S. Fasoula

Sample preparation optimization in fecal metabolic profiling

Metabolomic analysis of feces can provide useful insight on the metabolic status, the health/disease state of the human/animal and the symbiosis with the gut microbiome. As a result, recently there is increased interest on the application of holistic analysis of feces for biomarker discovery. For metabolomics applications, the sample preparation process used prior to the analysis of fecal samples is of high importance, as it greatly affects the obtained metabolic profile, especially since feces, as matrix are diversifying in their physicochemical characteristics and molecular content. However there is still little information in the literature and lack of a universal approach on sample treatment for fecal metabolic profiling. The scope of the present work was to study the conditions for sample preparation of rat feces with the ultimate goal of the acquisition of comprehensive metabolic profiles either untargeted by NMR spectroscopy and GC-MS or targeted by HILIC-MS/MS. A fecal sample pooled from male and female Wistar rats was extracted under various conditions by modifying the pH value, the nature of the organic solvent and the sample weight to solvent volume ratio. It was found that the 1/2 (wf/vs) ratio provided the highest number of metabolites under neutral and basic conditions in both untargeted profiling techniques. Concerning LC-MS profiles, neutral acetonitrile and propanol provided higher signals and wide metabolite coverage, though extraction efficiency is metabolite dependent.

Properties of the retention time of ionizable analytes in reversed-phase liquid chromatography under organic modifier gradients in different eluent pHs.

This work provides a theoretical basis for an empirical model proposed in Fasoula et al. (2013) [2] for the retention of monoprotic acids/bases during linear organic modifier gradient runs performed with a fixed change of organic content but for different gradient durations and in different eluent pHs. Based on the analytical expressions derived in the above publication for the retention of monoprotic solutes under organic modifier gradients in different mobile phase pHs it was proved that the dependences of retention time of all monoprotic acids/bases upon each of two factors governed the gradient elution under consideration are always the following: practically linear upon gradient duration (for constant pH) and sigmoid upon pH (for constant gradient duration).

Retention prediction and separation optimization of ionizable analytes in reversed-phase liquid chromatography by organic modifier gradients in different eluent pHs

The influence of eluent pH on retention of ionizable analytes during organic modifier gradients in reversed-phase high performance liquid chromatography is studied. Two approaches are examined for the retention prediction of solutes under organic modifier gradient conditions at any constant mobile phase pH: In the first approach an analytical solution of the fundamental equation of gradient elution in linear organic modifier gradients for monoprotic acids/bases under certain assumptions is proposed. The second approach is based on an empirical model arising from the evaluation of the gradient retention data. Both approaches were successfully applied to describe the retention behavior of 16 OPA derivatives of amino acids obtained in 19 simple mono-linear organic modifier gradient runs performed between two given acetonitrile contents with different gradient duration and at different eluent pHs, in a particular pH range where amino acids behave as weak monoprotic acids. Further, this study provides a reliable way for optimizing gradient separations of ionogenic compounds with respect to mobile phase pH.

Retention prediction and separation optimization under multilinear gradient elution in liquid chromatography with Microsoft Excel macros

A package of Excel VBA macros have been developed for modeling multilinear gradient retention data obtained in single or double gradient elution mode by changing organic modifier(s) content and/or eluent pH. For this purpose, ten chromatographic models were used and four methods were adopted for their application. The methods were based on (a) the analytical expression of the retention time, provided that this expression is available, (b) the retention times estimated using the Nikitas-Pappa approach, (c) the stepwise approximation, and (d) a simple numerical approximation involving the trapezoid rule for integration of the fundamental equation for gradient elution. For all these methods, Excel VBA macros have been written and implemented using two different platforms; the fitting and the optimization platform. The fitting platform calculates not only the adjustable parameters of the chromatographic models, but also the significance of these parameters and furthermore predicts the analyte elution times. The optimization platform determines the gradient conditions that lead to the optimum separation of a mixture of analytes by using the Solver evolutionary mode, provided that proper constraints are set in order to obtain the optimum gradient profile in the minimum gradient time. The performance of the two platforms was tested using experimental and artificial data. It was found that using the proposed spreadsheets, fitting, prediction, and optimization can be performed easily and effectively under all conditions. Overall, the best performance is exhibited by the analytical and Nikitas-Pappas methods, although the former cannot be used under all circumstances.

Expressions for multilinear combined pH/organic solvent elution of ionizable analytes in reversed-phase HPLC.

Expressions for the retention time of ionogenic analytes eluted under multilinear double pH/solvent-gradients in reversed-phase liquid chromatography are developed by dividing each gradient profile into a finite number of subportions, where the solute retention factors or their logarithms vary linearly with time. To test the theory, two series of experimental gradient retention data of amino acid OPA derivatives were analyzed: The first one was a monolinear or bilinear pH-gradient data set obtained in eluents with different but constant organic modifier contents, whereas the second data set comprised retention data of combined pH/organic solvent-gradients, where the organic content was changed linearly with time but the variation of pH exhibited a curved form approximated by five linear subportions. It was found that the derived expressions describe these experimental retention data with high accuracy, since under double pH/solvent-gradients the overall errors in the fitted and predicted retention times were 1.9% and 1.7%, respectively, whereas under simple pH-gradients these errors were 0.9% and 2%, respectively.

Retention modeling in combined pH/organic solvent gradient reversed-phase HPLC.

An approach for retention modeling of double pH/organic solvent gradient data easily generated by automatically mixing two mobile phases with different pH and organic content according to a linear pump program is proposed. This approach is based on retention models arising from the evaluation of the retention data of a set of 17 OPA derivatives of amino acids obtained in 27 combined pH/organic solvent gradient runs performed between fixed initial pH/organic modifier values but different final ones and for different gradient duration. The derived general model is a ninth parameter equation easily manageable through a linear least-squares fitting but it requires eighteen initial pH/organic modifier gradient experiments for a satisfactory retention prediction in various double gradients of the same kind with those used in the fitting procedure. Two simplified versions of the general model, which were parameterized based on six only initial pH/organic modifier gradients, were also proposed, when one of the final double gradient conditions, pH or organic content was kept constant. The full and the simplified models allowed us to predict the experimental retention data in simultaneous pH/organic solvent double gradient mode very satisfactorily without the solution of the fundamental equation of gradient elution.

Retention prediction of highly polar ionizable solutes under gradient conditions on a mixed-mode reversed-phase and weak anion-exchange stationary phase

In the present work the retention of three highly polar and ionizable solutes - uric acid, nicotinic acid and ascorbic acid - was investigated on a mixed-mode reversed-phase and weak anion-exchange (RP/WAX) stationary phase in buffered aqueous acetonitrile (ACN) mobile phases. A U-shaped retention behavior was observed for all solutes with respect to the eluent organic modifier content studied in a range of 5-95% (v/v). This retention behavior clearly demonstrates the presence of a HILIC-type retention mechanism at ACN-rich hydro-organic eluents and an RP-like retention at aqueous-rich hydro-organic eluents. Hence, this column should be promising for application under both RP and HILIC gradient elution modes. For this reason, a series of programmed elution runs were carried out with increasing (RP) and decreasing (HILIC) organic solvent concentration in the mobile phase. This dual gradient process was successfully modeled by two retention models exhibiting a quadratic or a cubic dependence of the logarithm of the solute retention factor (lnk) upon the organic modifier volume fraction (φ). It was found that both models produced by gradient retention data allow the prediction of solute retention times for both types of programmed elution on the mixed-mode column. Four, in the case of the quadratic model, or five, in the case of the cubic model, initial HILIC- and RP-type gradient runs gave satisfactory retention predictions of any similar kind elution program, even with different flow rate, with an overall error of only 2.5 or 1.7%, respectively.

Simple models for the effect of aliphatic alcohol additives on the retention in reversed-phase liquid chromatography

Four retention models for the effect of aliphatic alcohol additives on the retention of analytes in reversed-phase liquid chromatography have been developed following either a semi-thermodynamic treatment or an empirical approach. Their performance was tested using the experimental retention times of six non-polar analytes (alkylbenzenes) and ten o-phthalaldehyde derivatives of amino acids under different isocratic chromatographic runs when a small amount of ethanol, 1-propanol, 1-butanol, 1-pentanol, 1-hexanol or 1-heptanol was added to methanol/water mixtures containing a constant amount of methanol. It was shown that for the structurally simple alkylbenzenes all the models can be adopted for retention prediction with good results. In contrast, just one out of four models, that with the fewest approximations, predicts satisfactorily the retention properties of amino acids derivatives. However, the most interesting feature is that this model can predict the effect of an alcohol-additive on the retention properties of solutes, even if this additive was not used in chromatographic runs done for the fitting procedure, provided that it belongs to the same homologous series of alkanols. This feature is also observed in all models described the retention of alkylbenzenes.

Multivariate analysis of chromatographic retention data as a supplementary means for grouping structurally related compounds

In the present study a series of 45 metabolite standards belonging to four chemically similar metabolite classes (sugars, amino acids, nucleosides and nucleobases, and amines) was subjected to LC analysis on three HILIC columns under 21 different gradient conditions with the aim to explore whether the retention properties of these analytes are determined from the chemical group they belong. Two multivariate techniques, principal component analysis (PCA) and discriminant analysis (DA), were used for statistical evaluation of the chromatographic data and extraction similarities between chemically related compounds. The total variance explained by the first two principal components of PCA was found to be about 98%, whereas both statistical analyses indicated that all analytes are successfully grouped in four clusters of chemical structure based on the retention obtained in four or at least three chromatographic runs, which, however should be performed on two different HILIC columns. Moreover, leave-one-out cross-validation of the above retention data set showed that the chemical group in which an analyte belongs can be 95.6% correctly predicted when the analyte is subjected to LC analysis under the same four or three experimental conditions as the all set of analytes was run beforehand. That, in turn, may assist with disambiguation of analyte identification in complex biological extracts.

Retention prediction in reversed-phase liquid chromatography systems with methanol/water mobile phases containing different alkanols as additives

In an effort to gain enhancement of selectivity in reversed-phase liquid chromatography, retention was tuned in this study by introducing short and medium straight-chained-length alkanol additives (methanol (MeOH), ethanol, 1-propanol, 1-butanol, 1-pentanol, 1-hexanol or 1-heptanol) at low concentrations in mobile phases containing MeOH as the main organic solvent. A six-parameter retention model considering simultaneously the contents of the main organic modifier and of the alcohol additive as well as of the number of alkyl chain of additive was developed by a direct combination of equations expressing separately a linear dependence of the retention upon each of these factors. The effectiveness of the above model was tested in the retention prediction of a mixture of six alkylbenzenes under isocratic conditions with mobile phases containing as an additive any member of the homologues series of alkanols (with 1-7 carbon atoms) at different low concentrations in a wide range of MeOH-water mixtures. The prediction was excellent in all cases even when the alkanol additives used in experiments for the fitting procedure are different than those used in chromatographic runs done for testing the prediction ability of the proposed model.