Main

Secretion of FSH is regulated at the hypothalamic level by sex steroids and at the pituitary level by inhibin, a glycoprotein with two dissimilar subunits (1). Several recent findings suggest that inhibin B, consisting of α- and βB-subunits, is the endocrinologically most important form of inhibin in men. The serum inhibin B level increases in response to rhFSH (2,3), whereas the level is low in men with primary or secondary hypogonadism (2,4,5), suggesting that the amount of circulating inhibin B reflects the function of the Sertoli cells. Unequivocal evidence that inhibin B can suppress pituitary FSH secretion is lacking. However, a negative feedback appears likely, because in normal and infertile men, in men with primary or secondary hypogonadism, and in boys during puberty, serum inhibin B and FSH levels correlate negatively (4–9).

In addition to inhibin B, men have circulating inhibin precursor forms and free inhibin α-subunits, but the physiologic relevance of these proteins is unclear. Even in men without testes, the serum level of inhibin precursors that contain propeptide in the α-subunit (pro-αC inhibin) is measurable (2). The pro-αC inhibin assay could also detect dimeric inhibin precursor forms that contain the propeptide (10), but in men, it has been suggested to be unique in measuring circulating free inhibin (pro)-α-subunits (4,7,10).

We describe here the patterns of change in serum inhibin B and pro-αC inhibin levels during normal male development, and relate these findings to previously described parameters of puberty. Moreover, to further evaluate the role of FSH in the production of pro-αC inhibin in very early puberty, we studied serum pro-αC inhibin levels in boys with hypogonadotropic hypogonadism before and during their treatment with rhFSH.

METHODS

Subjects. The study material consisted of a longitudinal series of data gathered from an earlier follow-up study of healthy Finnish schoolboys (11). At the beginning of the study, the age (mean ± SEM) of the 38 boys studied was 11.7 ± 0.1 y. The subjects were healthy, apart from one boy in whom epilepsy was diagnosed and treated with oxcarbazepine medication. In short, the study covered 3 y and consisted of nine visits at 3-mo intervals during the first 2 y, the final check-up taking place 1 y later. At every visit, puberty was staged according to Tanner (12), and the length and width of the testes were measured with a ruler to the nearest millimeter. Testis volume was calculated from the formula 0.52 × length × width2 and converted to milliliters (13).

The study protocol was approved by the parents, the school authorities, and the ethics committee of the Hospital for Children and Adolescents, University of Helsinki.

Standard laboratory analyses. At every visit, venous blood samples were drawn from all the boys between 0830 and 1400 h. After clotting, the serum was separated by centrifugation and stored at -20°C until required. The serum testosterone concentration was measured by RIA after separation of the steroid fractions on a Lipidex-5000 microcolumn (Packard-Becker, B.V. Chemical Operations, Groningen, The Netherlands) as described previously (14). Serum LH and FSH concentrations were measured by time-resolved immunofluorometric assays, using reagents from Wallac OY (Turku, Finland), as described previously (15). Serum sex hormone-binding globulin was determined by immunofluorometric assays (DELFIA, Wallac, Turku, Finland). The free androgen index was calculated as serum testosterone/sex hormone-binding globulin × 100 (16).

Inhibin B and pro-αC inhibin assays. Serum inhibin B concentration was determined as described in detail previously (7,17,18). The sensitivity of the assay was 10 pg/mL, there being a 0.5% cross-reaction with inhibin A and 0.1% cross-reactions with human pro-αC precursor, activin A, activin B, and follistatin. The within- and between-plate coefficients of variation were less than 10%. Serum pro-αC inhibin concentrations were determined as previously described (19). The sensitivity of this assay was 3 pg/mL. Recombinant forms of activin A, activin B, and follistatin all showed cross-reactions of less than 0.02%. The within- and between-plate variations were less than 5 and 7%, respectively. All measurements were carried out in duplicate.

Statistics. For analyses of serum inhibin B and pro-αC inhibin concentrations at each Tanner's stage, the median value within each stage was calculated for each subject. In cases with an even number of observations for a stage, one of the two central observations was randomly selected. Spearman's rank correlation coefficients between serum inhibin B, pro-αC inhibin, and FSH levels were calculated for the boys at different Tanner's genital stages, to avoid the effect of nonlinear relationships between these variables. To examine the significance of the differences between mean serum hormone and pro-αC inhibin levels in boys at different Tanner stages, we used ANOVA, followed by Fisher's protected least significant difference test. Before ANOVA, a log transformation was applied to the serum hormone concentrations, as the variances and means increased simultaneously. Thereafter, the variances of the transformed variables at the different stages did not differ. All values are expressed as geometric means (±SD), unless stated otherwise. Statistical significance was accepted for p < 0.05.

RESULTS

The associations between serum FSH and inhibin B levels in boys at the different Tanner genital stages are presented in Fig. 1. In boys at stage G4, serum FSH and inhibin B levels correlated significantly (r = -0.57, p < 0.001, n = 37; Fig. 1). Individual patterns of serum inhibin B and FSH levels in six boys with similar rates of pubertal progression are presented in Fig. 2. The development of the inverse relationship between serum inhibin B and FSH levels is less obvious in boy 5, recognizable in boy 1, but can be particularly well seen in boys 2, 3, 4, and 6 (Fig. 2). The distributions of the serum inhibin levels and other parameters investigated in boys at different Tanner stages are given in Table 1. In boys at each stage, serum inhibin B and pro-αC inhibin levels did not correlate (r = -0.10-0.10).

Figure 1
figure 1

Association of FSH with inhibin B in healthy boys at different Tanner genital stages (from G1 to G4) during the 3-y follow-up. There were observations made on 16 subjects in stage G1 and on 37 subjects in stages G2, G3, and G4.

Full size image
Figure 2
figure 2

Individual serum (S) inhibin B and FSH concentrations in six boys during the 3-y follow-up. These boys were selected on the basis of similar tempo of puberty; they were followed up from Tanner stage G1 to stage G4, and during that time, had comparable testicular growth rates (range, 4.2-5.4 mL/y).

Full size image
Table 1 Distributions of clinical and hormonal parameters in 38 healthy boys during the course of puberty
Full size table

Because the numbers of boys in each Tanner stage differed, the longitudinal changes in the parameters examined were studied in the 16 boys who entered the study at stage G1, and were followed up until stage G4. Their mean serum inhibin B concentration increased from 109 (90-131) pg/mL at stage G1 to 134 (99-181) pg/mL at stage G4 (Fig. 3), the only significant change occurring between stages G1 and G2 (p < 0.02), after which there was a plateau. From G1 to G2, the mean serum FSH concentration did not change significantly (p = 0.18), but the mean testis volume (p < 0.001), and mean serum LH and testosterone concentrations were elevated (p < 0.001 and p < 0.01, respectively, Fig. 3). In this same subgroup of boys, the mean serum pro-αC inhibin level increased, but did not show the plateau in mid-puberty (Fig. 3). In contrast, the only statistically significant change between successive stages was from G3 to G4 (p < 0.05).

Figure 3
figure 3

Distributions of serum (S) inhibin B (upper left panel), pro-αC inhibin (upper right panel), gonadotropin (middle panels), and testosterone (lower left panel) levels, and testis volumes (lower right panel) in 16 subjects followed up from Tanner stage G1 to stage G4. The 10th, 25th, 50th, 75th, and 90th percentiles are shown.

Full size image

Serum pro-αC inhibin levels in boys at the earliest stages of puberty (G1 and G2) were not associated with age, but in boys at stage G1, the levels correlated positively with serum FSH concentrations (r = 0.50, p = 0.051, n = 16, Fig. 4, left panel). To clarify the causality behind this finding, we examined the serum pro-αC inhibin levels in three prepubertal (G1P1) gonadotropin-deficient boys from an earlier study (3). These boys (A, B, and C), were treated with rhFSH (1.5 IU/kg s.c. three times/wk) for 1 y from ages 12.8, 13.2 and 13.2 y, respectively. During the 12-mo treatment period, serum pro-αC inhibin levels, measured at 3-mo intervals, increased (p < 0.02, ANOVA). The values before and during the treatment are presented in Fig. 4 (right panel).

Figure 4
figure 4

(Left panel) Association of pro-αC inhibin with FSH in healthy boys (n = 16) at Tanner stage G1. (Right panel) Serum (S) pro-αC inhibin concentrations in three prepubertal gonadotropin-deficient boys from an earlier study (3), treated with recombinant human FSH (1.5 IU/kg three times/wk s.c.) for 1 y. Serum levels before (first data point for each subject), and measured with 3-mo intervals during the treatment are shown for boys A (○), B (▪), and C (▵).

Full size image

DISCUSSION

The recent development of specific immunoassays for measuring the different inhibin forms in human plasma and serum has significantly improved the understanding of the signaling between the pituitary gland and the gonads. The present study was undertaken to investigate the changes occurring in serum gonadotropin and inhibin levels during pubertal development of healthy boys. We found that, in prepubertal boys, FSH is likely to regulate the production of pro-αC inhibin, but in the course of puberty, the most dramatic findings were observed between serum FSH and inhibin B levels.

In early puberty, simultaneously with elevations in serum LH and testosterone levels, the mean serum inhibin B concentration increased. The relative roles of LH and FSH in this process are unclear. Although a significant production of inhibin B by the Leydig cells is unlikely (4), the LH-induced androgen production may indirectly stimulate inhibin B secretion from the Sertoli cells by, for example, enhancing the interaction between the Sertoli cells and germ cells (7). On the other hand, rhFSH appears to stimulate inhibin B production by immature testes (3), but in the present study, serum inhibin B and FSH levels in healthy prepubertal boys were not positively correlated. Nevertheless, the temporal relationship between FSH and inhibin B secretion is not known, and therefore, this lack of positive association in healthy prepubertal subjects does not rule out FSH-induced secretion of inhibin B.

After the early increase in serum inhibin B level, it remained relatively unchanged and correlated negatively with serum FSH level, suggesting, in agreement with the results of Crofton et al. (9), an establishment of a feedback loop between testicular inhibin B and pituitary FSH secretion. A significant correlation was observed relatively late, in boys at Tanner stage G4. However, an earlier feedback regulation has been proposed (9) and is further suggested by our previous finding that pre- and early pubertal boys with gonadal failure have higher serum FSH levels than do normal subjects (20).

In men, as evidenced by immunologic characterization of adult male serum (10), and suggested by the lack of correlation between serum inhibin B and pro-αC inhibin levels (4,7), the pro-αC assay mainly measures the amount of circulating free inhibin (pro)-α-subunits and not the inhibin B precursor forms (containing the propeptide in the α-subunit). In the present study, serum inhibin B and pro-αC inhibin levels were not correlated and displayed different patterns of change during puberty. These findings suggest that also in boys the pro-αC inhibin assay measures the amount of free inhibin (pro)-α-subunits. These proteins have different potential sources (2,4,21), which can explain the wide distributions of the serum pro-αC inhibin levels observed in boys at different Tanner stages. Nevertheless, in healthy prepubertal boys, serum pro-αC inhibin and FSH levels correlated, and during the rhFSH treatment of prepubertal gonadotropin-deficient boys, the serum pro-αC inhibin level increased. These findings suggest that FSH is an important factor stimulating pro-αC inhibin secretion by immature testes. Parenthetically, the rhFSH treatment related increase could also have been a normal age-related phenomenon. This was not, however, supported by the finding that in healthy pre- and early pubertal boys, serum pro-αC inhibin levels did not correlate with age.

The different precursor forms of free inhibin α-subunits are proposed to inhibit the binding of FSH to its receptor (22). However, there is no unequivocal evidence for any distinct endocrine function of these proteins in men (7,23), which is in agreement with our results obtained by studying boys during puberty. Nevertheless, our results suggest that if the serum inhibin level in boys is measured by an assay not specific for dimeric inhibin B, the presence of free inhibin (pro)-α-subunits in large amounts can potentially explain the lack of negative association between serum FSH and immunoactive inhibin levels (24).

In conclusion, the present work demonstrates that serum inhibin B level increases in early puberty, plateaus thereafter, and correlates negatively with the serum FSH level, supporting the view that, in healthy boys during puberty, pituitary FSH secretion is inhibited by inhibin B. During puberty, serum pro-αC inhibin level increases differently, displays wide variation, and is not negatively correlated with serum gonadotropin levels, suggesting that, in boys, the pro-αC inhibin assay measures mainly the serum level of free circulating inhibin (pro)-α-subunits.