| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Reproductive Endocrinology |
Endocrine Section, Department of Internal Medicine, and Institute of Chemistry (C.P.), University of Brescia, Brescia, Italy; College of Agriculture, Forestry, and Life Sciences, Clemson University (W.B.W.), Clemson, South Carolina 29634; and the Departments of Pediatrics (A.D.R.) and Internal Medicine (J.D.V.), University of Virginia Health Sciences Center, Charlottesville, Virginia 22908
Address all correspondence and requests for reprints to: Dr. Andrea Giustina, Endocrine Section-Clinica Medica c/o 2 Medicina, Spedali Civili, 25123 Brescia, Italy. E-mail: Giustina{at}master.cci.unibs.it
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In summary, our experimental model suggests that in males even very small increases in circulating testosterone occurring during the earliest stages of puberty are able to amplify pulsatile GH secretion. Our concomitant secretagogue data further suggest that testosterone exerts its action at different sites in the hypothalamo-somatotropic axis, i.e. directly at the pituitary level, and also at hypothalamic loci, possibly increasing both GHRH and somatostatin release.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
To our knowledge, the large majority of data available to date on sex hormone-mediated regulation of spontaneous and stimulated GH secretion in humans has been derived from cross-sectional population studies including healthy subjects of normal stature representing one or more stages of pubertal development (2, 5, 6, 7, 8, 9, 10). In relation to the specific actions of testosterone on spontaneous GH secretion, previous experiments indicate that increased serum testosterone concentrations in males during late puberty are linked to a greater mean serum GH concentration over a 24-h period than those in other males in earlier stages of puberty (2, 5, 7, 9). Other studies have suggested that the increase in the daily GH production rate is attributable mechanistically to an augmentation of the calculated maximal rate of GH release attained per secretory episode, resulting in a greater mass of GH released per secretory burst (6). In contrast, GH secretory burst duration and frequency or GH half-life did not vary in males at different pubertal stages (7). Considerable evidence also suggests a positive correlation between the magnitude of plasma GH responses to some provocative stimuli and androgen concentrations in human subjects (11, 12).
However, important aspects concerning the role and mechanism of testosterone’s action in the regulation of spontaneous and stimulated GH secretion remain to be clarified. These include 1) the time course and role of androgens in the heightened GH response to physiological and pharmacological stimuli on a longitudinal basis in the same male subject during puberty; and 2) the nature of the specific actions of testosterone on the development of GH responsiveness to hypothalamic GHRH and somatostatin acting alone or in combination. To address these issues, we performed a longitudinal study of nocturnal pulsatile GH secretion and GH secretory responses to GHRH alone or combined with L-arginine (a functional somatostatin antagonist) (13) in previously untreated patients with isolated hypogonadotropic hypogonadism (IHH), who were exposed to progressively higher testosterone levels designed to mimic the androgen environment recognized during the early stages of puberty.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Six boys were studied, all of whom were referred to our Endocrine Section for delayed sexual development and were subsequently diagnosed with IHH. All patients had normal 46,XY karyotypes. None of the patients had been previously treated with androgens or gonadotropins. The diagnosis of IHH was made when all of the following criteria were met: 1) failure of pubertal onset by 15 yr of age; 2) low serum immunoreactive LH and FSH concentrations in the presence of low serum testosterone levels; 3) normal pituitary-thyroid and pituitary-adrenal function, as reflected by normal serum TSH and free T4 levels and a normal cortisol response to a short ACTH stimulation test; and 4) normal magnetic resonance imaging of the sella and parasellar structures. Furthermore, during the 6 months after the end of the study, none of the subjects apparently entered spontaneous puberty.
The pretreatment clinical characteristics of the subjects are
summarized in Table 1
. All subjects had testicular
volumes of 3 mL or less, and there was a mean delay in bone age of 1.5
yr. Differences in body mass indexes (BMIs) before and during therapy,
which could contribute to the variability in pre- and posttreatment GH
parameters (14), were minimal, i.e. the mean
(±SD) posttreatment BMI (22.8 ± 1.1
kg/m2) increased by only 3% over the mean pretreatment
value (22.4 ± 1.1 kg/m2).
|
Informed consent, before entering the study, was obtained from
all the patients and their parents. The study protocol was approved by
the local ethical committee. All patients received three incremental
doses of testosterone enanthate (Testoenant, Geymonat, Italy): 25-,
50-, and 100-mg im injections; each dose was administered three times
over a period of 4 weeks (one injection every 2 weeks). All patients
underwent blood sampling at 0800 h for total serum testosterone
assay every week. Studies of GH secretion (nocturnal and stimulated)
were performed four times: before testosterone administration (T1) and
on the sixth day after the third administration of each dose of
testosterone (T2, T3, and T4; Fig. 1). At these times,
samples from three 24-h urine collections performed over a 1-week
period were obtained for urinary GH assay, and blood samples were drawn
for assays of serum free testosterone, 17ß-estradiol, sex
hormone-binding globulin (SHBG), dehydroepiandrosterone sulfate
(DHEA-S), insulin-like growth factor I (IGF-I), and IGF-I-binding
protein-3 (IGFBP-3). Concurrently, routine hematological and urinary
parameters were assessed.
|
The study was performed according to a protocol previously described (15). Briefly, after admission to the unit at 1700 h, the subjects had a light meal at 1730 h. At 2000 h, a catheter was inserted into a forearm vein. Subjects then rested in bed. The first blood sample was withdrawn for GH assay at 2200 h. Subsequent blood samples were obtained at 20-min intervals until 0600 h on the following day. Lights were turned off at 2230 h and turned on at 0600 h. Sleep was evaluated by visual inspection at 20-min intervals.
Stimulated GH secretion
GHRH and L-arginine stimulation studies (13) were also repeated four times (before testosterone administration and on the seventh and eighth days after the third administration of each testosterone dose) according to a single blind, cross-over design. Briefly, all subjects underwent the following two tests in random order and on consecutive days: 1) an infusion of L-arginine hydrochloride (Damor, Naples, Italy; 30 g, iv, in 100 mL saline), or 2) an infusion of saline (100 mL, iv) from -30 to 0 min. At 0 min, the subjects received on both occasions an iv bolus injection of human GHRH-(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29)NH2 (Geref, Serono, Milan, Italy; 100 µg in 1 mL saline). Blood samples for GH assays were withdrawn at -45, -30 (time of the start of saline or arginine infusion), 0 (time of GHRH injection), 15, 30, 45, 60, 90, and 120 min.
Assays
A commercial immunoradiometric assay was used for estimation of serum GH concentrations (Allegro hGH, Nichols Institute, San Juan Capistrano, CA; inter- and intraassay coefficients of variation, 5.4% and 2.3%, respectively; sensitivity limit of the assay, 0.06 µg/L). Commercial RIAs were used for the measurement of total IGF-I (Nichols Institute; after acid-ethanol extraction: inter- and intraassay coefficients of variation, 5.2% and 9.4%, respectively; sensitivity limit of the assay, 0.013 nmol/L), IGFBP-3 (Nichols Institute; inter- and intraassay coefficients of variation, 5.8% and 5.6%, respectively; sensitivity limit of the assay, 0.06 µg/mL), total testosterone (Diagnostic Product Corp., Los Angeles, CA; inter- and intraassay coefficients of variation, 7.3% and 5%, respectively; sensitivity limit of the assay, 0.14 nmol/L), free testosterone (Diagnostic Products Corp.; inter- and intraassay coefficients of variation, 3.2% and 3.4%, respectively; sensitivity limit of the assay, 0.52 pmol/L), 17ß-estradiol (Diagnostic Products Corp.; inter- and intraassay coefficients of variation, 4.9% and 4.5%, respectively; sensitivity limit of the assay, 18.4 pmol/L), SHBG (Radim, Liege, Belgium; inter- and intraassay coefficients of variation, 4.6% and 5.6%, respectively; sensitivity limit of the assay, 2.5 nmol/L), and DHEA-S (Immunotech, Marseille, France: inter- and intraassay coefficients of variation, 4.1% and 5.6%, respectively; sensitivity limit of the assay, 1.62 µmol/L). A commercial immunoenzymatic assay (enzyme-linked immunosorbent assay) was used for estimation of urinary GH concentrations (Novo Nordisk, Bagsvaerd, Denmark: inter- and intraassay coefficients of variation, 1.3% and 4.7%, respectively; sensitivity limit of the assay, 2 ng/L).
Routine hematological and chemical assays were performed using automated procedures (Technicon II analyzer, Beckman, Palo Alto, CA). All GH samples from individual subjects were quantitated in a single assay in duplicate.
Data analysis
Quantitative characteristics of pulsatile GH secretion basally and after the administration of secretagogues were assessed assuming a simple burst model of hormone secretion using multiparameter deconvolution analysis (16). A dose-dependent variance model consisting of a power function fit of intrasample variance vs. dose (GH concentration) for all 24 samples in each subject was used to estimate within-sample SDs. Statistical confidence intervals of 95% were applied to each estimated GH secretory burst amplitude to achieve at least 90% specificity of pulse enumerations, based on 30 synthetic GH profiles created by computer simulation (17). Confidence limits for each estimate (pulse amplitude and mass) were determined by the support plane procedure (18). The serum GH concentration responses to saline plus GHRH and to L-arginine plus GHRH also were deconvoluted (19), and additionally expressed as mean concentrations (micrograms per L) and peak (maximum) serum values (micrograms per L). Differences among basal and stimulated mean GH pulse properties within and between subjects were detected by ANOVA with a repeated measures design. Correlations were assessed by applying univariate linear regression analysis. Data are expressed as the mean ± SEM. P < 0.05 was considered statistically significant.
Approximate entropy (ApEn)
ApEn is a recently introduced model-independent regularity statistic that is designed to estimate the relative degree of orderliness of serial observations (20). This analysis is designed to quantify the logarithmic likelihood that runs of patterns (consecutive data values) of length m will be repeated within a tolerance of r. Here, we used m = 1 and r = 0.20 times the within-series (8-h GH profile) SD to normalize ApEn. This normalized version is scale invariant and will be designated ApEn (1, 20%). As this was calculated here based on 24 data values, we also estimated the SD of each ApEn value by Monte Carlo perturbations (300 simulations/GH series).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The height of each of the six subjects was close to the 50th
percentile. As expected, short term testosterone administration over 17
weeks was associated with minimally advancing pubertal status and no
increases in bone age (Table 1
) or testicular size. No significant
change in BMI was observed during testosterone treatment.
Sex hormone profiles
Serum testosterone. Before treatment (T1), serum testosterone
concentrations (mean, 1.21 ± 0.13 nmol/L) were low in all
subjects. Serum testosterone levels increased in all subjects during
therapy. The means of all total testosterone measurements during the
treatment periods were: T2, 3.99 ± 0.87 nmol/L (P
< 0.05 vs. baseline); T3, 7.07 ± 1.1 nmol/L
(P < 0.05 vs. T2); and T4, 11.57 ±
0.58 nmol/L (P < 0.05 vs. T3); the highest
total serum testosterone concentrations recorded during the various
phases of treatment were: T2, 5.8 ± 1.04 nmol/L
(P < 0.05 vs. baseline); T3, 9.46 ±
1.04 nmol/dL (P < 0.05 vs. T2); and T4,
15.7 ± 1.04 nmol/L (P < 0.05 vs. T3;
Fig. 1). The highest free serum testosterone levels measured during the
phases of treatment were: T2, 16.05 ± 3.61 pmol/L
(P < 0.05 vs. baseline); T3, 24.8 ±
5.8 pmol/L (P < 0.05 vs. T2); and T4,
46.18 ± 4.06 pmol/L (P < 0.05 vs.
T3).
17ß-Estradiol. Before treatment (T1) and after the 25-mg
dose of testosterone replacement (T2), serum estradiol was undetectable
(<18.4 pmol/L) in all subjects. Serum 17ß-estradiol levels became
detectable only in two of the subjects after 50 mg testosterone
administration (T3; 29.4 and 33 pmol/L, respectively). Estradiol levels
were detectable in all but one of the patients after the last stage of
treatment (T4; mean, 42.2 ± 9.5 pmol/L; P < 0.05
vs. T3, T2, and T1; Table 2).
|
). Spontaneous GH secretion
Mean serum GH concentration. Mean nocturnal serum GH concentrations were higher after the 25-mg dose of testosterone compared to pretreatment values (2.64 ± 0.6 vs. 3.72 ± 0.4 µg/L; P < 0.05). The further increases in the mean nocturnal serum GH concentration observed at the other treatment steps (T3, 4.19 ± 0.9; T4, 5.15 ± 0.77 µg/L) were significantly higher than that at T1, but they did not achieve statistical significance compared to the T2 value.
Pulsatile GH secretion. The nocturnal serum GH
concentration and secretion profiles of two patients (referred to as 2
and 3 in Tables 1 and 2) before testosterone administration and after
the various testosterone treatments are depicted in Fig. 2. Specific measures of deconvolution-estimated
pulsatile GH secretion and half-life are presented in Table 3.
|
|
). In
particular, the mass of GH secretory bursts was clearly greater during
the first stage of treatment (T2) compared to the values before
treatment (T1; P < 0.05); no further statistical
increases in burst mass were observed at the T3 and T4 stages of
treatment (Table 3). Also, the mean amplitude (maximal rate) of GH
secretory pulses was greater in the T3 and T4 stages than that in the
T1 and T2 stages (P < 0.05). A small, but significant,
rise in the mean number of detectable GH secretory pulses occurred
concurrently with the increase in pulse amplitude (Table 3). These
amplitude and mass changes were specific, as the GH half-life and the
estimated secretory half-duration (duration at half-maximal amplitude)
did not change. The total mass of pulsatile GH secretion rose
significantly (vs. pretreatment) during stage T2 of
treatment without further changes at higher treatment doses (Table 3). ApEn
ApEn was used as a measure of the relative regularity or
orderliness of serial GH release over the 8-h nocturnal sampling
period. By computer-assisted simulations, we estimated a SD
of ApEn for each of the 24 individual GH time series. These values
varied between 0.02–0.08 given the within-assay uncertainty in the GH
immunoradiometric assay. Mean values of ApEn rose significantly
(P < 0.05), indicating increased disorderliness or
irregularity of the GH release process at T2 (25 mg testosterone) and
higher doses of testosterone (50 and 100 mg) vs. that at T1
(before androgen treatment). Mean values were not significantly
different from one another at the T2, T3, and T4 stages of
treatment (Table 3).
24-h urinary GH
The mean 24-h urinary GH concentration (resulting from three 24-h urine collections for each subject) before androgen treatment (T1) was 3.35 ± 0.6 ng/L. Urinary GH levels did not significantly increase at T2 (4.58 ± 0.6 ng/L; P = NS vs. T1). Conversely, significant increases vs. T1 were observed at T3 (6.8 ± 1.7 ng/L) and T4 (10.3 ± 2.7 ng/L; P < 0.05 vs. T2 and T3).
Stimulated GH secretion
The kinetics of the GH responses to saline plus GHRH and to
L-arginine plus GHRH are illustrated in Fig. 3. The mean pre-GHRH injection serum GH levels were not
significantly different at the four stages of the study. In fact,
infusion of L-arginine alone did not increase mean time
zero serum GH levels in a similar fashion across the four different
stages of testosterone treatment.
|
). Peak
serum GH concentrations pretreatment (T1) occurred 15–60 min after
GHRH injection in all patients. Testosterone administration did not
change the timing of GH peaks significantly or alter the mean absolute
serum GH levels (Fig. 3) or the deconvolution-estimated GH secretion
after GHRH treatment. L-Arginine pretreatment induced a
significant increase in the GH responses to GHRH (compared to saline
pretreatment) both at baseline (T1; GH peak, 31.8 ± 4.7 µg/L;
P < 0.05 vs. saline plus GHRH) and during
treatment. Specifically, testosterone significantly (P
< 0.05) increased the GH responses to L-arginine plus GHRH
compared to those during T1 during stage T3 (GH peak, 49 ± 5.7
µg/L) and T4 (GH peak, 49 ± 8.3 µg/L) of treatment (Fig. 4a).
Deconvolution analysis revealed the greatest calculated total mass of
GH secreted after L-arginine plus GHRH at T3 and T4
compared to T1 stages (Fig. 4b). Mean absolute serum GH concentrations
at 15, 30, and 45 min after L-arginine plus GHRH at stages
T3 and T4 were also significantly higher than those observed at T1
(Fig. 4a). The ratio between the GH peak after L-arginine
plus GHRH vs. that after saline plus GHRH was significantly
higher at T3 and T4 compared to that at T1. All subjects experienced
facial flushing after GHRH injection. No side-effects were observed
during L-arginine infusion.
|
The plasma IGF-I concentrations for the six subjects before
treatment (T1) and at the end of the study are summarized in Table 2.
Values during treatment were always greater than those before treatment
(T1; 35.3 ± 6 nmol/L) and progressively increased to reach
maximal values in all subjects at the last stage of treatment (T2,
49.3 ± 6.67 nmol/L; T3, 57.92 ± 4.71 nmol/L; T4, 69 ±
6 nmol/L). Conversely, no significant changes in IGFBP-3 levels were
observed during treatment (T1, 3.33 ± 0.2 µg/mL; T2, 3.58
± 0.4 µg/mL; T3, 3.21 ± 0.3 µg/mL; T4, 3.53 ± 0.31
µg/mL).
Correlations
Serum testosterone concentrations vs. parameters of spontaneous GH
secretion and plasma IGF-I. Significant (P <
0.05) linear correlations were found between the mean nocturnal GH
concentration (r = 0.682), the GH secretory burst mass (r =
0.504), the number of GH peaks (r = 0.5), GH ApEn (r =
0.484), 24-h urinary GH (r = 0.703), and plasma IGF-I (r =
0.459) and the free serum testosterone recorded for each subject at
each stage of treatment. Most of these parameters also significantly
correlated with the peak total testosterone levels recorded for each
subject throughout the various phases of the study (Fig. 5). Serum IGF-I concentrations correlated significantly
with mean 24-h serum GH levels and total mass of GH secreted in pulses
(r = 0.86; P < 0.05). The global correlations
were confirmed by the z-score transformation of the six individual r
values contained within each of the four treatment and testing of the
null hypothesis of a random normal distribution of z-score above 0
(which would have indicated no correlation). We used the
Kolmogorov-Smirnov statistic to test this null hypothesis (21).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
3-fold) increase in serum testosterone
concentrations for only 6 weeks induced by the lowest dose (25 mg
biweekly) of testosterone amplified the GH secretory burst mass
significantly without altering detectable GH secretory frequency,
half-duration, or GH half-life, thus closely mimicking the GH response
during normal puberty in boys (7). Mechanistically, testosterone also
augmented pituitary GH secretion significantly after presumptive
somatostatin withdrawal combined with GHRH (L-arginine plus
GHRH) (13). Sex steroid hormones profoundly influence multiple endocrine axes and the GH axis in both rats (22) and humans. In the human, previous reports have described an increase, mediated by sex steroids, of 24-h plasma GH levels in both sexes during spontaneous puberty or exogenous sex hormone treatment (2, 3, 5, 6, 7, 8, 9, 10, 23, 24, 25). As these studies were cross-sectional, large interindividual variations in baseline plasma GH concentrations could have obscured the influence of androgens and/or estrogens, and the studies could not clarify the time course of GH axis responses. In healthy boys, the onset of adult androgen secretion in mid to late puberty is accompanied by a consistent increase (2- to 3-fold) in the mean (24-h) serum GH concentration (2, 5, 7, 9, 23). Mechanistic analysis by deconvolution techniques demonstrated that the increase in peak serum GH concentrations resulted from an amplified and maximal GH secretory rate attained within each discrete release episode (7). A previous study investigated the effects of parenterally injected testosterone in boys with constitutionally delayed puberty. In these subjects, testosterone administration doubled the height and area of serum GH concentration peaks due to a doubling or tripling of the mass of GH secreted per burst (24). These experimental observations in boys are consistent with the inference that androgen can positively regulate GH secretory burst mass or amplitude. To date, only one previous report has characterized the effects of testosterone on 24-h serum GH concentrations in a group of previously untreated young men with isolated gonadotropin deficiency (25). In these patients, when mean plasma testosterone levels were normalized, there was a 2-fold rise in mean 24-h serum GH concentrations and GH pulse amplitude. By contrast, mean 24-h GH pulse frequency did not change.
Our study offers insight into the maturation of mechanisms that control spontaneous pulsatile GH secretion in the young male, as we modulated the GH axis by progressively increased serum testosterone levels within the physiologically observed range for early puberty. Our data, obtained by frequent blood sampling and studied by means of deconvolution analysis, confirm that testosterone is able to significantly increase the amplitude and mass of spontaneous GH secretory events (25). Of considerable interest is our observation of augmented GH peak amplitude after 6 weeks of treatment with the lowest dose of testosterone (25 mg every 2 weeks). Indeed, blood concentrations of testosterone attained after this dose approximated only one fifth to one seventh of those anticipated in healthy young men (24). Therefore, we can hypothesize that burst mass is a highly regulated fundamental unit of spontaneous GH secretion and serves as the primary mechanism activated during the early stages of male puberty. Interestingly, the total mass of GH secreted was significantly increased compared to untreated levels after the 4-week regimen of 25-mg testosterone injections. These findings are confirmed by the observation that the 24-h urinary GH concentration was also significantly increased by the 50- and 100-mg testosterone doses.
There are conflicting reports in the literature concerning the effects of testosterone on the frequency of spontaneous GH peaks. Some researchers have observed, in cross-sectional studies, an increase throughout various pubertal stages in GH peak number (26), which has not been confirmed in either normal pubertal boys (2) or hypogonadal men (25) studied with 20-min sampling for 24 h. Our 20-min data suggest that both the 50- and 100-mg doses of testosterone increase the frequency of detectable spontaneous nocturnal GH peaks. However, additional studies will be required to corroborate this inference, given the lack of significant changes in GH intersecretory pulse interval. Indeed, previous studies comparing different subjects in various pubertal stages may have been unable to detect transient changes in GH peak frequency in early puberty. Independently of possible changes in GH burst frequency, the present clinical model of low dose testosterone administration unmasked a positive strict correlation between circulating total and free testosterone concentrations and total and individual GH secretory burst masses in the male. This finding supports experimental evidence and the earlier inferential thesis that testosterone is very potent in increasing both spontaneous GH secretory event amplitude and mass (7).
Our data show that the ApEn of GH secretion was significantly increased even by the 25-mg dose of testosterone. These data indicate that 6-week testosterone administration affects the quantifiable orderliness of GH release even at very low blood androgen concentrations. The visual impression that regularity of the GH secretory process is sexually dimorphic in both the human and the rat can be quantitatively confirmed, because females secrete GH with more irregularity than males (27). However, to date, no data are available on the effects of puberty in general or of testosterone during pubertal development in males in particular on the orderliness of spontaneous GH release. Our data show for the first time that testosterone acts specifically to increase the disorder of the GH secretory process in man. It could be hypothesized that this effect is simply the manifestation of an increase in GH secretion rather than an isolated or independent event. However, this hypothesis is unlikely because this parameter is not dependent on the hormone concentration per se; in fact, it is thought that ApEn, as normalized, is independent of scale (20, 21). Thus, this loss of regularity of GH secretion can be hypothesized to be a distinct feature of pubertal development in males, occurring in the very early stages of the process and possibly reflecting maturation, in the sense of an increased activity, of the hypothalamic (GHRH/somatostatin interplay) drive on GH secretion determined by testosterone.
IGF-I mediates most of the effects of GH at the peripheral level and is produced by the liver under the control of circulating GH (28). The biological activity of IGF-I may depend, in turn, on the fraction of IGF-I that is not bound by its binding proteins, the major one of which is IGFBP-3 (29). Association with IGFBP-3 accounts for about 95% of bound IGF-I, and, at least for some target tissues, acts as a physiological inhibitor of IGF-I action. Previous data have shown that plasma IGF-I levels increase during puberty (30, 31) or testosterone treatment in androgen-deficient males (32, 33). Conversely, less is known about the effects of either pubertal development or testosterone administration on IGFBP-3 levels (30). In fact, it has been reported that IGFBP-3 levels in young children are low and rise with increasing age to peak in 14- to 15 yr-old children. In children entering puberty, IGFBP-3 levels rise slightly earlier in females than in males, although the difference is significant only in the 10- to 11-yr age group (34). Interestingly, our data show that although testosterone potently and specifically increases IGF-I levels, it does not change IGFBP-3 levels over the same time course, even at the highest dose used. Therefore, the predicted testosterone-mediated increase in biologically active IGF-I may be hypothesized to play a role in the synergism between testosterone and GH in the growth process (35) even if, on the basis of our data, it cannot be excluded that concomitant testosterone-mediated changes in other IGFBPs could occur and, in turn, influence the levels of circulating free IGF-I.
Based on available experimental data in the rat (22), one can hypothesize that the increase in GH peak amplitude is caused either by a testosterone-mediated increase in effective GHRH release and action (36, 37) or by a direct sex steroid effect at the pituitary level (38, 39, 40, 41, 42). In the human, evidence is available that sex steroids are among the factors that enhance GH release after stimulation by numerous pharmacological agents. In fact, treatment of boys with delayed puberty with androgens leads to higher peak serum GH concentrations in response to hypoglycemia. This increase is identified as early as 2 days after a single injection of testosterone. When therapy is continued for 2–3 months, peak GH concentrations after hypoglycemic stimulation increase to values seen in pubertal children (43). Martin and associates (12) suggested that androgens enhance the GH response to L-arginine stimulation. L-Arginine does not influence either basal or GHRH-induced GH secretion from rat anterior pituitary cells in vitro (44), whereas this amino acid enhances the GH response to GHRH in men (13, 44). As L-arginine blocks glucocorticoid-induced GH suppression (45) and enhances the GH response to a maximal (46) GHRH dose, it has been hypothesized to stimulate GH secretion by inhibiting endogenous somatostatin secretion (45). Surprisingly, no data are available on changes in GH responses to either direct pituitary (GHRH) or hypothalamic (e.g. L-arginine) stimuli during pubertal development. Our observations indicate that the GH response to GHRH is not significantly influenced by the increase in testosterone levels that concomitantly stimulates pulsatile GH secretion. In fact, both the amplitude of the peaks and the total mass of GH secreted after GHRH stimulation were similar in our group of hypogonadal males at the beginning and end of the study, when serum levels of testosterone typical of late puberty were attained. Conversely, when the GHRH test was performed after pretreatment with L-arginine, the higher doses of testosterone (50 and 100 mg) significantly enhanced the GH responses, suggesting that testosterone can modulate somatostatinergic inhibitory tone.
Few data have been published on the behavior of the GH response to GHRH during the various stages of normal puberty in males. In humans, the GH response to GHRH is maximal during infancy and becomes less pronounced with advancing adult age (47). This decreased responsiveness may be due to an increase in somatostatinergic tone with age, as a restoration of the GH response to GHRH can be achieved by cholinergic pretreatment (48), which is thought to block somatostatin activity at the hypothalamic level (49). Moreover, available indirect in vivo studies in the rat strongly suggest that testosterone may have a stimulating effect on hypothalamic somatostatin tone (50, 51, 52, 53). Therefore, to reconcile available data in the rat with our clinical findings, we can hypothesize that testosterone does not change the GH response to GHRH alone in our patients because its direct stimulating actions at the pituitary level are obscured by an increase in hypothalamic somatostatin tone (45). To test this hypothesis, we performed an experiment investigating the effects of changes in circulating testosterone levels on a combined hypothalamic-pituitary stimulus of GH secretion. Our data clearly demonstrate that testosterone enhances the GH response to GHRH plus L-arginine in boys; this effect is evident even with a 50-mg replacement dose of this hormone. The finding that in the presence of L-arginine, an amino acid thought to decrease somatostatin tone (13, 44), the GH response to GHRH is increased significantly during testosterone replacement confirms the overall hypothesis of a dual action of testosterone at the hypothalamic level (increase in GHRH and somatostatin) combined with a direct GH-stimulating action at the pituitary level.
Aromatization to estrogen has been hypothesized to be important in testosterone-mediated stimulation of the GH axis in the human (54). Indeed, the available clinical data indicate that antiestrogens reduce testosterone’s stimulation of GH in healthy adolescent boys (55). Moreover, estradiol (56) has been reported to increase growth in prepubertal males, whereas the nonaromatizable androgen, dihydrotestosterone, has been suggested to be unable to mimic the stimulatory action of testosterone on spontaneous GH secretion in prepubertal males (57). In our study the first changes in GH secretion were evident in the presence of very small increases in serum testosterone concentrations concomitant with still undetectable circulating estrogen levels. Moreover, the various parameters of GH release only correlated with circulating testosterone. These findings support the thesis that the GH-stimulating action of testosterone occurs via the androgen receptor (58). Moreover, although these data cannot rule out completely possible significant biological effects either of small changes in circulating 17ß-estradiol concentrations undetectable with the assay we used or of intrahypothalamo-pituitary aromatization of testosterone, it should be noted that the testosterone-mediated changes in GH secretion that we observed occurred in the presence of a progressively decreased serum level of SHBG, which can be considered a reliable biological marker of the peripheral activity of testosterone via the androgen receptor (59).
In conclusion, our human studies suggest that the small increases in testosterone occurring during early puberty are able to promote an increase in GH secretory peaks, plasma IGF-I levels, and the irregularity of the GH release process. Mechanistically, our pharmacological data further suggest that testosterone may act directly at the pituitary level to increase the GH secretory capacity of somatotropes and also at the hypothalamic level to augment the hypothalamic activity of both GHRH and somatostatin.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received August 23, 1996.
Revised November 26, 1996.
Accepted December 5, 1996.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. L. Utz, A. Yamamoto, P. Sluss, J. Breu, and K. K. Miller Androgens May Mediate a Relative Preservation of IGF-I Levels in Overweight and Obese Women Despite Reduced Growth Hormone Secretion J. Clin. Endocrinol. Metab., October 1, 2008; 93(10): 4033 - 4040. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Giustina, G. Mazziotti, and E. Canalis Growth Hormone, Insulin-Like Growth Factors, and the Skeleton Endocr. Rev., August 1, 2008; 29(5): 535 - 559. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Paulo, M. Cosma, C. Soares-Welch, J. N. Bailey, K. L. Mielke, J. M. Miles, C. Y. Bowers, and J. D. Veldhuis Gonadal Status and Body Mass Index Jointly Determine Growth Hormone (GH)-Releasing Hormone/GH-Releasing Peptide Synergy in Healthy Men J. Clin. Endocrinol. Metab., March 1, 2008; 93(3): 944 - 950. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Muniyappa, J. D. Sorkin, J. D. Veldhuis, S. M. Harman, T. Munzer, S. Bhasin, and M. R. Blackman Long-term testosterone supplementation augments overnight growth hormone secretion in healthy older men Am J Physiol Endocrinol Metab, September 1, 2007; 293(3): E769 - E775. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Veldhuis, D. M. Keenan, A. Iranmanesh, K. Mielke, J. M. Miles, and C. Y. Bowers Estradiol Potentiates Ghrelin-Stimulated Pulsatile Growth Hormone Secretion in Postmenopausal Women J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3559 - 3565. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Veldhuis, J. N. Roemmich, E. J. Richmond, and C. Y. Bowers Somatotropic and Gonadotropic Axes Linkages in Infancy, Childhood, and the Puberty-Adult Transition Endocr. Rev., April 1, 2006; 27(2): 101 - 140. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D Veldhuis, D. M Keenan, K. Mielke, J. M Miles, and C. Y Bowers Testosterone supplementation in healthy older men drives GH and IGF-I secretion without potentiating peptidyl secretagogue efficacy Eur. J. Endocrinol., October 1, 2005; 153(4): 577 - 586. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. P A Rooman, L. O. De Beeck, M. Martin, J. van Doorn, S. Mohan, and M. V L Du Caju Ethinylestradiol and testosterone have divergent effects on circulating IGF system components in adolescents with constitutional tall stature Eur. J. Endocrinol., April 1, 2005; 152(4): 597 - 604. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Veldhuis, S. M. Anderson, A. Iranmanesh, and C. Y. Bowers Testosterone Blunts Feedback Inhibition of Growth Hormone Secretion by Experimentally Elevated Insulin-Like Growth Factor-I Concentrations J. Clin. Endocrinol. Metab., March 1, 2005; 90(3): 1613 - 1617. [Abstract] [Full Text] [PDF]
|
|
, Time of GH testing; 
, time of im
testosterone enanthate injection.
) or
L-arginine plus GHRH (