Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
Essential in biomedical research is the necessity of gathering statistically relevant data about large populations of specific biological entities, e.g. organisms, cells or molecules, while preserving detailed information about each single entity under investigation. This thesis deals with this need and proposes the combination of microfluidics and micro-arraying techniques in developing technological tools to conceive bio-assays at single molecule/cell/organism resolution. First, we propose an on-chip immunoassay technique, through which we demonstrated detection of the biomarker tumor necrosis factor alpha in serum down to concentrations in the attomolar range (10-18 M). In particular, we provide a comprehensive predictive model of the assay, which employs micro-arrays of superparamagnetic beads. We introduce the concept of magnetic particle-scanning, as a method for building immunoassays with extremely low limit of detection, down to the single-molecule level. Afterwards, we modified our bead micro-arraying technique, to make it suitable for the immobilization of particles and cells of various sizes and properties. Specifically, we present a method for the electrostatic self-assembly of dielectric microspheres in well templates, as a technique for fast and versatile fabrication of microlens arrays. By combining these arrays with microfluidics, we created a new tool for single-nanoparticle detection in flowing media, able to detect moving objects of sub-diffraction size through conventional low-magnification microscopes. An analogous micro-arraying method was then developed to seed large populations of non-adherent cells in isolated micro-compartments. In combination with an electrowetting-on-dielectric microfluidic platform, this technique allows implementing high-throughput cytotoxicity assays on yeast cells, at single-cell resolution. Subsequently, we conceived technological solutions for the automated analysis of Caenorhabditis elegans, one of the most employed model organisms in biomedical research. First, we developed a microfluidic platform for on-chip nematode culture and creation of synchronized C. elegans embryo micro-arrays. Long-term multi-dimensional imaging in our device allows systematic phenotyping studies at single-embryo resolution. We could discriminate embryonic development variations with unprecedented accuracy and we successfully analyzed the impact of perturbations of the mitochondrial functions on the embryogenesis. A second generation prototype of the device is then presented, enabling long-term automated studies on C. elegans at single-nematode resolution and over the whole organism development, from early embryogenesis to adulthood. Finally, we introduce a third generation prototype, which features: (i) a new microfluidic design tailored for the isolation of larvae at a desired developmental stage and for their successive culture and treatment; (ii) a method for reversible immobilization of nematodes, enabling long-term high-resolution imaging. We successfully employed this platform to analyze protein aggregation in a C. elegans model for human amyotrophic lateral sclerosis (ALS). The device allows precisely localizing protein aggregates within the nematode tissues, as well as monitoring the evolution of single aggregates over consecutive days at sub-cellular level.
Silvia Demuru, Brince Paul Kunnel