Exploration of the Low-Molecular Weight Proteome of Tissue and Serum, with Applications to Disease Biomarker Discovery

The low-molecular weight proteome is of great interest due to the vast number of smaller proteins and peptides present within it, some of which are likely to be disease-related and may provide insights into disease mechanisms or even prediction and diagnosis. Serum is often the most preferred specimen for disease prediction or diagnosis, as it can be obtained less-invasively and its collection is fairly routine in a clinical setting. Although serum samples are one of the most preferred specimens for disease studies, additional disease information may also be obtained using other specimens, such as tissue. The study of tissue specimens can be extremely valuable in disease studies as it likely contains the highest concentration of potential markers for tissue related diseases. Therefore, despite the difficulty of sample collection, as compared to serum specimens, tissue specimens can be invaluable in studying tissue-related diseases. Tissue proteomics approaches for studying the low-molecular weight proteome of tissue have been infrequently described in the literature, which suggests the need to develop new methods or improve upon current proteomics approaches to better study this subset of the proteome in tissue. In this dissertation, I first report on a low-molecular weight tissue proteomics approach that we have developed based on a previously reported serum approach. As reported here, we conclude that this tissue proteomics approach cannot only distinguish the low-molecular weight proteome of different tissue types, but it also appears to be within the reproducibility range of other currently used proteomic approaches. Additionally, we also report on the use of a serum proteomics approach for biomarker discovery and identification in preeclampsia and endometriosis. These studies reveal many candidate biomarkers for preeclampsia prediction and endometriosis diagnosis. In the preeclampsia study we discovered several candidate biomarkers early in pregnancies (12-14 weeks gestation) that were able to predict preeclampsia later in pregnancy with statistical significance. In the endometriosis study, several candidate biomarkers were discovered that were statistically significant in diagnosing endometriosis. Further, for both preeclampsia and endometriosis, combinations of 3 or 4 biomarkers yielded an improved sensitivity and specificity in disease prediction and diagnosis with combined area under the curves of greater than 0.8. Thus, the low-molecular weight proteome of both tissue and serum could potentially provide important information regarding disease physiology, prediction, and diagnosis that may supplement our current understanding of these processes.