Growth hormone-binding protein levels: Studies of children with short stature

Growth Hormone-Binding Protein Levels: Studies of Children With Short Stature Nelly Mauras,

Lena M.S. Carlsson,

Suzanne

Murphy,

and Thomas

J. Merimee

A high-affinity growth hormone-binding protein (GHBP) in serum is derived from the extracellular domain of the GH receptor. In an attempt to investigate the differences in GHBP levels in various conditions of poor growth, we measured GHBP levels by two methods-an Ultrogel chromatographic technique and a ligand-mediated immunofunctional assay (LIFA). The following three groups of children were studied: Turner’s syndrome (n = 7). idiopathic and/or familial short stature ([KS] n = 15). and organic or idiopathic hypopituitarism (n = 19). All groups were similar in age (Turner’s syndrome, 10.1 f 0.9 years; ISS, 10.0 f 0.7; hypopituitarism, 11.5 r 1.0) and height SEM score (Turner’s syndrome, -2.9 f 0.3; ISS, -3.0 & 0.4; hypopituitarism, -2.3 f 0.4). Their values were compared with those values of GHBP in healthy controls of similar age. lmmunofunctional assay values for GHBP were as follows: Turner’s syndrome, 235.4 -e 26.0 pmol/L; ISS, 122.4 f 11.0; and hypopituitarism, 157.1 -C 23.0. These results were significantly different in subjects with ISS and hypopituitarism as compared with a group of healthy controls between the ages of 9 and 12 years (N = 255; GHBP = 287.9 f 10.9 pMol/L; P < .OOl compared with both ISS and hypopituitarism). Similar changes were found using Ultrogel chromatography. This difference in GHBP levels is still significant even when more stringent criteria are applied to define hypopituitarism (ie, peak GH responses to stimuli ~6.0 ng/mL, instead of 510 ng/mL originally). Correlation analysis was performed for GHBP values in all three groups and several parameters of growth, including age, height, height SD score, weight, plasma insulin-like growth factor-l (IGF-I), peak GH response to stimuli, and growth velocity. The only consistent significant correlation found was between GHBP and weight in all three groups (Turner’s syndrome, r = 92; ISS, .61; hypopituitarism, .55; P c .Ol). In conclusion, girls with Turner’s syndrome have normal GHBP levels. However, in ISS a relative deficiency of the GHBP is present, similar to patients with hypopituitarism. This may represent a quantitative GH receptor deficiency, which might contribute to these patients’ poor linear growth. Copyright C 1994 by W.B. Saunders Company

T

HE HIGH-AFFINITY growth hormone-binding protein (GHBP) in serum derived from the extracellular domain of the GH receptor may reflect GH receptor number’,? even though it is not definitely shown. GHBP may confine GH to the intravascular compartment to protect it from degradation and to prolong its biological half-1ife.j GHBP levels have been reported to be both normal and low in patients with GH deficiency4 and absent in patients with Laron dwarfism.5J GHBP concentrations positively correlate with height in growing children’ and with po’oled GH values in growing females,x yet they were found to correlate negatively with 24-hour GH concentrations in boys in mid-puberty.’ This suggests that a higher level of GHBP would decrease the neuroendocrine secretory tone of GH and vice versa, implying that the interaction of GH with GHBP maybe more important in modulating the biologic effect of GH than the absolute levels alone. In an attempt to investigate the differences in GHBP levels in various syndromes of poor growth, we measured circulating GHBP levels in three groups of children with short stature and correlated their values with several parameters commonly used in the assessment of growth retardation. These data were compared with those obtained in children of comparable age and of normal stature.

SUBJECTS AND METHODS

Forty-one short children with a mean age of 10.7 2 1.7 years were selected from our pediatric endocrine clinics and evaluated prospeciively for 1 year. One group had Turner’s syndrome (n = 7), a second group had idiopathic and/or normal variant familial short stature ([ISS] n = 15). and the third had organic or idiopathic hypopituitarism (n = 19). Their clinical characteristics are summarized in Table 1. The peak GH response to stimuli was assessed using standard pharmacologic stimuli (arginine-insulin, arginine-levodopa stimulation tests) or peak spontaneous nocturMefaabo/;sm, Vol43, No 3 (March), 1994: pp 357.359

nal GH concentrations. A normal response was defined initially as being greater than 10 ng/mL. A group of 255 healthy children sampled at random during a school screening were used as controls. Their ages ranged between 9 and 12 years. and their stature was believed to represent the spectrum of variance of the general population. These subjects were part of a larger sample of controls reported previously.x Assuys

GH concentrations were measured using the Allegro Immunoradiametric assay of Nichols Institute, and plasma insulin-like growth factor-l (IGF-I) concentrations by a radioimmunoassay of the Nichols Institute, San Juan Capistrano, CA. GHBP levels were measured by two methods. a chromatographic assay using Ultrogel AcA44(‘.’ and a ligand-mediated immunofunctional assay (LIFA) described previously by Carlsson et al.9 For the chromatographic assay, 100 FL serum with [‘zSI]hGH (20,000 to 25.000 cpm) in 25 mmol/L Tris buffer was chromatographed on an Ultrogel AcA44 column measuring 0.9 x 15 cm. Each serum sample was also prepared in a similar manner with the addition of 2.5 kg uniodinated GH to determine the nonspecific binding of [‘liI]GH. Intraassay and interassay coefficients of variation were 14% and 11.6%. respectively. For the LIFA. a specific monoclonal antibody (MAb) directed against GHBP was used (MAb 263) as described previously.” Dilutions of recombinant human GHBP in phosphate-buffered

From the Nemours Children’s Clinic, Jacksonville. FL; Genentech, Inc. S. San Francisco. CA; and the Universil?,of Florida, Gainesville. FL. Submitted December 29, 1992; accepted April 1 I. 1993. Supported by the Nemours Foundation, Genentech, Inc, and National Institutes of Health Grant No. ROI DK 18130. Address reprint requests to Nellv Mauras. MD, Division of Endocrinology, Nemours Children’s Cl&. 807 Nira St, Jucksomille. FL 32207. Copyright 0 1994 by W.B. Saunders Company 0026049519414303-OO14$O.~.OOlO 357

358

MAURAS ET AL

Table 1. Clinical Characteristics of Patients Turner’s

Syndrome

ISS

(n = 7)

Height (cm)

Hypopituitarlsm

cn = 15)

10.1f 0.9

Age Iv1

Table 2. GHBP Levels (pmol/L)

121.2 & 1.6

(n = 19)

10.0 + 0.7

11.5 f 1.0

121.6 + 4.7

130.1 & 5.4

* 0.3

-3.0

+ 0.33

-2.3

Turner’s Patent

Height SD

-2.9

Weight (kg)

25.5 % 1.8

26.0 f 2.5

35.6 ? 4.9

IGF-I (IJlmL)

0.58 + 0.08

1.24 & 0.33

0.52 ? O.lO*

Peak GH (ng/mL)

28.4 2 7.7

13.7 + 1.7

4.8 2 0.7

No.

2 0.4

NOTE. Data are means + SE.

Syndrome GHBP

GHBP values in the control group did not meet the criteria fol normal distribution, and hence logarithmic transformation of the data was performed for the GHBP values in all groups using the Kolmogorov-Smirnov goodness-of-fit test.“’ Student’s t test was then applied to the log transforms of the GHBP data. Pearson Product-Moment Correlations”’ were performed to analyze the relationship between GHBP values in all three groups and several parameters, including age, height. height standard deviation score, weight. plasma IGF-I. peak GH response to stimuli, and growth velocity. Data are expressed as means + standard errors.

40

1

J

-0

5 0

m bp

20

10 ..

.

.

r = 0.78 .OOl

p<

-

0 0

2 SE

Hypopltuitarism GHBP

Patient No.

GHBP

8

95.1

23

67.6t

225.8

9

165.4

24

128.lt

231.4

10

90.5

25

134.4

164.7

11

149.0

26

473.1 110.3t

358.8

12

69.0

27

251.6

13

201.4

28

97.8

152.2

14

128.4

29

204.7

15

133.7

30

72.4t

16

74.3

31

158.6t 234.3

235.4 ? 26.0

17

47.9

32

18

90.7

33

58.6

19

139.1

34

266.3

20

159.8

35

21

154.1

36

93.2

22

137.9

37

198.0

38

130.0

122.4 2 11.0x

61.8t

39

72.9

40

207.1

41

216.3t 157.1 2 23.0*

*P c ,001 compared with controls after log transformation of data. tPatients considered as ISS when hypopituitarism is redefined as peak GH < 6.0 ng/mL.

Data Analysis

30

Mean

ISS Patient No.

263.3

*Results include IGF-I levels before treatment (n = 14).

saline containing bovine serum albumin, EDTA. Tween 20. and Thimesoral (assay buffer) were used as standards in this assay. In essence. the microliter plates were coated with MAb 263 directed against GHBP. blocked for 2 hours, and washed. and then the assay sample was added with either 0 or 200 +g/L rhGH and incubated for 2 hours. After a second wash step, horseradish peroxidaseconjugated anti-GH MAb was added to detect the bound GH. After 2 hours of incubation, the plates were washed and substrate was added. The reaction was terminated after 15 minutes, and an optical density of 490 nm was measured. The intraassay and interassay coeficients of variation were 7% and 1 IQ. respectively. Since, as shown in Fig 1. a high correlation existed between the two methods. results are presented in this report for the LIFA only.

by LIFA in All Groups

60

120

100

240

300

360

420

400

pMol/L Fig 1. Correlation between the two methods used to determine GHBP values, Ultrogel (expressed as % bound) and LIFA (pmoL/L). Both methods were highly correlated.

RESULTS GHBP levels in all groups studied are shown in Table 3. No patient with Turner’s syndrome had a GHBP level in serum less than 100 pmoL/L. The lowest value recorded in this group was 152.2 pmoL/L. Six of 15 children with ISS (40%) and seven of 19 with hypopituitarism (37%) had GHBP values less than 100 pmoL/L, in contrast to only 7.5% in controls of the same age. The mean GHBP value in the control group was 287.9 ) 10.9 pmol/L. which was significantly higher than the mean values in the ISS and hypopituitary groups (122.0 + 11 1’ 157.1 IT 23 pmoL/L, P < .Ol 13controls). Even when hypopituitarism is defined with more stringent criteria, ie, peak GH responses less than 6.0 ng/mL, and the subjects with responses between 6 and 10 ng/mL are moved to the ISS category, patients with hypopituitarism still have a significantly lower GHBP level than control subjects (Fig 2). Mean GHBP values in the redefined groups are as follows: ISS, 120.5 * 9.8 pmol/L (P < ,001); hypopituitary subjects, 180.9 +- 33.9 (P < ,003). With redefinition, nine of 22 patients (41%) with ISS have significantly decreased GHBP values (ie, 5 100 pmoL/L). When correlation analysis was performed between the GHBP level and several parameters of growth, the only consistent significant correlation found was with weight, as follows: Turner’s syndrome, r = .92. P < ,002; ISS, Y = .61, P < ,015; hypopituitarism, r = .55. P < ,015. When body mass index was correlated with GHBP values. all three groups had a significantly positive correlation with GHBP values (r = .94. .61, and .72 for Turner’s syndrome, ISS. and hypopituitarism, respectively: P < .Ol for all groups).

GHBP LEVELS IN CHILDREN WITH SHORT STATURE

359

Fig 2. GHEP values are compared in all groups. Hypopituitary patients were defined by both standard criteria (peak GH responses 510 ng/mL) and stringent criteria (GH < 6 ng/mL). lP c.01 v controls; ***redefined as peak GH ~6 ng/mL.

DISCUSSION

A complex cascade of events contributes to normal growth. Two hypothalamic peptides, the growth hormonereleasing factor and somatostatin, produce an ultradian and circadian rhythm of GH, which then binds to its binding proteins, (GHBP) and then to its receptors on the cells, affecting the generation of IGF-I. IGF-I in turn binds to its binding proteins (six now recognized so far) and stimulates its receptor in the target ~ell.~‘-‘” In this study we have found a relative, yet significant deficiency of the circulating levels of GHBP in children in whom no identifiable cause of short stature was found and who had been characterized as having normal GH secretory function. This confirms data in a different group of children reported by Carlsson et al.*j The understanding of the physiologic role of circulating GHBP .is presently evolving. Martha et al7 found a striking negative correlation between circulating GHBPs and 24hour mean GH concentrations in normal-stature healthy children, suggesting that perhaps the circulating level of

GHBP serves as a modulator of sorts for the tone of GH production by the somatotrope. If the latter is true, one could interpret the GH levels in ISS subjects as being inappropriately low for the level of circulating GHBP. However, the interaction of GH and GHBP modulating growth is undoubtedly more complex than plasma levels alone would indicate, since no correlation was found between circulating GHBP levels and height in this group of children. Our data in the hypopituitary subjects indicate that when the criterion used for GH deficiency is either the commonly applied or the stringent (children with very low GH responses to stimuli, ie, <6 ng/mL), the GHBP levels of these youngsters are still significantly different from those of the controls, suggesting that GH is at least one modulator of GHBP concentrations.j In all groups, a strong positive correlation was found between GHBP and weight and between body mass index and GHBP. This is similar to the positive correlation with body mass index found by others.” This suggests that adiposity may be yet another modulator of GHBP concentrations and hence GH secretory tone and effect. In conclusion, our results indicate that (1) children with Turner’s syndrome have normal GHBP levels; (2) children with hypopituitarism, either loosely or strictly defined, appear to have low GHBP levels; and (3) in ISS. there is in many subjects a quantitative deficiency of the GHBP. Whether this deficiency is reflecting a GH receptor deficiency, thereby contributing to these patients’ poor linear growth, and whether it can be overcome by exogenous pharmacologic doses of GH remains to be determined. ACKNOWLEDGMENT

The authors would like to thank Bill Tucker Diane Goin for typing the manuscript.

for the artwork

and

REFERENCES I. Leung DW, Spencer SA, Cachianes G: GH receptor and serum binding protein: Purification, cloning and expression. Nature 330 537-543, 1987 2. Spencer SA. Hammonds RG. Henzel WJ, et al: Rabbit liver GH receptor and serum binding protein. Purification. characterization and sequence. J Biol Chem 263:7862-7867.1988 3. Bauman G. Amburn KD, Buchanan TA: The effect of circulating GHBP on metabolic clearance, distribution and degradation of hGH. J Clin Endocrinol Metab 64:657-660, 1987 4. Postel-Vinay MC, Tar A, Hocquette JF. et al: Human plasma (GH)-binding proteins are regulated by GH and testosterone. J Clin Endocrinol Metab 73:197-202, 1991 5. Bauman G, Shaw MA, Winter RJ: Absence of the plasma GHBP in Laron-type dwarfism. J Clin Endocrinol Metab 65:814816.1987 6. Daughaday WH, Trivedi B: Absence of serum GHBP in patients with GH receptor deficiency (Laron dwarfism). Proc Natl Acad Sci USA 84:4636-4640. 1987 7. Martha PM. Rogol AD, Blizzard RM. et al: GHBP activity is inversely related to 24-h GH release in normal boys. J Clin Endocrinol Metab 73:175-181. 1991

8. Merimee TJ, Russell B, Quinn S. et al: Hormone and receptor studies: Relationship to linear growth in childhood and puberty. J Clin Endocrinol Metab 73:1031-1037, 1991 9. Carlsson LMS. Rowland AM. Clark RG. et al: Ligandmediated immunofunctional assay for quantitation of GHBP in human blood. J Clin Endocrinol Metab 73:1216-1223. 1991 10. Sokal R. Rohlf J: Introduction to Biostatistics. San Francisco, CA. Freemand, 1973 11. Kruhlich L. Dheriwal AP, McCann SM: Stimulatory and inhibitory effects of purified hypothalamic extracts on GH release from rat pituitary in vitro. Endocrinology 83:783-790. 1968 12. Brazeau P, Vale W. Burgus R: Hypothalamic polypeptide that inhibits immunoreactive pituitary GH. Science 179:77-79, 1973 13. Martin JB. Brazeau P, Tannenbaum GS: Neuroendocrine organization of GH regulation, in Reichlin S, Baldessarini RJ. Martin JB (eds): The Hypothalamus. New York, NY, Raven. 1978 14. Carlsson LMS. Attie KM, Compton PG: Reduced concentrations of serum GHBP in children with ISS. (submitted) 15. Martha PM Jr, Blizzard RM. Gesundheit N, et al: A longitudinal assessment of serum GHBP in normal boys during puberty. Pediatr Res 31:80A. 1992 (abstr 465)