Abstract Hematopoiesis involves the generation of pluripotent stem cells (HSCs) that have the ability to self-renew or differentiate into any of at least eight distinct lineages (Orkin, 1995). The HSCs undergo successive differentiation to multipotential progenitors, lineage-committed precursors, and, finally, terminally differentiated cells. Differentiation of HSCs along a particular lineage is determined, in part, by the availability of specific cytokines and their receptors and the utilization of signal transducing pathways that allow for select changes in gene expression to occur by modifying pre-existing nuclear factors.
p56lck, a member of the src family of cytoplasmic tyrosine protein kinases, is expressed primarily in lymphoid cells. Previous RNase protection data demonstrated the existence of at least two lck mRNAs (type I and type II) with different 5' untranslated regions in most T cells. These have been found here to arise from two separate promoters. S1 nuclease analysis and primer extension were used to locate the site of initiation of type I lck mRNA. The nucleotide sequence of the region upstream of this start site contains no classical promoter motifs. A cDNA clone of type II lck mRNA was isolated. The promoter of this mRNA must be more than 10 kilobases upstream of the type I promoter region. In two murine thymoma cell lines, LSTRA and Thy19, lck is expressed at elevated levels as a result of Moloney murine leukemia virus retrovirus promoter insertion. p56lck is encoded in these cells by a hybrid virus-lck mRNA containing the 5' untranslated region of Moloney virus mRNA. The structures and the sites of integration of the proviruses upstream of lck in these cells were examined by molecular cloning and Southern analysis. A truncated and rearranged provirus, flanked by 554 nucleotides (nt) of duplicated cellular sequences, was found 962 nt upstream of the start site for type I lck mRNA in LSTRA cells. What appears to be a Moloney mink cytopathic focus-forming provirus was found between 584 to 794 nt upstream of the start site for type I lck mRNA in Thy19 cells. Thus in both tumor cell lines, viral DNA is present between the promoters for type I and type II lck mRNAs. Comparison of the sequences of the 5' ends of the lck and c-src genes suggests that divergence of these two genes involved exon shuffling and that a homolog of the neuronal c-src(+) exon is not present in lck.
We have isolated and characterized Xenopus cDNA clones for a new transcription factor that represents an early marker for the developing heart. The cDNAs encode a protein that we have designated GATA-4; it contains the highly conserved DNA-binding domain that characterizes this family of cell-type restricted transcriptional activators. Whole-embryo in situ analysis of Xenopus embryos demonstrates that the GATA-4 gene is transcribed in presumptive cardiac ventral mesoderm at the time that bilateral progenitors fuse and form the cardiac tube. GATA-4 is therefore the earliest molecular marker of cardiogenesis yet characterized. By stage 30, the GATA-4 mRNA is expressed in the developing atria and ventricles; at stage 38, cross-sections reveal that the gene is active in the endocardial layer, but not in myocardium. By stage 40, GATA-4 message is detected in the great vessels. In the adult frog, the GATA-4 gene is highly transcribed in heart and gut; lower levels of message are detected in various endoderm-derived tissues and gonads. Expression in the stomach is largely confined to the epithelium. The GATA-4 gene is first activated at stage 11; mRNA is initially present throughout the marginal zone of explants and later partially localized to the ventral marginal zone. GATA-4 mRNA is also detected at high levels in cultured endodermal explants derived from the vegetal region of early embryos. In mesoderm induction experiments, GATA-4 transcription is not induced in animal caps treated with activin or bFGF. The GATA-4 gene may provide a new early marker for studying the inductive processes that lead to the formation of the cardiovascular system and for the specification of the endocardial lineage.
Collagen lattices containing bovine retinal pericytes (RPs), vascular smooth muscle cells (VSMCs), pulmonary microvessel endothelial cells (PMECs), or aortic endothelial cells (AECs) were prepared and contraction was quantitated by measuring the resulting change in lattice area. VSMCs were the most efficient at lattice contraction followed by RPs and then PMECs. AECs did not contract the lattices. To document further that these observations represent contraction, cells were grown on inert silicone rubber sheets. Substratum wrinkling was indicative of tension development and quantitated as percent of cells contracted. RPs were more contractile than PMECs, and AECs were incapable of developing tension. VSMCs were less contractile than RPs, unlike the comparative contractility observed with the lattice system. Alteration of actin-containing filaments by cytochalasin B significantly reduced RP contraction of silicone rubber and inhibited their contraction of collagen lattices in a dose-dependent manner. Rhodamine-phalloidin staining of contracting RPs revealed microfilament bundle orientations that suggested their association in the force applied for contraction. RP, VSMC and PMEC contraction of collagen lattices was directly proportional to the concentration of fetal calf serum. Also, RP contraction was greater in calf serum than calf plasma-derived serum, an indication that RPs respond to substances that appear continuously and episodically in blood. These in vitro findings support the theory that pericytes in vivo are contractile but that endothelial cells may also contribute to microvascular tonus.
