Immunoreactive inhibin levels during ovarian stimulation may predict granulosa cell maturity

OBJECTIVE: The aim of the study was to assess whether the immunoreactive inhibin response to ovarian stimulation in polycystic ovarian syndrome was of predictive value for the outcome of ovulation induction. DESIGN: Daily injections of purified FSH (Metrodin, Serono Laboratories, UK) were administered for the purpose of inducing development and ovulation of a single follicle. PATIENTS: All patients had anovulatory infertility secondary to polycystic ovarian syndrome and were resistant to clomiphene citrate. MEASUREMENTS: Alternate day serum samples were obtained for measurement of gonadotrophins, sex steroids and inhibin by radioimmunoassay. Alternate day ovarian ultrasound scans were carried out to monitor follicular development. RESULTS: There was a high incidence of multiple follicular development (MFD) (10.2 +/- 7.1 (standard deviation) follicles). There was a close relationship between the number of follicles on the day of the ovulatory trigger and serum oestradiol (E2) levels (R = 0.97, P < 0.001). This relationship was not seen with serum immunoreactive inhibin levels (R = 0.31). The cycles were divided into two groups depending whether there was or was not MFD (more than five follicles > 7 mm in diameter on day of hCG administration). The MFD group had 15.6 +/- 8.5 follicles and the non-MFD group 3.2 +/- 1.6 follicles. There was no difference in the circulating E2 concentration (pmol/l) per follicle between the two groups (non-MFD cycles, n = 6,345 +/- 28.9; MFD cycles, n = 8, 308 +/- 40.29). However, the immunoreactive inhibin concentration (U/ml) per follicle was lower, P < 0.001, in the MFD group (4.3 +/- 1.6 vs 6.9 +/- 0.5). The maximum and mean follicle diameter were lower, P < 0.01, in the MFD group (maximum follicle size 16.5 +/- 3.9 vs 21.2 +/- 1.5 mm, mean follicle size 11.5 +/- 2 vs 14.9 +/- 2.3). When individual cycles were examined E2 and immunoreactive inhibin secretion rose in parallel in cycles with less than five follicles with a rapid rise occurring when the follicles reached about 12 mm in diameter. In cycles where there was MFD there was a disparity between E2 and immunoreactive inhibin secretion with E2 levels rising 4.3 +/- 1.4 days before immunoreactive inhibin levels. CONCLUSION: These data suggest that in cycles where there are multiple small follicles, E2 secretion is maintained whereas immunoreactive inhibin secretion is substantially lower. Thus, in view of the disparity between E2 and immunoreactive inhibin secretion evident in the MFD group, measurement of immunoreactive serum inhibin concentration may be a better indicator of granulosa cell maturity. Immunoreactive inhibin secretion may occur only from healthy mature follicles.     

Immunoreactive inhibin concentrations in adult men: presence of a circadian rhythm

The circadian and chronological changes in inhibin secretion were studied in normal adult men. To determine the diurnal release of inhibin, blood samples were collected every hour for 24 h from five healthy young adult men, and serum inhibin was measured by RIA. During the evening and the night, serum inhibin concentrations were relatively low; the lowest value was observed at 2200 h. At 0700 h, inhibin started to increase, reached a peak at 0900 h, and then gradually decreased. These results suggest that inhibin secretion is circadian. Testosterone and cortisol also showed circadian rhythms. In some of the five volunteers, the serum concentrations of inhibin, testosterone, and cortisol were superimposable, but no significant relationship was observed between the serum inhibin and FSH concentrations. With regard to the age-related changes in basal inhibin levels, the highest values were observed in the twenties, and lower values were found with aging. This relationship suggests that increased FSH in elderly men might be due to the reduced amount of peripheral inhibin.     

Serum inhibin B levels in community-dwelling elderly men

BACKGROUND: AND OBJECTIVE: Ageing in men is accompanied by a decline of Leydig cell function, with a 50% decrease of the population means for serum free testosterone between age 25 and 75 years. Information on Sertoli cell function and spermatogenesis in the elderly is scarce. Studies on seminal parameters in ageing men have suggested that spermatogenesis may be fairly well maintained in the elderly, but they included mostly selected subjects and only few men over 60 years. More systematic studies are lacking. The aim of the present study was to assess serum inhibin B levels in elderly men as an index of global Sertoli cell function and spermatogenic activity. SUBJECTS AND MEASUREMENTS: Specific immunoassays were used to determine serum levels of inhibin B, gonadotrophins, testosterone and oestradiol in blood obtained between 0800 and 1000 h. from 189 ambulatory, community-dwelling elderly men (age: 70-85 years) and, for comparison, from 51 middle-aged (35-54 years) and 50 young (< 35 years) controls. RESULTS: All age groups combined, serum inhibin B was only weakly negatively correlated to age (Spearman correlation coefficient: - 0.17; P < 0.01) and more strongly to serum FSH (- 0. 52; P < 0.001). In a multiple regression analysis serum FSH, but not age or serum free testosterone, emerged as an independent determinant of serum inhibin B levels. An age-related decline of median inhibin B levels in the study population was essentially limited to the younger age groups, with stable levels between age 35 and 79 years, and only a modest further decrease thereafter. There was a progressive age-related increase of serum FSH across age groups with, consequently, a marked decrease of the serum inhibin B : FSH ratio. The prevalence of men presenting with low serum inhibin B (below 10th percentile for inhibin B levels in men < 35 years), indicative of deficient Sertoli cell function and spermatogenesis, increased most strikingly between men < 35 years and those 35-54 years, which contrasts with the more progressive increase at an older age of the prevalence of low serum (free) testosterone. CONCLUSION: Global testicular Sertoli cell function and spermatogenic activity, as assessed indirectly through serum inhibin B levels, appear to be well maintained in ambulatory elderly men, albeit there are age-related alterations at the level of the Sertoli cells as indicated by a progressive increase of testicular drive by pituitary FSH.     

Circulating levels of inhibin in cancer

Study on frequency distribution of serum inhibin levels--a peptide hormone involved in suppression of follicle stimulating hormone--was carried out using a specific radioimmunoassay developed in this laboratory, on 237 sera from normal men and women, from male and female patients with cancer prostate, lung, stomach and breast, and from patients with non-malignant disease of the respective organ. The mean serum inhibin levels in patients with malignant and non-malignant disease of prostate (90 +/- 19 ng/ml and 54 +/- 8 ng/ml) were significantly higher than found in normal men (31 +/- 9 ng/ml). However, the levels in cancer of stomach (70 +/- 9 ng/ml) and lung (45 +/- ng/ml) were higher than those in normal as well as benign condition of stomach and lung. The mean serum inhibin in case of women with cancer of breast (27 +/- 5 ng/ml) and lung (53 +/- 16 ng/ml) were significantly higher than those found in normal women (15 +/- 2 ng/ml) and in women with benign disease of breast (10 +/- 3 ng/ml) and lung (21 +/- 4 ng/ml). Considering the levels in individual patient basis, 66% of the patients with cancer of stomach had higher inhibin concentration than the uppermost limit of inhibin found in normal men and in contrast none in benign disease of stomach.     

Serum inhibin levels in normal men and men with testicular disorders

Serum concentrations of inhibin, FSH and LH were measured in 39 normal men and 127 men with testicular disorders resulting in infertility. The infertile men were divided into groups on the basis of their mean sperm count, FSH levels and karyotype. The mean (+/- S.D) serum concentrations of inhibin in the normal men was 554 +/- 156 U/l and did not differ significantly from those groups with oligospermia, azoospermia or Klinefelter's syndrome. Combined analyses of all groups did not reveal any significant correlation between serum concentrations of inhibin and FSH or with any other parameter measured. Serum concentrations of FSH and LH were positively correlated, and Leydig cell dysfunction, as evidenced by increased serum LH levels, low testosterone levels or a declining testosterone/LH ratio were found with severe spermatogenic damage. The failure of serum concentrations of inhibin to correlate with those of FSH levels or the degree of testicular damage raise questions as to the clinical value of this parameter alone.     

Immunoreactive alpha-MSH and ACTH levels in rat plasma and pituitary.

A radioimmunoassay is described for the measurement of alpha-melanocyte-stimulating hormone (alpha-MSH). The antibody was produced in rabbits by immunization with alpha-MSH coupled to bovine serum albumin with carbodiimide. The antibody did not react significantly with ACTH, beta-MSH, or 6 fragments of ACTH. The sensitivity and reliability of the assay were improved by employing a simple plasma extraction procedure. When applied to a 2 ml plasma sample, the detection limit of the radioimmunoassay was 6 pg/ml. ACTH was measured with a sensitive and specific radioimmunoassay previously described for humans and adapted for the rat. The anti-ACTH serum cross-reacted with the biologically active portion of alpha-p ACTH and not with alpha-MSH, beta-MSH or the alpha-p 17-39 and alpha-p 25-39 fragments of ACTH. The detection limit was 20 pg/ml. Plasma and pituitary alpha-MSH and ACTH had the same immunoreactivity as synthetic alpha-MSH and ACTH. alpha-MSH and ACTH contents of the rat neurointermediate lobe were 1398 +/- 360 (SE) ng and 28.2 +/- 2.9 ng, respectively, while in the anterior lobe they were 102 +/- 31 ng and 551 +/- 36 ng, respectively. The plasma alpha-MSH concentration at 8 AM in male rats was 64 +/- 8 pg/ml when the plasma ACTH concentration was 92 +/- 15 pg/ml. Over a 24-hour period two peaks of plasma alpha-MSH were observed, one at 4 AM (142 +/- 35 pg/ml) and the other at 4 PM (139 +/- 26 pg/ml). Plasma ACTH was higher at noon (151 +/- 43 pg/ml) and 4 PM (130 +/- 48 pg/ml). Short-term exposure to ether induced a transient increase in alpha-MSH level 5 min later and a rapid return to normal levels. Plasma ACTH increased significantly 2.5 min after the onset of ether stress and remained high for 30 min. Two hours' exposure to ether did not change plasma alpha-MSH, although a 3-fold increase in plasma ACTH was observed. Haloperidol injection was followed by a large increase in plasma alpha-MSH, whereas ACTH levels increased similarly after saline and Haloperidol injection. Corticoid administration reduced ACTH, but not alpha-MSH. Three weeks after adrenalectomy, alpha-MSH levels had not changed but ACTH levels had increased ten-fold. These data indicate that alpha-MSH is secreted in the rat, and that the regulation of its secretion is different from that of ACTH.     

Circulating immunoreactive inhibin and testosterone levels in men with critical illness

OBJECTIVE--We aimed to concurrently characterize serial changes in circulating immunoreactive inhibin (irINH) and testosterone (T) as reflections of Sertoli and Leydig cell responses to acute critical illness in man. DESIGN--Blood samples were drawn within 24 hours of admission to an Intensive Care Unit and at weekly intervals thereafter for up to 4 weeks while the patient remained in Intensive Care Unit or after discharge to a general ward. PATIENTS--We studied 13 male subjects with critical illness requiring intensive therapy. MEASUREMENTS--Plasma levels of irINH, T, LH, FSH and sex hormone binding globulin (SHBG) were analysed in relation to (i) the severity of illness as indicated by a sepsis score, acute physiology and chronic health evaluation score, and reverse triiodothyronine (rT3) levels and (ii) the outcome of illness as determined by discharge from Intensive Care Unit and the two-month mortality. RESULTS--Overall irINH levels remained normal and correlated negatively with rT3 (r = -0.63, P = 0.001) but not with sepsis, acute physiology and chronic health evaluation score, or gonadotrophin levels. Neither admission nor serial irINH levels significantly distinguished between the different clinical outcomes. In contrast, T levels were depressed and inversely correlated with both sepsis and acute physiology and chronic health evaluation scores (P less than 0.02), and positively with gonadotrophins (P less than 0.01), but not rT3 levels. Men eventually discharged from the Intensive Care Unit showed a rise, while those remaining showed a fall, in T levels (P = 0.02, time-course interaction). Similarly, T levels were lower in patients who died than in survivors, despite the comparable T levels on admission (P = 0.02, time-course interaction). Despite the fall in T levels, gonadotrophin levels remained inappropriately in the eugonadal range but higher in men who were discharged from Intensive Care Unit (P = 0.02, time-course interaction). FSH but not LH levels were correlated with sepsis score (P = 0.02) but not acute physiology and chronic health evaluation score or rT3. CONCLUSIONS--Sertoli cell function as judged by circulating irINH levels is much less affected by acute critical illness than is Leydig cell function as judged by circulating T levels. The suppressive effect of acute critical illness on Leydig cell function is consistent with a hypothalamic-pituitary lesion.     

Regulation of Sertoli cell inhibin production and of inhibin alpha-subunit mRNA levels by specific germ cell types

To elucidate the endocrine and paracrine regulation of testicular inhibin production, the effects of follicle-stimulating hormone (FSH), (Bu)2cAMP, germ cells (either crude or enriched preparations) and germ cell-conditioned media on inhibin production (immuno- and bio-activities) and the levels of alpha- and beta B-subunit mRNAs were assessed in cultured Sertoli cells isolated from 20-day-old rats. FSH and (Bu)2-cAMP stimulated both secreted and intracellular inhibin levels in a dose-dependent manner. Using cDNA probes corresponding to the alpha-subunit and the beta B-subunit of rat inhibin it was also shown that both FSH and (Bu)2cAMP markedly increased the level of alpha-subunit mRNA but had no effect on the beta B-subunit mRNA. Addition of a crude mixture of germ cells to Sertoli cell monolayers was found to enhance inhibin secretion. Of the different germ cell fractions tested in co-culture, early spermatids reproducibly stimulated both basal and (Bu)2cAMP-induced production of inhibin whereas pachytene spermatocytes only increased the latter; cytoplasts from elongated spermatids (CES) had no effect. Co-culture of Sertoli cells with liver epithelial cells (LEC) significantly enhanced (Bu)2cAMP-induced inhibin levels. Media conditioned by early spermatids consistently and dramatically stimulated the secretion of both bioactive and immunoactive inhibin by Sertoli cells while spent media from pachytene spermatocytes displayed less activity. CES-conditioned media had only minor stimulatory effects, which may have resulted from the contamination of this fraction by spermatids. Media conditioned by LEC had no effect on inhibin production, confirming that the activity of this cell line is not mediated via a diffusible factor. Early spermatids were found to increase levels of the alpha-subunit mRNA. The current study provides evidence for the involvement of germ cells, in particular of early spermatids, in the local testicular regulation of inhibin gene expression and production in the rat. This may be of crucial importance for the ontogeny of this parameter of Sertoli cell function, and has important implications with regard to the postulated endocrine and paracrine roles of inhibin.     

Decreased serum inhibin levels in normal elderly men: evidence for a decline in Sertoli cell function with aging

Compared to young men, normal elderly men have decreased sperm production despite elevated serum gonadotropin levels. To determine whether the seminiferous tubule defect in elderly men includes decreased Sertoli cell function, we measured serum immunoreactive inhibin concentrations in young and elderly men before and after clomiphene citrate (CC) administration. Thirty-eight healthy men, 19 young (aged 22-35 yr) and 19 elderly (aged 65-85 yr), were studied before CC administration. The mean baseline serum inhibin level was significantly lower (P less than 0.001) in the elderly men than in the young men [416 +/- 22 (+/- SE) vs. 588 +/- 30 U/L], while serum immunoreactive FSH and LH levels were higher in the older men, and bioactive FSH levels were similar in the two age groups. Eleven young men and 13 elderly men were studied after 1 week of CC administration. The mean serum inhibin level increased by 71%, from 566 +/- 36 to 970 +/- 82 U/L, in the young men, but it increased by only 24%, from 421 +/- 26 to 520 +/- 38 U/L, in the elderly men. Serum immunoreactive LH and bioactive and immunoreactive FSH concentrations increased to similar levels in both groups after CC administration. We conclude that the seminiferous tubule defect of elderly men includes decreased Sertoli cell function.     

Relationship of serum inhibin levels to serum follicle stimulating hormone and sperm production in normal men and men with varicoceles.

The purpose of this study was to examine the relationships between serum inhibin levels as measured by RIA and serum FSH and sperm concentration. Three groups of men were used for this study: group I, normal fertile men (n = 67); group II, fertile men with a varicocele (n = 57); and group III, infertile men with a varicocele (n = 21). There were no differences in mean serum inhibin levels between the three groups. The two groups of men with varicoceles exhibited higher serum FSH levels and FSH responses to GnRH than the normal men. Sperm counts in both groups II and III were significantly lower than group I. In the normal men there was an inverse correlation between baseline serum inhibin and serum FSH levels and GnRH stimulated FSH levels, r = -0.415 and 0.422, P less than 0.005, respectively. Furthermore, the normal men exhibited a positive correlation between serum inhibin measurements and sperm concentration and testicular volume, r = 0.35 and 0.26, P less than 0.01 and less than 0.05, respectively. In neither group of men with a varicocele were these relationships found. These data demonstrate that serum inhibin does correlate with FSH in a negative fashion, when the reproductive system is normal, as would be expected for a negative feedback factor. Finally, the relationship of serum inhibin levels to testicular size and sperm count in the normal men suggests that serum inhibin levels reflect to some extent the integrity of seminiferous tubule function.     

Hormonal control of inhibin B in men

Serum inhibin B (IB) and testosterone (T) levels, secreted by Sertoli cells (SC) and Leydig cells (LC), respectively, are parameters of the functional state of these cells. Whereas LC activity and, consequently, T secretion are regulated by serum LH, factors regulating IB secretion by SC are still partially unknown. There is evidence that under certain conditions such as puberty, aging or some spermatogenesis defects, LH levels or Gn-independent factors might contribute to regulating SC activity and IB secretion. Among these factors, GH and IGF-I as well as PRL might have a role. Therefore, in order to explore the possible effects of either LH alone and FSH alone or a combination of both Gn, respectively, on SC function, IB plasma levels and spermatogenesis, we studied their effects in 6 patients with hypogonadotropic hypogonadism (HH), whereas the effects of GH on these parameters were studied in 6 men with panhypopituitarism (PH). Finally, the possible effects of PRL on SC function and spermatogenesis were studied in 6 patients with hyperprolactinemia (HPRL); 24 normal, fertile adults served as control group. In men with HH, neither human chorionic Gn (hCG) nor FSH, respectively, were able to increase serum IB after 3 months of therapy, whereas combined Gn therapy for 24 months increased IB plasma levels and stimulated spermatogenesis in 4 out of 6 hypogonadal men. In panhypopituitaric men, GH added to the classical Gn therapy did not have an additional effect on serum IB levels or spermatogenesis. Surprisingly, in our hyperprolactemic men, IB plasma levels were increased and positively correlated (p<0.01) with serum PRL levels, whereas normalization of the latter by cabergoline treatment caused a decrease of IB levels and a moderate increase in T, LH and FSH. In conclusion, the lack of SC response to FSH therapy alone, as opposed to the response to combined Gn therapy, might indicate that normalization of serum T by hCG is required to obtain IB secretion by SC. Addition of GH did not affect SC function, serum IB levels or spermatogenesis. Finally, our data suggest that PRL plasma levels might have a direct role on IB secretion, suggesting that the hypogonadism found in patients with HPRL might be a consequence of both central (inhibition of Gn secretion) and peripheral (stimulation of IB secretion) origin.     

Restoration of plasma testosterone levels in uremic men with clomiphene citrate.

Five men with chronic renal failure and symptoms suggestive to androgen deficiency were treated with clomiphene citrate (Clomid) at a dose of 100 mg/day for a period of 5 to 12 months. The treatment resulted uniformly in increased libido, sexual potency, and a general sense of well-being. Circulating testosterone rose from mean basal value of 223 +/- 164 to 879 +/- 171 ng/dl (SD), representing a mean increment of 290%. Mean serum lutenizing hormone (LH) and follicle-stimulating hormone (FSH) values before treatment were 76 +/- 40 and 143 +/- 85 ng/ml (SD). During treatment, both LH and FSH increased dramatically to 518 +/- 302 and 787 +/- 291 ng/ml (SD), respectively. Both serum gonadotropin values are expressed as ng/ml of LER 907. The effect of clomiphene on spermatogenesis in these subjects was inconclusive as either improvement or deterioration occured. In these five patients, serum prolactin was not related in any way to testicular function as its values were consistently in the normal range throughout the entire study period. Serum total estrogen, however, was elevated in all; the significance of this high circulating estrogen in relation to gonadal dysfunction in uremia is not clear at the present time. However, we found that normalization of circulating androgen was beneficial to our patients and that long-term clomiphene treatment achieved this goal by increasing pituitary gonadotropin secretion and secondarily stimulating testicular hormonogenesis.     

Diurnal rhythm in serum levels of inhibin B in normal men: relation to testicular steroids and gonadotropins.

Inhibin B is a testicular glycoprotein that is secreted from the Sertoli cells and believed to play a role in FSH secretion. We characterized the diurnal profile of serum inhibin B and the relation to gonadotropins and testicular steroids. Serum inhibin B was measured in 13 healthy normal male volunteers (median age, 30 yr) by continuous blood drawing, with sampling every 30 min for 24 h. Blood samples were also analyzed for FSH, LH, testosterone, estradiol, and sex hormone-binding globulin. We found a significant diurnal variation in inhibin B, with peak values in the early morning and nadirs in the late afternoon, followed by gradual increasing nocturnal values. An average decline of 3%/h from 0900 until 1700 h was calculated. Significant cross-correlation was found between inhibin B and testosterone as well as estradiol, whereas no cross-correlation was found between inhibin B and FSH. Two-dimensional time-series analyses revealed a statistically significant influence of testosterone on inhibin B. In addition, estradiol and inhibin B had a significant influence on one another. In conclusion, we found a significant diurnal variation in inhibin B levels in normal men, with a pattern of higher values in the early morning hours and lower values in the late afternoon and evening. We did not find evidence for a role of FSH in this diurnal variation of inhibin B. However, covariation with serum levels of testosterone and estradiol suggested that these hormones might play a role in the diurnal rhythm of inhibin B, although some other common influence could not be excluded.     

Immunoreactive inhibin concentrations in serum throughout the menstrual cycle of the macaque: suppression of inhibin during the luteal phase after treatment with an LHRH antagonist

Concentrations of immunoreactive inhibin in serum samples collected daily from six adult stumptailed female macaques during normal menstrual cycles were measured with a heterologous radioimmunoassay. Serum inhibin concentrations were low during the follicular phase of the cycle. After ovulation they began to rise, reaching a plateau between 8 and 11 days, before falling in parallel with the decline in luteal progesterone secretion. The dependence of the inhibin secretion by the corpus lutem on pituitary gonadotrophins was investigated by the administration of an LHRH antagonist [N-Ac-D-Nal(2)1,D-pCl-Phe2,D-Trp3,D-hArg(Et2)6,D-Ala10 ]LHRH once daily for 3 days beginning on day 8 of the luteal phase in six macaques. LHRH antagonist treatment markedly suppressed serum levels of inhibin and progesterone and these remained at the level found in the follicular phase for the remainder of the luteal phase. These results show that inhibin in the macaque is secreted into the peripheral blood almost exclusively during the luteal phase, being highest when FSH is at its nadir. Suppression of serum inhibin concentrations during the luteal phase by LHRH antagonist suggests that its secretion is integrated with the LH control of the corpus luteum.     

Depressed plasma pyridoxal-5'-phosphate levels in tobacco-smoking men

Tobacco-smoking men (n = 106) showed considerably lower plasma pyridoxal-5'-phosphate levels (P less than 0.001) than non-smoking controls (n = 143). A previously reported inverse relationship between plasma pyridoxal-5'-phosphate levels and age is confirmed by using new and sensitive high-performance liquid chromatography methodology for pyridoxal-5'-phosphate determinations. Plasma pyridoxal levels were neither lower in tobacco-smoking men nor could any relationship between pyridoxal levels and age be demonstrated. Two major risk factors for coronary artery disease, tobacco-smoking and increasing age, are thus associated with lower plasma pyridoxal-5'-phosphate levels.     

Subnormal plasma adrenal androgen levels in men with uremia

The 24-h mean plasma concentrations of 8 hormones were measured in 11 men with chronic uremia and 32 normal men. Our findings confirm previous reports of subnormal levels of testosterone, T3, and T4 and elevated levels of LH, PRL, and cortisol. In addition, we observed a new finding: markedly subnormal levels of the adrenal androgens dehydroisoandrosterone (DHA) and DHA sulfate. The mean DHA level in the patients was 164 +/- 46 (SD) ng/dl, compared with 320 +/- 124 in age-matched controls (P < 0.0001); the geometric mean DHA sulfate level was 40 micrograms/dl (95% confidence limits, 11-113) in the patients and 76 micrograms/dl (95% confidence limits, 26-214) in age-matched controls (P = 0.005). The depression of adrenal androgen levels in the face of elevated cortisol levels suggests a biosynthetic block in the adrenal cortex at the step where the C-19 and C-21 pathways diverge, namely the removal of the 2-carbon side chain by C-17, 20-lyase. If a similar defect were present in the testes, it could account for the diminished synthesis of testosterone, which is a further metabolite of DHA in the testes.