Mutant WD-repeat protein in triple-A syndrome.

Triple-A syndrome (MIM 231550; also known as Allgrove syndrome) is an autosomal recessive disorder characterized by adrenocorticotropin hormone (ACTH)-resistant adrenal insufficiency, achalasia of the oesophageal cardia and alacrima. Whereas several lines of evidence indicate that triple-A syndrome results from the abnormal development of the autonomic nervous system, late-onset progressive neurological symptoms (including cerebellar ataxia, peripheral neuropathy and mild dementia) suggest that the central nervous system may be involved in the disease as well. Using fine-mapping based on linkage disequilibrium in North African inbred families, we identified a short ancestral haplotype on chromosome 12q13 (<1 cM), sequenced a BAC contig encompassing the triple-A minimal region and identified a novel gene (AAAS) encoding a protein of 547 amino acids that is mutant in affected individuals. We found five homozygous truncating mutations in unrelated patients and ascribed the founder effect in North African families to a single splice-donor site mutation that occurred more than 2,400 years ago. The predicted product of AAAS, ALADIN (for alacrima-achalasia-adrenal insufficiency neurologic disorder), belongs to the WD-repeat family of regulatory proteins, indicating a new disease mechanism involved in triple-A syndrome. The expression of the gene in both neuroendocrine and cerebral structures points to a role in the normal development of the peripheral and central nervous systems.

Induction of pituitary lactotrope differentiation by luteinizing hormone alpha subunit.

Addition of gonadotropin releasing hormone to cultures of fetal rat pituitary induced differentiation of lactotropes as revealed by immunocytochemistry. Antiserum to luteinizing hormone (LH) (recognizing native LH), but not antiserum to LH-beta (recognizing both native LH and its beta subunit), inhibited this induction. Further addition of highly purified LH-alpha subunit in culture medium also induced lactotrope differentiation. Thus, the alpha subunit may have a specific biological activity of its own with probable practical use in clinical investigations.

Hormonal regulation of the mouse adrenal melanocortinergic system.

Adrenocortical cells of several species have been reported to express significant levels of Agouti-related protein (Agrp) as well as melanocortin 4-receptor (MC4-R). In this study, we used the mouse tumoral adrenal cell line ATC7- L that secretes corticosterone in basal conditions with a 2- fold increase in response to ACTH treatment. We reported that these cells expressed functional MC4-R. They also expressed Agrp mRNA and secreted immunoreactive Agrp in the culture medium. Long-term treatment of ATC7-L with (Nle4,D-Phe7)-alpha MSH (NDP-alpha MSH) or forskolin as well as Agrp strongly reduced MC4-R level by more than 30%. On the contrary, leptin treatment did not modify this level although it significantly reduced MC2-R level. These results could be correlated to some data obtained in vivo on adrenal glands removed from diet-induced obese mice exhibiting a hyperleptinemia, where the level of both MC2-R and MC4-R appeared to be reduced as Agrp mRNA expression level was increased compared to Control mice. All these data would suggest the existence of a link between the metabolic status and the activation of the adrenal melanocortinergic system.

Regulation of corticotropin receptor number and messenger RNA in cultured human adrenocortical cells by corticotropin and angiotensin II.

The regulation of ACTH receptor binding sites and mRNA by ACTH and angiotensin II (A-II) was studied using cultured human adrenal fasciculata reticularis cells (HAC). These cells expressed two major ACTH receptor transcripts of 1.8 and 3.4 kb and three minor ones of 4, 7, and 11 kb. ACTH increased the levels of all these transcripts in a time- and dose-dependent manner. At a maximal concentration of 10(-8) M, ACTH enhanced 21- and 4-fold the level of ACTH receptor mRNA and the number of receptors per cell, respectively. Pretreatment of HAC with A-II produced a dose-dependent enhancement of ACTH receptor mRNA that was associated with an increase of both ACTH receptor number and responsiveness to this hormone. The effects of A-II were completely blocked by an AT1 receptor subtype antagonist but not by an AT2 antagonist. The effects of ACTH together with A-II on ACTH receptor mRNA were greater than those induced by each hormone alone. These results show that ACTH receptor number and mRNA are positively regulated by the two main hormones (ACTH and A-II) which, in vivo, regulate adrenocortical functions. In addition, they also show that HAC are a target for A-II. Thus, regulation of ACTH receptors may be one mechanism by which ACTH and A-II regulate adrenocortical functions under both normal and pathological conditions.

Identification of Urop11, a novel leptin-modulated gene that is upregulated in the hypothalamus of mice with virus-induced obesity.

Obesity results from disturbances of tightly regulated interactions between the nervous, endocrine, and metabolic systems that can be caused by external factors, such as viral infections. A mouse model of obesity induced by brain infection with a morbillivirus, canine distemper virus, allowed us to identify obesity-related genes. Using a subtractive library for the hypothalamus, the main brain structure regulating energy homeostasis, we identified a new gene on mouse chromosome 19 which we named upregulated obese product (Urop) 11 and, which has no homology with any known mRNA. A step-by-step molecular approach allowed us to isolate the full-length mRNA, predict the protein sequence, and identify consensus sites. Urop11 was mainly detected in the hypothalamus and adipocytes, and was dramatically upregulated in these central and peripheral structures in obese mice. Urop11 was also expressed in human neural and lymphoid samples and its expression seemed to be regulated by the state of lymphocyte activation. Interestingly, Urop11 expression was strongly upregulated both in vivo in mouse hypothalamus and in vitro in mouse neural cell lines, after leptin treatment. Taken together, our data show that Urop11 is a target of leptin, the satiety factor produced by adipocytes, in physiological and pathological conditions, including obesity. This new gene can be considered a key molecule in the hypothalamic integration pathway and demonstrates the importance of Urop11 as a target of leptin action.

Presence of multiple functional polyadenylation signals in the 3'-untranslated region of human corticotropin receptor cDNA.

We present 2.59 kb of the 3'-non-coding region of the ACTH receptor cDNA that contains seven potential polyadenylation signals. Among these signals, five are functional as detected by 3'-RACE and are consistent with the transcripts of 1.8, 3.4 and 4 kb visualized on Northern blots. We propose that the most likely molecular mechanism for the multiple ACTH-R mRNA transcripts is the alternative use of polyadenylation signals.

Expression of the human melanocortin-4 receptor gene is controlled by several members of the Sp transcription factor family.

The melanocortin-4 receptor (MC4-R) plays a key role in the hypothalamic control of food intake, lending importance to the understanding of the mechanisms that regulate its expression. To identify factors controlling the expression of the human (h) MC4-R gene, a fragment containing 1253 bp of the 5'-flanking region of the hMC4-R gene was isolated. A series of hMC4-R luciferase constructs were developed and used to transiently transfect HEK293 and GT1-7 cell lines, both expressing endogenous MC4-R mRNA. Deletion analysis of the 1253 bp fragment showed that the basal promoter activity is mainly restricted to the 179 bp upstream of the transcription start site in both cell types. Mutation of a putative Sp1-binding site located at position -76 bp resulted in a dramatic reduction of the luciferase activity in HEK293 and GT1-7 cells by 87 and 80% respectively. Both in vitro and in vivo studies (gel shift and chromatin immunoprecipitation analyses) revealed binding of both Sp1 and Sp3 to this site in HEK293 cells. Cotransfection with an Sp1 expression vector in Drosophila cells that do not express Sp1, in conjunction with treatment of HEK293 cells with mithramycin A, a specific inhibitor of Sp1, confirmed the role of Sp1. For the first time, we have demonstrated that the constitutive activity of the hMC4-R promoter is dependent upon Sp transcription factors.

Functional relationships between three novel homozygous mutations in the ACTH receptor gene and familial glucocorticoid deficiency.

Familial glucocorticoid deficiency (FGD) is an autosomal recessive disorder characterized by a glucocorticoid adrenal insufficiency without mineralocorticoid deficiency. Mutations of the ACTH receptor (MC2-R) gene have been reported in some FGD cases, but only a few of them have been functionally studied. We reported clinical features and MC2-R gene analysis in three families. For each proband, an homozygous mutation was identified after amplification and sequencing of the whole intronless MC2-R gene. One mutation converted Val-142 located in the second intracellular loop to Leu. Another mutation in the sixth transmembrane domain converted Ala-233 to Pro. The last mutation converted the negatively charged Asp-103 in the first extracellular loop to an uncharged Asn. Functional studies of these mutations as well as the S120R mutation were performed after stable transfection of M3 cells and measurement of ACTH-induced cAMP production. For the S120R, V142L, and A233P mutated MC2-R, cAMP production curves were similar to that obtained with M3 parental cells, confirming that these mutations are responsible for the FGD in the affected patients. The D103N-mutated MC2-R had an impaired cAMP response to physiological doses of ACTH, but the maximal response at very high concentrations of ACTH was similar to that obtained for the wild-type MC2-R. All these results demonstrated clear relationships based on functional studies between MC2-R homozygous mutations and FGD phenotype.

Genetics of ACTH insensitivity syndromes.

The last decade has seen remarkable progress in our understanding of the genetic causes of these potentially lethal conditions. Clearly, other genes that cause FGD remain to be discovered, and further work is required on the functions of MRAP and ALADIN in the expectation that they will provide insights into essential biological processes and perhaps identify key therapeutic strategies and targets for these diseases.

Regulation of guanine nucleotide binding regulatory proteins in cultured adrenal cells by adrenocorticotropin and angiotensin-II.

In addition to their steroidogenic effect on cultured bovine adrenal fasciculata cells ACTH and angiotensin-II (A-II) have a long term effect on the ability of these cells to respond to subsequent hormonal stimulation. The present work explores the effects of a 72-h pretreatment of adrenal cells with both hormones on the first steps of the mechanism of action of ACTH and A-II and on the amounts of the alpha-subunits of guanine nucleotide binding proteins Gs and Gi. ACTH but not A-II increased acute ACTH or cholera toxin-induced cAMP production. Moreover, ACTH but not A-II enhanced the amount of alpha S protein evaluated by cholera toxin ADP ribosylation, whereas both hormones elevated immunoblotted alpha S. Both hormones increased A-II induced phosphoinositide breakdown and Ca2+ uptake without modification of the A-II potentiating effect on ACTH-induced cAMP production. Treatment of cells with pertussis toxin (PT, 0.5 micrograms/ml) for the last 24 h reduced by 27% the A-II induced phosphoinositide breakdown in A-II pretreated cells but had no significant effect in ACTH-pretreated cells. No effect of PT was observed on A-II induced Ca2+ uptake or on its potentiating action on ACTH-induced cAMP production in ACTH as well as A-II-pretreated cells. Moreover, both hormones increased Gi proteins (40-41 kDa) evaluated by PT ADP ribosylation. Immunoblot analysis revealed that ACTH preferentially enhanced alpha i3, whereas the stimulatory effect of A-II was more marked on alpha i1 and alpha i2. These results indicate that in bovine adrenal fasciculata cells, peptide hormones settle target cell responsiveness not only by regulating the membrane-bound receptors, but also by modulating the level of G proteins coupling these receptors to the intracellular signals.

Stimulatory effect of insulin and insulin-like growth factor I on Gi proteins and angiotensin-II-induced phosphoinositide breakdown in cultured bovine adrenal cells.

Recent data have shown that pretreatment of bovine adrenal fasciculata cells with insulin-like growth factor I (IGF-I) or insulin enhances the steroidogenic response to angiotensin II (A-II). In the present work we have studied the effects of both peptides on the first steps of the mechanism of action of A-II and on the amounts of pertussis toxin (PT)-sensitive guanine nucleotide binding proteins (Gi proteins). Both peptides increased A-II-induced phosphoinositide breakdown without modification of either A-II-induced Ca2+ uptake or the A-II-potentiating effect on ACTH-induced cAMP production. The effects of IGF-I at a nanomolar concentration were higher than those induced by insulin at a micromolar concentration, which in turn was higher than those induced by a nanomolar concentration of this peptide. Treatment of cells with pertussis toxin (0.5 microgram/ml) for 24 h reduced by 25% of the A-II-induced phosphoinositide breakdown in control cells and 32% and 28% in cells pretreated with insulin at nanomolar and micromolar concentrations, respectively, but had no significant effect in cells pretreated with IGF-I. No effect of pertussis toxin was observed on A-II-induced Ca2+ uptake or on its potentiating action on ACTH-induced cAMP production. Moreover, both IGF-I and insulin enhanced the amounts of Gi protein(s) evaluated by pertussis toxin ADP-ribosylation or immunoblotting. Again, the effects of insulin at nanomolar concentrations were lower than those induced by the same concentrations of IGF-I or insulin at micromolar concentrations. These results suggest that, in bovine adrenal fasciculata cells, A-II receptors are coupled to the phosphoinositide pathway through pertussis toxin sensitive and insensitive Gp protein(s). Moreover, the findings also indicate that the enhanced A-II responsiveness of IGF-I or insulin treated cells is in part mediated through an increase in the amount of G protein(s).

Angiotensin II receptor subtypes AT1 and AT2 are down-regulated by angiotensin II through AT1 receptor by different mechanisms.

The regulatory effects of angiotensin II (AngII) on its receptor subtypes, AT1 and AT2, were studied using cultured bovine adrenal cells (BAC), which express both receptor subtypes, and PC12W and R3T3 cells, which express only AT2 receptors. In BAC, AngII caused a decrease in AT1- and AT2-binding sites and their corresponding messenger RNAs (mRNAs), but with different kinetics. AT1-binding sites decreased by more than 50% within the first 3 h, whereas AT1 mRNA started to decline after a lag period of 3 h. Both AT2-binding sites and mRNA remained stable within the first 6 h of AngII treatment. Then, AT2 mRNA decreased rapidly with an apparent half-life of 2-3 h, whereas AT2-binding sites declined with an apparent half-life of about 16 h. Measurement of transcription rate and mRNA half-life by the [3H]uridine-thiouridine method revealed that AngII reduced by 90% the rate of AT1 transcription, but had no effect on AT1 mRNA half-life, whereas it slightly reduced AT2 transcription, but markedly reduced AT2 mRNA stability. All of the effects of AngII on both AT1 and AT2 receptors were blocked by losartan, indicating that they were mediated exclusively through the AT1 receptor. In PC12W cells, AngII was unable to modify AT2-binding sites or mRNA. Moreover, in BAC, [125I]AngII was internalized through the AT1 receptor, whereas occupancy of AT2 receptors in either BAC or PC12W did not produce internalization of the hormone. These results indicate that AngII, through the AT1 receptor, down-regulates both AT1 and AT2, but by different mechanisms; AT1 receptor is regulated through internalization-degradation of the occupied receptor and inhibition of transcription, whereas AT2 receptor is regulated mainly by decreasing the stability of its mRNA. Moreover, the phorbol ester phorbol 12-myristate 13-acetate mimicked most of the effects of AngII in BAC and decreased both AT2-binding sites and mRNA on PC12W cells, indicating that the hormonal regulation of both AT1 and AT2 receptors is mediated through protein kinase C activation.

Fetal rat somatotropes in vitro: effects of insulin, cortisol, and growth hormone-releasing factor on their differentiation: a light and electron microscopic study.

Somatotropes first appear in the fetal rat pituitary just before term. These cells have never been detected in cultured fetal pituitaries. A modified culture medium has, however, enabled their differentiation in vitro. Hypophysial primordia were explanted on days 13-18 of gestation and cultured in different media until the equivalent of term. Immunoreactive somatotropes could be detected, by light and electron microscopy, in cultured primordia explanted on day 14 of gestation or later. The size and numbers of immunoreactive cells depended on culture medium composition. The control medium, containing insulin, cortisol, T3, and glucagon, proved favorable to somatotrope differentiation and proliferation. Increased insulin concentration reduced somatotrope numbers. In the presence of only insulin and cortisol (or corticosterone) somatotropes were more numerous than in the control. Culture medium enriched with insulin alone, with insulin and T3, or with insulin and glucagon, was not suitable for development of this cell type. Addition of GH-releasing factor ( GHRF ) to the medium during the first culture day did not accelerate the first appearance of the somatotropes but did significantly increase their size. GHRF addition 1/2 h before the end of culture did not modify their morphology. The ultrastructure of somatotropes in vitro is very similar to that observed in vivo on day 21 of gestation. The cells were characterized by their lamellar endoplasmic reticulum and immunoreactive secretory granules (300-400 nm maximal section diameter). Fetal somatotropes can, therefore, be successfully caused to differentiate in vitro. Their appearance depends on insulin and glucocorticoid concentration. T3 and/or glucagon may be inhibitory. GHRF may increase storage in somatotropes. These factors may also regulate the development of somatotropes in vivo.

Ovine adrenal fasciculata cells contain angiotensin-II receptors coupled to intracellular effectors but are resistant to the steroidogenic effects of this hormone.

Ovine adrenal fasciculata cells (OAC) responded to ACTH but were resistant to the steroidogenic action of angiotensin-II (A-II), while bovine adrenal fasciculata cells (BAC) responded to this hormone as well as to ACTH. However both cell types contained specific A-II binding sites (120,000 +/- 14,000 and 85,000 +/- 10,000 sites per cell for OAC and BAC, respectively) of similar high affinity [dissociation constant (KD) congruent to 2 x 10(-9) M]. Moreover, in both cell types, A-II receptors were coupled to intracellular effectors since A-II: 1) stimulated the accumulation of inositol phosphates, although the effects in BAC were higher than in OAC; 2) enhanced the influx and the efflux of 45Ca2+; 3) increased cytosolic free Ca2+ concentration ([Ca2+]i); 4) potentiated ACTH-induced cAMP production; and 5) induced A-II receptor loss. Both cell types appear to have an active protein kinase C since the phorbol ester 4 beta-phorbol 12-myristate-13-acetate potentiates ACTH-induced cAMP production and caused A-II receptor loss. In addition, 4 beta-phorbol 12-myristate-13-acetate and Ca2+ ionophore enhanced the steroid production by BAC but had no effect on OAC. These results indicated that the steroidogenic refractoriness of OAC to A-II might involve some step(s) beyond the initial activation of the two branches of the phosphoinositide pathway, activation of protein kinase C and increase of [Ca2+]i, and before conversion of cholesterol to pregnenolone.

Angiotensin II potentiates agonist-induced 3',5'-cyclic adenosine monophosphate production by cultured bovine adrenal cells through protein kinase C and calmodulin pathways.

Interactions between signal transducing systems may be important in the integrated control of cellular processes in basal and hormonally regulated cells. The cultured bovine adrenal fasciculata cell provides a model to study the interactions between the cAMP and calcium-sensitive phospholipid dependent protein kinase C. In this study, angiotensin II (A-II) and phorbol ester (PMA) potentiated the stimulatory actions of ACTH in a dose-dependent manner on cAMP production. At maximal concentrations, A-II and PMA also potentiated the effects of cholera toxin and forskolin on cAMP production. Both staurosporine, a protein kinase C inhibitor, and desensitization of protein kinase C by a 24-h pretreatment with PMA blunted the effect of PMA, but only partially inhibited (34%) the effect of A-II. Neither nifedipine, a specific calcium channel antagonist, nor pretreatment of cells with pertussis toxin modified the amplifying effects of A-II or PMA. In contrast, trifluoperazine, a calmodulin inhibitor, reduced the potentiating effect of A-II by about 35%, but association with staurosporine blunted its effects. Moreover, the steroidogenic effects of ACTH plus A-II were more than additive, but this synergism was blunted in the presence of both inhibitors. In conclusion, PMA and A-II potentiated agonist-induced cAMP production by bovine adrenal fasciculata cells. The data suggest that the effects of PMA were mediated exclusively by protein kinase C, whereas those of A-II were mediated by both protein kinase C and calmodulin.

Ontogeny of hypothalamic luteinizing hormone-releasing hormone (GnRH) and pituitary GnRH receptors in fetal and neonatal rats.

Although it is known that LH secretion starts at 17 days of gestation in the fetal rat and that this first LH release is most likely driven by hypothalamic GnRH, an earlier role for GnRH during fetal life has been postulated with the observation that presence of GnRH is important before day 13 of gestation for the differentiation of the pituitary anlage. In order to clarify the different roles of GnRH during fetal life, we have studied the first appearance of GnRH in the fetal brain, the expression of GnRH receptors in the fetal pituitary gland, and the presence of GnRH immunoreactivity within the fetal gonadotrophs. GnRH was present in the earliest brain tissue examined (12 days of gestation). From 12-17 days, GnRH content of fetal brain remained low and then increased markedly by the end of gestation. No immunoreactive GnRH-like material could be detected in rat placental tissue throughout gestation. Binding sites for GnRH were detected as early as 12 days of gestation in fetal pituitary glands. However, binding was very low until 16 days. At 17 days, Scatchard analysis indicated the presence of high affinity, low capacity binding sites [affinity constant (Ka) = 10(10) M-1]. Intracellular presence of GnRH as seen by immunocytochemistry using ultrathin sections prepared by cryoultramicrotomy was first visible at 14 days and started to increase at 16 days. LH was first detectable in the fetal pituitary by RIA at 17 days; FSH was first detectable at 21 days, and PRL at 1 day of postnatal life. Thereafter, neonatal pituitary contents of LH, FSH, and PRL increased linearly with-time, as did the number of pituitary GnRH receptors. At 10 days of postnatal life, pituitary contents of LH and FSH were significantly higher in females than in males. In summary, hypothalamic GnRH appears early in fetal life and potentially can induce differentiation of the pituitary anlage. Conversely, the presence at 15 days of gestation of specific binding sites for GnRH and of intracellular GnRH immunoreactivity in gonadotrophs indicates that the hypophysiotropic action of GnRH clearly precedes the start of LH biosynthesis.

Gonadotropin releasing hormone (GnrH) stimulates immunoreactive lactotrope differentiation.

To study lactotrope differentiation in the fetal rat, immunocytochemistry was performed on pituitary primordia explanted from 13-day-old fetuses and cultured in different synthetic media until the equivalent of 21 days. Lactotropes were detected only by antirat prolactin antiserum when the synthetic medium was enriched with GnRH (10(-9)M). These results indicate that lactotrope differentiation may partly depend on stimulatory factors such as GnRH.

Study of the alteration of gene expression in adipose tissue of diet-induced obese mice by microarray and reverse transcription-polymerase chain reaction analyses.

In the present study we developed a model of diet-induced obesity (DIO) in male C57 BL/6J mice using an 8-wk high fat diet. This model should better reflect the physiology of the majority of the human obese patients than mouse genetic models of obesity with defects in leptin or leptin signaling. At the end of the diet, DIO mice displayed an increased weight (20%) and higher leptin, insulin, glucose, and corticosterone plasma levels compared with mice fed a standard diet during the same period. Moreover, they became resistant to the central effect of peripheral administration of leptin. Oligonucleotide microarray studies were conducted in adipose tissue. They showed that a great number of genes are differentially expressed. The majority of these genes (69%) are down-regulated in DIO mice. Among those are genes encoding enzymes of the lipid metabolism or markers of adipocyte differentiation, enzymes involved in detoxification processes, as well as structural components of the cytoskeleton. Some other groups of genes displayed increased expression, such as those encoding inflammatory markers. The results of the microarray analysis were confirmed by semiquantitative RT-PCR studies run on a selected number of genes that were differentially expressed or not modified.

Evolution of lactotropes in normal and anencephalic human fetuses.

The presence of lactotropes in pituitaries removed from 15 normal fetuses, 1 normal premature infant, 3 normal infants, and 7 anencephalic fetuses or infants was studied by immunocytochemistry. These cells were first detected in normal fetuses at 18 weeks of gestation in both sexes. After 22 weeks of gestation, the number of cells increased sharply and regularly until term. Moreover, the cytoplasmic area of these cells increased significantly until term, but was decreased 4 and 5 months after birth. In all anencephalic fetuses, lactotropes were found; these cells were more numerous and often showed significantly greater cytoplasmic area than those in normal fetuses. A stimulatory effect of the fetal hypothalamus, therefore, is not essential for the development of this cell type.

Effects of transforming growth factor-beta 1 on human adrenocortical fasciculata-reticularis cell differentiated functions.

Transforming growth factor-beta 1 (TGF beta 1) has been reported to have a strong inhibitory effect on the specific function of adrenal cells of several species. In the present study, we examined the effects of TGF beta 1 on cultured human fasciculata-reticularis cells. TGF beta 1 alone had no effect on ACTH receptor messenger ribonucleic acid (mRNA) levels and was unable to reduce the strong stimulatory effects of ACTH on its own receptor. However, TGF beta 1 enhanced angiotensin-II type 1 receptor mRNA and binding sites. Treatment with TGF beta 1 increased significantly the levels of 3 beta-hydroxysteroid dehydrogenase mRNA, reduced those of cytochrome P-450 17 alpha-hydroxylase mRNA, and had no effect on cholesterol side-chain cleavage cytochrome P-450 mRNA. Whatever the experimental condition, TGF beta 1 did not reduce basal or ACTH-stimulated cortisol production, but the production of dehydroepiandrosterone sulfate of TGF beta 1-treated cells was always decreased. The effects of TGF beta 1 on 3 beta-hydroxysteroid dehydrogenase mRNA and dehydroepiandrosterone sulfate were opposite the change observed at the time of adrenarche. As adrenal cells express TGF beta 1 mRNA, it is tempting to postulate that a local diminution of TGF beta 1 might be involved in adrenarche. Our findings also illustrate the specific species differences and, therefore, the caution to extrapolate to humans the results observed in other species.