Mammary tumor modifiers in BALB/cJ mice heterozygous for p53

BALB/c mice are predisposed to developing spontaneous mammary tumors, which are further increased in a p53 heterozygous state. C57BL/6J mice are resistant to induced mammary tumors and develop less than 1% mammary tumors in both wild-type and p53 ^+/− states. To map modifiers of mammary tumorigenesis, we have established F₁ and F₂ crosses and backcrosses to BALB/cJ (N2-BALB/cJ) and C57BL/6J (N2-C57BL/6J) strains. All cohorts developed mammary carcinomas in p53 ^+/− females, suggesting that multiple loci dominantly and recessively contributed to mammary tumorigenesis. We mapped two modifiers of mammary tumorigenesis in the BALB/cJ strain. Mtsm1 (mammary tumor susceptibility modifier), a dominant-acting modifier, is located on chromosome 7. Mtsm1 is suggestive for linkage to mammary tumorigenesis (p = 0.001). We have analyzed the Mtsm1 region to locate candidate genes by comparing it to a rat modifier region, Mcs3, which shares syntenic conservation with Mtsm1. Expression data and SNPs were also taken into account. Five potential candidate genes within Mtsm1 are Aldh1a3, Chd2, Nipa2, Pcsk6, and Tubgcp5. The second modifier mapped is Mtsm2, a recessive-acting modifier. Mtsm2 is located on chromosome X and is significantly linked to mammary tumorigenesis (p = 1.03 × 10⁻⁷). Electronic supplementary material The online version of this article (doi: ) contains supplementary material, which is available to authorized users.     

Joint effects of germ-line <Emphasis Type="Italic">TP53</Emphasis> mutation, <Emphasis Type="Italic">MDM2</Emphasis> SNP309, and gender on cancer risk in family studies of Li–Fraumeni syndrome

Li–Fraumeni syndrome (LFS) is a rare familial cancer syndrome characterized by early cancer onset, diverse tumor types, and multiple primary tumors. Germ-line TP53 mutations have been identified in most LFS families. A high-frequency single-nucleotide polymorphism, SNP309 (rs2279744), in MDM2 was recently confirmed to be a modifier of cancer risk in several case-series studies: substantially earlier cancer onset was observed in SNP309 G-allele carriers than in wild-type individuals by 7–16 years. However, cancer risk analyses that jointly account for measured hereditary TP53 mutations and MDM2 SNP309 have not been systematically investigated in familial cases. Here, we determined the combined effects of measured TP53 mutations, MDM2 SNP309, and gender and their interactions simultaneously in LFS families. We used the method that is designed for extended pedigrees and structured for age-specific risk models based on Cox proportional hazards regression. We analyzed the cancer incidence in 19 extended pedigrees with germ-line TP53 mutations ascertained through the clinical LFS phenotype. The dataset consisted of 463 individuals with 129 TP53 mutation carriers. Our analyses showed that the TP53 germ-line mutation and its interaction with gender were strongly associated with familial cancer incidence and that the association between MDM2 SNP309 and increased cancer risk was modest. In contrast with several case-series studies, the interaction between MDM2 SNP309 and TP53 mutation was not statistically significant in our LFS family cohort. Our results showed that SNP309 G-alleles were associated with accelerated tumor formation in both carriers and non-carriers of germ-line TP53 mutations.     

The ROSA26 LacZ-neo R insertion confers resistance to mammary tumors in Apc Min/+ mice

B6.129S7-Gtrosa26 (ROSA26) mice carry a LacZ-neo ^R insertion on Chromosome (Chr) 6, made by promoter trapping with AB1 129 ES cells. Female C57BL/6J Apc ^Min /+ (B6 Min/+) mice are very susceptible to the induction of mammary tumors after treatment with ethylnitrosourea (ENU). However, ENU-treated B6 mice carrying both Apc ^Min and ROSA26 are resistant to mammary tumor formation. Thus, ROSA26 mice carry a modifier of Min-induced mammary tumor susceptibility. We have previously mapped the modifier to a 4-cM interval of 129-derived DNA that also contains the ROSA26 insertion. Here we report additional evidence for the effect of the ROSA26 insertion on mammary tumor formation. To test the hypothesis that the resistance was due to a linked modifier locus, we utilized two approaches. We have derived and tested two lines of mice that are congenic for 129-derived DNA within the minimal modifier interval and show that they are as susceptible to mammary tumors as are B6 mice. Additionally, we analyzed a backcross population segregating for the insertion and show that mice carrying the insertion are more resistant to mammary tumor development than are mice not carrying the insertion. Thus, the resistance is not due to a 129-derived modifier allele, but must be due to the ROSA26 insertion. In addition, the effect of the ROSA26 insertion can be detected in a backcross population segregating for other mammary modifiers. Received: 29 December 2000 / Accepted: 4 April 2001     

Mapping genetic modifiers of survival in a mouse model of Dravet syndrome

Epilepsy is a common neurological disorder affecting approximately 1% of the population. Mutations in voltage‐gated sodium channels are responsible for several monogenic epilepsy syndromes. More than 800 mutations in the voltage‐gated sodium channel SCN1A have been reported in patients with generalized epilepsy with febrile seizures plus and Dravet syndrome. Heterozygous loss‐of‐function mutations in SCN1A result in Dravet syndrome, a severe infant‐onset epileptic encephalopathy characterized by intractable seizures, developmental delays and increased mortality. A common feature of monogenic epilepsies is variable expressivity among individuals with the same mutation, suggesting that genetic modifiers may influence clinical severity. Mice with heterozygous deletion of Scn1a (Scn1a^+/?) model a number of Dravet syndrome features, including spontaneous seizures and premature lethality. Phenotype severity in Scn1a^+/? mice is strongly dependent on strain background. On the 129S6/SvEvTac strain Scn1a^+/? mice exhibit no overt phenotype, whereas on the (C57BL/6J × 129S6/SvEvTac)F1 strain Scn1a^+/? mice exhibit spontaneous seizures and early lethality. To systematically identify loci that influence premature lethality in Scn1a^+/? mice, we performed genome scans on reciprocal backcrosses. Quantitative trait locus mapping revealed modifier loci on mouse chromosomes 5, 7, 8 and 11. RNA‐seq analysis of strain‐dependent gene expression, regulation and coding sequence variation provided a list of potential functional candidate genes at each locus. Identification of modifier genes that influence survival in Scn1a^+/? mice will improve our understanding of the pathophysiology of Dravet syndrome and may suggest novel therapeutic strategies for improved treatment of human patients.     

Genetic modifiers affecting severity of epilepsy caused by mutation of sodium channelScn2a

Mutations in the voltage-gated sodium channels SCN1A and SCN2A are responsible for several types of human epilepsy. Variable expressivity among family members is a common feature of these inherited epilepsies, suggesting that genetic modifiers may influence the clinical manifestation of epilepsy. The transgenic mouse model Scn2a^Q54 has an epilepsy phenotype as a result of a mutation in Scn2a that slows channel inactivation. The mice display progressive epilepsy that begins with short-duration partial seizures that appear to originate in the hippocampus. The partial seizures become more frequent and of longer duration with age and often induce secondary generalized seizures. Clinical severity of the Scn2a^Q54 phenotype is influenced by genetic background. Congenic C57BL/6J.Q54 mice exhibit decreased incidence of spontaneous seizures, delayed seizure onset, and longer survival in comparison with [C57BL/6J × SJL/J]F₁.Q54 mice. This observation indicates that strain SJL/J carries dominant modifier alleles at one or more loci that determine the severity of the epilepsy phenotype. Genome-wide interval mapping in an N₂ backcross revealed two modifier loci on Chromosomes 11 and 19 that influence the clinical severity of of this sodium channel-induced epilepsy. Modifier genes affecting clinical severity in the Scn2a^Q54 mouse model may contribute to the variable expressivity seen in epilepsy patients with sodium channel mutations.     

Single‐cell analysis of p16INK4a and p21WAF1 expression suggests distinct mechanisms of senescence in normal human and Li‐Fraumeni Syndrome fibroblasts

Herein we used single‐cell observation methods to gain insight into the roles of p16^INK4A and p21^WAF1 (hereafter p16 and p21) in replicative senescence and ionizing radiation‐induced accelerated senescence in human [normal, ataxia telangiectasia (AT) and Li‐Fraumeni syndrome (LFS)] fibroblast strains. Cultures of all strains entered a state of replicative senescence at late passages, as evident from inhibition of growth, acquisition of flattened and enlarged cell morphology, and positive staining for senescence‐associated β‐galactosidase. In addition, proliferating early‐passage cultures of these strains exhibited accelerated senescence in response to ionizing radiation. Immunofluorescence microscopy revealed the heterogeneous expression of p16 in normal and AT fibroblast strains, with the majority of the cells exhibiting undetectable levels of p16 irrespective of in vitro culture age. Importantly, replicative senescence as well as accelerated senescence triggered by ionizing radiation were accompanied by sustained nuclear accumulation of p21, but did not correlate with p16 expression in p53‐proficient (normal and AT) fibroblasts. In p53‐deficient (LFS) fibroblasts, on the other hand, replicative senescence and ionizing radiation‐triggered accelerated senescence strongly correlated with expression of p16 but not of p21. Furthermore, senescence in LFS fibroblasts was associated with genomic instability encompassing polyploidy. Our findings are compatible with a model in which p16 serves as a backup regulator of senescence, triggering this response preferentially in the absence of wild‐type p53 activity. The possibility that one of the tumor‐suppressor functions of p16 may be associated with genomic instability, preventing the emergence of malignant progeny from polyploid giant cells, is also supported by these results. J. Cell. Physiol. 223: 57–67, 2010. © 2009 Wiley‐Liss, Inc.     

Scram1 is a modifier of spinal cord resistance for astrocytoma on mouse Chr 5

Tumor location can profoundly affect morbidity and patient prognosis, even for the same tumor type. Very little is known about whether tumor location is determined stochastically or whether genetic risk factors can affect where tumors arise within an organ system. We have taken advantage of the Nf1-/+;Trp53-/+cis mouse model of astrocytoma/glioblastoma to map genetic loci affecting whether astrocytomas are found in the spinal cord. We identify a locus on distal Chr 5, termed Scram1 for spinal cord resistance to astrocytoma modifier 1, with a LOD score of 5.0 and a genome-wide significance of P?相似文献   

Genetic basis of murine antibacterial defense to streptococcal lung infection

To evaluate the effect of genetic background on antibacterial defense to streptococcal infection, eight genetically diverse strains of mice (A/J, DBA/2J, CAST/Ei, FVB/NJ, BALB/cJ, C57BL/6J, 129/SvImJ, and C3H/HeJ) and tlr2-deficient mice (C57BL/6^tlr2−/−) were infected with three doses of Streptococcus zooepidemicus (500, 5,000, or 50,000 colony-forming units) by alveolar challenge. There was a range of susceptibility between the strains at each dose and time point (6, 24, and 96 h). At the lowest dose, the 129/SvImJ and C3H/HeJ strains had significantly higher bacterial counts at all time points after infection, when compared to A/J, DBA/2J, CAST/Ei, FVB/NJ, which were resistant to infection at the low dose of innoculum. At the medium dose, 129/SvImJ and C3H/HeJ had higher bacterial counts, while A/J, DBA/2J, and BALB/cJ showed reduced streptococcal growth. After the highest dose of Streptococcus, there were minimal differences between strains, suggesting the protective impact of modifier genes can be overcome. TLR2-deficient animals contained increased bacterial load with reduced cytokines after 96 h when compared to C57BL/6J controls suggesting a role of innate immunity in late antibacterial defense. Overall, we identify vulnerable (129/SvlmJ and C3H/HeJ) and resistant (A/J, FVB, and DBA) mouse strains to streptococcal lung infection, which demonstrate divergent genetic expression profiles. These results demonstrate that innate differences in pulmonary host defense to S. zooepidemicus are dependent on host genetic factors.     

Chr 19A/J modifies tumor resistance in a sex- and parent-of-origin-specific manner

Neurofibromatosis type 1 (NF1) is one of the most common human genetic diseases affecting the nervous system and predisposes individuals to cancer, including peripheral nerve sheath tumors (PNSTs) and astrocytomas. Modifiers in the genetic background affect the severity of the disease and we have previously mapped two modifier loci, Nstr1 and Nstr2, that influence resistance to PNSTs in the Nf1−/+;Trp53−/+cis mouse model of NF1. We report here the analysis of Nstr1 in isolation from other epistatic loci using a chromosome substitution strain, and further show that a modifier locus (or loci) on chromosome 19 influences resistance to both PNSTs and astrocytomas. This modifier locus interacts with sex, resulting in sex-specific modification of tumors. Allele variability on chromosome 19 affects both the timing and the penetrance of the growth of different tumor types associated with NF1, specifically PNSTs and astrocytoma. These results indicate that modifiers of cancer susceptibility interact and affect tumorigenesis under different genetic conditions and demonstrate the power of chromosome substitution strains to study genetic modifiers. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

ROSA26 mice carry a modifier of Min-induced mammary and intestinal tumor development

B6.129S7-Gtrosa26 (B6.R26) mice carry a LacZ-neoR insertion on Chromosome (Chr) 6, made by promoter trapping with 129 ES cells. Female C57BL/6J Apc ^Min /+ (B6Min/+) mice are highly susceptible to intestinal tumors and to the induction of mammary tumors after treatment with ethylnitrosourea (ENU). However, B6.R26/+Min/+ females develop fewer mammary and intestinal tumors after ENU treatment than do B6 Min/+ mice. B6.R26/+ mice from two independently derived congenic lines show this modifier effect. Each of these congenic lines carries approximately 20 cM of 129-derived DNA flanking the insertion, raising the possibility that the resistance is due to a linked modifier locus. To further map the modifier locus, we have generated several lines of mice carrying different regions of the congenic interval. We have found that resistance to mammary and intestinal tumors in ENU-treated Min/+ mice maps to a minimum 4-cM interval that includes the ROSA26 LacZ-neoR insertion. Therefore, the resistance to tumor development is due to either the ROSA26 insertion or a very tightly linked modifier locus. Received: 10 May 2000 / Accepted: 25 July 2000     

Effect of genetic background on the cardiac phenotype in a mouse model of Emery-Dreifuss muscular dystrophy

A-type lamins gene (LMNA) mutations cause an autosomal dominant inherited form of Emery-Dreifuss muscular dystrophy (EDMD). EDMD is characterized by slowly progressive muscle weakness and wasting and dilated cardiomyopathy, often leading to heart failure-related disability. EDMD is highly penetrant with poor prognosis and there is currently no specific therapy available. Clinical variability ranges from early onset with severe presentation in childhood to late onset with slow progression in adulthood. Genetic background is a well-known factor that significantly affects phenotype in several mouse models of human diseases. This phenotypic variability is attributed, at least in part, to genetic modifiers that regulate the disease process. To characterize the phenotype of A-type lamins mutation on different genetic background, we created and phenotyped C57BL/6JRj-Lmna^H222P/H222P mice (C57^{Lmna p.H222P}) and compared them with the 129S2/SvPasCrl-Lmna^H222P/H222P mice (129^{Lmna p.H222P}). These mouse strains were compared with their respective control strains at multiple time points between 3 and 10 months of age. Both contractile and electrical cardiac muscle functions, as well as survival were characterized. We found that 129^{Lmna p.H222P} mice showed significantly reduced body weight and reduced cardiac function earlier than in the C57^{Lmna p.H222P} mice. We also revealed that only 129^{Lmna p.H222P} mice developed heart arrhythmias. The 129^{Lmna p.H222P} model with an earlier onset and more pronounced cardiac phenotype may be more useful for evaluating therapies that target cardiac muscle function, and heart arrhythmias.     

Effects of FVB/NJ and C57Bl/6J strain backgrounds on mammary tumor phenotype in inducible nitric oxide synthase deficient mice

The ability to genetically manipulate mice has led to rapid progress in our understanding of the roles of different gene products in human disease. Transgenic mice have often been created in the FVB/NJ (FVB) strain due to its high fecundity, while gene-targeted mice have been developed in the 129/SvJ-C57Bl/6J strains due to the capacity of 129/SvJ embryonic stem cells to facilitate germline transmission. Gene-targeted mice are commonly backcrossed into the C57Bl/6J (B6) background for comparison with existing data. Genetic modifiers have been shown to modulate mammary tumor latency in mouse models of breast cancer and it is commonly known that the FVB strain is susceptible to mammary tumors while the B6 strain is more resistant. Since gene-targeted mice in the B6 background are frequently bred into the polyomavirus middle T (PyMT) mouse model of breast cancer in the FVB strain, we have sought to understand the impact of the different genetic backgrounds on the resulting phenotype. We bred mice deficient in the inducible nitric oxide synthase (iNOS) until they were congenic in the PyMT model in the FVB and B6 strains. Our results reveal that the large difference in mean tumor latencies in the two backgrounds of 53 and 92 days respectively affect the ability to discern smaller differences in latency due to the Nos2 genetic mutation. Furthermore, the longer latency in the B6 strain enables a more detailed analysis of tumor formation indicating that individual tumor development is not stoichastic, but is initiated in the #1 glands and proceeds in early and late phases. NO production affects tumors that develop early suggesting an association of iNOS-induced NO with a more aggressive tumor phenotype, consistent with human clinical data positively correlating iNOS expression with breast cancer progression. An examination of lung metastases, which are significantly reduced in PyMT/iNOS^−/− mice compared with PyMT/iNOS^+/+ mice only in the B6 background, is concordant with these findings. Our data suggest that PyMT in the B6 background provides a useful model for the study of inflammation-induced breast cancer.     

Fine mapping of an epilepsy modifier gene on mouse Chromosome 19

Mutations in voltage-gated sodium channels are associated with several types of human epilepsy. Variable expressivity and penetrance are common features of inherited epilepsy caused by sodium channel mutations, suggesting that genetic modifiers may influence clinical severity. The mouse model Scn2a ^Q54 has an epilepsy phenotype due to a mutation in Scn2a that results in elevated persistent sodium current. Phenotype severity in Scn2a ^Q54 mice is dependent on the genetic background. Congenic C57BL/6J.Q54 mice have delayed onset and low seizure frequency compared to (C57BL/6J × SJL/J)F1.Q54 mice. Previously, we identified two modifier loci that influence the Scn2a ^Q54 epilepsy phenotype: Moe1 (modifier of epilepsy 1) on Chromosome 11 and Moe2 on Chromosome 19. We have constructed interval-specific congenic strains to further refine the position of Moe2 on Chromosome 19 to a 5-Mb region. Sequencing and expression analyses of genes in the critical interval suggested two potential modifier candidates: (1) voltage-gated potassium channel subunit subfamily V, member 2 (Kcnv2), and (2) SWI/SNF-related, matrix-associated, actin-dependent regulator of chromatin, subfamily a, member 2 (Smarca2). Based on its biological role in regulating membrane excitability and the association between ion channel variants and seizures, Kcnv2 is a strong functional candidate for Moe2. Modifier genes affecting the epilepsy phenotype of Scn2a ^Q54 mice may contribute to variable expressivity and penetrance in human epilepsy patients with sodium channel mutations.     

Mice with reduced expression of the telomere‐associated protein Ft1 develop p53‐sensitive progeroid traits

Human AKTIP and mouse Ft1 are orthologous ubiquitin E2 variant proteins involved in telomere maintenance and DNA replication. AKTIP also interacts with A‐ and B‐type lamins. These features suggest that Ft1 may be implicated in aging regulatory pathways. Here, we show that cells derived from hypomorph Ft1 mutant (Ft1^kof/kof) mice exhibit telomeric defects and that Ft1^kof/kof animals develop progeroid traits, including impaired growth, skeletal and skin defects, abnormal heart tissue, and sterility. We also demonstrate a genetic interaction between Ft1 and p53. The analysis of mice carrying mutations in both Ft1 and p53 (Ft1^kof/kof; p53^ko/ko and Ft1^kof/kof; p53^+/ko) showed that reduction in p53 rescues the progeroid traits of Ft1 mutants, suggesting that they are at least in part caused by a p53‐dependent DNA damage response. Conversely, Ft1 reduction alters lymphomagenesis in p53 mutant mice. These results identify Ft1 as a new player in the aging process and open the way to the analysis of its interactions with other progeria genes using the mouse model.     

Gain-of-function mutation in Gnao1: A murine model of epileptiform encephalopathy (EIEE17)?

G protein-coupled receptors strongly modulate neuronal excitability but there has been little evidence for G protein mechanisms in genetic epilepsies. Recently, four patients with epileptic encephalopathy (EIEE17) were found to have mutations in GNAO1, the most abundant G protein in brain, but the mechanism of this effect is not known. The GNAO1 gene product, Gα_o, negatively regulates neurotransmitter release. Here, we report a dominant murine model of Gnao1-related seizures and sudden death. We introduced a genomic gain-of-function knock-in mutation (Gnao1 ^+/G184S) that prevents G_o turnoff by Regulators of G protein signaling proteins. This results in rare seizures, strain-dependent death between 15 and 40 weeks of age, and a markedly increased frequency of interictal epileptiform discharges. Mutants on a C57BL/6J background also have faster sensitization to pentylenetetrazol (PTZ) kindling. Both premature lethality and PTZ kindling effects are suppressed in the 129SvJ mouse strain. We have mapped a 129S-derived modifier locus on Chromosome 17 (within the region 41–70 MB) as a Modifer of G protein Seizures (Mogs1). Our mouse model suggests a novel gain-of-function mechanism for the newly defined subset of epileptic encephalopathy (EIEE17). Furthermore, it reveals a new epilepsy susceptibility modifier Mogs1 with implications for the complex genetics of human epilepsy as well as sudden death in epilepsy.     

Identification of Mom12 and Mom13, two novel modifier loci of ApcMin-mediated intestinal tumorigenesis

Colorectal cancer is a heterogeneous disease resulting from a combination of genetic and environmental factors. The C57BL/6J (B6) Apc^Min/+ mouse develops polyps throughout the gastrointestinal tract and has been a valuable model for understanding the genetic basis of intestinal tumorigenesis. Apc^Min/+ mice have been used to study known oncogenes and tumor suppressor genes on a controlled genetic background. These studies often utilize congenic knockout alleles, which can carry an unknown amount of residual donor DNA. The Apc^Min model has also been used to identify modifer loci, known as Modifier of Min (Mom) loci, which alter Apc^Min-mediated intestinal tumorigenesis. B6 mice carrying a knockout allele generated in WW6 embryonic stem cells were crossed to B6 Apc^Min/+ mice to determine the effect on polyp multiplicity. The newly generated colony developed significantly more intestinal polyps than Apc^Min/+ controls. Polyp multiplicity did not correlate with inheritance of the knockout allele, suggesting the presence of one or more modifier loci segregating in the colony. Genotyping of simple sequence length polymorphism (SSLP) markers revealed residual 129X1/SvJ genomic DNA within the congenic region of the parental knockout line. An analysis of polyp multiplicity data and SSLP genotyping indicated the presence of two Mom loci in the colony: 1) Mom12, a dominant modifier linked to the congenic region on chromosome 6, and 2) Mom13, which is unlinked to the congenic region and whose effect is masked by Mom12. The identification of Mom12 and Mom13 demonstrates the potential problems resulting from residual heterozygosity present in congenic lines.     

miR-34a is essential for p19Arf-driven cell cycle arrest

The Arf tumor suppressor gene product, p19^Arf, regulates cell proliferation in incipient cancer cells and during embryo development. Beyond its commonly accepted p53-dependent actions, p19^Arf also acts independently of p53 in both contexts. One such p53-independent effect with in vivo relevance includes its repression of Pdgfrβ, a process that is essential for vision in the mouse. We have utilized cell culture-based and mouse models to define a new role for miR-34a in this process. Ectopic expression of Arf in cultured cells enhanced the expression of several microRNAs predicted to target Pdgfrß synthesis, including the miR-34 family. Because miR-34a has been implicated as a p53-dependent effector, we investigated whether it also contributed to p53-independent effects of p19^Arf. Indeed, in mouse embryo fibroblasts (MEFs) lacking p53, Arf-driven repression of Pdgfrβ and its blockade of Pdgf-B stimulated DNA synthesis were both completely interrupted by anti-microRNA against miR-34a. Ectopic miR-34a directly targeted Pdgfrβ and a plasmid reporter containing wild-type Pdgfrβ 3′UTR sequence, but not one in which the miR-34a target sequence was mutated. Although miR-34a expression has been linked to p53—a well-known effector of p19^Arf—Arf expression and its knockdown correlated with miR-34a level in MEFs lacking p53. Finally, analysis of the mouse embryonic eye demonstrated that Arf controlled expression of miR-34a, and the related miR-34b and c, in vivo during normal mouse development. Our findings indicate that miR-34a provides an essential link between p19^Arf and its p53-independent capacity to block cell proliferation driven by Pdgfrβ. This has ramifications for developmental and tumor suppressor roles of Arf.