Title

Protein Recognition by Self-organizing Sensors

Date of Award

2013

Document Type

Dissertation

Degree Name

Doctor of Philosophy (Ph.D.)

Department

Photochemical Sciences

First Advisor

George Bullerjahn, Ph.D.

Second Advisor

Sheryl Coombs, Ph.D. (Committee Member)

Third Advisor

Milan Stojanovic, Ph.D. (Committee Member)

Fourth Advisor

H. Peter Lu, Ph.D. (Committee Member)

Fifth Advisor

John Cable, Ph.D. (Committee Member)

Abstract

There is a strong correlation between protein imbalance in human urine and diseases, which makes it an attractive target for sensors and diagnostics. However, detecting this imbalance is still challenging due to the multianalyte nature of human urine, which forms a complex matrix that complicates sensor design. Self-organizing receptors are the ideal base materials for such sensors as they are intrinsically cross-reactive and, as a result, can identify and quantify multicomponent analytes. Because self-organizing receptors do not display analyte complementarity, they do not reject varying binding partners even if their structures are dissimilar. However, the recognition process in the cross-reactive sensors can be guided by attaching certain moieties in order to increase affinity and selectivity for a desired analyte. In this study, we describe the preparation of fluorescent sensors based on biomimetic receptors capable of self-organization in the presence of proteins, thus serving as recognition elements. These sensors constitute fluorescein-coupled linear amine backbones (polyallylamine hydrochloride polymers) as well as amine-terminated poly-L-lysine dendrimers of different generations. Our sensors offer a wide dynamic range, tolerance for variable chemical backgrounds, room temperature operation, undemanding handling, reasonable size and mass, and low cost. The unique response patterns obtained from there urinary proteins (human serum albumin, uromodulin and transferrin) using these self-organizing sensors (SOS) resulted in a discriminatory accuracy approaching 100% in classifying proteins both qualitatively and quantitatively in buffer and human urine. This work also aims to show that the number of the necessary recognition elements in the SOS-based arrays can be decreased, thus simplifying the pattern recognition protocols without significantly compromising sensing reliability. In addition, we review the main concepts and approaches related to the use of self-organizing based sensor arrays. This dissertation lays the groundwork for further development of self-organizing sensors arrays for biomedical and chemical applications.