Khalid Khan (Presenter)
Waters Corporation
Authorship: Richard J Mbasu (1,2), Liam M Heaney (2), Billy Joe Molloy (3), Chris J Hughes (3), Leong L Ng (2), Johannes PC Vissers (3), James I Langridge (3), Donald JL Jones (1,2)
Department of Cancer Studies, University of Leicester, UK, (2) Department of Cardiovascular Sciences and NIHR Leicester Cardiovascular Biomedical Research Unit, UK, (3) Waters Corporation, UK
| Short Abstract The application of tandem quadrupole mass spectrometry with integrated microfuidic chromatography (IonKey) for the analysis of proteolytic peptides in human plasma is described. A tandem quadrupole platform was used for its performance in terms of sensitivity, precision, and linearity and IonKey was selected due to its balance of sensitivity and throughput. This LC-MS configuration was utilized to demonstrate that proteolytically digested, non-depleted plasma samples from heart failure patients could be classified with good discriminative power using a subset of proteins previously suggested as candidate biomarkers for cardiovascular disease. |
Long Abstract
Introduction:
Targeted LC-MS based assays are increasingly applied in the post-discovery proteomics area with emphasis on validation or in studies aimed at gaining understanding of biological systems, drug development and treatment. These experiments are technologically challenged since they require analyzing large sample numbers with high throughput, but also high sensitivity, high resolution, large dynamic range and excellent selectivity. Tandem quadrupole and high-resolution MS were systematically compared to contrast the performance of these analyzer types and how they could be operated for targeted protein quantitation experiments. One of the configurations was applied to characterise Heart failure (HF), representing a clinical population for which validated biomarkers are sparse and the aetiology of the disease means that the phenotypic changes are likely to be multifactorial. As a research proof of principle and to test the potential sensitivity to classify this disease, healthy controls, HF patients with preserved ejection fraction (HFPEF) and HF patient samples with reduced ejection fraction (HFREF) were analyzed.
Methods:
For the comparative part of the study, EDTA human plasma matrix was used, and for research proof of principle experiments, blood samples collected from a cohort of twenty healthy donors, twenty HFPEF patients, and twenty HFREF patients. All sera were mixed with ammonium bicarbonate in the presence of RapiGest, reduced, alkylated and digested overnight using trypsin. Fifteen SIL peptides, representing putative blood-based cardiovascular disease protein biomarkers, were spiked in the digested samples. The donor, HFPEF and HFREF samples were spiked post-digestion at four individual different levels of 0.25, 0.5, 2 and 10 fmol with the same SIL peptides. Nanoscale LC separation of tryptic peptides was performed with an Acquity M-class system equipped with a pre and an analytical column. After desalting and preconcentration, peptides were eluted from the precolumn to the analytical column and separated with a gradient of 3-40% B over 90 min at a flow rate of 300 nL/min. Higher throughput experiments were performed with 100 mm x 150 ìm ionKey microfluidics. Gradient conditions were from 3-40% B over 45 min at a flow rate of 1 ìL/min, followed by a column wash and re-equilibration. Samples were injected/loaded directly on-column or using a precolumn configuration. In this instance, the precolumn was 5 cm x 300 ìm and samples were loaded with a flow rate of 15 ìL/min for 1 min. MRM analysis was performed using two tandem quadrupole MS systems, Xevo TQ-S and Xevo-TQ-S micro, and two hybrid quadrupole orthogonal acceleration time-of-flight platforms, Xevo G2-XS Q-ToF and an IMS enabled Synapt G2-Si. All experiments were performed in positive mode ESI and targeted peptides were fragmented using CID. Endogenous and SIL peptides were targeted by at least three transitions with at least ten data points per chromatographic peak. Peptide MRM data were quantified with TargetLynx and Skyline. All statistical analyses were conducted with SIMCA and IBM SPSS statistics v22.
Results:
The MS response and assay sensitivities were tested from 6.25 amol to 12.5 fmol for the nanoscale LC experiments and from 62.5 amol to 12.5 fmol for the microfluidics separations. All MS instruments demonstrated linear behaviour with r2 values of 0.9909 or greater for all SIL peptides. The S/N values measured at the 12.5 amol and 125 amol spike level were found to be comparable for the majority of the SIL peptides measured with all LC-MS configurations, ranging from 2 to 30 for the nanoscale LC, and 2 to 100 for the microfluidics configuration. The median and average lower LOD values, across all MS platforms equalled 5 and 8 amol for nanoscale LC, and 20 and 39 amol for microfluidics LC-MS. The experiment also afforded calculation of the concentration of some of the endogenous peptides and estimation of assay precision of the different configurations. A tandem quadrupole platform in combination with microfluidics separation was chosen for the heart failure application part of the study based on its performance in terms of sensitivity, precision and linearity. Results obtained demonstrate that proteolytically digested non-depleted plasma samples from HF patients could be classified with good discriminative power. MVA showed that samples could be classified using OPLS-DA and near complete separation of healthy controls and HF (combined HFPEF and HFREF) patients observed. The discriminating proteins primarily contributing to the separation were ApoA1, CRP and plasma protease C1 inhibitor. Univariate analysis of these three proteins showed significant changes in levels between the groups. Good discriminating power was obtained by combining these protein surrogate peptides, with an AUC of 0.937 obtained for ROC curve analysis (p < 0.001). Separation between all three groups was not obtained with the selected subset of peptides, however, a partial separation model was developed for HFREF/HFPEF.
Conclusions:
The effect of both LC scale and choice of MS platform were investigated for potential use in translational biomarker studies. The results presented show that a larger 150 ìm ID microfluidic device had the required sensitivity, quantitative performance and throughput characteristics. In the case of the MS platform, sensitivity was comparable between modern tandem quadrupole and quadrupole time-of-flight systems, but the combination of a high sensitivity tandem quadrupole with a microfluidic inlet provided the best coefficient of variation, throughput and S/N. One of the LC-MS configurations was used for the analysis of selected peptides, and hence proteins, originating from heart failure disease patients, and compared to control samples. This preliminary study was designed to evaluate the overall application of LC-MS for translational research analysis and to classify this heterogeneous disease, resulting in separation of the two disease groups from the control group and partial separation of the two disease phenotypes.
References & Acknowledgements:
| Description | Y/N | Source |
| Grants | no | |
| Salary | yes | |
| Board Member | no | |
| Stock | yes | |
| Expenses | yes |
IP Royalty: yes
IP Desc:
| Planning to mention or discuss specific products or technology of the company(ies) listed above: | no |