2.2. NF1 gene

The NF1 gene is located in the long arm of chromosome 17 (17q11.2). (Cawthon et al. 1990, Viskochil et al. 1990). It is a very large gene spanning over 350 kb of genomic DNA (Marchuk et al. 1991) with 60 exons organized into four clusters, which are separated by four large introns (Fig. 1) (Li et al. 1995). The NF1 gene is ubiquitously expressed, resulting in 11-13 kb NF1 mRNA with many alternatively spliced variants (Fig. 1) (Gutmann et al. 1995, Gutmann et al. 1999, Vandenbroucke et al. 2002a). The NF1 gene is highly conserved between species. There is ~98% homology of the protein product between mouse and human, and ~60% homology between drosophila and human (Bernards et al. 1993, The et al. 1997).

The NF1 gene contains a domain (exons 21-27) related to Ras-GTPase-activating proteins (Ras-GAPs) (Fig. 1) (Ballester et al. 1990). The Ras-GAP proteins increase the intrinsic GTPase activity of proto-oncogene Ras. The NF1 gene has even longer homology to the yeast inhibitory regulator proteins (IRA), which are also involved in regulating the Ras-signaling pathway (Ballester et al. 1990). There are no other areas significantly homologous to other genes. One of the introns in the NF1 gene contains three embedded genes (OMGP, EVI2A and EVI2B), which are transcribed in an orientation opposite to NF1 gene (Fig. 1) (Viskochil et al. 1991). The functions of these three genes are unknown. The mutation rate of the NF1 gene is one of the highest known in the human genome, 3.1-6.5 x 10⁵ (Vogel & Motulsky 1997). Thus, ~50% of all NF1 patients lack a family history of the disease (Huson & Hughes 1994). It has been speculated that the high mutation rate is caused by the large size of the gene and the complexity of its processing (Upadhyaya et al. 1994). Recent advances in detecting mutations of the NF1 gene have made it possible to identify >95% of the mutations (Messiaen et al. 2000). There are no general hot spot areas for mutations in the NF1 gene, but the exons 10 and 37 have been shown to have the highest mutation rate and count for ~30% of the mutations (Messiaen et al. 2000).

The NF1 gene is ubiquitously expressed, as one would expect based on the diverse clinical manifestations of the disorder (DeClue et al. 1991, Gutmann et al. 1991, Daston & Ratner 1992). However, the highest level of NF1 gene expression is seen in neural tissues (Daston et al. 1992). There are three major alternatively spliced exons of the gene, but also many other splice variants with low expression levels (Fig. 1) (Gutmann et al. 1995, Gutmann et al. 1999, Vandenbroucke et al. 2002b). The most commonly expressed alternative isoform includes exon 23a (63bp), which is located in the GAP region. The isoform containing exon 23a has been shown to have decreased Ras-GAP activity (Andersen et al. 1993a). Two other common alternatively spliced exons are 9a (30bp) and 48a (54bp). The isoform containing exon 48a is highly expressed in muscle tissues (Gutmann et al. 1995), and expression of the 9a-containing isoform is seen during embryonic development of the brain (Gutmann et al. 1999). The functions of these two latter isoforms are not known. The 3‘ untranslated region of NF1 mRNA shares strong homology between human and mouse, suggesting that it is important for mRNA stability (Bernards et al. 1993). A few proteins have been identified that bind to the NF1 3‘ untranslated region, including the human RNA-binding protein (HuR), which is known to bind mRNAs of proto-oncogenes, cytokines and transcription factors (Haeussler et al. 2000). Furthermore, NF1 mRNA has recently been identified to be targeted towards the cell-cell adhesion zone (Yla-Outinen et al. 2002).