| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles |
Departments of Pediatrics (A.L.O.-S., D.B.D.) and Medicine and Clinical Biochemistry (M.A.S., S.O.), Addenbrooke’s Hospital, and University of Cambridge (M.A.S., D.B.D., S.O.), Cambridge, United Kingdom CB2 2QQ; and Department of Pediatrics, The John Radcliffe Hospital (S.J.H., M.Y.A.), Oxford, OX9 4BX, UK
Address all correspondence and requests for reprints to: Dr. A. L. Ogilvy-Stuart, Neonatal Unit, Rosie Hospital, Addenbrooke’s National Health Service Trust, Cambridge, United Kingdom CB2 2SW. E-mail: amanda.ogilvy-stuart{at}addenbrookes.nhs.uk
Abstract
Humans with congenital absence of the islets of Langerhans and mice
rendered null for the insulin receptor rapidly develop severe
hyperglycemia and ketoacidosis and, if untreated, die in the early
neonatal period. In contrast, children with homozygous or compound
heterozygous mutations of the insulin receptor gene, although
hyperglycemic postprandially, survive for many months without
developing ketoacidosis. Paradoxically, they often develop
hypoglycemia. The rarity of the condition and the difficulties of
undertaking metabolic studies in ill infants have limited the
physiological information that might explain the clinical features. We
studied a boy with Donohue’s syndrome who represents a further example
of the null phenotype, with two different and novel nonsense mutations
in the 
-subunit of the receptor. He survived for 8 months
without developing ketoacidosis, and fasting hypoglycemia was a
frequent problem. Despite the complete absence of insulin receptors,
evidence for persistent insulin-like effects on fat and liver was seen;
fasting plasma ß-hydroxybutyrate and nonesterified fatty acid
levels were low, fell further during the early postprandial period, and
failed to rise in response to hypoglycemia. The inverse relationships
between plasma insulin and insulin-like growth factor-binding protein-1
levels were maintained, suggesting persistent hepatic effects of
insulin. GH levels measured over a 6.5-h period were low
throughout.
Thus, the differences between congenital insulin deficiency vs. insulin receptor deficiency in humans may be explained by persistent insulinomimetic activity of the grossly elevated plasma insulin presumably being mediated through the type 1 insulin-like growth factor receptor. As GH plays a critical role in the regulation of ketogenesis during insulinopenia in humans, but not in rodents, this may contribute to the distinct phenotype of human vs. mouse insulin receptor knockouts.
DONOHUE’S SYNDROME (leprechaunism) results from the presence of homozygous or compound heterozygous mutations in the insulin receptor gene that produce a total or near-total absence of functional insulin receptors (1, 2, 3, 4, 5, 6, 7, 8, 9). Although the initial genetic studies described missense mutations (1), a number of examples of the null phenotype have been described (5, 8), including homozygous deletions of the insulin receptor gene (4). The absence of functional human insulin receptors is compatible with survival to term, albeit with low birth weight and marked phenotypic abnormalities, including low muscle and fat mass (10, 11, 12, 13, 14). Mice that have been rendered null for the insulin receptor by homologous recombination also have a low birth weight and a reduction in adipose tissue and muscle mass (15). However, all such mice die in a state of ketoacidosis within a few days of birth (15). In contrast, in human subjects with homozygous null mutations of the insulin receptor, ketoacidosis has not been reported (10, 11, 12, 13, 14). Remarkably, these and other subjects with Donohue’s syndrome can survive for months and sometimes years (10, 11, 12, 13, 14). They frequently suffer from fasting hypoglycemia, although all have been reported to have marked postprandial hyperglycemia. Thus, a paradox presents itself. Although congenital total deficiency of insulin (as is seen in congenital absence of the islets of Langerhans) in humans during the immediate postnatal period results in the development of fatal ketoacidosis (16), congenital total deficiency of insulin receptors is compatible with relatively prolonged, acidosis-free survival and is frequently associated with hypoglycemia in the fasting state. These observations provoke several questions. 1) Why do human infants who totally lack insulin receptors not rapidly develop diabetic ketoacidosis? 2) Why do such subjects frequently develop severe hypoglycemia? 3) Why is the null phenotype of the insulin receptor different in mice compared with humans?
We have studied a United Kingdom Caucasian subject with Donohue’s syndrome whose extreme insulin resistance was due to compound heterozygosity for two previously undescribed nonsense mutations in the extracellular domain of the insulin receptor. Despite having no functional insulin receptors, the infant survived for 8 months without developing diabetic ketoacidosis. In fact, fasting hypoglycemia was a major clinical problem. We have, within the limits of tolerance of an ill and poorly developed infant, examined intermediary metabolism in the fasting and postprandial state to gain insights into the possible physiological basis for the absence of ketoacidosis and the frequency of hypoglycemia in Donohue’s syndrome.
Case History
The baby was the first child of unrelated Caucasian parents, born at term after induction of labor, weighing 2.1 kg (less than third percentile). Breast-feeding was established by day 2, but irritability and progressive abdominal distension were observed during the first 4 weeks. His height and weight remained below the third percentile, and the lack of adipose tissue, coarse features, hirsutism, pigmentation, and abdominal distension together with symptomatic hypoglycemia documented at 2 months (glucose, 2 mmol/L) indicated a clinical diagnosis of Donohue’s syndrome. Daytime feedings were supplemented with 3% modified maize starch, and continuous overnight feedings of Polycal (glucose, maltose, and polysaccharides, providing 161 kJ/100 g) were given by nasogastric tube. At 3 months hypokalemia was documented together with increasing pigmentation, coarsening of facial features, and genital enlargement. A heart murmur was noted, and echocardiography revealed hypertrophic cardiomyopathy. Renal hypertrophy and hepatomegaly also developed before the patient’s death at 8 months from respiratory failure complicating a respiratory syncitial virus infection. No postmortem was performed.
Materials and Methods
Mutation detection
PCR-single strand conformational polymorphism (PCR-SSCP) analysis of the insulin receptor was carried out as previously described (17). Exons showing SSCP variant bands were sequenced as previously described (17) using the Sequenase II kit (U.S. Biochemical Corp., Cambridge, UK).
Exon 2 of the insulin receptor was amplified by PCR using Pfu DNA polymerase (Stratagene, Cambridge, UK) and subcloned into pBluescript SK (Stratagene) at the EcoRV site using standard protocols (18). Recombinant DNA was transformed into competent JM109 cells (Stratagene), and plasmids were purified using SNAP miniprep kits (Invitrogen, San Diego, CA). Purified plasmids were then sequenced in both directions using an automated ABI sequencer.
Insulin binding
Binding of insulin to cultured Epstein-Barr virus (EBV)-transformed lymphocytes was determined as previously described (19) by incubating 106 cells with [125I]insulin (3–4 x 10-11 mol/L) together with the indicated concentrations of cold insulin, insulin-like growth factor I (IGF-I), or receptor-specific monoclonal antibodies (total volume, 250 µL) for 3 h at 15 C.
Immunoprecipitation and Western blotting
EBV-transformed lymphocytes were lysed in 50 mmol/L HEPES (pH 7.4), 150 mmol/L NaCl, 10 mmol/L ethylenediamine tetraacetate, 0.5 mmol/L phenylmethylsulfonylfluoride, 2.5 mmol/L benzamidine, 1 µg/mL leupeptin, antipain, pepstatin, and 1% Triton X-100. Protein concentration was determined using the Bradford dye-binding procedure (Bio-Rad Laboratories, Inc., Hertfordshire, Hemel Hempstead, UK). Cell lysates were immunoprecipitated with a panel of insulin receptor-specific monoclonal antibodies or an irrelevant antibody for 2 h at 4 C. The immune complexes were then subjected to 6% reducing SDS-PAGE, proteins were transferred to Immobilon-P polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA), and membranes were probed with a polyclonal antibody raised against the insulin receptor C-terminal domain and a horseradish peroxidase-conjugated secondary antibody. Antibody binding was detected by enhanced chemiluminescence (Amersham Pharmacia Biotech, Little Chalfont, UK).
In vivo studies
All studies were undertaken with the full informed consent of the parents. Fasting blood samples were obtained from the subject (at 3 months) and from the parents for estimation of glucose and insulin and to permit extraction of DNA for genetic studies.
A standard oral glucose load (1.75 g/kg) was administered after a 2-h fast when the proband was 4 months of age. Fasting blood samples were taken for analysis of glucose, cholesterol, triglycerides, nonesterified free fatty acids (NEFA), sex hormone-binding globulin (SHBG), testosterone, ß-hydroxybutyrate, IGF-I, IGF-II, IGFBP-1, IGFBP-3, androstenedione, and leptin. After the glucose load, blood samples were taken half- hourly for 2 h and used for estimation of glucose, NEFA, and ß-hydroxybutyrate.
At 6 months of age, a 6.5-h plasma hormone profile was obtained using a microsampling technique (20). Samples of venous blood were collected through a 24-gauge cannula into 1:4 heparinized saline and aliquoted at 15-min intervals. GH, glucose, and insulin levels were measured every 15 min; the IGF-binding proteins, IGFBP-1 and IGFBP-3, were measured every 30 min. Triacylglycerol (TAG), NEFA, ß-hydroxybutyrate, glycerol, and lactate were measured each 15 min for the first 5 h, (encompassing three oral feedings).
Hormone and metabolite assays
Blood glucose was measured with a Stat Plus analyzer using a glucose oxidase method (model 2300, YSI, Inc., Farnborough, UK). C Peptide was measured using a double antibody RIA (EURO/Diagnostic Products, Glyn Rhonwy, UK). IGFBP-3 was measured using a coated tube immunoradiometric assay from Diagnostic Products (Webster, TX). Leptin was measured by RIA (Linco Research, Inc., St. Charles, MO). Insulin was assayed in duplicate using the Medgenix immunoenzymetric assay (Biosource Technologies, Inc., Fleurus, Belgium). Microsamples of insulin were measured using an amplified, end-point enzyme immunometric assay for specific measurement of human insulin (21). Proinsulin and 32/33 proinsulin were assayed in duplicate using a time-resolved fluorometric assay (DELFIA, Wallac, Inc., Turku, Finland). The solid phase antibody bound to a microtiter plate was the same in each case. This and the labeled antibody for intact proinsulin were previously described (22). TAG was measured using a micromethod capable of measuring subnanomolar quantities of TAG based on the formation of a colored formazan by reduction of 2-(-4-iodo-phenyl)-3-(-4-nitrophenyl)-5-phenyltetrazolium chloride. 2-(-4-Iodo-phenyl)-3-(-4-nitrophenyl)-5-phenyltetrazolium chloride was made up to 12 nmol/L solution and run on the Instrumentation Laboratories Multistat III microcentrifugal analyzer (Warington, UK). The TAG concentration is calculated from the difference total and free glycerol (23). Lactate, ß-hydroxybutyrate, and glycerol were analyzed using a micromethod for preparing percloric extracts of blood and run on a microenzymic assay for these analytes in an IL FLS Multistat III microcentrifugal analyzer (24).
NEFA levels were measured using an enzymic method (Alpha Laboratories, Eastleigh, UK). Plasma GH concentrations in the diluted plasma samples were determined by an immunoradiometric assay (NETRIA, St. Bartholomew’s Hospital, London, UK) (20). Plasma IGF-I levels were determined in undiluted plasma samples by RIA after acid-ethanol extraction as previously described (25). IGF-II was measured by RIA after acid-acetone extraction (26). IGFBP-1 was measured by RIA (27). SHBG was measured using an in-house saturation binding assay with tritium as label.
Normative data for the various hormone measurements are scarce at 3–8
months of age. We have provided in Table 2
the likely observed range in
this age group based on assays of samples from normal children ranging
from birth to 2 yr of age made in the same laboratory using the same
assay methods.
|
Detection of mutations in the insulin receptor gene
Two variant SSCP patterns were detected: one in exon 2 was found
in the patient and his father, whereas the other in exon 3 was found in
the patient and his mother (data not shown). Direct sequencing of exon
3 confirmed that both the patient and his mother were heterozygous for
a Gln272Stop codon mutation predicted to lead to
premature truncation of the insulin receptor (Fig. 1). Direct sequencing of exon 2 showed
multiple bands in the patient and his father, suggesting that they were
heterozygous for a frameshift mutation at Asn108.
Sequencing of the patient’s exon 2 DNA subcloned into the pBluescript
plasmid confirmed that he was heterozygous for a single base pair
deletion in codon 108, causing a frameshift and a stop codon at amino
acid 109. This patient is therefore a compound heterozygote for two
novel nonsense mutations in the extracellular domain of the insulin
receptor predicted to result in severely truncated receptors lacking
most of the extracellular domains, the transmembrane anchor, and all
the cytoplasmic domains necessary for biological activity (Fig. 1).
|
Expression of insulin receptors in EBV-transformed lymphocytes
from the patient, his parents, and controls was examined by insulin
binding and Western blotting. Cells from both parents bound insulin to
a similar extent as normal subjects (Fig. 2). In contrast, cells from the
patient bound insulin poorly (Fig. 2). Although in one experiment shown
in Fig. 2A, the low level of [125I]insulin
binding was inhibited by
10-9–10-8
mol/L unlabeled insulin, this was not confirmed in other experiments
(Fig. 2, A and B). Moreover, two different insulin receptor-specific
monoclonal antibodies (25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49) that had previously been shown
to inhibit insulin binding to its receptor (28) were
without effect on the patient’s cells, although they clearly inhibited
insulin binding to parental and control cells (Fig. 2B). As insulin can
bind with low affinity to the IGF-I receptor (29), it was
possible that the low level of insulin binding to the patient’s cells
was to the IGF-I receptor. To test this,
[125I]insulin binding was determined in the
presence of 10-8 mol/L
IGF-I or IGF-I receptor-specific antibodies known to inhibit IGF-I
binding to its receptor (30). Again, these were without
effect (Fig. 2C). Taken together, these results suggest that the low
level of [125I]insulin binding to the
patient’s EBV-transformed lymphocytes is not to the insulin receptor
or the IGF-I receptor.
|
), in rare cases exon skipping
or read-through of stop codons occur, resulting in a nearly full-length
protein lacking only the exon, or part of the exon, with the nonsense
mutation (31). To determine whether this was the case,
cell lysates from EBV-transformed lymphocytes were immunoprecipitated
with three insulin receptor-specific monoclonal antibodies that
predominantly recognize epitopes within exons 7 and 8 (), exon 12
(18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44), and exon 22 (CT-1) (32, 33, 34) and then blotted
with a polyclonal antibody against the C-terminal domain (Fig. 3). All three insulin receptor-specific
antibodies, but not an irrelevant antibody, immunoprecipitated a strong
band corresponding to the insulin receptor ß-subunit in cell lysates
from the parents and controls. However, the insulin receptor
ß-subunit was not detected in cell lysates from the patient.
Dilutions of a control lysate before immunoprecipitation with 83–14
indicated that the patient had less than 1% of the ß-subunit seen in
the control (Fig. 3A).
|
The nonfasting blood sample obtained at 2 months confirmed grossly
elevated insulin levels, which remained elevated despite a blood
glucose of 1.8 mmol/L in a fasting sample at 3 months (Table 2). In
contrast, glucose and insulin levels were normal in the parents (Table 1). In the proband, cholesterol
and triglyceride levels were not elevated in this fasting sample. SHBG
levels were inappropriately low for the child’s age, whereas
testosterone levels were probably normal, and androstenedione levels
were elevated. IGF-I and IGFBP-3 levels were normal and low normal,
respectively, whereas IGFBP-1 levels were elevated. The most
striking abnormality, however, was the very low IGF-II levels (Table 2).
|
. Fasting
blood glucose levels were normal, but an exaggerated postload rise was
observed at 30 min, with levels subsequently falling, resulting in
severe hypoglycemia at 120 min. ß-Hydroxybutyrate levels remained
suppressed throughout the test. NEFA levels were also low, and although
a modest rise was observed at 120 min, the ß-hydroxybutyrate and NEFA
responses to hypoglycemia were subnormal.
|
. At this stage the
patient’s blood glucose levels were elevated, particularly after
meals, but ß-hydroxybutyrate remained below the detection limit of
the assay throughout the sampling period. Grossly elevated insulin
levels were observed throughout the study period, but, nevertheless, a
further rise in levels was observed in relation to feedings. Glycerol,
TAG, and NEFA levels also rose after feedings, but levels fell
postprandially. GH levels remained low throughout the sampling period
(mean, 1.6 ± 0.45 µg/L). IGFBP-1 levels were elevated, but
showed the normal inverse relationship with plasma insulin levels
(r = -0.49; P = 0.001), whereas IGFBP-3 showed no
significant change during the profile in relation to feedings or
insulin level.
|
We have described a patient with Donohue’s syndrome (10), who has two novel nonsense mutations in exons 2 and 3 of the insulin receptor. Most premature stop codon mutations in the insulin receptor are associated with a reduction in messenger ribonucleic acid, but if expressed, truncated receptors lacking transmembrane and cytoplasmic domains (as predicted in this patient) are considered nonfunctional (35, 36). Although nonsense mutations occasionally give rise to exon skipping (31), this would result here in receptors lacking exons 2 and/or 3, which are necessary for insulin binding (37, 38), and so the receptors would most likely be nonfunctional. Moreover, as the patient’s EBV-transformed lymphocytes express very few, if any, mature receptors, it is unlikely, that exon skipping does take place. This patient, therefore, probably represents a further example of the null phenotype (5).
Glucose and insulin levels in the parents were normal, as has previously been reported in heterozygotes for this form of insulin receptor mutation. The rarity of the human null phenotype and the severity of illness present in these children mean that the metabolic sequelae of this genetic catastrophe are sparsely documented. However, as with all such experiments of nature, important information concerning normal human physiology can frequently be obtained from the detailed study of such rare patients. We have had the opportunity to undertake some, albeit limited, metabolic studies in this child, which have allowed us to begin to address some of the apparent paradoxes presented by this condition.
1) Why is the clinical phenotype associated with congenital absence of insulin receptors so different from that associated with congenital absence of insulin?
In subjects born with congenital absence of the islets of Langerhans, severe hyperglycemia and ketoacidosis develop rapidly. In contrast, subjects who are homozygous for null mutations in the insulin receptor, including the patient described here, are characterized by fasting hypoglycemia, postprandial hyperglycemia, and the absence of ketoacidosis. The studies performed in our patient provide evidence for the persistence of insulinomimetic effects despite the complete absence of insulin receptors.
Despite the grossly elevated fasting insulin levels, a marked postprandial rise in levels was still observed during the hormone profile. Although early postprandial glucose levels were abnormally elevated, the subsequent fall to hypoglycemic levels during the oral glucose tolerance test and the failure of NEFA to rise in response to hypoglycemia suggest that insulin was continuing to exert actions on glucose production/disposal and on lipolysis. Further evidence of continued insulin action is the maintenance of the normal inverse relationship between IGFBP-1 and insulin during the 6.5-h profile (39). Thus, in our patient, despite the absence of insulin receptors, IGFBP-I levels remained entrained to insulin throughout the study day. Although there is no formal proof that the postprandial insulin surges were causally linked to the above changes in circulating metabolites and binding proteins, this would appear to be the most likely explanation for their occurrence. The C peptide levels are lower than might be expected from the levels of insulin and proinsulin. Delayed clearance of insulin might therefore contribute to the marked hyperinsulinemia.
We were unable to directly measure the relative contributions of hepatic glucose production and peripheral glucose uptake to the changes in glucose level during the oral glucose tolerance test. However, the marked hypoglycemia at 2 h postglucose load indicates the suppression of hepatic glucose production, which could be mediated through direct effects of insulin at the liver or through a reduction in the supply of gluconeogenic precursors. The most likely candidate for a molecule mediating these effects of insulin is the type 1 IGF receptor, which has considerable structural and functional homology with the insulin receptor (40). Although type 1 IGF receptors disappear from the liver in adult life (41), they are present during fetal and early postnatal life and are also abundant in muscle (42). In support of the hypothesis that insulin could be signaling through the IGF-I receptor, both severe hypokalemia (43) and, in animal studies, proliferation of cardiac myocytes resulting in hypertrophic cardiomyopathy (44) have been reported with high dose IGF-I administration, and both were observed in our patient. There is a relative paucity of these receptors on the adipocyte (45), and this may explain the lack of adipose tissue development and reductions in fat stores in patients with Donohue’s syndrome. The modest rise in NEFA levels after feedings could reflect this lack of effect of insulin (acting through the type 1 IGF receptor) on lipid disposal. The subsequent reduction in NEFA, TAG, and glycerol levels postprandially might still be compatible with signaling through the type 1 IGF receptor at the adipocyte, but uptake by other tissues, such as muscle, is also possible. The genetic deletion of one particular receptor may result in the up-regulation of related receptors, therefore, interpretations relevant to normal physiology should be made with caution in knockout humans as well as mice.
Some of the phenotypic features (penile enlargement, hirsutism, and virilization) observed in our patient are commonly noted in Donohue’s syndrome and might also be explained by insulin action via the type 1 IGF receptor causing reduced SHBG levels and, thus, increased circulating levels of free androgen despite normal testosterone and modestly elevated levels of androstenedione, a weak androgen. Insulin regulation of SHBG in the isolated hepatocyte (46) and in the circulation (47) has been demonstrated, but as our patient has no functional insulin receptors, this suggests a role for the type 1 IGF receptor in the regulation of SHBG. In addition, both GH and insulin normally regulate hepatic IGF-I and, indirectly, IGFBP-3 production (48), yet circulating levels of these peptides were normal or only modestly reduced despite GH deficiency and complete insulin resistance. These findings lead us to speculate that during infancy, the hepatic type 1 IGF receptor may have an important role in regulating hepatic IGF-I and IGFBP-3. This is clearly not the case for IGF-II, as levels of this peptide were extremely low.
2) Why do the phenotypes of mouse and human insulin receptor knockouts vary so markedly?
It is striking that mice rendered null for the insulin receptor rapidly develop neonatal ketoacidosis, a phenomenon that is not seen in the human knockout. The different feeding patterns of neonatal mice and humans may play a role in this species divergence. Thus, from birth human infants have bolus feedings with rapid gastric emptying, whereas neonatal mice tend to feed near-continuously and rarely have an empty stomach. The different pattern of nutrient loading may thus contribute to the species differences in susceptibility to ketoacidosis.
We suggest another potential explanation for the species difference that involves an appreciation of the relative roles of GH in the regulation of lipolysis and ketogenesis in rodents and humans. In the rodent, GH does not have a major role in regulating lipolysis or ketogenesis during fasting or in the insulin-deficient state (49). In contrast, GH levels are elevated during fasting and during insulin deficiency in humans, and they have a critical role in regulating lipolysis and ketogenesis (50, 51). GH levels are high during the newborn period (52), with baseline and peak levels of 17.1 and 30.7 g/L, respectively, in appropriate for gestational age babies with higher levels (baseline, 25.9; peak, 45.0 g/L) in small for gestational age babies (52). In older prepubertal children (5.3–14.0 yr), baseline and peak GH levels are 0.35 and 16.5 g/L, respectively, with similar levels in prepubertal children born small for gestational age (53). This child with Donohue’s syndrome had GH levels that remained below 2.4 g/L throughout the 6.5-h study period. GH levels of similar magnitude to those described in our patient have been reported in other patients with Donohue’s syndrome (54), and they are thought to arise because of negative feedback resulting from insulin interaction with the type 1 IGF receptor at the level of either hypothalamus or pituitary. Thus, GH deficiency in the human insulin receptor knockout combined with a species difference in the dependency of ketogenesis on GH may explain the lack of ketoacidosis in human vs. rodent insulin receptor knockouts.
3) What underlies the transition from fasting hypoglycemia to fasting hyperglycemia?
The transition from fasting hypoglycemia to hyperglycemia with time is notable and is a common feature of Donohue’s syndrome. One possible explanation involves ß-cell exhaustion, with a failure of the islets to maintain the high levels of secretory activity required to preserve relative normoglycemia. This could relate to evolving changes in the GH/IGF-I axis; IGFBP-1 levels were lower at 6 months than at 3 months. Another potential mechanism for this is the deposition of islet amyloid, which has been described in patients with chronic hypersecretion of insulin due to extreme insulin resistance (55). Unfortunately, we were unable to obtain postmortem tissue from our patient to directly address this possibility.
Other potential explanations for the transition from hypo- to hyperglycemia are apparent. For example, if the maintenance of near-normoglycemia in the neonatal period is, as we suggest, critically dependent on type 1 IGF receptors, then the transition from hypo- to hyperglycemia could relate to developmental changes in the expression of type 1 IGF receptors. As the infant grows older, the subsequent development of fasting and postprandial hyperglycemia may reflect the disappearance of type 1 IGF receptors from the liver or the failure of insulin signaling through the type 1 IGF receptor to sustain normal growth and glucose homeostasis. More detailed information on the expression of IGF-I receptors in tissues such as liver, fat, and muscle at different developmental stages in the human would help to test this hypothesis.
In conclusion, we have had the rare opportunity to undertake limited metabolic evaluation of a human infant without insulin receptors. The pattern of responses of plasma glucose, intermediary metabolites, and IGFBP-1 to fasting and feeding strongly suggest that the high levels of circulating insulin present in this child are continuing to have insulinomimetic effects despite the complete absence of the cognate receptor. It appears most likely that the type 1 IGF receptor is playing a role in the mediation of these effects. Given the central role of GH in regulating human ketogenesis, the persistently low circulating levels of GH in this child may at least in part explain the absence of ketoacidosis. Future metabolic studies of this rare syndrome may shed further light on species differences in the control of intermediary metabolism and provide novel insights into the hepatic regulation of IGF-I, SHBG, and the IGFBPs during infancy.
Acknowledgments
We thank Dr. K. Frayn (Oxford Lipid Metabolism Group, Radcliffe Infirmary, Oxford, UK) and Prof. C. N. Hales (Department of Clinical Biochemistry, Addenbrooke’s Hospital, Cambridge, UK) for assisting with the assays, and Prof. K. Siddle (Department of Clinical Biochemistry, Addenbrooke’s Hospital) for the receptor-specific antibodies.
Footnotes
1 Supported by the Wellcome Trust. 
Received March 28, 2000.
Revised August 15, 2000.
Revised March 21, 2001.
Accepted March 26, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  A. McDonald, R. M Williams, F. M Regan, R. K Semple, and D. B Dunger IGF-I treatment of insulin resistance Eur. J. Endocrinol., August 1, 2007; 157(suppl_1): S51 - S56. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Dunger, C. J. Petry, and K. K. Ong Genetics of Size at Birth Diabetes Care, July 1, 2007; 30(Supplement_2): S150 - S155. [Full Text] [PDF]
|
|
|
|
 |
|
 I. Giurgea, C. Sempoux, C. Bellanne-Chantelot, M. Ribeiro, L. Hubert, N. Boddaert, J.-M. Saudubray, J.-J. Robert, F. Brunelle, J. Rahier, et al. The Knudson's Two-Hit Model and Timing of Somatic Mutation May Account for the Phenotypic Diversity of Focal Congenital Hyperinsulinism J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 4118 - 4123. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Hojlund, T. Hansen, M. Lajer, J. E. Henriksen, K. Levin, J. Lindholm, O. Pedersen, and H. Beck-Nielsen A Novel Syndrome of Autosomal-Dominant Hyperinsulinemic Hypoglycemia Linked to a Mutation in the Human Insulin Receptor Gene Diabetes, June 1, 2004; 53(6): 1592 - 1598. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
