QTL analyses of lineage-negative mouse bone marrow cells labeled with Sca-1 and c-Kit.

Differences in the number of functionally and/or phenotypically defined bone marrow cells in inbred mouse strains have been exploited to map quantitative trait loci (QTL) that determine the variation in cell frequency. To extend this approach to the differences in the stem/progenitor cell compartment in CBA/H and C57BL/6 mice, we have exploited the resolution of flow cytometry and the power of QTL analyses in 124 F(2) mice to analyze lineage-negative (Lin(-)) bone marrow cells according to the intensity of labeling with Sca-1 and c-Kit. In the Lin(-) Sca-1(+) c-Kit(+) enriched population, six QTL were identified: one significant and five suggestive. Whereas previous in vitro clonogenic, LTC-IC, day 35 CAFC, and flow cytometry each identified different QTL, our approach identified the same or very similar QTL at all three loci (chromosomes 1, 17, and 18) as well as QTL on chromosomes 6 and 10. In silico analyses implicate hematopoietic stem cell homing involving Cxcr4 and Cxcl12 as being the determining pathway. The mapping of the same or very similar QTL in independent studies using different assay(s) suggests a common genetic determinant, and thus reinforces the biological and genetic significance of the QTL. These data also suggest that mouse bone marrow cell subpopulations can be functionally, phenotypically, and genetically defined.

DNA methylation during mouse hemopoietic differentiation and radiation-induced leukemia.

OBJECTIVE: To examine DNA methylation in mouse hemopoiesis before and after in vivo exposure to a leukemogenic dose of x-rays, and address whether methylation levels are associated with the relative radiosensitivity of tissues in vivo. METHODS: The methylation status of control CBA/H and C57BL/6 mouse tissues before and after exposure to 3-Gy x-rays, and myeloid and lymphoid leukemias and lymphomas, was assessed by the direct analysis of the 5-methylcytosine (5-(Me)C) content of DNA, and by Southern blot analysis of genomic repeat sequences. RESULTS: The DNA 5-(Me)C content of bone marrow is 15% lower than spleen. Together with the analyses of stem (myeloid) and progenitor (lymphoid) leukemias and lymphomas, we found a trend of increasing methylation during hemopoietic differentiation. Exposure to x-rays induced greater cell death in the hypomethylated bone marrow (>80%) than spleen (50%) in vivo, supporting the observed correlation found between methylation status and radiosensitivity of other high-turnover hierarchical tissues. Furthermore, there was an 8% DNA 5-(Me)C content decrease in bone marrow after in vivo exposure to 3-Gy x-rays, but this was genotype dependent, being observed in AML-susceptible (CBA/H) but not AML-resistant (C57BL/6) inbred mice. CONCLUSION: Together these data suggest that methylation status may be related to the relative radiosensitivity of high-turnover hierarchical tissues such as bone marrow and that radiation-induced DNA hypomethylation has a role in radiation leukemogenesis.

Evidence for clustered tumour suppressor gene loci on mouse chromosomes 2 and 4 in radiation-induced acute myeloid leukaemia.

PURPOSE: To investigate the influence of genetic and epigenetic factors on allelic loss on chromosomes 2 and 4 in mouse radiation-induced acute myeloid leukaemia (r-AML). METHODS: r-AML that arose in (CBA/HxC57BL/6)F1xCBA/H and F1xC57BL/6 mice were screened for transcription factor PU1 (also known as SPI-1) gene mutations and methylation of the paired box gene 5 (Pax5) gene promoter. We have increased the statistical significance of a genetic linkage analysis of affected F1xCBA/H mice to test for linkage to loci implicated directly or indirectly with r-AML-susceptibility. RESULTS: There was a statistically significant difference ( p < 10-4) in the frequency of PU1 gene mutations in F1xCBA/H and F1xC57BL/6 r-AML, implicating a second linked but genotype-dependent myeloid leukaemia suppressor gene on chromosome 2. A suggestive CBA/H r-AML-resistance locus maps within 10 cM of the minimally deleted region on chromosome 4. The Pax5 gene promoter is subject to ongoing subclonal promoter methylation in the r-AML, evidence that Pax5 gene silencing confers a selective advantage during clonal expansion in vivo. CONCLUSIONS: Allelic loss in mouse r-AML and subsequent tumour suppressor gene mutation (PU1) or silencing (Pax5) is strongly influenced by genetic background and/or epigenetic factors, and driven by in vivo clonal selection.