i) One project involves the analysis of skin in a mouse model of osteogenesis imperfecta. This is motivated by the fact that diagnosis of mild osteogenesis imperfecta (OI) is invasive and time consuming. However, if the type I collagen abnormalities of OI are manifest in skin, MRI would represent a potential diagnostic approach. Accordingly, we used MRI to detect dermal abnormalities in oim/oim and oim/+ mice: these are mice models for OI. Differences between normal and oim/oim mice included the introduction of a hair follicle-rich layer (layer 2L) found below the dermis. MRI results were also consistent with decreased collagen content and a more highly hydrated collagen network. Histology of human fetal tissue indicated that these skin changes may also be manifest in human patients with OI. We conclude that characterization of phenotypic differences in the skin of oim/oim and oim/+ mice by MRI is feasible, and may potentially be extended to diagnosis of human OI. ii) Additional work centers on analysis of cardiovascular function in transgenic mice (TG++) overexpressing the alpha-1C22++ splice variant of a calcium channel. Magnetic resonance imaging (MRI) was used to investigate the cardiovascular phenotype in the basal state, as well as during stress induced by isoproterenol (IP) infusion over 7 days. Chronic IP stress significantly increased left ventricle (LV) mass and LV to body weight ratio in both wild type (WT) and TG22+/+ mice. In addition, LV end-systolic volume (EDV) in both normal and TG22+/+ were significantly elevated compared to vehicle-treated mice. Further, IP-induced stress in TG22+/+ mice significantly decreased LV ejection fraction and cardiac output. Overall, chronic IP stress in TG22+/+ mice resulted in markedly greater contractile dysfunction than in WT, indicating increased remodeling in response to beta-adrenergic stress. This observation may provide insight into a potential pathogenic role for this channel in cardiovascular disease. iii) MR spectroscopy of brain is an excellent adjunct to morphologic imaging studies. We use this approach to define cerebellar metabolites in the XRCC1-deficient (KO) mouse, in which this key participant in base excision repair and single-strand break repair is lacking. Both MR imaging and spectroscopic analyses were performed. In the imaging experiments, cerebellar volume was found not to differ between the two groups. In the spectroscopy experiments, we were able to edit lipid signal by use of a TE = 135 ms, permitting clear delineation of choline, creatine, and NAA. No difference in the NAA neuronal marker was observed between the two groups. For both MRI volume measurements and MR spectroscopic measurements of neuronal viability, a larger variability in the data was seen in the XRCC1 KO animals. Overall, in spite of the cerebellar ataxia demonstrated in the XRCC1 KO phenotype, we found no differences in cerebellar volume or in cerebellar neuronal viability between the normal and KO animals. Thus, the ataxic phenotype is present in spite of preservation of these important parameters of brain phenotype. We will continue to refine our techniques and apply them to other disease models, such as Huntington's Chorea.
