Synergy of Nf2 and p53 mutations in development of malignant tumours of neural crest origin

Article metrics

Abstract

Previously, we have mimicked human neurofibromatosis type 2 (NF2) in conditional Nf2 mutant (P0Cre;Nf2flox2/flox2) mice. Schwannomas, characteristic for NF2, were found at low frequency in older mice. Here, we report that these mice, upon additional hemizygosity for p53, rapidly develop multiple tumours showing features consistent with malignant peripheral nerve sheath tumours. Thus, p53 hemizygosity promotes tumorigenesis of mutant Nf2 peripheral nerve cells. In contrast, young P0Cre;Nf2flox2/+;p53+/− cis mice mainly succumb to Nf2/p53-related osteogenic tumours. Therefore, Cre-mediated early biallelic loss of Nf2 function in neural crest-derived cells hemizygous for p53 results in resistance to osteogenic tumours and increased susceptibility to peripheral nerve sheath tumours.

Introduction

Vestibular schwannomas are benign neural sheath tumours of the vestibular branch of the eighth cranial nerve. They occur either as bilateral tumours, the hallmark of the human hereditary cancer syndrome neurofibromatosis type 2 (NF2), or as sporadic unilateral tumours. Loss or mutational inactivation of both alleles of the NF2 gene in the Schwann cell is the rate-limiting step in neoplastic transformation. NF2 patients carrying one mutant NF2 allele are predisposed not only to bilateral vestibular schwannomas but also to schwannomas of other cranial, spinal and cutaneous nerves, to cranial and spinal meningiomas, as well as to presenile lens opacities and cerebral calcifications (McKusick, 1994). Loss of NF2 function is also frequently found in mesotheliomas (Bianchi et al., 1995). The product of the NF2 gene, called schwannomin or merlin (Rouleau et al., 1993; Trofatter et al., 1993), has structural similarity with members of the ezrin-radixin-moesin (ERM) family of proteins that link the cytoskeleton and the cell membrane (for a review, see, Bretscher et al., 2000). Together with ezrin and moesin, schwannomin and the transmembrane receptor CD44 form a molecular switch that specifies cell growth arrest or proliferation (Morrison et al., 2001).

The role of loss of Nf2 function in development and tumorigenesis has been studied in various knockout mouse models. Embryos homozygous for Nf2 mutations causing either indirect skipping of exon 3 or 2–3 or direct deletion of exon 2 are lethal (McClatchey et al., 1998; Giovannini et al., 2000). They fail to initiate gastrulation between embryonic days 6.5 and 7.0 (McClatchey et al., 1997). Heterozygous Nf2 mutant mice do not mimic human NF2 but develop later in life (10–30 months) mainly osteomas and osteosarcomas showing loss of heterozygosity (LOH) for Nf2 (McClatchey et al., 1998; Giovannini et al., 2000). In addition, indications for a role of Nf2 loss in metastatic potential have been found (McClatchey et al., 1998). Previously, we have mimicked NF2 in conditonal Nf2 mutant mice with P0 promoter-directed Cre recombinase activity in Schwann cells (Giovannini et al., 2000). P0Cre;Nf2flox2/flox2 mice develop Schwann cell hyperplasia and later in life schwannomas due to Cre-mediated biallelic Nf2 exon 2 deletion, whereas these features are absent in P0Cre;Nf2flox2/+ mice. The schwannomas occur incidentally or at a low frequency depending on the P0Cre strain. These data suggest that the loss of the Nf2 tumour suppressor function in man and P0Cre;Nf2flox2/flox2 mice is not observed in heterozygous Nf2 mutant mice possibly because of an insufficient rate of spontaneous loss of the Nf2 wild-type allele. P0Cre;Nf2flox2/flox2 mice also show neurocristopathy due to P0 promoter activity in tissues with neural crest-derived components. These include osseous metaplasia and lens cataract, two non-neoplastic features of NF2, and the infrequently found osteogenic tumours.

Interestingly, while in P0Cre;Nf2flox2/flox2 mice schwannomas occur late in life at low frequency, Schwann cell hyperplasia occurs early in life at a very high frequency. These observations strongly suggest that additional mutations are required for progression of Schwann cell hyperplasia to tumours. A good candidate may be loss of p53 as synergy between Nf2 and p53 mutations in tumorigenesis has been found in Nf2+/−;p53+/− mice (McClatchey et al., 1998). Nf2+/−;p53+/− cis mice (carrying the Nf2 and p53 mutation on the same chromosome 11) develop osteosarcomas and fibrosarcomas after a shorter latency period (5 months) than seen in either Nf2+/− or p53+/− mice. Loss of p53 may be important in tumorigenesis of P0-permissive murine cells in view of the observation that P0-SV40T antigen transgenic mice develop schwannomas (Messing et al., 1994).

We therefore assessed the role of p53 hemizygosity in conditional Nf2 knockout mice and found strong synergy between Nf2 and p53 mutations in development of malignant tumours of neural crest origin.

Results and discussion

Multiple MPNSTs in P0CreB;Nf2flox2/flox2;p53+/− mice

Previously, we have described the phenotypic effects of four types of Nf2 conditional knockout mice that were generated with different P0Cre transgenic lines (A–D) (Giovannini et al., 2000). The highest incidence of benign and malignant schwannomas was found in P0CreC;Nf2flox2/flox2 mice (35%), the lowest in P0CreB;Nf2flox2/flox2 mice (4%). We now refer to these malignant schwannomas and also to neurofibrosarcomas as malignant peripheral nerve sheath tumours (MPNSTs) based upon the latest WHO Classification of tumours of the nervous system (Kleihues and Sobin, 2000). Schwann cell hyperplasia occurred at high frequency in all four types of P0Cre;Nf2flox2/flox2 mice (75–100%). Based on the lowest percentage of Schwann cell tumours combined with a high percentage of Schwann cell hyperplasia, P0CreB;Nf2flox2/flox2 mice were chosen to investigate a possible cooperation of Nf2 and p53 mutations in Schwann cell tumorigenesis.

Mice lacking functional p53 succumb by the age of 10 months predominantly to lymphomas, p53+/− mice acquire tumours at older age, most often osteosarcomas and soft tissue sarcomas. These tumours show LOH for p53 (Donehower et al., 1992; Harvey et al., 1993; Jacks et al., 1994; Purdie et al., 1994).

P0CreB;Nf2flox2/flox2 and p53−/− mice (Donehower et al., 1992) were crossed to generate P0CreB;Nf2flox2/flox2;p53+/− mice, which were monitored over time for the appearance of tumours.

Nearly all P0CreB;Nf2flox2/flox2;p53+/− mice rapidly died starting at 2 months of age (median age 4.5 months), thus exhibiting a dramatically reduced survival compared to their P0CreB;Nf2flox2/flox2 littermates (Kaplan–Meier test: P=<0.0001) (Figure 1).

Figure 1
figure1

Survival of P0CreB conditional Nf2 knockout mice with or without mutant p53 allele(s). The viability of P0CreB;Nf2flox2/flox2 mice correlates with the number of p53 mutant alleles. P0CreB;Nf2flox2/+;p53+/− cis mice show a decreased survival rate compared to trans mice but live significantly longer than P0CreB;Nf2flox2/flox2;p53+/− mice. Survival is shown over a period of 24 months (numbers of considered moribund plus dead animals are in parentheses)

Of the 26 histologically examined P0CreB;Nf2flox2/flox2;p53+/− mice, 22 developed peripheral nerve tumours (85%). In contrast, only one of 27 P0CreB;Nf2flox2/flox2 mice within a follow-up period of 24 months developed a neurofibroma (at 9.5 months) (Giovannini et al., 2000). To compare the phenotypic abnormalities within a time window similar to that of P0CreB;Nf2flox2/flox2;p53+/− mice, eight young P0CreB;Nf2flox2/flox2 mice were killed before the age of 10 months. None of these developed peripheral nerve tumours (Table 1).

Table 1 Summary of the phenotypic consequences of mutant p53 allele(s) in P0CreB;Nf2flox2/flox2 mice

Macroscopically, the multiple tumours (one to seven per P0CreB;Nf2flox2/flox2;p53+/− mouse) usually originated from the peripheral nerves of the limbs (16 of 52 tumours), and from dorsal root ganglia (17 of 52 tumours). Microscopically, nearly all tumours (44) showed features that were consistent with MPNSTs. These were found in 20 of 26 animals (77%). They occurred as independent primary tumours that did not metastasize but aggressively invaded adjacent tissues. They were only slightly fibrotic, contained multiple mitotic cells, showed some nuclear pleiomorphism and were often highly anaplastic (Figure 2). Benign Schwann cell tumours were also seen, three schwannomas in three of 26 (11%) and four neurofibromas in two of 26 (8%) animals (the distinction between the two entities in the mouse could not always be established with certainty following the criteria of human pathology). In addition, four of eight (50%) P0CreB;Nf2flox2/flox2 mice in a p53−/− background showed four MPNSTs and four undifferentiated sarcomas mainly in ganglia already at 2–3.5 months of age (Table 1).

Figure 2
figure2

Histological analysis of phenotypic abnormalities in P0Cre;Nf2flox2/flox2;p53+/− and P0Cre;Nf2flox2/+;p53+/− cis mice. (a–c) Main nerve MPNST in a 5-month-old P0CreB;Nf2flox2/flox2;p53+/− mouse. (a) Cross-section of lower spine with main nerve of a paralysed mouse, showing tumour growth in perineurium. (b) Between arrow heads extensive tumour growth in perineurium of a single nerve; tumour cells are markedly hyperchromatic and slightly pleiomorphic. (c) Strong p75 immunoreactivity in nerve tumour, consistent with a tumour of neuroectodermal origin. Many tumour cells show variable positivity for p75, but not for S100. (d, e) Trigeminal nerve MPNST in a 5-month-old P0CreC;Nf2flox2/flox2;p53+/− mouse. (d) Diffuse sarcomatoid growth pattern with focally intracellular slit formation, suggestive of a neurogenic neoplasm. (e) Ultrastructural examination shows a high grade tumour consisting of cells with very elongated thin cell processes, basement membrane (arrow heads), and a few pinocytotic vesicles, features of human MPNST. (fh) Osteosarcomas and an MPNST in P0CreB;Nf2flox2/+;p53+/− cis mice. (f, g) Osteosarcoma emanating from the trigeminal nerve of a 5.5-month-old mouse. (f) Cross-section of the base of the scull (white arrow head: trigeminal nerve), showing a well-differentiated osteosarcoma unilaterally (black arrow head). (g) Margin of the same moderately invasive osteosarcoma showing well-differentiated bone lined by osteoblasts and osteoclasts. (h) Osteosarcoma and MPNST bulging into the nasal cavities of a 13.5-month-old mouse. Cross-section of partially obstructed nasal cavities (of a mouse with breathing difficulties) showing a well-differentiated osteosarcoma at one side (*), and an MPNST contralaterally (**), evolving from the trigeminal branches. The nasal septum is indicated by an arrow head. (a, b, d and fh) H&E stains. Magnification: (a, f, h) × 25; (c, g) × 100; (b) × 200; (d) × 400; (e) × 4400

Next, we investigated whether the peripheral nerve tumours were Nf2 and/or p53-related. All 24 peripheral nerve tumours of P0CreB;Nf2flox2/flox2;p53+/− mice that were analysed displayed deletion of Nf2 exon 2, and 23 of these tumours showed loss of the p53 wild-type allele (Figure 3a). One (benign) Schwann cell tumour in the intestine developed as a result of biallelic Nf2 inactivation only (Table 1). Also, the single peripheral nerve tumour and all four undifferentiated sarcomas that could be analysed in P0CreB;Nf2flox2/flox2;p53−/− mice exhibited deletion of Nf2 exon 2 (Table 1).

Figure 3
figure3

Southern blot analysis of tumour DNA. (a) Three representative tumours (lanes 1–3) of P0CreB;Nf2flox2/flox2;p53+/− mice show Cre-mediated Nf2 gene inactivation and LOH for p53. Left panel: XbaI–BamHI-digested DNA (probe B). The bands corresponding to the Nf2flox2 or Nf2+ allele (5.0 kb), and Nf2Δ2 allele (3.0 kb) are indicated. Right panel: BamHI-digested DNA (p53-specific probe). The bands corresponding to the p53+ allele (5.0 kb), p53 allele (6.5 kb), and the p53 pseudogene are indicated. (b) Four representative tumours (lanes 4–7) of P0CreB;Nf2flox2/+;p53+/− cis mice show Cre-mediated Nf2 gene inactivation and LOH for Nf2 and p53. Left panel: XbaI–BamHI-digested DNA (probe B). Middle panel: XbaI–SacI-digested DNA (probe A). The bands corresponding to the Nf2flox2 (8.0 kb) and Nf2+ allele (9.3 kb) are indicated. Right panel: BamHI-digested DNA (p53-specific probe). Positions of marker bands are indicated by arrows

To ascertain that the increased frequency of MPNST development was independent of the P0Cre transgenic line used, we generated P0CreC;Nf2flox2/flox2;p53+/− mice. Also these mice showed a strongly reduced survival compared to their P0CreC;Nf2flox2/flox2 littermates (Kaplan–Meier test: P=<0.0001, data not shown). Four of 12 mice succumbed to MPNSTs (33%), three of which at 5 months of age (data not shown). All four tumours showed deletion of Nf2 exon 2, and loss of the p53 wild-type allele was found in all three tumours that we could analyse by Southern blotting.

In man, the histogenesis of malignant peripheral nerve (sheath) tumours is controversial. In addition to undifferentiated cells, heterogeneity of peripheral nerve components has been observed based on ultrastructural and immunohistochemical features (Erlandson and Woodruff, 1982). Ultrastructural examination of two tumours regarded histologically as MPNST was performed, in one case tumour cells were anaplastic but based on location in the sciatic nerve it was regarded as high grade MPNST, the other case, a trigeminal tumour had features consistent with peripheral nerve tumour because of the presence of elongated Schwann cell processes, basement membrane, and a few pinocytic vesicles (Figure 2). We further characterised the tumours by immunohistochemistry. Nearly all MPNSTs (32/34) and three schwannomas analysed were S100/p75+ (Figure 2), demonstrating that these tumours originated from the neural crest (Zorick and Lemke, 1996).

Although additional p53 hemizygosity did not alter the high frequency of Schwann cell hyperplasia as observed in ganglia in P0CreB;Nf2flox2/flox2 mice, it increased the incidence of diffuse Schwann cell hyperplasia in major peripheral nerve trunks (from 1/8 to 8/26). In particular, these were the nerves of the brachial plexus, the sciatic and trigeminal nerves that became frequent sites of early malignant tumorigenesis in P0Cre;Nf2flox2/flox2;p53+/− mice. Also, in P0Cre;Nf2flox2/flox2;p53−/− mice we observed Schwann cell hyperplasia in both nerves and ganglia.

Concerning other neural crest-derived cells (e.g. osteoblasts and odontoblasts), no increased incidence of tumours or hyperplasia was found in P0Cre;Nf2flox2/flox2;p53+/− mice (Table 1).

Osteogenic tumours predominate in P0Cre;Nf2flox2/+;p53+/− cis mice

In a p53 wild-type background, heterozygous conditional Nf2 knockout mice (P0Cre;Nf2flox2/+;p53+/+) do not develop Schwann cell tumours (Giovannini et al., 2000). We investigated the tumorigenic phenotypes of 23 P0CreB;Nf2flox2/+;p53+/− cis and 53 P0CreB;Nf2flox2/+;p53+/− trans mice carrying Nf2flox2 and p53 alleles on the same and opposite chromosomes, respectively. Indeed, P0CreB;Nf2flox2/+;p53+/− cis mice showed a decreased survival rate compared to trans mice (P<0.0001) but lived significantly longer than P0CreB;Nf2flox2/flox2;p53+/− mice (P<0.0215) (Figure 1). None of 13 histologically examined mice with the trans configuration developed Schwann cell tumours, indicating that the presence of one p53 mutant allele was not sufficient for neoplastic transformation of Nf2A2/+ Schwann cells. However, in three of 16 mice analysed with the cis configuration we found MPNSTs (19%) in peripheral nerves at 3.5, 5, and 13.5 months of age (Figure 2, data not shown). The one tumour that was available for analysis showed both deletion of Nf2 exon 2 and LOH for Nf2 and p53. Two MPNSTs analysed by immunostaining were S100/p75+.

Interestingly, the great majority of the P0CreB;Nf2flox2/+;p53+/− cis mice (81%) showed osteogenic tumours, both osteosarcomas (63%, 11 in 10 mice) and osteomas (25%, six in four mice) that all arose in the neural crest-derived bone at a mean age of 5.5 months. Three bone tumours were analysed and showed loss of functional Nf2 and p53 (Figure 2, data not shown). Eight of the 13 mice with osteogenic tumours showed this type of tumour within the thin bones lining the nasal passages. This site is also predominantly affected in Nf2+/−;p53+/− cis mice but seldom in Nf2+/−;p53+/− trans mice (McClatchey et al., 1998). Since these neural crest-derived bone tumours already blocked the airways when still small in size, both P0CreB;Nf2flox2/+;p53+/− cis and Nf2+/−;p53+/− cis mice died at young age.

Also, P0CreB;Nf2flox2/+;p53+/− trans mice developed osteogenic tumours. However, these tumours arose after 12 months of age and were not restricted to neural crest-derived bone: Two tumours analysed showed LOH for p53 but no deletion of Nf2 exon 2 (data not shown). Also 22 of 27 other tumours of different types found in these mice were exclusively p53-related (the remaining five tumours were background tumours, data not shown). The lack of early bone tumours in mice with the trans configuration is in agreement with the fact that heterozygous p53 mutant mice have a relatively long tumour-free period (Donehower et al., 1995). Apparently, the probability to develop osteogenic tumours due to loss of both wild-type Nf2 and p53 alleles in trans is low in neural crest-derived osteogenic cells.

Intriguingly, in contrast to the P0CreB;Nf2flox2/+;p53+/− cis mice, P0CreB;Nf2flox2/flox2;p53+/− mice did not develop osteogenic tumours (Table 1). This could not be attributed to their killing prior to occurrence of these tumours, since the first osteogenic tumour in P0CreB;Nf2flox2/+;p53+/− cis mice was found at 4.5 months of age, an age at which 50% of the histologically analysed P0CreB;Nf2flox2/flox2;p53+/− mice were still alive (data not shown). We speculate that early loss of schwannomin leads to a recalibration of the CD44 signalling network in which schwannomin functions. CD44 is necessary for limb development by presentation of the signalling molecule FGF-8 to FGF receptors on the surface of limb mesenchymal cells (Sherman et al., 1998). It has been suggested that CD44 null embryos can compensate for the lack of CD44, since deficiencies in limb outgrowth are not seen in adult stage. Based on the results of Kaya et al. (1997) and Sherman et al. (1998), this compensation may no longer be possible after differentiation of the FGF-8-presenting apical ectodermal ridge cells and early limb outgrowth. Similarly, early loss of schwannomin may lead to compensatory signalling, thereby preventing the generation of osteogenic tumours. Loss of schwannomin at a later developmental stage may disable this compensatory signalling leading to the generation of osteogenic tumours. Anyhow, early biallelic loss of Nf2 function in neural crest-derived cells of the P0Cre;Nf2flox2/flox2;p53+/− mice shifts the tumour spectrum from osteogenic to peripheral nerve sheath tumours. Although it has been demonstrated that the developmental stage of Schwann cells may determine their susceptibility to tumour-initiating alterations (Jin et al., 1993; Sherman et al., 1999), a concomitant modification of the tumour spectrum has not been described yet.

Interestingly, P0CreB;Nf2flox2/+;p53+/− cis mice succumbed to malignant peripheral nerve tumours after a significantly longer latency period compared to P0CreB;Nf2flox2/flox2;p53+/− mice, although both categories of mice required loss of (part of) the chromosome carrying the wild-type p53 allele for malignant tumorigenesis. Our observations may indicate that the proliferation rate and, hence, the probability of LOH is higher in homozygous than in heterozygous Nf2 mutant cells or that functional loss of Nf2 increases the risk of abnormal chromosomal segregations. Further analyses have to resolve this.

Altogether, our results indicate that p53 suppresses malignant Schwann cell tumorigenesis initiated by biallelic Nf2 inactivation. Although synergism of Nf2 and p53 in the development of MPNST was anticipated based on the results of McClatchey et al. (1998), the outcome was not obvious since we recently observed lack of cooperation of these two genes in mouse meningeal cell tumorigenesis (Kalamarides et al., 2002). Recent data from other mouse models support a synergistic effect of neurofibromatosis type 1 (Nf1) and p53 genes in MPNST development (Cichowski et al., 1999; Vogel et al., 1999). In NF1 patients, mutation of p53 is associated with malignant transformation of plexiform neurofibromas into MPNSTs (Menon et al., 1990). In NF2 patients, however, benign Schwann cell tumours do not progress into MPNSTs. Thus, in mice, loss of function of either Nf1 or Nf2 can induce malignant Schwann cell tumorigenesis in cooperation with a p53 alteration.

Finally, the potential role of compensatory responses in the shift of the tumour spectrum in P0CreB;Nf2flox2/flox2;p53+/− mice as compared to P0CreB;Nf2flox2/+;p53+/− cis mice needs to be addressed in future studies. Our observations indicate that the time window in which gene inactivations occur is of critical importance for accurate tumour modelling.

Materials and methods

Generation and genotyping of conditional Nf2 and p53 mutant mice

The generation and genotyping of conditional Nf2 mutant mice of P0Cre strain B and C have been described (Giovannini et al., 2000). Genotyping of p53+/− mice, kindly provided by L. Donehower, Baylor College of Medicine, Houston, TX, USA, was performed by PCR analysis as described (Donehower et al., 1992).

P0Cre;Nf2flox2/flox2 mice were mated to p53+/− mice, producing P0Cre;Nf2flox2/+;p53+/− trans mice, which were then crossed with Nf2flox2/flox2 mice to generate P0Cre;Nf2flox2/flox2;p53+/− mice. The Nf2flox2/flox2;p53+/− mice of the latter offspring were mated with P0Cre transgenic mice to produce P0Cre;Nf2flox2/+;p53+/− cis mice.

Analysis of Nf2 exon 2 deletion, Nf2 and p53 LOH

Deletion of Nf2 exon 2 was detected by Southern analysis of XbaI–BamHI-digested tumour DNA using probe B and Nf2 LOH analysis as performed by Southern blotting of XbaI–SacI-digested DNA using probe A have been described (Giovannini et al., 2000).

LOH for p53 was determined by Southern analysis of BamHI-digested DNA using a probe corresponding to exons 2–6 of the p53 cDNA (Harvey et al., 1993).

Histological and electron microscopical analysis, immunohistochemistry

Mice were killed when moribund or held until 24 months of age. Histological analysis was performed as described (Giovannini et al., 1999).

For electron microscopical analysis, tissues were left overnight in Karnovsky's fixative, rinsed in s-collidine buffer, and postfixed for 1 h in 1% osmium tetroxide as described (Erlandson and Woodruff, 1982). Specimens then were embedded in epoxy resin and examined with a transmission electron microscope. For indirect immunoperoxidase assay with DAB substrate, the rabbit polyclonal antisera anti-mouse p75 LNGFR (Chemicon International) and antibovine S-100 protein (DAKO) were used and subsequently horse radish peroxidase-conjugated goat anti-rabbit Ig antibody (Amersham) as described (Giovannini et al., 1999).

References

  1. Bianchi AB, Mitsunaga S-I, Cheng JQ, Klein WM, Jhanwar SC, Seizinger B, Kley N, Klein-Szanto AJP and Testa JR . (1995). Proc. Natl. Acad. Sci. USA, 92, 10854–10858.

  2. Bretscher A, Chambers D, Nguyen R and Reczek D . (2000). Annu. Rev. Cell Dev. Biol., 16, 113–143.

  3. Cichowski K, Shih TS, Schmitt E, Santiago S, Reilly K, McLaughlin ME, Bronson RT and Jacks T . (1999). Science, 286, 2172–2176.

  4. Donehower LA, Harvey M, Slagle BL, McArthur MJ, Montgomery Jr CA, Butel JS and Bradley A . (1992). Nature, 356, 215–221.

  5. Donehower LA, Harvey M, Vogel H, McArthur MJ, Montgomery Jr CA, Park SH, Thompson T, Ford RJ and Bradley A . (1995). Mol. Carcinog., 14, 16–22.

  6. Erlandson RA and Woodruff JM . (1982). Cancer, 49, 273–287.

  7. Giovannini M, Robanus-Maandag E, Niwa-Kawakita M, van der Valk M, Woodruff JM, Goutebroze L, Merel P, Berns A and Thomas G . (1999). Genes Dev., 13, 978–986.

  8. Giovannini M, Robanus-Maandag E, van der Valk M, Niwa-Kawakita M, Abramowski V, Goutebroze L, Woodruff JM, Berns A and Thomas G . (2000). Genes Dev., 14, 1617–1630.

  9. Harvey M, McArthur MJ, Montgomery Jr CA, Bradley A and Donehower LA . (1993). FASEB J., 7, 938–943.

  10. Jacks T, Remington L, Williams BO, Schmitt EM, Halachmi S, Bronson RT and Weinberg RA . (1994). Curr. Biol., 4, 1–7.

  11. Jin JJ, Yu NA and Rajewsky MF . (1993). Cell Growth Differ., 4, 227–237.

  12. Kalamarides M, Niwa-Kawakita M, Leblois H, Abramowski V, Perricaudet M, Janin A, Thomas G, Gutmann DH and Giovannini M . (2002). Genes Dev., 16, 1060–1065.

  13. Kaya G, Rodriguez I, Jorcano JL, Vassalli P and Stamenkovic I . (1997). Genes Dev., 11, 996–1007.

  14. Kleihues P and Sobin LH . (2000). Cancer, 88, 2887.

  15. McClatchey AI, Saotome I, Mercer K, Crowley D, Gusella JF, Bronson RT and Jacks T . (1998). Genes Dev., 12, 1121–1133.

  16. McClatchey AI, Saotome I, Ramesh V, Gusella JF and Jacks T . (1997). Genes Dev., 11, 1253–1265.

  17. McKusick VA . (1994). Mendelian Inheritance in Man: A Catalogue of Human Genes and Genetic Disorders. The John Hopkins University Press: Baltimore & London.

  18. Menon AG, Anderson KM, Riccardi VM, Chung RY, Whaley JM, Yandell DW, Farmer GE, Freiman RN, Lee JK, Li FP, Barker DF, Ledbetter DH, Kleider A, Martuza RL, Gusella JF and Seizinger BR . (1990). Proc. Natl. Acad. Sci. USA, 87, 5435–5439.

  19. Messing A, Behringer RR, Wrabetz L, Hammang JP, Lemke G, Palmiter RD and Brinster RL . (1994). J. Neurosci., 14, 3533–3539.

  20. Morrison H, Sherman LS, Legg J, Banine F, Isacke C, Haipek CA, Gutmann DH, Ponta H and Herrlich P . (2001). Genes Dev., 15, 968–980.

  21. Purdie CA, Harrison DJ, Peter A, Dobbie L, White S, Howie SE, Salter DM, Bird CC, Wyllie AH, Hooper ML and Clarke AR . (1994). Oncogene, 9, 603–609.

  22. Rouleau GA, Mérel P, Lutchman M, Sanson M, Zucman J, Marineau C, Hoang-Xuan K, Demczuk S, Desmaze C, Plougastel B, Pulst SM, Lenoir G, Bijlsma E, Fashold R, Dumanski J, de Jong P, Parry D, Eldridge R, Aurias A, Delattre O and Thomas G . (1993). Nature, 363, 515–521.

  23. Sherman L, Sleeman JP, Hennigan RF, Herrlich P and Ratner N . (1999). Oncogene, 18, 6692–6699.

  24. Sherman L, Wainwright D, Ponta H and Herrlich P . (1998). Genes Dev., 12, 1058–1071.

  25. Trofatter JA, MacCollin MM, Rutter JL, Murrell JR, Duyao MP, Parry DM, Eldridge R, Kley N, Menon AG, Pulaski K, Haase VH, Ambrose CM, Munroe D, Bove C, Haines JL, Martuza RL, MacDonald ME, Seizinger BR, Short MP, Buckler AG and Gusella JF . (1993). Cell, 72, 791–800.

  26. Vogel KS, Klesse LJ, Velasco-Miguel S, Meyers K, Rushing EJ and Parada LF . (1999). Science, 286, 2176–2179.

  27. Zorick TS and Lemke G . (1996). Curr. Opin. Cell Biol., 8, 870–876.

Download references

Acknowledgements

We thank L.A. Donehower for the p53−/− mice; F. van der Ahé, K. Ankama, N. Bosnie, T. Maidment, H. Raasø, L. Rijswijk, and A. Zwerver for animal care; the CDTA (Orléans, France) for housing part of the mouse colony; R. Regnerus for genotyping of the mice; J. Bulthuis, K. de Goeij, D. Hoogervorst, L. Kuijper-Pietersma, and E. van Muylwijk for histotechnical assistance; R. Erlandson for electron microscopy; H. te Riele for critically reading the manuscript. This work was supported by Grants from the Commission of the European Communities (BMH4-CT96-1518), Ligue Nationale Française contre le Cancer, Association pour la Recherche sur le Cancer, Human Frontier Science Program (M.G.). United States Army Medical Research and Material Command Grant DAMD 17-00-0594 (M.G.).

Author information

Correspondence to Anton Berns.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • MPNST
  • neural crest
  • P0 promoter
  • conditional knockout mice
  • compensatory signalling

Further reading