In contrast to skeletal muscle, the mechanisms responsible for activation and maintenance of tissue-specific transcription in cardiac muscle remain poorly understood. A family of hormone-encoding genes is expressed in a highly specific manner in cardiac but not skeletal myocytes. This includes the A-and B-type natriuretic peptide (ANP and BNP) genes, which encode peptide hormones with crucial roles in the regulation of blood volume and pressure. Since these genes are markers of cardiac cells, we have used them to probe the mechanisms for cardiac muscle-specific transcription. Cloning and functional analysis of the rat BNP upstream sequences revealed unexpected structural resemblance to erythroid but not to muscle-specific promoters and enhancers, including a requirement for regulatory elements containing GATA motifs. A cDNA clone corresponding to a member of the GATA family of transcription factors was isolated from a cardiomyocyte cDNA library. Transcription of this GATA gene is restricted mostly to the heart and is undetectable in skeletal muscle. Within the heart, GATA transcripts are localized in ANP-and BNP-expressing myocytes, and forced expression of the GATA protein in heterologous cells markedly activates transcription from the natural cardiac muscle-specific ANP and BNP promoters. This GATA-dependent pathway defines the first mechanism for cardiac muscle-specific transcription. Moreover, the present findings reveal striking similarities between the mechanisms controlling gene expression in hematopoietic and cardiac cells and may have important implications for studies of cardiogenesis.The discovery of the MyoD family of myogenic factors has resulted in great advances in the understanding of the mechanisms of skeletal muscle commitment and differentiation (reviewed in references 48 and 68). In contrast, the mechanisms controlling cardiac determination and differentiation remain essentially unknown (49). So far, no MyoD-like factors have been detected in cardiac muscle (54), and mice homozygous for inactivated MyoD (53), Myf-5 (13), or myogenin (27) loci have a normal cardiac phenotype. These observations suggest that tissue-specific transcription and cell differentiation are controlled by distinct regulatory pathways in the two striated muscles. This would be consistent with the fact that skeletal and cardiac myocytes have distinct spatial and temporal origins in the developing embryo, although they both arise from mesoderm.Skeletal muscle cells originate from the somites of the dorsal (paraxial) mesoderm, whereas cardiac muscle cells are derived from the splanchnic mesenchyme of the anterior lateral plate mesoderm (10, 32). Commitment of mesodermal cells to the cardiac lineage occurs very early, when cells migrate to form the cardiogenic area at the beginning of the third week (days 16 to 18) of human embryonic development (or at 18 to 20 h in chicken embryogenesis [35]). By the end of the third week (days 21 to 22), the tubular-or primitive-heart is formed and joined by blood vessels. Thus, the heart ...

Commitment of mesodermal cells to the cardiac lineage is a very early event that occurs during gastrulation, and differentiation of cardiac muscle cells begins in the presomite stage prior to formation of the beating heart tube. However, the molecular events, including gene products that are required for differentiation of cardiac muscle cells, remain essentially unknown. GATA-4 is a recently characterized cardiac muscle-restricted transcription factor whose properties suggest an important regulatory role in heart development. We tested the role of GATA-4 in cardiac differentiation, using the pluripotent P19 embryonal carcinoma cells, which can be differentiated into beating cardiac muscle cells. In this system, GATA-4 transcripts and protein are restricted to cells committed to the cardiac lineage, and induction of GATA-4 precedes expression of cardiac marker genes and appearance of beating cells. Inhibition of GATA-4 expression by antisense transcripts blocks development of beating cardiac muscle cells and interferes with expression of cardiac muscle markers. These data indicate that GATA-4 is necessary for development of cardiac muscle cells and identify for the first time a tissue-specific transcription factor that may be crucial for early steps of mammalian cardiogenesis.

We have studied the effects of glucocorticoids on the activity and viability of neonatal rat osteoclasts in vitro. In the bone slice assay, glucocorticoids caused a dose-dependent decrease in the amount of bone resorbed, which was accompanied by a parallel decrease in osteoclast number. Loss of osteoclasts was due to their death, which occurred by the process of apoptosis. Evidence for the latter was obtained by a range of techniques, including time-lapse video microscopy, acridine orange staining, DNA fragment detection and transmission electron microscopy. Immunocytochemistry revealed the presence of glucocorticoid receptors in osteoclasts, and glucocorticoid-induced cell death could be prevented by the glucocorticoid receptor antagonist, RU486. These observations suggest that glucocorticoids promote receptor-mediated apoptosis of rat osteoclasts in vitro. This finding may help to explain recent data indicating that, in sharp contrast with their effects on the human skeleton, glucocorticoids inhibit bone resorption and increase bone mass in rats in vivo.

Primary pigmented nodular adrenocortical disease (PPNAD) is a rare cause of ACTH-independent adrenal Cushing's syndrome (CS), which is often associated with Carney complex (CNC). We have recently described a paradoxical increase in cortisol excretion after dexamethasone administration in most patients with PPNAD. In the present study we investigated the hypothesis that this phenomenon is due to a primary abnormality of the tissues affected by PPNAD, rather than a defect of the patients' hypothalamic-pituitary-adrenal axis; as such it should be replicated in vitro by adrenal slices exposed directly to dexamethasone. We were able to study adrenal tissues from eight patients with CS caused by PPNAD; two patients were also studied in vivo according to a protocol first described in ACTH-independent macronodular adrenal hyperplasia (AIMAH) for the clinical detection of aberrant hormone receptor expression. Their DNA has been previously screened for inactivating mutations of the PRKAR1A gene, the most frequent molecular defect leading to PPNAD and/or CNC. We also investigated whether glucocorticoid receptor (GR) expression underlies paradoxical dexamethasone responses in PPNAD by immunohistochemistry and semiquantitative PCR, and we correlated GR expression with that of other markers for PPNAD (e.g. synaptophysin). Indeed, we demonstrated that dexamethasone induced cortisol secretion in vitro in five of these tumors; no such increase was seen in adenomatous or AIMAH tissues that were treated in the same manner. GR mRNA was expressed, and GR immunoreactivity was detected in PPNAD nodular cells. Staining for GR was not seen in surrounding cortical cells, and hence, it correlated with synaptophysin, which also stains PPNAD in a similar manner. In normal adrenal tissue, GR was detected mostly in medullary areas, whereas GR immunoreactivity was weak in adenomatous and AIMAH tissues. We conclude that 1) dexamethasone produces an increase in glucocorticoid synthesis by PPNAD adrenal slices in vitro, suggesting a direct effect on adrenocortical tissue, and 2) this phenomenon is accompanied by increased expression of the GR in PPNAD nodules. PPNAD and/or CNC patients with and without mutations leading to protein kinase A activation demonstrated in vitro and/or in vivo paradoxical dexamethasone responses and GR expression, indicating that PRKAR1A alterations are not necessary for these phenomena.

The product of a single gene encoding proopiomelanocortin (POMC) is differentially processed to produce corticotropin and a-melanotropin in anterior and intermediate pituitary cells, respectively. Hormonal control of POMC gene transcription and of corticotropin or amelanotropin release is also tissue-specific; for example, glucocorticoids specifically inhibit anterior but not intermediate pituitary POMC transcription. Outside the pituitary gland, very low levels of POMC mRNAs are present in brain, testes, ovaries, and placenta. We have used transgenic mice to identify POMC 5' flanking sequences that are sufficient for tissuespecific expression and glucocorticoid regulation in anterior and intermediate pituitary cells. Three lines of transgenic mice were established, each carrying 50-75 copies (per cell) of a chimeric rPOMCneo gene constituted of rat POMC promoter sequences and of bacterial neomycin-resistance coding sequence. High levels of rPOMCneo transcripts were detected in pituitaries of mice from all three lineages. In situ hybridization revealed that the ratio of intermediate to anterior pituitary transcripts was similar for the transgene and endogenous POMC mRNA. rPOMCneo transcripts were not detected in any other tissue except at very low levels in the testes in two transgenic lines. Endogenous mouse POMC mRNA increased in response to depletion of plasma glucocorticoids (adrenalectomy) and decreased after glucocorticoid treatment; rPOMCneo transcripts were altered to the same extent by these treatments in all three lines. Intermediate pituitary and testicular rPOMCneo transgene expression was not altered by these treatments. Thus, no more than 769 base pairs of the rat POMC promoter are required for pituitary-specific expression and for specific glucocorticoid inhibition of the POMC gene in the anterior pituitary.