PhD abstract
Sepsis is a syndrome caused by a dysregulated immune system response to an infection (bacterial, viral, fungal), leading to organ failure. In 2017, 48.9 million cases of sepsis were reported worldwide, resulting in 11 million deaths. Early and accurate diagnosis is critical for implementing a therapeutic treatment and improving patient survival. Currently, sepsis diagnosis is based on clinical evaluation of the patient, which can be influenced by nonspecific symptoms. The individual quantification of biomarkers such as procalcitonin (PCT) or C-reactive protein (CRP) using immunological methods serves as an aid in diagnosis. Since quantifying a single biomarker may lack specificity, simultaneous monitoring of a panel of biomarkers appears to be a significant advantage for early diagnosis.
In the present work, a list of nine protein biomarkers of interest for sepsis diagnosis was established in collaboration with clinicians and in vitro diagnostic manufacturers. The objective was then to develop a targeted proteomic method using high-resolution mass spectrometry (HRMS) to enable specific, sensitive, and multiplexed quantification of these biomarkers in routine, in clinical serum samples.The proteins targeted in this study have different molecular masses and serum concentrations, with a dynamic range of approximately 105 in healthy patients' serum. Thus, the quantification of these proteins by HRMS is challenging. The first step was to select two proteotypic peptides per protein, considering various criteria such as molecular mass or amino acid composition. These peptides were then used to optimise high-performance liquid chromatography conditions (chromatographic columns, mobile phases, elution gradients), ionisation, fragmentation, and HRMS analysis, as well as sample preparation workflow (protein precipitation, enzymatic digestion, peptide extraction), allowing specific and sensitive quantification of the targeted proteins. These developments led to the identification of increasingly selective protocols, resulting in a decrease in the limits of quantification. Thus, three methods were developed: i) the first targeting five proteins and meeting clinical (covering physiological and pathological concentrations) and metrological (metrological traceability) needs, ii) the second targeting eight proteins and meeting the clinical need for a highly multiplexed method, and iii) the last targeting PCT, allowing comparison with an existing reference method.To ensure reliable quantification and traceability of measurement results to SI units, a protein-based calibration was selected for proteins with existing certified reference materials (recombinant proteins. In contrast a peptide-based calibration was selected for the other proteins. Therefore, it was necessary to rigorously characterise peptide standards (purity assessment and quantification through amino acid analysis) during this work.
The validation of the first method and the evaluation of measurement uncertainties for the quantification of the five biomarkers in human serum, followed by the analysis of clinical samples, allowed the assessment of its potential implementation for early and specific diagnosis of sepsis in routine practice. Finally, the development of a robust LC-HRMS/MS method and rigorously characterised standards represents a significant advancement in establishing metrological references.
Key words
Multiplex quantification, ID-LC-hrMS/MS, sepsis, biomarker, metrology, sample preparation