Date of Award

Fall 2010

Document Type


Degree Name

Doctor of Philosophy (PhD)


Biomedical Engineering

First Advisor

Steven A. Jones


Layer-by-layer self-assembly (LbL) is a technique that generates engineered nano-scale films, coatings, and particles. These nanoscale films have recently been used in multiple biomedical applications. Concurrently, microfabrication methods and advances in microfluidics are being developed and combined to create "Lab-on-a-Chip" technologies. The potential to perform complex biological assays in vitro as a first-line screening technique before moving on to animal models has made the concept of lab on a chip a valuable research tool.

Prior studies in the Biofluids Laboratory at Louisiana Tech have used layer-by-layer and in vitro biological assays to study thrombogenesis in a controlled, repeatable, engineered environment. The reliability of these previously established techniques was unsatisfactory for more complex cases such as chemical and shear stress interactions.

The work presented in this dissertation was performed to test the principal assumptions behind the established laboratory methodologies, suggest improvements where needed, and test the impact of these improvements on accuracy and repeatability.

The assumptions to be tested were: (1) The fluorescence microscopy (FM) images of acridine orange-tagged platelets accurately provide a measure of percent area of surface covered by platelets; (2) fibrinogen coatings can be accurately controlled, interact with platelets, and do not interfere with the ability to quantify platelet adhesion; and (3) the dependence of platelet adhesion on chemical agents, as measured with the modified methods, generally agrees with results obtained from our previous methods and with known responses of platelets that have been documented in the literature.

The distribution of fibrinogen on the final LbL surface generated with the standard, static process (s-LbL) was imaged by tagging the fibrinogen with an anti-fibrinogen antibody bound to fluorescein isothiocyanate (FITC). FITC FM images and acridine orange FM images were taken sequentially at selected surface locations to generate a composite overlap of presumed platelet adhesion as a function of fibrinogen distribution. The method was unable to distinguish the surface from the adhered cells. The surface inhomogeneity and porosity retained a large amount of acridine orange stain, even in the absence of platelets, and components in the platelet-rich plasma (PRP) were found to fix acridine orange in a mode that fluoresced in the FITC imaging FM. Both of these problems obfuscated the platelet adhesion FM results when using s-LbL surfaces and acridine orange staining of platelets.

A dynamic process (d-LbL) was developed in which a solution of the molecule to be layered was constantly washed over the surface, and was constantly mixed to maintain a more homogeneous distribution of solute relative to the surface during the layering process. The d-LbL surfaces were tested as described above, and found to reduce the size and number of regions of anomalous acridine orange pooling trapped by the surface, providing a greater consistency and reliability in identifying platelets.

The improved surface was then used in a series of platelet adhesion experiments under static and dynamic flow conditions, and with and without the chemical additive L-arginine. The complex microcharmel system used in prior studies was replaced with a simpler system involving fewer nuisance variables for these tests.

The tests were performed on both collagen and fibrinogen surfaces. Collagen has been used as a thrombogenic surface in multiple studies in the literature, but produces additional variables in thrombogenesis control that are avoided when fibrinogen is used. In these tests, fibrinogen was found to be as thrombogenic as collagen, and platelet coverage of both biointerfaces was reduced by L-arginine in a manner similar to previously reported work.

The simpler system differed from the previous microchannel system in important factors: (1) It exposed the platelets to much lower shear stresses; (2) It introduced an oscillatory flow, which introduced a higher degree of variability in the adhesion than previously reported; (3) the previous work had not removed the acridine orange surface problems. Therefore, a direct comparison between results was not possible.

The new d-LbL surface process was successful in testing the basic assumptions. Testing showed that the new process eliminated the anomalous acridine orange retention problem during fluorescence imaging. This improvement in fluorescence response meant that the FM results matched the platelet adhesion on plain glass slides and adhesion reported by others in microfluidic flows. The chemical additive responses behaved as expected, with an increase in L-arginine contributing to a decrease in thrombogenesis under dynamic conditions, but not under static conditions.