A patient with a novel homozygous missense mutation in FTO and concomitant nonsense mutation in CETP

Abstract

The fat mass and obesity associated gene (FTO) has previously been associated with a variety of diseases and conditions, notably obesity, acute coronary syndrome and metabolic syndrome. Reports describing mutations in FTO as well as FTO animal models have further demonstrated a role for FTO in the development of the brain and other organs. Here, we describe a patient born of consanguineous union who presented with microcephaly, developmental delay, behavioral abnormalities, dysmorphic facial features, hypotonia, and other various phenotypic abnormalities. Whole exome sequencing revealed a novel homozygous missense mutation in FTO and a nonsense mutation in the cholesteryl ester transfer protein (CETP). Exome CNV analysis revealed no disease causing large duplications or deletions within coding regions. Patient’s, her parents’ and non-related control’ fibroblasts were analyzed for morphologic defects, abnormal proliferation, apoptosis and transcriptome profile. We have shown that FTO is located in nucleus of cells from each tested samples. Western blot analysis demonstrated no changes in patient FTO. Q-PCR analysis revealed slightly decreased levels of FTO expression in patient cells compared to controls. No morphological or proliferation differences between the patient and control fibroblasts were observed. There is still much to be learned about the molecular mechanisms by which mutations in FTO contribute to such severe phenotypes.

Introduction

The fat mass and obesity associated (FTO) gene was the first gene locus reported to be associated with body weight and metabolic disorders ^1–5. Since then, a cluster of common variants in the first, and largest, intron of FTO have been described across multiple populations of different ethnicities ^6–9. Recently, Boissel et al. described a consanguineous Palestinian-Arab family with a homozygous non-synonymous FTO mutation that led to an inherited life-threatening disease in nine family members – this was the first report of a homozygous FTO mutation in the literature. The identified mutation resulted in an arginine to glutamine change at position 316 (R316Q) thereby rendering FTO catalytically inert ¹⁰. The affected family members suffered from postnatal growth retardation, head and face dysmorphisms, severe psychomotor delay, functional cognitive deficits, and, in some patients, brain malformations, cardiac defects, genital abnormalities, and cleft palates. In all affected individuals, death occurred within the first 30 months of life. In this report, we describe a patient with a novel homozygous missense mutation in FTO. We further discuss the phenotypic expression of this mutation.

Materials and Methods

Ethics Statement

The study protocol was approved by the Yale School of Medicine Human Investigation Committee (HIC) (protocol number 0908005592). Institutional review board approval for genetic and MRI studies and written consent from all study subjects, were obtained by the referring physicians at participating institutions.

DNA Extraction

Blood samples were collected from patients and their parents. DNA was extracted from the blood using the commercially available Gentra Puregene Blood Kit from Qiagen.

Whole Genome Genotyping

Samples were analyzed using 610K Quad Bead Chips according to the manufacturer’s protocol (Illumina).

Whole Exome Capture and Sequencing

Exome capture for the index case was performed using the NimbleGen 2.1M human exome array (Roche Nimblegen, Inc.) according to the manufacturer’s protocol along with modifications previously described in the literature ^{11, 81}. Exome library sequencing was performed using the HiSeq2000 with barcoding technology, paired end analysis, and six samples per lane. Image analysis and subsequent base calling was performed using the Illumina pipeline (version 1.8).

Exome Data Analysis

Analysis of the sequencing data was performed according to the previously described bioinformatics pipeline devised by our research team ⁸².

Sanger Sequencing

Coding regions and exon-intron boundaries of FTO were evaluated by Sanger sequencing using standard protocols. Amplicons were cycle sequenced on ABI 9800 Fast Thermo cyclers, and post cycle sequencing clean-up was carried out with the CleanSEQ System (Beckman Coulter Genomics). The amplicons were analyzed on 3730×L DNA Analyzer (Applied Biosystems Inc.).

Copy Number Variation (CNV) Analysis

The depth of coverage log ratio between the patient and control samples was calculated using the GATK-Depth of Coverage tool. Segments with CNVs were identified from the log ratio of the depth of coverage using the ExomeCNV R package ⁸³. False positive CNV events were identified and corrected for by calculating minor allele frequencies (BAF) in each CNV segment.

Co-Expression Analysis

Co-expression patterns were analyzed using the Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.7 ⁸⁴. Interrogation and procurement of results were performed using previously established protocols ^{12, 13}.

Skin Biopsy and Fibroblast Culture

Four millimeter skin punch biopsies were obtained from the umbilical area of the patient (NG1305-1), her parents (NG1305-2 and NG1305-3) as well as from control individuals using a standardized procedure ^{85, 86}. Samples were maintained in 50 ml conical tubes filled with Dulbecco’s Modified Eagle Medium (DMEM; Gibco, cat. no. 11965-084) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gibco, cat. no. 10438-026), 1% (1x) L-glutamine (Gibco, cat. no. 25030-081), and 2% (1x) Penicillin-Streptomycin (Gibco, cat. No. 15140-122), and subsequently transported to the laboratory for culture. Once they arrived in the laboratory, samples were washed at least three times in PBS (Sigma Chemical Co., Saint Louis, USA). Samples were cut into small fragments and placed in 100 mm² Petri dishes, which were maintained for 30 minutes semi-opened in laminar flow to allow specimens to adhere to the dish surface. Afterwards, 10 ml of culture medium containing DMEM supplemented with 10% heat-inactivated fetal bovine serum, 1% (1x) L-glutamine, 1% (1x) Penicillin-Streptomycin at 37°C, was added to the dish. Cultures were maintained in a humidified incubator at 37°C, under 5% CO2 in air. The culture medium replaced initially after two days and, subsequently changed three times a week. Fibroblasts began to emerge from the biopsy fragments within in 7–9 days. Subculture (passage) of fibroblasts was performed at 70% cellular confluence.

Once there were sufficient cells, enzymatic detachment was performed using 0.25% Trypsin-EDTA (1X) (Gibco, cat. no. 25200-056) and cells were plated in 25 cm² culture flasks for additional proliferation. Primary fibroblasts were grown at 37°C in 5% CO2 and medium was changed every two days. Cells were passaged using trypsin at approximately 70% confluence and either harvested for cryopreservation or reseeded for further proliferation. Early passages of fibroblasts were used for cell-based functional assays.

Population Doubling Assay

Cells were trypsinized upon confluency and 2000 cells per well were seeded in 135 microliters of medium. There were nine wells per individual (NG1305-1, NG1305-2, NG1305-3, CTRL-1, CTRL-2 and CTRL-3) included on a 96 well-plate. A separate plate of cells was also included using serial dilutions of cells for calibration curve. After 2^nd, 5^th, 7^th, 9^th and 11^th days of incubation, we performed a CellTiter-Glo® Luminescent Cell Viability Assay (Promega). Luminescence was read using the GloMax®-Multi Detection System (Promega) and data was analyzed with Graphpad Prism software.

Fibroblast Morphology and Apoptosis Assay

Fibroblasts were seeded in 24-well plates and images were obtained on the 2^nd, 5^th, 7^th, 9^th and 11^th days using an inverted microscobe at 20× magnification. Apoptosis was evaluated with an Annexin V-FITC and 7-amino-actinomysin (7-AAD) dye kit (Ebioscience) according to manufacturer’s protocol. Briefly, cells were harvested and washed in 1X PBS, and then resuspended in 1X Binding Buffer at 1–5×10⁶ cells/mL. 5 μL of fluorochrome-conjugated Annexin V was added to 100 μL of the cell suspension, which was then incubated 10–15 minutes at room temperature and protected from light. Cells were subsequently washed with 2 mL of 1X Binding Buffer and resuspended in 200 μL of 1X Binding Buffer. 5 μL of 7-AAD viability staining solution was added to the solution, which was stored at 2–8°C in the dark, and analyzed within 4 hours using flow cytometry.

Immunohistochemistry and Immunostaining

For immunohistochemistry of fibroblasts from extracted skin biopsies of index case, her parents and control sample were fixed in 3.6% formaldyhyde and washed three times with PBS. Then cells were exposed to 0.3% Triton X-100 (Sigma Chemical) in phosphate buffered saline for 1 minutes and then washed two times with PBS. Cells were blocked in blocking solution containing 1X phosphate buffered saline (PBS), 5% normal bovine serum, 5% normal goat serum, 5% normal donkey serum and 0.1% Tween for half an hour at room temperature. After blocking, specimen were extensively washed in PBS and incubated with primary antibodies diluted in blocking solution on a horizontal shaker overnight at 4°C, and washed in PBS at room temperature for three times. Cells were then incubated with secondary antibodies in blocking solution for 1 hours and washed in PBS for three times at room temperature, then washed and imaged with Zeiss microscopy system. Primary antibody used for immunohistochemistry are: a custom rabbit anti–recombinant mFTO antibody (1:100 dilution). Alexa Fluor®488 Donkey Anti-Rabbit IgG (H+L) Antibody (A-21206, Life technologies) was used as secondary antibody.

Western Blot Analysis

Protein expression was analyzed by Western blot analysis as described previously ⁸⁷. Proteins were extracted from fibroblasts of patient, her father and mother and control subject, then protein blots were performed on 40 μg of total proteins. The primary antibodies used were as follows: a custom rabbit anti–recombinant mFTO antibody (1:100 dilution), anti-glyceraldehyde 3-phosphate dehydrogenase (1:100 dilution, Rabbit, catalog no.: sc-25778, Santa Cruz Biotechnology, Santa Cruz, CA). The horseradish peroxidase (HRP)-conjugated secondary antibody used was as AffiniPure goat-anti-rabbit IgG FC-fragment was used (Jackson, West Grove, PA #111-035-046).

RNA Extraction, cDNA Synthesis and Quantitative PCR analysis

FTO mRNA expression in affected individual, both parents and control sample (individual of the same ethnicity and without a FTO sequence variant is used as control) was assessed by real-time-PCR (RT-PCR). Total RNA was isolated using miRNeasy Mini Kit (Qiagen; catalog no.: 24104) from fibroblasts according to the manufacturer’s instructions and transcribed into complementary DNA (cDNA) as described previously (Applied Biosystems; catalog no.:4374966). RNA quantity was assessed using a NanoDrop spectrophotometer for cDNA synthesis. Quantitative PCR (qPCR) analysis was performed using FastStart Universal SYBR Green Master Mix (Roche; catalog no.: 04913914001). Two different FTO and the reference gene TBP (TATA box-binding protein) primers were used and the PCR efficiency of >90% (slope= −3.2 and −3.6) and R² 99% obtained. Relative changes in gene expression were analyzed with the ΔCT method ⁸⁸.

Transcriptome Analysis

To further characterize the effects of the FTO mutation in the family, RNA was extracted from these fibroblast cultures and RNA expression analysis using Illumina HumanHT12.v4 chips was performed on patient, father and mother samples and analyzed using the DAVID platform. Briefly, data is normalized using normal-exponential convolution model-based background correction and quantile normalization using the limma R package. The normalized data for 3 samples were used to perform an unsupervised hierarchical clustering using the euclidean distance as the dissimilarity metric and the average agglomerative method for clustering.

Results

Clinical Report

We report the case of a five year-old female (NG1305-1) presented with a dysmorphic face and developmental delay. She was the first child born of a consanguineous union (parents were second cousins) (Fig. 1). She was born at 40 weeks gestation with 2460 g (3^rd percentile) and 50 cm long (50^th percentile). Her head circumference at birth was not recorded. She was immediately admitted to the neonatal intensive care unit after birth due to respiratory distress and concern given her dysmorphic facial features. During her hospitalization she began to have seizures that were managed with antiepileptic medications. She was subsequently discharged, but was later admitted at nine and a half months-old. At this time, her weight was 7650 g (10^th percentile), height was 66 cm (3^rd percentile), and head circumference was 41 cm (<2^nd percentile). On examination, she was hypertonic and lacked the ability to control her head movements. She was microcephalic, had sparse hair and eyebrows, a prominent metopic ridge, an asymmetric skull, esotropia of her right eye, hypotelorism, a long philtrum, microretrognathia, a high palate, a prominent alveolar ridge, anteverted nostrils, a thin vermillion and lips, and a short neck (Fig. 1B). She was noted to have hepatosplenomegaly and diastasis recti. Laboratory workup revealed high AST (70 U/L), ALT (57 U/L), and CK (1191 U/L) levels. Eye examination demonstrated bilateral nystagmus, iris nodules, posterior synechia, and astigmatism. Assessment of auditory brainstem responses demonstrated bilateral conductive and sensorineural hearing loss. At 19 months of age the patient’s weight was 11 kg (<3^rd percentile), height was 83.5 cm (<3^rd percentile), and head circumference was 45cm (<3^rd percentile). In addition to the previously noted physical exam findings, she was now also found to have difficulty swallowing. An echocardiogram was performed and demonstrated normal cardiac anatomy and function. Upper endoscopy revealed grade 1 esophageal varices. A head CT demonstrated prematurely closed metopic sutures (craniosynostosis). A brain MRI revealed a thin corpus callosum. Chromosomal analysis demonstrated a normal female karyotype. In addition to the physical exam abnormalities, the patient also demonstrated developmental delay with respect to reaching milestones – she did not develop head control until after 10 months of age, began sitting without support at 2 years-old, and walking at 4 years-old. Throughout this time, her AST, ALT and CK all remained abnormally elevated. A liver biopsy was performed but did not show any abnormalities; EMG was unremarkable. At 3 years-old, Denver testing confirmed severe developmental delay and the Wechsler Intelligence Scale for Children-Revised showed profound intellectual disability (IQ:23). At 4 years and 9 months, the patient’s weight was 15 kg (10^th percentile), height was 97cm (3^rd-10^th percentile), and head circumference was 46cm (<3^rd percentile). She was microcephalic, could walk with support, and spoke a few words. The patient also continued to have bilateral nystagmus, brachydactyly and hepatosplenomegaly. Labs demonstrated a high CK (816 U/L, normal <170), normal total cholesterol (173 mg/dl, normal 114–200), low LDL cholesterol (30 mg/dl, normal 63–130), and high HDL cholesterol (136 mg/dl, normal 30–75) (S1 Table).

Figure 1. I Phenotypic and molecular studies has been performed on materials obtained from index patient, parents and control samples.

A. Pedigree of the family. B. Clinical pictures of the patient (upper right panels) when she was 9 month-old (upper panels), 2 and half years old (left lower panel) and 4 and nine month-old (right lower panel). C and D. Representative sequence alignment figures cover the mutations in FTO and CETP, respectively. The top line in each panel represents the non-mutated reference sequence. The subsequent lines below the reference lines depict the results from exome sequence. Each line represents a distinct coverage read. Mean 20x coverage of all bases was above 81% for index patient. E and F. Chromatogram illustrations of FTO and CETP obtained via Sanger sequencing analysis of the index patient analyzed via whole-exome sequencing (NG1305-1) and her parents. Note that the respective mutations identified via whole-exome sequencing were confirmed as homozygous mutations in the index patient and as heterozygous in her parents. DNA from healthy individuals was also Sanger sequenced and these results are included as controls. G. CNV Segment Detection of NG1305-1: The log ratio comparing NG1305-1 and control sequence depths of coverage for each exon are depicted as gray dots. The black lines demonstrate regions of segmented copy neutral events, green lines are segmented deletion events and red lines are amplification events. H. mRNA levels of FTO in fibroblast cells were extracted from homozygous patient, heterozygous parents and control individual-same ethnicity and without a sequence variant- were analyzed using RT-PCR. I. The figure depicts FTO staining in patient fibroblasts. Significant nuclear accumulation is noted. J. Western blot of FTO and as an internal control GAPDH in patient, parental and control cells. No difference was detected.

She was last time evaluated 5.5 years of age. She developed seizures at 5 years old and electroencephalogram demonstrated that high amplitude and irregular waves and spikes being more prominent on the right side of posterior temporal occipital region. After seizures she lost the ability to assisted walk assisted and spoken a few words. Six months after seizure when she was 5 years, 6 months old, she gained the ability of walk with support, however speech never came back. Her weight was 19kg (25–50^th percentile), height was 100.5cm (3^rd–10^th percentile), and head circumference was 47cm (<3^rd percentile). She had also suck/swallowing disorders, frequent drooling and protruding tongue. Physical exam was also remarkable for enlargement of both her liver and spleen (5 cm and 4 cm below costal margin, respectively). Repeated blood tests demonstrated a normal AST (29 U/L) and ALT (29 U/L) levels, high CK (587 U/L, normal <174), normal total cholesterol (153 mg/dl, normal 114–200), and high HDL cholesterol (127 mg/dl, normal 30–75). In addition, she demonstrated abnormal behavior phenotype such as repetitive hand movements, frequent laughter/smiling, apparent happy demeanor and short attention span. Evoked response audiometry (ERA) revealed that moderate hearing loss at right ear and mild hearing loss at left ear. Intriguingly, her father and mother had obesity, both weight were 100 kg. The father’s height was 169 cm and the mother’s height was 161 cm, father’s head circumference was 58 cm, and mother’s was 55 cm. Blood biochemistry analysis for parents of the patient revealed high total cholesterol levels (father 264 mg/dl, and mother 240 mg/dl), and high LDL cholesterol levels (father 169 mg/dl, mother 142.2 mg/dl) in both parents, however in father triglyceride levels also was found to be high (180 mg/dl). We also have found high HgA1c level (6.4) in father which indicates increased risk for type 2 diabetes. Both parents’ blood HDL, glucose, cortisol and insulin levels were found within normal limits. Blood CPK level was normal for mother however was found to be high in father (438 mcg/L) which may be found high levels in different situations such as heart attack, myaocarditis, myopathy, hypo/hyperthyroidism etc. where problems occured normally found organs in where is mainly in the heart, brain, and skeletal muscle.

Whole Genome Genotyping and Whole Exome Sequencing

Whole genome genotyping was performed and the inbreeding coefficient for the patient (NG1305-1) was determined to be 0.01. Homozygous genomic segments (>2.5 cM each) were identified (S2 Table) and whole exome capture and sequencing of the patient’s germ line DNA (S3 Table) was used to search for disease causing mutations within these regions ¹¹.

Variant analysis identified only 2 homozygous novel mutations located within the aforementioned regions of homozygosity (S4 Table) (Fig. 1). The first was a homozygous missense mutation (ENST00000471389.1:c.812A>C) within the FTO (fat mass and obesity associated) coding sequence at position 53,878,127 on chromosome 16. This mutation resulted in a histidine to proline change at position 271 (ENSP00000418823.1:p.His271Pro) in the amino acid sequence of the FTO protein. This FTO mutation has not been previously reported in the dbSNP, NHLBI GO ESP Exome Variant Server, or 1000 Genomes databases, nor has it been observed within the cohort of 3,000 subjects with non-neurological diseases who have been whole-exome sequenced at the Yale School of Medicine. When we interrogated these databases for the mutational burden of FTO, we found no instances of homozygous deleterious mutations in FTO.

The other detected mutation (ENST00000566128.1:c.1207C>T) was in the CETP (cholesteryl ester transfer protein, plasma) gene at position 57,017,318 on chromosome 16, resulting in an early protein truncation and likely subsequent nonsense mediated decay (ENSP00000456276.1:p.Arg403X).

Sanger sequencing of the entire coding region of FTO and CETP in the patient’s parents revealed that both were heterozygous for the identified mutations (Fig. 1). The parents were then reexamined for metabolic findings previously reported in association with FTO mutations but both were healthy and had no evidence of abnormal metabolic functioning. The patient’s exome data was analyzed for large scale CNV events, and no disease causing large duplications or deletions within coding regions were identified (Fig. 1).

Co-Expression Data Analysis

Using the Human Brain Transcriptome database ¹², we investigated the spatial and temporal changes in the FTO expression during human cortical development. FTO mRNA is expressed throughout the entire brain with marked expression in the fetal cortex and cerebellum. This expression remains robust in the adult brain (S1 Fig.) ¹².

We next performed co-expression analysis on FTO ^{12, 13} (S5 and S6 Tables). We observed that FTO expression patterns positively correlate with the expression patterns of genes involved in the ubiquitin mediated proteolysis pathway such as HUWE1, UBE3B, and others. We also found that FTO expression patterns positively correlate with genes in which mutations cause either neurodevelopmental disorders such as MECP2 (Rett syndrome, MIM:312750) and GDI1 (Mental retardation, X-linked 41, MIM:300849), or hearing impairment such as HARS (Usher syndrome type 3B, MIM:614504).

Q-PCR, Western Blot, and Immunohistochemistry

We have shown that FTO is located in nucleus of cells from the patient, her parents, and non-related controls (Fig. 1). Western blot analysis demonstrated no changes in patient FTO, as expected given the observed mutation. Q-PCR analysis revealed slightly decreased levels of FTO expression in patient cells compared to controls (Fig. 1).

Transcriptome Analysis

Interestingly, we found that type 2 diabetes associated genes (insulin-like growth factor binding protein 2, 36kDa and neurocanthocytosis), retinol metabolism genes (ALDH1A1, DHRS3 and RDH10), renin angiotensin system genes (AGT and MME), and genes related to metabolism of xenobiotics by cytochrome P450 (ALDH1A3, GSTT1 and MGST1) demonstrated at least a twofold increase in relative expression in the patient’s fibroblasts compared to her parents’ and controls’ fibroblasts. Conversely, cell cycle genes (E2F2, BUB1, CDC20 and CCNB1), and obesity susceptibility genes (including ADRB2, ENPP1) demonstrated a relative decrease level of expression of at least 2-fold in the patients cells relative to her parents’ and controls’ samples.

Fibroblast Morphology, Population doubling, and Apoptosis Assays

Fibroblast morphology was compared among the patient (NG1305-1), parent (NG1305-2 and NG1305-3), and control cultures at 11 days incubation. No morphological differences between the patient and control fibroblasts were observed (Fig. 2). The rates of apoptosis were assessed at five and nine days incubation; again, no differences were appreciated between patient and control fibroblasts (Fig. 3). In addition, fibroblasts were counted on the 2^nd, 5^th, 7^th, 9^th and 11^th days and a growth curve was drawn (number of cells versus time) in order to calculate the population doublings (PD) from the exponential phase of the curve.

Figure 2. I Phenotypic and proliferation comparisons of fibroblasts extracted from index case, parents and control subjects.

Upper panel shows comparison of fibroblast morphology of the patient, her parents and control fibroblasts. Images were taken after 2^nd, 5^th, 7^th, 9^th and 11^thdays incubation under 20x objectives. Lower panel compares number of cells versus time in each study subject and there is no statistically difference between the patient and control samples.

Figure 3. I The flow cytometric analysis of apoptosis in fibroblast cells after three days incubation using FITC-annexin V and 7AAD double staining.

Quadrant analysis of the gated cells in FL-1 versus FL-3 channels was from 1,000 events. Annexin V+/7AAD− (lower left quadrant) areas stand for early apoptotic cells, and Annexin V+/7AAD+ (upper right quadrant) areas stand for late apoptotic or necrotic cells.

The PD, which occurred between the 5^th and 11^th days, was obtained using the equation: PD = (logN1/log2) - (logN0/log2), where N1 was the final and N0 was the initial cell count. There were no differences in proliferation between patient and control fibroblasts (Fig. 2, S7 Table).

Discussion

The FTO gene encompasses a large genomic region whose nine exons span more than 400 kb on chromosome 16q12.2 and encodes 505 amino acids FTO protein composed of two domains: an N-terminal domain carrying a catalytic core and a C-terminal domain of unknown function ¹⁴. In our patient, as well as in the previously reported case, the homozygous FTO mutation affected the catalytic domain of the protein. Using both bioinformatic and in vitro biochemical data, FTO was predicted and confirmed to be an AlkB family DNA/RNA demethylase ^15–17. It has been known that the FTO protein localizes to the nucleus ^{15, 18, 19}. However, recently Gulati et al. demonstrated that a fraction of FTO located in the cytoplasm and N-terminus of FTO is necessary for its ability to shuttle between the nucleus and cytoplasm ^{20, 21}. Our IHC data demonstrating that FTO was localized to the nucleus of fibroblasts are consistent with this observation.

FTO expression begins early embryogenesis when it is ubiquitously expressed throughout the body, but it most highly expressed in the brain, which is consistent with the observation that multiple organ systems are affected by FTO deficiency ^{10, 15, 22}. Similar to the FTO phenotype previously reported by Boissel et al., our patient also demonstrated intrauterine growth retardation, failure to thrive, hypertonicity, seizures, severe developmental delay, postnatal microbrachycephaly, dysmorphic craniofacium, sensory-neural hearing loss, and optic disc abnormality, a short neck, cutis marmorata, a PDA, drumstick fingers, brachydactyly. Interestingly, not all of the affected family members demonstrated all of these traits and our case’s clinical findings are less aggressive than Boissel et al cases’ especially the ones who dies due to recurrent infections. Our patient additionally had abnormal behavior phenotype including repetitive hand movements, frequent laughter/smiling, apparent happy demeanor and short attention span as well as other clinical findings such as hip dislocation, strabismus, a trigonocephaly, thin upper and lower lips, a long philtrum, and osteopenia.

Because of its well-known association with obesity, FTO is well-studied ^23–49. Previous studies, have shown that even though some are predicted to have deleterious effects on FTO function, nonsynonymous mutations were equally common in both the obese and lean cohorts ^50–52. Other studies in lean and obese cohorts of children brought similar findings. In both African American and Chinese Han populations, variants were identified in FTO, but the overall frequencies were similar in case and control, with none conferring risk of obesity ^{53, 54}.

Further, Boissel et al. reported none of the parents were appeared to be obese and detailed phenotypic data on the extended family has not been reported, in our case both of the parents of the affected children were obese, interestingly based on international diabetes foundation’s (IDF) metabolic syndrome criteria, patients’ father diagnosed as a metabolic syndrome due to his increased wrist circumference and blood pressure and high trigliceride level. Since patient’s mother is also carrier for same FTO gene variant, father’s metabolic syndrome may be caused by either environmental factors and/or sex selection for the metabolic result of FTO variant.

Variants in FTO and its association with various human phenotypes continue to emerge ^{6, 7, 9, 11, 55–63}. Intriguingly, in humans with loss-of-function mutations in FTO share phenotypic features of cases with partial trisomies of chromosome 16q including the FTO gene such as learning difficulties and behavioral problems ^{64, 65}.

Boissel et al. demonstrated changed morphology, impaired proliferation and accelerated senescence in cultured skin fibroblasts from affected subjects. Although we did not test senescence, we were not able to detect any abnormalities in either morphology or proliferation in the patient fibroblast cultures compared to either parent or control fibroblasts. Given that we did not detect any defects in proliferation, we also tested for variability in the rates of apoptosis, and again did not see any statistically significant differences between the patient and parental.

Our patient was found to have a novel homozygous nonsense mutation in CETP, resulting in a truncated 403 amino acid protein (compared to the normal 476 amino acid protein), and likely causing a complete CETP deficiency. Human CETP is a 74kDa plasma glycoprotein that facilitates the transfer of cholesteryl esters between lipoproteins ^{68, 69}. Genetic deficiency of cholesteryl ester transfer protein (CETP) (OMIM 607322) is well known cause of primary, also known as familial, hyperalphalipoproteinemia (HALP), and results in elevated plasma HDL (2–5× increase in the homozygous and 25–80% greater in the heterozygous genotype) and decreased LDL ^70–72 and it is mostly detected in the Japanese population more than Caucasians ^72–74. In Caucasians only a few cases of HALP due to CETP mutations have been reported ^75–79. Despite the high HDL-C levels in patients with CETP deficiency, an increased risk of atherosclerotic coronary artery disease (CAD) has been reported ⁸⁰. Our patient’s mutation in CETP was novel and laboratory findings fit with the genetic diagnosis.

Boissel et al. were the first to report its homozygous loss of function mutation as an example of a human disorder related to a defect in an AlkB-related protein. We describe the second novel homozygous missense mutation in FTO to be reported in the literature, in addition to a nonsense mutation in CETP. We further discuss some of the phenotypic consequences of such an FTO mutation. However based on studies conducted in human and other organisms fully functional FTO is critical for normal physiology and there is still much to be learned about the molecular mechanisms by which a mutation FTO leads to such a severe phenotype, as observed in our and previously reported patients.

Supplementary Material

1. S1 Table.

Plasma lipid and protein levels of the NG1305 were correlated with CETF mutation that was found patient by whole exome sequencing.

2. S2 Table.

Blocks of genomic homozygosity were identified via whole-genome genotyping obtained from index patient (NG1305-1). Chr – chromosome; SNP – single nucleotide polymorphism; cM – centimorgans; Mb – million base pairs.

3. S3 Table.

Summary of sequencing quality metrics achieved by whole-exome sequencing performed on index patient (NG1305-1).

4. S4 Table.

Summary of whole-genome genotyping and whole-exome sequencing results for index patient NG1305-1. Blocks of genomic homozygosity were identified via whole-genome genotyping. Homozygous variants were identified via whole-exome sequencing.

5. S5 Table.

List of the top 100 genes with expression patterns that are positively correlated with FTO expression.

6. S6 Table.

List of the top 100 genes with expression patterns that are negatively correlated with FTO expression.

7. S7 Table.

Comparison of doubling time between the patient, parents and control samples in different incubation time points revealed no statistical differences.

8. S1 Fig.

Graph of FTO expression in various regions of the normal developing human brain over time. These data were obtained using the Human Brain Transcriptome database. Note that FTO is expressed throughout the all brain regions. Expression levels, except cerebellum, peak early in development, decrease slightly during fetal life, and then remain relatively constant throughout adult life. In cerebellum, expression levels of FTO decrease slightly during fetal life and remain high after birth and adult life.