Capillary Electrophoresis-Mass Spectrometry at Trial by Metabo-Ring: Effective Electrophoretic Mobility for Reproducible and Robust Compound Annotation

Abstract

Capillary zone electrophoresis-mass spectrometry (CE-MS) is a mature analytical tool for the efficient profiling of (highly) polar and ionizable compounds. However, the use of CE-MS in comparison to other separation techniques remains underrepresented in metabolomics, as this analytical approach is still perceived as technically challenging and less reproducible, notably for migration time. The latter is key for a reliable comparison of metabolic profiles and for unknown biomarker identification that is complementary to high resolution MS/MS. In this work, we present the results of a Metabo-ring trial involving 16 CE-MS platforms among 13 different laboratories spanning two continents. The goal was to assess the reproducibility and identification capability of CE-MS by employing effective electrophoretic mobility (μ_eff) as the key parameter in comparison to the relative migration time (RMT) approach. For this purpose, a representative cationic metabolite mixture in water, pretreated human plasma, and urine samples spiked with the same metabolite mixture were used and distributed for analysis by all laboratories. The μ_eff was determined for all metabolites spiked into each sample. The background electrolyte (BGE) was prepared and employed by each participating lab following the same protocol. All other parameters (capillary, interface, injection volume, voltage ramp, temperature, capillary conditioning, and rinsing procedure, etc.) were left to the discretion of the contributing laboratories. The results revealed that the reproducibility of the μ_eff for 20 out of the 21 model compounds was below 3.1% vs 10.9% for RMT, regardless of the huge heterogeneity in experimental conditions and platforms across the 13 laboratories. Overall, this Metabo-ring trial demonstrated that CE-MS is a viable and reproducible approach for metabolomics.

Figure 1

Typical profile obtained for the analysis of a standard mixture of cationic metabolites by CE-MS (platform 19, see Table S2) using 10% acetic acid as BGE.

Figure 2

Individual bias to the average and interlaboratory reproducibility of μ_eff as obtained for the analysis of cationic metabolite standards by 19 CE-MS platforms. Green and red lines mark bias thresholds at ±5 and ±10%, respectively. Metabolites are presented in decreasing mobility order from left to right.

Figure 3

Individual bias and interlaboratory variability of μ_eff (A) and RMT (B) as determined for the analysis of standard metabolite mixture by 16 CE-MS platforms. The green and red lines denote a 5% and 10% threshold, respectively. Metabolites are presented in decreasing mobility order from left to right.

Figure 4

Individual bias and interlaboratory variability of μ_eff for cationic metabolites in complex biological matrixes: plasma (A, C, and E) and urine (B, D, and F) at three different concentrations (undiluted: A and B; 5-times dilution: C and D; 10-times dilution: E and F) by 16 CE-MS platforms. Metabolites are presented in decreasing mobility order from left to right.