| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Institute of Reproductive and Developmental Biology (C.L., J.H.S., M.G.P.) and Department of Histopathology (B.S.-D., E.-N.L.), Imperial College, London W12 ONN, United Kingdom; and Cancer Research UK (S.C., I.R.), Clare Hall Laboratories, South Mimms, Potters Bar, Hertfordshire EN6 3LD, United Kingdom
Address all correspondence and requests for reprints to: Malcolm G. Parker, Institute of Reproductive and Developmental Biology, Imperial College, Faculty of Medicine, Du Cane Road, London W12 ONN, United Kingdom. E-mail: m.parker{at}imperial.ac.uk.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Numerous proteins have been implicated in target gene activation by NRs, and generally many of these coregulators are large in size, contain multiple activation- and receptor-interacting domains, and appear to function as components of large, multiprotein complexes (5, 6). A number of coactivator complexes have been implicated in the activation of the AF-2 domain, including the p160/steroid receptor coactivator (SRC) complex and the thyroid hormone receptor-associated protein (TRAP) complex. The p160/SRC family of coactivators includes SRC-1, transcriptional intermediary factor 2/glucocorticoid receptor interacting protein 1/SRC-2, and p300/CBP interacting protein/amplified in breast cancer 1 (AIB1)/activator of TR and RAR/receptor-associated coactivator 3/TR-activator molecule 1/SRC-3 (7), which are encoded by three distinct genes. These highly homologous proteins exhibit a common domain structure and serve as adapter molecules that recruit chromatin remodeling activities to hormone-responsive promoters. The p160/SRC coactivators contain a central receptor-interacting domain containing three copies of a consensus leucine-rich motif, LXXLL, responsible for the interaction with the ligand-bound NRs (8, 9), and two conserved activation domains, AD1 and AD2, which are responsible for the recruitment of cAMP response element binding protein (CREB)-binding protein/p300 (10, 11, 12, 13) and coactivator-associated arginine methyltransferase-1 (14), respectively. These proteins that possess histone acetyltransferase and arginine methyltransferase activities, respectively, are important in chromatin remodeling at hormone-responsive promoters.
The TRAP complex was initially identified in a screen to isolate cofactors that could interact with and modulate the activity of thyroid hormone receptor-
(TR) in the presence of thyroid hormone (15). Unlike the p160/SRC coactivator complex, the TRAP complex possesses no intrinsic histone acetyltransferase activity, but could markedly activate TR-mediated transcription in vitro on naked DNA templates (16). This suggests that the TRAP complex mediates a novel activation step distinct from those mediated by p160/SRC complexes, possibly in the recruitment of the RNA polymerase II holoenzyme complex to activated promoters. Since its discovery, a number of other transcriptional regulatory complexes that resemble the TRAP complex, including vitamin D receptor-interacting protein (17), cofactor required for Sp1 activation (18), negative regulator of activated transcription (19), activator-recruited cofactor (20), and SRB/MED- containing cofactor complex (21), have been described. These are typically large megadalton complexes, the composition of which varies from 7–18 subunits, some of which are homologous to yeast mediator and others restricted to metazoans, suggesting additional roles for mediator complexes in multicellular organisms (22, 23). Interestingly, these mediator complexes can potentiate transcriptional initiation from distinct families of transcription factors, including NRs (15, 17, 24), p53 (25, 26), Sp1 (27), sterol regulatory element-binding protein-1a, and nuclear factor 
B (NFB) (20); and Sox9 (28), suggesting that mediator complexes are multifunctional complexes that can integrate a variety of activation signals.
The TRAP220 subunit of the TRAP coactivator complex is capable of interacting directly with the ligand-binding domain of NRs by means of two centrally located LXXLL motifs in a ligand-dependent manner (29, 30). Thus TRAP220, also termed, TRIP2 (31), proliferator activator receptor-binding protein (32), RB18A (33), and vitamin D receptor-interacting protein 205 (34), in addition to binding the liganded TR, have been reported to interact with and/or enhance the ligand-dependent activity of vitamin D receptor, peroxisome proliferator-activated receptor- (PPAR), PPAR
, estrogen receptor-, estrogen receptor-ß, retinoid X receptor- (RXR), retinoic acid receptor-, androgen receptor, and hepatocyte nuclear factor 4 (24, 29, 35, 36, 37, 38, 39, 40). Collectively, these findings suggest that the TRAP coactivator complex plays a key role in the initiation of transcription by NRs.
Since its discovery, two laboratories have generated mice devoid of the TRAP220 gene, and in both cases, TRAP220 null mice were found to die at embryonic day 11.5 (E11.5). However, although these mice exhibit defects in a number of tissues, including the placenta, heart, liver, central nervous system, and eye (41, 42, 43), the precise cause of embryonic lethality has remained elusive. To extend this work and define more precisely the in vivo role of TRAP220, we have generated mice that are either completely devoid of the gene, express the gene at reduced levels (
10% normal levels), or have their trophoblast lineage replaced by tetraploid methodology (44). In this paper we report that complete ablation of the gene results in embryonic lethality at E11.5 due to impaired placental function, because supplementation of TRAP220 null embryos with wild-type placentas via aggregation with tetraploid embryos corrects placental insufficiencies. However, these rescued embryos exhibit another lethal phase and die at E13.5 due to severe hepatic necrosis, coupled with poor heart development. Collectively, these results suggest that TRAP220 expression is essential in extraembryonic tissues before E11.5, and in the embryo thereafter.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Each ES cell mutant clone was verified by Southern blotting of genomic DNA, and two independent clones were then microinjected into recipient blastocysts to generate chimeric mice. Finally, germline-transmitting chimeras were crossed to wild-type C57BL/6J mice to produce heterozygous TRAP220 mutants (Fig. 1B). All heterozygous TRAP220 mutant mice were phenotypically normal.
|
). To determine the stage of embryonic lethality, embryos recovered at different time points were genotyped from TRAP220+/con crosses. Up to E11.5, TRAP220con/con embryos were viable, morphologically indistinguishable from wild-type and heterozygous controls, and recovered at the expected Mendelian ratio of 25%. At E12.5, a heartbeat could be detected in 65% of TRAP220con/con embryos, but some had started to degenerate and displayed overall growth retardation compared with wild-type controls. By E13.5, a heartbeat could be detected in only 30% of TRAP220con/con embryos, and all displayed severe growth retardation. By E14.0, all TRAP220con/con embryos were necrotic and resorbing (Table 1 and Fig. 2, A–C).
|
|
). Genotype analysis from TRAP220+/- matings revealed TRAP220-/- mice die at E11.5. At E10.5 and earlier gestational stages, TRAP220-/- embryos were viable, externally indistinguishable from wild-type embryos in gross morphology, and recovered at the expected Mendelian ratio of 25%. At E11.0, a heartbeat could be detected in 75% of TRAP220-/- embryos, but some displayed overall growth retardation and had started to degenerate. By E12.0, all TRAP220-/- embryos were necrotic and resorbing (Table 2 and Fig. 2
, A–C).
|
, D and E). When embryos taken from TRAP220+/con intercrosses were analyzed, Northern blotting revealed that TRAP220con/con embryos expressed approximately 80% less TRAP220 mRNA than wild-type controls (Fig. 2D). Presumably, this residual 20% expression accounts for the additional survival of TRAP220con/con embryos by two additional gestational days, and furthermore, suggested that these mutants are hypomorphic, expressing both TRAP220 mRNA and protein at very low levels. This was confirmed by Western blotting, in which TRAP220 protein was only just detectable in E10.5–E12.5 TRAP220con/con embryos in comparison to wild-type and heterozygous controls, and therefore embryos homozygous for the conditional allele will now be referred to as TRAP220hyp/hyp (Fig. 2E).
Hypoplasia of the Heart in TRAP220 Homozygous Mutants
TRAP220-/- hearts were indistinguishable from wild-type siblings at E8.5 when circulation in the embryo begins. However, from E9.5 onward, the hearts of TRAP220-/- embryos showed prominent ventricular hypoplasia in comparison to wild-type controls (Fig. 3, compare A with F), similar to that previously reported (41, 42). In particular, the compact zone of the ventricular myocardium was severely affected. At E9.5, trabecular development was impaired, and at later stages a significant proportion of the TRAP220-/- endocardium was detached from the myocardium (Fig. 3G, asterisks). At E10.5 the hearts of TRAP220-/- embryos exhibited severe noncompaction of the ventricular myocardium, manifested by thin ventricular walls (Fig. 3, compare C with H, arrows). All other heart structures were developing normally in TRAP220-/- embryos; no abnormal chamber dilation was evident, endocardial cushion development was normal, and all TRAP220-/- hearts analyzed had a normal pericardial sac. In contrast, hearts of TRAP220hyp/hyp embryos appeared normal at E10.5 and earlier gestational stages. Active proliferation was noted in the myocardium and developing subendocardial cushions, whereas trabecular development of the cardiac chambers appeared normal. However, at E11.5 and beyond, embryonic development of the heart was retarded in comparison to wild-type controls by approximately 1.5 d at around the time of death at E13.5. Development of the ventricular myocardium was particularly affected at E12.5, manifested by thin ventricular walls at the apex of the heart (see Fig. 3L), although some trabecular development and peripheral compaction was evident (Fig. 3M, arrows).
|
Defective Hematopoiesis in TRAP220 Homozygous Mutants
The liver primordia of TRAP220-/- and TRAP220hyp/hyp mutants were indistinguishable from wild-type controls at E10.5, before the shifting of hematopoiesis to the fetal liver, and abundant hemangioblasts were observed in the peripheral circulation. However, while active proliferating hematopoietic cells could be detected in the liver sinusoids, mature megakaryocytes were difficult to identify in TRAP220-/- embryos, suggesting that hematopoiesis was impaired. At E11.5 in TRAP220-/- and E12.5 in TRAP220hyp/hyp embryos, the liver could be visualized with excessive amounts of nucleated erythrocytes in dilated sinusoids (Fig. 3, compare D with I and N). In addition, megakaryocytes remained immature and difficult to identify (Fig. 3, compare E with J and O), suggesting that the control of megakaryocytic and erythrocytic lineages require TRAP220.
In contrast to TRAP220-/- and TRAP220hyp/hyp embryos, heterozygous mutant embryos and adults displayed no gross or histological abnormalities in relation to hepatic development (data not shown).
Normal Ontogeny of Other Organs in TRAP220 Homozygous Mutants
Histological examination of TRAP220-/- and TRAP220hyp/hyp embryos during midgestation showed normal development in a number of other tissues including the fore-, mid-, and hindbrain, neuroepithelium where active glial differentiation was observed, and dorsal root ganglia. Ear and eye development were also normal; differentiation of the inner (neural) and outer (future retinal) layers was evident, although retarded in TRAP220hyp/hyp mutants at later stages. The metanephros appeared viable, and gut development appeared normal also (data not shown).
Expression of TRAP220 during Midgestational Development
To study the expression of TRAP220 mRNA around the point of embryonic failure, in situ hybridization was performed on paraffin sections of mouse embryos and placentas at different embryonic stages. As shown in Fig. 4, TRAP220 mRNA was widely expressed during embryonic development. At E9.5–E10.5, TRAP220 mRNA expression was low and diffuse, but particularly strong in the neural epithelium of the neural tube (Fig. 4A, Fb and inset), while other regions showed low expression. At E11.5–E12.5, TRAP220 mRNA expression was abundant throughout many embryonic tissues, being strongest in the developing liver and primitive gut, nasopharynx, and developing limb buds. Moderate expression could be detected in the brain and optic stalk, branchial arch, and urogenital ridge, whereas expression in the heart was low (Fig. 4B). By E13.5–E14.5, TRAP220 mRNA became very specific as embryonic tissues differentiated into specialized structures. TRAP220 was expressed abundantly in the optic recess and forebrain, inferior ganglion of the vagus nerve, nasopharynx including the vomeronasal organ, tongue, and lower jaw. The developing liver, dorsal root ganglia, lung, pancreas, intestine, and genital tubercle also showed strong expression, whereas moderate expression could be detected in the developing midbrain and neural tube. TRAP220 mRNA expression remained low in the heart and large blood vessels (Fig. 4C). In the developing placenta, TRAP220 mRNA expression was moderate in the giant and spongiotrophoblast cell layers and strongest in the labyrinthine portion at E11.5 (Fig. 4F), and this pattern of expression was maintained throughout E9.5–E13.5 (data not shown).
|
, compare A with E and I). At E10–E11.5, hematopoietic cells (including megakaryocytes) were observed clustered in the labyrinthine portion of both wild-type and homozygous mutants, and branching villi were observed in the spongiotrophoblast portion. Furthermore, the labyrinthine, spongiotrophoblast, and giant cell layers were all well defined and developed. At later stages of development, no gross pathological lesions were detected in the placentas or in the vascularization of TRAP220hyp/hyp mutants (data not shown).
|
2–5% giant cells, 15% spongiotrophoblast cells, and 30–40% labyrinthine cells being positive for BrdU) with no noticeable differences between each genotype (compare Fig. 5B with Figs. 5F and 5J). Similarly, by TUNEL no increase in apoptosis was detected in any TRAP220 mutant placentae between E10–E11.5 in comparison to wild-type controls (data not shown). Finally, RNA in situ hybridization was performed against placental lactogen 2 (Pl2), a marker of giant cells, and Tpbp, a marker of spongiotrophoblast cells (48) to analyze the integrity of the placenta at E10–E11.5. Expression of both genes remained unchanged in TRAP220 mutant placentae in comparison to wild-type controls (compare Fig. 5, C and D, with Fig. 5, G and H, and K and L), and furthermore, the thickness of the spongiotrophoblast and labyrinthine cell layers was similar between each TRAP220 genotype (0.2–0.3 and 0.5–0.6 mm, respectively). This suggests, therefore, that the overall formation of the placenta is normal, but differences in specific placental function cannot be discounted.
Death during midgestational development can be a common consequence of defects in extraembryonic tissues (49). To investigate the importance of TRAP220 expression in the placenta, aggregation chimeras were generated between wild-type tetraploid embryos and diploid embryos derived by intercrossing TRAP220+/- parents, and TRAP220+/hyp parents, to produce TRAP220 null
tetraploid and TRAP220 hypomorph tetraploid chimeras, respectively (44). In generating tetraploid aggregation chimeras, tetraploid morulas fail to contribute to the embryo, which is therefore derived entirely from the diploid partner of the chimera. However, tetraploid cells maintain an unrestricted potential to develop into the extraembryonic lineages, such that the placenta will be invariably wild type for TRAP220. Therefore, if the embryonic lethality of TRAP220-/- or TRAP220hyp/hyp embryos at E11.5 and E13.5, respectively, is attributable to placental defects, the tetraploid partners should rescue this.
Genotype analysis of the resulting chimeras confirmed this proposal. Viable TRAP220 null tetraploid chimeras were recovered at E12.0 (one mutant of a total of two), E13.0 (three mutants of a total of four), and E13.5 (two mutants of a total of four), demonstrating undoubtedly that TRAP220 is essential in the placenta during embryogenesis (Fig. 6), and moreover, that the death of TRAP220-/- embryos at E11.5 is due primarily to placental insufficiency and not cardiac failure. However, although these rescued embryos appeared viable, a second lethal phase was discovered later in embryonic development at E13.5. When hypomorph chimeras were genotyped, almost all TRAP220 hypomorph tetraploid chimeras (four of five) were viable at E13.5, but could not be rescued any further, reinforcing the absolute requirement of TRAP220 at E13.5 in the embryo (Fig. 6).
|
tetraploid chimeras retrieved at E13.5 were severely degenerate, demonstrating that, in addition to its role in the placenta, TRAP220 plays an essential role in the continuation of embryogenesis. On gross examination, both TRAP220 null and hypomorph tetraploid chimeras appeared severely growth retarded. In particular, limb and eye development were evidently delayed by at least 1 d in comparison to wild-type controls (Fig. 6, compare A with F and K), and in some cases a pericardial effusion was evident (Fig. 6F, inset). By histological analysis, it was revealed that the death of TRAP220 mutant tetraploid chimeras could be attributed to hepatic hemorrhagic necrosis, coupled with poor myocardial development. The livers of all TRAP220 mutant tetraploid chimeras exhibited hepatic necrosis, and in some cases this was visible on macroscopic inspection (Fig. 6, compare panel D with panels I and N). Nucleated erythrocytes remained disproportionately abundant in comparison to other hematopoietic cells (Fig. 6, compare panel E with panels J and O), and furthermore, no necrosis or apoptosis was present in other organs of the embryo, indicating a specific hepatic lesion rather than generalized resorbing. The hearts of all TRAP220 mutant tetraploid chimeras were also very degenerate, appearing dilated, congested with blood, and hypoplastic (Fig. 6, compare panel B with panels G and L). In addition to changes already observed in TRAP220-/- embryos at E11.5, myocardial fibers were disrupted, myocytes appeared swollen, and necrosis could be noted in both the pericardium and outer myocardium. Therefore, as before, the ventricular myocardium of TRAP220 mutant tetraploid chimeras failed to stratify into the multilayer structure that is necessary to maintain normal cardiac function as observed in wild-type controls (Fig. 6, compare panel C with panels H and M).
Normal Ontogeny of Other Organs in TRAP220 Mutant Tetraploid Chimeras
Strikingly, many other regions of the TRAP220 mutant tetraploid chimeras appeared completely normal in the presence of a defective cardiac and hepatic system. The central nervous system was normal with no obvious focal lesions in any TRAP220 mutant tetraploid chimeras. In the forebrain, normal structures visualized included the lateral and third ventricles, and diencephalon. Eye development was visualized with good retinal pigmentation and lens differentiation, consistent with E13.5 development, whereas in the hindbrain, the metencephalon and falx were developing normally, again consistent with a developmental stage of E13–E13.5. The nasal cavity, oropharynx, and tongue were developing normally, whereas normal structures could also be seen in the upper thoracic cavity. The aorta was visualized with active muscle proliferation, and the pulmonary vein was also normal. Lung development was proceeding normally. Finally, in the abdominal cavity, in contrast to aberrant liver development, gut development (stomach and small and large intestine) was entirely normal for E13.5 (data not shown).
Expression Analysis
In an attempt to understand the molecular mechanisms associated with the TRAP220 phenotypes, real-time PCR was performed on a number of candidate genes known to be important in hematopoiesis and cardiac development. Through this analysis we found that vascular endothelial growth factor (VEGF), an important angiogenic mediator that is known to cause noncompaction of the ventricular myocardium when overexpressed in mice (50), was up-regulated at significantly higher levels throughout midgestation in TRAP220-/- embryos than in wild-type and heterozygous controls from E10.5 onward, as was one of its receptor isoforms VEGFR-1 (flt-1), which is known to be essential during embryogenesis in regulating the formation of blood vessels (51) (Fig. 7, A and B). Similarly, erythropoietin (Epo), a glycoprotein hormone that is responsible for erythrocyte maturation, was up-regulated by at least 2-fold in TRAP220-/- embryos during midgestational development, and to a lesser extent its receptor, EpoR (Fig. 7, C and D), which may account for the abundance of nucleated erythrocytes and lack of megakaryocytes in TRAP220-/- embryos. Other mRNAs analyzed included VEGFR-2, VEGFR-3, hypoxia-inducible factor 1, hypoxia-inducible factor 2, aryl hydrocarbon receptor nuclear translocator, GATA-1, and GATA-2, which were expressed at levels similar to wild-type and heterozygous controls (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The formation of the placenta is fundamental to all mammalian development, and in mice begins at approximately E4.5 after implantation. Placental function must be sufficient by E7.5 onward when the metabolic requirements of the growing embryo approach the capacity of the yolk sac. In addition to its maternal-fetal exchange role, the placenta is a highly specialized organ that is responsible for the production of hormones, angiogenic growth factors, and tissue-remodeling factors (52). A number of nuclear receptor null mice display defects in placental development and die during midgestational development, including PPAR (53), RXR (54), and RXR/RXRß (55) mutants. More importantly, TRAP220 has been shown to be a bone fide coactivator for all these receptors (34, 37, 56, 57), and, furthermore, it has recently been demonstrated that mice devoid of AIB3 (58) or PRIP (59), two other coactivators of PPAR and other nuclear receptors, also exhibit defective placentation.
The placentae of TRAP220-/- embryos were histologically normal at E10–E11.5 in comparison to wild-type controls, with all three placental layers being well defined and developed and undergoing normal proliferation. However, ultrastructural analysis of TRAP220-/- placentae has previously suggested that defects are apparent in the vascularization of the labyrinthine portion (41), where TRAP220 placental expression is greatest, and which overlaps phenotypes observed in PPAR (53), AIB3 (58), and PPAR-interacting protein (59) null mutants, strongly suggesting that these four proteins may function in a common pathway in vivo. Indeed, the failure to establish an efficient chorioallantoic placenta at approximately E10–E11.5 is a common cause of death during midgestation (49) and is typically manifested by defects in intraembryonic circulation and exchange, and growth retardation of the embryo.
To determine whether TRAP220-/- embryos died at E11.5 as a consequence of placental failure, TRAP220 null tetraploid chimeras were generated. Through this assay, we have unequivocally confirmed that TRAP220 is critical in placental function during midgestation, because TRAP220 null tetraploid chimeras are partially rescued from E11.5–E13.5, and furthermore, revealed that TRAP220 dependency initially resides within extraembryonic tissues, but thereafter in embryonic tissues. Unlike PPAR (53) and estrogen-related receptor-ß (60) null tetraploid chimeras, TRAP220 null tetraploid chimeras are not rescued to term and die at E13.5 of severe hepatic hemorrhagic necrosis coupled with insufficient myocardial development. Interestingly, TRAP220 hypomorph tetraploid chimeras, which also die at E13.5, could not be rescued any further, reinforcing the absolute requirement of TRAP220 in the embryo at E13.5. Thus, it is conceivable that some of the placental abnormalities in TRAP220-/- embryos may involve PPAR pathways; however, the exact causes of these abnormalities are yet to be determined.
The cardiovascular system is the first functional organ to develop in the vertebrate embryo and must develop quickly in anticipation of the demands of the growing embryo for oxygen and nutrients. Cardiovascular development requires numerous maturation steps that are controlled by genes that may not all function in the cardiomyocyte population. Many growth factors and angiogenic mediators are responsible for controlling cardiac development, and their expression and modes of action are complex. Furthermore, even when hypoplastic phenotypes are observed exclusively in the cardiomyocyte lineage of the heart, it is noteworthy that signals from the developing endocardium and epicardium influence cardiomyocyte proliferation (61). TRAP220 mutants and TRAP220 mutant tetraploid chimeras both exhibited hypoplasia of the heart, which is likely to disrupt fetal circulation. Trabeculation of the heart is impaired and ventricular function is hampered as myocardial cells fail to stratify into a multilayered structure needed to maintain normal cardiac function. In PPAR mutants, thin myocardial wall syndrome can be completely rescued through replacement of their trophoblast lineage by tetraploid chimera methodology (53). However, in contrast to this, TRAP220 null tetraploid chimeras still exhibit poor trabeculation and myocardial wall thinning after rescue with tetraploid methodology, implying that defective PPAR pathways are not contributing to the cardiac phenotype observed in TRAP220-/- embryos. Interestingly, single or double RAR and RXR knockouts exhibit (among other abnormalities) growth defects in cardiac tissues. However, in most cases the phenotypes observed in TRAP220 mutants are clearly distinct from these mutants. Ablation of RXR results in death at E14.5 due to severe cardiac muscle defects. Like TRAP220-/- embryos, expansion of the compact zone fails; however, trabeculation remains normal (62, 63), suggesting that RXR pathways may not be impaired in TRAP220 mutant mice.
It is likely that the hypoplastic myocardial wall and trabecular developmental abnormalities in TRAP220-/- hearts are secondary to defective placentation, because in the developing heart, PPAR is not expressed (53) and TRAP220 is expressed at very low levels. Although we cannot exclude the potential role of TRAP220 in the regulation of cardiomyocyte proliferation, cardiac development could be controlled by exogenous factors that depend on the expression of TRAP220 elsewhere. Given that VEGF and VEGFR-1 are up-regulated in TRAP220-/- embryos, it is conceivable that this may be inducing a cardiac phenotype similar to that of thin myocardial wall syndrome. VEGF is a key regulator of angiogenesis, and its biological effects are mediated by two distinct tyrosine kinase receptors, VEGFR-1 (flt-1) and VEGFR-2 (KDR/flk-1). VEGF is essential for early development of the vasculature to the extent that inactivation of even a single allele results in embryonic lethality in mice (64, 65). Likewise, disruption of the genes encoding VEGFR-1 (51) and VEGFR-2 (66) also results in severe abnormalities of blood vessel formation. In humans, thin myocardial wall syndrome has been linked to an arrest of endocardial-myocardial development (67), and the specific overexpression of VEGF in mice perturbs this junction and induces noncompaction of the ventricular myocardium (50).
The development of the hematopoietic system is a complex process that takes place in several microenvironments. Primitive hematopoiesis appears in the blood island of the embryo yolk sac at E7.5 in mice and, subsequently, definitive hematopoiesis arises in the aorta-gonad-mesonephros region at E10.5 before shifting to the fetal liver at E11.5, where the major production of various hematopoietic cells begins. Finally, at around the time of birth, hematopoiesis shifts to the bone marrow and spleen (68). Both embryonic and definitive hematopoiesis require a plethora of both broad-spectrum as well as lineage-specific transcription factors that modulate the expression of downstream genes to mediate the formation, survival, proliferation, and differentiation of pluripotent progenitor cells (68). Any disruption in the hematopoietic cascade could lead to an arrest of hematopoiesis and hepatic failure.
At E11.5, TRAP220 mutant embryos displayed abnormalities in megakaryocyte and erythroid differentiation, while at later stages both TRAP220hyp/hyp and TRAP220 null tetraploid chimeras die as a result of severe hepatic necrosis. Although a direct role for TRAP220 in hematopoiesis remains uncertain, TRAP220 expression is strongest in the fetal liver at E11.5 when definitive hematopoiesis begins and, interestingly, TRAP220 has been shown recently to interact with GATA-1 (43), a key regulator of myeloid lineage maturation (68). GATA-1 null mice display complete ablation of embryonic erythropoiesis and, more importantly, display blocked megakaryocyte maturation and die at E11.5 (69). Furthermore, mice lacking GATA-1 specifically in megakaryocytes have defects that reveal the absolute requirement of this transcription factor in thrombopoiesis (70). Platelet shape is abnormal, and platelet counts are low in these mice, leading to prolonged bleeding. Thus, it is conceivable that the livers of TRAP220 null tetraploid chimeras may be hemorrhagic as GATA-1 pathways are impaired, resulting in blocked megakaryocyte differentiation and impaired thrombopoiesis.
Finally, as Epo is up-regulated in TRAP220-/- embryos, it is conceivable that this may be causing polycythemia, i.e. the abundance of erythrocytes in the liver. Epo is a glycoprotein hormone that activates intracellular signaling pathways by binding to its receptor, EpoR, to induce the up-regulation of globins, transferrin receptor, and other erythroid proteins that enhance the viability and maturation of erythroid progenitor cells (68, 71). Epo null mice die of severe anemia at E13.5, when primitive erythroblasts die and are not replaced by definitive erythropoiesis, implying that Epo is critical for the survival, proliferation, and differentiation of definitive erythroid progenitors (72). In contrast, hypoxia induced by cardiac and pulmonary deficiencies causes Epo levels to rise to inappropriate levels in the blood and induce polycythemia (73). The overproduction of nucleated erythrocytes in the fetal liver may cause the sinusoids to become dilated and congested with blood and, furthermore, induce intrahepatic hemorrhage.
To conclude, we have shown here that TRAP220 plays an important role in the formation of a functional chorioallantoic placenta and is essential for the normal ontogeny of the cardiac and hepatic systems. Together, these results both confirm and expand the known spectrum of physiological processes regulated by the TRAP complex and in particular TRAP220 and, in addition, demonstrate the absolute requirement of this coactivator during the continuation of embryogenesis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
RNA Preparation and Analysis
Total RNA was isolated using Trizol (Invitrogen Life Technologies, San Diego, CA) according to the manufacturer’s instructions, and Northern blotting analysis was performed as described previously (74). For Northern analysis a 551-bp BamHI fragment from the 3'-part of the TRAP220 cDNA was used as a probe, and a 500-bp EcoRI GAPDH fragment was used as a loading control. Quantification was accomplished with a Storm PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, CA) and ImageQuant software.
Antibody Production and Western Analysis
Purified GST-tagged TRAP220 protein (aa 1126–1311) was used to immunize mice. TRAP220 antibody generation and purification were performed as described previously (75). Antibody specificity was tested using recombinant proteins and whole-cell extracts from a variety of mouse tissues. For immunoblotting, whole-embryo extracts were separated by 6% SDS-PAGE and blotted onto nitrocellulose. The membranes were blocked in TBS-T [20 mM Tris-HCl (pH 7.6), 137 mM NaCl, 0.1% Tween 20] containing 3% nonfat milk powder, washed with TBS-T, and incubated for 2 h with anti-TRAP220 mouse monoclonal serum. After washing, the membranes were incubated with biotinylated goat antimouse IgG (DAKO Corp., Carpinteria, CA), washed with TBS-T, incubated with streptavidin-horseradish peroxidase (DAKO Corp.) and washed again with TBS-T. Bound immunoglobulins were visualized using the enhanced chemiluminescence detection system (Amersham Pharmacia Biotech, Arlington Heights, IL).
Histology and in Situ Hybridization
For embryo histology, 5-µm sections of paraffin-embedded concepti were stained with hematoxylin and eosin. Specific localization of TRAP220 mRNA was accomplished by in situ hybridization using antisense riboprobes synthesized using [-35S]UTP (800 Ci/µmol: Amersham Pharmacia Biotech) with T3 and T7 polymerases (Promega Corp., Madison, WI). In total, four mouse TRAP220 riboprobes were generated, two directed against the 5'-, and two directed against the 3'-part of the TRAP220 cDNA, and all produced similar results when used separately. The relative positions of these probes are available on request. DNA fragments were subcloned into pBluescript (Stratagene, La Jolla, CA), linearized with XbaI to synthesize probes of approximately 650 nucleotides. In addition, several probes of nonrelated mRNAs with known expression patterns, similar length, and similar GC content were used as controls to verify the specificity of the hybridizations (data not shown). Embryonic mice and placentae (E9.5–E13.5) were excised from pregnant females, fixed in 4% paraformaldehyde (pH 7.2) overnight, dehydrated through an ethanol series, and embedded in paraffin wax. Tissue sectioning and in situ hybridizations were carried out as described previously (76). Slides were dipped in Ilford K5 emulsion and autoradiographed at 4 C before being developed in Kodak D19 (Eastman Kodak, Rochester, NY) and counterstained in Giemsa. Sections were examined under conventional or reflected light dark-field conditions (Nikon UK Ltd, Kingston Upon Thames, Surrey, UK).
Nonisotopic in situ hybridizations were carried out essentially as previously described (77). Digoxigenin-labeled antisense probes were synthesized to Pl2 and Tpbp (a gift from J. C. Cross) and hybridized overnight at 60 C. Signals were detected using 4-nitroblue tetrazolium chloride and 5-bromo-4-chloro-3-indolyl-phosphate (Roche Clinical Laboratories, Indianapolis, IN) and counterstained with nuclear fast red (Vector Laboratories, Inc., Burlingame, CA).
Analysis of Placental Proliferation and Apoptosis
BrdU (100 µg/g body weight) was injected ip into pregnant female mice. The mice were killed 2 h later, and the embryo and placental tissues were removed and fixed in 4% paraformaldehyde (pH 7.2) and processed for immunohistochemistry. BrdU labeling was detected using a BrdU in situ detection kit (Pharmingen, San Diego, CA) according to the manufacturer’s instructions. Sections were counterstained with hematoxylin.
Apoptotic cells were detected on 5-µm sections of paraffin-embedded placentae using the ApopTag Plus peroxidase in situ detection kit (Intergen, Purchase, NY), according to the manufacturer’s instructions.
Generation of Tetraploid-Aggregation Chimeras
Procedures followed the published protocol (78) with some modifications. Briefly, wild-type, two-cell stage blastomeres (E2.0) were recovered from superovulated CBAxC57BL/6J F1 females fertilized by C57BL/6J males. Recovered blastomeres were fused in 0.3 M mannitol drops using a CF150 genepulsar (Biochemical Laboratory Service, Budapest, Hungary), and the resulting tetraploid embryos were incubated at 37 C overnight, 5% CO2 in K-simplex optimized media. After two cell divisions, four-cell tetraploid compacted morulas, as well as six- to eight-cell stage morulas (E3.5) recovered freshly from TRAP220+/- x TRAP220+/- matings were treated with acid tyrode’s solution to remove the zona pellucida. Subsequently, overnight aggregates of one diploid embryo and two wild-type tetraploid embryos were assembled in K-simplex optimized media-covered microdepressions in a cell culture incubator. Aggregation-derived blastocysts were transferred into 2.5 d post coitus pseudopregnant CBAxC57BL/6J F1 females, and chimeric embryos were recovered for analysis at different midgestational embryonic stages.
Expression Profiling
Total RNA was extracted from embryos using Trizol according to the manufacturer’s instructions, after which 1 µg total RNA was converted into single-stranded cDNA using SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen Life Technologies) according to the manufacturer’s instructions. After synthesis, single-stranded cDNA was subjected to real-time PCR analysis.
Real-Time PCR and Data Analysis
Real-time PCR was performed and data were analyzed as described previously (79) using an ABI Prism 7700 sequence detector following the protocol of PE Applied Biosystems’ Taqman SYBR Green Master Mix using gene-specific primers and the constitutively expressed L19 gene as an internal control (80). The optimal primer concentrations were determined according to the manufacturer’s guidelines, and single amplicon generation was checked by agarose gel electrophoresis. Data were analyzed during the linear phase of the PCR and the results were plotted using Microsoft Excel. TaqMan primer sequences were generated using PE Applied Biosystems software and the sequences were: Epo, 5'-TCATCTGCGACAGTCGAGTTCT-3' and 5'-TGCACAACCCATCGTGACAT-3'; EpoR, 5'-CCCTGTGACTATGGATGAAGCTT-3' and 5'-TCTGGCCTGGGCTTTGAG-3'; VEGF, 5'-CATCTTCAAGCCGTCCTGTGT-3' and 5'-CTCCAGGGCTTCATCGTTACA-3'; VEGFR-1, 5'-CGGCTGTCCATGAAAGTGAA-3' and 5'-TGTTGCAGGCGAGCCAT-3'.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: AF, Activation function; AIB1, amplified in breast cancer 1; BrdU, 5-bromo-2'-deoxyuridine; E11.5, embryonic d 11.5; Epo, erythropoietin; EpoR, Epo receptor; ES, embryonic stem; PPAR, peroxisome proliferator-activated receptor; NR, nuclear hormone receptor; RXR, retinoid X receptor; SRC, steroid receptor coactivator; TR, thyroid hormone receptor; TBS-T, Tris-buffered saline-Tween 20; TRAP, TR-associated protein.
Received for publication March 20, 2003. Accepted for publication September 10, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
) and conditional (
) embryos die at E11.5 and E13.5, respectively, and are retarded in growth and development in comparison to heterozygous (
) and wild-type (
) siblings. More than 25 different conditional and null litters were compared, with a minimum of six embryos per litter. D, Northern blotting analysis of TRAP220 mRNA and GAPDH mRNA in pooled E10.5 embryos of the indicated genotypes. Each blot, containing 10 µg total RNA per lane, was hybridized simultaneously with probes for TRAP220 and GAPDH. Shaded bars show RNA quantification; note the reduced levels of TRAP220 mRNA in conditional embryos in comparison to wild-type message. E, Western blotting analysis of TRAP220 protein in pooled E10.5 embryos of the indicated genotypes. Note the reduced levels of TRAP220 protein in hypomorph embryos in comparison to wild-type protein. All genotypes were determined by Southern blotting using the external 3'-probe, or by PCR analysis with primers specific to each mutant allele (see Materials and Methods for details). +/+, Wild type; con/con, conditional; +/-, heterozygous null; +/hyp, heterozygous hypomorph; hyp/hyp, hypomorph; and -/-, null.
