Children with neurofibromatosis type 1 (NF1) have a markedly increased risk for juvenile chronic myelogenous leukemia (JCML) and monosomy 7 (Mo7) syndrome. Both JCML and Mo7 syndrome have a poor prognosis, with either progression to acute myeloid leukemia or death from incurrent problems. Using a mouse model of NF1, we have demonstrated that the lack of the Nf1 gene in hematopoietic cells is sufficient to cause chronic myeloid leukemia, but not acute myeloid leukemia. This argues that additional genetic events are responsible for the progression from chronic to acute disease. The goal of this proposal is to identify and isolate genes involved in this leukemia progression. Toward this end, we have developed a mouse model system which takes advantage of mice that harbor one mutant allele of the Nf1 gene but requires a further mutation for transformation to neoplasia.

The general strategy is to breed the mutant Nf1 allele onto a strain of mouse that expresses murine leukemia virus (MuLV) and exhibits a high incidence of acute myeloid leukemia. In this system, the MuLV acts as a somatic mutagen to activate cooperating cellular proto-oncogenes or inactivate tumor suppressor genes, resulting in accelerated tumor development. Since MuLVs activate proto-oncogenes by integrating nearby or inactivate tumor suppressor genes by integrating within the gene. Thus, the affected genes can be identified and cloned using these somatically acquired viruses as signposts. The most promising loci are the ones that have sustained proviral insertion in tumors of multiple mice. The identification of a so-called Acommon site of viral integration@ strongly indicates that the region harbors a gene, that when mutated by the virus, is directly involved in the development of myeloid leukemia.

Preliminary studies using this system have demonstrated that:

1. Tumor progression is accelerated in these heterozygous Nf1 mutant mice relative to wild-type controls.

2. Tumor tissue isolated from the majority of these heterozygous mice lacks the normal Nf1 allele due to either loss of heterozygosity or to viral insertion.

3. Each of these tumors contains at least one somatically acquired provirus which can serve as a tag to identify genes involved in Nf1-dependent tumor progression.

4. Finally, after isolating a site viral integration from one tumor, we determined that this site is similarly affected in two other independently-derived tumors. This strongly indicates that this region harbors a gene that is involved in tumor progression.

Technical Objective 1: identify regions of the genome that cooperate with loss of Nf1 in myeloid leukemia to cause tumor progression. We propose to create a large panel of murine acute myeloid tumors that lack Nf1 expression. This panel will be used to identify additional common sites of viral integration.

Technical Objective 2: characterize a locus containing a gene involved in tumor progression. Having already identified one common site, we proposed to identify and characterize the affected gene.