Clinical Endocrinology (1990), 33, 399-406

THE INTERACTION BETWEEN CLONIDINE AND GROWTH HORMONE RELEASING HORMONE IN THE STIMULATION OF GROWTH HORMONE SECRETION IN MAN D. SURI, P. C. H I N D M A R S H , C. E. BRAIN, P. J . P R I N G L E C. G . D. BROOK

AND

The Endocrine Unit, The Middlesex Hospital, London, UK (Received 5 February 1990; returned.for revision 9 March 1990;$nally revised 28 March 1990; accepted 23 April 1990)

SUMMARY

Six normal adult males were given clonidine and GHRH either separately, or in combination, in random order. The peak serum G H concentrations elicited by clonidine or GHRH were variable but one factor influencing the G H response to GHRH was the G H secretory status in the hour prior to the administration of the GHRH. Peak serum G H concentrations attained were significantly greater when serum G H concentrations were rising (mean 52.9 mU/l, SD 17.2) than if they were falling (mean 27.5 mU/l, SD 13.3) or unchanged/undetectable (mean 20.6 mU/1, SD 9.8) (one-way ANOVA, F= 8.77; P = 0.004). The G H response to clonidine was not influenced by the secretory status in the hour prior to administration of clonidine. Pretreatment with clonidine did not augment the peak serum G H response to GHRH but the direction of response was more predictable than when GHRH was administered separately or repeatedly. Prior treatment with GHRH(1-29)-NH* led to a marked attenuation of the peak serum G H response to clonidine. These results suggest that the alpha-2 adrenergic agonists probably stimulate G H secretion through pathways other than just GHRH. G H secretion is under the dual control of growth hormone releasing hormone (GHRH) and somatostatin (SS). These hypothalamic peptides are in turn regulated by a number of neurotransmitters. Noradrenergic pathways acting via alpha-2 receptors stimulate G H secretion (Durand et al., 1977; Willoughby & Day, 1981). Administration of alpha adreno-receptor antagonist drugs reduces the G H response to hypoglycaemia and L-dopa while alpha adrenoceptor agonists, such as clonidine, increase basal GH levels (Wass, 1983). G H responses to insulin-induced hypoglycaemia can be specifically blocked by the alpha-2 antagonist yohimbine (Tatar & Vigas, 1984). It is unclear which of the two hypothalamic peptides mediates the alpha-2 adrenergic effects. In the rodent, the stimulatory effect of clonidine on G H secretion is abolished after Correspondence: Professor C. G . D. Brook, Endocrine Unit, Middlesex Hospital, Mortimer Street, London WIN SAA, UK.

399

400

D . Suri et al.

passive immunization with anti GHRH antibodies (Miki et al., 1984), although a mechanism operating to inhibit hypothalamic S S release cannot be excluded (Cella et al., 1987). In humans, pretreatment with GHRH 1-44 abolished the G H response to the subsequent administration of GHRH 1-44 without affecting the G H response to clonidine (Dieguez et al., 1988). Further, alpha adrenergic receptor agonism increases plasma G H levels at a time when the G H response to supramaximal doses of GHRH 1-44 is absent (Valcavi et al., 1988). Clonidine has been used to promote growth in short children (Pintor et al., 1987), although not all groups have achieved significant improvements in growth rate (Pescovitz &Tan, 1988). As an enhancement of GHRH-induced secretion has been proposed as one of the mechanisms of this improved growth rate (Pintor et al., 1987; Alba-Roth et al., 1989), we have reinvestigated the interaction of clonidine and GHRH on G H secretion in the human. If clonidine were to augment GHRH secretion then this would have important implications in the management of short, slowly growing children. Clonidine could be used either to enhance GHRH tone and G H secretion or as a supplement to GHRH treatment. METHODS Subjects Six healthy adult male volunteers aged between 21 and 23 years participated in the studies. All were of normal height and weight. The protocol was approved by the Ethics Committee of the Middlesex Hospital. Informed consent was obtained in all cases.

Protocol Following an overnight fast an indwelling intravenous cannula was inserted in the forearm at 0800 h. Thirty minutes later a sample was drawn for estimation of serum G H concentration and the process repeated thereafter at 15-min intervals for 300 min. Sixty minutes after the first sample was drawn the studies shown in Fig. 1 were performed. The dose of oral clonidine used was 0.15 mg/m2 body surface area and the intravenous bolus injection of GHRH( 1-29)-NHz (Kabi, Stockholm) was 100 pg. The studies were carried out with a 2-3 day interval between each of the protocols. Assay

Samples from each individual were spun, separated and stored at -20°C prior to measurement in the same assay. Serum G H concentrations was measured using the Tandem-R Immunoradiometric kit (Hybritech, Europe). The interassay coefficient of variation was 10.8% at 5.9 mU/1 and the intra-assay coefficients of variation at serum GH concentrations of 1.4, 7.1,26.4 and 99.4 mU/1 were 10.6, 7.8,4.9 and 4.9% respectively. The sensitivity of the assay was 0.5 mU/L The standard used was HS2443E (NIH) which had been recalibrated to mU/1 with the 1st International Reference Preparation 66/217. Statistical analysis

The Wilcoxon Rank Sign Sum Test was used to compare the peak serum G H responses to the various stimuli. Linear regression was used to determine whether the serum G H

401

Clonidine and growth hormone release Saltne/placebo Clonidine (CLON) (0.15 mg)

t

4

GHRH ( I -29) -NH, ( 100 p d CLON 4- GHRH ( 1 - 2 9 ) -NHE GHRH ( I -29) -NH,

4

+CLON

+

GHRH ( 1 - 2 9 ) -NH, GHRH (1-29) -NH,

I -60

+t +

4 t

4

I

I

I

I

I

0

60

120

180

240

Time ( m i n )

Fig. 1. The study protocol performed with either saline, clonidine (0.15 mg/m2) alone, GHRH (1 -29)-NH2 (100 pg i.v.) alone, or combinations.

response to clonidine or GHRH( 1 -29)-NH2 predicted the response on a subsequent occasion to the same stimulus. The criteria for assessing pre-intervention GH secretory status were similar to those of Devesa et al. (1989). Serum GH concentrations in the preceding hour were described as: (a) unchanged/undetectable, if there was no significant change in serum GH concentration or the values were all below the sensitivity of the assay; (b) rising, if the serum concentration increased by 5 mU/l or more; and (c) falling, if serum GH concentration decreased by 5 mU/I or more. One-way analysis of variance (ANOVA) with Duncan’s after test applied was used to assess the influence of pre-intervention GH secretory status on the subsequent serum GH response to an intravenous bolus injection of GHRH (Snedecor & Cochran, 1980). RESULTS

No effect of saline on serum GH concentration was observed. The serum GH concentrations during this observation period were all below the limit of detection of the assay. Clonidine studies The oral administration of clonidine (0.15 mg/cm2)produced a wide range of peak serum GH response (median 14.7 mU/I, range 1.0-35.7) (Fig. 2a). When clonidine was administered on a separate occasion a similarly wide range of response was observed (median 4.2 mU/l, range 0.5-30.6) (Wilcoxon test, P=O. 17). There was no correlation between the peak serum GH responses to two doses of oral clonidine ( r = -0.1 1, P=0.8 I).

D.Suri et al.

402

50

-y 5 ;

-

40-

3

E

30-

20-

10

-

0-

u Day I

u Day2

u GHRH

Clonid ine

u CLON 2h post

GHRH

Fig. 2. Peak serum GH concentrations following clonidine (0.15 mg/rn2)given a, on two separate occasions; b, 2 h following a previous dose of GHRH (100 pg i.v.).

The peak serum GH concentration obtained in response to clonidine was markedly reduced by the prior administration of an intravenous bolus injection of GHRH( 1-29)NH2 (Fig. 2b) (Wilcoxon test, P=0.03).

GHRH studies An intravenous bolus injection of GHRH( 1 -29)-NH2 produced a prompt increase in serum GH concentration and there was no significant difference between the peak serum GH response to GHRH(l-29)-NH2 obtained on two separate occasions (median day 1, 27-6 mU/l, range 5.4-31.0; median day 2, 40.3 mU/1 range 13.7-17.9) (Wilcoxon test P=0.75) (Fig. 3a). The peak serum GH response to a second intravenous bolus injection of GHRH(l-29)-NH2 given 2 h after the first was not significantly different from that observed after the first injection (median 1st injection 25.8 mU/l, range 5 - 4 - 5 1 ; median second injection 22.0 mU/1, range 5.4-75.1) (Wilcoxon test, P=0.25) (Fig 3b). There was no correlation between the peak serum GH response observed following an injection of an intravenous bolus of GHRH( 1 -29)-NH2 on one occasion compared to that observed on a second ( r = -0.33, P=0.61). Although the peak serum GH response to GHRH( 1-29)-NH2 administration after prior injection of clonidine was greater than the peak serum GH response to clonidine (Wilcoxon test, P=O.O5) (Fig. 3c) the peak response was no different when compared to that observed if GHRH( 1-29)-NH2 was administered on two separate occasions or if GHRH( 1 -29)-NH2 was administered 2 h after a dose of GHRH( 1 -29)-NHz (Wilcoxon test, P > 0.05 for all combinations). However, when the direction of serum GH response

403

Clonidine and growth hormone release ‘lo[

70

-

60

-

50

-

80.

2

\ 3

E

( a )

( b )

( C )

40 0

3020 10

-

I

-

: 0

0’

U

Dav 2 GHRH

u u GHRH

GHRH 2h later

1111 CLON GHRH

2 h after CLON

Fig. 3. Peak serum GH concentrations following GHRH (100 pg i.v.) given a, on two separate occasions; b, 2 h following a previous dose of GHRH (100 pg i.v.);c, 2 h following a previous dose of clonidine (0.15 mg/m2).

to two bolus injections of GHRH (three out of six increase) are compared to oral clonidine followed by GHRH (five out of six increase) regularization of the GH response is observed. Pre-administration G H secretory status A total of three GHRH studies using GHRH only were performed on each subject. The effectsof the pre-administration GH secretory status on the peak serum GH response to an intravenous bolus injection of GHRH(1-29)-NHz are shown in Fig. 4. The values

50-2 I 40. L5

E 30-

20m

10

-

Rising

Falling Unchanged/ undetectable

Fig. 4. Peak serum GH concentration achieved following GHRH when pre-administrationserum GH concentrations are: rising (n = 5); falling (n= 5); or unchanged/undetectable (n = 8). Data shown as mean+SEM.

404

D.Suri et al.

obtained when serum G H concentrations were rising were significantly greater than those obtained when serum G H concentrations (Fig. 4) were decreasing or unchanged/ undetectable (one-way ANOVA, F= 8.77; P = 0.004: Duncan procedure: rising versus decreasing P < 0.05, rising versus unchanged/undetectable P < 0.01). DISCUSSION These data demonstrate that the pituitary gland is capable of responding to successive doses of GHRH given 2 h apart. This differs from the observations of Shibasaki et a/. (1985) and Valcavi eta/. (1988) who reported that two successive doses of GHRH failed to increase the serum G H levels on the second occasion. Closer inspection of those data show a small elevation in serum G H concentration at the time a second episode might be expected. The descending limb of the GHRH-induced G H pulse was shifted to the right at the time that the second stimulus was applied suggesting that a secretory episode had taken place. Deconvolution analysis of such data reveals that two distinct secretory episodes had to have taken place in order to generate the serum GH concentrations measured (Hindmarsh e t a / . , 1990). Further, both Shabasaki et a/. (1985) and Valcavi et a/. (1988) used supramaximal stimulatory doses which may have influenced the results obtained. Clonidine did not appear to enhance the G H response to GHRH but it does seem to regularize it. These findings would support the observation of Valcavi et a/. (1988) but are different from those of Alba-Roth et al. (1989). The dose of clonidine we used was less than that used by Alba-Roth et al. (1989) but this does not explain the differences since several groups have demonstrated that 150 pg of clondine given orally is adequate to stimulate GH secretion (La1 et a/. 1975; Gil-Ad et a/., 1979; Hoffman et a/., 1989). No control data were provided in the studies of Alba-Roth et a/. (1989) in that single and double administrations of GHRH were not performed. Our data show that this information is essential for the interpretation of responses when applied in a repetitive fashion. Finally, their proposal that clonidine leads to stimulation of G H via endogenous GHRH in man was based on studies performed on isolated rat anterior pituitary cells. In the rat there is good evidence that activation of alpha-2 adrenoceptors stimulates G H secretion by a direct or indirect effect on endogenous GHRH release (Cella et a/., 1987; Miki et a/., 1984; Bruhn et al., 1989). The situation would appear to be different in the human as the data of ourselves and Dieguez et a/. (1988) indicate. The lack of G H stimulation by clonidine after pretreatment with GHRH is not due to depletion of G H stores in the pituitary gland as repeated stimulation of the pituitary gland with GHRH elicited a G H response. This suggests that the GHRH-stimulated pathway within the pituitary is functioning at this stage. A more likely explanation is that clonidine does not act predominantly through GHRH but rather by reducing SS tone. Studies in rabbits demonstrating involvement of the alpha-2 adrenergic system in the regulation of GHRH and S S secretion (Minamitani et a/., 1989) would support this hypothesis. Reduced SS tone would explain the regularization of the GH response to GHRH, an effect recently observed also in rodents (Sato e t a / . , 1989) and pigs (Dubreuil et al., 1989). The magnitude of the response would not be altered if clonidine acted through SS as this peptide has not been implicated in GH synthesis. The reduced response to clonidine after GHRH would also be consistent with SS mediation as pretreatment with GHRH stimulates SS release from the hypothalamus (Aguila & McCann, 1985).Clonidine would

Clonidine and growth hormone release

40 5

be able to reduce hypothalamic SS tone to some extent but insufficiently to restore full responsiveness. This study has demonstrated a wide variation in response to both clonidine and GHRH. The GH response to either of these two agents has poor predictive value, which confirms other studies (Hoehe et af.,1988). The reason for this discordance is unclear but as GH secretion is pulsatile and as an endogenous rhythm of GHRH and SS has been postulated (Plotsky & Vale, 1985; Tannenbaum & Ling, 1984; Brain et al., 1988), interaction of the stimulatory agents with these endogenous rhythms is possible. Devesa et ai. ( I 989) have demonstrated the importance of the serum GH concentration in the hour preceding administration of GHRH on the subsequent GH response to GHRH, and our data support their findings. Administration of GHRH during a period when serum GH concentrations were rising led to a better response than when serum GH concentrations were falling or remained unchanged/undetectable. This would accord with the hypothesis of a falling SS tone coupled with rising portal serum GHRH concentrations leading to the GH secretory burst. Disordered GH secretion might result, therefore, not simply from insufficient GHRH secretion but from alterations in SS tone. In conclusion, clonidine would apear to stimulate GH secretion mainly through a GHRH-independent mechanism in normal human subjects. The serum G H response to clonidine and GHRH is variable and not easily reproducible. An important factor in explaining the wide variation in response to GHRH appears to be the GH secretory status in the hour preceding the administration of GHRH. We suggest that the interpretation of pharmacological tests of GH secretion need to take account of the secretory status prior to administration of the secretagogue.

REFERENCES

AGUILA,M.C. & MCCANN.S.M. (1985) Simulation of somatostatin release in-vitro by synthetic growth hormone-releasing factor by a non dopaminergic mechanism. Endocrinology, 117, 762-765. ALBA-ROTH, J., LOSA,M., SPIESS,Y., SCHOPOHOL, J . , MULLER, A. & VON WERDER, K. (1989) Interaction of clonidine and GHRH in GH secretion in-vivo and in-vitro. Clinical Endocrinology, 30, 485-491. BRAIN, C., HINDMARSH, P.C., BROOK,C.G.D. & MATTHEWS, D.R. (1988) Continuous subcutaneous growth hormone releasing factor analogue augments growth hormone secretion in normal male subjects with no desensitization of the somatotroph. Clinical Endocrinology, 28, 543-549. BRUHN,T.D.. TRESCO,P.A., MULLER, G.P. & JACKSON, I.M.D. (1989) Beta-endorphin mediates clonidine stimulated growth hormone release. Neuroendocrinology, 50,460-463. CELLA, S.G., LOCATELLI, V.. WEHRENBERG, W.B. & MULLWE, E.E. (1987) Pharmacological manipulation of alpha-adrenocepters in the infant rat and effects on growth hormone secretion. Study of the underlying mechanisms of action. Endocrinology, 120, 1639-1643. DEVESA, J.L.. LIMA,N., LOIS,C., FRAGA,M.J., LECHUGA, V. & ARCETRESGUERRES, J.A.F. (1989) Reasons for the variability in growth hormone (GH) responses to GHRH challenge: the endogenous hypothalamicsomatotroph rhythm (HSR). Clinical Endocrinology, 30, 367-377. DIEGUEZ, C., PAGE,M.D. & SCANLON, M.F. (1988) Growth hormone neuroregulation and its alterations in disease states. Clinical Endocrinology, 28, 109-143. DUBREUIL, P.. PELLETIER, G . , PETITCLERC, D . , LAPIERRE, H., GAUUREAU, P. & BRAZEAU,P. (1989) Effects of active immunisation against somatostatin on serum growth hormone concentration in growing pigs: influence of fasting and repetitive somatostatin injections. Endocrinology, 125, 1378- 1384. DURANU, D.. MARTIN, J.B. & BRAZEAU, P. (1977) Evidence for a role of alpha-adrenergetic mechanisms in regulation of episodic growth hormone secretion in the rat. Endocrinology, 100, 722-728. GIL-AD,I., TROPPER, E. & LARON, 2.(1979) Oral clonidine as a growth hormone stimulation test. Lancet, ii, 278-280.

D . Suri et al. HINDMARSH, P.C., MAITHEWS, D.R., BRAIN,C.E., PRINGLE, P.J. & BROOK,C.G.D. (1990) The application of deconvolution analysis to elucidate the pulsatile nature of growth hormone secretion using a variable halflife of growth hormone. Clinical Endocrinology, 32,739-747. HOEHE,M., VALDO,G. & MATUSSEK, N. (1988) Growth hormone, noradrenaline, blood pressure and cortisol responses to clonidine in healthy adult male volunteers: dose response relations and reproducibility. Psychoneuroendocrinology, 13,409418. R.L., ARRENDALE, R.F. & HAHN,D.A. (1989) Relation of plasma HOFFMAN, W.H., DI PIRO,J.T., TACKETT, clonidine to growth hormone concentrations in children and adolescents. Journal of Clinical Pharmacology, 29,538-542. LAL,S . , TOLIS,G., MARTIN, J.B., BROWN, G.M. & GUYDA, H.(1975) Effects ofclonidine onGH, LH, TSH in the serum of normal men. Journal of Clinical Endocrinology and Metabolism, 41,827-832. MINAMITANI, N., CHIHARA, K., KAJI,H., KODAMA, H., KITA,T. & FUJITA, T. (1989) Alpha-2-adrenergic control of growth hormone (GH) secretion in conscious male rabbits-involvement of endogenous GH-releasing factor and somatostatin. Endocrinology, 125,2839-2845. K. (1984) Evidence that opiatergic and alpha-adrenergic mechanisms stimulate MIKI,N., ONO,M. & SHIZUME, rat growth hormone release via growth hormone releasing factor. Endocrinology, 114, 1950-1952. PESCOVITZ, O.H. & TAN,E. (1988) Lack of benefit of clonidine treatment for short stature in a double-blind, placebo-controlled trial. Lancet, ii, 974-877. PINTOR, G., Locm, R., CELLA,S., PUGGIONI, R., LOCATELLI, V. & MULLER, E.E. (1987) Clonidine treatment for short stature. Lancet, i, 1226-1230. PLOTSKY, P.M. &VALE,W. (1985) Patterns of growth hormone-releasing factor and somatostatin secretion into the hypophysialportal circulation of the rat. Science, 320,461-463. K., KITA,T., KASHIO, Y ., OKIMURA, Y ., KITAJIMA, N. & FUJITA, T. (1989) Physiological role SATO,M., CHIHARA, of somatostatin-mediated autofeedback regulation for growth hormone: Importance of growth hormone in triggering somatostatin release during a trough period of pulsatile growth hormone release in conscious male rats. Neuroendocrinology, 50, 139-1 5 1. SHIBASAKI, I., HOTTA,M., MASUDA, A., IMAKI, T., OBARA, N., DEMURA, H., LING,N. & SHIZUME, K. (1985) Plasma GH responses to GHRH and insulin induced hypoglycaemia in man. Journal of Clinical Endocrinology and Metabolism, 60, 1265-1276. SNEDECOR, G.W. & COCHRAN, W.G. (1980) Statistical Methods. Seventh edition. Iowa State University Press, Arnes, Iowa. TANNENBAUM, G.S. & LING,N. (1984) The interrelationship of growth hormone (GH)-releasing factor and somatostatin in generation of the ultradian rhythm of GH secretion. Endocrinology, 115, 1952-1957. TARTAR,P. & VIGAS,M. (1984) Role of alpha-I and alpha-2 adrenergic receptors in the growth hormone and prolactin response to insulin induced hypoglycaemia in man. Neuroendocrinology, 39,275-280. C., PAGE, M.D., ZINI,M., CASOLI,P., PORTIOLO, I. & SCANLON, M.F. (1988) Alpha-2VALCAVI, R., DIEGUEZ, adrenergic pathways release growth hormone via a non-GRF-dependent mechanism in normal human subjects. Clinical Endocrinology, 29, 309-316. WASS,J.A.H. (1983) Growth hormone neuroregulation and the clinical relevance for somatostatin. Clinics in Endocrinology and Metabolism, 12,695-724. WILLOUGHBY, J.O. & DAY,T.A. (1981) Central catecholamine depletion: effects on physiological growth hormone and prolactin secretion. Neuroendocrinology, 32,65-69.