Mechanisms of Development 120 (2003) 549–560 www.elsevier.com/locate/modo

Genetic ablation of the tumor suppressor menin causes lethality at mid-gestation with defects in multiple organs Philippe Bertolinoa,b, Ivan Radovanovicc, Huguette Casseb, Adriano Aguzzic, Zhao-Qi Wanga,*, Chang-Xian Zhangb a

b

International Agency for Research on Cancer (IARC), 150 Cours Albert-Thomas, F-69008 Lyon, France Laboratoire Ge´ne´tique et Cancer, CNRS, UMR5641, Faculte´ de Medicine, Universite´ de Lyon 1, 8 Av. Rockefeller, F-69373 Lyon, France c Institute of Neuropathology, University Hospital Zurich, Schmelzbergstrasse 12, CH-8091 Zurich, Switzerland Received 23 September 2002; received in revised form 23 January 2003; accepted 20 February 2003

Abstract Patients suffering from multiple endocrine neoplasia type 1 (MEN1) are predisposed to multiple endocrine tumors. The MEN1 gene product, menin, is expressed in many embryonic, as well as adult tissues, and interacts with several proteins in vitro and in vivo. However, the biological function of menin remains largely unknown. Here we show that disruption of the Men1 gene in mice causes embryonic lethality at E11.5 – E13.5. The Men1 null mutant embryos appeared smaller in size, frequently with body haemorrhages and oedemas, and a substantial proportion of them showed disclosure of the neural tube. Histological analysis revealed an abnormal development of the nervous system and heart hypotrophy in some Men1 null embryos. Furthermore, Men1 null livers generally displayed an altered organization of the epithelial and hematopoietic compartments associated with enhanced apoptosis. Chimerism analysis of embryos generated by injection of Men1 null ES cells, showed that cells lacking menin do not seem to have a general cell-autonomous defect. However, primary Men1 null embryonic fibroblasts entered senescence earlier than their wild-type counterparts. Despite normal proliferation ability, Men1 null ES cells exhibited a deficiency to form embryoid bodies, suggesting an impaired differentiation capacity in these cells. The present study demonstrates that menin plays an important role in the embryonic development of multiple organs in addition to its proposed role in tumor suppression. q 2003 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Men1 Embryonic lethality; Fetal liver defects; Neurotube disclosure; Heart hypotrophy; Chimerism analysis

1. Introduction Multiple endocrine neoplasia type 1 (MEN1) is a hereditary syndrome characterized by the multiple occurrence of endocrine tumors of the parathyroids, pancreas, and anterior pituitary. The tumor deletion mapping based on the detection of loss of heterozygosity (LOH) in MEN1 related endocrine tumors (Bystrom et al., 1990; Larsson et al., 1988) contributed to the localization of the MEN1 locus (to chromosome 11 in humans), and also suggested that the gene responsible for the disease plays a role as a tumor suppressor. The MEN1 gene, identified by positional cloning (Chandrasekharappa et al., 1997; Lemmens et al., 1997), encodes a major transcript of about 3 kb, with an ORF composed of 610 codons. The protein encoded by the * Corresponding author. Tel.: þ 33-4-7273-8510; fax: þ33-4-7273-8329. E-mail address: [email protected] (Z.Q. Wang).

MEN1 gene, menin, principally located in the nucleus (Guru et al., 1998; Wautot et al., 2000), is well conserved from Drosophilae to human, indicating that it may possess a fundamental biological role. MEN1 expression has been found in many embryonic and adult tissues. Manickam et al. have shown that at the end of somitogenesis in zebrafish (24 h post fertilization, h.p.f), the Men1 gene is strongly transcribed in proerythroblasts and, to a lesser extent, in the neural tube and the ventral mesenchyme (Manickam et al., 2000). At 48 h.p.f., equivalent to the stage of 10.5 – 11.5 post coitum (dpc) in the mouse, the Men1 expression is no longer apparent in the blood, but can be detected in endoderm, including the liver premordium. In the mouse, the Men1 gene is generally transcribed at as early as 7 dpc, whereas at a later gestational stage, the expression is more readily found in certain tissues such as brain, thymus and liver (Guru et al., 1999; Stewart et al., 1998). We have previously analyzed the expression of

0925-4773/03/$ - see front matter q 2003 Elsevier Science Ireland Ltd. All rights reserved. doi:10.1016/S0925-4773(03)00039-X

550

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

menin in 20-week-old human fetal tissues and found that the protein can be detected in eight out of the 12 tissues tested, including the brain cortex, adrenal gland, heart, kidney, pituitary, testis, thymus and thyroid, with the most abundant expression in the brain cortex (Wautot et al., 2000). These data indicate that MEN1 is expressed far beyond the tissues involved in MEN1 pathology and suggest that it may have other unknown biological functions. Biochemical studies have identified several partners for menin, including JunD (Agarwal et al., 1999; Gobl et al., 1999), Smad3 (Kaji et al., 2001), nm23 (Ohkura et al., 2001), Pem (Lemmens et al., 2001), and NF-kB family components p50, p52 and p65 (Heppner et al., 2001). In vitro studies have shown that by direct interaction, menin inhibits JunD-activated transcription as well as the NF-kB pathway (Agarwal et al., 1999; Gobl et al., 1999; Heppner et al., 2001). On the other hand, a TGF-b intracellular signal transducer, Smad3, is also capable of interacting with menin (Kaji et al., 2001), suggesting a role for menin in mediating the signals from the TGF-b pathway. Despite biochemical links between menin and its partners, the in vivo biological importance of menin remains elusive. The mouse model generated for MEN1 has shown that heterozygous Men1 mutant mice develop multiple endocrine tumors (Crabtree et al., 2001) mimicking the human cases. However, the lethal phenotype of Men1 homozygous mutant mice has not been characterized. In order to study the biological function of the Men1 gene in mouse development, we generated a Men1 knock-out mouse model and investigated the cause of the embryonic lethality. Our data show that homozygous Men1 mutant embryos die between E11.5 and E13.5, with defects in the development of multiple organs, including the neural tube, heart and liver. Generation of Men1 null ES cells and embryonic fibroblasts (EFs) indicates that menin is dispensable for proliferation and survival; however, null mutation in the gene compromised the differentiation capacity of these ES cells and caused early senescence in EFs. Furthermore, analysis of chimeric fetuses generated by injection of Men1 null ES cells showed no cell-autonomous proliferation defects in the absence of menin. The present study demonstrates that the tumor suppressor menin is required for mouse embryonic development likely by its role in cell differentiation and/or senescence.

2. Results 2.1. Targeted disruption of the Men1 allele in ES cells and mice To disrupt the Men1 gene in mice, a conditional approach was used to allow the generation of various mutant alleles in embryonic stem (ES) cells (Fig. 1A). The targeting vector was designed to delete the third exon of the Men1 gene and contained a neomycin/thymidine-kinase (neoTK) selection

cassette flanked by two loxP sites that was inserted into intron 2 of the gene, and a third loxP in the intron 3. After electroporation of the targeting vector into E14.1 ES cells and G418 selection, the nine clones containing targeted (T) allele (Men1 þ/T) were obtained. Two of these clones were subjected to the transfection with vector expressing the Cre recombinase which deleted neoTK and exon 3 to generate ES clones containing the deleted (D) allele (Men1 þ/D). In order to generate null mutant ES cells, gene targeting in the remaining wild-type allele of Men1 þ/D ES cells was performed. Among 600 clones analyzed by Southern blotting, seven Men1 T/D clones were found, two of which were subsequently subjected to transfection with a Cre-expressing vector which gave rise to Men1 D/D ES clones. Western blot analysis of nuclear protein extracts prepared from Men1 T/D or Men1 D/D ES cells showed absence of menin expression in these ES cells (Fig. 1C), suggesting that the targeted allele is a null allele and deletion of exon 3 causes null mutation in the Men1 locus. The generation of viable Men1 null ES cells indicated that menin is not essential for cell survival and proliferation (see below). 2.2. Disruption of the Men1 gene results in lethality during mid-gestation To generate mutant Men1 mice, three Men1 þ/T and five Men1 þ/D ES clones were respectively injected into blastocysts to generate chimeric mice. Several chimeras transmitted these Men1 mutant alleles to their offspring. Heterozygous Men1 knock-out mice (Men1 þ/T and Men1 þ/D) of both sexes developed normally, and were fertile and healthy during the monitoring period of 7 months. To generate homozygous Men1 mutant mice, heterozygous Men1 þ/T mice in a mixed background (129/Sv £ C57BL/6) were intercrossed and pups were genotyped by Southern blot analysis. Among the 552 pups analyzed, no homozygous mutant offspring was found (Table 1), suggesting Table 1 Genotype of Men1 T/T embryos and micea Age

Number of offspring analysed

Genotype of live offspring (dead embryos) þ/þ

E11.5 E12.5 E13.5 E15.5 E17.5–18

132 224 227 83 38

4 weeks

552

a

28 65 (1) 72 (1) 20 14 192

þ /T

T/T

64 (1) 107 (3) 111 (2) 42 (2) 20

35 (4) 30 (18) 10 (31) 0 (19) 0 (4)

360

0

Genotypes analysis of offspring generated from Men1 heterozygote intercrosses. DNA was isolated from the yolk sac of the embryos between days 11.5 and 18 of gestation (E11.5–E18) or from the tails of 4 weeks old mice and analysed by Southern blotting. þ/þ, wild-type; þ/T, heterozygotes; T/T, homozygotes.

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

551

Fig. 1. Inactivation of the Men1 gene in ES cells and in mice. (A) Structure and restriction map of the wild-type murine Men1 allele (þ ), targeting vector, targeted allele (T) and deleted allele (D). Exons are indicated by a black box with numbers, Neo, TK and DTA correspond to the neomycin, thymidine kinase and diphethera toxin A-chain genes, respectively. The probe designed for the Southern blot analysis is shown by an open box. Arrows represent the position of primers (p1 and p2) used for RT-PCR analysis. Abbreviations for the restriction site are as follows, G: Bgl II; B: Bam HI. (B) Southern blot analysis of ES cell clones. The wild-type allele produced a 5.6 kb fragment, the targeted allele a 3.6 kb fragment and the deleted allele a 2.6 kb fragment after digestion of genomic DNA with Bam HI and Bgl II. The genotype of each ES clone is indicated on the top of the gel, size of each allele is marked. (C) Western blot analysis of menin expression in wild-type (þ /þ ) and mutant ES cells (þ/T, T/D and D/D). The protein loading was monitored by p300. RT-PCR (D) and Western blot (E) analyzes of menin expression in E12.5 embryos of wild-type (þ/þ), heterozygous (þ /T) and homozygous (T/T) genotypes. HPRT was used as an internal control for RT-PCR analysis and anti-b-tubulin antibody was used for protein loading control. The size of RT-PCR products is shown in parentheses (D).

that the disruption of the Men1 gene results in embryonic lethality. To define the stage at which the Men1 null embryos died, embryos ranging from E11.5 to E18 were collected and genotyped. While a normal Mendelian ratio of Men1 T/T embryos was observed at E11.5, Men1 T/T embryos started to die at E12.5 and E13.5 (Table 1) and no viable Men1 null embryos were found at the stage of E15.5 and E18 (Table 1). We conclude that the embryonic lethality of Men1 T/T mutant embryos occurs during the midgestation between E11.5 and E13.5. The timing at which homozygous mutant embryos died was reproducible in a pure 129/Sv background. Similar lethal phenotype was also seen in null embryos obtained from intercrossing heterozygous Men1 þ/D mice (data not shown), consistent with the

protein data showing a null mutation in Men1 D/D ES cells (Fig. 1C). RT-PCR and Western blot analyzes of Men1 T/T embryos showed that the insertion of a neoTK cassette into intron 2 of the Men1 locus (the T allele) disrupted the Men1 expression at RNA and protein levels (Figs. 1D,E), consistent with the previous study (Crabtree et al., 2001). 2.3. Men1 null mutant embryos have multiple defects in the organogenesis of the neural tube, heart and liver By macroscopic examination, Men1 T/T embryos at E11.5– 12.5 were generally found smaller in size compared to wild-type embryos (Figs. 2A,B), often accompanied by body haemorrhages (Figs. 2C,D) or oedemas. In particular,

552

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

Fig. 2. Macroscopic examination of Men1 T/T embryos. Upper panel shows the macroscopic view of, respectively, an E12.5 wild-type embryo (A); a normallooking E12.5 Men1 T/T embryo with smaller size (B); and an E12.5 Men1 T/T embryo exhibiting extensive haemorrhages, and exencephaly indicated by an arrow (C). Lower panel shows E11.5 Men1 T/T embryos presenting abdominal haemorrhages (D) and abnormal dorsal closure of the neural tube resulting in a ‘zig-zag’ pattern indicated by arrows (E).

nearly 30% of the null mutant embryos had an abnormal neural tube closure with a ‘zig-zag’ shape (Fig. 2E) and, though rarely, exencephaly (Fig. 2C). Histological analysis of Men1 T/T embryos macroscopically having a ‘zig-zag’ dorsal pattern showed an incomplete closure of the dorsal neural tube with a mild invagination of the overlying ectoderm (Fig. 3B) when compared with wild-type counterparts (Fig. 3A). In addition, in contrast to wildtype littermates (Fig. 3C), microscopic analysis of Men1 T/T embryos having an abnormal cephalic shape typically revealed severe exencephaly with opened and protruding neuro-ectodermal structures of the forebrain and midbrain (Fig. 3D). Although macroscopic examination revealed apparently normal circulation in E11.5 – 12.5 embryos (data not shown), histological analysis of Men1 T/T embryos showed a clear myocardial hypotrophy in the majority of embryos resulting in decreased thickness and density of the ventricular trabeculations (Fig. 3F), as well as a thinner

intraventricular septum and ventricular walls (Fig. 3H) in comparison with wild-type embryos (Figs. 3E,G). Another predominant phenotype from macroscopical analysis is that about 42% (37 out of 89 examined) of the null embryos exhibited severe growth retardation of the liver (Fig. 4A, upper panel). While Men1 null embryos were generally not anemic, about 17% of Men1 null livers were pale at E12.5. Microscopic analysis demonstrated that while some Men1 null embryos showed an apparently normal liver structure, the majority exhibited an abnormal organization of the epithelial and hematopoietic compartments, with a loosening of intercellular contacts and nuclear condensation of hematopoitic and epithelial cells, occasionally accompanied by the presence of cell debris (Figs. 4A,d,e). We further examined apoptosis in Men1 null livers at E11.5 by TUNEL staining and found a substantially high number of TUNEL-positive cells in Men1 null fatal livers compared with that in wild-type controls (Fig. 4B). These data indicate an increased apoptosis in fetal liver lacking menin. These

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

553

developmental defects were also observed in a pure 129/Sv background. 2.4. No general cell-autonomous proliferation defects in the absence of menin

Fig. 3. Histological analysis of neural tube closure and heart development in Men1 T/T embryos. (A, B) Axial sections of the dorsal neural tube in E11.5 Men1 þ/þ and Men1 T/T littermates. The Men1 deficient embryo does not show an appropriate closure of the neural tube neuroepithelium (arrows in B) compared with normal morphology in wild-type animals (arrow in A). nt: neural tube, e: ectoderm, do: dorsal root ganglia. Magnification: 100£. (C, D) Axial sections of midbrain structures in E11.5 Men1 þ/þ and Men1 T/T embryos. (C) Normal brain structures in E11.5 Men1 þ/þ embryo with a well regular midbrain neuroepithelium overlying the ventricular space. mb: midbrain, v: ventricule. (D) This Men1 T/T embryo shows a poorly organized neuroepithelium morphology with an aberrant bulging growth towards the outer sides of the cranium and irregular thickness. ne: neuroepithelium. Magnification: 40£ . (E) Sagital sections of E11.5 Men1 þ/þ embryos showing the normal morphology of the left ventricule (lv). (F) In Men1 T/T embryos the overall size of the heart appears normal, however a myocardial hypotrophy is present: the ventricular wall is thinner and the trabeculations density is lower than in the wild-type littermate (see E). li: liver. (G) Transverse section of the Men1 þ/þ heart showing the right (rv) and left (lv) ventricules separated by the interventricular septum (ivs), which is significantly thinner in the Men1 T/T littermate (H) compared to the wild-type counterpart (G). Magnification: 100£.

To further investigate whether the observed abnormal organogenesis is due to an autonomous defect in menindeficient cells, we generated chimeric embryos by injecting Men1 D/D ES cells into wild-type blastocysts using two independent clones. Tissues were collected from E12.5 and E18.5 chimeric embryos and the contributions of wild-type and Men1 D/D ES cell derivatives was quantified by PhosphoImage analysis after Southern blotting. At E12.5, a high contribution of Men1 null ES cell derivatives was found in all the tissues analyzed, including the fetal liver (Fig. 5A). However, all six E18 chimeric embryos showed nearly no Men1 D/D ES derivatives in the liver (Fig. 5B), significantly reduced contribution of Men1 D/D cells in the thymus, and to a lesser extent, in the cortex and yolk sac in some embryos, although a high percentage of Men1 D/D cells (30 –60%) could find in most of the organs. The high contrast in the contribution of Men1 D/D cells in the livers of E12.5 and E18.5 chimeric embryos suggests that Men1 D/D cells can form all cell types in early development and that differentiation and/or proliferation of some cell types, particularly in the liver, require menin after E12.5. On the other hand, we noticed that, despite the abnormal heart organogenesis in E11.5 Men1 null embryos, Men1 D/D ES cells contributed efficiently to heart formation in E12.5 or E18.5 embryos. Although Men1 D/D embryos showed neural tube closure defects, differentiation of Men1 D/D ES cells into neuronal lineage was possible albeit at a low efficiency. However, the resolution of chimerism analysis could not rule out that menin was crucial for some specific cell types in the heart and neural tube. Taken together, these data suggest that menin may be important for the differentiation of fetal liver cells, but dispensable for the development of most tissues or organs including the heart and neural tissues, which however can be regulated by menin. 2.5. Proliferation of Men1 null mutant ES and EF cells is not affected To determine whether Men1 null mutation would affect cell proliferation, Men1 þ/þ , Men1 þ/D, and Men1 D/D ES cells were examined for their proliferation capacity. As shown in Fig. 6A, there was no difference in the growth curve of the three genotypes. We further investigated the impact of Men1 mutation in other cell types, namely primary EF cells. As with the ES cells, no significant difference of proliferation was observed among Men1 þ/þ , Men1 þ/T, and Men1 T/T EF cells during the period of examination (Fig. 6A). These data indicate that menin is dispensable for normal proliferation of cells at least in these two cell types. However, we found that Men1 null EF cells

554

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

Fig. 4. Fetal liver defects in Men1 T/T embryos. (A) Macroscopic view of wild-type (a); and Men1 T/T (b) livers showing reduced size of Men1 T/T livers. (c– e) Phase contrast view of liver histology showing a similar liver architecture with normal organization of the epithelial compartment in Men1 þ/þ (c). However Men1 T/T livers (d, e) show a variable degree of decreased cellularity compared with wild-type littermates. Magnification: 200X. (B) Increased apoptosis in E11.5 Men1 T/T livers. Men1 T/T livers (right panel) show a highly increased number of TUNEL positive cells (green), compared with that in wild-type livers. Nuclei were counter-stained in red by propidium iodide. Magnification: right, 400£ , and left, 200£.

entered the senescence stage from two to three passages earlier than their wild-type counterparts in serial passaging experiments (Fig. 6B). This phenomenon was observed in more than four independent Men1 null EF lines. 2.6. In vitro differentiation is altered in Men1D/D ES cells To further study the differentiation capacity of Men1 D/D ES cells, we performed an embryoid body (EB) formation assay, which recapitulates some features of early embryogenesis in vitro. Morphological analysis showed that EBs formed by the Men1 null ES cells were much smaller in size

compared with those formed by wild-type ES cells, and lacked hemoglobinization (Fig. 6C). Furthermore, the number of EBs derived from Men1 D/D ES cells (E6 and G8) was greatly reduced when compared with wild-type ES cells (G7 and H6) (Fig. 6D). These data suggest that ES cell differentiation is impaired in the absence of menin. To further test this hypothesis, we tried to rescue the phenotype by introducing menin cDNA into Men1 null ES cells. Appropriate expression levels of exogenous menin in Men1 null ES cells were confirmed by Western blot analysis (Fig. 6E). Interestingly, genetically complemented Men1 mutant ES cells exhibited a normal capacity to form EBs

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

555

Fig. 5. Chimerism analysis of Men1 D/D ES cell derivatives in tissues of chimeric embryos. Contribution of Men1 D/D cells to various tissues was analyzed in five E12.5 and six E18.5 chimeric embryos. Individual chimeric fetus is represented by different colors. Tissue abbreviations are as follows, Bo: bone, Br: brain, Bs: brain stem, Ce: cerebellum, Co: cortex, He: heart, Int: intestine, Ki: kidney, Li: liver, Lu: lung, Sk: skin, Sp/Pa: spleen and pancreas, St: stomach, Th: thymus, YS: yolk sac.

and hemotoglobinization (Figs. 6C,D) compared to that of wild-type counterparts, indicating an essential role for menin in this process.

3. Discussion Generation of mice with a null-mutation in the Men1 gene revealed a crucial role for its product, menin, in embryonic development. We show that embryos lacking menin die between E11.5 and E13.5 with defects in multiple organs, including the nervous system, heart and liver. The timing of Men1 null embryos’ death, and the growth retardation phenotype are similar to the previous report by Crabtree et al. (2001). However, the embryonic tissues affected that could contribute to the lethal phenotype in these animals were not characterized in the previous report. In the present study, we found that Men1 null embryos are associated with neural tube disclosure, and occasionally with exencephaly, reminiscent of what was observed in mutant mice displaying a failure in the elevation of the neural folds and in roofplate induction/formation (see review by Lee and Jessell, 1999). Although no abnormality was noted in major heart structures, the obvious developmental cardiac defects, such as a thin myocardium wall as well as septal and trabeculation abnormalities, were found in Men1 null mutant embryos, suggesting that menin plays a role in the differentiation of some cardiac cell types. However, we cannot rule out that heart hypotrophy may be

secondary to other metabolic problems, such as hypoxia or ischimia. Histopathological analysis and TUNEL staining of Men1 null livers showed that the altered organization of hematopoietic and epithelial compartments was associated with enhanced apoptosis, suggesting an important role for menin in fetal liver function. Although we were unable to define the molecular basis of the lethal phenotype, the normal development of these organs are vital to embryogenesis. Among the defects observed in Men1 null embryos, while neural tube disclosure and heart hypotrophy can cause embryonic lethality (Sabapathy et al., 1999; Crispino et al., 2001), the severe defects in fetal liver function, which may result not only in the dysfunction of blood circulation but also in internal bleeding (Eferl et al., 1999; Hilberg et al., 1993; Pandolfi et al., 1995), also contribute to embryonic lethality. Phenotype specificity observed in Men1 null mutant embryos correlates with the spectrum of menin expression including these organs during development. In human fetal tissues, we have shown that menin expression can be readily detected in the brain cortex, adrenals, heart, kidney, pituitary, testis, thymus and thyroid, with the most important expression found in brain cortex (Wautot et al., 2000). In the mouse, the men1 gene is expressed in the entire embryo at E7, and at E13.5, the expression becomes restricted to certain tissues, including the forelimbs, gut, brain, liver and lung (Stewart et al., 1998; Guru et al., 1999). In zebrafish at the equivalent stage, the men1 gene is expressed mainly in the eyes, brain and limb buds as well as liver premordium (Manickam et al., 2000). Therefore, it is very possible that

556

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

menin is of vital importance to the organogenesis of these tissues. Indeed, the present study has identified menin as a crucial factor for the development of multiple organs, in addition to its proposed role in the development of endocrine tumors in MEN1 patients. Another striking feature in Men1 null embryos is the severe growth retardation of the liver and heart. Moreover, a large proportion of Men1 null embryos is generally smaller in size. Although we cannot rule out the possibility that proliferation of specific cell types in these organs may be affected, this growth retardation may not link to a proliferation defect because Men1 null ES cells and primary EFs proliferate normally in vitro. Moreover, generation of E12.5 chimeras via blastocyst injection of Men1 D/D ES cells did not reveal a compromised proliferation capacity of mutant cells in vivo. Thus menin may not be essential for general cell proliferation. However, it is interesting to note that Men1 null EFs exhibit an early senescence phenotype, arguing for the limited lifespan of specific cell types lacking menin. This phenotype may suggest a link between menin and telomere function and the loss of menin may affect telomere-mediated cellular function, resulting in earlier senescence. In this regard, the MEN1 product has been localized in telomeres (Suphapeetiporn et al., 2002). In addition, a recent study has shown that menin interacts with a subunit of replication protein A (RPA) (Sukhodolets et al., 2003), suggesting the possible involvement of menin in replication or recombination. It is possible that the absence of menin may stall cell replication (senescence) due to its fundamental function in DNA replication, cell cycle progression or cell death. On the other hand, complete loss of menin, e.g. via LOH, in adult tissues or specific cell types, such as endocrine cells, would favor cells having accumulated genetic alterations or pre-malignant events that overcome cell proliferation block, to acquire a proliferation advantage leading to tumorigenesis. This hypothesis may explain endocrine tumors that arise in MEN1 patients and also in Men1 heterozygous knockout mice (Crabtree et al., 2001; our unpublished observations).

557

In addition, we found that the formation of embryoid bodies (EB) from Men1 null ES cells was greatly reduced compared to wild-type ES cells, suggesting that menin may be required for the differentiation of specific cell types. In this regard, we found that although hepatoblasts of Men1 null fetal livers grow normally in cultures and Men1 null cells participated in the formation of chimeric fetal livers at E12.5, a much smaller contribution of Men1 null cells was detected in E18.5 chimeric livers, suggesting a differentiation defect during the period of E12.5 and E18.5. Since differentiation of hepatoblasts into hepatocytes and bile duct epithelial cells occurs around E14.5 (Desmet et al., 1998), the difference in the contribution of Men1 null ES cell derivatives in fetal livers between E12.5 and E18.5 may be explained, at least in part, by the temporal requirement of menin for the differentiation of hepatoblasts during this period. Perhaps consistent with this observation, we found that the fetal liver failure of Men1 null embryos was closely associated with an increased apoptosis. These data suggest that a lack of menin may induce apoptosis in a cell typespecific manner during organogenesis, leading to defects in the differentiation of the cells involved. However, how menin regulates apoptosis requires further investigation. Biochemical studies have shown that menin interacts with several proteins including JunD (Agarwal et al., 1999), Pem (Lemmens et al., 2001), Smad3 (Kaji et al., 2001) and the major components of NF-kB (p50, p52 and p65) (Heppner et al., 2001). However, the biological relevance of these interactions remains largely unknown. Homozygous deletion of JunD results in viable and apparently normal embryos with a mild phenotype in adulthood (Thepot et al., 2000). Interestingly, the fact that Men1 null embryos die around E11 – 13.5 and exhibit several developmental defects, including liver and heart failure, is reminiscent of the phenotype of knock-out mice lacking c-Jun (Eferl et al., 1999; Hilberg et al., 1993). However, Western blot analysis of different members of the AP-1 family showed no difference in expression of c-Jun, JunD and JunB proteins between Men1 null mutants and wild-type embryos (data

Fig. 6. Men1 mutant ES and EF cells exhibit normal growth in cell culture but show defects in embryoid body (EB) formation. (A) The growth curves of Men1 null mutant ES (Men1 D/D) and EFs cells (Men1 T/T) show no difference compared with their heterozygous and wild-type counterparts in 8-day cultures. Cells were plated at 1 £ 105 (ES cells) or 0.5 £ 105 (EFs) per 35 mm dishes and the cell number was counted every 2 days. The number of ES cells (left panel) corresponds to the mean of duplicate values ^ standard variation of at least two cell lines of each genotype. The proliferation of EFs (right panel) has been examined in duplicates using the EFs of three independent littermates. A representative result from one of them with two independent Men1 null cell lines (1 and 2) is shown. These proliferation experiments were repeated twice. (B) Serial passaging of EFs derived from littermates of E12.5 wild-type (Men1 þ/þ ) or homozygous (Men1 T/T) embryos at an interval of 3 days at a density of 5 £ 105 cells/well. EFs at passage 4 were used to start the experiment as passage ‘0’. Each plot represents a mean of two EF lines originating from two embryos, with cell numbers on the Y axis against passage numbers in culture on the X axis. Each point of analysis is the mean of duplicates of each cell line. It is noted that after three passages, Men1 null cells entered senescence, whereas wild-type cells entered senescence after passage 5. (C) Phase contrast microscopy shows that EBs formed from Men1 D/D ES cells were smaller and non-haemoglubinized (middle panel) compared with the ones derived from wild-type counterparts (left panel). This differentiation defect was rescued by ectopically expressed menin in Men1 D/D ES cells (right panel). Magnification: 10 £ . (D) Summary of EB numbers from all three genotypes of ES cells as indicated after 10 days of in vitro differentiation in a semi-solid medium (see Section 4). The experiment was repeated twice each time using two independent cell lines of each genotype. (E) Western blot analysis of ectopic expression of the Men1 gene in Men1 null ES cells. Note comparable menin expression in rescued Men1 D/D ES cells (marked as pCI-Men1 þ ) compared with wild-type cells. b-tubulin was used as a loading control. G7 and H6 are wild-type ES cells, whereas E6 and G8 are Men1 D/D ES cells.

558

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

not shown), suggesting that AP-1 factors may not be responsible for the embryonic phenotype in Men1 mutant embryos. Moreover, genetic ablation of other known partners for menin, such as Pem (Pitman et al., 1998), Smad3 (Zhu et al., 1998; Yang et al., 1999) and the major components of NF-kB, p50, p52 and p65 (Franzoso et al., 1997; Ishikawa et al., 1997; Beg et al., 1995), has not generated phenotypes similar to mice lacking menin. Thus, it seems that these interacting factors may not be biologically relevant for the function of menin in embryonic development and that other potential partners of menin are involved in the cellular function of these organs or cell types. However, we cannot exclude at this stage the possibility that the developmental defects seen in Men1 mutant embryos might also be secondary to the dysregulation of the above menin partners. In fact, chimerism analysis of embryos generated by injection of Men1 null ES cells showed that menin does not seem to have a cell-autonomous defect in many tissues. Because of the early death of Men1 null mice, the function of menin in adult life has not been delineated. A recent study has demonstrated that heterozygous Men1 mutant mice develop various tumors after 8 months, including in endocrine tissues (Crabtree et al., 2001), mimicking the tumor phenotype of MEN1 patients and confirming its tumor suppressor function. Conditional knock-out mice that are viable and have the Men1 gene inactivated in specific tissues or at specific stages, would allow further assessment of menin and its interacting partners in various biological processes, which, ultimately, will advance the understanding of the mechanism underlying tumorigenesis related to MEN1 disease.

allele produced a 5.6 kb fragment, the targeted allele a 3.6 kb fragment and the deleted allele a 2.6 kb fragment after digestion of genomic DNA with Bam HI and Bgl II. Men1 þ/T and Men1 þ/D ES clones were injected into blastocysts of C57BL/6 background and transferred into pseudopregnant females. The resulting chimeras were bred with C57BL/6 or 129/Sv mice. Germline transmission was screened by Southern blot analysis.

4. Experimental procedures

4.3. Histological analysis and immunohistochemistry

4.1. Gene targeting of Men1 in ES cells and mice

Embryos were collected at different embryonic stages between E11.5 and E13.5, and fixed in 4% buffered formalin for at least 48 h and embedded in paraffin. Sections (7 mm) were stained with hematoxylin and eosin (H&E) for histological examination. Apoptosis was examined on tissue sections by TUNEL staining using the ‘in situ cell death detection’ kit (Roche, Mannheim, Germany) and propidium iodide (Sigma, St Louis, USA) for nuclear counterstaining. Sections were analyzed with confocal laser scanning microscopy (Leica-TCS-SP2).

To construct the gene targeting vector, a mouse genomic DNA corresponding to the Men1 gene was isolated from an 129/Sv genomic library using a mouse Men1 cDNA probe. A 5.4 kb fragment from the Pvu II restriction site at the 50 non-coding region to the Xho I site in exon 10 was cloned into pBluescript-SK. A cassette expressing the neomycin resistant gene (Neo) and thymidine kinase (TK) flanked by two loxP sites was placed in intron 2 and another loxP site was inserted into intron 3. For negative selection, the diphethera toxin A-chain gene (DTA) was cloned into the vector. The targeting vector was electroporated into E14.1 ES cells and targeted ES clones were identified by Southern blotting. To generate different mutant alleles of Men1, the targeted ES clones (Men1 þ/T) were transfected with a Creexpressing plasmid (pMC-Cre) to obtain the deleted allele (D allele) with deletion of both the NeoTK cassette and exon 3. Men1 þ/T and Men1 þ/D ES clones were genotyped by Southern blotting using an upstream probe. The wild-type

4.2. Generation of Men1 null mutant ES cells and chimerism analysis To generate null ES cells, two Men1 þ/D ES clones were electroporated with the targeting vector and targeted disruption to the remaining wild-type allele was identified using the strategy described above. Men1 T/D clones were subjected to transfection with a Cre-expressing vector to delete the NeoTK cassette and exon 3 to generate Men1 D/D ES clones. To establish stable Men1 D/D ES cell lines ectopically expressing menin, 2 Men1 D/D ES cell lines (E6 and G8) were electroporated with the pCI-Men1 vector (Wautot et al., 2000), selected for neo resistance and subcloned. Menin expression was checked by Western blot analysis. To generate chimeric embryos, two clones of Men1 D/D ES cells were injected into blastocysts of C57BL/6 mice. Tissues of chimeric embryos were respectively dissected at E12.5 and E18.5 for DNA isolation. The contribution of Men1 D/D ES cell derivatives to the tissues of chimeric fetuses was determined by Southern blot analysis after Bam HI and Bgl II digestion followed by PhosphorImager scanning and quantification using the ImageQuant Software. The quantification was performed twice from samples obtained with two independent Men1 D/D ES clones.

4.4. RT-PCR and Western blot analyzes For RT-PCR analysis, total RNA was isolated from E12.5 wild-type and Men1 T/T embryos, as well as from control cell lines (wild-type EFs) with single step RNA extraction system (TRI-REAGENT, Sigma, l’Ile d’Abeau, France). cDNA was synthetized with Superscript II reverse transcriptase (Invitrogen, Cergy Pontoise, France) and the amount of cDNAs were adjusted by dilution to produce

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

equal amounts of hypoxanthine guanine phosphoribosyl transferase (HPRT) amplicon. PCR was performed by inclusion of primers of HPRT, as an internal control, and of Men1 in the same reaction mixture. HPRT primers: 50 -GCTGGTGAAAAGGACCTCT-30 and 50 -CACAGGACTAGAACACCTGC-30 . Men1 primers, p1: 50 -CGGATGTCATATGGAACAGC-30 and p2: 50 -CAGATGTCCGAGGTCATACA-30 , located in exons 2 and 4 of the Men1 gene, respectively. To prepare enriched nuclear protein fractions from embryonic tissues and cells, samples were placed in five volumes of hypotonic buffer containing 0.2 M sucrose, 3 mM CaCl2, 2 mM magnesium acetate, 0.1 mM EDTA, 10 mM Tris –HCl (pH 7.5), and disaggregated by passing them through 23- and 26-gauge needles. One volume of lysis buffer (the above hypotonic buffer supplemented with 1% Nonidet P-40) was added. After being centrifuged, the pellets were resuspended in a solution containing 20 mM Hepes (pH 8.0), 1.5 mM MgCl2, 450 mM KCl, and 25% glycerol, and incubated for 30 min at 48C with agitation. After extraction and protein quantification, protein were heated at 958C for 5 min, and subjected to Western blotting as previously described (Wautot et al., 2000). The primary antibodies used were raised against menin (C-19, a polyclonal antibody raised against the C-terminus of the protein; 1:7500), b-tubulin (1:5000) and p300 (N-15, 1:2000). All antibodies mentioned above were from Santa Cruz Biotechnology, Inc., (California, USA), except b-tubulin (Roche, Mannheim, Germany).

559

4.6. Embryoid body formation of ES cells For embryoid (EB) body formation, ES cells were passed and kept without feeder cells on gelatin-coated dishes 24 h before experimentation. Cells were then counted and 2500 cells were seeded (day 0) in a semi-solid medium onto 35 mm non-adherent petri dishes. Semi-solid medium is based on Iscove-modified DMEM and 0.9% methylcellulose (ES-Cult M3120, StemCell Technologies Inc., Vancouver Canada) complemented with 15% fetal calf serum, 2 mM L -glutamine, 150 mM MTG (Sigma) and 40 ng/ml murine stem cell factor (R&D systems, Abington, UK). After 10 days EBs were counted and examined for their morphology under a phase contrast microscope. The experiment was performed in duplicate and repeated twice.

Acknowledgements We are grateful to D. Galendo and N. Lyandrat for technical assistance in the maintenance of the mouse colonies and histology preparation, and to V. Wautot for her participation in the beginning of this study. We are also grateful to M. Billaud, A. Calender and R. Eferl for critical reading of the manuscript and discussion. PB is the recipient of a fellowship from the French Ministry of Education, and this study is supported by the Association for International Cancer Research, UK, the Association pour la Recherche Contre le Cancer, France, the Programme Emmergence from the Re´gion Rhoˆne-Alpes and La Ligue Contre le Cancer du Rhoˆne.

4.5. Proliferation and serial passaging of ES and embryonic fibroblasts References ES cells were maintained with a low passage number on a murine fibroblastic feeder cell layer in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 15% fetal calf serum, 2 mM L -glutamine, 1 mM sodium pyruvate, 0.1 mM non-essential amino acids, 100 U/ml penicillin, 100 mg/ml streptomycin, 0.1 mM mercaptoethanol and 10 ng/ml murine leukemia inhibiting factor (LIF). Embryonic fibroblasts were isolated from E12.5 embryos derived from intercrosses of Men1 þ/T or Men1 þ/D mice according to protocols published previously (Wang et al., 1995). For the proliferation assay, cells at the concentration of either 0.5 £ 105 (for EFs) or 1 £ 105 (for ES cells) were plated onto 35 mm dishes as duplicates. EFs at passage 2 or 3, originated from littermates of wild-type, heterozygous and homozygous mutant embryos, were used. At the indicated time period, cell numbers of each genotype were counted using a Coulter ZI cell counter. For immortalization, primary EFs at passage 4, originating from littermates of E12.5 wild-type (Men1 þ/þ ) or homozygous (Men1 T/T) embryos, were used to start the serial passaging (designated as passage 0), according to a 3T3 protocol.

Agarwal, S.K., Guru, S.C., Heppner, C., Erdos, M.R., Collins, R.M., Park, S.Y., Saggar, S., Chandrasekharappa, S.C., Collins, F.S., Spiegel, A.M., Marx, S.J., Burns, A.L., 1999. Menin interacts with the AP1 transcription factor JunD and represses JunD-activated transcription. Cell 96, 143–152. Beg, A.A., Sha, W.C., Bronson, R.T., Ghosh, S., Baltimore, D., 1995. Embryonic lethality and liver degeneration in mice lacking the RelA component of NF-kappa B. Nature 376, 167– 170. Bystrom, C., Larsson, C., Blomberg, C., Sandelin, K., Falkmer, U., Skogseid, B., Oberg, K., Werner, S., Nordenskjold, M., 1990. Localization of the MEN1 gene to a small region within chromosome 11q13 by deletion mapping in tumors. Proc. Natl. Acad. Sci. USA 87, 1968–1972. Chandrasekharappa, S.C., Guru, S.C., Manickam, P., Olufemi, S.E., Collins, F.S., Emmert-Buck, M.R., Debelenko, L.V., Zhuang, Z., Lubensky, I.A., Liotta, L.A., Crabtree, J.S., Wang, Y., Roe, B.A., Weisemann, J., Boguski, M.S., Agarwal, S.K., Kester, M.B., Kim, Y.S., Heppner, C., Dong, Q., Spiegel, A.M., Burns, A.L., Marx, S.J., 1997. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 276, 404–407. Crabtree, J.S., Scacheri, P.C., Ward, J.M., Garrett-Beal, L., Emmert-Buck, M.R., Edgemon, K.A., Lorang, D., Libutti, S.K., Chandrasekharappa, S.C., Marx, S.J., Spiegel, A.M., Collins, F.S., 2001. A mouse model of multiple endocrine neoplasia, type 1, develops multiple endocrine tumors. Proc. Natl. Acad. Sci. USA 98, 1118– 1123.

560

P. Bertolino et al. / Mechanisms of Development 120 (2003) 549–560

Crispino, J.D., Lodish, M.B., Thurberg, B.L., Litovsky, S.H., Collins, T., Molkentin, J.D., Orkin, S.H., 2001. Proper coronary vascular development and heart morphogenesis depend on interaction of GATA-4 with FOG cofactors. Genes Dev. 15, 839–844. Desmet, V.J., van Eyken, P., Roskams, T., 1998. Histopathology of vanishing bile duct diseases. Adv. Clin. Pathol. 2, 87–99. Eferl, R., Sibilia, M., Hilberg, F., Fuchsbichler, A., Kufferath, I., Guertl, B., Zenz, R., Wagner, E.F., Zatloukal, K., 1999. Functions of c-Jun in liver and heart development. J. Cell Biol. 145, 1049–1061. Franzoso, G., Carlson, L., Xing, L., Poljak, L., Shores, E.W., Brown, K.D., Leonardi, A., Tran, T., Boyce, B.F., Siebenlist, U., 1997. Requirement for NF-kappaB in osteoclast and B-cell development. Genes Dev. 11, 3482–3496. Gobl, A.E., Berg, M., Lopez-Egido, J.R., Oberg, K., Skogseid, B., Westin, G., 1999. Menin represses JunD-activated transcription by a histone deacetylase-dependent mechanism. Biochim. Biophys. Acta 1447, 51–56. Guru, S.C., Goldsmith, P.K., Burns, A.L., Marx, S.J., Spiegel, A.M., Collins, F.S., Chandrasekharappa, S.C., 1998. Menin, the product of the MEN1 gene, is a nuclear protein. Proc. Natl. Acad. Sci. USA 95, 1630–1634. Guru, S.C., Crabtree, J.S., Brown, K.D., Dunn, K.J., Manickam, P., Prasad, N.B., Wangsa, D., Burns, A.L., Spiegel, A.M., Marx, S.J., Pavan, W.J., Collins, F.S., Chandrasekharappa, S.C., 1999. Isolation, genomic organization, and expression analysis of Men1, the murine homolog of the MEN1 gene. Mamm. Genome 10, 592–596. Heppner, C., Bilimoria, K.Y., Agarwal, S.K., Kester, M., Whitty, L.J., Guru, S.C., Chandrasekharappa, S.C., Collins, F.S., Spiegel, A.M., Marx, S.J., Burns, A.L., 2001. The tumor suppressor protein menin interacts with NF-kappaB proteins and inhibits NF-kappaB-mediated transactivation. Oncogene 20, 4917–4925. Hilberg, F., Aguzzi, A., Howells, N., Wagner, E.F., 1993. c-jun is essential for normal mouse development and hepatogenesis. Nature 365, 179– 181. Ishikawa, H., Carrasco, D., Claudio, E., Ryseck, R.P., Bravo, R., 1997. Gastric hyperplasia and increased proliferative responses of lymphocytes in mice lacking the COOH-terminal ankyrin domain of NF-kappaB2. J. Exp. Med. 186, 999– 1014. Kaji, H., Canaff, L., Lebrun, J.J., Goltzman, D., Hendy, G.N., 2001. Inactivation of menin, a Smad3-interacting protein, blocks transforming growth factor type beta signaling. Proc. Natl. Acad. Sci. USA 98, 3837–3842. Larsson, C., Skogseid, B., Oberg, K., Nakamura, Y., Nordenskjold, M., 1988. Multiple endocrine neoplasia type 1 gene maps to chromosome 11 and is lost in insulinoma. Nature 332, 85 –87. Lee, K.J., Jessell, T.M., 1999. The specification of dorsal cell fates in the vertebrate central nervous system. Annu. Rev. Neurosci. 22, 261–294. Lemmens, I., Van de Ven, W.J., Kas, K., Zhang, C.X., Giraud, S., Wautot, V., Buisson, N., De Witte, K., Salandre, J., Lenoir, G., Pugeat, M., Calender, A., Parente, F., Quincey, D., Gaudray, P., De Wit, M.J., Lips, C.J., Hoppener, J.W., Khodaei, S., Grant, A.L., Weber, G., Kytola, S., Teh, B.T., Farnebo, F., Phelan, C., Hayward, N., Larsson, C., Pannet, A.A.J., Forbes, S.A., Duncan Bassett, J.H., Thakker, R.V., 1997. Identification of the multiple endocrine neoplasia type 1 (MEN1) gene. The European Consortium on MEN1. Hum. Mol. Genet. 6, 1177–1183. Lemmens, I.H., Forsberg, L., Pannett, A.A., Meyen, E., Piehl, F., Turner,

J.J., Van de Ven, W.J., Thakker, R.V., Larsson, C., Kas, K., 2001. Menin interacts directly with the homeobox-containing protein Pem. Biochem. Biophys. Res. Commun. 286, 426 –431. Manickam, P., Vogel, A.M., Agarwal, S.K., Oda, T., Spiegel, A.M., Marx, S.J., Collins, F.S., Weinstein, B.M., Chandrasekharappa, S.C., 2000. Isolation, characterization, expression and functional analysis of the zebrafish ortholog of MEN1. Mamm. Genome 11, 448–454. Ohkura, N., Kishi, M., Tsukada, T., Yamaguchi, K., 2001. Menin, a gene product responsible for multiple endocrine neoplasia type 1, interacts with the putative tumor metastasis suppressor nm23. Biochem. Biophys. Res. Commun. 282, 1206–1210. Pandolfi, P.P., Roth, M.E., Karis, A., Leonard, M.W., Dzierzak, E., Grosveld, F.G., Engel, J.D., Lindenbaum, M.H., 1995. Targeted disruption of the GATA3 gene causes severe abnormalities in the nervous system and in fetal liver haematopoiesis. Nat. Genet. 11, 40 –44. Pitman, J.L., Lin, T.P., Kleeman, J.E., Erickson, G.F., MacLeod, C.L., 1998. Normal reproductive and macrophage function in Pem homeobox gene-deficient mice. Dev. Biol. 202, 196–214. Sabapathy, K., Hu, Y., Kallunki, T., Schreiber, M., David, J.P., Jochum, W., Wagner, E.F., Karin, M., 1999. JNK2 is required for efficient T-cell activation and apoptosis but not for normal lymphocyte development. Curr. Biol. 9, 116 –125. Stewart, C., Parente, F., Piehl, F., Farnebo, F., Quincey, D., Silins, G., Bergman, L., Carle, G.F., Lemmens, I., Grimmond, S., Xian, C.Z., Khodei, S., Teh, B.T., Lagercrantz, J., Siggers, P., Calender, A., Van de Vem, V., Kas, K., Weber, G., Hayward, N., Gaudray, P., Larsson, C., 1998. Characterization of the mouse Men1 gene and its expression during development. Oncogene 17, 2485–2493. Sukhodolets, K.E., Hickman, A.B., Agarwal, S.K., Sukhodolets, M.V., Obungu, V.H., Novotny, E.A., Crabtree, J.S., Chandrasekharappa, S.C., Collins, F.S., Spiegel, A.M., Burns, A.L., Marx, S.J., 2003. The 32-kilodalton subunit of replication protein A interacts with menin, the product of the MEN1 tumor suppressor gene. Mol. Cell. Biol. 23, 493 –509. Suphapeetiporn, K., Greally, J.M., Walpita, D., Ashley, T., Bale, A.E., 2002. MEN1 tumor-suppressor protein localizes to telomeres during meiosis. Genes Chromosomes Cancer 35, 81 –85. Thepot, D., Weitzman, J.B., Barra, J., Segretain, D., Stinnakre, M.G., Babinet, C., Yaniv, M., 2000. Targeted disruption of the murine junD gene results in multiple defects in male reproductive function. Development 127, 143–153. Wang, Z.Q., Auer, B., Stingl, L., Berghammer, H., Haidacher, D., Schweiger, M., Wagner, E.F., 1995. Mice lacking ADPRT and poly(ADP-ribosyl)ation develop normally but are susceptible to skin disease. Genes Dev. 9, 509 –520. Wautot, V., Khodaei, S., Frappart, L., Buisson, N., Baro, E., Lenoir, G.M., Calender, A., Zhang, C.X., Weber, G., 2000. Expression analysis of endogenous menin, the product of the multiple endocrine neoplasia type 1 gene, in cell lines and human tissues. Int. J. Cancer 85, 877–881. Yang, X., Letterio, J.J., Lechleider, R.J., Chen, L., Hayman, R., Gu, H., Roberts, A.B., Deng, C., 1999. Targeted disruption of SMAD3 results in impaired mucosal immunity and diminished T cell responsiveness to TGF-beta. EMBO J. 18, 1280– 1291. Zhu, Y., Richardson, J.A., Parada, L.F., Graff, J.M., 1998. Smad3 mutant mice develop metastatic colorectal cancer. Cell 94, 703–714.