Defective haematopoiesis and vasculogenesis in ... - Semantic Scholar

2 downloads 0 Views 1MB Size Report
his expert technical assistance, and Dr Anita Roberts for constructive review of the manuscript. All animal experimentation was done under licence from the UK ...
Development 121, 1845-1854 (1995) Printed in Great Britain © The Company of Biologists Limited 1995

1845

Defective haematopoiesis and vasculogenesis in transforming growth factorβ1 knock out mice Marion C. Dickson1,*, Julie S. Martin 1, Frances M. Cousins1, Ashok B. Kulkarni2, Stefan Karlsson2 and Rosemary J. Akhurst1,† 1Department of Medical Genetics, University of Glasgow, Duncan Guthrie Institute, Yorkhill, Glasgow, G3 8SJ, UK 2Developmental and Metabolic Neurology Branch, National Institute of Neurological Disorders and Stroke, National

Institutes of

Health, Bethesda, Maryland, 20892, USA *Present address: Glaxo Research and Development Ltd., Gunnels Wood Road, Stevenage, Herts SG1 2NY, UK. †Author for correspondence

SUMMARY Transforming growth factor β1 (TGFβ1) is shown here to be required for yolk sac haematopoiesis and endothelial differentiation. Mice with a targeted mutation in the TGFβ1 gene were examined to determine the cause of prenatal lethality, which occurs in 50% of homozygous TGFβ1 null (TGFβ1−/−) conceptions. 50% of TGFβ1−/− and 25% of TGFβ1+/− conceptuses were found to die at around 10.5 dpc. The primary defects were restricted to extraembryonic tissues, namely the yolk sac vasculature and haematopoietic system. The embryos per se showed developmental retardation, oedema and necrosis, which were probably secondary to the extraembryonic lesions. The defect in vasculogenesis appeared to affect endothelial differentiation, rather than the initial appearance and outgrowth of endothelial cells. Initial differentiation of yolk sac mesoderm to endothelial cells occurred, but defective

differentiation resulted in inadequate capillary tube formation, and weak vessels with reduced cellular adhesiveness. Defective haematopoiesis resulted in a reduced erythroid cell number within the yolk sac. Defective yolk sac vasculogenesis and haematopoiesis were present either together, or in isolation of each other. The phenotypes are consistent with the observation of abundant TGFβ1 gene expression in both endothelial and haematopoietic precursors. The data indicate that the primary effect of loss of TGFβ1 function in vivo is not increased haematopoietic or endothelial cell proliferation, which might have been expected by deletion of a negative growth regulator, but defective haematopoiesis and endothelial differentiation.

INTRODUCTION

rescue of these TGFβ1−/− mice (Letterio et al., 1994). However, only 50% of TGFβ1−/− conceptuses actually reach parturition (Shull et al., 1992; Kulkarni et al., 1993), suggesting that there is some prenatal lethality of TGFβ1−/− conceptuses, and that TGFβ1 is essential for embryonic or fetal development in these cases. During embryogenesis, the first detectable expression of TGFβ1, as observed by reverse transcriptase (RT) polymerase chain reaction (PCR) analysis, occurs at the pre-implantation stage (Rappolee et al., 1988). However, using in situ hybridisation analysis, TGFβ1 transcripts were not detectable until 7.5 days post coitum (dpc), when expression was seen in the extraembryonic blood islands of the yolk sac, mesodermal cells of the allantois and the pro-angioblast progenitors within the cardiogenic mesoderm of the embryo (Akhurst et al., 1990). The mesodermal haemangioblasts of the blood islands are progenitors of both endothelial and haematopoietic lineages (Sabin, 1920; Wilt, 1965; Miura and Wilt, 1969), and cellular descendants of both lineages were shown to express TGFβ1 RNA and protein (Akhurst et al., 1990). Furthermore, within the embryo proper, at later stages of development, TGFβ1

TGFβ1 is a member of a large superfamily of proteins, including the activins, inhibins, bone morphogenetic proteins, and Müllerian inhibitory substance, which are thought to be important for normal development (Lyons et al., 1991; Akhurst, 1994). The distribution of TGFβ1 RNA and protein suggests that it plays an important function throughout postimplantation mammalian development (Heine et al., 1987; Lehnert and Akhurst, 1988; Gatherer et al., 1990; Akhurst et al., 1990). Nevertheless, targeted disruption of the TGFβ1 gene does not necessarily lead to embryonic lethality or congenital malformation (Shull et al., 1992; Kulkarni et al., 1993). The absence of a developmental phenotype in TGFβ1 homozygous null (TGFβ1−/−) neonates born to a cross of TGFβ1 heterozygotes (TGFβ1+/−) has been attributed either to the expendable role of TGFβ1 in normal development or to redundancy with other TGFβ isoforms (Erikson, 1993). Recently, it has been demonstrated that there is transplacental and lactational transfer of maternal TGFβ1 from TGFβ1+/− mothers to their TGFβ1−/− fetuses and lactating pups, which might account for

Key words: TGFβ1, endothelial differentiation, haematopoiesis, yolk sac, embryogenesis, mouse

1846 M. C. Dickson and others RNA was detected in both fetal blood cells of the liver and early endothelial cells; thus implicating this growth factor in the control of haematopoiesis, vasculogenesis and angiogenesis (Heine et al., 1987; Lehnert and Akhurst, 1988; Gatherer et al., 1990; Akhurst et al., 1990). TGFβ1 is widely reported to be a potent inhibitor of both endothelial (Heimark et al., 1986; Muller et al., 1987) and haematopoietic (Ohta et al., 1987; Sing et al., 1988; Ottman and Pelus, 1988) proliferation. However, although TGFβ1 can inhibit endothelial cell growth in a 2-dimensional cell culture system, in a three dimensional culture system this ‘growth regulator’ had no effects on microvascular endothelial cell proliferation, but primarily affected differentiation (Madri et al., 1988, 1992; Merwin et al., 1990; Basson et al., 1992). The role of TGFβ1 in vasculogenesis and angiogenesis in vivo, might therefore be to check cellular proliferation rates at a time of most rapid endothelial cell division, or to induce the differentiation of, or stabilise, endothelial tubes. The endogenous function of TGFβ1 in haematopoiesis is equally contentious. Most reports conclude that TGFβ is an inhibitor of haematopoiesis (Ohta et al., 1987; Sing et al., 1988; Ottman and Pelus, 1988), though most of these studies were performed on cultured cells, so the exact function of TGFβ1 on haematopoiesis in vivo remains debatable. The TGFβ1 knock out mice provide an important resource for studies on the developmental function of TGFβ1. It has previously been reported that roughly 50% of TGFβ1−/− conceptions reach partuition, but then die of a multifocal inflammatory disorder three weeks post-partum (Shull et al., 1992; Kulkarni et al., 1993). In this study, we have investigated the cause of prenatal lethality of TGFβ1−/− conceptuses which do not reach partuition. The findings suggest an important role for TGFβ1 as a positive regulator of endothelial differentiation and yolk sac haematopoiesis.

MATERIALS AND METHODS TGFβ1 knock-out mice All the animals examined in this study were from a mixed genetic background. The TGFβ1 null allele had originally been generated on a 129/Sv genetic background (Kulkarni et al., 1993). Mice were bred for 2 generations on to a C57BL/6J background, and then crossed once to inbred NIH/Olac (Harlan Olac Ltd., UK). Subsequent matings were obtained by F1 intercrosses. Embryos and pups were generated from intercrosses of TGFβ1+/− mice. Embryos were harvested from natural timed matings. The presence of the vaginal plug was regarded as 0.5 dpc. Preimplantation embryo transfers For the embryo transfers, 2 cell stage mouse embryos were harvested from superovulated TGFβ1+/− intercrosses, and transferred to pseudopregnant wild-type mothers, using standard techniques (Allen et al., 1987). Genotype analysis was performed on the viable 18 dpc or 19 days post-partum (dpp) offspring. Preparation and characterisation of TGFβ-depleted embryo culture medium Blood was harvested from the dorsal aorta of ether-anaesthetised rats, as described by (Cockcroft, 1990). The blood was gently transferred to tubes containing heparin-coated beads (Sarstedt), inverted and left for up to 2 hours on ice. The tubes were then centrifuged at 2.4 K rpm

for 20 minutes to separate the blood, and the upper plasma layer collected. The plasma was centrifuged for a further 10 minutes, and heat-treated at 56°C for 40 minutes. Before use in the culture system, the plasma was decanted into 1.5 ml Eppendorf tubes and centrifuged at 14 K rpm for 15 minutes to deposit a diffuse pellet. The supernatant was removed and this process was repeated once more before the plasma was ready to be used. A standard CCL64 assay (Danielpour et al., 1989) demonstrated no detectable TGFβ in this plateletdepleted plasma (i.e.