Nf1 deficiency causes Ras-mediated granulocyte/macrophage colony stimulating factor hypersensitivity and chronic myeloid leukaemia.

The Ras signal transduction pathway is often deregulated in human myeloid leukaemia. For example, activating point mutations in RAS genes are found in some patients with juvenile chronic myelogenous leukaemia (JCML), while other patients with JCML show loss of the neurofibromatosis type 1 (NF1) gene, a Ras GTPase activating protein. By generating mice whose haematopoietic system is reconsituted with Nf1 deficient haematopoietic stem cells we show that Nf1 gene loss, by itself, is sufficient to produce the myeloproliferative symptoms associated with human JCML. We also provide evidence to indicate that Nf1 gene loss induces myeloproliferative disease through a Ras-mediated hypersensitivity to granulocyte/macrophage-colony stimulating factor (GM-CSF). Finally, we describe a genetic screen for identifying genes that cooperate with Nf1 gene loss during progression to acute myeloid leukaemia.

Maternal imprinting of the mouse Snrpn gene and conserved linkage homology with the human Prader-Willi syndrome region.

Prader-Willi syndrome (PWS) is associated with paternal gene deficiencies in human chromosome 15q11-13, suggesting that PWS is caused by a deficiency in one or more maternally imprinted genes. We have now mapped a gene, Snrpn, encoding a brain-enriched small nuclear ribonucleoprotein (snRNP)-associated polypeptide SmN, to mouse chromosome 7 in a region of homology with human chromosome 15q11-13 and demonstrated that Snrpn is a maternally imprinted gene in mouse. These studies, in combination with the accompanying human mapping studies showing that SNRPN maps in the Prader-Willi critical region, identify SNRPN as a candidate gene involved in PWS and suggest that PWS may be caused, in part, by defects in mRNA processing.

De novo deletions of SNRPN exon 1 in early human and mouse embryos result in a paternal to maternal imprint switch.

Prader-Willi syndrome (PWS) is a neurogenetic disease characterized by infantile hypotonia, gonadal hypoplasia, obsessive behaviour and neonatal feeding difficulties followed by hyperphagia, leading to profound obesity. PWS is due to a lack of paternal genetic information at 15q11-q13 (ref. 2). Five imprinted, paternally expressed genes map to the PWS region, MKRN3 (ref. 3), NDN (ref. 4), NDNL1 (ref. 5), SNRPN (refs 6-8 ) and IPW (ref. 9), as well as two poorly characterized framents designated PAR-1 and PAR-5 (ref. 10). Imprinting of this region involves a bipartite 'imprinting centre' (IC), which overlaps SNRPN (refs 10,11). Deletion of the SNRPN promoter/exon 1 region (the PWS IC element) appears to impair the establishment of the paternal imprint in the male germ line and leads to PWS. Here we report a PWS family in which the father is mosaic for an IC deletion on his paternal chromosome. The deletion chromosome has acquired a maternal methylation imprint in his somatic cells. We have made identical findings in chimaeric mice generated from two independent embryonic stem (ES) cell lines harbouring a similar deletion. Our studies demonstrate that the PWS IC element is not only required for the establishment of the paternal imprint, but also for its postzygotic maintenance.

A mouse model for Prader-Willi syndrome imprinting-centre mutations.

Imprinting in the 15q11-q13 region involves an 'imprinting centre' (IC), mapping in part to the promoter and first exon of SNRPN. Deletion of this IC abolishes local paternally derived gene expression and results in Prader-Willi syndrome (PWS). We have created two deletion mutations in mice to understand PWS and the mechanism of this IC. Mice harbouring an intragenic deletion in Snrpn are phenotypically normal, suggesting that mutations of SNRPN are not sufficient to induce PWS. Mice with a larger deletion involving both Snrpn and the putative PWS-IC lack expression of the imprinted genes Zfp127 (mouse homologue of ZNF127), Ndn and Ipw, and manifest several phenotypes common to PWS infants. These data demonstrate that both the position of the IC and its role in the coordinate expression of genes is conserved between mouse and human, and indicate that the mouse is a suitable model system in which to investigate the molecular mechanisms of imprinting in this region of the genome.

Learning deficits, but normal development and tumor predisposition, in mice lacking exon 23a of Nf1.

Neurofibromatosis type 1 (NF1) is a commonly inherited autosomal dominant disorder. Previous studies indicated that mice homozygous for a null mutation in Nf1 exhibit mid-gestation lethality, whereas heterozygous mice have an increased predisposition to tumors and learning impairments. Here we show that mice lacking the alternatively spliced exon 23a, which modifies the GTPase-activating protein (GAP) domain of Nf1, are viable and physically normal, and do not have an increased tumor predisposition, but show specific learning impairments. Our findings have implications for the development of a treatment for the learning disabilities associated with NF1 and indicate that the GAP domain of NF1 modulates learning and memory.

A candidate mouse model for Prader-Willi syndrome which shows an absence of Snrpn expression.

The best examples of imprinting in humans are provided by the Angelman and Prader-Willi syndromes (AS and PWS) which are associated with maternal and paternal 15q11-13 deletions, respectively, and also with paternal and maternal disomy 15. The region of the deletions has homology with a central part of mouse chromosome 7, incompletely tested for imprinting effects. Here, we report that maternal duplication for this region causes a murine imprinting effect which may correspond to PWS. Paternal duplication was not associated with any detectable effect that might correspond with AS. Gene expression studies established that Snrpn is not expressed in mice with the maternal duplication and suggest that the closely-linked Gabrb-3 locus is not subject to imprinting. Finally, an additional new imprinting effect is described.

The human pseudoautosomal GM-CSF receptor alpha subunit gene is autosomal in mouse.

The gene encoding the granulocyte macrophage colony stimulating factor receptor alpha subunit (CSF2RA) has previously been mapped to the pseudoautosomal region of the human sex chromosomes. In contrast, we report that the murine locus, Csf2ra, maps to an autosome in the laboratory mouse. By in situ hybridization and genetic mapping, Csf2ra maps at telomeric band D2 of mouse chromosome 19. This first instance of a pseudoautosomal locus in human being autosomal in mouse, indicates incomplete conservation between the human and mouse X chromosomes and suggests that the genetic content of the pseudoautosomal region may differ between species of eutherian mammals due to chromosomal rearrangements.

Maternal methylation imprints on human chromosome 15 are established during or after fertilization.

Prader-Willi syndrome (PWS) is a neurogenetic disorder that results from the lack of transcripts expressed from the paternal copy of the imprinted chromosomal region 15q11-q13 (refs. 1,2). In some patients, this is associated with a deletion of the SNURF-SNRPN exon 1 region inherited from the paternal grandmother and the presence of a maternal imprint on the paternal chromosome. Assuming that imprints are reset in the germ line, we and others have suggested that this region constitutes part of the 15q imprinting center (IC) and is important for the maternal to paternal imprint switch in the male germ line. Here we report that sperm DNA from two males with an IC deletion had a normal paternal methylation pattern along 15q11-q13. Similar findings were made in a mouse model. Our results indicate that the incorrect maternal methylation imprint in IC deletion patients is established de novo after fertilization. Moreover, we found that CpG-rich regions in SNURF-SNRPN and NDN, which in somatic tissues are methylated on the maternal allele, are hypomethylated in unfertilized human oocytes. Our results indicate that the normal maternal methylation imprints in 15q11-q13 also are established during or after fertilization.

Mechanisms of genomic imprinting.

A small number of mammalian genes undergo the process of genomic imprinting whereby the expression level of the alleles of a gene depends upon their parental origin. In the past year, attention has focused on the mechanisms that determine parental-specific expression patterns. Many imprinted genes are located in conserved clusters and, although it is apparent that imprinting of adjacent genes is jointly regulated, multiple mechanisms among and within clusters may operate. Recent developments have also refined the timing of the gametic imprints and further defined the mechanism by which DNA methyltransferases confer allelic methylation patterns.

The product of the H19 gene may function as an RNA.

The mouse H19 gene was identified as an abundant hepatic fetal-specific mRNA under the transcriptional control of a trans-acting locus termed raf. The protein this gene encoded was not apparent from an analysis of its nucleotide sequence, since the mRNA contained multiple translation termination signals in all three reading frames. As a means of assessing which of the 35 small open reading frames might be important to the function of the gene, the human H19 gene was cloned and sequenced. Comparison of the two homologs revealed no conserved open reading frame. Cellular fractionation showed that H19 RNA is cytoplasmic but not associated with the translational machinery. Instead, it is located in a particle with a sedimentation coefficient of approximately 28S. Despite the fact that it is transcribed by RNA polymerase II and is spliced and polyadenylated, we suggest that the H19 RNA is not a classical mRNA. Instead, the product of this unusual gene may be an RNA molecule.

Molecular cloning and expression of the type 1 and type 2 murine receptors for tumor necrosis factor.

Clones encoding the type 1 (p80) and type 2 (p60) forms of the murine receptors for tumor necrosis factor (TNF) were isolated by cross-hybridization using probes derived from the cloned human TNF receptors. Each of the murine receptors shows strong sequence homology to the corresponding human receptor (approximately 65% amino acid identity) throughout the molecule but only modest homology, limited to ligand-binding domains, between themselves. The ligand-binding characteristics of the recombinant murine receptors mirror those of the human homologs: both receptor types bind TNF-alpha and -beta with multiple affinity classes, and the ligands cross-compete. Analysis of the murine transcripts encoding these receptors revealed the presence of RNAs for one or both forms of the receptors in all cells examined. It was also demonstrated that for both types of human TNF receptor, variably sized transcripts are observed in different cells. The murine cDNAs were further used to determine the chromosomal locations of the TNF receptor genes. They are not linked, in contrast to the ligands, and map to chromosomes 4 (type 1) and 6 (type 2).

The rhombotin gene family encode related LIM-domain proteins whose differing expression suggests multiple roles in mouse development.

The rhombotin (RBTN1 or Ttg-1) gene was first identified at a chromosome translocation in a T-cell acute leukaemia and later used to isolate two related genes (RBTN2 or Ttg-2 and RBTN3). Complete characterization of these genes in man and mouse shows that all three encode cysteine-rich proteins with typical LIM domains. RBTN1 and RBTN3-derived proteins have 98% identity in the LIM domains but are located on separate chromosomes in man and in mouse while RBTN1 and RBTN2, both located on human chromosome 11p but are on separate chromosomes in mouse, are only 48% identical in this part of the protein. The exon organization of RBTN1 and RBTN3 genes are similar, both having an intron, absent from the RBTN2 gene, in the LIM2-encoding region. The remarkable similarity between rbtn-1 and rbtn-3 proteins is parallelled in their expression patterns in mouse development, since both genes show high expression in restricted areas of the brain, but little lymphoid expression. rbtn-2 expression, however, is more ubiquitous. This gene shows a low level of thymus expression but high expression in fetal liver, adult spleen and B-cell lines, consistent with a role in B-cell development. These results suggest multiple cellular targets for the action of these proteins during development.

Neurofibromin expression and astrogliosis in neurofibromatosis (type 1) brains.

Patients with type 1 neurofibromatosis (NF1) have mutations in the gene encoding the protein neurofibromin. Immunocytochemistry on sections of cortex and cerebellum of unaffected and NF1 individuals and wild-type and NF1-deficient mice showed that the distribution of neurofibromin was similar to that reported for rat. However, dystrophic neurofibromin-expressing neurons were found in human but not rodent brain. Intensity of anti-neurofibromin reactivity was reduced in NF1-deficient mice but not in human brains. GFAP was upregulated in three NF1 brains studied by immunocytochemistry; a 4-18-fold increase in GFAP levels was documented by Western blot analysis in three brains. GFAP content/cell and the number of GFAP-immunoreactive astrocytes was increased in NF1 brains as compared to the controls. These results suggest that mutations in the NF1 gene do not grossly alter the pattern of neurofibromin expression, but activation of astrocytes may be common in NF1. Presence of degenerative debris in one of two brains using the cupric silver method suggests that degeneration is not always detectable in NF1 brains.

The Prader-Willi syndrome imprinting center activates the paternally expressed murine Ube3a antisense transcript but represses paternal Ube3a.

The imprinted UBE3A gene exhibits maternal-only expression in specific cell types in the brain, but exhibits biallelic expression in other cell types. UBE3A is located adjacent to a cluster of imprinted, paternally expressed genes that are known to be positively regulated by the Prader-Willi syndrome imprinting center (PWS-IC). Here, we examined the effect of the PWS-IC on the UBE3A locus. Using intersubspecific crosses, we found that deletion of the PWS-IC causes an upregulation of the paternal Ube3a allele. This indicates that unlike its positive effect on all the other paternally expressed transcripts in the region, the PWS-IC negatively regulates the levels of paternal UBE3A. Interestingly, we found that like the human UBE3A locus, the murine Ube3a locus includes an imprinted, paternally expressed antisense transcript. We show that this paternal antisense transcript is positively regulated by the PWS-IC. These results are consistent with a model in which the PWS-IC mediates activation and maintenance of paternal gene expression in the 15q11-q13 region, with repression of the paternal UBE3A gene occurring as an indirect result of expression of the antisense transcript.

The structure and expression of a novel gene activated in early mouse embryogenesis.

We report the cloning and sequence determination of the mouse H19 gene. This gene is under the genetic control of two trans-acting loci in the mouse, termed raf and Rif. These loci determine the adult basal and inducible levels, respectively, of H19 mRNA, as well as the mRNA for alpha-fetoprotein. By elucidating the sequence and structure of the H19 gene we show that it is unrelated to the alpha-fetoprotein gene, and therefore must have acquired its regulation by raf and Rif independently. The sequence also indicates that the H19 gene has a very unusual structure. It is composed of five exons, 1307, 135, 119, 127 and 560 bp in size, along with four very small introns whose combined lengths are 270 bases. The largest open reading frame of the gene, sufficient to encode a protein of approximately 14 kd, is contained entirely within the first large exon, 680 bases downstream of the cap site of the mRNA. Preceding the translation initiation codon are four ATG codons, each of which is followed shortly thereafter by translation terminator codons. The rest of the gene, which encompasses all five exons, is presumed to be untranslated. That the long 5' untranslated region may be used to regulate the translation of the mRNA is suggested from in vitro translation studies. Experiments which utilized tissue culture cell lines of the mesodermal lineage suggest that the gene is activated very early during muscle cell differentiation.

Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis.

Fas antigen is a cell-surface protein that mediates apoptosis. It is expressed in various tissues including the thymus and has structural homology with a number of cell-surface receptors, including tumour necrosis factor receptor and nerve growth factor receptor. Mice carrying the lymphoproliferation (lpr) mutation have defects in the Fas antigen gene. The lpr mice develop lymphadenopathy and suffer from a systemic lupus erythematosus-like autoimmune disease, indicating an important role for Fas antigen in the negative selection of autoreactive T cells in the thymus.

Analysis of murine Snrpn and human SNRPN gene imprinting in transgenic mice.

The SNRPN gene is known to be expressed exclusively from the paternal allele and to map to the critical region for the neurobehavioral disorder, Prader-Willi syndrome (PWS). As a means to investigate the mechanism of imprinting for the SNRPN gene, we have sought to recapitulate the imprinted expression of the endogenous gene. Using an 85-kb murine Snrpn clone, containing 33 kb of 5' and 30 kb of 3' flanking DNA, we obtained two intact transgenic lines. One line, containing two copies of the Snrpn transgene, recapitulated the imprinted expression pattern of the endogenous locus, whereas the other transgenic line, containing a single copy, was expressed upon both maternal and paternal inheritance. This suggests that a 6.6-kb region of maternal-specific DNA methylation that we have identified may be sufficient to confer imprinted expression, but not in a copy-number independent manner. Finally, we produced five lines of transgenic mice using a 76-kb human SNRPN clone containing 45 kb and 7 kb of 5' and 3' flanking DNA, respectively. We found all the lines were expressed upon both maternal and paternal inheritance, regardless of copy number, suggesting that the imprinting machinery in mouse and human may have diverged.

Expression and imprinting of MAGEL2 suggest a role in Prader-willi syndrome and the homologous murine imprinting phenotype.

Prader-Willi syndrome (PWS) is caused by the loss of expression of imprinted genes in chromosome 15q11-q13. Affected individuals exhibit neonatal hypotonia, developmental delay and childhood-onset obesity. Necdin, a protein implicated in the terminal differentiation of neurons, is the only PWS candidate gene to reduce viability when disrupted in a mouse model. In this study, we have characterized MAGEL2 (also known as NDNL1), a gene with 51% amino acid sequence similarity to necdin and located 41 kb distal to NDN in the PWS deletion region. MAGEL2 is expressed predominantly in brain, the primary tissue affected in PWS and in several fetal tissues as shown by northern blot analysis. MAGEL2 is imprinted with monoallelic expression in control brain, and paternal-only expression in the central nervous system as demonstrated by its lack of expression in brain from a PWS-affected individual. The orthologous mouse gene (Magel2) is located within 150 kb of NDN:, is imprinted with paternal-only expression and is expressed predominantly in late developmental stages and adult brain as shown by northern blotting, RT-PCR and whole-mount RNA in situ hybridization. Magel2 distribution partially overlaps that of NDN:, with strong expression being detected in the central nervous system in mid-gestation mouse embryos by in situ hybridization. We hypothesize that, although loss of necdin expression may be important in the neonatal presentation of PWS, loss of MAGEL2 may be critical to abnormalities in brain development and dysmorphic features in individuals with PWS.

Loss of neurofibromin results in neurotrophin-independent survival of embryonic sensory and sympathetic neurons.

Mutations at the neurofibromatosis 1 (NF1) locus in humans and mice result in abnormal growth of neural crest-derived cells, including melanocytes and Schwann cells. We have exploited a targeted disruption of the NF1 gene in mice to examine the role of neurofibromin in the acquisition of neurotrophin dependence in embryonic neurons. We show that both neural crest- and placode-derived sensory neurons isolated from NF1(-/-) embryos develop, extend neurites, and survive in the absence of neurotrophins, whereas their wild-type counterparts die rapidly unless nerve growth factor (NGF) or brain-derived neurotrophic factor (BDNF) is added to the culture medium. Moreover, NF1 (-/-) sympathetic neurons survive for extended periods and acquire mature morphology in the presence of NGF-blocking antibodies. Our results are consistent with a model wherein neurofibromin acts as a negative regulator of neurotrophin-mediated signaling for survival of embryonic peripheral neurons.