It may seem counterintuitive to study the function of human disease genes in fruit flies, and it is true that not all disease-related questions can be studied in what biologists call model systems. Part of our effort to identify factors that affect neurofibromatosis type 1 (NF1) disease progression involves a genetic study of 1,200 human patients. However, there are enormous advantages to studying gene function in organisms such as the fruit fly, Drosophila melanogaster, because Drosophila has been investigated for more than 90 years by thousands of geneticists who have created technologies and genetic resources that are among the best in biology. Another important factor is that fruit flies are genetically less complex, meaning that flies typically have only one gene of each kind, whereas mammals often have whole families of related genes. All this would be meaningless if flies and man tended to use different biological solutions to the same problems, but this is not the case and many biological pathways are used over and over again throughout evolution with only minor variations. Therefore, when we found that the Drosophila NF1 protein is 60% identical to human neurofibromin, we took this high level of identity as a further indication that the fly and human proteins may serve similar functions.

We created Drosophila NF1 loss-of-function mutations (the kind found in human patients) and observed that NF1-deficient flies are viable and normally patterned, but 20% reduced in size. The mutants also have a learning defect, lack a day-night activity rhythm, and in combination with certain other mutations show a thickening of the glial layer of peripheral nerves. Thus, several defects of Drosophila NF1 mutants resemble human disease symptoms. However, although the only known activity of neurofibromin is to serve as a negative regulator (a brake) of the Ras signaling protein, several defects of NF1 mutant flies were entirely insensitive to manipulating Ras signaling strength. Rather, these defects were rescued (i.e., cured) by stimulating the activity of a different signaling pathway, namely one that involves cyclic-AMP and Protein Kinase A. We also found that expressing human neurofibromin rescued all tested Drosophila mutant defects, and the overall growth defect of mutants was fully restored by expressing NF1 in just a few larval brain neurons. Based on this latter finding we believe that NF1 may control overall growth by regulating the production of a brain-derived growth promoting hormone.

Our proposed project aims to continue to exploit our Drosophila NF1 model to investigate questions that are not easily answered in other systems. For example, for technical reasons it has been impossible to conduct so-called structure-function studies with mammalian neurofibromin. In flies, however, we can easily generate and express various mutant transgenes and test whether they rescue any defects. This is important because it provides the only feasible way to assess the functional importance of protein segments, including those implicated as potentially important by those working on human neurofibromin. Therefore, the first of our four specific aims is to continue our work to assess the ability of mutant NF1 transgenes to rescue five different defects (reduced overall size, a biochemical abnormality in brain extracts, the learning defect, the rest-activity rhythm defect, and the glial thickening). Some of this work will be done by three collaborators using transgenic flies provided by us.

Among mammalian Ras-related proteins that can be inactivated by neurofibromin are three so-called conventional Ras proteins, which are often mutated in cancer, and three or more R-Ras family members, whose role in signal transduction is less well understood. Because it is important to know what precise pathways are regulated by NF1, our second aim is to determine whether Drosophila Ras1 or Ras2 (close relatives of conventional and R-Ras proteins, respectively) are the relevant targets of NF1 in various defects. We also will attempt to identify the exact mechanism underlying the NF1 growth deficiency. We have identified a small number of genes that are misexpressed in NF1 mutant brain, among which are some that code for hormones. We will investigate whether these hormones are directly responsible for the growth deficiency and also analyze why this defect is rescued by increasing cyclic-AMP. We are interested in this because similar hormonal defects may underlie some human symptoms, and we would like to confirm whether cyclic-AMP is a potential target for therapy in NF1. Finally, we have found that large parts of the NF1 protein are dispensable for size rescue, but only when the mutant proteins are overexpressed. This suggests that some NF1 patients may benefit from strategies that aim to increase NF1 gene expression. Unfortunately, determining what drives human NF1 gene expression in the central and peripheral nervous system is a daunting task, given the large size of the human gene. However, the Drosophila gene is 20 times smaller than the human gene and also expressed in the nervous system. Thus, we propose to identify factors that drive its expression and explore whether manipulating NF1 gene expression provides a viable strategy of interfering with disease development.