Department of Pediatrics (A.R.), University of Florida College of Medicine, Childrens Medical Services Center, Gainesville Florida 32608; Instituto Endocrinologia, Metabolismo y Reproduccion (J.G-A.), Quito, Ecuador; Department of Pediatrics (R.G.R.), Oregon Health Sciences University, Portland Oregon 97201-3098; and Department of Genetics and Howard Hughes Medical Institute (U.F.), Stanford University Medical Center, Stanford, California 94305-5323
Address correspondence to: Arlan Rosenbloom, Department of Pediatrics, Childrens Medical Services Center, 1701 SW 16th Avenue, Room 2163, Gainesville, Florida 32608.
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Background and History |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The recognition of this population could not have come at a more propitious time. In 1987, GH binding protein (GHBP) was identified as being identical to the extracellular domain of the GH receptor (GHR) and to be absent in several patients with Laron syndrome. During the year that we were identifying the first 45 patients, 1989, the human GHR gene was characterized and a partial gene deletion found in two Israeli patients with Laron syndrome, and a single nucleotide substitution resulting in a missense mutation was described in four patients in a Mediterranean family (reviewed in Ref. 2). The unique concentration in time and place of the Ecuadorian clinical experience and the finding that a single mutation of the GHR accounted for GHRD in all but one of the patients permitted description of new clinical and biochemical features, the quantification of other features, appreciation of phenotypic variability with the same GHR mutation, and provided opportunity for treatment study. This review is of the findings in the Ecuadorian population. Other reviews are available covering the basic physiology, genetics, and global experience with this condition (1, 2, 3, 4, 5).
![]() |
Epidemiology |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Mortality with GHRD has only been studied in the Ecuadorian cohort. Childhood mortality before the age of 7 yr was 19% for affected children (15 of 79) compared with 9.7% (21 of 216) for unaffected siblings (P < 0.05). Causes of death (meningitis, pneumonia, diarrhea) did not differ between affected and unaffected children, suggesting greater vulnerability of the children with GHRD.
Among adults over 50 yr of age followed for 7 yr, two died of heart disease (an uncommon problem in the Andes): a man at 55 yr and a woman at 67 years, (5). Compared with sibling and community controls, adults with GHRD had significantly elevated total cholesterol and low-density lipoprotein (LDL) cholesterol, which was unrelated to adiposity [determined by dual-energy x-ray absorptiometry (DEXA)] or insulin resistance [inferred by fasting insulin-like growth factor binding protein (IGFBP)-1 concentration]. These findings were thought to reflect decreased activity of the hepatic LDL receptor, which is influenced by GH action (8).
![]() |
An Unusual Mutation in the GHR Gene Causes GHRD in Ecuador |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The GHR gene spans 86 kilobase pairs and includes nine exons, numbered 210, that encode the receptor and four additional exons in the 5' untranslated region. Exon 2 encodes a secretion signal sequence, exons 3 through 7 the large extracellular GH binding domain, exon 8 the transmembrane domain, and exons 9 and 10 the cytoplasmic domain and 3' untranslated region (3). To look for mutations in the GHR gene, Mary Anne Berg, in the laboratory of UF, first excluded major structural rearrangements by Southern blot analysis with a human GHR complementary DNA probe. Because automated sequencing was not readily available to look for more subtle mutations, Berg applied a new mutation screening technique, denaturing gradient gel electrophoresis; this most sensitive of mutation detection methods was perfected in the laboratory of Richard Myers at University of CaliforniaSan Francisco (San Francisco, CA), who provided generous advice. Based on the published sequences of the GHR gene exons and flanking introns, primers were designed to amplify each coding exon. PCR products from genomic DNA of obligate heterozygotes contain a mixture of normal and mutant amplimers. Heteroduplex DNA molecules show altered melting behavior under denaturing gradient gel electrophoresis conditions. By this analysis, only GHR exon 6 amplimers produced evidence of a sequence variant (10). On sequencing of cloned PCR products containing exon 6, a single nucleotide substitution was identified that changed the glutamic acid codon GAA in position 180 of the amino acid sequence to a GAG codon, also encoding glutamic acid (10, 11).
Single nucleotide polymorphisms leading to such synonymous changes are rather common in the human genome and are considered normal variants. Therefore, the mechanism by which this mutation could be pathogenic was not immediately apparent. Inspection of the surrounding nucleotide sequence, however, suggested that this A to G substitution could have activated a cryptic splice site. The normal exon 6/intron 6 splice junction is ATG/GTAAGT. The mutation creates a sequence GAG/GTAAAT within the exon 6 coding sequence. Therefore, if splicing occurred after the substituted G nucleotide, the resultant shortened messenger RNA would lack the 24 nucleotides encoding the most C-terminal eight amino acids of exon 6. Expression of the GHR gene in immortalized lymphoblastoid cell lines allowed us to test this hypothesis directly. RT-PCR analysis of cell lines from affected individuals and heterozygotes led to the surprising result that the new splice site created by the mutation was exclusively used, although the original site remained intact (10, 11). It is now known that similar mutations in other genes often lead to transcript heterogeneity that may modulate the clinical phenotype (12). No evidence for transcript heterogeneity was found at the molecular level for the GHR E180 splice mutation, a finding that is consistent with the rather uniform clinical picture.
The eight deleted amino acids are part of the extracellular domain and are located near the receptor dimerization site. They also include a putative glycosylation site. We consider it most likely that the mutant protein is misfolded and degraded intracellularly (13). This hypothesis is consistent with the lack of serum GHBP in most of the Ecuadorean GHRD patients (2). In contrast, a putative stable mutant protein that may lack the ability to transmit signal would be expected to exert a dominant negative effect. No effect on growth, however, was observed in individuals heterozygous for this mutation (14).
Once the E180 splice mutation was identified, rapid assays for its
detection were developed by use of allele-specific oligonucleotide
hybridization (10, 11) and by an allele-specific restriction enzyme
analysis assay with MnlI, for which the mutation generates a
new cleavage site (15). To test all available GHRD patients in Ecuador
and their relatives for the presence of this mutation, we used dried
blood spots collected onto filter papers. The DNA that was
PCR-amplified from these blood spots appeared to be stable at room
temperature for several months. The results, to date, indicate that all
but one of the individuals with GHRD from Ecuador were homozygous for
the E180 splice mutation. The single exception, an offspring of a
consanguineous marriage, was homozygous for a recurrent R43X nonsense
mutation in exon 4. This mutation involves a CpG hotspot and has
occurred independently in different populations as indicated by
different haplotypes (3, 16). Genotyping potential heterozygotes by PCR
from blood spots enabled the systematic comparison of carrier and
noncarrier siblings for a variety of clinical parameters (8, 14, 17).
The carrier detection results were also used for genetic counseling
(Fig. 2) (1).
|
The gender bias in the initially identified population from Loja was not due to the GHRD, as the sex ratio was normal in patients from El Oro province, and they carry the same mutation. An independent X-linked lethal allele segregating in the inbred Loja population could reduce the number of viable male births. This possibility has not been systematically pursued. The molecular genetic GHR studies of this population revealed an instructive example of a founder effect, where a rare recessive mutation is amplified in a geographically isolated population and leads to an unusually high disease frequency.
![]() |
Clinical Phenotype and Spectrum with Homogeneity for the GHR Mutation |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Despite a characteristic clinical appearance at birth indicating
an effect of GHRD in utero (patient 1 in Fig. 3), affected newborns are of normal
weight and most are normal in length, similar to what is seen with
severe GH deficiency due to GH gene deletion (20). There is rapid
postnatal decline in SD
score), and growth velocity is
approximately half normal. In both sexes there is absence of the
pubertal growth spurt, with persistent growth beyond the normal time of
adolescence and substantial delay in the onset of puberty in 50% (20).
Adult stature varies from -5.3 to -12 SDS, using United States
standards, a range of 95124 cm for women and 106141 cm for men.
This wide variation, despite homogeneity for the GHR mutation, was seen
within families, as well as within the entire population, and is
comparable to the range of SDS for height in a genetically diverse
group of patients (18).
|
As noted above, heterozygosity for the E180 splice mutation did not affect stature among 53 relatives of probands compared with 37 homozygous normal relatives. Furthermore, if heterozygosity were to influence stature, the heights of heterozygous parents of probands would be expected to correlate with those of probands and of carriers who are their offspring and not with heights of their homozygous normal children. In fact, parental height SDS did not correlate with height SDS of affected offspring, but correlated strongly with those of both heterozygous and homozygous normal offspring (14).
Craniofacial characteristics
As with statural effects, there is wide variability of
craniofacial features among affected individuals, despite identity of
the GHR mutation (Fig. 3). Affected children are often recognized by
knowledgeable family members at birth because of frontal prominence,
depressed nasal bridge, sparse hair, small hands or feet, and
hypoplastic fingernails (patient 1 in Fig. 3
). Because of the prominent
forehead, small face, prominence of the scalp veins with thin skin and
hypotrichosis, and setting sun sign, patients may be subjected to
evaluation for hydrocephalus (patient 2 in Fig. 3
). Decreased vertical
dimension of the face is demonstrable by computer analysis of the
relationships between facial landmarks and is present in all patients,
compared with their relatives, including those with reasonably normal
facial appearance (patient 10 in Fig. 3
) (21). The occasional
occurrence of unilateral ptosis with facial asymmetry may result from
positional deformity due to decrease in muscular activity in
utero (5).
Musculoskeletal and body composition
Delayed walking, despite normal time of speech onset, is considered to be the result of hypomuscularity, which is apparent on roentgenograms. Osteopenia is also demonstrated on radiographs and by DEXA (5, 6). In a detailed study of 11 adult probands matched to 11 unaffected siblings, areal bone mineral density was reduced in the patients with GHRD, but estimated volumetric bone density (bone mineral apparent density) was normal. Iliac crest dynamic histomorphometric studies from bone biopsies confirm the interpretation of preserved bone mineral apparent density, with the only difference being a reduced trabecular connectivity in GHRD (22). It remains uncertain whether or not these patients have true osteoporosis; fractures do not seem to be more frequent, including in the elderly.
Limited elbow extensibility seen in 85% of Ecuadorian patients with GHRD over 5 yr of age and with increasing severity with age has been described in at least one other individual with GHRD and with familial anterior hypopituitarism due to PROP1 deficiency (4). The reason for this acquired problem with severe IGF-I deficiency is unknown.
Children are underweight to normal weight for height despite the
appearance of obesity (Fig. 1), whereas all females and most males
after puberty are substantially overweight with marked decrease in lean
to fat ratios by DEXA (8, 20).
Intellectual function
We noted exceptional school performance with seriously constrained social and economic independence of most women, but only a few of the men in the Ecuadorian patient population (23). The only controlled evaluation of intellectual ability of patients with GHRD was carried out in Ecuador with intelligence tests that had been validated in cross-cultural research, designed to minimize the effects of physical size, motor coordination, and cultural background. The intellectual ability of the patients with GHRD did not differ from that of their relatives or of community controls, and there was no effect of heterozygosity for the E180 splice mutation (17). The studies indicate that neither fetal nor postnatal brain growth or intellectual development is dependent on GH stimulated IGF-I production.
![]() |
Biochemical Features of IGF-I Deficiency Resulting from GHR Failure
(Table 2![]() |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ecuadorian children with GHRD have elevated random GH levels that may be as high as 200 ug/L, and their response to stimulation is increased, with paradoxical elevations after oral or iv glucose, as in acromegaly (2, 6). Diurnal variation is normal with normal suppression by exogenous recombinant human IGF-I (24). Thus, the normal sensitivity of the GH secretion is preserved, despite elevated levels and lack of feedback suppression from IGF-I.
|
GHBP
The ligand-mediated immunofunction assay used to measure GHBP serum levels since 1990, uses an anti-GH monoclonal antibody to determine the amount of GH bound to GHBP. This largely functional assay should not detect structurally abnormal, but expressed, GHBP (2).
Despite in vitro evidence for failure of production of normally spliced receptor, four children and four adults of 49 Ecuadorian GHRD patients initially studied had serum GHBP levels higher than 40% of the sex-specific lower limit for controls, and one adult male had a level in the lower portion of the normal adult male range. The presence or amount of GHBP measured did not relate to stature. There were no age-dependent changes, indicating that the difference in IGF values between children and adults was not related to the GHBP levels (7).
IGF
Serum IGF-I concentrations are profoundly reduced, with prepubertal patients having levels consistently less than 10 ng/mL and adults less than 100 ng/mL (6, 7). Thus, prepubertal IGF-I levels in GHRD are totally dependent on GH secretion and action, with the elevations in sexually mature individuals indicating that sex steroids may be able to stimulate IGF-I production independently of GH. Patients with GHRH receptor deficiency, however, have identical extremely low IGF-I levels as children and as sexually mature adults, which argues against this hypothesis (25).
Serum IGF-II concentrations are also significantly reduced, but not as dramatically as those of IGF-I. The reduction in IGF-II levels likely reflects the marked decrease in serum levels of IGFBP-3 and ALS (acid labile subunit).
IGFBP
The principal binding protein for circulating IGF, IGFBP-3, is considered to serve as a reservoir and delivery mechanism for IGF to tissues. In both prepubertal and adult patients with GHRD, serum IGFBP-3 concentrations were found to be less than 1 ug/mL, approximately 20% of normal adult concentrations (7). ALS, which combines with IGFBP-3 and IGF to form the circulating ternary complex and is also GH dependent, is profoundly reduced, averaging less than 1 ug/mL, compared to normal levels of 1620 ug/mL (26). Modest reductions were observed in serum concentrations of IGFBP-4 and IGFBP-5; because IGFBP-5 is able to complex with ALS, this reduction may reflect ALS deficiency rather than specific GH dependency for this IGFBP (26). IGFBP-I is elevated in GHD and GHRD; in GHRD it is the most abundant IGFBP and is strongly inversely related to insulinemia. IGFBP-2 is present at a mean 300 percent of control concentrations in children with GHRD and 175 percent of control in affected adults (7).
![]() |
Administration of IGF-I to Ecuadorian Subjects with GHRD |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Before undertaking treatment of the large number of children in the Ecuadorian population with GHRD, Vaccarello et al. (24) studied six Ecuadorian adults with GHRD on the Clinical Research Center at the University of Florida. They administered IGF-I in a sc dose of 40 ug/kg body weight every 12 h over 7 days. Mean integrated 24-h GH levels were suppressed, as were the number of peaks, the area under the curve, and clonidine-stimulated GH release. The mean peak serum IGF-I levels were 253 ± 11 ng/mL, reached between 26 h after injection, and mean trough levels were 137 ± 8 ng/mL before the next injection, values not significantly different from those of normal control Ecuadorian adults. There were no significant changes in the half-life or metabolic clearance of IGF-I between days 1 and 7, but the distribution volume did increase over this time. Although IGFBP-3 levels did not increase, elevated baseline IGFBP-2 levels (153% of control) increased 45% (P < 0.01). In these short-term studies, there was no increased risk of hypoglycemia despite low levels of IGFBP-3. There remained, however, concern whether the low IGFBP-3 levels would result in more rapid clearance of IGF-I and blunting of the growth response.
Additional studies were carried out on composition and distribution of IGF and IGFBP in patients from this study. It was found that the two forms of IGFBP-3 associated with IGF and ALS, which are able to form the ternary 150-kDa complex are abnormally distributed in GHRD patients. This distribution was unchanged by IGF-I treatment, which again demonstrated failure to increase IGFBP-3, as well as ALS, levels (26, 27).
Long-term treatment
In addition to concern about replacement therapy with IGF-I related to the failure of short-term therapy to increase IGFBP-3 levels, there was a question whether catch-up growth in children with GHRD would be as substantial as occurs with GH replacement therapy in patients with GHD, in the absence of a direct effect of GH on bone. This direct effect is considered to be the differentiation of prechondrocytes into early chondrocytes, which then secrete IGF-I, that in turn stimulates clonal expansion and maturation of the chondrocytes, or growth, thought to account for 20% of GH-induced growth (4).
The first IGF-I treatment report from the large Ecuadorian cohort was
of growth and body composition changes in two adolescent patients
treated with a combination of IGF-I (120 mcg/kg bid) and LRH analog to
forestall puberty. A girl age 18 yr and boy age 17.2 yr, with bone ages
of 13.5 and 13 yr, experienced an approximate tripling of growth
velocity, increased bone mineral density, and maturation of facial
features on IGF-I for 1 year. There was initial hair loss, followed by
recovery of denser and curly hair with filling of the fronto-temporal
baldness, the appearance of axillary sweating, loss of deciduous teeth,
and appearance of permanent dentition. The boy had coarsening of his
facial features indicative of supraphysiologic tissue levels of IGF-I
(Fig. 5). The fading of premature facial
wrinkles was noted in one patient. The other patient had submaxillary
gland swelling. Serum IGF-I levels were seen to increase into the
normal range during the first 28 h after IGF-I injection (28).
Studies were done at doses of 40, 80, and 120 ug/kg body weight with
pharmacokinetic profiles suggesting a plateau effect between 80 and 120
ug/kg per dose. It was considered that the carrying capacity of the
IGFBP was saturated at this level. Mean serum IGF-II levels decreased
concurrently with the increase in IGF-I, and serum IGFBP-3 levels did
not respond to IGF-I treatment. There was no apparent change in the
half-life of IGF-I during the treatment period, indicating no
alteration of IGF-I pharmacokinetics induced by prolonged treatment
(29).
|
Seventeen prepubertal Ecuadorian patients were entered into a randomized double-blind, placebo-controlled trial of IGF-I at 120 ug/kg sc twice daily for 6 months, following which all subjects received IGF-I. Such a study was considered necessary because of the observation of spontaneous periods of normal growth in these youngsters, the suggestion that nutritional changes that might accompany intervention would be an independent variable, and the need to control for side effects, particularly hypoglycemia, which occur in the untreated state. The nine placebo-treated patients had a modest, but not significant, increase in height velocity from 2.8 ± 0.3 to 4.4 ± 0.7 cm/yr, entirely attributable to three individuals with 6 month velocities of 6.6 to 8 cm/yr. Although this response was attributed to improved nutritional status, there was no accompanying increase in IGFBP-3, as was noted with nutrition-induced catch-up growth by Crosnier et al. (31) in their GHRD patient with anorexia. For those receiving IGF-I, the height velocity increased from 2.9 ± 0.6 to 8.8 ± 0.6 cm/yr and all 16 patients had accelerated velocities during the second 6 month period when all were receiving IGF-I. Again, no changes or differences in IGFBP-3 were noted. There was no difference in the rate of hypoglycemia events, nausea or vomiting, headaches, or pain at the injection site between the placebo and IGF-I-treated groups. Six of the seven IGF-I-treated patients experienced hair loss (32).
Two-year treatment results in the Ecuadorian group have been reported, comparing the 120 ug/kg bid dosage to 80 ug/kg bid and also comparing responses to those of GH-treated GHD in the Ecuadorian population. There were no baseline differences between the low- and high-dose groups for growth velocity, bone age, SDS for height, or mean percent body weight for height. No differences were seen between the two dosage groups in growth velocity or changes in height SDS, height age, or bone age. A group of six subjects receiving the higher dose followed for a 3rd yr continued to maintain 2nd yr growth velocities. The annual changes in height age in both the 1st and the 2nd yr of treatment correlated with IGF-I trough levels, which tended to be in the low normal range despite a failure of serum IGFBP-3 levels to increase. The comparable growth responses to the two dosage levels and the similar IGF-I trough levels were thought to indicate a plateau effect at or below 80 mcg/kg body weight twice daily (33).
When all 22 of these patients were compared to 11 GH-treated GHD subjects, the GHD group was found to have a greater change in SDS for height and height age but did not differ in the ratio of height age to bone age changes over the 2-yr period. There was, however, a greater change in mean percent body weight for height in the GHRD group treated with IGF-I, thought to indicate comparable effects on lean body mass of the two treatment programs, without the lipolytic effects of GH in the IGF-I-treated GHRD subjects. The difference in growth response between IGF-I-treated GHRD and GH-treated GHD was consistent with the hypothesis that 20% or more of GH-influenced growth is due to the direct effects of GH on bone (20). Nonetheless, the comparable ratios of height age change to bone age change suggested similar long-term effects for replacement therapy in these two conditions (33).
Unlike the results in the short-term adult studies, long-term administration of IGF-I in children was associated with elevated but not normal trough levels (12 h after previous injection) (26). This reduction in half-life of IGF-I compared to normals is almost certainly the result of deficient IGFBP-3 and ALS. In addition to a lack of change in serum concentrations of IGFBP-3 and ALS, there was no significant change in IGFBP-4 or IGFBP-5 concentrations (26).
|
![]() |
Footnotes |
---|
Received September 13, 1999.
Accepted October 14, 1999.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|