Apr 23, 2024
5:00pm - 7:00pm
Flex Hall C, Level 2, Summit
Richard Lee1,Oliver Nakano-Baker1,Shalabh Shukla1,John MacKenzie1
University of Washington1
Richard Lee1,Oliver Nakano-Baker1,Shalabh Shukla1,John MacKenzie1
University of Washington1
Rapid detection of VOC signatures from exhaled breath provides a simple and non-invasive way for determining states of health by assessing the breathomics signatures of conditions, diseases and chemical toxins that can endanger human performance and health. In response, we are advancing the fundamental science and engineering of biologically-inspired molecular probe-based sensing towards developing a compact, multiplexed sensor that can be rapidly deployed to detect and quickly diagnose existing and emerging infectious disease threats, toxicants, and biomarkers at low cost. The key to our approach is computationally-engineered chimeric peptides and modified protein biomolecules derived from computational derivation of target molecule binding sites and homology from natural odorant-binding olfactory receptor protein libraries. These bioinspired molecules are multifunctional and bind to specific airborne volatile and dissolved targets whose binding events are transduced directly into electrical signals in carbon-based field effect transistor (FET) sensors. These probes create a selective, compact, and customizable platform to directly sense disease or identify the human body’s response following exposure to a toxin. For complex VOC profiles such as the trace components of human breath, multichannel FET sensor chips can be functionalized with sets of peptide or protein probes to sensitize the transistor array to a disease, condition or intoxicant’s multicomponent signature thereby creating high selectivity. The FET sensor allows for detecting voltage threshold shifts versus target exposure and the gate voltage tuning of the sensor bias to achieve maximum sensitivity. The transistor-based biosensor identifies diseases and human body responses to antagonists using machine learning (ML) trained on typically multicomponent target profiles, recognizing their often complex volatile organic compound (VOC) signature profile in exhaled breath or ambient air using computationally designed and optimized sets of solid-binding peptides and protein probe molecules that can achieve high selectivity and exclude confounding factors. The platform technology could be deployed as handheld, wearable, or drone-mounted eNoses and as broadly capable chemical sensors. We have demonstrated the ability of our system to directly detect VOCs indicative of COVID-19 infection in simulated breath as well as sensing of human stress biomarkers in solution using sensor-adapted cortisol-binding proteins. Solution titration studies of protein-functionalized carbon nanotube FET sensors have enabled the determination of a target detection threshold of at least 10 parts per trillion (10 femtomolar) in a simple, direct chemical/electronic transduction signal. We are currently exploring the targeted detection of biotoxins, such as saxitoxin, with our molecular-probe functionalized FET sensor platform.