Neurofibromatosis type 1 and Ras-mediated signaling: filling in the GAPs

The neurofibromatosis type I pre-mRNA is a novel target of CELF protein-mediated splicing regulation

The CUG-BP and ETR-3 like factors (CELF) are a family of six highly conserved RNA-binding proteins that preferentially bind to UG-rich sequences. One of the key functions of these proteins is to mediate alternative splicing in a number of tissues, including brain, heart and muscle. To fully understand the function of CELF proteins, it is important to identify downstream targets of CELF proteins. In this communication, we report that neurofibromatosis type I (NF1) exon 23a is a novel target of CELF protein-mediated splicing regulation in neuron-like cells. NF1 regulates Ras signaling, and the isoform that excludes exon 23a shows 10 times greater ability to down-regulate Ras signaling than the isoform that includes exon 23a. Five of the six CELF proteins strongly suppress the inclusion of NF1 exon 23a. Over-expression or siRNA knockdown of these proteins in cell transfection experiments altered the levels of NF1 exon 23a inclusion. In vitro binding and splicing analyses demonstrate that CELF proteins block splicing through interfering with binding of U2AF⁶⁵. These studies, combined with our previous investigations demonstrating a role for Hu proteins and TIA-1/TIAR in controlling NF1 exon 23a inclusion, highlight the complex nature of regulation of this important alternative splicing event.

Regulation of neuron-specific alternative splicing of neurofibromatosis type 1 pre-mRNA

Neurofibromatosis type 1 (NF1) is one of the most common heritable autosomal dominant disorders. Alternative splicing modulates the function of neurofibromin, the NF1 gene product, by inserting the in-frame exon 23a into the region of NF1 mRNA that encodes the GTPase-activating protein-related domain. This insertion, which is predominantly skipped in neurons, reduces the ability of neurofibromin to regulate Ras by 10-fold. Here, we report that the neuron-specific Hu proteins control the production of the short protein isoform by suppressing inclusion of NF1 exon 23a, while TIA-1/TIAR proteins promote inclusion of this exon. We identify two binding sites for Hu proteins, located upstream and downstream of the regulated exon, and provide biochemical evidence that Hu proteins specifically block exon definition by preventing binding of essential splicing factors. In vitro analyses using nuclear extracts show that at the downstream site, Hu proteins prevent binding of U1 and U6 snRNPs to the 5' splice site, while TIAR increases binding. Hu proteins also decrease U2AF binding at the 3' splice site located upstream of exon 23a. In addition to providing the first mechanistic insight into tissue-specific control of NF1 splicing, these studies establish a novel strategy whereby Hu proteins regulate RNA processing.

Recurrent chromosomal imbalances and structurally abnormal breakpoints within complex karyotypes of malignant peripheral nerve sheath tumour and malignant triton tumour: a cytogenetic and molecular cytogenetic study

BACKGROUND

Cytogenetic studies of malignant peripheral nerve sheath tumours (MPNSTs) and malignant triton tumours (MTTs) are rare.

AIMS

To undertake cytogenetic analysis of these tumours.

METHODS

Conventional cytogenetic analysis of 21 MPNSTs and MTTs from 17 patients (nine with peripheral neurofibromatosis (NF1)) was carried out using standard culture and harvesting procedures. For a more precise identification of composite structural rearrangements and marker chromosomes, spectral karyotypic analysis (SKY) was applied to a subset of cases. In addition, EGFR gene copy number was assessed by fluorescence in situ hybridisation (FISH) analysis in a subset of cases.

RESULTS

Cytogenetic analysis revealed predominantly complex karyotypes. SKY analysis was useful in further defining many structural anomalies. Structural aberrations most frequently involved chromosomal bands or regions 1p31-36, 4q28-35, 7p22, 11q22-23, 19q13, 20q13, and 22q11-13. Overall, loss of chromosomal material was much more common than gain. Loss of chromosomes or chromosomal regions 1p36 (48%), 3p21-pter (52%), 9p23-pter (57%), 10 (48%), 11q23-qter (48%), 16/16q24 (62%), 17(43%), and 22/22q (48%), and gains of 7/7q (29%) and 8/8q (29%) were most prominent. These gains and losses were distributed equally between MPNST and MTT, demonstrating that these entities are similar with respect to recurrent genomic imbalances. Similarly, none of the recurrent chromosomal breakpoints or imbalances was restricted to either NF1 associated or sporadic MPNSTs. FISH analysis was negative for amplification.

CONCLUSIONS

These cytogenetic and molecular cytogenetic findings expand the knowledge of chromosomal alterations in MPNST and MTT, and point to possible recurring regions of interest.

Somatic Mitochondrial DNA Mutations in Neurofibromatosis Type 1-Associated Tumors

Neurofibromatosis type 1 is an autosomal dominantly inherited disease predisposing to a multitude of tumors, most characteristically benign plexiform neurofibromas and diffuse cutaneous neurofibromas. We investigated the presence and distribution of somatic mitochondrial DNA (mtDNA) mutations in neurofibromas and in nontumor tissue of neurofibromatosis type 1 patients. MtDNA alterations in the entire mitochondrial genome were analyzed by temporal temperature gradient gel electrophoresis followed by DNA sequencing. Somatic mtDNA mutations in tumors were found in 7 of 19 individuals with cutaneous neurofibromas and in 9 of 18 patients with plexiform neurofibromas. A total of 34 somatic mtDNA mutations were found. All mutations were located in the displacement loop region of the mitochondrial genome. Several plexiform neurofibromas from individual patients had multiple homoplasmic mtDNA mutations. In cutaneous neurofibromas, the same mtDNA mutations were always present in tumors from different locations of the same individual. An increase in the proportion of the mutant mtDNA was always found in the neurofibromas when compared with nontumor tissues. The somatic mtDNA mutations were present in the Schwann cells of the analyzed multiple cutaneous neurofibromas of the same individual. The observed dominance of a single mtDNA mutation in multiple cutaneous neurofibromas of individual patients indicates a common tumor cell ancestry and suggests a replicative advantage rather than random segregation for cells carrying these mutated mitochondria.

Somatic inactivation of Nf1 in hematopoietic cells results in a progressive myeloproliferative disorder

The NF1 tumor suppressor gene encodes a guanosine triphosphotase (GTPase)-activating protein that negatively regulates Ras signaling and is inactivated in a subset of juvenile myelomonocytic leukemias (JMMLs). Adoptive transfer of fetal liver cells from Nf1 mutant mice models JMML; however, this system has important limitations as a platform for performing biologic and preclinical studies. We have exploited the interferon-inducible Mx1-Cre transgene to ablate a conditional mutant Nf1 allele in hematopoietic cells. Somatic inactivation of Nf1 induces a myeloproliferative disorder with 100% penetrance that is associated with a sub-acute clinical course, tissue infiltration by myeloid cells, hypersensitivity to granulocyte-macrophage colony stimulating factor, hyperproliferation, and resistance to apoptosis. These Mx1-Cre, Nf1flox/flox mice establish a tractable experimental model for testing therapeutics and for identifying mutations that cooperate with hyperactive Ras in myeloid leukemogenesis. (Blood. 2004;103:4243-4250)

Quantitative effects of Nf1 inactivation on in vivo hematopoiesis

The NF1 tumor-suppressor gene is frequently inactivated in juvenile myelomonocytic leukemia, and Nf1 mutant mice model this myeloproliferative disorder (MPD). Competitive repopulation assays were performed to quantify the proliferative advantage of Nf1(-/-) hematopoietic cells in vivo. Nf1 mutant stem cells demonstrated a growth advantage that was greatest in myeloid lineage cells and least pronounced in T lymphocytes. Surprisingly, although low numbers of Nf1-deficient cells consistently outcompeted wild-type cells, levels of chimerism were stable over months of observation, and MPD was not observed unless threshold numbers of mutant cells were injected. These data showing that normal competitor cells can strongly modulate the growth of mutant populations in vivo have general implications for modeling cancer in the mouse. In particular, strains in which cancer-associated mutations are expressed in fields of target cells may not accurately model early events in tumorigenesis because they eliminate the requirement for a mutant clone to outcompete resident normal cells.

The angiogenic factor midkine is aberrantly expressed in NF1-deficient Schwann cells and is a mitogen for neurofibroma-derived cells

Loss of the tumor suppressor gene NF1 in neurofibromatosis type 1 (NF1) contributes to the development of a variety of tumors, including malignant peripheral nerve sheath tumors (MPNST) and benign neurofibromas. Of the different cell types found in neurofibromas, Schwann cells usually provide between 40 and 80%, and are thought to be critical for tumor growth. Here we describe the identification of growth factors that are upregulated in NF1-/- mouse Schwann cells and are potential regulators of angiogenesis and cell growth. Basic fibroblast growth factor (FGF-2), platelet-derived growth factor (PDGF) and midkine (MK) were found to be induced by loss of neurofibromin and MK was further characterized. MK was induced in human neurofibromas, schwannomas, and various nervous system tumors associated with NF1 or NF2; midkine showed an expression pattern overlapping but distinct from its homolog pleiotrophin (PTN). Immunohistochemistry revealed expression of MK in S-100 positive Schwann cells of dermal and plexiform neurofibromas, and in endothelial cells of tumor blood vessels, but not in normal blood vessels. Furthermore, MK demonstrated potent mitogenic activity for human systemic and brain endothelial cells in vitro and stimulated proliferation and soft agar colony formation of human MPNST derived S100 positive cells and fibroblastoid cells derived from an NF1 neurofibroma. The data support a possible central role for MK as a mediator of angiogenesis and neurofibroma growth in NF1. Oncogene (2001) 20, 97 - 105.

Kinase inhibitors in cancer therapy: a look ahead

The most essential kinases involved in cell membrane receptor activation, signal transduction and cell cycle control or programmed cell death and their interconnections are reviewed. In tumours, the genes of many of those kinases are mutated or amplified or the proteins are overexpressed. The use of key kinases offers the possibility to screen in vitro for synthetic small molecule kinase inhibitors. In view of the many interconnections of cellular kinases, their role in preventing or inducing programmed cell death and the possibility that a considerable number of signal transducing proteins are still unknown, cellular test systems are recommended in which the respective key kinase or one of its main partner molecules are overexpressed.

Characterization of Saccharomyces cerevisiae strains expressing ira1 mutant alleles modeled after disease-causing mutations in NF1

The 2818 amino acids of neurofibromin, the product of the human NF1 gene, include a 230 amino acid Ras-GAP related domain (GRD). Functions which may be associated with the rest of the protein remain unknown. However, many NF1 mutations in neurofibromatosis 1 patients are found downstream of the GRD, suggesting that the C-terminal region of the protein is also functionally important. Since the C-terminal region of neurofibromin encompassing these mutations is homologous with the corresponding regions in the two Saccharomyces cerevisiae Ras-GAPs, Ira1p and Ira2p, we chose yeast as a model system for functional exploration of this region (Ira-C region). Three missense mutations that affect the Ira-C region of NF1 were used as a model for the mutagenesis of IRA1. The yeast phenotypes of heat shock sensitivity, iodine staining, sporulation efficiency, pseudohyphae formation, and GAP activity were scored. Even though none of the mutations directly affected the Ira1p-GRD, mutations at two of the three sites resulted in a decrease in the GAP activity present in ira1 cells. The third mutation appeared to disassociate the phenotypes of sporulation ability and GAP activity. This and other evidence suggest an effector function for Ira1p.

p21 ras and phosphatidylinositol-3 kinase are required for survival of wild-type and NF1 mutant sensory neurons

Nerve growth factor (NGF) is a required differentiation and survival factor for sympathetic and a majority of neural crest-derived sensory neurons in the developing vertebrate peripheral nervous system. Although much is known about the function of NGF, the intracellular signaling cascade that it uses continues to be a subject of intense study. p21 ras signaling is considered necessary for sensory neuron survival. How additional intermediates downstream or in parallel may function has not been fully understood yet. Two intracellular signaling cascades, extra cellular regulated kinase (erk) and phosphatidylinositol-3 (PI 3) kinase, transduce NGF signaling in the pheochromocytoma cell line PC12. To elucidate the role these cascades play in survival and differentiation, we used a combination of recombinant adenoviruses and chemical inhibitors to perturb these pathways in sensory neurons from wild-type mice and mice deficient for neurofibromin in which the survival and differentiation pathway is constitutively active. We demonstrate that ras activity is both necessary and sufficient for the survival of embryonic sensory neurons. Downstream of ras, however, the erk cascade is neither required nor sufficient for neuron survival or overall differentiation. Instead, the activity of PI 3 kinase is necessary for the survival of the wild-type and neurofibromin-deficient neurons. Therefore, we conclude that in sensory neurons, NGF acts via a signaling pathway, which includes both ras and PI 3 kinase.

Apoptosis of human BEL-7402 hepatocellular carcinoma cells released by antisense H-ras DNA--in vitro and in vivo studies

Recent findings suggest that over-expression of activated H-ras inhibited apoptotic cell death by blocking the activity of apoptotic endonuclease(s). This study was designed using antisense H-ras oligodeoxynucleotides (ODN) to evaluate whether alterations of H-ras expression in BEL-7402 human hepatocellular carcinoma cells could influence the induction of apoptosis in vitro and in vivo. We found that, in vitro, continuous suppression of H-ras expression could decrease the proliferation of BEL-7402 cells and inhibit H-ras-induced entry into S phase. In situ end labeling showed that a large number of cells underwent apoptotic cell death after treatment with antisense H-ras ODN (P < 0.01), and gel electrophoresis of DNA extracted from these cells demonstrated a typical DNA ladder, characteristic of apoptosis. In vivo study indicated that pretreatment with antisense H-ras significantly retarded tumor growth in comparison with the untreated controls or tumors treated with non-specific ODN (P < 0.01, P < 0.01). In situ end-labeling revealed that pronounced apoptotic nuclei were also present in the tissue treated with antisense H-ras ODN (P < 0.01). Immunocyto-histochemical study showed that expression of p21H-ras was significantly decreased after treatment with antisense H-ras. These results indicate that suppression of H-ras over-expression by antisense ODN could effectively inhibit tumor growth and revive the apoptotic pathway by releasing the activity of apoptotic endonuclease(s). The data also suggest the need for further studies to elucidate molecular events involved in antisense H-ras-released apoptosis and evaluate its therapeutic implications.