Function of the Growth Hormone-Insulin-like Growth Factor I Axis in the Profoundly Growth-Retarded Diabetic Child: Evidence for Defective Target Organ Responsiveness in the Mauriac Syndrome Nelly Mauras, Thomas Merimee, and Alan D. Rogol Mauriac syndrome (MS) consists of a triad of poorly controlled diabetes, profound growth retardation, and hepatomegaly. The mechanisms involved in the growth retardation of those patients are not well understood. In an attempt to determine whether the growth retardation was secondary to somatroph secretory failure, abnormal pulsatile secretion, deletion of the growth hormone (GH) receptor, inadequate insulin-like growth factor I (IGF-I) generation, or abnormal IGF-I binding proteins (IGFBPs) two patients with MS were studied and their results compared with those of age-matched diabetic boys of similar glycemic control who were growing well. Overnight GH profiles in the MS and normally growing diabetics were analyzed by the CLUSTER program. The mean 12-hour GH concentrations, pulse amplitude, and pulse frequency were not different in either group of patients and did not change during acute normalization of the serum glucose overnight in the MS patients. The GH-binding proteins (GHBPs) relative binding were found to be the same in both groups of patients and did not differ from normal nondiabetic sera (62% 2 8.0% relative specific binding in MS patients, v 53% + 4.3% in diabetic controls). The IGF-I concentrations were normal and comparable in both groups of patients (1.1 ? 0.1 U/mL MS, v 1.1 f 0.3 diabetic controls). The IGFBPs were comparable in both groups of patients as well. One of the patients with MS had no meaningful increase in his growth velocity after 1 year on GH therapy despite good compliance. In conclusion, our data show normal hypothalamicpituitary function, normal GHBP, IGF-I generation, and IGFBPs in two patients with MS when compared with normally growing diabetic children. These data, and the lack of linear growth in response to exogenous GH therapy in one patient, suggest a GH-resistant state, either secondary to impaired bioactivity of IGF-I, or a defect at or distal to the IGF-I receptor. Copyright 0 1991 by W.B. Saunders Compky
M
AURIAC’S SYNDROME (MS) was described in the 1930s as a triad of poorly controlled diabetes mellitus, profound growth retardation, and hepatomegaly.’ Following the wide availability of insulin and intensification of diabetic control, this entity has become quite rare. It is well established that good metaboliciglycemic control in children with insulin-dependent diabetes mellitus (IDDM) plays an important role in their growth; however, data regarding the ultimate height of these children remain conflicting. Studies in twin pairs discordant for diabetes show that if diabetes develops before the onset of puberty, the diabetic twin is shorter than the sibling.’ Other investigators have found that the younger the child at diagnosis, the lesser the loss of height SD score.’ The same investigators did not fmd a correlation between HgA,C levels and either short- or long-term growth in these children. Poorly controlled diabetic children with normal height have been reported to have a significant acceleration of their growth with intensification of the insulin therapy.4 However, most data indicate that children with poorly controlled diabetes achieve an adult height within 2 SD for the general population. MS may represent one extreme of the spectrum of growth retardation found in poorly controlled diabetic children and the mechanisms responsible for this phenomenon are poorly understood. Circulating growth hormone (GH) con-
From the Nemours Children’s Clinic, Jacksonville, FL; the University of Virginia Medical Center, Charlottesville, VA; and the Universily of Florida College of Medicine, Gainesville, FL. Supported by Nemours Foundation Research Grant and National Institutes of Health Grant No. RR 00847 GCRC. Address reprint requests to Nelly Mamas, MD, Nemours Children S Clinic, PO Box 5720, Jacksonville, FL 32247. Copyright 0 1991 by W.B. Saunders Company 00260495/9114010-0020$03.00l0 1106
centrations (both basal and stimulated) are elevated in diabetic patient? and they tend to normalize with intensive insulin therapy and better metabolic control. Insulinlike growth factor I (IGF-I) concentrations (basal and stimulated) are normal or low despite elevated GH concentrations, suggesting a relative defect in GH action in poorly controlled diabetic patients.‘.” The effect of GH on cell growth is not only dependent on adequate GH secretion, but on the action of binding proteins and receptors. Recently, GH binding protein (GHBP) has been characterized” and its structure seems to be analogous to the extracellular domain of the GH receptor.‘-’ The full extent of its biologic action in different disease states is yet to be characterized. Studies have shown absent or reduced levels of GHBP in Laron-type dwarfs.“,” GH-dependent somatomedins or IGFs mediate many of the anabolic and mitogenic effects of GH.” IGF-I is bound to different molecular weight binding proteins, which themselves can affect the biological function of this important growth factor.“-“’ One of these binding proteins, IGFBP,. is directly related to GH status.” ” In this study, we evaluated the steps of the GH cascade in two patients with classical MS and profound growth retardation. Five diabetic boys of comparable age and glycemic control were also studied in a similar manner. GH secretion, GHBP levels (as an estimate of GH receptors), plasma IGF-I, and IGFBPs were examined in all subjects.
CASE REPORTS
Two patients with physical features of MS participated in this study. Their clinical characteristics are listed in Table 1. Both youngsters had markedly dilated abdomens and significant hepatomegaly with normal transaminases, negative viral hepatitis workup, and abdominal computed tomography (CT) scans and magnetic resonance images (MRIs). Metabolism,
Vol40,No10(0ctober),1991:
pp1106-1111
GH AND IGF-I AXIS IN MAURIAC SYNDROME
Table 1. Clinical Characteristics
1107
of MS Patients MS-1
MS-2
Chronological age (yr)
130/12
148112
Bone age (yr)
10 6/12
136/12
Height age (yr)
106/12
11 o/12
GrowIh velocity (cm/yr) Height SD score Mean HgA,C (preceding 6 mo)
2
2.6
-2
-3
cisco, CA) at a dose of 0.3 mgikgiwk divided daily. Growth assessed at the end of the l-year treatment period.
Assays Serum GH concentrations were measured in duplicate in the same assay run using the Nichols Institute human GH immunoradiometric
assay
kit (San
12%
15.9%
Duration of diabetes (yr)
8
10
coefficient
Tanner stage (genital)
I
III
with a sensitivity
of variation
Juan
Capistrano,
measured
The two boys with MS described above, as well as five age- and Tanner stage-matched normally growing insulin-dependent diabetic boys, were studied after informed written consent was obtained. All subjects were in good health and had no evidence of diabetic retinopathy, nephropathy, or neuropathy. All had normal blood counts and thyroid function. The clinical characteristics of the control boys are summarized in Table 2. The delay in skeletal maturation of the MS subjects was probably secondary to delayed entry into puberty. Studies were conducted at the Nemours inpatient research unit at Jacksonville Wolfson Children’s Hospital. The morning of the admission, subjects took their usual amount of insulin and ate their regular meals. After dinner, subjects were fasted until the following morning. In the evening of the study, an intravenous (IV) heparin lock was placed in a forearm vein for frequent blood sampling. Baseline determinations of glycosylated hemoglobin, serum testosterone, plasma somatomedin C (IGF-I). and serum C-peptide determinations were obtained. Blood was withdrawn at 20-minute intervals from 8:00 PM to 8:00 AM for GH determinations. Plasma glucose concentrations were measured hourly by Accucheck (Boehringer Mannheim. Indianapolis, IN) and Beckman Glucose Analyzer (Palo Alto, CA). At 8:00 AM, a hxed bolus dose of the GH-releasing factor (hpGRF-44), followed by 100 kg of the luteinizing hormone-releasing hormone (LHRH; Factrel, Ayerst Pharmaceuticals, New York, NY), were given IV as a bolus dose and blood samples withdrawn sequentially every 15 minutes for GH determinations for the next 90 minutes. After the study was completed, all subjects took their usual amount of insulin, ate breakfast, and were discharged home. The two subjects (no. 1 and 2) with MS returned to our inpatient unit 2 weeks after the first study for an identical study as described above. However, an open loop insulin delivery system was used as described beforeLJ to keep the plasma glucose concentration at nearly 5.6 mmol/L throughout the study. After the study was completed, MS patient no. 1was treated for 1 year with human GH (Protropin, Genentech, South San Fran-
of the 0.5 p.g/L. Plasma
by Nichols
GHBP
The
intraassay
IGF-I
(somatomedin
C)
by Nichols Institute using the et al.Li.‘h Serum testosterone was
Institute
(RIA) and glycosylated Lab, Akron, OH). modified
CA).
(CV) was less than 10% above 1.5 kg/mL,
concentrations were measured modified method of Furlanetto
Methods
rate was
by standard
hemoglobins
radioimmunoassay
by Glyc-Athn
was measured
by the method
by Daughaday
and Trivedi.”
method
(ISO-
of Herington
et al,‘7 as
Two assay tubes
are pre-
pared for each sample, the first containing 100 )LI. of serum with “I-GH (20 to 25,000 cpm); the second prepared in an identical manner,
but
mixtures
are brought
also
(Tris, 25 mmol/L; serum
albumin,
column. the
containing
CaCl,, 10 mmol/L; pH 7.5)
The ultragel
percent
2.5 up of unlabeled
to a total volume
GH.
These
of 250 uL with Tris buffer
0.02% NaN,; and 0.1% bovine
and then transferred
to an appropriate
column
separatory
method
for determining
bound
has
described
previously
of GH
been
only after correction for nonspecific binding. An adult reference standard and appropriate age-based pool of control sera were assayed with the serum from each patient.
The data are reported
relative specific binding, with 100% being the normal nondiabetic adult control sera.
binding
AM) were electrophoresed
through
a sodium dodecyl
sulfate (SDS)
polyacrylamide gel using a 5% to 15% linear gradient.“’ Separated proteins were electroblotted onto nitrocellulose filters (0.45 urn pore size) using a BioTrans Unit (Gelman Sciences. Ann Arbor. MI). Filters were blocked, labeled with ?-IGF-I or overnight at 4°C and visualized Hossenlopp
by autoradiography,
according
to the method
Data Analysis The GH concentration analysis?
profiles were analyzed
The pulse parameters
obtained
using CLUSTER
in the MS subjects
compared with the five diabetic controls using standard metric testing (Wilcoxon signed rank tests).
11
Mean HgA,C (6 mo) Duration of diabetes(y) Tanner stage
-1 9.7% 1.5 I
14
15
12.1 k 1.0
15
16.5
12.7 k 1.3
5.7
6.0 2 0.6
9.5
11
10
SD score
Mean k SE
12.1 + 1.1
13.6
11 4.2
5
15.5
9.7
Bone age (yr) Height age (vr) Growth velocity (cm/y)
4
3
10.9
5.9
11 6.3
0
+1
10%
10%
1
I
0.75
I
7.8 +1
+1
13.6% 4 IV
10.4% 1.25 IV
were
nonpara-
Subject No.
10.8
of
et al?”
2
Chronological age (yr)
as for
Serum IGF-binding proteins (IGFBPs) were measured by Western ligand blot analysis. Unreduced plasma samples (1 uL at 8:00
Table 2. Clinical Characteristics of Diabetic Controls
1
by
Daughaday et a12*;for the current report. bound (B) and free (F) GH after separation on a 0.9 x 17-cm ultragel column are shown
10.7% + 0.7% 1.7 + 0.6
1108
MAURAS, MERIMEE, AND ROGOL
Table 3. GH Pulse Analysis by Cluster
Mean GH
Mean Pulse
Mean Pulse
FK!qUWlCy
Amplitude
(pulses/12
kg/L)
h)
InterpUlSe
Area Under
Interval
(&g/L)
(min)
GH Curve
(pg/L min)
MS patients MS-l,
26.0
6
37.1
108
1,345
MS-l,
28.0
5
43.5
140
2,480
MS-2,
9.0
7
12.2
83
399
MS-2,
13.0
4
23.9
113
1,045
19.0 f 4.7
5.5 k 0.6
29.2 k 7.0
111 + 11
1,317 2 434
Mean r SE Controls 1
9.5
3
2.5
150
1,745
2
4.7
5
11.7
110
432
3
8.8
5
17.0
115
813
4
26.6
4
64.9
153
3,349
18.6
6
31.3
80
870
13.6 2 4.0 (NS)
4.6 lr 0.5 (NS)
30.0 2 9.3 (NS)
122 + 13 (NS)
1,446 t- 521 (NSJ
5 Mean + SE
Abbreviations: A, patient at baseline; 6, patient during glucose clamp. NS, not significant as compared with MS patients.
RESULTS
GH Pulses
There were no significant differences between the GH concentration profiles in the patients with MS before and after acute normalization of the plasma glucose concentrations. The mean glucose concentration of both MS patients during the baseline study was 11.0 2 1.4 mmol/L and 5.5 ? 0.4 mmol/L during the closed-loop insulin delivery. The mean overnight glucose concentration was 8.7 mmol/L on the controls. Table 3 summarizes the data of the 12-hour GH concentration profiles as analyzed by CLUSTER.j’ The mean GH concentrations of the MS patients, although higher than the controls, were not significantly different. The mean GH pulse frequency, pulse amplitude, and areas under the curve were all comparable in the MS patients, as in the normally growing diabetic controls. The differences in the individual control subjects (no. 1,2, and 3) with lower mean 1Zhour GH concentrations as compared with MS subject no. 1 reflect differences in pubertal status (see Tables 1 and 2). Figure 1A shows the GH pulse profile of MS-l,, and Fig
$
80 70 80
y
50
-
40
= (3
30
lB, the profile on a normally growing diabetic boy (no. 5). Both patients were Tanner stage III to IV of sexual development. The GH responses to GRF were comparable in both groups of patients (data not shown). The plasma IGF-I concentrations were the same on both groups (Table 4). GHBP Data
The percentage of binding of GHBP is shown for a patient with MS (no. 2) in Fig 2. Binding of ‘“I-GH for the pooled sera of the control diabetic boys, as well as adult nondiabetic pool, are shown for comparison. Table 5 gives the GHBP results in all subjects, with each subject’s serum being assayed with a known adult nondiabetic control. No significant differences from normal were noted in either the patients with MS or normally growing diabetics. IGF BPS
IGFBPs in plasma from MS and normally growing diabetics were assessed by Western ligand blotting. Figure 3 shows the migration of IGFBPs from plasma after SDS60
20 10 0
I
,
1
I
I
,‘I1
,
I
012345678
TIME
I 012345678
(min)
x10*
TIME
(min)
x10’
Fig 1. Serum GH concentrations over 12 hours for MS patient no. 1 (A) and control subject no. 5 (6). The deflections at the top represent the pulses detected by the CLUSTER program; error bars are derived from duplicate measurements for each data point. The mean GH concentration, area under the curve, pulse amplitude. and frequency were not different (see Table 3). Note the differences in the ordinates.
GH AND IGF-I AXIS IN MAURIAC
SYNDROME
Table 4. IGF-I Concentrations
1109
(U/mL)
Table 5. GHBP Results
RelatwSpecific
COIItds
MS
Subjects
1:l.Z
MS-1,:l.O
MS patients
2:0.5
MS-2,:0.8
3:0.44
MS-1
10.6%
54%
MS-2,: 1.O
4:2.0
MS-2
13.7%
70%
Mean
12.2%
62% 66%
1.1 + 0.1
Controls
1.1 * 0.3
polyacrylamide gel electrophoresis (nonreducing conditions), transfer onto nitrocellulose, and incubation with [‘*‘I]IGF-I. In standard human plasma (lane A), multiple specific bands were observed, two major bands at 42,000 and 38,000 M,, and additional bands at 34,000, 28,000, and 24,000 M,, similar to previously reported Western ligand blotting patterns of human plasma.“~‘2 The 42,000 and 38,000 M, bands have been shown to represent different glycosylation states of IGFBP,, the binding subunit of the GH-dependent 150,000 M, major plasma complex.“~” This IGFBP? doublet is elevated in plasma from acromegalic patients (lane K). We could detect no consistent differences in IGFBP band pattern and/or intensity among normal, MS (lanes B, C, E, F), and normally growing diabetics (lanes D, G-J) using this technique. Growth Data
One of the patients with MS, MS-l, was treated with recombinant DNA GH (Protropin) for a full year at a dose of 0.3 mg/kg/wk divided daily. The pretreatment growth velocity was 1.7 cmiyr. After 1 year of GH therapy, he grew 2.6 cm. The somatomedin dose while receiving GH therapy was 1.0 U/mL. DISCUSSION
The mechanisms responsible for poor growth in patients with MS are poorly understood. We have analyzed the multiple sequence of steps involved in the complex pathway that controls growth, starting with a positive signal in the hypothalamic-pituitary axis and resulting in pulsatile GH release, leading to the binding of GH to circulating proteins, and to the generation of IGF-I and its associated
jC X
4
bound
12
=
,
13%
26
26
FRACTION
Binding
MS-1 B:1.4
5:1.3 Mean 5 SE
Specific Binding
36
% bound =
12
20
t4.W.
26
,
36
V. bound = 21%
16
24
COLLECTED
Fig 2. Elution of ‘“I-GH after incubation with serum and buffer (see Methods) is shown for MS patient no. 2 (A), a pool of the normally growing diabetic control sera (B), and a nondiabetic adult pool (C). F, free; 6, bound after correction for nonspecificity. There were no significant differences among the groups.
1
13.0%
2
10.9%
55%
3
9.7%
49%
4
10.9%
55%
5
7.8%
40%
10.5% + 0.9%
53% 2 4.3%
Mean ‘_ SE
binding proteins, in two youngsters with MS, and compared their results to those of normally growing diabetic controls. We found similar ultradian rhythms of GH secretion in the diabetics who were growing poorly as compared with those growing well. The mean GH concentrations, pulse frequencies, amplitudes, and areas under the curve of the computer-identified pulses, were not different in the two groups of patients and did not change in the MS patients after acute normalization of the blood glucose concentrations. The relatively high GH concentrations, as previously noted by other investigators,S.h confirm normal pituitary function in these patients, irrespective of their growth performance. The elevated GH concentrations are not fully understood, but such elevation could be as much as cause as consequence of their poor metabolic control.” The GHBP has been found to be homologous to the extracellular domain of the GH receptor”.‘4 and quantitative and/or qualitative deficiencies may be responsible for poor growth in Laron dwarfs.14 Despite the variability in the patients’ levels, they were all essentially normal as compared with control data in children.” We found normal circulating plasma IGF-I concentrations in both groups of subjects. However, the IGF-I assay used in this study only measures immunoactivity. Using Phillip’s assay for IGF-1 bioactivity, investigators have found reduced IGF-1 activity at the cartilage level in the serum of patients with poorly controlled diabetes, which improves with better metabolic control.x Hence, RIAs for IGF-I, although normal, may be misleading with respect to the biological activity in patients with MS. It is possible that there is a defect in end-organ cartilagenous response to normal levels of IGF-I in these patients. We also found normal levels of circulating IGFBPs. which regulate IGF-I bioavailability and bioactivity.“.“’ The GH-dependent 42,000/38,000 M, IGFBP_, , the binding subunit of the major plasma IGFBP, shows little diurnal variation” and changes can be detected by Western ligand blotting. However, other IGFBPs, which comprise less than 10% of plasma IGFBPs and whose levels change rapidly with fluctuations in metabolic and nutritional status (eg. IGFBP-I),‘6 may not be adequately evaluated by single time point determinations. Indeed, one of the diabetic inhibitors in sera of diabetic rats is IGFBP-I.” However, in that study,
MAURAS,
1110
MERIMEE, AND ROGOL
MR x 1O-3
69
46
ABC
DEFGHI
the investigators were able to detect a major increase in circulating IGFBP-I in diabetic rats by Western ligand blotting. One of the two patients with MS was treated with GH for a full year without any meaningful increase in this linear growth on exogenous therapy, despite normal IGF-I levels. This is further evidence for a distal resistance to the effect of GH at the site of growing chondrocyte and may be due to defective IGF-I receptors. An alternative, yet untested, explanation could be the presence in the MS patients of increased levels of IGF-I inhibitors. The latter are found to be present in the sera of diabetic rats38,3Yand appear to be a general inhibitor of growing cartilage.4” In conclusion, our data on two patients with MS demonstrate normal hypothalamic-pituitary function, and normal GHBPs, IGF-I generation, and the IGF-I binding proteins when compared with normally growing diabetic children.
J
K
Fig 3. Western ligand blot analysis of IGFBPs. Lanes 6, C, E. and F represent plasma from patients no. 1 and 2 with MS before and after a glucose clamp, respectively. Lanes D, G, H, I, and J represent plasma from five diabetic controls. Lane A represents standard human plasma, lane K acromegalic plasma.
The lack of response to exogenous GH therapy suggests a relative GH-resistant state either secondary to defective IGF-I receptors or impaired IGF-I bioactivity, or due to the presence of circulating IGF-I inhibitors. Whether these abnormalities are directly related to glycemic control and are potentially reversible, or whether they represent an intrinsic abnormality in this syndrome, remains to be determined. ACKNOWLEDGMENT We are grateful to Sharon Treakle and the expert nursing staff of the Wolfson Children’s Hospital for their dedicated care of our patients; to Dr Cheryl A. Conover from the Mayo Clinic for Western ligand blot analysis of the IGFBPs; to Margaret Bowen for coordinating this study; to Ginger Bauler and Catherine Kern for the measurement of the RIA; to Bill Tucker for the art work, Emilie Olsen for her assistance in data analysis, and to Karen Richard for the excellent typing of the manuscript.
REFERENCES 1. Mauriac P: Gros Ventre, hepatomegalie, troubles de la croissance chez les enfants diabetiques, traites depuis plusieurs annees par l’insuline. Gax Hebd Sci Med Bordeaux 51:402.1930 2. Tattersall RB, Pyke DA: Growth in diabetic children: Studies in identical twins. Lancet 2:1105-1109, 1973 3. Herber SM, Dunsmore IR: Does control affect growth in diabetes mellitus? Acta Pediatr Stand 77:303-305,1988 4. Rudolf MCJ, Sherwin RS, Marcowitz R, et al: Effect of intensive insulin treatment on linear growth in the young diabetic patients. J Pediatr 101:333-339, 1982 5. Hayford JT, Danney MM, Hendrix JA, et al: Integrated concentration of growth hormone in juvenile-onset diabetes. Diabetes 29:391-398,198O 6. Asplin CM, Faria ACS, Carlsen EC, et al: Alterations in the pulsatile mode of GH release in men and women with IDDM. J Clin Endocrinol Metab 69:239-245, 1989 7. Lee RGL, Bode HH: Stunted growth and hepatomegaly in diabetes mellitus. J Pediatr 91:82-84. 1977 8. Winter RJ, Phillips LS, Klein MN, et al: Somatomedin
activity and diabetic control in children with insulin-dependent diabetes. Diabetes 28:952-954, 1975 9. Winter RJ, Phillips LS. Green OC. et al: Somatomedin activity in the Mauriac syndrome. J Pediatr 97598-600, 1980 10. Amiel SA, Sherwin RS, Hintz RL, et al: Effect of diabetes and its control on insulin-like growth factors in the young subject with type I diabetes. Diabetes 33:1175-l 179, 1984 11. Lanes R, Reeker B, Fort P, et al: Impaired somatomedin generation test in children with IDDM. Diabetes 34:156-160, 1985 12. Baumann G, Stolar MW, Amburn K, et al: A specific GH-binding protein in human plasma: Initial characterization. J Clin Endocrinol Metab 62:134-141, 1986 13. Baumann G, Shaw MA: Immunochemical similarity of the human plasma GH-binding protein and the rabbit liver GH receptor. Biochem Biophys Res Commun 152:573-578.1988 14. Baumann G, Shaw MA, Winter RJ: Absence of the plasma GH-binding protein in Laron-type dwarfism. J Clin Endocrinol Metab 65:814-816, 1987 15. Daughaday WH, Trivedi B: Absence of serum GH binding
GH AND IGF-I AXIS IN MAURIAC
SYNDROME
protein in patients with GH receptor deficiency. Proc Natl Acad Sci USA 84:4636-4640,1987 16. Hintz RL: The somatomedins, in Barness LA (ed): Advances in Pediatrics, vol 28. Chicago, IL, Yearbook Medical, 1981. pp 293-3 17 17. Hintz RL: Plasma forms of somatomedin and the binding protein phenomenon. Clin Endocrinol Metab 13:31-42, 1984 18. Nessley SP, Rechler MM: Insulin-like growth factors: Biosynthesis, receptors, and carrier proteins, in Li CH (ed): Hormonal Proteins and Peptides. vol 12. San Diego, CA, Academic, 1984, pp 127-203 19. DeMellow JSM, Baxter RC: Growth hormone-dependent insulin-like growth factor (IGF) binding proteins with inhibits and potentiates IGF-I stimulated DNA synthesis in human skin fibroblasts. Biochem Biophys Res Commun 156:199-204,1988 20. Liu F. Powell DR, Hintz RL: Characterization of insulin-like growth factor binding proteins in human serum from patients with chronic renal failure. J Clin Endocrinol Metab 70:620-628, 1990 21. Hardouin S. Hossenlopp P, Segovia B, et al: Heterogeneity of insulin-like growth factor binding proteins and relationships between structure and affinity. I. Circulating forms in man. Eur J Biochem 170:121-132,1987 22. Martin JL, Baxter RC: Insulin-like growth factor-binding protein from human plasma: purification and characterization. J Biol Chem 261:8754-8760, 1986 23. Baxter RC, Martin JL: Radioimmunoassay of growth hormone-dependent insulin like growth factor binding protein in human plasma. J Clin Invest 78:1504-1512,1986 24. White NH, Skor D, Santiago JV: Practical closed-loop insulin delivery. Ann Intern Med 97:210-213, 1982 25. Furlanetto RW, Underwood LE, VanWyk JJ, et al: Estimation of somatomedin-C levels in normals and patients with pituitary disease by RIA. J Clin Invest 60:648-657, 1977 26. Furlanetto RW: Pitfalls in the somatomedin-C RIA. J Clin Endocrinol54:1084-1086, 1982 27. Herington AC, Ymer S, Stevenson J: Identification and characterization of specific binding proteins for GH in normal human sera. J Clin Invest 77:1817-1823,1986 28. Daughaday WH, Trivedi B, Andrews BA: Ontogeny of serum GH binding protein in man: A possible indicator of hepatic
1111
GH receptor development. J Clin Endocrinol Metab 65:1072-1074, 1987 29. Laemmli UK Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 227:680-685, 1970 30. Hossenlopp P. Seurin D, Segovia Quinson B, et al: Analysis of serum insulin-like growth factor binding proteins using Western blotting. Use of the method for titration of the binding proteins and competitive binding studies. Anal Biochem 154:138-143, 1986 31. Veldhuis JD, Johnson ML: A novel general biophysical model for stimulating episodic endocrine gland signaling. Am J Physiol250:E486-E493, 1986 32. Conover CA, Liu F, Powell D, et al: Insulin-like growth factor binding proteins from cultured human fibroblasts: Characterization and hormonal regulation. J Clin Invest 83:852-859, 1989 33. Press M, Tamborlane WV, Sherwin RS: Importance of raised GH levels in mediating the metabolic derangements of diabetes. N Engl J Med 310:810-815,1984 34. Leung DW, Spencer SA, Cachianes G, et al: Growth hormone receptor and serum binding protein: Purification, cloning and expression. Nature 330:537-543, 1987 35. Merrimee TJ, Baumann G. Daughaday W: Growth hormonebinding protein: II. Studies in pygmies and normal statured subjects. J Clin Endocrnol Metab 71:1183-1188,199O 36. Cotterill AM, Cowell CT, Baxter RC, et al: Regulation of the growth hormone-independent growth factor-binding protein in children. J Clin Endocrinol Metab 67:882-887, 1988 37. Unterman TG, Pate1 K, Kumar Mahathre V, et al: Regulation of low molecular weight insulin-like growth factor binding protein in experimental diabetes mellitus. Endocrinology 126:26142624.1990 38. Phillips LS, Belosky DC, Reichard LA: Nutrition and somatomedin. V. Action and measurement of somatomedin inhibitor(s) in serum from diabetic rats. Endocrinology 104:1513-1518, 1979 39. Phillips LS, Bajaj VR, Fusco AC, et al: Nutrition and somatomedin: Studies of somatomedin inhibitors in rats with streptozotocin-induced diabetes. Diabetes 32:1117-l 125,1983 40. Phillips LS, Vassilopoulou-Sellin R, Reichard LA: Nutrition and somatomedin: The somatomedin inhibitor in diabetic rat serum is a general inhibitor of growing cartilage. Diabetes 28:919924, 1979