Michele Harvey

Fibroblast cultures were derived from mouse embryos containing either one (p53+/-) or two (p53-/-) inactivated p53 alleles and compared to normal embryo fibroblasts for a number of growth parameters. Early passage p53-deficient embryo fibroblasts (p53-/-) divided faster than normal embryo fibroblasts, achieved higher confluent densities, and had a higher fraction of division-competent cells under conditions of low cell density. Flow cytometry studies of early passage embryo fibroblasts showed that the percent of p53-deficient cells in G0/G1 was lower than in normal cells, consistent with the argument that p53 mediates a G1 block. When p53-deficient and normal cells were passaged for long periods of time, the homozygote (p53-/-) fibroblasts grew at a high rate for over 50 passages and never entered a non-growing senescent phase characteristic of the heterozygote (p53+/-) and normal (p53+/+) cells. The p53-deficient fibroblasts were genetically unstable during passaging, with the p53-/- cells showing a high degree of aneuploidy and the p53+/- cells displaying a moderate level of chromosomal abnormalities by passage 25. Surprisingly, the heterozygote cells lost their single wild type allele very early during culturing and in spite of this loss most heterozygote lines entered into senescence. We conclude that the loss of p53 by itself is insufficient to confer immortality on a cell, but does confer a growth advantage. Taken together, the findings confirm that the absence of p53 promotes genomic instability, which in turn may result in genetic alterations which directly produce immortality.

In a large scale mutagenesis screen for embryonic mutants in zebrafish, we have identified 63 mutations in 24 loci affecting the morphogenesis of the zebrafish brain. The expression of marker genes and the integrity of the axonal scaffold have been studied to investigate abnormalities in regionalization, neurogenesis and axonogenesis in the brain. Mutants can be broadly classified into two groups, one affecting regionalization along the anterior-posterior or dorsal-ventral axis, and the other affecting general features of brain morphology. The first group includes one locus that is required to generate the anlage of the midbrain-hindbrain boundary region at the beginning of somitogenesis. Four loci were identified that affect dorsal-ventral patterning of the brain, including the previously described cyclops locus. Mutant embryos of this class show a reduction of ventral neuroectodermal structures and variable fusion of the eyes. The second group includes a large class of mutations affecting the formation of brain ventricles. Analysis of this class reveals the requirement of a functional cardiovascular system for ventricle enlargement during embryogenesis. Mutations in one locus lead to the formation of supernumerary primary neurons, a phenotype reminiscent of neurogenic mutants in Drosophila. Other mutant phenotypes described here range from abnormalities in the fasciculation and outgrowth of axons to defects in the diameter of the neural tube. The identified loci establish the genetic foundation for a further analysis of the development of the zebrafish embryonic brain.

Mice with disrupted germline p53 alleles have been engineered by us and others and have been shown to have enhanced susceptibility to spontaneous tumors of various types. We monitored a large number of p53-deficient mice (p53+/- and p53-/-) and their wild-type littermates (p53+/+) of two different genetic backgrounds (129/Sv and mixed C57BL/6 x 129/Sv) up to 2 yr of age. p53+/- and p53-/- 129/Sv mice show accelerated tumorigenesis rates compared with their p53-deficient counterparts of mixed C57BL/6 x 129/Sv genetic background. The tumor spectra of the two strains of mice are similar except that almost half of 129/Sv p53-/- males develop malignant teratomas, whereas these tumors are rarely observed in C57BL/6 x 129/Sv mice and never in 129/Sv p53+/- males. In the study reported here, we further characterized the lymphomas that arose in the p53-nullizygous mice and found that over three-quarters of the lymphomas were of thymic origin and contained primarily immature (CD4+/CD8+) T-cells, whereas the remainder originated in the spleen and peripheral lymph nodes and were of B-cell type. The high incidence of early-onset lymphomas in the nullizygous mice makes these animals a good lymphoma model, whereas the heterozygous mice may be a useful model for Li-Fraumeni syndrome, a human inherited cancer predisposition.