Lucile Packard Foundation for Childrens Health (R.G.R.), Palo Alto, California 94304; Department of Pediatrics, Stanford University (R.G.R.), Stanford, California 94305; and Oregon Health & Science University (R.G.R., V.H.), Portland, Oregon 97201
Although research over the last 20 yr has led to the elucidation of genetic etiologies for isolated GH deficiency and combined pituitary hormone deficiencies (1), our understanding of growth failure in the presence of normal GH secretion remains primitive (2). In part, this reflects the intricacies of human growth, which is most certainly the product of a multigenic process that results in the broad range of growth patterns observed throughout our species. Despite this complexity, investigations involving targeted gene disruption of critical components of the growth axis (3), supported by reports of corresponding human mutations (4, 5, 6, 7), have convincingly demonstrated the central role of the GH-IGF system in mammalian growth. In the process, they have provided important insights into the molecular basis of non-GH-deficient growth failure.
A variety of studies have indicated that approximately 25% of children evaluated for idiopathic short stature (ISS) have primary IGF deficiency (IGFD), i.e. abnormally low serum concentrations of IGF-I in the face of normal GH secretion (8, 9). In its purest and most dramatic forms, primary IGFD has been identified with three classes of molecular defects: 1) GH insensitivity resulting from mutations or deletions of the gene for the GH receptor (GHR) (classical Laron syndrome) (4); 2) genetic defects affecting the Janus kinase 2 (JAK2)-signal transducer and activator of transcription 5b (STAT5b) GH signaling pathway (5); and 3) deletions or mutations of the gene for IGF-I (6). Such cases appear, at least at this time, to be quite rare; whereas several hundred patients with mutations of the GHR gene have been identified, only one subject with a STAT5b mutation and two patients with IGF-I gene defects have been reported to date. It is to be anticipated, however, that such cases represent the tip of the iceberg for ISS, and, as has been observed with other receptor and postreceptor defects, identification of the molecular defects in the most severe cases will be followed by characterization of more subtle GHR and/or signaling abnormalities.
The report by Lewis et al. (10) of a dysfunctional GH variant raises interesting questions concerning a form of short stature poised somewhere midway between the realm of GH deficiency and that of primary IGFD. Two patients were identified in a study of "familial short stature," (although ISS appears to be a more appropriate definition, because parental heights were not always abnormal). Significantly, all subjects had normal GH secretion, as defined by peak serum GH concentrations more than 10 ng/ml after pharmacological provocation. Subjects were screened for mutations of GH1; five heterozygous GH1 gene lesions were identified in four different individuals. Two mutations were novel: Ile179Met and -360AG. Secretion of Ile179Met GH from rat pituitary cells was normal, as were receptor binding characteristics and sensitivity to proteolysis. The Ile179Met variant was found to induce phosphorylation of STAT5 normally, but activation of ERK was reduced to 50% of that observed with wild-type GH.
Taken at face value, these data would appear to support the hypothesis that the patients short stature resulted from secretion of a GH variant that resulted in differential activation of the ERK and STAT5 pathways, which, as the authors rightly note, would be a highly unusual, if not unprecedented, effect of a mutant receptor agonist. It is, therefore, all the more important to take a critical perspective of the authors conclusions. Of note: 1) serum GH concentrations in the patient were not elevated (peak provocative GH of 10.4 ng/ml), an atypical characteristic of patients with some degree of GH insensitivity (whether to exogenous or endogenous GH); 2) the serum IGF-I was not particularly low (-1.2 SD), necessitating the unlikely conclusion that the short stature resulted from something other than simple IGFD; 3) the father, heterozygous for the same GH mutation, was not particularly short (height SD of -1.4, compared with the mother with a height SD of -3.9, despite wild-type GH); and 4) interpretation of studies of the STAT5 pathway were limited by the use of antibodies incapable of distinguishing between STAT5a and STAT5b.
Most significant in interpreting these findings is the observation that the growth characteristics of the one patient reported with homozygosity for a mutation of STAT5b are indistinguishable from those observed in patients with GH insensitivity resulting from homozygosity for mutations of GHR (5); in other words, selective defects in the STAT5b pathway apparently could duplicate all of the growth characteristics of total absence of wild-type GHR. This patient had a height SD of -7.5, extremely low serum IGF-I, and a minimal response of the IGF axis to GH administration. One would infer from these observations that, at least in humans, the growth-promoting actions of GH and the GHR are mediated largely, if not exclusively, through the JAK2/STAT5b/IGF pathway, which was not impaired in the patient with the Ile179Met GH variant. To the credit of Lewis et al. (10), the authors fully recognize that the Ile179Met variant did not cosegregate with the short stature phenotype in the family reported and suggest that it may have been operating in concert with other, as-yet-unidentified, genetic loci.
In many ways, the GH variant described by Lewis et al. resembles the cases of biologically inactive GH reported by Takahashi and colleagues (11, 12). The concept of bioinactive GH was proposed by Kowarski et al. (13) in 1978, but the original reported cases lacked convincing biochemical data, nor was a definitive molecular lesion in those initial patients ever identified. The cases from Japan are somewhat more persuasive: 1) a child heterozygous for a R77C mutation (11), which appeared to behave as a GH antagonist, as in the case reported by Lewis et al. (10); however, the same mutation was identified in the normal-statured father; 2) a child heterozygous for a D112G mutation, which appeared to inhibit GHR dimerization (12). Thus, whereas it appears likely that true cases of bioinactive endogenous GH will be identified, some uncertainties surround each of the cases reported to date.
Whether or not the findings specific to the Ile179Met GH variant hold up is not, however, the critical point of the provocative study by Lewis et al. (10). Although it must be recognized that ISS represents a heterogeneous group of disorders, many such cases are likely to prove to have subtle defects of the multiple genes involved in the regulation of the skeletal response to endogenous (and, in some cases, exogenous) GH. Investigation of the biochemical and molecular basis for ISS, i.e. abnormal growth in the face of apparently normal GH production and secretion, will be a daunting task (14). For the 25% of ISS patients with IGFD, molecular defects of the GH-IGF axis represent fertile ground for genetic investigation. Appropriate candidate genes include GH1 (as proposed by Lewis et al.) (10); GHR (both mild homozygous and heterozygous states); JAK2, STAT5b, and other components of the GH signaling cascade; and IGF-I. Heterozygosity for GHR mutations has already been proposed by Goddard et al. (15) as a cause of ISS. In addition to the patient with the STAT5b mutation (5), Salerno et al. (16) identified abnormal tyrosine phosphorylation in two of 14 children with ISS, and Freeth et al. (17) have also observed abnormal GH signaling in fibroblasts from patients with severe growth failure and normal GHRs. For patients with normal serum IGF-I, defects involving IGF receptors, IGF binding proteins, the IGF signaling cascade, and epiphyseal responsiveness to IGFs and other growth factors are likely to be uncovered. A molecular basis for ISS, which has for too long been vaguely and noncriticially viewed as a "normal variant," represents, perhaps, the most exciting current challenge for growth research.
Footnotes
Address all correspondence and requests for reprints to: Ron Rosenfeld, M.D., Senior Vice-President for Medical Affairs, Lucile Packard Foundation for Childrens Health, 770 Welch Road, Suite 350, Palo Alto, California 94304. E-mail: ron.rosenfeld{at}lpfch.org.
This work was supported by National Institutes of Health Grants CA58110 and DMD17-00-0042 (to R.G.R.).
Abbreviations: IGFD, IGF deficiency; ISS, idiopathic short stature; JAK2, Janus kinase 2; STAT5b, JAK2, signal transducer and activator of transcription 5b.
Received January 20, 2004.
Accepted January 20, 2004.
References