Plasma proteomic landscape and patient stratification in response to severe infection

Thesis

Rationale: A dysregulated host response to infection can cause life-threatening organ dysfunction, presenting as the clinical syndrome of sepsis. Sepsis is an unmet global health challenge, with substantial clinical and molecular heterogeneity hindering the development of targeted therapies. Subphenotypes in sepsis have been identified at the clinical and molecular levels, especially using the leukocyte transcriptome. Due to the limited throughput of proteome profiling, there has not been a comprehensive analysis of the sepsis plasma proteome that enables understanding of the individual response. In severe COVID-19, a maladaptive immune response to SARS-CoV-2 infection can also lead to organ failure and adverse outcome. This thesis characterises the proteomic landscape in sepsis and COVID-19 to investigate disease mechanisms and inter-individual variation in the response.

Methods: The plasma proteome was profiled using high-throughput mass spectrometry for 2622 samples from 1612 individuals, including ICU sepsis patients with community acquired pneumonia or faecal peritonitis and non-septic controls. Higher-depth proteomics and cytokine data were also generated for subsets of the sepsis patients to complement the analysis. For understanding the COVID-19 blood proteome, 395 plasma and serum samples from patients and controls were analysed.

Results: Various aspects of the immune response distinguished sepsis from control conditions, highlighting alterations in innate immunity, acute-phase response, neutrophil function and extracellular matrix organisation, with sepsis-specific proteins identified. Three sepsis patient subgroups based on the plasma proteome were identified and validated in independent samples. The subgroups were associated with differential clinical severity and distinct molecular characteristics, including one subgroup showing greater activity in immune pathways and predictive of higher mortality. The proteomic and previously defined leukocyte transcriptomic patient subgroups exhibited significant interaction, shared and distinct molecular mechanisms, and improved risk stratification when combined. Proteomic features of COVID-19 severity were identified, including acute-phase response, complement cascade, tissue necrosis and shifted lipoprotein metabolism.

Conclusion: This thesis has identified key biological processes of the sepsis response and clinically informative patient subgroups at the plasma proteome level, highlighting opportunities for patient stratification and development of a precision medicine approach.

Authors: Mi, Y

• DOI: 10.5287/ora-pr50xzvnj
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: plasma; proteomics; mass spectrometry; sepsis; precision medicine; COVID-19; critical illness; sub-phenotyping; transcriptomics
• Subjects: Septicemia; Critical care medicine; Functional genomics; COVID-19 (Disease); Proteomics; Genomic medicine