The neurofibromatoses: when less is more

The study of cancer predisposition syndromes presents unique opportunities to gain insights into the genetic events associated with tumor pathogenesis. Individuals with two inherited cancer syndromes, neurofibromatosis 1 (NF1) and neurofibromatosis 2 (NF2), develop both benign and malignant tumors. The corresponding genes mutated in these two disorders encode tumor suppressor proteins, termed neurofibromin (NF1) and merlin (NF2), which function in novel ways to regulate cell growth and differentiation. Neurofibromin inhibits cell proliferation, at least in part, by modulating mitogenic pathway signaling through inactivation of p21-ras. In contrast, merlin may act as a membrane-associated molecular switch that regulates cell-cell and cell-matrix signals transduced by cell surface receptors. Significant progress in our understanding of the genetics and biology of NF1 and NF2 has elucidated the roles of these genes in tumor initiation and progression.

Tumorigenesis in neurofibromatosis: new insights and potential therapies

The neurofibromatoses NF1 and NF2 are inherited cancer predisposition syndromes in which affected individuals are prone to development of mostly benign, but occasionally malignant, tumors. The NF1 and NF2 genes function as tumor suppressor genes (negative growth regulators), such that their loss of expression predisposes to tumor formation. Neurofibromin, the protein product of the NF1 gene, acts as a negative regulator of the ras proto-oncogene, to reduce cell growth. Merlin, the NF2 gene product, is involved in regulating cell proliferation and motility, and probably plays a role in integrating multiple cell-signaling pathways. By understanding the function of these tumor suppressors, we have a unique opportunity to develop targeted pharmacotherapeutic interventions for these disorders.

The NF2 interactor, hepatocyte growth factor-regulated tyrosine kinase substrate (HRS), associates with merlin in the "open" conformation and suppresses cell growth and motility

The neurofibromatosis 2 tumor suppressor protein, merlin or schwannomin, functions as a negative growth regulator; however, its mechanism of action is not known. In an effort to determine how merlin regulates cell growth, we analyzed a recently identified novel merlin interactor, hepatocyte growth factor-regulated tyrosine kinase substrate (HRS). We demonstrate that regulated overexpression of HRS in rat schwannoma cells results in similar effects as overexpression of merlin, including growth inhibition, decreased motility and abnormalities in cell spreading. Previously, we showed that merlin forms an intramolecular association between the N- and C-termini and exists in "open" and "closed" conformations. Merlin interacts with HRS in the unfolded, or open, conformation. This HRS binding domain maps to merlin residues 453-557. Overexpression of C-terminal merlin has no effect on HRS function, arguing that merlin binding to HRS does not negatively regulate HRS growth suppressor activity. These results suggest the possibility that merlin and HRS may regulate cell growth in schwannoma cells through interacting pathways.

Neurofibromatosis 1 (NF1) heterozygosity results in a cell-autonomous growth advantage for astrocytes

Individuals with neurofibromatosis 1 (NF1) develop low-grade astrocytomas at an increased frequency. To gain insight into the function of the Nf1 gene product as a growth regulator for astrocytes, we examined mice heterozygous for a targeted Nf1 mutation. In our previous studies, we demonstrated increased numbers of proliferating astrocytes in Nf1 heterozygote (Nf1+/-) mice in vivo. We now show that cultured Nf1+/- astrocytes exhibit a cell-autonomous growth advantage in vitro associated with increased p21-ras pathway activation. Furthermore, we demonstrate that Nf1+/-;wild-type N-ras mice have a similar astrocyte growth advantage in vitro and in vivo as either oncogenic N-ras or Nf1+/-; oncogenic N-ras mice. Lastly, mice heterozygous for targeted defects in both Nf1 and p53 as well as Nf1 and Rb exhibit 3- and 2.5-fold increases in astrocyte proliferation in vivo, respectively, suggesting that abnormalities in Nf1- and p53/Rb-regulated pathways cooperate in the heterozygous state to confer a growth advantage for brain astrocytes. Collectively, these results provide evidence for a cell-autonomous growth advantage in Nf1+/- astrocytes and suggest that some of the brain pathology in individuals with NF1 might result from reduced, but not absent, NF1 gene function.

Merlin, DAL-1, and progesterone receptor expression in clinicopathologic subsets of meningioma: a correlative immunohistochemical study of 175 cases

The molecular pathogenesis of meningiomas is poorly characterized. Loss of NF2 (merlin) expression has been reported in 30%-80% of all sporadic meningiomas. Recently, we found that loss of expression for a second Protein 4.1-family tumor suppressor. DAL-1, is also common. A biologically important role for progesterone receptor (PR) has also been proposed based on its reported inverse relationship with tumor grade. In order to better define the pathogenetic roles of these proteins, we studied the merlin, DAL-1, and PR immunoprofiles in 175 fully characterized meningiomas, including nonrecurring versus recurring benign, proliferative versus brain invasive atypical and anaplastic subtypes. Loss of expression for either Protein 4.1-family tumor suppressor (merlin or DAL-1) was almost universal (92%), with combined losses being common (58%). Individually, absence of merlin or DAL-1 protein was detected in 74% and 76% respectively, with no significant differences among the 5 subsets. PR immunoreactivity was commonly associated with retained DAL-1 expression (p < 0.001) and with tumor grade, with 51% of benign, 21% of atypical, and 11% of anaplastic tumors staining positive (p < 0.001). We conclude that PR immunohistochemistry may have diagnostic utility in meningothelial neoplasms. Protein 4.1-family tumor suppressor losses are likely important early events in meningioma pathogenesis, whereas PR expression is associated with benignity.

Loss of neurofibromin is associated with activation of RAS/MAPK and PI3-K/AKT signaling in a neurofibromatosis 1 astrocytoma

Neurofibromatosis 1 (NF1) is a common autosomal dominant cancer predisposition syndrome, in which 15% to 20% of affected individuals develop astrocytomas. Neurofibromin, the protein product of the NF1 gene, functions as a tumor suppressor, largely by inhibiting Ras activity. While loss of neurofibromin has been implicated in the molecular pathogenesis of other NF1-associated tumors, there is no formal evidence demonstrating loss of neurofibromin function in NF1-associated astrocytomas. In this report, we describe an NF1 patient from whom both astrocytoma tumor tissue as well as corresponding non-neoplastic white matter were available for analysis. Loss of neurofibromin expression was observed in the tumor and was associated with elevated levels of Ras-GTP. However, elevated Ras-GTP levels were not the result of oncogenic Ras mutations, altered p120-GAP function, growth factor receptor activation, or abnormal p53, Rb, or p16 expression. Furthermore, increased Raf-MAPK and PI3-K/Akt activity was detected in the NF1 astrocytoma compared with the corresponding normal white matter. These results support a role for neurofibromin as the critical GAP in the molecular pathogenesis of NF1 astrocytomas.

Malignant peripheral nerve sheath tumors in neurofibromatosis 1

One of the most clinically aggressive cancers associated with neurofibromatosis 1 (NF1) is the malignant peripheral nerve sheath tumor (MPNST). To determine the incidence and relative risk (RR) of MPNSTs in individuals with NF1, 1,475 individuals with NF1 were included from a cohort of patients examined by a single experienced geneticist from 1977 to 1996. The end points were incidence of MPNST, relative risk of MPNST, and relative risk associated with specific NF1 physical findings. Thirty-four individuals were identified with MPNST (2%). The relative risk of MPNST was higher than expected with an RR value of 113 (95% confidence interval [CI] = 78-158). The average 10-year annual incidence of MPNST between the second and fifth decade of life was roughly the same with a range of 0.0013 and 0.0068 MPNST per patient year. Most lesions occurred in the limbs (n = 18; 53%), and those with limb lesions survived longer than those with nonlimb MPNSTs. Pain associated with a mass was the greatest risk factor associated with MPNST development (RR = 31.4; 95% CI = 13.2-75.1). Further biological and epidemiological studies are needed to determine other factors that influence the risk of MPNST development in individuals affected with NF1. Am. J. Med. Genet. 93:388-392, 2000. Published 2000 Wiley-Liss, Inc.

Loss of neurofibromatosis 1 (NF1) gene expression in NF1-associated pilocytic astrocytomas

The critical role of the neurofibromatosis 1 (NF1) gene as a tumour suppressor has been clearly demonstrated for malignancies arising in NF1 patients. However, little is known about the more common benign tumours, such as the pilocytic astrocytoma. Most NF1-associated astrocytomas are benign and clinically non-progressive, though aggressive tumours are occasionally encountered. In this study, eight pilocytic astrocytomas from six individuals affected with NF1 were analysed for NF1 expression. All eight tumours demonstrated loss of neurofibromin expression by immunohistochemistry, which was confirmed in one case using Western blot analysis. Microsatellite analysis showed loss of a single NF1 allele (LOH) in two of four NF1-associated tumours. These results demonstrate that, in contrast to sporadic astrocytomas, loss of NF1 expression is an important primary genetic event in the pathogenesis of NF1-associated pilocytic astrocytomas.

The neurofibromatosis 2 tumor suppressor protein interacts with hepatocyte growth factor-regulated tyrosine kinase substrate

The neurofibromatosis 2 tumor suppressor protein schwannomin/merlin is commonly mutated in schwannomas and meningiomas. Schwannomin, a member of the 4.1 family of proteins, which are known to link the cytoskeleton to the plasma membrane, has little known function other than its ability to suppress tumor growth. Using yeast two-hybrid interaction cloning, we identified the HGF-regulated tyrosine kinase substrate (HRS) as a schwannomin interactor. We verified the interaction by both immunoprecipitation of endogenous HRS with endogenous schwannomin in vivo as well as by using bacterially purified HRS and schwannomin in vitro. We narrowed the regions of interaction to include schwannomin residues 256-579 and HRS residues from 480 to the end of either of two HRS isoforms. Schwannomin molecules with a L46R, L360P, L535P or Q538P missense mutation demonstrated reduced affinity for HRS binding. As HRS is associated with early endosomes and may mediate receptor translocation to the lysosome, we demonstrated that schwannomin and HRS co-localize at endosomes using the early endosome antigen 1 in STS26T Schwann cells by indirect immunofluorescence. The identification of schwannomin as a HRS interactor implicates schwannomin in HRS-mediated cell signaling.

Loss of DAL-1, a protein 4.1-related tumor suppressor, is an important early event in the pathogenesis of meningiomas

Meningiomas are common nervous system tumors, whose molecular pathogenesis is poorly understood. To date, the most frequent genetic alteration detected in these tumors is loss of heterozygosity (LOH) on chromosome 22q. This finding led to the identification of the neurofibromatosis 2 (NF2) tumor suppressor gene on 22q12, which is inactivated in 40% of sporadic meningiomas. The NF2 gene product, merlin (or schwannomin), is a member of the protein 4.1 family of membrane-associated proteins, which also includes ezrin, radixin and moesin. Recently, we identified another protein 4.1 gene, DAL-1 (differentially expressed in adenocarcinoma of the lung) located on chromosome 18p11.3, which is lost in approximately 60% of non-small cell lung carcinomas, and exhibits growth-suppressing properties in lung cancer cell lines. Given the homology between DAL-1 and NF2 and the identification of significant LOH in the region of DAL-1 in lung, breast and brain tumors, we investigated the possibility that loss of expression of DAL-1 was important for meningioma development. In this report, we demonstrate DAL-1 loss in 60% of sporadic meningiomas using LOH, RT-PCR, western blot and immunohistochemistry analyses. Analogous to merlin, we show that DAL-1 loss is an early event in meningioma tumorigenesis, suggesting that these two protein 4.1 family members are critical growth regulators in the pathogenesis of meningiomas. Furthermore, our work supports the emerging notion that membrane-associated alterations are important in the early stages of neoplastic transformation and the study of such alterations may elucidate the mechanism of tumorigenesis shared by other tumor types.

Clinic-based study of plexiform neurofibromas in neurofibromatosis 1

Individuals with neurofibromatosis 1 (NF1) develop both benign and malignant tumors at an increased frequency. One of the most common benign tumors in NF1 is the plexiform neurofibroma. These tumors cause significant morbidity and mortality on account of their propensity to grow and affect adjacent normal tissues. To determine the clinical profile of plexiform neurofibromas in NF1, we conducted a retrospective review of 68 NF1 patients with plexiform neurofibroma. In our series, 44% of tumors were detected by 5 years of age and most were located in the trunk and extremities. Only two patients developed malignant peripheral nerve sheath tumors in their preexisting plexiform neurofibromas. Lastly, we demonstrate that there were no specific clinical features of NF1 associated with the presence of plexiform neurofibroma. These results underscore the importance of careful serial examinations in the evaluation of patients with NF1.

Genetic and cellular defects contributing to benign tumor formation in neurofibromatosis type 1

Neurofibromatosis type 1 (NF1) is a common inherited cancer predisposition syndrome. The NF1 gene product, neurofibromin, is hypothesized to function as a tumor suppressor and nearly all NF1 patients develop benign peripheral nerve tumors. These neurofibromas presumably arise from NF1 inactivation in S100(+)Schwann cells, but there is no formal proof for this mechanism. We demonstrate that fibro-blasts isolated from neurofibromas carried at least one normal NF1 allele and expressed both NF1 mRNA and protein, whereas the S100(+)cells typically lacked the NF1 transcript. Our findings further indicate that additional molecular events aside from NF1 inactivation in Schwann cells and/or other neural crest derivatives contribute to neurofibroma formation.

Expression of the tuberous sclerosis complex gene products, hamartin and tuberin, in central nervous system tissues

Tuberous sclerosis complex (TSC) is a common genetic disorder in which affected individuals can develop mental retardation, developmental brain defects, and seizures. Two genetic loci are responsible for TSC: TSC1 on chromosome 9q and TSC2 on chromosome 16p. Here, we report our analysis of TSC1 (hamartin) and TSC2 (tuberin) protein expression in the central nervous system (CNS). Both tuberin and hamartin are expressed in neurons and astrocytes where they physically interact. In the mouse cerebellum in vivo, tuberin predominantly localizes to the perinuclear region of the Purkinje cell, whereas hamartin is distributed along neuronal or astrocytic processes. In contrast, both hamartin and tuberin demonstrate similar neuronal expression patterns in pure neuronal cultures in vitro. Additionally, hamartin is highly expressed in astrocytes in mixed neuron-glia cultures in vitro, suggesting that hamartin may be important for astrocyte growth control. Unlike tuberin, loss of hamartin expression was not observed in sporadic astrocytomas. These results suggest that tuberin and hamartin may differentially contribute to the CNS pathology in TSC.

Neurofibromatosis 2 tumor suppressor protein, merlin, forms two functionally important intramolecular associations

The neurofibromatosis 2 (NF2) tumor suppressor gene product, merlin (schwannomin) forms an intramolecular association that is required for negative growth regulation in vitro and in vivo. In an effort to develop a molecular model for merlin relevant to its tumor suppressor function, we further characterized merlin intramolecular folding. We now demonstrate that merlin forms two intramolecular associations, one between the amino terminal (N-term) domain and the carboxyl terminal (C-term) domain and another within the amino terminal domain (N-term/N-term) itself. The N-term/C-term domain interaction requires contact between residues 302-308 in the N-term and an intact exon 17 (residues 580-595) in the C-term domain. In addition, we demonstrate that the N-term/N-term domain self-interaction is required for N-term/C-term domain association. Lastly, we identify NF2 patient mutations that dramatically reduce each of these interactions in vitro. Based on these findings, we propose a model for merlin folding critical to its ability to function as a tumor suppressor protein.

Developmental regulation of a neuron-specific neurofibromatosis 1 isoform

We have identified a protein isoform of the neurofibromatosis 1 (NF1) gene (neurofibromin) containing the alternatively spliced exon 9a that is expressed in forebrain neurons. Exon 9a neurofibromin is localized in the cytoplasm, sediments in a P100 fraction, and is expressed throughout the soma and processes in cortical neurons in vitro. Expression of exon 9a neurofibromin is developmentally regulated, with expression first detected after postnatal day 2. The identification of this novel protein isoform with restricted neuronal expression suggests novel functions for neurofibromin in the postmitotic brain that are perhaps relevant to the learning disabilities observed in children with NF1.

Differential effects of cAMP in neurons and astrocytes. Role of B-raf

Mitogen-activated protein kinase (MAPK) activation provides cell type-specific signals important for cellular differentiation, proliferation, and survival. Cyclic AMP (cAMP) has divergent effects on MAPK activity depending on whether signaling is through Ras/Raf-1 or Rap1/B-raf. We found that central nervous system-derived neurons, but not astrocytes, express B-raf. In neurons, cAMP activated MAPK in a Rap1/B-raf-dependent manner, while in astrocytes, cAMP decreased MAPK activity. Inhibition of MAPK in neurons decreased neuronal growth factor-mediated survival, and activation of MAPK by cAMP analogues rescued neurons from death. Furthermore, constitutive expression of B-raf in astrocytoma cells increased MAPK activation, as seen in neurons, and enhanced proliferation. These data provide the first experimental evidence that B-raf is the molecular switch which dominantly permits differential cAMP-dependent regulation of MAPK in neurons versus astrocytes, with important implications for both survival and proliferation.

Overexpression of bax in human glioma cell lines

OBJECT

Cells that lose their ability to undergo apoptosis may promote the development of neoplasms and result in resistance to clinical treatment with DNA-damaging modalities such as radio- and chemotherapy. Four established human glioma cell lines that are resistant to apoptosis were transfected with the proapoptotic gene bax and assessed for their sensitivity to a proapoptotic stimulus.

METHODS

Two cell lines had a wild-type p53 genotype (U87 and D247MG) and two had mutant p53 genotypes (U138 and U373). Constitutive overexpression of murine bax was achieved in U138 and U373 only, which resulted in an increased sensitivity of these lines to the apoptosis-inducing effect of cytosine arabinoside (ara-C). Multiple attempts to produce constitutive overexpression of bax in U87 and D247MG cells resulted in spontaneous, near-complete cell loss. Vector-only control transfections were successful in all four cell lines. Inducible overexpression of bax was achieved in the U87 cells and elevated levels of BAX were observed as early as 6 hours after gene induction. This overexpression of BAX resulted in the spontaneous induction of apoptosis in these cells.

CONCLUSIONS

Overexpression of BAX in four human glioma cell lines resulted in increased sensitivity to apoptosis. In the two lines that had a wild-type p53 genotype, overexpression of BAX produced spontaneous apoptosis. In contrast, the lines that had mutant, nonfunctional P53 did not undergo spontaneous apoptosis, but they were rendered more sensitive to the apoptosis-inducing effect of ara-C. Modulation of BAX expression may be a useful therapeutic modality for gliomas, regardless of p53 genotype.

Haploinsufficiency for the neurofibromatosis 1 (NF1) tumor suppressor results in increased astrocyte proliferation

Individuals affected with neurofibromatosis 1 (NF1) harbor increased numbers of GFAP-immunoreactive cerebral astrocytes and develop astrocytomas that can lead to blindness and death. Mice heterozygous for a targeted Nf1 mutation (Nf1+/-) were employed as a model for the human disease to evaluate the hypothesis that reduced NF1 protein (neurofibromin) expression may confer a growth advantage for astrocytes, such that inactivation of only one NF1 allele is sufficient for abnormal astrocyte proliferation. Here, we report that Nf17+/- mice have increased numbers of cerebral astrocytes and increased astrocyte proliferation compared to wild-type littermates. Intriguingly, primary Nf1+/- astrocyte cultures failed to demonstrate a cell-autonomous growth advantage unless they were cocultured with C17 neuronal cells. This C17 neuronal cell-induced Nf1+/- increase in proliferation was blocked by MEK inhibition (PD98059), suggesting a p21-ras-dependent effect. Furthermore, mice heterozygous for a targeted mutation in another GAP molecule, p120-GAP, demonstrated no increases in cerebral astrocyte number. These findings suggest that reduced NF1 expression results in a cell context-dependent increase in astrocyte proliferation that may be sufficient for the development of astrocytic growth abnormalities in patients with NF1.

Molecular analysis of malignant triton tumors

Triton tumors are rare variants of malignant peripheral nerve sheath tumor (MPNST) with muscle differentiation, often seen in patients with neurofibromatosis 1 (NF1). Individuals affected with NF1 harbor mutations in the NF1 tumor suppressor gene and develop neurofibromas and MPNSTs. The NF1 gene is expressed in Schwann cells and its expression is lost in schwannian neoplasms, suggesting a role in malignant development. Separately, there is evidence that p53 suppressor gene mutations are involved in MPNSTs. To determine the role of the NF1 and p53 genes in the development of the malignant Triton tumor we examined 2 such tumors, 1 from a 3-year-old boy without clinical manifestations of NF1 and another from a 24-year-old man with NF1. Histological analysis of these tumors showed both neural and muscle differentiation with S-100 and desmin immunoreactivity, respectively. Reverse transcribed RNA polymerase chain reaction (RT-PCR) of NF1 mRNA showed NF1 expression in the sporadic tumor. Strong nuclear immunoreactivity for p53 was observed throughout the malignant population in both tumors. This was confirmed by loss of heterozygosity for p53 in the non-NF1 patient, suggesting that p53 is involved in both hereditary and sporadic Triton tumors. The finding of preserved NF1 gene expression in the non-NF1-related Triton tumor suggests that different genetic events predispose to the development of this rare neoplasm in sporadic cases.

Intracranial gliomas in neurofibromatosis type 1

Optic pathway gliomas and brainstem gliomas are the predominant intracranial neoplasms associated with neurofibromatosis type 1 (NF1). Before the past 15 years, studies of optic pathway gliomas in NF1 were hampered by the inaccurate diagnosis of NF1, the unavailability of noninvasive neuroimaging techniques, and the frequent rendering of what would now be considered unnecessary, overly aggressive therapy. When studied systematically, these tumors behave in a much more benign fashion than their counterparts in children who do not have NF1. While they may cause symptoms in as many of 50% of cases, progression to the point where specific intervention is deemed necessary is unusual. Consequently, screening neuroimaging of asymptomatic patients is unwarranted. Because optic pathway tumors universally arise in children younger than 7 years of age, all such children should undergo yearly ophthalmologic evaluations and annual assessments of growth to monitor for signs of precocious puberty. Am. J. Med. Genet. (Semin. Med. Genet.) 89:38-44, 1999.

Establishment of primary vestibular schwannoma cultures from neurofibromatosis type-2 patients

Primary cultures were established from vestibular schwannomas of NF2 patients. The cultured tumor cells were selectively amplified by growth factor supplemented medium and characterized by immunocytochemistry. NF2 cDNA was amplified by RT-PCR and mutations were detected by both the non-isotopic RNase cleavage assay and direct DNA sequencing, no detectable wild-type NF2 transcript was found in cDNA from the cultured cells. Distinguishable morphology and growth rate differences have been observed in different passages of the primary cells. The data suggest that a pure schwannoma primary culture can be established and could be very useful in vitro model for further understanding the NF2 gene function in Schwann cells.