No abstract

No abstract

We have previously shown that familial septo-optic dysplasia (SOD), a syndromic form of congenital hypopituitarism involving optic nerve hypoplasia and agenesis of midline brain structures, is associated with homozygosity for an inactivating mutation in the homeobox gene HESX1/Hesx1 in man and mouse. However, as most SOD/congenital hypopituitarism occurs sporadically, the possible contribution of HESX1 mutations to the aetiology of these cases is presently unclear. Interestingly, a small proportion of mice heterozygous for the Hesx1 null allele show a milder SOD phenocopy, implying that heterozygous mutations in human HESX1 could underlie some cases of congenital pituitary hypoplasia with or without midline defects. Accordingly, we have now scanned for HESX1 mutations in 228 patients with a broad spectrum of congenital pituitary defects, ranging in severity from isolated growth hormone deficiency to SOD with panhypopituitarism. Three different heterozygous missense mutations were detected in individuals with relatively mild pituitary hypoplasia or SOD, which display incomplete penetrance and variable phenotype amongst heterozygous family members. Gel shift analysis of the HESX1-S170L mutant protein, which is encoded by the C509T mutated allele, indicated that a significant reduction in relative DNA binding activity results from this mutation. Segregation analysis of a haplotype spanning 6.1 cM, which contains the HESX1 locus, indicated that only one HESX1 mutation was present in the families containing the C509T and A541G mutations. These results demonstrate that some sporadic cases of the more common mild forms of pituitary hypoplasia have a genetic basis, resulting from heterozygous mutation of the HESX1 gene.

Recent discoveries of human genetic leptin deficiency have demonstrated its importance in regulating weight gain in early childhood. To investigate whether normal variation in leptin and insulin levels in cord blood could influence infancy growth, we assayed samples from 197 infants from a representative birth cohort, who were measured at birth, 4, 8, 12 and 24 months. Cord leptin levels correlated most closely with weight and ponderal index (kg/m3) at birth, but also with length and head circumference (all p<0.0005). Independent of birth size, females had higher leptin levels than males (p<0.0005). Cord levels of leptin, but not insulin, were negatively related to weight gain (p<0.005) from birth to 4 months, and accounted for 9.4% of the variance in weight gain, compared with breast/bottle feeding (3.5%) and early/late introduction of solids (1%). The effect of leptin levels on weight gain was independent of birthweight, and was still evident at 24 months. The wide variation in infancy growth ('catch-up' or 'catch-down') may be partly determined by leptin levels preset in utero. Our data support a role for leptin in the regulation of infancy weight gain, and suggest a mechanism whereby infants may 'catch-up' in growth postnatally following an adverse intrauterine environment.

GH insensitivity syndrome (GHIS) is associated with many different mutations of the GH receptor (GHR) gene. We examined the phenotypic and biochemical features in 82 GHIS patients from 23 countries, each fulfilling diagnostic criteria of GHIS. There were 45 males and 37 females [mean age, 8.25 yr; mean height, -6.09 SD score, and mean insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3), -7.99 SD score]. Sixty-three were GH-binding protein (GHBP) negative; 19 were GHBP positive (> 10% binding). The mean height in GHBP-negative subjects was -6.5 SD score, and that in GHBP-positive patients was -4.9 SD score (P = < 0.001). Clinical and biochemical heterogeneity was demonstrated by the wide range of height (-2.2 to -10.4 SD score) and IGFBP-3 (-1.4 to -14.7 SD score) values, which were positively correlated (r2 = 0.45; P = < 0.001). This contrasted with the lack of correlation between mean parental height SD score and height SD score (r2 = 0.01). Fifteen different GH receptor gene mutations were identified in 27 patients. All had homozygous defects, except 1 who had a compound heterozygous defect. The mutations were 5 nonsense, 2 frame shift, 4 splice, 4 missense, and 1 compound heterozygote. There was no relationship between mutation type or exon of the GHR gene involved and height or IGFBP-3 SD score. In conclusion, GHIS is associated with wide variation in the severity of clinical and biochemical phenotypes. This variation cannot clearly be accounted for by defects in the GHR gene. Other genetic and/or environmental factors must, therefore, contribute to phenotype in GHIS.