living RAW 264.7 cells.Display Omitted
Keywords: Boron dipyrromethene; Rhodamine-thiohydrazide; Ratiometric; Fluorescence; Hypochlorous acid; Living cells;

High-performance liquid chromatographic enantioseparation of cyclic β-aminohydroxamic acids on zwitterionic chiral stationary phases based on Cinchona alkaloids by Gyula Lajkó; Tímea Orosz; Nóra Grecsó; Beáta Fekete; Márta Palkó; Ferenc Fülöp; Wolfgang Lindner; Antal Péter; István Ilisz (84-94).
Cyclic β-aminohydroxamic acid enantiomer pairs were stereoselectively separated by high-performance liquid chromatography on the recently developed Cinchona alkaloid-based zwitterionic chiral stationary phases Chiralpak ZWIX(+)™, ZWIX(−)™, ZWIX(+A) and ZWIX(−A). The results of variation of the applied chromatographic conditions, such as the bulk solvent composition, the concentrations and ratio of the acid and base additives, the presence of water as mobile phase additive and the counter-ion concentration furnished a better understanding of the retention mechanism. A thermodynamic study in the temperature range 5–50 °C revealed enthalpy-controlled enantiodiscrimination in all cases. The structure–selectivity relationships clearly indicated the importance of the strereochemistry of the functional groups. From an enantiorecognition aspect, the diexo position of the functional groups always proved more favorable than the diendo position. The elution sequence was determined in all cases and was found to reversed when ZWIX(+)™ was changed to ZWIX(−)™ or ZWIX(+A) to ZWIX(−A).Display Omitted
Keywords: HPLC; Direct chiral separations; Enantiomers of cyclic β-aminohydroxamic acids; Temperature dependence of chiral separation; Chiral zwitterionic ion-exchangers;