Splanchnic extraction and interconversion of testosterone and androstenedione in man.

A constant infusion of (3)H-testosterone and (14)C-androstenedione was administered to four human subjects, two males and two females, until the concentrations of radioactive testosterone and androstenedione in systemic plasma became constant. At that time the concentrations of radioactive testosterone and androstenedione in hepatic vein plasma were determined. Splanchnic extraction of testosterone and androstenedione and the contribution of the splanchnic system to the blood interconversion of testosterone and androstenedione were calculated. Androstenedione is extracted by the splanchnic system more efficiently than testosterone since 82.3% of androstenedione and 44% of testosterone were removed from the plasma after one passage. The fraction of testosterone entering the splanchnic system that is transferred to blood androstenedione was 0.011 and the maximum possible transfer due to recirculation was 0.026. This was 28% of the total blood transfer from testosterone to androstenedione. The fraction of androstenedione entering the splanchnic system that is transferred to blood testosterone after one passage was 0.005, whereas the maximum possible transfer in this system was 0.006. This represented only 16% of the total transfer from androstenedione to testosterone. Therefore, a large fraction of the interconversion of testosterone and androstenedione in vivo occurs outside the splanchnic system.

Familial gynecomastia with increased extraglandular aromatization of plasma carbon19-steroids.

We evaluated a family in which gynecomastia occurred in five males in two generations. In each affected subject, gynecomastia and male sexual maturation began at an early age. The ratio of the concentration of plasma estradiol-17 beta to that of plasma testosterone was elevated in each affected subject. In the three siblings with gynecomastia, the transfer constant of conversion of androstenedione to estrone (i.e., the fraction of plasma androstenedione that was converted to estrone as measured in the urine) was 10 times that of normal persons. The transfer constant of conversion of testosterone to estradiol-17 beta in the one subject studied also was 8-10 times that of normal men, whereas the transfer constants of conversion of estrone to estradiol-17 beta and of estradiol-17 beta to estrone were normal. Despite the elevation in extraglandular aromatase activity, there was a normal response of the hypothalamic-pituitary axis to provocative stimuli. This is the second documentation of gynecomastia that is associated with increased extraglandular aromatase activity, and the first time that the defect was found to be familial with a probable X-linked (or autosomal dominant, sex limited) mode of inheritance.

Plasma aldosterone concentration at delivery and during the newborn period.

Aldosterone concentrations in plasma of women on normal sodium intake undergoing cesarean section were 3.7+/-1.4 ng/100 ml (mean+/-1 SD). These values were significantly lower (P < 0.001) than those observed in mothers on normal sodium diet, delivered by the vaginal route (14.9+/-7.0 ng/100 ml). A significant elevation (P < 0.001) of the concentrations was found if the mothers had been on sodium restriction and/or diuretics (44.9+/-24.2 ng/100 ml). In supine position, adult nonpregnant subjects have aldosterone concentrations in plasma of 1.7+/-1.4 ng/100 ml on normal sodium intake and of 16.7+/-8.1 ng/100 ml on low sodium diet.Simultaneous determinations of aldosterone levels in cord blood showed that cord values were significantly higher than those of the corresponding mother (P < 0.01 by paired t test). However, values in cord blood of infants born to mothers on a normal sodium intake were significantly lower (P < 0.005) than those of infants whose mothers had required low sodium diet and/or diuretics during their pregnancy. Aldosterone concentrations in plasma of infants 1-72 hr of age and born to mothers on normal sodium intake were 25.9+/-11.7 ng/100 ml (mean +/-1 SD). These values were significantly lower (P < 0.005) than those of infants born to mothers on restricted sodium intake with or without diuretics (80.3+/-54.4 ng/100 ml). The concentrations at birth were not significantly different from those observed during the first 3 days of life (P > 0.6).

The regulation of aldosterone secretion in anephric man.

The regulation of aldosterone secretion in anephric man was investigated in studies on nephrectomized patients who were being intermittently hemodialyzed while awaiting renal transplantation. The effects of supine and upright posture on the concentration of plasma aldosterone on the 1st day postdialysis and on a 3rd or 4th day postdialysis were compared to the effects of postural variation in normal subjects who were on a low sodium intake and on a high sodium intake. In contrast with the normal subjects who exhibited higher concentrations of plasma aldosterone after 2 hr of upright posture than in the supine position and low concentrations of plasma aldosterone on a high sodium intake, the anephric patients showed less consistent variations in plasma aldosterone due to changes in posture and exhibited higher concentrations of plasma aldosterone on the 3rd or 4th day postdialysis, despite an increase in body weight, than on the 1st day postdialysis. The increase in the concentration of plasma aldosterone in the anephric patients between the 1st day postdialysis and the 3rd or 4th day postdialysis indicates that aldosterone secretion is not responding primarily, in this situation, to volume-related stimuli. There was a high degree of correlation between the concentration of plasma aldosterone and the corresponding levels of serum potassium concentration, which also rose significantly between the 1st day postdialysis and the 3rd or 4th day postdialysis. Furthermore, when potassium accumulation between dialyses was prevented in three of these patients, the concentration of plasma aldosterone fell to minimally detectable levels. The results of these studies suggest that the primary regulator of aldosterone secretion in the absence of the kidneys is potassium.

Transplacental passage and fetal secretion of aldosterone.

The transplacental passage and the production of aldosterone were studied in late pregnancy during a constant infusion of 1,2-aldosterone-(3)H to mothers at the time of elective cesarean section. It was found that, while maternal aldosterone crossed the placenta, there was a significant secretion of aldosterone by the fetus. The aldosterone concentration in fetal plasma was 2-12 times higher than that of the corresponding mothers. Pregnancy had no effect on the metabolic clearance rate of aldosterone, but it increased the rate of production of this steroid. However, the increments that we observed were smaller than those reported in previous reports. The discrepancy was probably due to differences in body posture, our subjects being supine for at least 10 hr at the time of study.

Genetic analysis of familial isolated growth hormone deficiency type I.

Nuclear DNA from individuals belonging to nine different families in which two sibs were affected with isolated growth hormone deficiency type I were studied by restriction endonuclease analysis. By using 32P-labeled human growth hormone or the homologous human chorionic somatomammotropin complementary DNA (cDNA) sequences as a probe, the growth hormone genes of affected individuals from all families yielded normal restriction patterns. Polymorphic restriction endonuclease sites (HincII and MspI), which are closely linked to the structural gene for growth hormone on chromosome 17, were used as markers in linkage analysis of DNA of family members. Of the nine affected sib pairs two were concordant, three were possibly concordant, and four were discordant for both linked markers. Since only concordant sib pairs would have inherited the same growth hormone alleles, further studies to identify mutations of the growth hormone genes should be limited to this subgroup. It is unlikely that the discordance observed in four of the sib pairs is due to recombination, because the polymorphic HincII site is only 116 base-pairs from the -26 codon of the growth hormone gene. Thus, in at least four of the nine families, the mutation responsible for isolated growth hormone deficiency is not within or near the structural gene for growth hormone on chromosome 17.

Exon 7 Ncol restriction site within CYP21B (steroid 21-hydroxylase) is a normal polymorphism.

A point mutation within exon 7 producing an amino acid coding change and a recognition site for the endonuclease Ncol has been reported in the HLA-Bw47-linked CYP21A pseudogene and some mutant CYP21B (steroid 21-hydroxylase) genes of patients with congenital adrenal hyperplasia (CAH). Whether this mutation is deleterious was not demonstrated. We analyzed DNA from various subjects for the presence of the exon 7 Ncol site: group 1, 10 normal subjects; group 2, 11 patients with salt-losing CAH; and group 3, 18 members of an Amish pedigree in which 10 expressed HLA-Bw47 not linked to CAH. Southern blots of Ncol-digested genomic DNA which were hybridized with CYP21 cDNA showed that four subjects of group 1 had a heterozygous Ncol pattern. In group 2, seven patients had the Ncol site; two of them were homozygous for the site and had deletions of both CYP21B genes. The other five were heterozygous for the Ncol site, which was linked to a CYP21B deletion and a HLA-Bw47 haplotype. In group 3, no one exhibited the exon 7 Ncol site. To map the Ncol sites to CYP21A or CYP21B in the normal subjects, DNA from the four Ncol heterozygous subjects was double digested with Ncol and Mbol and hybridized with CYP21 cDNA. Ncol-Mbol fragments unique to CYP21A were identified in all four, but the smaller CYP21B-specific fragments were not detected. Their genomic DNA in the region of exon 7 (bases +1167 to +2058) was then amplified, cloned, and sequenced.(ABSTRACT TRUNCATED AT 250 WORDS)

Functional characterization of naturally occurring mutant androgen receptors from subjects with complete androgen insensitivity.

Mutations in the androgen receptor (AR) are thought to cause complete androgen insensitivity (CAIS) in 46,XY human subjects who have a female phenotype despite normal adult male concentrations of plasma testosterone. Assays of AR binding in cultured skin fibroblasts from subjects with CAIS show either an apparent absence of AR (AR-) or normal levels of AR (AR+) binding. In several subjects with CAIS, AR-, no gross AR mutation was detected by Southern blot analyses of genomic DNA and normal sized 10 kilobase mRNA was present on Northern blots of poly(A+) RNA from cultured genital skin fibroblasts. We have used the polymerase chain reaction to amplify individual exons within the human AR gene of subjects with CAIS and have identified point mutations in three subjects. In one AR- subject (R774C), amino acid 774 was changed from arginine (CGC) to cysteine (TGC), in another AR- subject (R831Q), arginine (CGA) was changed to glutamine (CAA) at position 831, and in an AR+ subject (V866M) a methionine (ATG) was substituted for valine (GTG) at position 866. Transfection of wild type and mutant AR cDNA clones into COS cells results in detection of AR protein by immunoblotting. AR ligand binding activity is absent in cells transfected with AR mutants R774C and R831Q, but present with AR mutant V866M. Androgen binding in cells transfected with AR mutant V866M has a 6-fold lower apparent binding affinity than that of wild-type AR. Transcriptional activation of the MMTV-CAT reporter gene was androgen dependent and specific and nearly maximal at physiological concentrations (10(-10) M) of androgen when wild-type AR was transfected into cells, whereas neither AR mutants R774C nor R831Q were able to stimulate CAT activity even at 10(-8) M androgen. AR mutant V866M was able to stimulate CAT activity but the androgen dose dependency was shifted toward pharmacological concentrations of steroid that exceed in vivo levels. The molecular basis of CAIS in humans exhibits genetic heterogeneity. Our study shows that some cases of CAIS are explained by an inability to form a functional AR-steroid complex and hence, the AR is unable to activate transcription of genes essential for male sex differentiation during fetal development.

Androgen binding in nuclear matrix of human genital skin fibroblasts from patients with androgen insensitivity syndrome.

Specific sex steroid-binding sites are associated with the salt-insoluble nuclear matrix from which lipids, histones, and chromatin have been extracted. In intact cultured normal human genital skin fibroblasts incubated for 1 h at 37 C with a saturating concentration (2 nM) of [3H]dihydrotestosterone [( 3H]DHT), approximately 50% of the total intracellular androgen receptor-steroid complexes were found in the nucleus. Within isolated nuclei from such cells, 28-49% of the specific androgen receptor binding was associated with the nuclear matrix. The antiandrogen cyproterone acetate inhibited DHT binding within the nuclear matrix. Cultured genital skin fibroblasts from two unrelated patients with receptor-positive complete androgen insensitivity (CAIS, AR+), had normal (approximately 50%) nuclear binding of DHT, and 35% and 45% of it was localized to the nuclear matrix. Genital skin fibroblasts from a patient with receptor-negative complete androgen insensitivity (CAIS, AR-) had no specific DHT binding in isolated nuclei or nuclear matrix. Scatchard analysis of specific DHT binding in the nuclear matrix isolated from cells of normal subjects after an in vitro exchange assay (0 C; 24 h) revealed the presence of saturable (maximum binding, approximately equal to 200 fmol/mg nuclear DNA), high affinity (Kd approximately equal to 1.0 nM) binding sites. By contrast, in the nuclear matrix isolated from cells of a patient with CAIS, AR+, the binding affinity for DHT was 3-fold lower (Kd approximately equal to 3.0 nM). When cytosolic androgen receptor-DHT complexes prepared from cells preincubated at 37 C for 1 h with [3H]DHT were incubated at 0 C for 1 h with isolated nuclei and nuclear matrix in the presence of 0.15 M KCl, 40-60% of specific nuclear binding was associated with the nuclear matrix. In these cell-free in vitro experiments, radiolabeled DHT-receptor complexes prepared from normal or mutant cells were mixed with isolated nuclei and nuclear matrix prepared from cells of normal subjects or patients with CAIS, AR+ or CAIS, AR-. Under these conditions, specific DHT binding in nuclei and nuclear matrix was quantitatively similar in the presence of a mutant (CAIS, AR+) receptor-steroid complex or in the presence of nuclei or nuclear matrix from the mutant cells (CAIS, AR- or AR+) when compared simultaneously with the same subcellular fractions prepared from the cells of normal subjects.(ABSTRACT TRUNCATED AT 400 WORDS)

Puberty in boys: correlation of plasma levels of gonadotropins (LH, FSH), androgens (testosterone, androstenedione, dehydroepiandrosterone and its sulfate), estrogens (estrone and estradiol) and progestins (progesterone and 17-hydroxyprogesterone).

Mean serum concentration dehydroepiandrosterone (DHA), DHA sulfate (DHAS), progesterone (P), 17-hydroxyprogesterone (17-OH-P), estrone (E1), estradiol (E2), and androstenedione (A) were compared from 43 boys followed longitudinally for as long as 4 yr during puberty. These data were also compared with serum levels of LH, FSH, and testosterone. Elevation is recognized early in puberty for DHAS, late in puberty for P and A, and gradually throughout puberty for E1, 17-OH-P, and DHA. When compared by age, the same general pattern is apparent with adult levels of E1 reached at age 12, DHAS and E2 by 13, and DHA, P, 17-OH-P, and A not until after age 15. Significant elevations of DHA occurred with the onset of pubic hair and voice change; elevations of DHAS occurred with the onset of genital and axillary hair growth; and testosterone increased with the onset of genital and pubic hair growth and voice change.

Integrated concentration of plasma aldosterone in normal subjects: correlation with cortisol.

The integrated concentration (I. C.) of plasma aldosterone and cortisol was determined every 30 min during a 24-h period, using a blood collection system composed of a nonthrombogenic catheter and a small, portable withdrawal pump. The experiments were carried out in 8 normal adult men during daily routine life, and repeated in 2 of the subjects while recumbent in bed most of the day. The following conclusions were made: a) The 30-min I. C. of aldosterone fluctuated widely throughout the day. b) Although peaks of increased concentration occurred after a change in posture from supine to erect, there were many peaks of concentration that occurred during the supine posture. The 24-h I.C. of aldosterone in 2 subjects that were studied a second time while staying in bed most of the day was lower than the I. C. observed during normal activity. Furthermore, there was a significant correlation between 24-h I. C. and percentage of time spent in supine position. c) A weak, but significant correlation was found between the 30-min I. C. of aldosterone and cortisol in 4 out of the 7 subjects tested. The overall correlation for all experiments was also significant (R = 0.3. p smaller than 0.001). D) The 4-h I. C. of aldosterone and cortisol showed that the lowest mean values were between 4 PM and 4 AM and the highest values between 4 AM and 4 PM for both steroids.

Preferential binding of testosterone over epitestosterone by human plasma.

The main purpose of this study was to investigate if the specificity of the binding of testosterone to plasma proteins could be defined as a preferential binding of this steroid over epitestosterone. The amount of testosterone that is specifically bound was calculated using the formula: (see article) concentration, where "Ri" is the ratio (14C) testosterone: (3H) epitestosterone in plasma prior to centrifugation, "Ru" is the isotope ratio in the protein-free supernatant obtained after ultracentrifugation (149,000 x g, at 0 C, for 18 h) and "T concentration" is the testosterone concentration in plasma resulting from addition of (14C) testosterone, the endogenous steroids having been removed by preliminary charcoal extraction. The theoretical separation of the binding sites for testosterone into two populations, one non-specific with no preference for testosterone over epitestosterone, and another with absolute specificity for testosterone over its physiologically inactive stereoisomer, proved to be useful. Ovalbumin was found to be an example for non-specific, non-preferential binding. Determination of the ratio (14C) testosterone: (3H) epitestosterone in the successive fractions of various ultracentrifuged preparations showed a small but significant preference for (14C) testosterone by human and bovine serum albumin, while alpha1-acid glycoprotein had a preference for (3H) epitestosterone. Saturation curves showed at least two components: the first one, presumably corresponding to TeBG, had a higher affinity and lower capacity. This binding capacity can be accurately determined by extrapolation to the ordinate or the second component, a straight line corresponding to a binding of somewhat lower affinity and much larger capacity.

Effect of food intake on the metabolic clearance rate of aldosterone.

The effect of eating on the metabolic clearance rate (MCR) of aldosterone was investigated in 10 adult individuals. 3H-aldosterone was infused continuously over a period of 5 h while the subjects remained supine. Three h after the start of the infusion, each subject ate a bowl of soup. The MCR of aldosterone, before and after the intake of food, was calculated by dividing the rate of infusion of 3H-aldosterone by the mean concentration of under the same conditions but without eating. The MCR of aldosterone (mean +/- 1 SD) was 1284 +/- 513 L/24 h before food intake and 2182 +/- 180 L/24 h after food intake in the 10 individuals who ate. The MCR in the 11 subjects who did not eat was 1363 +/- 446 and 1357 +/- 434 during the same periods (p greater than 0.05). The 29% increase in the MCR induced by eating was highly significant (p less than 0.001); it was similar in magnitude and duration to a previously reported effect of food intake on the hepatic blood flow (13).

Time-dependent biphasic response of aromatase to dexamethasone in cultured human skin fibroblasts.

Human genital skin fibroblasts grown in cell culture possess aromatase activity and, therefore, provide a model to investigate the molecular mechanisms that control aromatase in extraglandular tissues. Following the observation by other investigators that glucocorticoids stimulated aromatase activity in cultured stromal-vascular cells from adipose tissue, we examined the influence of dexamethasone (DEX) on aromatase in cultured skin fibroblasts. Preincubation of skin fibroblasts with DEX stimulated aromatase expression in all cell strains. In time-course studies, aromatase activity showed a biphasic curve, with peak levels at 12 h and a return to baseline levels by 72 h. When DEX was removed after 12 h, aromatase activity could be completely restimulated by DEX only after a period of 60-72 h. The DEX stimulation appeared to involve glucocorticoid receptor function, since the concentration of DEX required for half-maximal stimulation of aromatase activity (4.2 nM) was similar to the dissociation constant (Kd, 4.3 nM) of the receptor (for DEX). Actinomycin D and cycloheximide (CHX) inhibited DEX stimulation of aromatase when they were present in the preincubation and assay media. When cells were preincubated with DEX and CHX and then washed free of CHX and DEX before the assay, superinduction of aromatase activity occurred. Our data concerning the time course and superinduction of aromatase activity by DEX are in contrast to the findings reported by others for adipose tissue stromal-vascular cells and suggest that the mechanisms for the control of aromatase in extraglandular tissue may vary significantly in different tissues.

Induction and superinduction of messenger ribonucleic acid specific for aromatase cytochrome P-450 in cultured human skin fibroblasts.

Aromatase cytochrome P-450 (cytochrome P-450AROM) catalyzes the formation of aromatic C18 estrogenic steroids from C19 androgens. Several studies indicate that cytochrome P-450AROM activity is induced by glucocorticoids such as dexamethasone (DEX) and superinduced by DEX plus cycloheximide (CHX). We have used cultured human skin fibroblasts as a model system to investigate the regulation of aromatase gene expression. Whereas Northern blot analysis of total cellular RNA or poly (A)+ RNA from untreated strains of normal human skin fibroblasts failed to demonstrate any hybridization with a specific human placental cytochrome P-450AROM complementary DNA, analysis of RNA from cells treated with DEX demonstrated hybridization of the cytochrome P-450AROM complementary DNA to two transcripts of about 2.5 and 3.0 kilobases. Incubation of cells with DEX plus CHX resulted in a further increase in levels of cytochrome P-450AROM messenger RNA (mRNA) when compared to cells treated with DEX alone, suggesting that inhibition of protein synthesis superinduces transcription of the cytochrome P-450AROM gene. By contrast, levels of beta-actin mRNA were not affected by treatment with DEX and CHX. Treatment of cells with CHX alone did not produce a change in either aromatase activity or levels of cytochrome P-450AROM mRNA transcripts. These results indicate that aromatase activity is regulated by changes in the concentration of cytochrome P-450AROM mRNA, and imply that control of cytochrome P-450AROM gene expression is at the level of gene transcription. We conclude that the cytochrome P-450AROM gene is regulated by a complex mechanism that includes both positive and negative transcription factors.

Lack of correlation between gonadotropin and adrenal androgen levels in agonadal children.

To evaluate the relationship between the secretion of gonadotropins and adrenal androgens, patients with gonadal agenesis were evaluated by (a) administering human luteinizing hormone (hLH) for 5 days with or without estrogen pretreatment to agonadal patients who had prepubertal LH levels; (b) correlating circulating gonadotropin levels with adrenal androgens in 45 patients; and (c) comparing adrenal androgens with gonadotropins after long-term administration of estrogen or androgens. Results are as follows: (a) No alteration in serum concentrations of dehydroepiandrosterone (DHA), dehydroepiandrosterone sulfate (DHAS), estrone (E1), testosterone (T), or in excretion of urinary 17-ketosteroid (17 KS) occurred after the administration of hLH. (b) No clearcut relationship between endogenous level of LH or FSH and DHA OR DHAS was demonstrated although a coincident increase of all hormones with age occurred. (c) Administration of estrogen to patients with gonadal agenesis did not affect their levels of DHA and DHAS although those patients given androgen developed higher DHAS, but not DHA, levels. Hence, increasing gonadotropin concentrations would not appear to be a primary etiologic factor in the maturation of the adrenal.

Ascertainment of 21-hydroxylase deficiency in individuals with HLA-B14 haplotype.

21-Hydroxylase activity was measured in North American caucasian individuals with the HLA-B14 antigen to estimate the frequency of heterozygosity for the attenuated congenital adrenal hyperplasia trait. A 30-min iv ACTH stimulation test was administered to 9 normal HLA-B14-positive subjects and to a comparable HLA-B14-negative control group. Changes in plasma progesterone and 17-hydroxyprogesterone over 30 min were summed and expressed as a combined rate of rise. Six of 9 HLA-B14-positive individuals had a rate of rise greater than 2 SD above the mean control value. On this basis, about two thirds of B-14-positive individuals are heterozygote carriers for 21-hydroxylase deficiency. Thus, the frequency of the attenuated form of congenital adrenal hyperplasia linked to the HLA-B14 locus in women is approximately 1 in 6000 if there is only 1 B14, and 1 in 2000 if there are 2 B14s in the HLA type.

Human complete androgen insensitivity with normal dihydrotestosterone receptor binding capacity in cultured genital skin fibroblasts: evidence for a qualitative abnormality of the receptor.

Complete androgen insensitivity syndrome (CAIS), or so-called testicular feminization, results from the lack of androgen action on target organs. Within this syndrome, two major variants have been described. In the first variant, the specific intracellular androgen receptors are undetectable (CAIS, AR-), whereas normal levels of androgen receptors are measured in the second variant (CAIS, AR+). From studies with cultured labial skin fibroblasts of three CAIS, AR+ patients from the same family, we have demonstrated that their androgen receptors present qualitative differences when compared to normal receptors: 1) the apparent affinity constant (Kd) of 5 alpha-dihydrotestosterone for the receptor is higher than normal; 2) the in vitro dissociation rate of the receptor-steroid complex is faster than normal: 3) the cellular androgen receptor is more thermolabile than normal when cells are exposed to superphysiological temperatures; and 4) the relative binding affinity of the androgen receptor for progesterone is greater than normal. These findings suggest the presence of a structural abnormality of the androgen receptor in patients with CAIS, AR+. However, these changes (e.g. slightly decreased affinity of the receptor for 5 alpha-dihydrotestosterone) probably do not explain the total lack of androgen action in these patients. Finally, the present data from this family along with those from the literature suggest the presence of heterogeneity as to the cause of the defect in CAIS, AR+.

The pituitary-gonadal axis in men with protein-calorie malnutrition.

The pituitary-gonadal axis was studied in 28 men with severe protein-calorie malnutrition in a Calcutta hospital. The men were selected for the severity of their malnutrition and for absence of other diseases. They had clinical findings of hypogonadism and low total and unbound plasma testosterone. During 2-5 months of refeeding there was clinical recovery and increase of plasma testosterone to normal. Plasma LH was high in malnutrition, decreased during refeeding, and remained above normal after refeeding. Some patients failed to show LH elevation in malnutrition despite low plasma tesosterone. Plasma FSH was high in malnutrition, decreased during refeeding, and was near the level of normal Indian men after refeeding. HCG 4000 IU im per day for 3 days produced subnormal increments in plasma testosterone both in malnutrition and after refeeding; corresponding decreases in FSH occurred. It is concluded that the hypogonadism of protein-calorie malnutrition is primarily on the basis of diminished Leydig cell function. Appropriate pituitary LH response is intact in some patients, but is either absent or inadequate in others. Subclinical Leydig cell insufficiency, indicated by LH elevation and subnormal response to HCG, persists after refeeding has produced recovery from malnutrition and clinical hypogonadism. FSH elevation in malnutrition may be secondary to the reduced Leydig cell function.

Endocrine features of an adrenal-like tumor of the ovary.

Detailed in vivo endocrine studies of a virilizing adrenal-like tumor of the ovary are reported. A nonspecific steroidogenic baseline profile was observed in the face of clearcut histological adrenal-like features. Dynamic testing disclosed incomplete dexamethasone-related steroidogenic suppression suggestive of partial tumoral autonomy. At the same time, the observation of sizable dexamethasone-related steroidogenic decrements, unaccounted for by normal adrenal contribution, may be interpreted to suggest a direct, pharmacological, inhibitory effect of dexamethasone on tumoral and/or ovarian stromal steroidogenesis. These observations confirm the limited role of baseline steroid profiling and dynamic testing in the preoperative localization of androgen-secreting tumors. Reversible dysfunction of hypothalamic gonadotropic feedback mechanisms was suggested by combined single dose LRH-clomiphene testing. Complete resolution followed tumor removal.