Diagnostic and Prognostic Profiling of Infectious Diseases Through Multiplexed Protein Assays

Abstract

Laboratory-based biomarkers can indicate that normal biologic functions, pathogenic processes, or immunologic responses have occurred. Diagnostic biomarkers are used diagnose patients with a specific disease or condition, while prognostic biomarkers provide information on disease progression, recurrence, or adverse outcome. Host inflammatory proteins, such as cytokines, can be employed as diagnostic and prognostic biomarkers; however, due to the complex nature of an inflammatory response, multi-biomarker signatures increase clinical utility. The need to identify and validate protein signatures for specific diseases is driving a need for analytical techniques amenable to multiplexed protein detection. This thesis applies silicon photonic microring resonators as a biomolecular sensing platform. The instrument and assay method are amenable for multiplexed detection of sixteen analytes in two samples simultaneously, using automated reagent handling and featuring a time-to-result of 45 minutes. Herein, I describe the development and application of multiplexed protein biomarker assays to diagnose and identify biomarker signatures for various infections. To improve rapid diagnostic approaches in the field of filoviral infection, we developed a two-plex biomarker assay for diagnosis of Zaire ebolavirus and Sudan ebolavirus infections using the pathogen-specific soluble glycoprotein (sGP). The sGP is a potential early diagnostic and prognostic biomarker of infection and, when coupled with the microring resonator assay, provide a fast, sensitive method for early Ebola infection diagnosis. The assay achieved limits of detection in the ng/mL range, exhibited no cross-reactivity, and successfully detected sGP in clinically relevant specimens (Chapter 2 and 3). Exploration of host inflammatory biomarker signatures was applied to expand the diagnostic and prognostic toolbox of neonatal conditions (Chapter 4) and tuberculosis infections (Chapters 5 and 6). We employed a seven-plex panel to temporally measure cytokine profiles in over sixty preterm neonates and identified altered cytokine trends in neonates exposed to inflammation in utero. Our work highlights the potential of longitudinal profiling studies to identify prognostic or monitoring biomarkers associated with infection status. Extending this multiplexed profiling approach to an infectious disease, we applied a fourteen-plex cytokine panel to profile over 500 QuantiFERON stimulated plasma samples from patients with latent tuberculosis infection (LTBI) with the goal of generating multi-biomarker signatures for LTBI diagnosis and disease reactivation prognosis. In comparing two independent patient cohorts, we showed a high overlap in important biomarkers, mainly IP-10, IL-2, and CCL8, towards classifying LTBI and high-risk of reactivation status. After merging cohorts to increase sample size, we uncovered a set of biomarkers capable of classifying LTBI status with 87% accuracy. Importantly, we identified a panel of biomarkers that can be used to stratify low versus high risk of reactivation and conclude that a group of eight host inflammatory protein biomarkers should be considered for future LTBI diagnostic platform development. Further analysis of multiplexed assay calibrations yielded insight into expected analyte- and sensor batch-related variability (Chapter 7). Targets IL-6, CCL3, and CCL8 were the most robust assays and should be used as comparisons for newly generated calibration curves. Taken together, the work presented in this thesis applies analytical techniques to clinically relevant challenges through rapidly detecting an early biomarker of Ebola virus infection, monitoring immunological changes in preterm neonates, and identifying host immune biomarkers correlated to disease phase and reactivation risk in LTBI. The assays and biomarker signatures presented here have the potential to impact diagnosis, prognosis, and clinical management of infections and infectious diseases.