The studies described in this dissertation involve the use and comparison of two mouse strains: one sensitive (CBA/CaJ) and another resistant (C57BL/6J) to radiation-induced acute myeloid leukemia (AML). The purpose of these studies was to identify factors that may account for the large difference in the susceptibility of these strains to radiation-induced AML.
The present study was initiated to determine whether the distances between breakpoint clusters on chromosome 2 are in closer proximity in the bone marrow cells of the CBA/CaJ mouse strain than in the C57BL/6J strain. Bacterial artificial chromosomes (BACs) were selected as markers of the central portion of the proximal and distal deletion breakpoint clusters as well as mdr on chromosome 2, where the preponderance of breaks occurs. Distance measurements were made by three dimensional fluorescent in situ hybridization (3DFISH) image analysis of hundreds of cells using Metamorph and ImageJ for data collection and Autoquant software for deconvolution and reconstruction of the three dimensional cell nuclei. Comparing bone marrow cells of CBA/CaJ and C57BL/6J mice, no differences were found between the proximity of the two regions represented for the selected markers compared in both murine strains. For the markers chosen the distribution of the distances showed similarities between the same cell types from both mouse strains; namely, fibroblasts, whole bone marrow (WBM), and hematopoietic stem cells (HSC).
However, there was not found a change in the distance distributions toward the closer distances expected between the clusters in HSC and WBM compared with fibroblasts in both mouse strains. There was; however, a tissue-dependent distance distribution between the markers Specifically, the average distances of the clusters in fibroblasts (2.55 um for CBA/CaJ and 3.09 um for C57BL/6) were larger than the distance in blood cells (1.74 um in BM and 1.53 um in HSC for CBA/CaJ; and 1.79 um in BM and 1.77 um in HSC for C57BL/6). This tissue-dependency is consistent with the concept of tissue predisposition to certain kind of cancers, in which, for instance blood cells contain specific characteristics or nuclear organization not present in fibroblasts that could lead to AML.
Using AML cells from actual radiation-induced tumors, the measurements done within the intact chromosome 2 from these AML samples showed a high proportion of cells with distances between the clusters markers that were similar to the distances seen for the small domain from normal BM cells. Therefore, from our data, deletion of chromosome 2 seemed to occur mainly in a non-random fashion because the PU.1 gene was deleted from the large domain in 8 out of 10 cases in an average proportion of ∼74% of the analyzed cells considering all AML cases.
To explore and test the possible effect of the genomic imprinting on the structure and organization of the chromatin in both small and large domain from mouse chromosome 2, a different mouse model was used that allowed us to differentiate the parental origin of each chromosome 2 inherited after fertilization for the hybrid offspring (F1) obtained from crosses between a C3H/HeNCrl and Tirano/EiJ mouse strain.
The latter has a Robertsonian translocation that involved chromosome 2 and 8, which allows tracking of a paternal or maternal copy of chromosome 2 in the F1 mice. Although such a CBA strain was not available, the C3H mouse strain is similarly sensitive to AML induction after radiation treatment, and chromosome 2 in this mouse model is hyper-radiosensitive as well. Then, if the small or closed and large or open configuration of the chromatin that was observed in the interphase is due to the genomic imprinting, we should be able to determine its parental origin. The experimental data did not show evidence of any influence in the chromosomal domain conformation in relation to the genomic imprinting occurring in mouse chromosome 2. No difference was seen for the maternal and paternal copies of chromosome 2 within interphase cells. All chromosome 2 domains from C3H/HeNCrl showed breakpoint clusters distances and organization of the domains similar to the small domain in both maternal and paternal copies. (Abstract shortened by UMI.)