Fluorine-18-labeled estrogens, progestins and corticosteroids for receptor-based imaging of breast tumors and target areas of the brain

Estrogens can be labeled with the positron-emitting radionuclide fluorine-18 (t$\sb{1/2}$ = 110 min) by fluoride ion (n-Bu$\sb4$N$\sp{18}$F) displacement of a 16$\beta$-trifluoromethanesulfonate (triflate) derivative of the corresponding estrone 3-triflate, and purification by HPLC. That sequence has been used to synthesize the 11$\beta$-methoxy 1 and 11$\beta$-ethyl 2 analogues of the breast tumor imaging agent, 16$\alpha$-($\sp{18}$F) fluoro-17$\beta$-estradiol (FES). Tissue distribution studies of 1 and 2 in immature female rats show high selectivity for target tissue (T, uterus) vs non-target (NT, muscle and lung), with T/NT ratios being 43 and 17 at one hour after injection for 1 and 2, respectively. The parent estrogen FES has previously been shown to display an intermediate value for tissue selectivity.

Vibrational characteristics of brain magnetic resonance elastography

This thesis aims to investigate vibrational characteristics of magnetic resonance elastography (MRE) of the brain. MRE is a promising, non-invasive methodology for the mapping of shear stiffness of the brain. A mechanical actuator shakes the brain and generates shear waves, which are then imaged with a special MRI sequence sensitive to sub-millimeter displacements. This research focuses on exploring the profile of vibrations utilized in brain elastography from the standpoint of ultimately investigating nonlinear behavior of the tissue. The first objective seeks to demonstrate the effects of encoding off-frequency vibrations using standard MRE methodologies. Vibrations of this nature can arise from nonlinearities in the system and contaminate the results of the measurement. The second objective is to probe nonlinearity in the dynamic brain system using MRE. A non-parametric decomposition technique, novel to the MRE field, is introduced and investigated.

Characterizing neural tissues with mass spectrometry imaging via enhanced computational approaches

Neuropeptides affect the activity of the myriad of neuronal circuits in the brain. They are under tight spatial and chemical control and the dynamics of their release and catabolism directly modify neuronal network activity. Understanding neuropeptide functioning requires approaches to determine their chemical and spatial heterogeneity within neural tissue, but most imaging techniques do not provide the complete information desired. To provide chemical information, most imaging techniques used to study the nervous system require preselection and labeling of the peptides of interest; however, mass spectrometry imaging (MSI) detects analytes across a broad mass range without the need to target a specific analyte. When used with matrix-assisted laser desorption/ionization (MALDI), MSI detects analytes in the mass range of neuropeptides. MALDI MSI simultaneously provides spatial and chemical information resulting in images that plot the spatial distributions of neuropeptides over the surface of a thin slice of neural tissue. Here a variety of approaches for neuropeptide characterization are developed. Specifically, several computational approaches are combined with MALDI MSI to create improved approaches that provide spatial distributions and neuropeptide characterizations. After successfully validating these MALDI MSI protocols, the methods are applied to characterize both known and unidentified neuropeptides from neural tissues. The methods are further adapted from tissue analysis to be able to perform tandem MS (MS/MS) imaging on neuronal cultures to enable the study of network formation. In addition, MALDI MSI has been carried out over the timecourse of nervous system regeneration in planarian flatworms resulting in the discovery of two novel neuropeptides that may be involved in planarian regeneration. In addition, several bioinformatic tools are developed to predict final neuropeptide structures and associated masses that can be compared to experimental MSI data in order to make assignments of neuropeptide identities. The integration of computational approaches into the experimental design of MALDI MSI has allowed improved instrument automation and enhanced data acquisition and analysis. These tools also make the methods versatile and adaptable to new sample types.