The formation of neurofibromas and their subsequent transformation to malignant peripheral nerve sheath tumors (MPNST) are downstream consequences of NF1 mutation in the Schwann cell lineage. However, at present we have only limited insight into the molecular changes and pathways that underlie neurofibroma formation and tumor progression. Recent advances in diagnostic pathology, molecular genetics and, in particular, genomics have provided new techniques to classify, analyze, and generate molecular models of tumors. We propose to use microarray gene expression profiling with Affymetrix GeneChip arrays to test the hypothesis that sequential Schwann cell gene expression changes delineate neurofibroma formation and peripheral nerve tumor progression.
Our hypothesis is that subsequent to NF1 mutation, stages in neurofibroma formation and transformation to malignancy are marked by distinct changes in Schwann cell gene expression. We propose to define changes across a wide spectrum of human and mouse NF1 models, and then to test if certain mouse models accurately model stages in human neurofibroma/MPNST pathogenesis.
Five well-established NF1 investigators (Ratner, Wallace, Lázaro, Giovannini, Stemmer-Rachamimov) have joined forces to work with a bioinformatics core (Aronow) and a statistical core (Page) to assemble and facilitate mining of multi-experiment integrated datasets. The cores will help organize and coordinate experiments, control chip data quality and identify expression patterns, gene groups and specific genes reflective of specific pathophysiologic processes relevant to NF1. The bioinformatics core will work with the statistical core to combine investigators' data and evaluate it using statistical programs. Other, critical, roles for the cores will be to 1) gather data from publicly accessible datasets, to provide complementary power and 2) make accumulated data available, through a website and a public database, to the scientific community.
Develop a Molecular Staging Profile of Peripheral Nerve Tumors in Human NF1, using gene expression profiling, by testing the following hypotheses:
Relate changes in gene expression in Nf1 mouse models to human tumors.
This group effort will cross-compare gene expression profile data over all existing NF1 peripheral nerve model systems to create a database that exceeds the sum of the individual investigations. Data integration and mining will identify specific and shared gene relationships among models. This group proposal is designed to overcome inadequate experimental design, poor technical performance, and rate limiting bioinformatics that can lead to cumbersome follow-through validation of key findings, allowing NF1 investigators to obtain maximal benefit of advanced genomics resources. The studies will improve understanding of how NF1 models are related or distinct and so identify the most appropriate models (mouse or cell culture) for pre-clinical trials. We will identify molecules associated with tumorigenesis; each will mark or be a candidate therapeutic target in NF1 disease. In addition, the correlation of gene expression patterns to tumor progression can lead to identification of prognostic or early detection markers in tumors (novel immunohistochemical markers) or in plasma (biochemical markers, similar to PSA in prostate cancer).