In Vivo and In Vitro Analysis of TGF-beta1 Knockout Embryos

The transforming growth factors-type beta (TGF-betas) are a group of proteins expressed in the developing embryo and the adult. Three TGF-beta's exist in vertebrates, TGF-beta1, TGF-beta2, and TGF-beta3, all of which have growth inhibitory and stimulatory roles in vitro, depending on the target cell type. In the mouse embryo TGF-beta1 RNA expression is first detected in the haemangioblasts of the yolk sac at approximately 7 dpc. These are small cell populations which give rise to the vascular endothelial cells and the haematopoietic cells. Later in development expression continues in tissues undergoing haematopoiesis and vasculogenesis, and is also found during osteogenesis, and in epithelia involved in epithelial / mesenchymal transformations (Akhurst et al, 1990). This project was undertaken to investigate the function of TGF-beta1 during development by depletion of the protein using molecular methods, such as antisense oligonucleotide technology, in conjunction with post-implantation embryo culture. However, to ensure that any resulting phenotypes were as a result of the removal of TGF-beta1, the medium also had to be depleted of TGF-beta. Dulbeccos modified Eagles medium (DMEM) plus rat serum is commonly used for embryo culture, and was also examined in this project. However, analysis of the TGF-beta content in the rat serum using the CCL64 assay revealed that considerable amounts of the protein were present (≥15ng/m1), thereby rendering serum unsuitable for use in this project. Therefore, another medium had to be found. Although previous reports had found plasma unsuitable for embryo culture, it was also reported that this biological fluid contained negligible amounts of TGF-beta. Analysis of rat plasma using the CCL64 assay in this project detected ≤10pg/ml, suggesting that plasma was ideal for use as a medium depleted of TGF-beta1. However, when the rat plasma was combined with DMEM, embryonic development was severely disrupted. Therefore, a completely novel medium was devised, utilising Foetal Mouse Palate medium, in combination with rat plasma. This medium contained no, or undetectable levels of TGF-beta, and supported normal development of 8.5 dpc embryo for 24 hours in vitro. Before the culture system could be used with antisense oligodeoxynucleotides to remove TGF-beta1 from mouse embryos, two other groups generated transgenic mice with targeted disruption of the TGF-beta1 gene (Shull et al, 1992; Kulkarni et al, 1993). 50% of the TGF-beta1 homozygous null mice were born, but only survived to approximately 3 weeks of age before dying of a wasting syndrome involving a multifocal, mixed inflammatory cell response, and tissue necrosis. No other gross phenotypes were observed. The other 50% of TGF-beta1 nulls died in utero. In collaboration with Dr.M.Dickson, I determined the reason for this embryonic lethality. By performing detailed studies at different developmental stages, it was concluded that the embryonic lethality was due to yolk sac abnormalities arising at 9-9.5 dpc. Both the vasculature and the blood was affected. The vasculature defects included weak vessels, a disorganised vasculature, or a complete lack of vessels altogether. Anaemia was observed in a number of yolk sacs. At 9.5 cipc, only if the yolk sac was severely defective was the development of the embryo per se affected. Dr.M.Dickson observed that by 10.5 dpc, the embryos appeared necrotic and were close to death (Dickson et al, 1995). Further experiments were performed to examine the dichotomy in embryonic survival. It was suggested that maternal TGF-beta1 was rescuing those TGF-beta1 null embryos that survived gestation (Letterio et al, 1994). This was examined in the current study by culturing embryos of all genotypes in the medium depleted of TGF-beta. The intact conceptuses were cultured from 8.5 dpc, when they are all phenotypically normal, to 9.5 dpc, when the defective phenotypes are observable. An increase in the number of abnormal phenotypes was expected if maternal TGF-beta was responsible for embryonic rescue. This was not observed. However, it was possible that the maternal TGF-beta had already had effect before 8.5 dpc. The initial effect of TGF-beta1 therefore appears to be in yolk sac vasculogenesis and haematopoiesis, coinciding with the expression data. As TGF-beta1 was thought to be a potent growth inhibitor of many cell types from in vitro studies (Roberts & Sporn, 1990, for review), BrDU incorporation was employed to examine cellular proliferation rates in the TGF-beta1 knockout conceptuses. The cells of the yolk sac, and two types of embryonic epithelial cell types were examined. There was no increase in cell proliferation in the TGF-beta1 null samples in comparison with the wild-type samples. It is therefore likely that TGF-beta1 has a role in the control of yolk sac endothelial and haematopoietic cell differentiation. TGF-beta1 therefore appears to be crucial for normal yolk sac development. Without a functioning yolk sac, the embryo does not receive nutrients and oxygen, and is unable to survive this critical gestation period before development of the chorioallantoic placenta