Main

Long-acting GnRH-a can reversibly suppress gonadal steroid production, when given continuously(1). After an initial stimulation of gonadotropin release, the pituitary response diminishes, and the resulting decrease in LH and FSH levels leads to a reduction in gonadal steroid synthesis(2, 3). The mechanism of reversible chemical castration by GnRH agonists is not fully understood, but both the number of GnRH receptors(4) and postreceptor signaling events(5) may be involved. Clinically, GnRH-a have proved useful in the treatment of prostate cancer, endometriosis, and central precocious puberty(6), especially with the advent of depot preparations of superagonists(7).

In precocious puberty, GnRH-a suppress the acceleration in epiphyseal maturation, which otherwise would result in short adult stature(8, 9), and there is now increasing interest(10, 11) in the potential value of using GnRH-a in other children with an impaired predicted adult height, in the hope that delaying puberty might extend the time available for growth by delaying epiphyseal closure. However, the advantages of this approach are by no means clear(11–13), because gonadal steroids also affect growth and the growth plate directly, as well as modulating pituitary GH secretion, hepatic GHR, and the related GHBP. There are only a few reports evaluating GH secretion and GHBP in GnRH-a-treated children with precocious puberty, but results are contradictory, and studies often lack untreated control groups(14–18).

The purpose of this study was to use prolonged GnRH-a treatment in an experimental animal model in which the effects on growth, GH secretion, GHR, and GHBP can all be directly assessed. In addition, the effectiveness of depot GnRH-a treatment could be compared with a state of complete gonadal steroid deficiency induced by gonadectomy.

The advantages of using the rat for such experiments is that it shows sexual dimorphism in growth and GH secretion(19–21), in its expression of hepatic GH-receptors(22) and that GHBP(23) can be measured, and it is modulated by exogenous gonadal steroids(24, 25). We have now investigated the effects of the widely used slow release GnRH-a preparation triptorelin in prepubertal male and female rats, measuring growth, gonadal weight, hepatic GHR, and GHBP levels. We have also compared the effects of GnRH-a treatment with surgical gonadectomy on the 24-h pattern of GH secretion in both male and female animals.

METHODS

Animals. Normal rats of the NIMR:AS strain were obtained from our SPF colony at the National Institute for Medical Research at Mill Hill, London, UK. Rats were housed in light- and temperature-controlled rooms(23-25°C, 14-h light) with food and water available ad libitum. Experiments were approved by the local ethical committee for animal experiments.

Experiment 1. To assess the effects of a GnRH-a, prepubertal, 4-wk-old male and female rats (n = 6-8) were injected with a slow release preparation of triptorelin (Decapeptyl-CR, Ferring AB, Malmö, Sweden) at a dose of 1 mg/kg s.c. every 3 wk. A control group received excipient alone. Weekly measurements of body weight and body length (= nose-anus length)(26, 27) were taken, and after 15 wk of treatment, rats were killed, and blood and liver samples were taken for measurement of gonadal steroids, GHBP, and hepatic GHR binding. In addition, tibial length was measured, using callipers.

Experiment 2. At 4 wk of age, groups (n = 6-22) of rats were gonadectomized or sham-operated under halothane anesthesia. Sham-operated rats immediately received either 1 mg/kg triptorelin or excipient s.c., and this treatment was repeated every 3 wk as above. At the age of 10-11 wk, rats were transferred to individual metabolic cages. Under halothane anesthesia, a catheter was inserted in the jugular vein and attached via a swivel to an automatic blood microsampling system, previously described(20). Cannulas were flushed twice daily with heparinized(20 units/mL) saline. After 2 d of recovery, serial blood samples (20 μL every 10 min) were collected for 24 h. GH was measured directly by RIA (using NIDDK reagents) in diluted whole blood(21) and expressed as nanograms/mL in terms of the rat GH-RP-2 standard. The minimal detectable concentration was 0.4 ng/mL, and the combined sampling and interassay variability was estimated at 10%.

Experiment 3. In two other experiments, groups of rats(n = 5-7) were sham-operated, gonadectomized, or treated with triptorelin at the age of 4 wk as before. In the first experiment, they were weighed and measured weekly until the age of 7 wk, when they were killed, and plasma and liver samples were taken for GHBP and GH-radioreceptor assays. In the second experiment, they were weighed and measured weekly until the age of 10.3 wk.

Analysis of GH profiles. Pulsar analysis (PC Pulsar Program version 2.0)(28) was used to analyze the 24-h GH profiles, with the following parameter settings: peak weight 0.05, splitting parameter 2.7, smoothing time 0.5, threshold value 0.01, assay parameters A, B, C, 0, 10, and 0, respectively, and peak detection parameters G 1-5 set to 5, 3, 2, 1.5, and 1, respectively.

GHR and GHBP assays. GHR binding was assayed as previously described using radioiodinated recombinant bovine GH(27), essentially following the method of Maiter et al.(29). Results are presented as percent specific GH binding per mg of protein(30). Plasma GHBP level was measured by RIA(31). Samples were assayed in duplicate. The detection limit was 1.2 μg/L, and the interassay coefficient of variation was 10%.

Estrogen and testosterone assays. Gonadal steroids were measured by RIA, using commercial kits (RIA-mat Byk Sangtec, Dietzenbach, Germany for testosterone; RSL direct, ICN Biomedicals, Zoetermeer, The Netherlands for 17β-estradiol). The detection limit of the testosterone RIA was 0.1 ng/mL. 17β-Estradiol was measured after extraction of the plasma on C18 columns (3 mL/200 mg) (Varian), concentrating the samples 4-fold by freeze drying. The detection limit of the RIA was 1 pg/mL.

Statistics analysis. Results are presented as mean ± SEM. Data of all groups in an experiment were subjected to ANOVA followed by Student Neuman Keuls test or Fisher's protected LSD test. Growth curves were analyzed using a multivariate ANOVA for repeated measures, followed by t tests.

RESULTS

Experiment 1. Figure 1,a and b, shows the effects of chronic triptorelin treatment on body weight and body length of male and female rats. At 4 wk of age, growth rates begin to diverge in male and female rats. In females, triptorelin treatment accelerated both gain in weight and longitudinal growth. The effect became significant at 7 wk of age and was maximal by 11 wk of age, after which the growth rates were similar. Acceleration of growth in triptorelin-treated females was expected, because ovariectomy also results in increased growth rate. In contrast, the expected deceleration of growth in males, which is seen after surgical gonadectomy, was not observed after triptorelin treatment. Triptorelin in males even increased weight gain after 9 wk of treatment and gain in length after 11 wk of treatment. Tibial length (at sacrifice) was increased by triptorelin in females (3.52 ± 0.02 versus 3.62 ± 0.02 cm (saline versus triptorelin, respectively), p < 0.01), but not in males (3.98 ± 0.01 versus 3.97 ± 0.02 cm (saline versus triptorelin, respectively), p > 0.05).

Figure 1
figure 1

Growth curves for (a) weight and(b) body length for saline and triptorelin (1 mg/kg s.c., once every 3 wk)-treated male and female rats. *p < 0.05, **p < 0.01 vs saline-treated rats of the same sex (indicating significantly different curves).

Full size image

As expected, both testis and ovarian weights were significantly reduced after 15 wk of triptorelin treatment (Fig. 2a), but the effect was more marked in females than in males. None of the triptorelin-treated females showed vaginal opening, whereas all saline-treated females did, confirming that triptorelin effectively delayed puberty.

Figure 2
figure 2

The effect of 15-wk triptorelin treatment on(a) gonadal weight, (b) hepatic GHRs, and (c) GHBP. Rats were treated with triptorelin (1 mg/kg s.c. once every 3 wk) from the age of 4 wk. *p < 0.05, **p < 0.01 vs saline-treated rats of the same sex. ††p < 0.01vs saline-treated males.

Full size image

As expected, plasma GHBP levels were higher in females than in males (Fig. 2c). Triptorelin treatment in females reduced GHBP levels completely to male values, whereas this treatment had no effect on GHBP in males (Fig. 2c). Hepatic bovine GH binding was also higher in females than in males (Fig. 2b), but there were no statistically significant effects of triptorelin in either group.

Testosterone levels were significantly lower in triptorelin-treated male rats compared with saline-treated ones (0.73 ± 0.03 nmol/L (range 0.6-0.8) versus 4.2 ± 1.3 nmol/L (range 1.2-7.2) p< 0.01). Estrogen levels, sampled at various stages of the estrus cycle, ranged from 2.6 to 41.97 pg/mL (mean 15.24 ± 6.1) in saline-treated females, compared with 1.55-2.41 pg/mL (mean 1.83 ± 0.29) in triptorelin-treated females, but, due to the large variation, this did not reach statistical significance.

Experiment 2. Figure 3 shows representative 24-h profiles of GH secretion obtained from groups of sham-operated, gonadectomized, or triptorelin-treated male and female rats. The sham-operated male rats (Fig. 3a) showed 3-h episodic secretory bursts of GH secretion separated by periods of low baseline GH output, typical of normal male rats. At the other extreme, sham-operated female rats (Fig. 3f) exhibited a much more continuous irregular pattern with frequent, smaller GH peaks over a higher GH baseline, and an increase in pulse height after the onset of darkness, typical of normal female rats. Pulsar analysis (Fig. 4) confirmed that males had significantly higher GH peak heights and widths, but secreted fewer pulses with a longer interpulse interval and a lower baseline than do females. Despite these differences in pattern, the mean GH levels (and thus area under the curve) were not different between sham-operated males and females.

Figure 3
figure 3

Representative spontaneous 24-h GH secretion in a(a) sham-operated saline-treated (intact) male, (b) triptorelin-treated male, (c) gonadectomized male, (d) ovariectomized female, (e) triptorelin-treated female, and(f) sham-operated saline-treated female. Note the gliding scale from typical male (a) to typical female (f) GH secretion pattern, especially for GH baseline and peak amplitude and interval. Gonadectomy, sham-operation, or start of triptorelin treatment (1 mg/kg s.c. once every 3 wk) was at 4 wk of age and blood microsampling at 10-11 wk of age.

Full size image
Figure 4
figure 4

Pulsar analysis of 24-h GH secretion profiles of control, triptorelin-treated, or gonadectomized male and female rats. Gonadectomy, sham-operation, or start of triptorelin treatment (1 mg/kg s.c. once every 3 wk) was at 4 wk of age and 10-min blood microsampling at 10-11 wk of age. *p < 0.05, **p < 0.01 vs saline-treated rats of the same sex. †††p < 0.001, ††p < 0.01 vs saline-treated males.

Full size image

Figure 3b-e show the transition between these GH patterns affected by chemical or surgical gonadectomy. Triptorelin treatment in males (Fig. 3b) had no effect on the episodic pattern of GH, but the amplitude of the GH peaks was significantly reduced(p < 0.05, Fig. 4d). Gonadectomy in males had a little more pronounced effect (Fig. 3c). The regular GH pattern was somewhat disturbed, shown by a marginal increase in number of peaks (p = 0.058, Fig. 4c) and decrease in peak length (p < 0.05, Fig. 4f) and peak interval (p < 0.05, Fig. 4e). Changes in baseline (p = 0.09, Fig. 4b) and peak amplitude (p = 0.10, Fig. 4d) did not reach statistical significance. Ovariectomy in female rats (Fig. 3d) resulted in a profound decrease of baseline GH level (p < 0.01, Fig. 4b), and therefore decreased mean GH levels (p < 0.05, Fig. 4a), but none of the other parameters changed significantly. Triptorelin treatment in female rats (Fig. 3e) also decreased baseline levels (p < 0.01), but to a lesser extent than ovariectomy. Triptorelin-treated females also had a higher number of GH peaks(p < 0.01, Fig. 4c) than did sham-operated or ovariectomized females.

Experiment 3. Because triptorelin treatment was less effective than gonadectomy in altering GH secretion, we compared the short-term effects of surgical and chemical gonadectomy on other parameters. Male and female rats were gonadectomized or injected with triptorelin at the age of 4 wk and killed 3 wk later. Gonadectomy decreased gain in weight (less so in length) in male rats and increased gain in weight and length in female rats (Fig. 5,a and b). Triptorelin treatment was less effective than surgical gonadectomy in both males and females, and increased weight gain only in female rats significantly (p < 0.01). The longer experiment (until the age of 10.3 wk) gave similar results: in males, gonadectomy decreased weight and length gain (p < 0.05), but triptorelin had no effect (Table 1A). In females, ovariectomy and triptorelin treatment increased weight and length gain compared with sham-operated rats (p < 0.01). Weight and length of tritptorelin-treated females was halfway between that of sham-operated and ovariectomized rats until the age of 7 wk, after which they caught up to the level of ovariectomized rats (Table 1B). When looking more in detail to the weight data in the male rats, a similar trend was seen: weight was in between that of sham-operated and gonadectomized rats until the age of 7 wk, after which triptorelin-treated rats accelerated their growth.

Figure 5
figure 5

Effect of short-term gonadectomy or triptorelin treatment on (a) weight gain, (b) body length gain,(c) hepatic GHRs, and (d) plasma GHBP. Treatment started at 4 wk of age, and sacrifice was at 7 wk of age. *p < 0.05,**p < 0.01 vs saline-treated rats of the same sex.†††p < 0.001, ††p< 0.01 vs saline-treated males.

Full size image
Table 1 Treatment results
Full size table

Females had higher GHBP levels than males. Neither surgical nor chemical gonadectomy changed levels significantly, though GHBP levels in triptorelin-treated female rats lay between those of sham-operated and ovariectomized rats. Hepatic bovine GH binding was not significantly different between groups but showed the same tendency (Fig. 5,c and d). Again, bovine GH binding in triptorelin-treated rats was close to binding in gonadectomized animals though less so in males than in females.

DISCUSSION

Continuous GnRH exposure leads to a desensitization of the pituitary and thus to a diminished gonadotropin release and reduced gonadal steroid production(1). Long-acting GnRH-a are therefore used to delay puberty and prevent premature closure of growth plates(8, 9). However, such treatment also modulates height velocity, which could be due to an effect of GnRH-a treatment on GH secretion, hepatic GHRs or GHBP. This is difficult to assess in children in a controlled manner. It has been reported that GnRH-a do not change GH secretion(16) or decrease peak amplitude(15) and that GHBP is not changed(15) or increased(18) during GnRH-a treatment in children with precocious puberty. We therefore decided to investigate the effects of GnRH-a treatment on this GH-GHR/GHBP growth axis in a rat model, because the regulation of GH secretion, number of GHRs, and GHBP levels by gonadal steroids can be directly assessed in the rat(24, 25, 32, 33).

Rats of the AS strain are in puberty between the age of approximately 28 and 42 d. Females have their vaginal opening around 35-40 d of age, which characterizes the end of puberty. Male rats start to grow faster than females from the age of 4-5 wk, and at that age GH secretion also becomes sexually dimorphic. Neonatal gonadectomy in rats reduces growth in males and increases growth in females from the age of 30-35 d. The effect of prepubertal gonadectomy (25 or 30 d of age) in males is much less pronounced but still present, and prepubertal ovariectomy in females clearly stimulates growth(24).

This is the first report to evaluate the growth of rats treated with triptorelin from prepubertal age until adulthood. Like prepubertally ovariectomized rats(24), triptorelin-treated female rats showed enhanced growth, both in terms of weight and longitudinal growth. This enhancement of growth started at the age of 6 wk, approximately 2 wk after initiation of treatment, and continued for about 5 wk, after which the growth curve was parallel to that of saline-treated rats. A time lag was expected given the initial stimulatory phase of such GnRH-a before desensitization of the pituitary occurs(1). However, it also probably reflects a time course of steroid disappearance because surgical ovariectomy also does not affect growth immediately. Triptorelin treatment had less effect on the growth rate in male rats than in females. This agrees with earlier findings that prepubertal orchidectomy has little effect on growth(24), and suggests that growth during and after puberty is more sensitive to estrogen than to testosterone in rats.

We then wished to know whether changes in GH secretion could be responsible for the differential effects of triptorelin on growth in the two sexes. As is well known(19, 34), females have a higher baseline GH and more frequent shorter pulses of lower amplitude than do males. Pulsar analysis showed that triptorelin treatment in males affected only peak amplitude. We previously demonstrated that increases in GH peak amplitude accelerated growth in GH-deficient rats(31). Despite the decreased GH peak amplitudes, triptorelin did not inhibit growth. In females, baseline GH levels were most clearly affected by triptorelin. We recently showed that varying GH baseline independently of GH peak had little effect in growth in gonadally intact GH-deficient rats(31). Given these relatively small effects on GH secretion, it is more likely that the effects of triptorelin on growth are due to a reduction of the direct inhibiting effect of estrogen on growth(35) or to the deceleration of epiphyseal maturation in the absence of estrogen. Support for the latter option is given by the fact that a mutation in the human estrogen receptor caused a failure in growth plate closure and continuing growth in a man(36).

Although longitudinal growth, measured by body length and tibia length, was only slightly affected by long-term triptorelin treatment in male rats, their body weight was increased, the opposite to what occurs after gonadectomy in males. This might point to an increased fat mass. Although its relevance is doubtful, we calculated rat BMI as BMI = weight/nose-anus length2. BMI was increased after triptorelin treatment in both males and females (results not shown). Such increase in BMI has also been observed during triptorelin treatment in children(9, 37). Apart from fat mass, BMI is dependent on bone and muscle mass. Because it is unexpected that triptorelin treatment will increase muscle or bone mass, fat mass probably increases during triptorelin treatment. From our experiments we cannot explain this effect but leptin, the hormone involved in body weight regulation and reducing food intake in leptin-deficient mice, also stimulates the reproductive system and increases LH and FSH levels in those mice(38). Thus, a link between regulation of the GnRH-LH/FSH axis and leptin might exist.

When looking at the growth curves of Figure 1, there seems to be a time difference in the effect of triptorelin on growth in the two sexes with a relatively early effect in females compared with males. However, the effect of triptorelin on growth is the sum of positive and negative effects of triptorelin on growth: if triptorelin indeed has a fat mass and weight-increasing effect, as described above, then this would counteract the effect of reduced testosterone levels (weight-decreasing effect) in males but be additive to the effect of decreased estrogen levels(weight increasing effect) in females. This could be an explanation for the apparent time difference in the effect of triptorelin in males versus females. Experiment 3 shows that, until about 3-4 wk after the start of triptorelin treatment, the effect of triptorelin is in the direction of that of gonadectomy in both sexes, but that afterwards an acceleration in weight gain in both genders appears, suggesting a second way of action of triptorelin, besides suppression of gonadal steroid production.

Gonadectomy affected different components of GH secretion in males and females(24). Ovariectomy decreased baseline GH without altering peaks, whereas orchidectomy affected peak parameters without changing baseline GH significantly. One interpretation of these data could be that ovariectomy affects mainly somatostatin tone, whereas the most prominent effects of orchidectomy in males is on GRF. This agrees with experiments in which GH profiles were studied in sham-operated and gonadectomized male rats with and without estrogen treatment(39) and gains support from in vitro experiments showing that orchidectomy reduces the responsiveness of somatotrophs to GRF(40). Furthermore, manipulation of testosterone levels affects the number of GRF neurons and GRF mRNA levels per neuron(41), whereas both estrogen and testosterone affect somatostatin mRNA(42).

Triptorelin treatment had a much smaller impact on GH secretion than did surgical gonadectomy. In males, the only factor affected was peak amplitude, and in females baseline GH levels fell, but not to the same extent as in ovariectomized rats. Inspection of the profiles shows that the GH patterns of triptorelin-treated animals lie somewhere between the extremes of the patterns of intact and gonadectomized animals. One exception is that the number of GH peaks was much higher in the triptorelin-treated females than in any of the other groups. GnRH-a treatment may affect GH pulsatility via mechanisms other than suppression of gonadal steroid production. The neuroendocrine status is different in gonadectomized animals, which have high gonadotropin and GnRH levels(43) compared with triptorelin-treated animals which have blocked GnRH signaling and suppressed gonadotropins(1–3). This different neuroendocrine status could result in a different effect on the hypothalamus-pituitary-GH axis in gonadectomized compared with triptorelin-treated rats and may account for the differences we found between the GH profiles in these groups.

We also measured hepatic GHR binding and plasma levels of GHBP, which is an alternatively spliced product of the GHR in the rat(44). Triptorelin had no effect on GHR binding, but did after long-term treatment reduce the higher plasma GHBP levels seen in normal female rats(23) to those in males, confirming earlier results for ovariectomy(25). Because GHBP levels are selectively sensitive to baseline GH exposure independent of GH peaks in the rat(31), this effect of triptorelin is probably mediated via the reduction in baseline GH exposure. We have previously commented on the inverse relationship between growth rate and GHBP that can be observed with different estrogen states in the rat(25), but the significance for growth of such changes in GHBP is unclear because GHBP can both enhance and inhibit GH action(45, 46) and the mechanism of generation of GHBP and its sensitivity to gonadal steroids is very different in humans(33, 47, 48). Triptorelin treatment in male rats did not affect GHBP levels, in line with results for orchidectomy and experiments showing that GHBP levels are not affected by testosterone(49).

Because the effects of triptorelin on growth and bone maturation in children can also be quite variable, one can question whether triptorelin suppresses gonadal steroids completely(50, 51), and whether the effectiveness of triptorelin treatment on growth parameters equals that of complete removal of gonadal steroid production.

One possible explanation for the less marked or different effects of triptorelin than of gonadectomy is that the dose was inadequate to achieve complete gonadal suppression. Surprisingly, there is little literature about the doses of long-acting triptorelin necessary for suppression of gonadal steroid production in rats. One article states that a dose of 300 μg/kg i.m. provokes pituitary desensitization in rats(4). We chose a dose of 1 mg/kg s.c. every 21 d, which is thus probably supramaximal, also compared with the usual human dose of 3.6 mg per injection, independent of body weight, every month. This dose was clearly effective, because both testis and ovarian weights were reduced by 3 wk of treatment (data not shown) and remained low over 15 wk of treatment, and plasma steroid measurements confirmed the suppression of the pituitary gonadal axis. However, ovarian weight was more reduced than testes weight (70% versus 20% decrease). Testis weight, determined by germ cell epithelium and Sertoli cells, depends mainly on the number of Sertoli cells produced during perinatal development(52) so that our later GnRH-a treatment might not be very efficient to reduce testis weight. Furthermore, testis weight and Sertoli cell number is regulated by FSH(53), which may not be completely suppressed by GnRH-a treatment(54).

Another possible explanation for the difference between surgical and chemical gonadectomy is the presence of other factors in the gonads affecting the hypothalamus-pituitary-growth axis, which are not regulated by LH/FSH and are thus not suppressed by GnRH-a. For example it is not clear how GnRH-a will affect the inhibin/activin system, which might have significant impact on GH production(55). Conversely, there may be factors other than LH that stimulate gonadal steroid production and are less affected by GnRH-a. One factor produced in the testis that stimulates steroidogenesis by Leydig cells has recently been isolated from Sertoli cells and is composed of TIMP-1 and the proenzyme form of cathepsin L(56). TIMP-1 is also produced in luteal and granulosa cells of the ovary and stimulates steroidogenesis in the ovary. Because FSH regulates TIMP-1, and may not be completely suppressed by GnRH-a(54, 57), TIMP-1 might continue to stimulate low level gonadal steroid production in the triptorelin-treated rats.

A third possibility is that extrapituitary GnRH receptors are affected differently by triptorelin treatment than pituitary GnRH receptors. There are GnRH receptors in the gonads and the hypothalamus, but also in extrahypothalamic regions, including the hippocampus and the limbic system, via which GnRH might modulate behavior(58, 59). In addition, GnRH acts as a neuromodulator/neurotransmitter in some brain areas, and GnRH is, apart from the nerve fibers in the median eminence, also released from nerve fibers near the third and lateral ventricle, maybe into the cerebrospinal fluid enabling to act at distant sites(60). It is unknown how continuous GnRH-a treatment will affect these processes. To further investigate these possibilities, more experiments are needed, for example experiments in which castrated rats will be treated with GnRH agonists.

In conclusion, we have for the first time investigated the effect of the GnRH-a triptorelin on growth in the rat and related it to changes in GH secretion, hepatic GHR, and GHBP. Triptorelin treatment was not as effective as surgical gonadectomy on most parameters tested and affected different aspects of the GH secretory pattern in male and female rats. Triptorelin treatment stimulated growth in female rats, although decreasing their GHBP and baseline GH secretion, but the stimulation of growth might be largely due to a reduction of the direct inhibiting effect of estrogen on growth. Triptorelin treatment had much less impact in males and slightly stimulated growth, although decreasing the amplitude of GH pulses. The stimulation of weight gain by triptorelin in male rats, which is opposite to the effect of surgical gonadectomy and unexpected given the decrease in GH secretion, could indicate an interference of triptorelin in the control of body weight.