Precise, high-throughput quantification of nanoscale biological targets remains a key challenge in diagnostics, one that engineered particle technologies are well-equipped to address. My work addresses this gap through the development of a modular Janus Particle (JP) platform in which stochastic rotational dynamics serve as a means to measure nanovesicle capture in complex biological media.
I fabricate JPs by depositing a monolayer of fluorescently intercalated, micron-sized polystyrene beads with a 30 nm layer of gold via physical vapor deposition (PVD). This gold hemisphere facilitates simple anisotropic antibody conjugation and creates distinguishable ‘bright’ and ‘dim’ fluorescent states based on bead orientation. This fabrication method produces highly uniform particles, and is scalable for high-throughput production. Inter-batch variation is minimal, ensuring reproducibility across technical and biological replicates.
Once in solution, thermal motion induces a characteristic particle blinking via rotation. Particle-fluid interactions subject particles to rotational Brownian motion, which scales proportionally to media viscosity and inversely with the cube of particle diameter. A custom particle tracking algorithm extracts intensity fluctuations over time, and a continuous wavelet transformation (CWT) identifies the dominant rotational frequency for each particle. When JPs bind to a nanovesicle (>50nm) via conjugated antibodies, the increased drag measurably slows down rotation and shifts the blinking frequency relative to unbound control JPs.
I led experimental efforts to apply this platform for profiling small extracellular vesicles (sEVs), which carry diverse surface biomarkers indicative of cellular origin and disease state. Our first study evaluated the relative expressions of aEGFR, CEA, and GPC1 on sEVs to distinguish glioblastoma, colorectal cancer, and pancreatic cancer from healthy samples, without sample pretreatment. This resulted in a cumulative AUC 0.919. The modularity of the platform is easily scalable for other clinically relevant disease subclassifications and progressions.
Expanding this project’s scope, I integrated the JP platform into a multimodal study combining lipidomics, proteomics, vesicles sizing, and activity/affinity assays to study cancer metastasis. In a cohort of 43 patient samples, we achieved 100% precision and recall in classifying healthy, pre-metastatic, and metastatic stages. A key finding emerged while studying ADAM10, a metastasis-associated disintegrin. While enzymatic activity decreased in cancer plasma samples, the concentration of ADAM10+ sEVs increased, suggesting the formation of protease inhibitory complex, a new insight in the mechanism for cancer signaling and immune evasion.
To extend the platform’s utility to virology, I independently developed magnetic Janus Particles (mJPs) to pre-concentrate low-titer viral samples. The mJPs capture viruses from larger sample volume, followed by magnetically-facilitated concentration, improving sensitivity by two orders of magnitude. When combined with a smartphone clip-on lens, the system is translatable to a point-of-care screening tool for resource-limited settings. Additional preliminary results demonstrate the potential for early Alzheimer's Disease detection via phosphorylated Tau-181 on sEVs, and for monitoring chemotherapeutic response by quantifying cancer-associated sEV fractions over time.
The work integrates nano- and microfabrication, diagnostic design, and computational signal analysis with relevance to clinical applications. By leveraging dynamic particle behavior in complex fluids, it offers a novel framework for advancements in disease profiling and biological characterization. I have independently led particle design, fabrication, functionalization, and experimentation, and I intend to utilize this breadth of technical expertise in an industry role focused on diagnostics, product and process development, or pharmaceutical technologies.