Multiple particle tracking to assess the microstructure of biological fluids

Abstract

Tracking the Brownian motion of colloids was first used about a hundred years ago to demonstrate the molecular nature of matter. Today's colloidal scientists perform particle tracking experiments to assess the structural and mechanical properties of complex materials at a micron length scale. Indeed, the dynamics of micron sized probe particles embedded in a material can be related to the local mechanical response of the system. This probing technique, called microrheology, has received much interest in the last few decades due to the importance of a materials local properties in its function and its macro-scale characteristics. These new assessments are especially relevant in soft matter sciences such as biophysics. Video microscopy particle tracking is an easy technique to implement experimentally. Movies of the fluctuating particles in a sample are recorded and analyzed off-line using custom algorithms. For this reason, it is widely used in studies of soft matter properties and in fluid dynamics. However, behind this apparent simplicity lie a number of subtle limitations that can alter significantly the validity of the measurements. The focus of the parts of this thesis is an exhaustive characterization of the errors incurred in the standard video microscopy particle tracking setup.

(cont.) Detailed understanding of these errors led to new methods to circumvent some of the intrinsic limitations. The trajectories extracted from particle tracking are used to compute the means-squared displacement that characterizes the dynamics of the probe particles. This measurement suffers from two kinds of limitations: the finite spatial resolution in the particle localization and statistical uncertainties. The source of localization errors was separated into two separate contributions. A "static error" arises in the position measurements of immobilized particles. A "dynamic error" comes from the particle motion during the finite exposure time that is required for visualization. We calculated the propagation of these errors on the mean-squared displacement and examined the impact of our analysis on theoretical model fluids used in biorheology. These theoretical predictions were verified for purely viscous fluids using simulations and a multiple particle tracking technique performed with video microscopy. We showed that the static contribution could be confidently eling the sampling design, we derived estimators for the mean and variance of particle's dynamics that are independent, under well-efined conditions, of the peculiar statistics of the measurement output.

(cont.) These estimators serve to quantify a material heterogeneity. Having gained a full characterization of the technique, we applied video multiple particle tracking to study a complex time-evolving system of self-assembling peptides. This material undergoes a transition from a purely viscous solution to an elastic hydrogel through the molecular assembly of the peptides into a fibrous network. We used the oligo-peptide KFE8 as a model self-assembling peptide and assessed the dependency of the gelation kinetics with the pH of the solution. We were able to develop a theoretical model for this dependency by using the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory for the interaction between the peptides.