Expanding Genetic and Functional Diagnoses of IGF1R Haploinsufficiencies

Abstract

Background

The growth-promoting effects of IGF-I is mediated through the IGF-I receptor (IGFIR), a widely expressed, cell-surface tyrosine kinase receptor. IGF1R copy number variants (CNV) can cause pre- and postnatal growth restriction or overgrowth.

Methods

Whole exome sequence (WES), chromosomal microarray and targeted IGF1R gene analyses were performed on three unrelated children who share features of SGA, short stature, and elevated serum IGF-I, but otherwise had clinical heterogeneity. FACS (fluorescence-activated cell sorting) analysis of cell surface IGF1R was performed on live primary cells derived from the patients.

Results

Two novel IGF1R CNV and a heterozygous IGF1R nonsense variant were identified in the three patients. One CNV (4.492 Mb) was successfully called from WES, utilizing eXome-Hidden Markov Model (XHMM) analysis. FACS analysis of cell surface IGFIR on live primary cells derived from the patients demonstrated ~50% reduction in IGFIR availability associated with the haploinsufficiency state.

Conclusion

In addition to conventional methods, IGF1R CNVs can be identified from WES data. FACS analysis of live primary cells is a promising method for efficiently evaluating and screening for IGFIR haploinsufficiency. Further investigations are necessary to delineate how comparable IGFIR availability lead to the wide spectrum of clinical phenotypes and variable responsiveness to rhGH therapy.

INTRODUCTION

Normal human intrauterine and postnatal growth requires an intact insulin-like growth factor I (IGF-I) and IGF-I receptor (IGFIR) axis. IGF1R haploinsufficiency due to molecular defects is associated with impaired growth, and indicates that a single allele of the IGF1R gene is insufficient for normal growth and development. A complete loss of IGF1R has yet to be reported in humans, and may be lethal, as was shown in rodent studies in which targeted ablation of the Igf1r gene resulted in perinatal death (1,2). Indeed, for the most hypomorphic of IGF1R mutations described to date (two homozygous missense mutations (3,4) and two compound heterozygous IGF1R mutations (5,6)), residual expression and function likely ensured survival, although clinical presentations were more severe than in patients who were IGF1R haploinsufficent (7,8). Current treatment options of recombinant human (rh) GH or rhIGF-I therapy, approved for multiple indications including idiopathic short stature and SGA (small for gestational age), remain a conundrum, as IGF1R deficient patients exhibit IGF-I resistance, often with serum IGF-I concentrations well above the normal ranges.

Synthesized as a single polypeptide precursor, the IGFIR undergoes proteolytic cleavage into α- and β-chains and forms a tetramer (α₂β₂), with the extracellular α_2-subunits involved in ligand binding and the intracellular β₂-subunits carrying the intrinsic tyrosine kinase functions necessary for signal transduction (9). Ligand association leads to IGFIR autophosphorylation and activation of multiple downstream signaling pathways, including the phosphatidylinositol 3-kinase/AKT and MAPK/ERK pathways important for cell survival and growth (10). In an IGF1R haploinsufficiency state, IGF-I-induced IGF1R signaling is reduced (11–13), although binding of IGF-I may remain normal (14,15), further stressing the necessity of bi-allelic expression of the IGF1R gene for full biological activity.

In this study, we report three new IGF1R haploinsufficiency cases and highlight: (a) the unique application of whole exome sequencing (WES) data for assessing copy number variation in the genetic workup of short stature; (b) the utility of fluorescence-activated cell sorting (FACS) analysis as a promising method for efficiently evaluating cell surface IGFIR availability using live primary cells derived from patients; and (c) the effectiveness of rhGH therapy for treating some patients. Our study expands the genetic and functional diagnosis for an IGF1R haploinsufficiency state, emphasizes the importance of including CNV analysis in the initial evaluation of children who present with clinical and biochemical features suggestive of IGFIR insufficiency, and raises questions of how comparable IGF1R availability result in variable clinical phenotypes and responsiveness to rhGH therapy.

CLINICAL REPORTS

Proband 1

P1, an 8-yr-old Chilean girl with short stature, was born at 35 weeks of gestation with a birth weight of 2.1 kg, and birth length of 42 cm (SDS −2.61), to non-consanguineous parents. She showed catch-up growth (to 25^th percentile for height) until age 5 yrs when she started to deviate below this channel. At age 8.3 yrs, her height was 115.5 cm (SDS −2.06), weight was 22.1 kg (SDS −0.92), body mass index was 16.3 kg/m² (SDS +0.25) and head circumference was 51 cm (SDS −1.0). The patient had proportional body segments and no dysmorphic features.

Her endocrine evaluations at the age of 8 yr revealed elevated serum IGF-I (488 ng/mL, RR, 120–385) and serum IGFBP-3 (6.2 mg/L, RR, 2.2–6.5) for her prepubertal, chronological age. Her basal serum GH was 1.1 ng/L, which peaked at 11.2 ng/L after stimulation with clonidine. A diagnosis of IGF-I insensitivity was made.

During this period, rhGH therapy starting at a relatively low dose of 0.117 mg/kg/week, led to a clinically significant increase in her growth velocity. During rhGH treatment, her height shifted from below the 3rd percentile to approximately the 10^th percentile at 12 years of age (Figure 1A). As expected, her serum IGF-1 concentrations on rhGH therapy were high (811 to 1145 ng/mL). Her pubertal development advanced normally, and she underwent menarche at 12.5 years of age.

Figure 1.

Growth profiles of Probands. (A) Growth chart of Proband 1 (P1), arrow indicates start of GH treatment. (B) Growth chart of Proband 2 (P2). (C) Growth chart from Proband 3 (P3) during GH treatment.

Her family history reveals that her father is modestly short, with a height of 163 cm (SDS −1.93), whereas her mother’s height is normal (157 cm, SDS −0.98) as is her 18 yr old brother (168 cm SDS −1.2). A summary of the endocrine studies for P1 is shown in Table 1.

Table I.

Summary of biochemical data of affected Probands.

	P1	P2	P3
Gender	Female	Male	Female
Gestational age, wk	35	36	Term delivery
Birth length, cm	42 (SDS −2.61)	43.2 (SDS −2.3)	Not recorded
Birth weight, kg	2.1 (SDS −1.57)	1.93 (SDS −2.25)	1.65 (SDS −3.0)
Head CF, cm	NA	32.5 (SDS −1.5)	NA
First consult
Age, yr	8.3	8	6
Bone age	7 yrs	7yrs	5 yrs 9 mo
Height, cm	115.5 (SDS −2.06)	(SDS −3.4)	(SDS −5.2)
Weight, kg	22.1 (SDS −0.92)	(SDS −2.5)	(SDS −5.0)
Head CF, cm	51 (SDS −1.0)	(SDS −2.0)	-----
Glucose, mg/dL	85	—	—
TSH, μu/ml	4.35	—	0.88
T4 free, ng/dL	1.22	2.1	1.7
GH, stimulated (Clonidine), ng/ml	11.2	15.5	35
IGF-I, ng/mL	488 (RR, 120–385)	161 (RR, 52–231)	269 (RR, 52–297)
IGFBP-3, mg/L	6.2 (RR, 2.2–6.5)	4.03 (RR, 2.1–4.2)	4.5 (RR, 1.3–5.6)
ALS, mg/L	–	8.8 (RR, 4.2–13)	–
GHBP pmol/L	–	916 (RR, 267–1638)	–
Cerebral RNM	Normal	Stable lesion in left inferior cerebellar peduncle	–
Post-rhGH treatment		NA
Age, yr	12.08		10 yr
Height, cm	−1.4		−4.1
IGF-1, ng/mL	1145 (RR, 143–693)		709 (RR, 65–457)
IGFBP-3, mg/L	8 (RR, 2.4–8.4)		6.5 (RR, 2.9–5.2)

Proband 2

P2, is an 8-yr-old Caucasian boy with neurofibromatosis type 1 (NF-1) who presented with severe growth failure. He was born small for gestational age (SGA) at 36 weeks with a birth weight of 1.930 kg (SDS −2.25), birth length of 43.2 cm (SDS −2.3) and head circumference of 32.5 cm (SDS −1.5). His growth chart is shown in Figure 1B. He required therapy for mild motor and language delays in the toddler years. His brain magnetic resonance imaging showed a stable lesion (likely a glioma) in the left inferior cerebellar peduncle, with normal pituitary and optic nerve tracts.

The initial endocrine evaluations (Table 1) showed slow weight gain (SDS −2.5), growth failure (SDS −3.4) and mild microcephaly (head circumference SDS−2.0). Physical exam showed café-au-lait spots, thin upper lip, fifth finger clinodactyly and subcutaneous neurofibromas. His development was age appropriate; only symptoms of attention deficit and hyperactivity were observed.

His endocrine evaluation at the age of 8 yr revealed serum IGF-1 of 161 ng/ml and serum IGFBP-3 of 4.03 mg/L, both normal for his chronological age. His GH binding protein was 916 pmol/L; acid labile subunit was 8.8 mg/L and thyroxine was 2.1 ng/dL, all within the normal range. His serum GH peaked at 15.5 ng/mL after stimulation with arginine/clonidine.

His family history revealed his maternal grandmother is modestly short, with a height of 152 cm (SDS −1.8), whereas his paternal grandparents are normal statured. His mother, who also has NF-1, had a normal height of 160 cm (SDS −0.5) and his father’s height is 178 cm (SDS 0.2). Clinical data for P2 are shown in Table 1.

Proband 3

P3, is a 6 year old Chinese girl with severe postnatal growth failure. She was born SGA from a term delivery with a birth weight of 1.65 kg (SDS −3.0); birth length was not recorded. She underwent repair of a ventricular septal defect and an aortic coarctation in the neonatal period. She did not speak until age 3 years and was subsequently diagnosed with a central auditory processing disorder.

Her initial endocrine evaluation showed that her height was −5.2 SDS and weight, −5 SDS. The physical exam showed mild mid-face hypoplasia, crowded dentition and ocular hypertelorism. Her serum IGF-I was 296 ng/ml (reference range, 52–297) and IGFBP-3 was normal. GH stimulation test with clonidine a peaked of 35 ng/ml.

During this period, rhGH therapy was started at a dose of 0.30 mg/kg/week (Figure 1C). From age 6.3 yrs to 9.5 yrs, her height improved to −4.1 SDS with persistently elevated IGF-I and IGFBP-3 concentrations. rhGH was discontinued at age 10 yrs after growth failed to improve on a higher rhGH dose of 0.42 mg/kg/week, despite an elevated serum IGF-I concentration of 709 ng/ml (RR, 65–457) and an IGFBP-3 of 6.5 mg/L (RR, 2.9–5.2). At age 12.8 yrs, she had Tanner 2 breast and Tanner 3 pubic hair, irregular menses, a bone age of 13, and an improved growth rate of 7.1 cm/yr off rhGH, with a markedly high serum IGF-I concentration of 1183 ng/ml. At age 14.2 yrs, her height was 136.4 cm (−3.7 SDS). Clinical data for P3 are summarized in Table 1.

Her family history reveals that her father’s height is normal at 170 cm (SDS −0.2) and her mother’s height is also normal (163 cm; SDS −0.2). She has one sister also reported with normal linear growth.

MATERIALS AND METHODS

Samples from probands and family members

Blood samples from the probands and all immediate family members were collected for genetic studies with informed consent and in compliance with the respective institutional review boards. Serum assays for P1 were performed by standard techniques in the IDIMI laboratory. Serum IGFBP-3 was measured by an immunoradiometric assay from Diagnostic System Laboratories (Webster, TX, USA). Serum assays for P2 and P3 were measured as previously described (11).

Primary dermal fibroblast cell cultures

Primary fibroblast cultures were established from skin biopsies taken from the forearm of the index patients (P1, P2), and from a control subject with normal pre- and postnatal growth. Samples were collected in compliance with the study protocol approved by the institutional review boards of San Borja-Arriarán Clinical Hospital, Cincinnati Children’s Hospital Medical Center and Oregon Health and Science University. Fibroblast cultures were maintained as previously described (11).

Genomic DNA and cDNA

Extraction of genomic DNA from whole blood or primary fibroblast cultures, total RNA from primary fibroblast cultures, synthesis of cDNA, Sanger sequencing of the IGF1R gene, were previously described (11). The primers for PCR amplification of exon 13: Forward, 5′-GCCAAGGGTGTGGTGAAAGATGAA-3′; Reverse, 5′-ACCACATGGTGACAATTGAACTCCTTCATC-3′

Fluorescent In Situ Hybridization (FISH) analysis

The RP11-602 FISH probe and standard cytogenetic techniques were used to confirm the deletion observed in P2.

Chromosomal Microarray

Single nucleotide polymorphism (SNP) microarray analysis was performed on genomic DNA samples from P2. The Illumina Infinium assay was performed on DNA samples using the CytoSNP-850K BeadChip platform. B-allele frequency and Log2R ratio were analyzed with the Illumina GenomeStudio analysis software, and DNA copy number changes were prioritized using cnvPartition software at Cincinnati Children’s Hospital Medical Center Human Genetics Diagnostic Laboratories.

Whole Exome Sequencing (WES)

Whole exome sequencing was performed on DNA extracted from whole blood from P3 and her parents. After library construction, hybridization and capture were performed using the Illumina’s Rapid Capture Exome Kit and following the manufacturer’s suggested protocol. Sequencing, downstream sequence alignment, variant calling, and variant annotation were performed as previously described (16). Exomes were sequenced with mean target coverage >80×. Variants were then filtered for rare (Minor Allele Frequency, MAF <1%) nonsynonymous variants that segregated with the phenotype as previously described (16). Copy number variation was called using the eXome-Hidden Markov Model (XHMM) algorithm (17) using default parameters and adjustment for the first 5 principal components. CNVs were jointly called and genotyped along with 189 additional samples sequenced on the same platform, including samples from the parents of P3.

Flow cytometry analysis by fluorescence-activated cell sorting (FACS)

Flow cytometric analysis by fluorescence-activated cell sorting (FACS) of cell surface IGF1R on live primary fibroblasts were performed (for detailed procedure, see Supplemental Methods). Briefly, adherent fibroblasts were gently lifted with 0.05% trypsin/5mM EDTA after IGF-I treated (100 ng/mL, 24h), trypsin activity were neutralized with fetal bovine serum supplemented media, and RPMI-washed suspended cells in aliquots of 100μl (10⁶ cells/mL RPMI) were processed for FACS analysis as for PBMCs (see below). All experiments were performed in triplicates, at least 2 independent times.

PBMCs were prepared for FACS of cell surface IGFIR (18). Briefly, cells, treated for 1 hour with or without 50 ng/mL IGF-I (GroPep Ltd, Thebarton, South Australia, Australia), were incubated with phycoerythrin (PE)–conjugated anti-human IGFIR-α (CD221; BD Biosciences, San Jose, California) for 30 minutes (4°C in the dark), and stained with 0.25% propidium iodide (PI). For each sample, a total of 100,000 live PBMCs or 20,000 live fibroblasts (PI-negative cells) were acquired via a FACSCalibur flow cytometer (BD Biosciences), and the fluorescence emitted by IGFIR-PE–labeled cells (gated as shown in Supplemental Figure 1) was analyzed using FCS Express 3 analysis software (De Novo Software, Los Angeles, California).

Western immunoblot analysis

Cultured fibroblasts were serum starved overnight prior to IGF-I treatment (15 or 20 min), cell lysates collected and subject to Western immunoblot analyses (19). Rabbit polyclonal IgG against phospho-AKT (dilution 1:1000) and rabbit monoclonal IgG against Akt (dilution 1:2000) were purchased from Cell Signaling Technologies (Beverly, MA, USA). The anti-rabbit IgG secondary antibody was purchased from Amersham-Pharmacia Biotech (Uppsala, Sweden).

RESULTS

Identification of novel heterozygous IGF1R variants

Clinical and biochemical profiles of the 3 probands were consistent with resistance to the growth-promoting effects of IGF-I. Defective IGFIR expression was considered. Targeted sequencing of the IGF1R gene of genomic DNA (whole blood or cultured fibroblasts) from P1 identified a heterozygous IGF1Rc.2629C>T variant (rs150221450; minor allele frequency of 8.237e-06, Exome Aggregation Consortium Browser) in exon 13 (Figure 2A), a transition variant confirmed in Sanger sequencing of the IGF1R cDNA. Analysis of the genomic DNA from the other family members revealed that the father was also heterozygous for c.2629C>T. The c2629C>T changed codon Arg877 (CGA) to a stop codon (TGA), p.R877*, generating a predicted truncated IGFIR protein which, if expressed, would be unable to anchor to the cell membrane.

Figure 2.

Molecular defects identified in Probands. Probands, indicated by arrow. Height SDS, indicated. Family members not genetically analyzed, “?”. (A) Electropherogram of the heterozygous IGFIR c.2629C>T variant in P1, compared to normal sequences, and segregation of the variant (half shaded) in the family. (B) For P2: heterozygous 15q23.3 deletion involving the 0.282 Mb region which included exon 4 to 21 of the IGFIR gene. Co-segregation of the heterozygous 15q23.3 deletion and NF1 phenotype in the family, half-shaded. (C) De novo heterozygous 4.492 Mb deletion on chromosome 15 [46 XX del(15)(q26.2:qter)(97,970,832–102,463,314)] identified in P3.

In contrast to P1, probands P2 and P3 were found to carry copy number variants of IGF1R. For P2, SNP microarray analysis identified an unique heterozygous interstitial deletion of approximately 0.282 Mb on chromosome 15 [46 XY del(15)q26.3(99,438,083–99,720,341)]. This region includes exons 4 to 21 of the IGF1R gene (Figure 2B), the deletion of which resulted in loss of ~75% of the IGF1R coding sequences, in addition to three other genes that have unknown phenotypic effects (PGPEP1L, SYNM and TTC23). Parental study of chromosome 15q26.3 by FISH analysis confirmed a maternally inherited deletion (Figure 2B).

The IGF1R copy number variant in P3 was identified through re-evaluating WES data for CNVs, an analysis procedure not routinely performed with WES. Initial analysis of WES (P3 and parents) utilizing stringent filtering of called rare variants did not reveal an obvious genetic cause for the phenotypic presentation. Analysis was subsequently expanded to CNV discovery, applying the XHMM algorithm (17). A total of 27 CNVs (16 deletions and 11 duplications) were successfully detected in the patient’s sample, including a large 4.492 Mb heterozygous deletion on chromosome 15 [46 XX del(15)(q26.2:qter)(97,970,832–102,463,314)] (Figure 2C). The CNV was not present in either the parents, or in the 189 additional exomes that were simultaneously analyzed, suggesting the 4.492 Mb deletion was de novo in P3. All other CNVs were either inherited from the parents and/or were present in additional individuals in the CNV call set.

The 4.492 Mb CNV partially or fully overlapped a total of 28 genes, including the IGF1R gene (OMIM 270450). Five of the other genes within the deletion are associated with Mendelian phenotypes that are autosomal recessive (OMIM 613195, 615023, 614340, 615113, 605282) and we did not find additional rare protein-altering variants within their sequences. While it is possible that any individual gene, or a combination of these 5 genes in the 4.492 Mb deletion, could have contributed to the complex phenotype, data base search of each gene for functions and association with diseases (OMIM, Pubmed and UniProt) supports deletion of IGF1R as the primary contributor. Of note, no other rare protein-altering sequence variants in IGF1R were identified. P3 was treated with rhGH, and demonstrated only a modest growth response to GH therapy (Figure 1C), despite markedly increased serum IGF-I and IGFBP-3 concentrations. Samples for further analysis of IGFIR protein expression and functions, were, unfortunately, not available.

Fluorescent-activated cell sorting (FACS) of cell-surface IGFIR on Live primary cells

An IGFIR haploinsufficiency state was indicated for each of our 3 probands. We had previously demonstrated that IGF1R haploinsufficiency due to nonsense-mediated mRNA decay, IGF1Rc.3348_3366dup (11), and an hypomorphic IGF1R p.E121K/E234K compound heterozygous mutation (6), resulted in significantly reduced total IGFIR protein expressions with concomitant reduction in IGF-I-induced signaling. To assess if IGFIR expression was similarly altered in our present patients, we employed flow cytometry analysis by fluorescent-activated cell sorting (FACS) of live cells derived from P1 and P2, probing for the presence of cell-surface IGFIR (Figure 3). Unlike methods for assessing total IGFIR expression, FACS analysis provides relative quantitation of expressed IGFIR that had translocated appropriately to the cell surface, thus reflecting in vivo availability of IGFIR.

Figure 3. IGFIR Expression and functional analysis.

FACS analysis of cell surface IGFIR, labeled by PE-conjugated anti-human IGFIR-alpha antibody, on live primary cells derived from affected patients. (A) The geometric mean fluorescent intensity (MFI), Y-axis, emitted by labeled primary fibroblasts, untreated (black bars) and IGF-I treated (100 ng/ml, 24h; white bars), is presented relative to MFI of normal, untreated, C1 fibroblasts, which was arbitrarily assigned a value of 100%. Experiments were performed at least 3 independent times, each time in duplicates. P1: R877X; P2: 15q26.3 deletion (0.282 MB). (B) Representative collated emitted fluorescence (log scale Fluorescence Intensity, FI, X-axis) of live fibroblasts (Counts, Y-axis), untreated (black graphs) and IGF-I treated (red graphs) for C1, P1 and P2. Gray-shaded region, background fluorescence emitted by unlabeled and untreated fibroblast control. (C) Detection of cell surface IGFIR on live PBMCs from P2 (red) and parents (Mother, teal; Father, blue) compared with normal PBMCs (black). Gray-shaded region, background fluorescence of unlabeled and untreated PBMC control. The geometric mean of the fluorescent intensity, MFI, detected in normal PBMCs was given an arbitrary unit of 100%. (D) The MFI, Y-axis, of labeled PBMC in (C) upon IGF-I treatment (50 ng/mL, 1 hour), compared to untreated (black bars), performed 2 independent times, each time in duplicates. Normal PBMCs, untreated, was given an arbitrary unit of 100%. (E) Representative Western immunoblot of primary fibroblasts treated with IGF-I (100 ng/ml, 20 min).

The geometric mean fluorescent intensities (MFI) emitted by anti-IGFIR-PE-probed live dermal fibroblasts from P1 and P2, were compared to the MFI from our panel of normal (C1) and previously characterized IGFIR-deficient fibroblasts. Fibroblasts carrying the previously described IGF1Rc.3348_3366dup or IGF1R p.E121K/E234K mutations had basal MFI that were significantly reduced (46±9% SD and 37±4% SD, respectively) relative to untreated C1 (arbitrarily assigned a MFI of 100%; Figure 3A). The MFI of fibroblasts carrying the heterozygous IGFIR p.R877* variant (P1), or IGF1R CNV (P2) were similarly reduced at 39± 9% SD and 46±13% SD, respectively (Figure 3A), and comparable to our proven IGFIR haploinsufficient and hypomorphic cells. When C1 and IGFIR haploinsufficient fibroblasts (c.3348_3366dup, p.R877* or IGF1R CNV) were treated with IGF-I (100ng/ml, 24h), the emitted MFI were found to be reproducibly, and comparably, further reduced (Figure 3A and 3B). In contrast, compound heterozygous IGFIR p.E121K/E234K fibroblasts, demonstrated poor responsiveness to IGF-I (Figure 3A). Altogether, the results suggest that availability of cell-surface wild-type IGFIR significantly decreases after exposure to IGF-I, possibly due to ligand-receptor internalization, and this phenomenon was observed for normal as well as IGFIR haploinsufficient cells.

We had previously shown that a deficiency in IGFIR could be detected by FACS analysis of PBMCs (18), a biological sample much more readily accessible than primary dermal fibroblasts established from skin biopsies. Since we had access to whole blood samples from P2 and his parents, FACS analysis of isolated live PBMCs for cell-surface IGFIR was performed, to determine whether there might be variations in IGFIR expression, depending on cell type. Of note, the mother of P2 carries the same IGF1R CNV while the father was wild-type for the CNV. As shown in Figure 3C and 3D, the FI emitted by anti-IGFIR-PE-labeled PBMCs from P2 (52±6% SD) and mother (50±7% SD) were similarly and markedly reduced when compared to both normal PBMCs and PBMCs from the father (87±18% SD). For all PBMC samples analyzed, exposure to 50 ng/ml IGF-I for 1 hour, was sufficient to further reduce detectable cell surface IGFIR (Figure 3D).

IGF-I-induced IGFIR signaling

To confirm that IGFIR haploinsufficiency in P1 and P2 correlated with reduced IGFIR signaling, primary dermal fibroblast cultures were stimulated with 100 ng/mL of IGF-I for 20 minutes and phosphorylation of Thr308 on AKT, phospho-T³⁰⁸-AKT, assessed. Primary fibroblasts from both patients demonstrated decreased T³⁰⁸-AKT phosphorylation compared to C1 cells (Figure 3E).

DISCUSSION

The importance of having two intact alleles of the IGF1R gene for normal intrauterine and postnatal growth is supported by the recognition of IGF1R haploinsufficiency as causal of pre- and post-natal growth retardation (8). IGF1R haploinsufficiency can be the consequence of allelic loss of IGF1R due to chromosomal 15q26 deletions (first reported in 1991 (20); P2 and P3 in this report) or due to specific allelic IGF1R mutations that abrogate mRNA (11) or protein expression (P1 in this report). Here, we demonstrated that cell-surface IGFIR levels were reduced by at least 50% when live, primary cells, confirmed genetically for IGFIR haploinsufficiency, were analyzed by FACS, and this reduction correlated to reduced IGF-I-induced IGFIR signaling. This implies that bi-allelic protein expression is critical for maintaining normal cell surface IGFIR availability and function. These results also serve to highlight FACS analysis of primary cells derived from patients as an efficient method for evaluating IGFIR haploinsufficiency. Future studies will include verifying the uniformity of cell-surface IGF1R expression in the different sub-populations of normal PBMCs and whether expression might be influenced by age or gender.

Cell-surface IGFIR availability was indistinguishable amongst our IGFIR haploinsufficient patients, who shared features of SGA and poor postnatal growth despite normal to markedly elevated serum IGF-I concentrations. Other clinical and biochemical features, however, including responsiveness to rhGH therapy (8,21), were highly variable as has been previously reported (8,22). Both P1, carrying the heterozygous nonsense variant IGF1R p.R877*, and her father who carried the same variant, exhibited mild short stature (HtSDS of −2.06 and −1.93, respectively) but had no other apparent phenotypic anomalies. The p.R877* nonsense mutation abrogated C879-mediated disulfide formation necessary for a functional IGFIR α-ß monomer, which, together with a lack of detectable truncated p.R877* peptides (either cell-associated or secreted, data not shown), support their IGFIR haploinsufficiency state. Of note, the p.R877* variant (c.2629C>T, rs150221450), although not novel, is extremely rare as it was found in only 1 in 121408 IGF1R alleles sequenced (NCBI dbSNP and ExAC databases) with clinical significance unknown. Our results provide strong evidence that this variant is likely to be the underlying cause of the mild short stature observed in this family.

In contrast to P1, probands P2 and P3 exhibited pronounced short stature and other phenotypic anomalies. Both probands carry IGF1R CNV, although P3, who carried a significantly larger chromosomal 15q26 deletion (4.5 Mb), was the more severely affected of the two, with a phenotype consistent with Chromosome 15q26-qter deletion syndrome (MIM 612626) including cardiac defects, facial dysmorphism, auditory processing disorder, crowded dentition and ocular hypertelorism. P2, in addition to carrying a maternally inherited IGF1R CNV, was also diagnosed with NF-1, but his growth abnormalities were not typical of those more commonly associated with NF-1 (macrocrania, reduced pubertal growth spurt in boys, and in some, overweight status)(23). Interestingly, it remains unclear why the mother of P2, who was IGFIR haploinsufficient by both genetic and functional assessments, was of normal stature. Altogether, the diverse spectrum of presentations by P2 and P3 may be consequences of CNVs of adjacent genes/chromosomal regions in addition to the IGF1R gene (22,24–26).

The inability of targeted Sanger sequencing or WES variant analysis to detect gross allelic deletions of IGF1R illustrates the importance of including CNV analysis in the initial evaluation of children who present with pre- and post- natal growth failure in the presence of normal-to-elevated serum IGF-I concentrations suggestive of IGFIR insufficiency. Chromosomal IGF1R deletions can be readily identified by conventional methods, including karyotype (for large deletions), chromosomal microarrays and SNP genotyping arrays, as was performed for P2, although it is of note that the 0.282 Mb deletion was on the border of resolution for microarray analysis. Multiplex ligation probe amplification (MLPA) applications have also detected IGF1R CNV in affected patients (27,28). The limited resolution of all these methods, however, requires additional methodologies to delineate boundaries of CNV. For P3, we leveraged existing WES data to serve as a basis for CNV discovery, an application that provides much finer resolution (less than 1 kb) than conventional CNV methods. WES data generated from patient P3 and her parents had not identified candidate single nucleotide variations of significance, but through XHMM analysis (17), which is a relatively sensitive and specific algorithm for detecting rare, large, CNVs (29–31), 27 CNVs in P3 were uncovered. Of these, only one, a 4.492 Mb heterozygous chromosome 15 deletion that spans 28 genes, including the entire IGF1R gene, was unique to P3. Our finding of an IGFIR deletion in P3, therefore, supports the efficacy and efficiency of using exome sequencing for both variant and CNV detection, and is in-line with the proposed “exome-first” paradigm for clinical investigations of genetic defects (32). While successfully applied to our present case, additional algorithm refinement and testing will likely be necessary before these approaches can be widely applied clinically in the genetic workup for patients with rare diseases.

Finally, therapeutic options for patients with IGFIR haploinsufficiency are currently limited to rhGH as the majority of these patients are born SGA without catch-up growth, a clinical condition approved for rhGH therapy. While long-term treatment, starting at a young age, has been successful for improving stature for some patients (5,13,21), including for P1 who had mild short stature, overall responsiveness has been highly variable (8,21). P3, for example, improved height by only 1 SD with 3 years of rhGH therapy, reaching a height SDS of −4.1 at age 9.5 yrs, and no catch up growth was reported for the patient with IGFIR haploinsufficiency due to c.3348_3366dup19 mutation (11). P2 was not treated with rhGH because of concerns about potential impact on his NF-1 condition. The reason(s) for variable responsiveness to rhGH remains to be fully elucidated, with treatment tempered by concerns of extreme elevations of serum IGF-I concentrations, as was observed for P1 and P3. Continued monitoring of serum IGF-I and adjustment of dosage is highly recommended.

In summary, the molecular basis of IGF1R haploinsufficiency can be established by multiple methods, including utilizing genome-wide data for uncovering IGF1R CNV, as demonstrated in this report. Our application highlights how genome-wide data can serve the dual purpose of variant and CNV discovery. FACS analyses of live primary cells derived from IGFIR haploinsufficient patients supported the genetic findings and, moreover, provide valuable insights into cell-surface receptor availability critical for growth-promoting responses. Further investigations are necessary to delineate mechanisms of how seemingly comparable IGFIR availability can result in a wide spectrum of clinical phenotypes and variable responsiveness to rhGH therapy.

Established facts.

• Copy number variations (CNV) involving the insulin-like growth factor 1 receptor gene (IGF1R) result in varying degrees of pre- and postnatal growth restriction.

Novel Insights.

• CNV discovery based on whole exome sequencing (WES) expands the spectrum of information that can be derived from analysis of WES: out of 27 CNV uncovered, a 4.492 Mb de novo CNV encompassing IGF1R was identified.

• Allelic IGF1R haploinsufficiency translated to a reproducible ~50% decrease in cell surface IGFIR protein availability when primary cells from patients were analyzed by fluorescence-activated cell sorting (FACS), thus providing functional evidence for the critical importance of bi-allelic protein expression.

• FACS analysis of live primary cells is a promising method to efficiently evaluate and screen for IGFIR haploinsufficiency.