Cited1 Is Required in Trophoblasts for Placental Development and for ...

MOLECULAR AND CELLULAR BIOLOGY, Jan. 2004, p. 228–244 0270-7306/04/$08.00⫹0 DOI: 10.1128/MCB.24.1.228–244.2004 Copyright © 2004, American Society for Microbiology. All Rights Reserved.

Vol. 24, No. 1

Cited1 Is Required in Trophoblasts for Placental Development and for Embryo Growth and Survival† Tristan A. Rodriguez,1‡ Duncan B. Sparrow,2 Annabelle N. Scott,2 Sarah L. Withington,2 Jost I. Preis,2 Jan Michalicek,3 Melanie Clements,1 Tania E. Tsang,2 Toshi Shioda,4 Rosa S. P. Beddington,1§ and Sally L. Dunwoodie1,2,5* Mammalian Development Division, National Institute for Medical Research, London, United Kingdom1; Developmental Biology Program2 and Molecular Cardiology Program,3 Victor Chang Cardiac Research Institute, and Department of Biotechnology and Biomolecular Sciences, University of New South Wales,5 Sydney, Australia; and Laboratory of Tumor Biology, Massachusetts General Hospital Cancer Center, Charlestown, Massachusetts4 Received 4 November 2002/Returned for modification 13 December 2002/Accepted 18 September 2003

Cited1 is a transcriptional cofactor that interacts with Smad4, estrogen receptors ␣ and ␤, TFAP2, and CBP/p300. It is expressed in a restricted manner in the embryo as well as in extraembryonic tissues during embryonic development. In this study we report the engineering of a loss-of-function Cited1 mutation in the mouse. Cited1 null mutants show growth restriction at 18.5 days postcoitum, and most of them die shortly after birth. Half the heterozygous females, i.e., those that carry a paternally inherited wild-type Cited1 allele, are similarly affected. Cited1 is normally expressed in trophectoderm-derived cells of the placenta; however, in these heterozygous females, Cited1 is not expressed in these cells. This occurs because Cited1 is located on the X chromosome, and thus the wild-type Cited1 allele is not expressed because the paternal X chromosome is preferentially inactivated. Loss of Cited1 resulted in abnormal placental development. In mutants, the spongiotrophoblast layer is irregular in shape and enlarged while the labyrinthine layer is reduced in size. In addition, the blood spaces within the labyrinthine layer are disrupted; the maternal sinusoids are considerably larger in mutants, leading to a reduction in the surface area available for nutrient exchange. We conclude that Cited1 is required in trophoblasts for normal placental development and subsequently for embryo viability. 75). In the mouse, Cited1, Cited2, and Cited4 (Cited3 is not present in mammals) each display restricted and distinct expression profiles; however, there is some overlap in the expression of Cited1 and Cited2. We have previously shown that Cited1 transcripts are localized to progenitors of the heart, limb, axial skeleton, and placenta (17). Cited1 is also expressed in the heart and mammary gland in adults, as well as melanocytes, melanoma cells, and papillary thyroid carcinoma (17, 28, 34, 60, 73). To investigate the function of Cited1 during embryonic development, we generated a Cited1 null mutant mouse. Here we report that the penetrance of the Cited1 null phenotype is dependent on the genetic background and that the placenta develops abnormally in Cited1 mutants. We show that Cited1 (an X-linked gene) is expressed in placental trophoblasts and that it is required in these cells for normal placental development. In addition, we demonstrate that the spongiotrophoblast layer is irregular in shape and that the maternal blood sinusoids of the labyrinth are greatly enlarged in Cited1 mutants. Consistent with this, Cited1 mutants are small in the perinatal period, with the mutation resulting in death on the day of birth or shortly after. The exact cause of death is unknown; however, it appears to be a consequence of aberrant placental function late in gestation.

The CITED (for CBP/p300-interacting transactivators with glutamic acid [E]/aspartic acid [D]-rich carboxyl-terminal domain) gene family is represented by four genes: Cited1 (formerly Msg1), Cited2 (formerly Mrg1), Cited3, and Cited4 (2, 12, 17, 59, 60, 75). CITED proteins share three conserved regions of homology (CR1, CR2, and CR3), outside of which they are divergent. Given that they lack sequence identity to known protein domains, they therefore represent a novel family of proteins. When tethered to heterologous DNA-binding domains, CITED proteins activate transcription, a function that is dependent on the CR2 domain that binds the transcriptional coactivators CBP/p300 (11, 73). No evidence exists that CITED proteins bind DNA; however, they can interact with DNA-binding proteins and potentiate the activation of reporter constructs in vitro. For example, Cited1 binds Smad4, and estrogen receptors ␣ and ␤ (ER␣/␤) (61, 74); Cited2 binds the LIM domain of Lhx2; and Cited2 and Cited4 bind TFAP2, as does Cited1 but weakly (6, 12, 24, 75). Embryonic expression of the CITED family during various stages of vertebrate development has been analyzed (2, 17, 55,

* Corresponding author. Mailing address: Victor Chang Cardiac Research Institute, 384 Victoria St., Darlinghurst, NSW 2010, Australia. Phone: 612 9295 8513. Fax: 612 9295 8501. E-mail: s.dunwoodie @victorchang.unsw.edu.au. † Dedicated to the memory of Rosa Beddington (23 March 1956 to 18 May 2001). ‡ Present address: MRC Clinical Sciences Centre, Imperial College, London, United Kingdom. § Deceased.

MATERIALS AND METHODS Targeting vectors and generation of the Cited1flox and Cited1neo mutant mouse lines. Two genomic clones spanning the Cited1 locus were isolated using Cited1 cDNA (17). A 20-kb Cited1 genomic clone (Cited1/sv) was isolated from a 129sv library (Stratagene), and a 17-kb Cited1 genomic clone (Cited1/Ola) was isolated from a 129Ola genomic library. The targeting construct contained a 6.5-kb 5⬘

228

VOL. 24, 2004

Cited1 IS REQUIRED FOR NORMAL PLACENTAL DEVELOPMENT

229

FIG. 1. Generation of Cited1 null mutant mice. (A) Schematic representation of the Cited1 locus, targeting vector, and Cited1flox and Cited1neo alleles. The exon (black box), translation start site (ATG), LoxP site (empty box), GFP, flip recombinase target (FRT), phosphoglycerate kinase I promoter (PGK), neomycin gene (neo) are shown. (B) Southern blot analysis distinguishing between wild-type ES cells (XY) and those carrying the targeted allele (XfloxY). The 5⬘ probe hybridizes with a 9.5-kb fragment following BamHI restriction in wild-type (XY) ES cells and an 8.5-kb fragment in correctly targeted XfloxY ES cells. (C) Southern blot analysis distinguishes between the Cited1 targeted (Xflox) and null (Xneo) alleles in mice. A neo-derived probe hybridizes with a 9.5-kb fragment identifying the XfloxX⫹ progeny and a 4-kb fragment identifying the XneoX⫹ progeny. (D) Genotype and sex of embryos and mice as determined using PCR. Three primers distinguish between the Cited1 wild-type (Cited1⫹) and null (Cited1neo) alleles. Males possess the Sry gene located on the Y chromosome. Heterozygous female, Cited1⫹/neo, wild-type male, Cited1⫹/Y; null male, Cited1neo/Y; null female, Cited1neo/neo.

230

RODRIGUEZ ET AL.

MOL. CELL. BIOL.

TABLE 1. Genotype of offspring from Cited1⫹/Y ⫻ Cited1⫹/neo and Cited1neo/Y ⫻ Cited1⫹/neo intercrosses with a mixed genetic background (129Ola; C57BL/6; C57BL/10; CBA)a Intercross

Cited1⫹/Y ⫻ Cited1⫹/neo Cited1neo/Y ⫻ Cited1⫹/neo

No. of offspring with Cited1 genotype of: ⫹m/Y neo/Y ⫹p/⫹m ⫹p/neo neo/neo

49 110

43 81

64 NA

59 115

NAb 71

P

0.17 0.002

a Crosses were performed as designated, and the genotype was established on postnatal day 10. Cited1 is on the X chromosome; therefore, wild-type males are indicated by ⫹/Y; hemizygous (null) males are indicated by neo/Y; wild-type females are indicated by ⫹/⫹; heterozygous females are indicated by ⫹/neo, and homozygous null females are indicated by neo/neo. Ratios of genotypes were tested for goodness of fit to expected Mendelian segregation (1:1:1:1) by chisquare analysis, calculated with 3 degrees of freedom. ⫹p, paternal wild-type allele; ⫹m, maternal wild-type allele. b NA, not applicable.

homology arm (derived from Cited1/Ola) interrupted by a LoxP sequence 51 bp 5⬘ of the Cited1 ATG. A second LoxP sequence, preceded by translation stop codons in all reading frames, was placed 5⬘ of the coding sequence for the green fluorescent protein (GFP), the Pgk-neo expression cassette flanked by FRT sites, and a 2.5-kb 3⬘ homology arm (derived from Cited1/sv) (18, 79) (Fig. 1A). The Pgk-Tk expression cassette was used for negative selection. This vector was linearized with XhoI and electroporated into E14.2 embryonic stem (ES) cells as described previously (62). Following double selection with G418 and ganciclovir, 209 ES cell clones were picked, expanded, and frozen by standard methods (27). Homologous recombinants were identified following BamHI restriction and hybridization with sequences located 5⬘ and external to the homology arms (Fig. 1A and B). A total of 13 targeted clones were identified, and chimeric males were generated from 1 of these clones following blastocyst injection by standard methods (27). These mice were mated with C57BL/6 females to establish F0 heterozygous females. Animals carrying this conditional allele of Cited1 were designated Cited1flox. The Cited1 coding region was deleted in vivo to generate the Cited1neo mouse line as follows: CBAxC57BL/10 F1 females were mated to F1 Cited1flox males generated from the cross between F0 Cited1flox females and C57BL/6 males. A CMV-Cre plasmid was introduced by pronuclear injection into fertilized eggs from these mice by standard procedures (27, 43). Digestion with BamHI and hybridization with a probe specific for the neo gene allowed identification of mice carrying the null Cited1 allele (Cited1neo) (Fig. 1C). Genotyping of the wild-type and modified Cited1 alleles. Mice and embryos were genotyped by PCR. DNA samples were prepared from tails, yolk sacs, or whole embryos as described previously (27). The PCR primers used to distinguish between the Cited1⫹ and Cited1neo alleles were Cited1-F (5⬘-TTACTTG CAGACCAACAGGC-3⬘), Cited1-R (5⬘-TGCTTCTTTGACCCATTTCC-3⬘), and GFP-R (5⬘-TGTTGCATCACCTTCACCCT-3⬘). The fragment sizes generated were 367 bp for the Cited1⫹ allele and 206 bp for the Cited1neo allele. The primers were used at a ratio of 1:2:1, respectively. Embryos were sexed by using Sry primers: Sry F (5⬘-TTCAGCCCTACAGCCACATGA-3⬘) and Sry R (5⬘-A TGTGGGTTCCTGTCCCACTG-3⬘) (63). Histological analysis. In all cases, the central region of the placenta was examined since this is where the morphology is well defined and most consistent. For histological analysis, embryos were fixed in Bouin’s fixative, dehydrated, embedded in paraffin wax, sectioned (7-␮m-thick sections), and stained with hematoxylin-eosin (29). Placentas for cryosection were dissected at 14.5 days postcoitum (p.c.), fixed in 4% paraformaldehyde (in phosphate-buffered saline [PBS]) overnight at 4°C, incubated at 4°C overnight in 30% sucrose in PBS, incubated in OCT (BDH) for 10 min at room temperature, and frozen at ⫺80°C. Sections (10 to 12 ␮m thick) were cut, allowed to dry, and either refrozen (at ⫺80°C) or immediately processed for RNA in situ hybridization (16). RNA probes were generated as previously described: Cited1 (17), 4311 (Tpbp) (14), Mest (Peg1) (38), proliferin (PLF) (33), and mouse placental lactogen II (mPLII) (58). The sections were counterstained with eosin. For immunohistochemistry, placentas were dissected on day 14.5 or 16.5 p.c., fixed in 4% paraformaldehyde (in PBS), paraffin embedded, and serially sectioned. Dewaxed sections were incubated with an affinity-purified rabbit antiCITED1 antibody raised against residues 188 to 202 at the carboxy terminus of CITED1 (60), and the antigen was visualized with a Vectastain ABC kit as specified by the manufacturer (Vector Labs). Alkaline phosphatase activity was detected by dewaxing sections and incubating them in nitroblue tetrazolium–5-

bromo-4-chloro-3-indolyl phosphate (NBT/BCIP) substrate (Roche) for 15 min at room temperature in the dark. The sections were subsequently counterstained with nuclear fast red. Vascular casts. To make fetal vascular casts, resin injections were performed using Batson’s no. 17 plastic replica and corrosion kit (containing methyl methacrylate casting resin) (Polysciences Inc.). Aliquots (2 ml) of base solution containing either red or blue pigment were prepared, and either 1 ml of catalyst or 10 ␮l of promoter was added. Embryos aged 16.5 days p.c. were dissected into M2 medium (27) plus 10% fetal calf serum, taking care to remove all membranes surrounding the umbilical vessels. At this point, the aliquots of blue base solution were mixed, drawn into a 1-ml syringe, attached to a 30.5-gauge needle, and placed on ice. Each embryo was placed into a 3-cm petri dish containing M2 medium, and the placenta was pulled over the side of the dish so that the umbilical vessels were dry and under tension. Chilled heparinized xylocaine (1% xylocaine in 0.9% NaCl with 1 IU of heparin/ml) was first injected into the umbilical artery, to dilate the vasculature and clear the placenta of blood. Blue resin was then also injected into the umbilical artery; this was repeated for every placenta in the litter. The aliquots of red base solution were then mixed, drawn up into a syringe, and injected into each umbilical vein, also using a 30.5-gauge needle. Following the injections, the placentas were washed in PBS and left at 4°C at least overnight to allow the resin to set. Fetal vascular casts were generated and analyzed for 21 wild-type (Cited1⫹/Y or Cited1⫹/⫹), 3 heterozygous females (Cited1⫹p/neo), and 9 null male (Cited1neo/Y) placentas. Maternal vascular casts were made by methods previously described by Adamson et. al. (1) and modified as follows. Female mice on day 16.5 of pregnancy were anesthetised with an intraperitoneal dose of ketamine (100 mg/kg of body weight) and xylazine (20 mg/kg of body weight). Each mouse was ventilated with O2 at 120 breaths/min and a volume of 0.5 ml. Occlusive catheters filled with heparinized saline were inserted into the right carotid and the femoral artery for drug infusion and continuous pressure monitoring, respectively. The chest was opened, and 0.2 ml of 10-IU/ml heparin at 0.1 ml/min was infused into the right carotid artery. This was then followed by infusion of 3 mM KCl solution (0.1 to 0.3 ml) to stop the heart. For arterial casts, a polyethylene catheter (0.80 mm [outer diameter] and 0.50 mm [inner diameter]) was placed into the aortic arch. The right atrium was cut so that it could serve as an exit point. An infusion pump (sp200i syringe pump; KD Scientific) was used to perfuse the animal’s circulation with 10 ml of warm (40 to 50°C) heparinized saline (0.9% NaCl, 1 IU of heparin/ml, 1% xylocaine) at 2 ml/min, followed by 10 ml of the same perfusate chilled to 4°C. The prechilled mixed resin was infused at 0.2 ml/min. This rate was reduced during infusion, so that the femoral arterial pressure never exceeded the physiologically comparable pressure of 100 mm Hg. A total volume of 3 ml of plastic mixture was infused, and then a ligature was tightened around the intrathoracic vena cava. The infusion syringe was kept under pressure (20 mm Hg) until the plastic had set. For venous casts, a polyethylene catheter was inserted into the intrathoracic vena cava in a retrograde direction and then the infusion procedure described above was followed; however, only 2 ml of resin was used. Maternal vascular casts were generated and analyzed for 15 wild-type (Cited1⫹/Y or Cited1⫹/⫹), 10 heterozygous female (Cited1⫹p/neo), and 10 null male (Cited1neo/Y) placentas. After both fetal and maternal injections, the internal cast was visualized by digesting the surrounding placental tissue with 20 to 30% KOH in H2O. The casts were washed thoroughly in distilled water, examined by light microscopy (Leica MZ8 stereo dissecting microscope), air dried, sputtered with gold, and analyzed with a scanning electron microscope (Cambridge S360). Morphometric and statistical analysis. Morphometric analysis was performed using NIH Image 1.62 software. For comparison of the areas of the labyrinthine and spongiotrophoblast layers of the placentas, cryosections processed by RNA in situ hybridization for 4311 expression were analyzed. The cross-sectional area of each of these layers was determined for each of 12 sections (taken every eighth section, 12 ␮m wide, starting from the center of the placenta) from two wild-type (Cited1⫹/Y) and two null (Cited1neo/Y) placentas at 14.5 days p.c. These data were compared using analysis of variance. For comparison of the maternal sinusoid size within the labyrinthine layer, placental sections processed for alkaline phosphatase activity were analyzed. The length around the sinusoids and the cross-sectional area of the sinusoids were measured. The sinusoids were measured from three sections (720 by 530 ␮m), 160 ␮m apart, taken from a wild-type (Cited1⫹/Y or Cited1⫹/⫹), a heterozygous female (Cited1⫹p/neo), and a null male (Cited1neo/Y) placenta. A total of 151 to 255 measurements were made for each placenta. These data were not normally distributed, either before or following transformation; therefore, the nonparametric Wilcoxon test was employed. For comparison of the sizes of the fetal capillaries within the placenta, the diameter of the capillaries was measured from scanning electron micrographs.


VOL. 24, 2004

231

FIG. 2. Loss of Cited1 results in reduced embryo weight at birth. Cited1neo/Y ⫻ Cited1⫹/neo crosses were performed. Weights are shown on PND0 (birth) (A) and PND160 (B). Genotype annotation is as described for Table 1. The mean weight (and standard error of the mean) was plotted against genotype. Nested analysis of variance demonstrates that the weight of Cited1 mutants (Cited1neo/neo and Cited1neo/Y) was significantly different from that of controls (Cited1neo/⫹ and Cited1⫹/Y) at birth (P ⫽ 0.0022). At 160 days, although the weights are different between the sexes, there is no statistical difference between genotypes within each sex.

The diameter was measured 20 ␮m from the tip of the capillary in 16 to 72 capillaries from two to six different capillary beds in three wild-type (Cited1⫹/Y or Cited1⫹/⫹) and four null male (Cited1neo/Y) individuals. These data were compared using nested analysis of variance. Primordial germ cell analysis. Embryos were collected on day 8.5 p.c., and alkaline phosphatase-expressing cells were identified by the method of Lawson et al. (30) with the following modifications. The number of somites was recorded before fixation in 4% paraformaldehyde (more than 2 h at 4°C). Embryos were dehydrated in 70% ethanol overnight and washed three times in PBS at room temperature and twice with NTMT (100 mM NaCl, 100 mM Tris-HCl [pH 9.5], 50 mM MgCl2, 0.1% Tween 20) with 3 mM levamisole (Sigma). Primordial germ cells were stained in NBT/BCIP (GIBCO BRL). Whole embryos were embedded in paraffin, serially sectioned at 8 ␮m thick, and processed for histological analysis. Primordial germs cells were counted from 6 wild-type (Cited1⫹/Y or Cited1⫹/⫹), 5 heterozygous female (Cited1⫹p/neo), and 16 null (Cited1neo/Y or Cited1neo/neo) embryos.

RESULTS Targeted disruption of the Cited1 gene and generation of null mutant mice. The conditional Cited1flox allele was generated by homologous recombination in E14.2 mouse ES cells. The Cited1 locus, targeting vector, targeted allele, and deleted allele are represented in Fig. 1A. Homologous recombination in ES cell clones was identified by Southern blot analysis (Fig. 1B). The entire Cited1 coding region was subsequently deleted via Cre-mediated recombination in vivo following pronuclear injection of a Cre-expressing plasmid. Founder mice carrying the Cited1 null allele (Cited1neo) were identified by Southern blot analysis (Fig. 1C). Initial confirmation of genotype was achieved using Southern blot analysis. After this, the genotype was determined using PCR (Fig. 1D). The targeting vector was designed so that Cre-mediated recombination would bring the coding region of GFP under the transcriptional control of the Cited1 locus. However, no expression of GFP could be detected in either ES cell-derived embryoid bodies or embryos (data not shown). To investigate this lack of expression, genomic DNA from mice carrying the Cited1neo allele was sequenced and a single nucleotide insertion was detected that caused a frameshift in the GFP coding region. This frameshift

would result in the production of a premature termination codon (data not shown). To establish that the Cre-mediated recombination had occurred as predicted, deletion of the Cited1 locus (from 51 nucleotides upstream of the Cited1 translation start site) was confirmed by sequencing the Cited1neo allele from genomic DNA (data not shown). On a mixed genetic background Cited1neo affects viability and growth. In the mouse, Cited1 maps to the X chromosome at 40.1 centimorgans (cM) (20). Therefore, mice had to be intercrossed in two distinct ways to generate progeny that represent each possible genotype. Crosses between wild-type Cited1 males (Cited1⫹/Y) and heterozygous Cited1 females (Cited1⫹/neo) were expected to produce the following offspring with equal probability: Cited1⫹/Y, Cited1neo/Y, Cited1⫹/neo, and Cited1⫹/⫹ (Table 1). This intercross showed that although the genotype of progeny (on postnatal day 10) did not significantly deviate from the Mendelian ratio, 12% fewer Cited1 null males (Cited1neo/Y) survived compared to wild-type males. However, in the cross between Cited1 null males (Cited1neo/Y) and heterozygous Cited1 females (Cited1⫹/neo), the ratio of progeny that survived (Cited1⫹/Y, Cited1neo/Y, Cited1⫹/neo, and Cited1neo/neo) deviated significantly from the Mendelian ratio; there were 20% fewer Cited1 null mice than wild-type mice (Table 1). Surviving null mice were smaller than their littermates; therefore, mice were weighed at birth and at 32 regular time points until they reached 160 days old. On the day of birth, Cited1neo/Y males were 23% lighter than Cited1⫹/Y control littermates while Cited1neo/neo females were 19% lighter than Cited1⫹/neo controls (Fig. 2A). After 160 days, Cited1 null mice were not significantly different in weight from control littermates (Fig. 2B). Taken together, these data demonstrate that some Cited1 null mice die prior to postnatal day 10 and that those that do survive are smaller at birth than the controls are. Since the growth deficit is not apparent after 160 days, this suggests that Cited1 null individuals do not suffer from a growth defect per se but, rather, that they are likely to be subjected to an impaired uterine environment.

232

RODRIGUEZ ET AL.

MOL. CELL. BIOL.

TABLE 2. Genotype of offspring from Cited1neo/Y ⫻ Cited1⫹/neo and Cited1⫹/Y ⫻ Cited1⫹/neo matings after six generations of backcrossing with C57BL/6 wild-type animalsa No. of offspring with Cited1 genotype of:

Backcross and day

⫹m/Y

Cited1neo/Y ⫻ Cited1⫹/neo Postnatal day 28 Postnatal day 0 Prenatal (total) 18.5 days p.c. 16.5 days p.c. 15.5 days p.c. 14.5 days p.c. 12.5 days p.c. 9.5 days p.c. 8.5 days p.c.

305 27 31 1 6 2 8 5 3 6

Cited1⫹/Y ⫻ Cited1⫹/neo Postnatal day 28 Postnatal day 0 Prenatal (total) 18.5 days p.c. 16.5 days p.c. 15.5 days p.c. 14.5 days p.c. 13.5 days p.c. 12.5 days p.c. 11.5 days p.c. 10.5 days p.c. 9.5 days p.c.

51 31 (1 dead) 124 17 67 9 9 3 8 1 6 4

neo/Y

⫹m/⫹p

⫹m/neo

56 18 (2 dead) 21 1 1 3 7 2 4 3

NAb

279 32 27 3 1 2 5 6 6 4

9 15 (4 dead) 136 16 79 10 11 2 7 3 4 4

51 31 (1 dead) 121 16 64 7 12 4 9 0 5 4

⫹p/neo

27 17 (2 dead) 125 17 66 11 13 7 4 2 3 2

neo/neo

P

89 18 (2 dead) 23 0 1 1 1 8 6 5

⬍0.0001 live 0.04 0.51 0.28 0.04 0.95 0.14 0.31 0.70 0.77

NA

⬍0.0001 live 0.003 0.80 1.00 0.57 0.81 0.85 0.32 0.57 0.34 0.77 0.84

a Crosses were performed as designated, with genotype annotation as described for Table 1. Wild-type males (⫹/Y) mated with heterozygous females (⫹/neo) produce Cited1 heterozygous females that carry no active copy of the Cited1⫹ allele in trophectoderm derivatives and are annotated ⫹p/neo. Ratios of genotypes were tested for goodness of fit to the expected Mendelian segregation (1:1:1:1) by chi-square analysis, calculated with 3 degrees of freedom. The paternal wild-type allele (⫹p), and maternal wild-type allele (⫹m) are indicated. Postnatal day 0 is the day of birth. b NA, not applicable.

Death due to the Cited1 null allele occurs at birth, is dependent on genetic background, and is due to a defect in trophoblast-derived extraembryonic tissues. In combination with an inbred genetic background, death occurred in the majority of Cited1 null individuals and in a considerable proportion of heterozygous females (Table 2). The loss of heterozygous females represents a parent-of-origin effect since those that succumbed to the mutation did so only when the Cited1 wild-type allele was inherited from their father. Penetrance of the phenotype associated with the Cited1 null allele was increased in the presence of a C57BL/6 genetic background. The Cited1neo allele was backcrossed six times with C57BL/6 individuals. Subsequently, when Cited1 null males (Cited1neo/Y) and heterozygous Cited1 females (Cited1⫹/neo) were crossed, null mutants were underrepresented by 78% compared with controls (Cited1⫹/Y) on postnatal day 28. The numbers of Cited1⫹/neo heterozygous females were not different from the numbers of control mice derived from this cross. However, when wild-type Cited1 males (Cited1⫹/Y) and heterozygous Cited1 females (Cited1⫹/neo) were crossed, there were 82% fewer Cited1 null males (Cited1neo/Y) than controls (Cited1⫹/Y) on postnatal day 28. In addition, this cross revealed that there were 47% fewer heterozygous Cited1 females (Cited1⫹/neo) than control females (Cited1⫹/⫹) (Table 2). Since viability is reduced in individuals carrying the Cited1neo allele, we therefore determined the time when death occurred by identifying the genotype of embryos at different stages of gestation. Table 2 demonstrates that regardless of the type of parental cross, death did not

significantly occur during gestation. This was the case for both Cited1 null (Cited1neo/Y and Cited1neo/neo) and heterozygous (Cited1⫹P/neo) females. In fact, the majority of these individuals died on the day of birth or shortly thereafter. The underrepresentation of heterozygous Cited1 females (Cited1⫹/neo), derived from the Cited1⫹/neo ⫻ Cited1⫹/Y cross, demonstrates that there is a parent-of-origin effect on allele behavior since the proportion of heterozygous females (Cited1⫹/neo) on this background was not reduced when Cited1 null males (Cited1neo/Y) were crossed with heterozygous females (Cited1⫹/neo). Specifically, if a heterozygous Cited1 female (Cited1⫹/neo) inherited a wild-type Cited1 allele from the mother, then the expected number of Cited1⫹/neo individuals survived; however, if the wild-type allele was inherited from the father, a significant proportion (47%) of heterozygous females died. This parent-of-origin effect affects embryo viability in heterozygous Cited1 females (Cited1⫹/neo) and is likely to be due to the distinct regimes applied to X chromosome inactivation in specific tissues of the conceptus. In females, X chromosome inactivation is random both in the embryo and in mesodermally derived extraembryonic tissues. By contrast, in extraembryonic ectoderm and endoderm, only the paternal X chromosome is inactivated (21, 47, 66). Therefore, heterozygous females that inherit the Cited1 wild-type allele from their mother (Cited1⫹m/neo) are both genotypically heterozygous and phenotypically wild type in extraembryonic ectoderm and endoderm. However, heterozygous females that inherit the Cited1 wild-type allele from their father (Cited1⫹p/neo) are ge-

VOL. 24, 2004


233

FIG. 3. Loss of Cited1 results in reduced weight perinatally. Placenta (A) and embryo (B) weights are shown at various stages. All data were derived from the crossing of Cited1⫹/Y ⫻ Cited1⫹/neo individuals, except for embryo weights designated PND0b, which were taken from the progeny of a Cited1neo/Y ⫻ Cited1⫹/neo cross. Genotype annotation is as described for Table 1. For cross Cited1⫹/Y ⫻ Cited1⫹/neo, at 14.5, 16.5, and 18.5 days p.c. and PND0a, wild-type males (n ⫽ 4, 12, 13, and 27) and females (n ⫽ 4, 12, 9, 31 [open squares]) were pooled since there was no statistical difference between these two groups. These were compared to heterozygous females (Cited1⫹p/neo, n ⫽ 6, 10, 12, and 17 [solid diamonds]) and hemizygous (null) males (n ⫽ 3, 14, 14, and 15 [solid squares]). For cross Cited1neo/Y ⫻ Cited1⫹/neo, wild-type males (n ⫽ 27) and heterozygous females (n ⫽ 32 [open squares]) were pooled since there was no statistical difference between the two groups. These were compared to null females (n ⫽ 18 [solid diamonds]) and hemizygous (null) males (n ⫽ 18; solid squares). The mean weight (and standard error of the mean) was plotted against genotype. Analysis of variance was used to identify differences between genotypes. Asterisks indicate a statistical difference between wild-type embryo weight and the weight of heterozygous females or hemizygous (null) males (P ⫽ ⬍ 0.0001). Pound signs indicate statistical difference between control (wild-type males and heterozygous females [Cited1⫹m/neo]) and null females (P ⫽ 10⫺19) or hemizygous (null) males (P ⫽ 10⫺13).

notypically heterozygous but phenotypically null in extraembryonic ectoderm and endoderm since the paternal Cited1⫹P allele is inactivated. We conclude that Cited1 inactivation in extraembryonic ectoderm and/or endoderm is responsible for the observed deaths. Cited1 is required to maintain embryonic growth late in gestation. Since Cited1 mutants die shortly after birth, we determined embryo and placenta weight during the perinatal period (Fig. 3). Embryos and placentas were weighed on days 14.5, 16.5, and 18.5 (day 18.5 p.c. is the day before birth). In addition, pups were weighed on the day of birth (postnatal day 0 [PND0]). Placental weights were the same for all genotypes at all stages examined (Fig. 3A). The same held true for embryo weights on days 14.5 and 16.5 p.c. However on day 18.5 p.c. and on PND0, both Cited1 null (Cited1neo/Y and Cited1neo/neo) and heterozygous (Cited1⫹P/neo) females weighed significantly less than did the controls (Fig. 3B). These data demonstrate that the Cited1 mutant phenotype is manifest in the embryo, late in gestation, but does not lead to death in utero (Table 2; Fig. 3). Cited1 is expressed in trophectoderm-derived cells of the placenta. The preceding genetic analysis demonstrates that Cited1 is required in extraembryonic ectoderm and/or endoderm for embryo survival. Using in situ analysis, we previously reported that the Cited1 transcript is localized to extraembryonic ectoderm, visceral endoderm, and visceral yolk sac endoderm (6.5 to 8.5 days p.c.); by Northern analysis, we found that it is also localized to the placenta (11.5 days p.c.) (16, 54). It is most likely that loss of Cited1 in the extraembryonic ectoderm-derived cells of the placenta (rather than in extraembryonic tissues during early development) is the cause of the late-onset death. To further examine Cited1 expression

in the placenta, we used in situ RNA hybridization and immunohistochemistry (Fig. 4). The mature placenta is established in mouse by day 10 p.c. and consists of trophoblasts (extraembryonic ectoderm) and endothelial and stromal cells (extraembryonic mesoderm). It is composed of three principal layers: an outer layer of secondary trophoblast giant cells, a middle spongiotrophoblast layer, and the innermost labyrinth. The labyrinth contains both trophoblasts and mesodermally derived cells that are embryonic in origin. Cited1 was expressed in all trophoblast-derived tissues of the placenta (Fig. 4). Cited1 transcripts were localized to secondary giant cells and spongiotrophoblasts (Fig. 4A to C). Expression was also detected within the labyrinthine layer in cuboidal and elongated trophoblasts (Fig. 4D) and in lining maternal blood spaces at the base (embryo side) of the labyrinth (Fig. 4E). Expression in cuboidal and elongated trophoblasts was particularly clear when nuclear ␤-galactosidase, targeted to the Cited1 locus, was analyzed (data not shown). Like the RNA, the Cited1 protein was detected in each of these trophoblast cell types (Fig. 4F). In giant cells and spongiotrophoblasts, the protein was localized predominantly to the nucleus but was also detected in the cytoplasm (Fig. 4G and H), while protein was localized throughout the cell in the labyrinthine layer (Fig. 4I and J). These data demonstrate that Cited1 is expressed in trophoblast-derived placental cells but not in mesoderm-derived cells (such as the fetal vasculature). This supports the hypothesis that Cited1 heterozygous females in which the Cited1⫹ allele is inherited from the father (Cited1⫹P/neo) die due to inactivation of the paternally acquired Cited1⫹P allele in trophoblast-derived cells of the placenta. Cited1 is required for normal placental morphology. Reduced embryonic weight on day 18.5 p.c. indicated that pla-

234

RODRIGUEZ ET AL.

MOL. CELL. BIOL.

VOL. 24, 2004


cental function was likely to be abnormal late in gestation; therefore, placentas were examined each day from days 14.5 to 18.5 p.c. Hematoxylin-and-eosin staining of sections showed that secondary giant cells, spongiotrophoblasts, and labyrinthine trophoblasts were present in placentas carrying the Cited1neo allele as well as in control placentas (data not shown). Gene expression analysis on day 14.5 p.c. demonstrated that the border between the spongiotrophoblast and labyrinthine layers was irregular in Cited1 null males (Cited1neo/Y) and heterozygous females (Cited1⫹P/neo). Fingerlike projections of spongiotrophoblasts (identified by 4311, PLF, and mPLII expression) often extended through the labyrinthine layer in mutants, unlike in controls (Fig. 5 and data not shown). Serial sections showed that the fetal vasculature (identified by Mest expression) was disrupted (Fig. 5A to D) due to the projection of these spongiotrophoblasts (Fig. 5E to L). Although the total size of each placenta was the same, there was a significant increase in the spongiotrophoblast layer and a concomitant decrease in the labyrinthine layer in Cited1 null males (Cited1neo/Y) compared with controls (Cited1⫹/Y) (Table 3). It was also clear that organization of the labyrinth was disrupted in the mutant placentas. The size of trophoblastlined maternal sinusoids (trophoblasts identified by alkaline phosphatase activity) was compared between Cited1 null males (Cited1neo/Y), heterozygous females (Cited1⫹P/neo), and controls (Cited1⫹/Y) (Fig. 6, Table 4, and data not shown). The length of the surface surrounding the maternal sinusoid and its cross-sectional area were measured from placental sections. The mean values of surface length and the cross-sectional area were up to 62 and 262% greater in Cited1 mutants than in controls. This defect is consistent with the fact that Cited1 is expressed in labyrinthine trophoblasts (Fig. 4) and is required in the extraembryonic ectoderm, such as placental trophoblasts, for neonate survival. Cited1 is required for normal organization of the fetal and maternal blood in the labyrinth. To determine the arrangement of fetal blood vessels and maternal blood sinusoids in the labyrinthine layer of Cited1 mutants, we generated vascular casts on day 16.5 p.c. since this was the earliest stage at which a placental defect was detected using histological analysis. The arterial and venous systems were filled with resin, and a combination of light microscopy and scanning electron microscopy (SEM) showed that placentas carrying the Cited1neo allele were different from the controls. Maternal vascular casts were generated following resin infusion from the arterial (Fig. 7) or venous (data not shown) side. A lateral view of the vascular cast filled from the arterial side showed the arrangement of the maternal arteries and trophoblast-lined compartments of the placenta. The organization of the radial arteries, spiral arteries, and central canals was similar to that described by Adamson et al. (1) and appeared to be the same regardless of the genotype. However, the labyrinth was irregular in shape, and

235

when it was viewed from the base of the placenta (proximal to the embryo), it was clear that the arterial side casts differed between Cited1 mutants and controls. In all wild-type (Cited1⫹/Y or Cited1⫹/⫹ [n ⫽ 15]) placentas, canal branches and anastomosing sinusoids spread out producing an even circular base, whereas in heterozygous females (Cited1⫹P/neo [n ⫽ 10]) and null males (Cited1neo/Y [n ⫽ 10]) the base was very irregular in shape. Although this was clearly evident by observing with the naked eye, light microscopy and SEM revealed the extent of the irregularity (Fig. 7A to F). In addition, the sinusoids were examined at several sites within the labyrinth (centrally, laterally, at the edges, and internally following breaking the cast) using SEM. The degree of branching and the size of the sinusoids appeared to be uniform regardless of their position within the labyrinth (data not shown), and in the controls they were well branched and fairly regular in size. In contrast, the sinusoids in all mutants showed less branching and were considerably larger (Fig. 7G to I). Due to the irregular shape of the sinusoid casts, it was not possible to accurately quantify size differences; however, mutant sinusoids were approximately 2- to 3.5-fold greater in diameter than were those in control placentas. Sinusoids filled from the venous side were similarly large and poorly branched in mutant placentas (data not shown). The increase in the size of the sinusoids shown here in the resin casts was similar to the increases measured following histological analysis (Fig. 6). Infusion of resin into the umbilical artery (blue) and vein (red) on day 16.5 p.c. generated casts of the fetal vasculature (Fig. 8). These dually filled casts showed that the umbilical vessels extended into two planes of the placenta: within the plane of the base of the placenta, and through the labyrinthine layer toward the spongiotrophoblast layer. Within the base of the placenta, the artery (blue) and vein (red) branched in a radial pattern; the artery branched a few times, stopping at the periphery of the placenta, while the vein branched more extensively and stopped before reaching the periphery (Fig. 8A and B). The arterioles that expanded through the labyrinthine layer extended toward the spongiotrophoblast layer, where they branched before anastomosing into a dense mass of capillaries that extended back toward the base of the placenta. The venuoles extended only a short distance into the labyrinthine layer, without branching, and formed an equally dense capillary mass which extended up toward those derived from the arterioles (Fig. 8C to H and data not shown) (1). Although there was a degree of variation in the casts generated, we were unable to correlate this with the Cited1 genotype; therefore, the overall arrangement of fetal vasculature in both control and mutant placentas appeared to be essentially the same. The width of the resin-filled arterioles that extended into the labyrinthine layer was measured, but no difference was observed (data not shown). Since fetal capillaries are surrounded by maternal sinusoids, which are enlarged in Cited1 mutants, the

FIG. 4. Cited1 transcript and protein are expressed in trophectoderm-derived cells of the placenta. RNA in situ hybridization (A to E) and immunohistochemistry (F to J) show Cited1 transcript and protein localization on day 14.5 p.c. Cited1 is expressed in secondary giant cells (arrow in panels B and G), spongiotrophoblasts (C and H), and cuboidal (arrow) and elongated (arrowhead) trophoblasts within the labyrinthine layer (D and I) and lining maternal blood spaces at the base (embryo side) of the labyrinthine layer (arrow in panels E and J). Fetal blood vessel (v), maternal sinus (s), and mesenchyme surrounding fetal blood vessel (m) are also shown. Bar, 1.2 mm (A and F), 310 ␮m (B, E, G, and J), and 150 ␮m (C, D, H, and I).

236

RODRIGUEZ ET AL.

MOL. CELL. BIOL.


VOL. 24, 2004

TABLE 3. Morphometric analysis of mouse placentas on day 14.5 p.c. following RNA in situ hybridizationa

Location

Total placenta Spongiotrophoblasts Labyrinthine layer

Cross-sectional area (mm2) for Cited1 genotype: Wild type (⫹m/ Y, ⫹p/⫹m)

Null (neo/Y)

10.84 ⫾ 0.19 3.86 ⫾ 0.09 6.74 ⫾ 0.14

10.61 ⫾ 0.18 4.35 ⫾ 0.09 5.94 ⫾ 0.12

P

0.4151 0.0004 ⬍0.0001

a The cross-sectional area was measured from placentas that had been cryosectioned and hybridized with spongiotrophoblast gene marker 4311 (see Materials and Methods). Mean areas are presented with standard error of the mean. Analysis of variance showed that wild-type (Cited1⫹p/⫹m, Cited1⫹m/Y) and null (Cited1neo/Y) placentas were always significantly different from each other, except when the total placental area was compared.

diameter of the fetal capillaries was determined. Those filled from the venous side were most similar to the controls, while those filled from the arterial side showed a small but significant increase in diameter (Fig. 8 G and H; Table 5). The resin casts of wild-type placentas show that the maternal sinusoids and fetal capillaries are approximately the same size (Fig. 7G and 8E and G). This presumably allows the efficient exchange of gases and nutrients to occur between maternal and fetal blood. In contrast, the surface area available for exchange in Cited1 mutants between the fetus and mother is greatly reduced due to the increased size of the maternal sinusoids (Fig. 7H and I and 8F and H). Cited1 is not required for primordial germ cell production or migration. In an effort to further analyze the Cited1 null phenotype, we examined other tissues in which Cited1 is expressed. Expression is detected in extraembryonic ectoderm adjacent to embryonic ectoderm and posterior to the primitive streak (17). These sites are significant with respect to the induction and collection of primordial germ cells (PGCs) respectively. PGCs are induced in the proximal embryonic ectoderm (6.5 days p.c.) due to inductive signals (BMP4 derived) from the adjacent extraembryonic ectoderm. Once induced, they move through the primitive streak (30–32, 67). Having passed through the primitive streak, they reside in a cluster posterior to the streak, in a position that will later become the base of the allantois (7.2 days p.c.) (23, 46). It is at this stage that PGCs can be identified as large cells possessing high levels of alkaline phosphatase (15, 23). The PGCs continue to express alkaline phosphatase as they proliferate and populate the developing hindgut (8.5 to 10.5 days p.c.) and as they migrate along the dorsal mesentery into the paired gonadal primordia (10.5 to 13.5 days p.c.) (39). Due to the coincidence of Cited1 expression with the site of induction and subsequent location of PGCs (between 6.5 and 7.5 days p.c.) and the ability of Cited1 to bind Smad4 (required for transforming growth factor ␤ [TGF-␤] and bone morphogenetic protein signaling) (5, 61),

237

we examined PGCs in embryos carrying the Cited1neo allele at 8.5 days p.c. Large and readily discernible PGCs were resident in the hindgut endoderm of Cited1⫹/Y, Cited1⫹/neo, and Cited1neo/neo embryos (Fig. 9). Although variation in the number of PGCs consistent with previous reports was observed (30, 69), there was no correlation between PGC number and the presence of the Cited1neo allele. These findings demonstrate that Cited1 is not required for the induction or maintenance of PGCs in mouse embryos up to day 8.5 p.c. DISCUSSION Cited1 gene function is affected by genetic background and modifier loci. The Cited1 locus, like many other loci, is subject to modification by elements within the genome that are yet to be determined (42). This is clear since penetrance of the phenotype (postnatal lethality) is influenced by the genetic background. For example, on a mixed background (129Ola-C57BL/ 6-C57BL/10CBA) there was a 20% reduction in Cited1 null mice compared to the wild type, while on a background derived largely from C57BL/6 (produced following six crosses back to C57BL/6), there was a 78% reduction in Cited1 null mice compared to the wild type (Tables 1 and 2). Cited1 is required for normal development of the labyrinthine layer. The Cited1 mutant placenta has an increased number of spongiotrophoblasts that penetrate into the labyrinthine layer, and late in gestation the maternal blood sinusoids are significantly larger than controls. In the mouse, the labyrinthine layer of the placenta consists of fetal blood vessels and maternal blood sinusoids enmeshed in trophoblasts. This layer begins to function as a nutrient transport unit from about day 10.5 p.c. It is here that gas, nutrient, and waste exchange occurs between the mother and fetus, and so abnormal development or function of the labyrinth results in impaired fetal development. Mutational analysis in the mouse has revealed that numerous genes are required for normal formation of the labyrinthine layer (25, 53). In the majority of cases, labyrinth formation is very limited and embryonic death occurs by day 10.5 p.c.; this correlates well with the time when the labyrinth begins to function as a nutrient transport unit (22). In some mutants, development of the labyrinthine layer is relatively extensive, but fetal death occurs later in gestation (HGF; days 13.5 to 15.5 p.c.) or during the perinatal period (Pdgfra, Pdgfb, Lifr, Wnt2, Esx1, and p185/Cul7) (4, 35, 41, 44, 45, 70, 71). The Cited1 mutant placenta associates most closely with this latter group since the labyrinth is well developed and death is postnatal. However, Cited1 mutants develop dilated maternal sinusoids, in contrast to fetal vessel dilation (Pdgfra, Pdgfb, Esx1, and p185/Cul7) and maternal vascular lesions (Lifr and Wnt2). This indicates that the Cited1 mutant placental phenotype is unique among those studied.

FIG. 5. Placental structure is affected by the loss of Cited1. RNA in situ hybridization of serial placental sections on day 14.5 p.c. shows the localization of Mest (A to D), 4311 (E to H), and PLF (I to L) transcripts in Cited1⫹m/Y (A, C, E, G, I, and K) and Cited1neo/Y (B, D, F, H, J, and L) mice. Panel C and D (higher magnifications of panels A and B) show the extent of fetal vasculature in the labyrinthine layer, with arrows indicating where the vasculature is interrupted. In panels G and H and panels K and L (higher magnifications of panels E and F and panels I and J, respectively), arrows show where the fetal vasculature is interrupted by spongiotrophoblasts expressing 4311 and PLF. Bar, 1.4 mm (A, B, E, F, I, and J) and 454 ␮m (C, D, G, H, K, and L).

238

RODRIGUEZ ET AL.

MOL. CELL. BIOL.

FIG. 6. Maternal blood sinusoids are enlarged in the absence of Cited1. (A to F) Alkaline phosphatase-expressing trophoblasts surround maternal blood sinusoids in the labyrinthine layer of Cited1⫹m/Y (A and B), Cited1⫹p/neo (C and D), and Cited1Y/neo (E and F) on day 16.5 p.c. Bar, 1.26 mm (A, C, and E) and 100 ␮m (B, D, and F). (G) Length around sections of maternal blood sinusoids plotted against genotype shows that the sinusoids are larger in mutants (Cited1⫹p/neo and Cited1Y/neo) than in the control (Cited1⫹m/Y).

The mouse placenta, like that of humans, is hemochorial; maternal blood is in direct contact with placental trophoblasts (72). The sinusoid spaces, which are filled with maternal blood, are formed as the chorioallantoic interface branches. The transcription factor Gcm1 is required for trophoblast branching; in its absence, the labyrinthine layer does not form and the maternal sinusoids are large (1, 3, 56). It is unlikely that the large maternal sinusoids, apparent in Cited1 mutants, are associated

with impaired Gcm1 function because the Cited1 mutant phenotype develops for up to 7 days after the Gcm1 phenotype becomes apparent. If Gcm1 is involved, Cited1 must be required late in gestation for Gcm1 expression or function. Since Cited1 mutants have an enlarged spongiotrophoblast layer and projections of these cells extend into the labyrinthine layer, it is possible that the physical presence of spongiotrophoblast projections into the labyrinthine layer interferes with tropho-

TABLE 4. Morphometric analysis of maternal blood sinusoids in placentas on day 14.5 p.c. following staining for alkaline phosphatase activitya Value in Cited1 genotype Measurement

Surrounding length (␮m) Cross sectional area (␮m2)

Wild type (⫹m/Y)

Heterozygous (⫹p/neo)

Null (neo/Y)

52.22 ⫾ 1.86 243.94 ⫾ 14.22

84.69 ⫾ 5.16 639.03 ⫾ 64.17

63.06 ⫾ 2.53 389.65 ⫾ 26.49

P

⬍0.0001 ⬍0.0001

a The surrounding length and cross-sectional area of maternal blood sinusoids was quantified from placental sections on day 14.5 p.c. (see Materials and Methods). Mean areas are presented with standard error of the mean. These data were not normally distributed, either before or following transformation, and so the nonparametric Wilcoxon test was used to compare these data. The surrounding length and cross-sectional area of maternal sinusoid sections were significantly different in Cited1 mutants (Cited1⫹p/neo, Cited1neo/Y) compared with controls (Cited1⫹m/Y).

VOL. 24, 2004


239

FIG. 7. Resin casts of maternal blood spaces reveal enlarged sinusoids in the absence of Cited1. (A to C) Light micrographs show the lateral view of resin casts filled from the arterial side. (D to I) Scanning electron micrographs of the placental base (D to F) and sinusoids (G to I). Cited1⫹/⫹ (A, D, and G), Cited1⫹p/neo (B, E, and H), and Cited1Y/neo (C, F, and I) at day 16.5 p.c. are shown. Bar, 1.6 mm (A to C), 1.8 mm (D to F), and 36 ␮m (G to I).

blast branching and hence with the size and shape of the maternal blood sinusoids. This potentially also explains why the overall shape of the labyrinthine layer in the maternal side resin casts is highly irregular. Alternatively, the spongiotrophoblast projections, rather than having a physical impact on labyrinthine development, may secrete a factor(s) that influences trophoblast behavior. Clues to how this phenotype develops may come from examining the known molecular interactions of Cited1. Cited1 is a transcriptional cofactor that associates with Smad4, estrogen receptors ␣ and ␤ (ER␣/␤), and TFAP2 (12, 61, 74). Consequently, the number of genes whose expression could be affected in Cited1 mutants is potentially very large. Some of these genes include those encoding TGF-␤1, TGF-␤3, activin A, inhibin, follistatin and nodal (dependent on Smad4), TGF-␣ (dependent on ER␣/␤), and adenosine deaminase and 3␤-hydroxysteroid dehydrogenase VI (dependent on TFAP2) (36, 48, 51, 57). These factors have opposing effects on key processes during placental development: trophoblast proliferation and differentiation, and the promotion and inhibition of hormone and vasoactive factor secretion (76). We examined

the expression of PLF and mPLII (which is potentially downstream of ER␣/␤), both of which are vasoactive factors, but could see no difference in the level of expression by RNA in situ hybridization. Since both of these factors are expressed by spongiotrophoblasts, a cell type overrepresented in Cited1 mutants, it is possible that absolute levels of PLF and mPLII are increased. An in situ approach is not the most efficient way to show the effect of the loss of Cited1 on placental gene expression; microarray analysis provides a broader and more quantifiable result. Effects of X chromosome inactivation on the Cited1 allele and embryo viability. The X chromosome in placental mammals is subject to a unique system of developmental regulation, involving the coordinate activation and inactivation of the chromosome during female development. This mechanism has evolved to make the X-linked gene dosage equivalent in males and females. X chromosome inactivation is random in cells derived from the inner cell mass. Therefore, all cells of the embryo, plus the mesodermal components of the extraembryonic tissues, have either the paternally or maternally derived X chromosome inactivated. This occurs approximately on day

240

RODRIGUEZ ET AL.

MOL. CELL. BIOL.


VOL. 24, 2004

TABLE 5. Quantification of the capillary diameter from fetal vascular casts on day 16.5 p.c.a Capillary diameter (␮m) for Cited1 genotype Side filled

Arterial Venous

P

Wild-type (⫹m/Y, ⫹p/ ⫹m )

Null (neo/Y)

9.66 ⫹ 0.11 10.43 ⫹ 0.17

10.677 ⫹ 0.12 10.82 ⫹ 0.13

⬍0.0001 0.0401

a The diameter of resin casts of fetal capillaries filled from the arterial and venous sides on day 16.5 p.c. was determined (see Materials and Methods). Mean diameters are presented with standard error of the mean. Nested analysis of variance showed that the diameters of capillaries filled from the arterial side were significantly different between wild-type (Cited1⫹p/⫹m, Cited1⫹m/Y) and null (Cited1neo/Y) mice.

6 p.c. (40, 50), although it is not complete in some somatic tissues until day 11.5 p.c. (68). In contrast, X inactivation is nonrandom in cells derived from the trophectoderm and occurs much earlier, between days 3.5 and 4.5 p.c. (19, 64–66). Thus, in the ectodermal and endodermal components of the extraembryonic tissues, only the paternal X is inactive. Cited1 is located on the X chromosome; this allowed us to determine, using heterozygous females in which the Cited1 wild-type allele was derived from the father (Cited1⫹p/neo), that the function of Cited1 function in trophectoderm-derived components of the conceptus is essential for survival. Cited1 is expressed in several of these components: visceral/yolk sac endoderm, extraembryonic ectoderm, and the placenta (17) (Fig. 3). Despite the expression in various extraembryonic tissues, we consider that embryo survival requires Cited1 in the placenta, since defects in this organ are apparent late in gestation and at a time when Cited1 expression in other trophectoderm-derived tissues would be expected to have no effect. The labyrinthine layer of the placenta is composed of cells with distinct origins: those derived from the trophectoderm and fetal blood vessels derived from the embryo. The trophectoderm-derived cells (chorionic trophoblasts) normally express Cited1, but in Cited1⫹p/neo females the paternally derived wildtype Cited1 allele is inactive; therefore, despite the presence of a wild-type Cited1 allele, we see a defect in the size of the trophoblast-lined maternal sinusoids. In addition to the parent-of-origin effects in extraembryonic cells, heterozygous females are mosaic, with random X inactivation causing approximately half of the somatic cells to lack Cited1 expression. However, this does not appear to affect embryo viability since no detectable loss of heterozygous females occurs when the wild-type allele is maternally derived (Table 1). Cited1 is not the only X chromosome-linked gene to affect placental development. Abnormal placental growth is observed when crosses and backcrosses between different mouse (Mus) species are performed. Placental dysplasia is X chromo-

241

some linked and is likely to be due to multiple loci (78). The main placental cell type to be affected in these dysplastic placentas is the spongiotrophoblast (52, 78). On the X chromosome, the genetic interval associated with hyperplastic placentas is large (between the centromere and 50 cM) (26). Esx1 (57 cM) may represent one of the genes associated with this phenotype since Esx1 mutants have a hyperplastic placenta and the expression of Esx1 is reduced in the placentas of interspecific hybrids (35, 77). It would be interesting to examine the expression of Cited1 in the placentas of interspecific hybrids since the gene is located at 40.1 cM and there is an increase in spongiotrophoblasts in the Cited1 mutant placenta. Could the function of Cited1 in the embryo be masked by functional redundancy? The only phenotypic effect of the Cited1 loss-of-function mutation reported here is a defect in the placenta; this is despite the fact that Cited1 is expressed in a number of tissues during mouse embryo development (17). It is possible that functional redundancy is masking other roles of Cited1; this can occur if there is redundancy at the genetic pathway level or if a single gene can substitute for Cited1 function. It is possible that other Cited genes can functionally compensate for the loss of Cited1 during embryonic development. Cited1, Cited2, and Cited4 are each expressed during embryonic development in the mouse (17, 37, 75). In the embryo, both Cited1 and Cited2 are expressed in nascent mesoderm, myocardium, cranial neural crest cells, presomitic mesoderm, and somites. No overlap in expression of Cited1 and Cited4 has been identified in the embryo; however, these genes are both expressed in adult mammary epithelial cells (74, 75). Therefore, it is possible that the Cited1 loss-of-function mutation results in embryonic defects but that compensatory effects of Cited2 mask these. This would require that they play similar molecular roles during embryonic development, a proposition supported by the observations that both Cited1 and Cited2 bind CBP/p300, ER␣/␤, and TFAP2 (6, 11, 73, 74). Generation of compound mutant embryos that lack both Cited1 and Cited2 could resolve this issue. In summary, the X-linked gene Cited1 is required in the mouse for normal placental development and for embryonic growth and viability. Cited1 mutants develop large maternal blood sinusoids, a phenotype consistent with its expression in trophoblasts including those that line these sinusoids. The Cited1 mutant mouse line that we have generated will allow further dissection of the function of this gene during embryonic and adult life. Importantly, it may prove to be a model for intrauterine growth restriction (IUGR) in humans. IUGR is a significant cause of infant mortality and morbidity, and it has been suggested that infants with IUGR exhibit higher rates of coronary heart disease, type 2 diabetes, hypertension, and stroke as adults (7–10, 13, 49). Therefore, fetal growth not only impacts the outcome of the perinatal period but also may affect adult well-being. The etiologies of IUGR are numerous but are

FIG. 8. Resin casts reveal the arrangement of the fetal vasculature. Light micrographs of Cited1⫹/Y (A, C, E, and G) and Cited1neo/Y (B, D, F, and H) showing the base (A and B) and side (C and D) of casts filled with red resin from the venous side and blue resin from the arterial side on day 16.5 p.c. Scanning electron micrographs of capillaries filled from the arterial side, at the top of the labyrinthine layer having recently emerged from the arterioles (E and F) and at the tips of the casts where the capillaries are projecting into the center of the labyrinthine layer (G and H), are shown. Bar, 1.25 mm (A and B), 1.2 mm (C and D), and 40 ␮m (E to H).

242

RODRIGUEZ ET AL.

MOL. CELL. BIOL.

often associated with abnormalities in placental structure and function. Since it is difficult to fully assess the role of genes during the development of placental insufficiency in the human, there is a need for the development and analysis of animal models. The Cited1 mutant mouse line reported here may serve as a useful model for placental insufficiency and thus contribute to our understanding of the etiology of this disorder. ACKNOWLEDGMENTS We are extremely grateful to Austin Smith for the 129/Olac genomic library and the E14.2 ES cells; Margaret Budanovic from the University of New South Wales Electron Microscopy Unit; James Cross and Michelle Tallquist for methods; Janet Rossant, Reinald Fundele, and Daniel Linzer for reagents; Stuart Gilchrist for help with the statistical analyses; Christine Biben for critical assessment of the manuscript; and unknown reviewers for their very helpful comments. Animal work complied with all relevant governmental and institutional policies. D.B.S. is a Westfield-Belconnen Postdoctoral Fellow. S.L.W. is a Royal Society of London International Postdoctoral Fellow and a Wellcome Trust International Travelling Research Fellow. T.E.T. is an NHMRC Peter Doherty Postdoctoral Fellow. S.L.D. is a Pharmacia Foundation of Australia Senior Research Fellow. T.S. is supported by the U.S. National Cancer Institute (R01-CA82230) and the AVON Project on Breast Cancer Research. REFERENCES

FIG. 9. Primordial germ cells are detected in mice carrying the Cited1⫹/neo allele. (A to C) Cross sections through the caudal region of Cited1⫹/Y (A), Cited1⫹/neo (B), and Cited1neo/neo (C) embryos with seven or eight somites are shown. Primordial germ cells are present (arrow) in the gut endoderm (ge). The dorsal aorta (da), neural tube (nt), and lateral plate mesoderm (1pm) are also shown. (G) A scatter plot shows the total number of primordial germ cells present in embryos with somite numbers between zero and eight. Cited1⫹/Y (squares), Cited1⫹/neo (circles), and Cited1neo/neo Cited1neo/Y (solid squares). Bar, 50 ␮m.

1. Adamson, S., Y. Lu, K. Whiteley, D. Holmyard, M. Hemberger, C. Pfarrer, and J. Cross. 2002. Interactions between trophoblast cells and the maternal and fetal circulation in the mouse placenta. Dev. Biol. 250:358–373. 2. Andrews, J. E., M. J. O’Neill, M. Binder, T. Shioda, and A. H. Sinclair. 2000. Isolation and expression of a novel member of the CITED family. Mech. Dev. 95:305–308. 3. Anson-Cartwright, L., K. Dawson, D. Holmyard, S. J. Fisher, R. A. Lazzarini, and J. C. Cross. 2000. The glial cells missing-1 protein is essential for branching morphogenesis in the chorioallantoic placenta. Nat. Genet. 25: 311–314. 4. Arai, T., J. S. Kasper, J. R. Skaar, S. H. Ali, C. Takahashi, and J. A. DeCaprio. 2003. Targeted disruption of p185/Cul7 gene results in abnormal vascular morphogenesis. Proc. Natl. Acad. Sci. USA 100:9855–9860. 5. Attisano, L., and S. Tuen Lee-Hoeflich. 2001. The Smads. Genome Biol. 2:3010.1–3010.8 6. Bamforth, S. D., J. Braganca, J. J. Eloranta, J. N. Murdoch, F. I. Marques, K. R. Kranc, H. Farza, D. J. Henderson, H. C. Hurst, and S. Bhattacharya. 2001. Cardiac malformations, adrenal agenesis, neural crest defects and exencephaly in mice lacking Cited2, a new Tfap2 co-activator. Nat. Genet. 29:469–474. 7. Barker, D. J., A. R. Bull, C. Osmond, and S. J. Simmonds. 1990. Fetal and placental size and risk of hypertension in adult life. Br. Med. J. 301:259–262. 8. Barker, D. J., P. D. Gluckman, K. M. Godfrey, J. E. Harding, J. A. Owens, and J. S. Robinson. 1993. Fetal nutrition and cardiovascular disease in adult life. Lancet 341:938–941. 9. Barker, D. J., C. N. Hales, C. H. Fall, C. Osmond, K. Phipps, and P. M. Clark. 1993. Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia 36:62–67. 10. Barker, D. J., C. Osmond, J. Golding, D. Kuh, and M. E. Wadsworth. 1989. Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease. Br. Med. J. 298:564–567. 11. Bhattacharya, S., C. L. Michels, M. K. Leung, Z. P. Arany, A. L. Kung, and D. M. Livingston. 1999. Functional role of p35srj, a novel p300/CBP binding protein, during transactivation by HIF-1. Genes Dev. 13:64–75. 12. Braganca, J., T. Swingler, F. I. Marques, T. Jones, J. J. Eloranta, H. C. Hurst, T. Shioda, and S. Bhattacharya. 2002. Human CREB-binding protein/p300-interacting transactivator with ED-rich tail (CITED) 4, a new member of the CITED family, functions as a coactivator for transcription factor AP-2. J. Biol. Chem. 277:8559–8565. 13. Brar, H. S., and S. E. Rutherford. 1988. Classification of intrauterine growth retardation. Semin. Perinatol. 12:2–10. 14. Carney, E. W., V. Prideaux, S. J. Lye, and J. Rossant. 1993. Progressive expression of trophoblast-specific genes during formation of mouse trophoblast giant cells in vitro. Mol. Reprod. Dev. 34:357–368. 15. Chiquoine, A. D. 1954. The identification, origin and migration of primordial germ cells in the mouse embryo. Anat. Rec. 118:135–146. 16. Dunwoodie, S. L., D. Henrique, S. M. Harrison, and R. S. Beddington. 1997.

VOL. 24, 2004

17. 18. 19. 20.

21. 22. 23. 24. 25. 26.

27. 28.

29. 30.

31. 32. 33. 34.

35. 36. 37.

38.

39. 40. 41. 42. 43.


Mouse DII3: a novel divergent Delta gene which may complement the function of other Delta homologues during early pattern formation in the mouse embryo. Development 124:3065–3076. Dunwoodie, S. L., T. A. Rodriguez, and R. S. Beddington. 1998. Msg1 and Mrg1, founding members of a gene family, show distinct patterns of gene expression during mouse embryogenesis. Mech. Dev. 72:27–40. Dymecki, S. M. 1996. Flp recombinase promotes site-specific DNA recombination in embryonic stem cells and transgenic mice. Proc. Natl. Acad. Sci. USA 93:6191–6196. Epstein, C. J., S. Smith, B. Travis, and G. Tucker. 1978. Both X chromosomes function before visible X-chromosome inactivation in female mouse embryos. Nature 274:500–503. Fenner, M. H., J. E. Parrish, Y. Boyd, V. Reed, M. MacDonald, D. L. Nelson, K. J. Isselbacher, and T. Shioda. 1998. MSG1 (melanocyte-specific gene 1): mapping to chromosome Xq13.1, genomic organization, and promoter analysis. Genomics 51:401–407. Frels, W. I., J. Rossant, and V. M. Chapman. 1979. Maternal X chromosome expression in mouse chorionic ectoderm. Dev. Genet. 1:123–132. Georgiades, P., A. C. Ferguson-Smith, and G. J. Burton. 2002. Comparative developmental anatomy of the murine and human definitive placentae. Placenta 23:3–19. Ginsburg, M., M. H. Snow, and A. McLaren. 1990. Primordial germ cells in the mouse embryo during gastrulation. Development 110:521–528. Glenn, D. J., and R. A. Maurer. 1999. MRG1 binds to the LIM domain of Lhx2 and may function as a coactivator to stimulate glycoprotein hormone alpha-subunit gene expression. J. Biol. Chem. 274:36159–36167. Hemberger, M., and J. C. Cross. 2001. Genes governing placental development. Trends Endocrinol. Metab. 12:162–168. Hemberger, M. C., R. S. Pearsall, U. Zechner, A. Orth, S. Otto, F. Ruschendorf, R. Fundele, and R. Elliott. 1999. Genetic dissection of X-linked interspecific hybrid placental dysplasia in congenic mouse strains. Genetics 153: 383–390. Hogan, B., R. Beddington, F. Costantini, and E. Lacy. 1994. Manipulating the mouse embryo. A laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. Huang, Y., M. Prasad, W. J. Lemon, H. Hampel, F. A. Wright, K. Kornacker, V. LiVolsi, W. Frankel, R. T. Kloos, C. Eng, N. S. Pellegata, and A. de la Chapelle. 2001. Gene expression in papillary thyroid carcinoma reveals highly consistent profiles. Proc. Natl. Acad. Sci. USA 98:15044–15049. Kaufman, M. H. 1992. The atlas of mouse development. Academic Press, Ltd., London, United Kingdom. Lawson, K. A., N. R. Dunn, B. A. Roelen, L. M. Zeinstra, A. M. Davis, C. V. Wright, J. P. Korving, and B. L. Hogan. 1999. Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev. 13:424–436. Lawson, K. A., and W. J. Hage. 1994. Clonal analysis of the origin of primordial germ cells in the mouse. Ciba Found. Symp. 182:68–84. Lawson, K. A., J. J. Meneses, and R. A. Pedersen. 1991. Clonal analysis of epiblast fate during germ layer formation in the mouse embryo. Development 113:891–911. Lee, S. J., F. Talamantes, E. Wilder, D. I. Linzer, and D. Nathans. 1988. Trophoblastic giant cells of the mouse placenta as the site of proliferin synthesis. Endocrinology 122:1761–1768. Li, H., N. U. Ahmed, M. H. Fenner, M. Ueda, K. J. Isselbacher, and T. Shioda. 1998. Regulation of expression of MSG1 melanocyte-specific nuclear protein in human melanocytes and melanoma cells. Exp. Cell Res. 242:478–486. Li, Y., and R. R. Behringer. 1998. Esx1 is an X-chromosome-imprinted regulator of placental development and fetal growth. Nat. Genet. 20:309– 311. Ma, G. T., V. Soloveva, S. J. Tzeng, L. A. Lowe, K. C. Pfendler, P. M. Iannaccone, M. R. Kuehn, and D. I. Linzer. 2001. Nodal regulates trophoblast differentiation and placental development. Dev. Biol. 236:124–135. Martinez-Barbera, J. P., T. A. Rodriguez, N. D. Greene, W. J. Weninger, A. Simeone, A. J. Copp, R. S. Beddington, and S. Dunwoodie. 2002. Folic acid prevents exencephaly in Cited2 deficient mice. Hum. Mol. Genet. 11:283– 293. Mayer, W., M. Hemberger, H. G. Frank, R. Grummer, E. Winterhager, P. Kaufmann, and R. Fundele. 2000. Expression of the imprinted genes MEST/ Mest in human and murine placenta suggests a role in angiogenesis. Dev. Dyn. 217:1–10. Mintz, B., and E. S. Russell. 1957. Gene-induced embryological modifications of primordial germ cells in mouse. J. Exp. Zool. 134:207–230. Monk, M., and M. I. Harper. 1979. Sequential X chromosome inactivation coupled with cellular differentiation in early mouse embryos. Nature 281: 311–313. Monkley, S. J., S. J. Delaney, D. J. Pennisi, J. H. Christiansen, and B. J. Wainwright. 1996. Targeted disruption of the Wnt2 gene results in placentation defects. Development 122:3343–3353. Muller, U. 1999. Ten years of gene targeting: targeted mouse mutants, from vector design to phenotype analysis. Mech. Dev. 82:3–21. O’Gorman, S., N. A. Dagenais, M. Qian, and Y. Marchuk. 1997. Protamine-

44.

45.

46. 47. 48.

49. 50. 51.

52. 53. 54.

55. 56.

57. 58. 59. 60. 61.

62. 63. 64. 65. 66. 67. 68. 69.

243

Cre recombinase transgenes efficiently recombine target sequences in the male germ line of mice, but not in embryonic stem cells. Proc. Natl. Acad. Sci. USA 94:14602–14607. Ogura, Y., N. Takakura, H. Yoshida, and S. I. Nishikawa. 1998. Essential role of platelet-derived growth factor receptor alpha in the development of the intraplacental yolk sac/sinus of Duval in mouse placenta. Biol. Reprod. 58:65–72. Ohlsson, R., P. Falck, M. Hellstrom, P. Lindahl, H. Bostrom, G. Franklin, L. Ahrlund-Richter, J. Pollard, P. Soriano, and C. Betsholtz. 1999. PDGFB regulates the development of the labyrinthine layer of the mouse fetal placenta. Dev. Biol. 12:124–136. Ordzenski, W. 1967. Observations on the origin of primordial germ cells in the mouse. Zool. Pol. 17:367–379. Papaioannou, V. E., D. D. West, T. Bucher, and I. M. Linke. 1981. Nonrandom X-chromosome expression early in mouse development. Dev. Genet. 2:305–315. Peng, L., and A. H. Payne. 2002. AP-2 gamma and the homeodomain protein distal-less 3 are required for placental-specific expression of the murine 3 beta-hydroxysteroid dehydrogenase VI gene, Hsd3b6. J. Biol. Chem. 277: 7945–7954. Pollack, R. N., and M. Y. Divon. 1992. Intrauterine growth retardation: definition, classification, and etiology. Clin. Obstet. Gynecol. 35:99–107. Rastan, S. 1982. Timing of X-chromosome inactivation in postimplantation mouse embryos. J. Embryol. Exp. Morphol. 71:11–24. Roberts, H. J., S. Hu, Q. Qiu, P. C. Leung, I. Caniggia, A. Gruslin, B. Tsang, and C. Peng. 2003. Identification of novel isoforms of activin receptor-like kinase 7 (ALK7) generated by alternative splicing and expression of ALK7 and its ligand, Nodal, in human placenta. Biol. Reprod. 68:1719–1726. Rogers, J. F., and W. D. Dawson. 1970. Foetal and placental size in a Peromyscus species cross. J. Reprod. Fertil. 21:255–262. Rossant, J., and J. C. Cross. 2001. Placental development: lessons from mouse mutants. Nat. Rev. Genet. 2:538–48. Sado, T., M. H. Fenner, S. S. Tan, P. Tam, T. Shioda, and E. Li. 2000. X inactivation in the mouse embryo deficient for Dnmt1: distinct effect of hypomethylation on imprinted and random X inactivation. Dev. Biol. 225: 294–303. Schlange, T., B. Andree, H. Arnold, and T. Brand. 2000. Expression analysis of the chicken homologue of CITED2 during early stages of embryonic development. Mech. Dev. 98:157–160. Schreiber, J., E. Riethmacher-Sonnenberg, D. Riethmacher, E. E. Tuerk, J. Enderich, M. R. Bosl, and M. Wegner. 2000. Placental failure in mice lacking the mammalian homolog of glial cells missing, GCMa. Mol. Cell. Biol. 20:2466–2474. Shi, D., and R. E. Kellems. 1998. Transcription factor AP-2gamma regulates murine adenosine deaminase gene expression during placental development. J. Biol. Chem. 273:27331–27338. Shida, M. M., L. L. Jackson-Grusby, S. R. Ross, and D. I. H. Linzer. 1992. PLacental-specific expression from the mouse placental lactogen II gene promoter. Proc. Natl. Acad. Sci. USA 89:3864–3868. Shioda, T., M. H. Fenner, and K. J. Isselbacher. 1997. MSG1 and its related protein MRG1 share a transcription activating domain. Gene 204:235–241. Shioda, T., M. H. Fenner, and K. J. Isselbacher. 1996. msg1, a novel melanocyte-specific gene, encodes a nuclear protein and is associated with pigmentation. Proc. Natl. Acad. Sci. USA 93:12298–12303. Shioda, T., R. J. Lechleider, S. L. Dunwoodie, H. Li, T. Yahata, M. P. de Caestecker, M. H. Fenner, A. B. Roberts, and K. J. Isselbacher. 1998. Transcriptional activating activity of Smad4: roles of SMAD hetero- oligomerization and enhancement by an associating transactivator. Proc. Natl. Acad. Sci. USA 95:9785–9790. Skarnes, W. C., J. E. Moss, S. M. Hurtley, and R. S. Beddington. 1995. Capturing genes encoding membrane and secreted proteins important for mouse development. Proc. Natl. Acad. Sci. USA 92:6592–6596. Sturm, K. S., C. N. Berger, S. X. Zhou, S. L. Dunwoodie, S.-S. Tan, and P. P. L. Tam. 1997. Unrestricted lineage differentiation of parthenogenetic ES cells. Dev. Genes Evol. 206:377–388. Takagi, N. 1974. Differentiation of X chromosomes in early female mouse embryos. Exp. Cell Res. 86:127–135. Takagi, N., and M. Sasaki. 1975. Preferential inactivation of the paternally derived X chromosome in the extraembryonic membranes of the mouse. Nature 256:640–642. Takagi, N., O. Sugawara, and M. Sasaki. 1982. Regional and temporal changes in the pattern of X-chromosome replication during the early postimplantation development of the female mouse. Chromosoma 85:275–286. Tam, P. P., and S. X. Zhou. 1996. The allocation of epiblast cells to ectodermal and germ-line lineages is influenced by the position of the cells in the gastrulating mouse embryo. Dev. Biol. 178:124–132. Tan, S. S., E. A. Williams, and P. P. Tam. 1993. X-chromosome inactivation occurs at different times in different tissues of the post-implantation mouse embryo. Nat. Genet. 3:170–174. Tsang, T. E., P. L. Khoo, R. V. Jamieson, S. X. Zhou, S. L. Ang, R. Behringer, and P. P. Tam. 2001. The allocation and differentiation of mouse primordial germ cells. Int. J. Dev. Biol. 45:549–555.

244

RODRIGUEZ ET AL.

70. Uehara, Y., O. Minowa, C. Mori, K. Shiota, J. Kuno, T. Noda, and N. Kitamura. 1995. Placental defect and embryonic lethality in mice lacking hepatocyte growth factor/scatter factor. Nature 373:702–705. 71. Ware, C. B., M. C. Horowitz, B. R. Renshaw, J. S. Hunt, D. Liggitt, S. A. Koblar, B. C. Gliniak, H. J. McKenna, T. Papayannopoulou, B. Thoma, L. Cheng, P. J. Donovan, J. J. Peschon, P. F. Bartlett, C. R. Willis, B. D. Wright, M. K. Carpenter, B. L. Davison, and D. P. Gearing. 1995. Targeted disruption of the low-affinity leukemia inhibitory factor receptor gene causes placental, skeletal, neural and metabolic defects and results in perinatal death. Development 121:1283–1299. 72. Wooding, F. B. P., and A. P. F. Flint. 1994. Placentation, p. 233–460. In G. E. Lamming (ed.), Marshall’s physiology of reproduction, 4th ed. Chapman & Hall, New York, N.Y. 73. Yahata, T., M. P. de Caestecker, R. J. Lechleider, S. Andriole, A. B. Roberts, K. J. Isselbacher, and T. Shioda. 2000. The MSG1 non-DNA-binding transactivator binds to the p300/CBP coactivators, enhancing their functional link to the Smad transcription factors. J. Biol. Chem. 275:8825–8834. 74. Yahata, T., W. Shao, H. Endoh, J. Hur, K. R. Coser, H. Sun, Y. Ueda, S. Kato, K. J. Isselbacher, M. Brown, and T. Shioda. 2001. Selective coactivation of estrogen-dependent transcription by CITED1 CBP/p300-binding protein. Genes Dev. 15:2598–2612.

MOL. CELL. BIOL. 75. Yahata, T., H. Takedatsu, S. L. Dunwoodie, J. Braganc¸a, T. Swingler, S. L. Withington, J. Hur, K. R. Coser, K. J. Isselbacher, S. Bhattacharya, and T. Shioda. 2002. Cloning of mouse Cited4, a member of the CITED family p300/CBP-binding transcriptional coactivators: induced expression in mammary epithelial cells. Genomics 80:601–613. 76. Yamaguchi, M., L. Ogren, H. Kurachi, K. Hirota, T. Imai, and F. Talamantes. 1995. Opposite effects of transforming growth factor alpha and epidermal growth factor on mouse placental lactogen I secretion. Proc. Natl. Acad. Sci. USA 92:2830–2834. 77. Zechner, U., M. Hemberger, M. Constancia, A. Orth, I. Dragatsis, A. Luttges, H. Hameister, and R. Fundele. 2002. Proliferation and growth factor expression in abnormally enlarged placentas of mouse interspecific hybrids. Dev. Dyn. 224:125–134. 78. Zechner, U., M. Reule, A. Orth, F. Bonhomme, B. Strack, J.-L. Guenet, H. Hameister, and R. Fundele. 1996. An X-chromosome linked locus contributes to abnormal placental development in mouse interspecific hybrid. Nat. Genet. 12:398–403. 79. Zernicka-Goetz, M., J. Pines, S. McLean Hunter, J. P. Dixon, K. R. Siemering, J. Haseloff, and M. J. Evans. 1997. Following cell fate in the living mouse embryo. Development 124:1133–1137.