SEARCH

SEARCH BY CITATION

Keywords:

  • Familial hypercholesterolemia (FH);
  • low-density lipoprotein receptor (LDLR);
  • in silico;
  • novel variants;
  • genetics

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information

Familial hypercholesterolemia (FH) is an autosomal dominant disease with a frequency of 1:500 in its heterozygous form. To date, mutations in the low-density lipoprotein receptor gene (LDLR) are the only identified causes of FH in the Greek population, causing high levels of low-density lipoprotein (LDL) and total cholesterol and premature atherosclerosis. The Greek FH population is genetically homogeneous, but most previous studies screened for the most common mutations only. The study aimed to characterize and assess novel LDLR variants. LDLR was examined by whole-gene DNA sequencing in 561 FH patients from 262 families of Greek origin. Novel LDLR variants were analyzed in silico using various software predicting pathogenicity and changes in protein stability. Twelve novel LDLR variants were identified, six of which are putative disease-causing variants: c.977C>G in exon 7, c.1124A>C in exon 8, c.1381G>T in exon 10, c.628_643dup{636del}, c.661–673dup in exon 4, and 13 c.1987+1_+33del in intron 13. All six putative variants were confirmed in the hypercholesterolemic members of the family. The results show that in silico analysis is a valuable tool to predict potential pathogenicity of novel variants, especially for populations that have not been extensively studied. The identification of novel pathogenic variants will facilitate the molecular diagnosis of FH from early childhood.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information

Familial hypercholesterolemia (FH) is a common autosomal dominant disease with a frequency of 1:500 individuals in its heterozygous form (Soutar & Naoumova 2007). FH patients have high levels of total cholesterol (T-C) and low-density lipoprotein cholesterol (LDL-C) in the plasma and usually present with arcus corneae, xanthelasmas, xanthomas, and premature atherosclerosis, even from childhood (Hobbs et al., 1992). Mutations in the low-density lipoprotein receptor (LDLR) (2), apolipoprotein B-100 (apoB-100) (Innerarity et al., 1990), and proprotein convertase subtilisin/kexin type 9 (PCSK9) (Abifadel et al., 2003) have been found to cause FH.

The clinical phenotype of FH patients is variable. There are three sets of criteria used for the clinical diagnosis of FH. These are the Simon Broome Register (UK), the Dutch Lipid Clinic Network (Netherlands), and the MEDPED Program (USA) criteria. However, although these criteria have been proven to be useful in identifying adult FH patients, they rely mainly on detection of clinical signs of heterozygous FH (xanthelasma, tendinous xanthomata, and corneal arcus) or manifestations of coronary heart disease, which occur in adulthood. Such clinical signs are rarely present in affected children or even in their young parents and therefore the clinical diagnosis may miss a considerable proportion of FH pediatric patients in whom the phenotype is not apparent yet. In addition, a proportion of children at high risk for FH, even carriers of mutations causing FH, may initially present lipid levels within the normal range and later develop hypercholesterolemia consistent with the diagnosis of FH (Kessling et al., 1990). Therefore, lipid levels within the normal range for childhood do not necessarily exclude FH. The presence of elevated lipid levels in a member of the family and a positive history of cardiovascular disease are suggestive of FH in children. It is also important to note that children in families with hypercholesterolemic parents are most likely to follow a particular low-fat diet, indicating that borderline lipid measurements in these children cannot exclude FH.

Since the clinical phenotype of FH patients is variable, molecular characterization of a gene defect is necessary for an indisputable diagnosis, particularly in children for which the clinical diagnosis can be difficult. Therefore, in cases in which novel variants are identified, distinguishing between pathogenic mutations and polymorphisms is crucial for molecular diagnosis, especially for early diagnosis. Although in vitro studies are a good approach to investigate pathogenicity of novel variants they are often laborious to perform, especially when numerous novel variants are identified in populations that have not been extensively studied. As an alternative, bioinformatics methods can be used to obtain useful information about the putative effects of novel variants on protein function and stability (Thusberg & Vihinen, 2009).

The LDLR gene, located on chromosome 19p13.1–3, consists of 18 exons spanning 45kb and codes for a precursor protein of 860 amino acids (Hobbs et al., 1992). The spectrum of LDLR mutations varies between populations with 1288 mutations reported worldwide (Leigh et al., 2008). To date, only mutations in the LDLR gene have been found to cause FH in patients of Greek origin. Previous studies identified a total of 27 LDLR mutations in Greek FH patients (Hobbs et al., 1992; Webb et al., 1996; Kotze et al., 1997, Mavroidis et al., 1997; Traeger-Synodinos et al., 1998; Xenophontos et al., 2000; Miltiadous et al., 2001; Dedoussis et al., 2004a, b, 2006; Laios & Drogari, 2006; Glynou et al., 2008). However, most of these studies screened for the most common mutations, using mainly methods that can identify known mutations. No mutations have been found in apoB-100 (Dedoussis et al., 2004a, b) or PCSK9 (Diakou et al., 2011) in FH patients of Greek origin, suggesting that the Greek population is genetically homogeneous or that there are other mutations that have not been identified yet. The present study specifically aimed to characterize the newly detected LDLR variants in Greek FH children and their families and assess their pathogenicity by in silico analysis. We anticipate that the present study will facilitate the molecular diagnosis of FH in Greece.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information

Subjects

In total, 561 FH patients from 262 families were studied. These included 262 unrelated index cases who were children and adolescents aged 2–17 years and 299 relatives. They were all clinically and biochemically diagnosed as heterozygous FH. The index cases were referred to the Children's Hospital “Aghia Sophia,” Athens (from 2007 to 2010). All the subjects were of Greek origin. The clinical diagnosis of heterozygous FH in the index cases was based on a uniform protocol including: (a) the presence of primary hypercholesterolemia T-C ≥ 200 mg/dl and plasma or serum LDL-C ≥ 120 mg/dl in the index case and (b) the presence of primary hypercholesterolemia T-C ≥ 200 mg/dl and plasma or serum LDL-C ≥ 120 mg/dl in the probands’ siblings, parents, grandparents, and great grandparents with an autosomal dominant mode of inheritance of hypercholesterolemia in the family or (c) family history of coronary artery disease at <55 years for men and <60 years for women in a first degree relative or (d) the presence of tendon and cutaneous xanthomas in the probands at an early age with hypercholesterolemia or in their parents with hypercholesterolemia. Secondary causes of hypercholesterolemia were excluded in the probands. The above criteria are based on careful examination of lipid levels as well as examination of the family history of cardiovascular disease and vertical transmission of hypercholesterolemia in three generations.

Informed consent was obtained from all the study participants. The molecular diagnosis protocol for FH was approved by the Scientific Committee and the Medical Ethics Committee of the Children's Hospital “Aghia Sophia.”

Molecular Analysis

Genomic DNA was isolated from leucocytes using the QIAGEN QIAMP DNA Blood Kit according to the manufacturer´s protocol (Qiagen Hamburg, Germany). The reference sequence NM_000527.3 was used for LDLR and the novel variants were numbered according to the Human Genetic Variation Society guidelines, where “+1” is the A of the ATG translation initiation codon of the coding DNA (den Dunnen & Antonarakis, 2000). LDLR was amplified by polymerase chain reaction (PCR) including the 18 exons, the proximal promoter region and the intron-exon boundaries (Table S1). In all index cases, the mutation hot spot regions of LDLR were sequenced first, and if no mutation was found, whole-gene sequencing was performed. In cases where no known LDLR mutation was detected in the index cases, the whole gene was also sequenced in their relatives. For index cases found to carry a previously known mutation, their relatives were directly sequenced in the region containing the mutation. If no mutation was identified in this region in their relatives, the whole gene was subsequently sequenced. PCR amplicons were purified using the NucleoFast 96 PCR plate (Macherey-Nagel GmbH, Düren, Germany) and sequenced using the BigDye Terminator version 3.1 Cycle sequencing kit and the 3730 DNA Analyzer (Applied Biosystems, Foster City, CA, USA) with the primers described in Supplementary Table S1. All novel variants were confirmed with sequencing in both orientations.

In Silico Analysis

The effect of amino acid substitutions on the mature LDLR peptide and the protein stability was assessed by 14 methods predicting pathogenicity and stability changes: PolyPhen-2 (Ramensky et al., 2002), SIFT (Ng & Henikoff 2003), PANTHER (Mi et al., 2007), Mutation Taster (Schwarz et al., 2010), Condel (González-Pérez & López-Bigas, 2011), PMut (Ferrer-Costa et al., 2005), I-Mutant (Capriotti et al., 2005), SNPs3D (Yue & Moult, 2006), Eris (Yin et al., 2007), nsSNPAnalyzer (Bao et al., 2005), mutation assessor (Reva et al., 2011), Mupro (Cheng et al., 2006), PhD-SNP (Desmet et al., 2009), and PopMuSiC (Dehouck et al., 2009). DiANNA was used to predict disulfide connectivity for novel variants introducing cysteine residues (Ferrè & Clote, 2005).

HOPE was used to analyze the effect of novel variants on the protein structure (Venselaar et al., 2010), and the structural information was obtained from the analysis of the Protein Data Bank file 3M0C (The x-ray crystal structure of PCSK9 in complex with the LDL receptor).

Species conservation was checked by comparing the human LDLR amino acid sequence with LDLR protein sequences from 12 other species (Zebrafish-ENSDARG00000029476, Xenopus-ENSXETG00000009352, Fugu-ENSTRUG0000-0002369, Chicken-ENSGALG00000004944, Cow-ENS-BTAG00000012314, Guinea pig-ENSCPOG0000000681-6, Dog-ENSCAFG00000017539, Rat-ENSRNOG0000-0009946, Rabbit-ENSOCUG0000000065, Mouse-ENS-MUSG00000032193, Macaque-ENSMMUG0000000361-1, Chimpanzee-ENSPTRG000000104-90), using CLU-STAL W (Thompson et al., 1994).

The Berkeley Drosophila Genome Project (BDGP) (Reese 2001) and the NetGene2 (Brunak et al., 1991) algorithms were used to predict potential splicing effects.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information

Clinical Phenotype of Index Cases

The biochemical characteristics of the index cases carrying novel variants are described in Table 1. Table 1 also shows whether the novel variant was found in the hypercholesterolemic 1st degree relatives that were examined. With the exception of an adopted index case, all index cases carrying a novel variant had at least one hypercholesterolemic parent (T-C ≥ 200 mg/dl), and most of them had at least one hypercholesterolemic grandparent. None of the index cases carrying a novel variant or their first degree relatives showed tendon and cutaneous xanthomas. Details of the family history of the index cases are described in Table S2.

Table 1. Biochemical characteristics of the index cases carrying the novel variants and carriage in relatives
Index case(s)1SexAge2T-CLDL-CHDL-CTGNovel variantRelatives analyzed3
  1. Plasma lipid levels are untreated levels (mg/dl); T-C, total cholesterol; LDL-C, low density lipoprotein cholesterol; HDL-C, high density lipoprotein cholesterol; TG, triglycerides.

  2. 1In cases where a novel variant was identified in more than one index case, the number of patients is given and the lipid levels are the mean ± SD.

  3. 2Age in years at the time of cholesterol measurement.

  4. 3Indicates the hypercholesterolemic family member(s) that were analysed to examine whether they carry the same variant or not.

  5. n/d indicates that segregation could not be determined since other members of the family were not available.

  6. *Subject 160 is the father of the index case, who was a carrier of a known mutation (c.1291G>A). The brother of the index case was a compound heterozygote who died at the age of 13 years and was carrying the c.1291G>A and the novel c.628_643dup{636del} variant.

160*M4632223133111c.628_643dup{636del}Carried by compound heterozygote son
974M84413577451c.661_673dupCarried by sister
605M72281615561c.810C>TCarried by mother
100M112111322738c.977C>GCarried by sister, father, grandfather
863F83252326276c.1124A>CCarried by mother
952F102251337586c.1381G>TCarried by brother
1121M42381645787c.1381G>TCarried by father
631F102491864874c.190+361C>TCarried by mother
8 homozygotes323.3 ± 98.3213.0 ± 127.769 ± 27.1116.5 ± 101.9c.191–254_-249del
770M82001335159c.941–38C>TCarried by father
990F42912077352c.1060+59A>CNot carried by father
1077M72651985940c.1987+1_+33delCarried by mother
742F72451429634c.2389+23G>ACarried by father
892M82862056260c.2389+57C>TCarried by mother

Molecular Analysis

Novel LDLR variants and in silico prediction of pathogenicity

In total, 98 index cases (51%) were found to be positive for one of the previously known LDLR mutations. Additionally, 12 novel variants were identified, six within the coding DNA sequence and six in intronic DNA sequence (Table 2). None of the variants described as novel have been previously submitted to dbSNP.

Table 2. In silico analysis of novel LDLR variants
GeneNucleotidePredictedPredictionSpeciesNormalOverallAllele
locationchangeeffectscoreconservationsplicingpredictiondesignation
  1. Prediction score: n/a indicates not applicable.

  2. Species conservation: H, human; Z, zebra fish; X, xenopus; F, fugu; Ck, chicken; B, cow; Gp, guinea pig; D, dog; Rt, rat; Rb, rabbit; M, mouse; Rh, Rhesus macaque; Ch, chimpanzee.

  3. Allele designation: As given by the present study.

Exon 4c.628_643dup{636del}p.(Trp214_Arg215 insHisProLeuSerTrp)n/an/aYesYesPathogenicFH-Kastoria
Exon 4c.661_673dupp.(Lys225Argfs*7)n/an/aYesYesPathogenicFH-Thrace
Exon 5c.810C>TSynonymous changen/an/aYesYesNonpathogenic 
Exon 7c.977C>Gp.(Ser326Cys)10/14H, Z, F, Ck, B, Gp, D, Rt, Rb, M, Rh, ChYesYesPathogenicFH-Leros
Exon 8c.1124A>Cp.(Tyr375Ser)14/14H, Z, X, F, B, Gp, D, Rb, M, Rh, ChYesYesPathogenicFH-Corfu
Exon 10c.1381G>Tp.(Gly461Cys)9/14H, B, Gp, Rb, ChYesYesPathogenicFH-Messinia
Intron 2c.190+361C>T n/an/aYesYesNonpathogenic 
Intron 2c.191–254_-249del n/an/aYesYesNonpathogenic 
Intron 6c.941–38C>T n/an/aYesYesNonpathogenic 
Intron 7c.1060+59A>C n/an/aYesYesNonpathogenic 
Intron 13c.1987+1_+33del n/an/aNoNoPathogenicFH-Heraklio
Intron 16c.2389+23G>A n/an/aYesYesNonpathogenic 
Missense variants

Pathogenicity prediction of missense variants was defined by comparing the results from 14 different methods. A prediction score of ≥7/14 was necessary to characterize the variant as pathogenic (Table 2). Detailed results of all the prediction methods are given in Table S3.

c.977C>G. The c.977C>G p.(Ser326Cys) variant in exon 7 was predicted as deleterious by 10 out of 14 prediction algorithms (prediction score 10/14) (Table 2). Residue 326 is highly conserved between species (12/13 species). DiANNA predicted no changes in disulphide bond formation introduced by the Cysteine residue. HOPE predicted that the altered residue is in close proximity and could affect a cysteine bond made by another residue in the region. The wild type Serine residue forms a hydrogen bond with the Histidine at position 327 and the difference in hydrophobicity introduced by the novel variant will affect hydrogen bond formation. The novel variant is located in the EGF-like 1 domain, and can disrupt this domain and abolish its interaction with other molecules. The c.977C>G novel variant was carried by an 11 year-old boy with T-C 211 mg/dl and LDL-C 132 mg/dl. Although his lipid levels are apparently borderline, he had a very strong family history of hypercholesterolemia, and a sister who was later clinically diagnosed as homozygous FH. His 41 year-old father had T-C 321 mg/dl, his 38 year-old mother had T-C 262 mg/dl, and his younger sister presented T-C 547 mg/dl at the age of two years. The boy's parents did not present with coronary and vascular disease at that time, probably due to their young age, but his paternal grandfather died by myocardial infarction at the age of 55 (Table S2). The proband's father, sister, and maternal grandfather all carried the c.977C>G novel variant.

c.1124A>C. The c.1124A>C p.(Tyr375Ser) variant was predicted as deleterious by all 14 methods used (prediction score 14/14) (Table 2), suggesting that this is a definite pathogenic variant. This residue is highly conserved between species (11/13 species). According to HOPE, this residue is located within the EGF-like 2, calcium-binding domain. The novel variant introduces an amino acid with different properties, which can disrupt and abolish its function. The variant was carried by an 8-year-old girl who presented with T-C 325 mg/dl and LDL-C 232 mg/dl. Her mother, who also carried the novel variant, had T-C 260 mg/dl and her paternal grandfather who presented with T-C 550 mg/dl had a bypass surgery at the age of 60 years (Table S2).

c.1381G>T. In silico analysis of the c.1381G>T p.(Gly461Cys) variant in exon 10 predicted it as deleterious by 9 out of 14 methods (prediction score 9/14) (Table 2), suggesting that it is disease-causing. Based on pathogenicity prediction by the 14 algorithms, this variant was assigned as deleterious, although species conservation was not high (5/13 species). According to the DiANNA predictions, the introduced cysteine residue results in a disulphide bond between residues 461Cys and 379Cys, with a high score of 0.99691. This residue is located within the LDL-receptor class B repeat, which forms a beta-propeller structure and is critical for ligand release and recycling of the receptor. HOPE predicted that the novel variant results in a Cysteine, which might disrupt the core structure of the LDLR class B repeat. The wild-type residue was buried in the core of the protein, and the introduced Cysteine residue is bigger and will probably not fit. The substituted Glycine is the most flexible amino acid and it is possible that it is needed at this position to make a special backbone conformation or to facilitate movement of the protein. The variant introduces a less flexible residue thereby disrupting this conformation or movement. The novel variant was carried by a 10-year-old girl showing T-C 225 mg/dl and LDL-C 133 mg/dl. Although the lipid levels are borderline, the girl had a strong family history of hypercholesterolemia and coronary disease. Her brother, who also carried the novel variant, presented with T-C 231 mg/dl and her father, who died by myocardial infarction at the age of 34, had T-C 300 mg/dl. In addition her paternal grandfather, who also presented with hypercholesterolemia, died by myocardial infarction at the age of 55 years (Supplementary Table S2).

In-frame insertion and truncating variants

Two small rearrangements were identified in exon 4 of LDLR (Table 2).

c.628_643dup{636del}. The c.628_643dup{636del} is a duplication of 16 nucleotides with a simultaneous deletion of a single nucleotide at position 636 (c.636C) in the duplication. It results in the predicted amino acid change p.(Trp214_Arg215insHisProLeuSerTrp), causing the in-frame insertion of five amino acids, which is likely to affect LDLR function. The c.628_643dup{636del} variant was carried by a 46-year-old father who presented T-C 322 mg/dl and LDL-C 231 mg/dl and had a myocardial infarction at the age of 59 years. His son had T-C 800 mg/dl and died by myocardial infarction at the age of 13 years (Supplementary Table S2). He was later found to be a compound heterozygote carrying the novel variant c.628_643dup{636del} and the previously described c.1291G>A mutation, which was inherited from his mother.

c.661_673dup. The duplication of 13 nucleotides c.661_673dup results in a p.(Lys225Argfs*7) change at the amino acid level. It is predicted to cause a frame shift introducing a stop codon leading to a truncated protein, and therefore is predicted to have a major effect on the protein function. The c.661_673dup variant was carried by an 8-year-old boy who presented with T-C 441 mg/dl and LDL-C 357 mg/dl. His mother presented with T-C 350 mg/dl, his sister with T-C 335 mg/dl and his maternal grandfather T-C 400 mg/dl (Table S2).

Splice site variants

None of the identified missense, in-frame insertion, and truncating variants affected normal splicing. In addition, the single synonymous change c.810C>T that was characterized did not cause a potential splicing effect (Table 2). In silico analysis of the intronic variants to assess whether they affect normal splicing (Table 2), showed that c.1987+1_+33del is the only variant that potentially alters normal splicing. It is predicted to disrupt a donor site leading to nonsplicing of intron 13. As a result, the c.1987+1_+33del variant may result in a 694 amino acid protein. The novel variant was carried by a 7-year-old boy with T-C 265 mg/dl and LDL-C 198 mg/dl. His father showed T-C 234 mg/dl, his mother, who was found to carry the variant, presented with T-C 359 mg/dl, and his maternal grandmother had T-C 299 mg/dl (Table S2).

The c.191-254_-249del variant was found in eight index cases in a homozygous state (Table 1). However, the clinical phenotype of these patients was in accordance with the heterozygous state, suggesting that the c.191-254_-249del variant is most probably a polymorphism.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information

To date, there are 1288 LDLR variants reported in FH patients (Usifo et al., 2012). They were all originally reported as pathogenic mutations, although functional studies were not performed for all of them. Usifo and colleagues examined in silico their potential pathogenicity and showed that 79% of the reported 1288 variants are likely to be disease-causing, demonstrating that in silico tools may be useful in pathogenic predictions when functional data are unavailable (Usifo et al., 2012).

The present study aimed to identify novel variants in FH patients of Greek origin, and assess their pathogenicity by in silico analysis. Previously reported mutations were identified in 51% of the unrelated index cases. Taking into account that the present study was based on patient screening and not cascade screening, this mutation detection rate is relatively high. Indeed, the mutation detection rate in other studies performing patient screening in various populations has a range of 13–55.6%, with the average mutation detection rate at 35.7% (Fahed & Nemer, 2011). Additionally, 12 novel LDLR variants were identified, six of which were predicted to be pathogenic. It must be noted that we did not perform in vitro studies to investigate the effect of novel variants on protein function and stability, and we did not examine healthy controls to confirm their presence or absence in healthy individuals. However, a combined use of fourteen different computer programs was applied as a strategy for the most reliable prediction of pathogenicity of novel variants that would differentiate between putative mutations and polymorphisms. In addition, the present study investigated carriage of novel variants within families, and a novel variant was not characterized as pathogenic unless it was confirmed in all of the hypercholesterolemic relatives that were examined.

The c.977C>G p.(Ser326Cys) variant causes a conservative change of a polar to polar amino acid with increased hydrophobicity. It occurs in the epidermal growth factor-like repeat A (EGF-A), which is also involved in LDL binding. The UCL LDLR variant database lists a mutation p.Ser326Phe in the same position, which has been reported once (García-García et al., 2001) and is listed as damaging for LDLR as predicted by PolyPhen and SIFT (Leigh et al., 2008).

The c.1124A>C p.(Tyr375Ser) variant is a conservative change of a polar to polar amino acid substitution. The resulting residue however is smaller in size with reduced hydrophobicity. The substitution lies in the EGF-B, which is not involved in ligand binding (Esser et al., 1988). The LDLR mutation database lists a mutation in the same residue, p.Tyr375Cys, which has been reported three times in different populations (Assouline et al., 1997; García-García et al., 2001; Damgaard et al., 2005) and is listed as damaging for LDLR as predicted by PolyPhen and SIFT (Leigh et al., 2008). This indicates that residue 375 of LDLR is commonly mutated, further strengthening the suggestion that the p.(Tyr375Ser) variant is potentially pathogenic.

The c.1381G>T p.(Gly461Cys) variant causes a nonconservative substitution of a nonpolar by a polar amino acid with increased size and hydrophobicity. No other LDLR mutations have been previously reported in the same residue in the UCL LDLR variant database (Leigh et al., 2008). This residue is located within the LDL-receptor class B repeat, which forms a beta-propeller structure and is critical for ligand release and recycling of the receptor (Jeon et al., 2001). This part of the receptor harboring the EGF and YWTD repeats controls the related processes of lipoprotein release at low pH and recycling of the receptor to the cell surface (Davis et al., 1987). The YWTD repeats have been predicted to fold into a YWTD domain—that is, a six-bladed β-propeller domain. Although the altered residue is located outside the YWTD domain, further studies suggest a structural requirement for maintaining the integrity of the interdomain interface for ligand release in the endosome (Jeon et al., 2001). Additionally, the introduced cysteine residue at position 461 is predicted to form a disulphide bridge with 379Cys, indicating that it could alter the secondary structure of the protein. Taking into consideration the above, the c.1381G>T p.(Gly461Cys) variant may disrupt the function of this region.

The exon 4 variant c.628_643dup{636del} lies within the LA repeat 5 involved in binding of LDL on the cell surface (Esser et al., 1988). The variant inserts five amino acids p.(Trp214_Arg215insHisProLeuSerTrp) resulting in an elongated protein. The change occurs in a highly conserved region, which is frequently mutated, as shown by the UCL LDLR variant database (Leigh et al., 2008), suggesting that this may be a mutation hot spot.

Variant c.661_673dup p.(Lys225Argfs*7) occurs in the LA repeat 5 which is involved in binding of LDL on the cell surface (Esser et al., 1988). The UCL LDLR variant database lists ten single nucleotide changes and 12 small DNA rearrangements in the same region, suggesting that this is a mutation hot spot (Leigh et al., 2008). In addition, Usifo and colleagues analysed in silico three variants that were identified in the same region [p.(Asp221_Asp227dup), p.(Asp224_Lys225delinsPhe), and p.(Asp227del)] and predicted that they disrupt calcium binding and are probably pathogenic (Usifo et al., 2012). Indeed, this region includes the D-x-S-D-E motif spanning positions 224–228, which is highly conserved and involved in calcium binding (Jeon et al., 2001; Jeon & Blacklow, 2005). The c.661_673dup p.(Lys225Argfs*7) duplication substitutes Lysine225 with Arginine, which are both polar and positively charged amino acids and therefore this change could have been tolerated under certain circumstances. However, the subsequent residues 226–231 are changed, with a premature stop codon introduced at position 231. Thus, the highly conserved D-x-S-D-E motif is disrupted. These data suggest that the c.661_673dup p.(Lys225Argfs*7) most likely causes a disruption in calcium binding and a truncated protein of 231 amino acids, and therefore it is a definite pathogenic variant.

Concerning the intronic sequence variants, only the c.1987+1_+33del variant in intron 13 was predicted to affect normal splicing, leading to an altered, truncated protein of 694 amino acids lacking exons 14 to 18 that encode part of the extracellular (EGF-like) domain, as well as the transmembrane and the cytoplasmic domain of LDLR. Therefore, this novel variant is expected to have a major effect on the LDLR protein. The remaining intronic variants were not predicted to disrupt normal splicing, although most of the changes occur close to the exon/intron boundaries. Characterization of the LDLR mRNA and in vitro splicing assays or quantitative RT-PCR may shed light on the effect of these intronic variants on splicing of the LDLR transcript.

It is not surprising that the remaining six novel variants, which were predicted as nonpathogenic, cannot explain the observed FH phenotype. First of all, molecular analysis only detects a mutation in fewer than 50% of patients clinically diagnosed with FH (Marks et al., 2003). Second, the FH phenotype may be explained by mutations in the apoB-100 or PCSK9 genes. It should be noted however that no mutations have been found in the apoB-100 (Dedoussis et al., 2004a, b) or PCSK9 (Diakou et al., 2011) genes in FH patients of Greek origin and therefore, analysis of apoB-100 and PCSK9 is not used in routine diagnosis of FH in Greece. Third, one cannot exclude other, unknown genetic factors that may lead to the FH phenotype. Indeed recent studies suggest a polygenic nature of FH (Talmud et al., 2013). Fourth, these novel variants may be in allelic association with a pathogenic variant that is yet to be identified in the family. For example, direct DNA sequencing of the proximal promoter region, exons and intron/exon boundaries fail to detect large rearrangements that can lead to apparent homozygosity for two normal LDLR alleles.

Previous studies reported 27 LDLR mutations in the Greek FH population (Hobbs et al., 1992; Webb et al., 1996; Kotze et al., 1997, Mavroidis et al., 1997; Traeger-Synodinos et al., 1998; Xenophontos et al., 2000; Miltiadous et al., 2001; Dedoussis et al., 2006, 2004a, b; Laios & Drogari, 2006; Glynou et al., 2008). The present study characterized six potentially pathogenic novel variants, and their identification may lead to earlier molecular diagnosis of FH. Defining pathogenic novel variants is particularly essential when a cascade screening strategy is followed, which is possible in this instance where children and their families are being investigated. Although a more accurate prediction of biological function of DNA variants may be gained by in vitro studies, these are costly and time consuming. In silico analysis of novel variants could be useful in identifying potential disease-causing variants. We applied a method combining fourteen computer programs assessing pathogenicity and changes in protein stability and created a prediction score based on the number of software programs giving a positive result. Evidently, the use of multiple prediction programs outweighs the use of just one or two programs, enabling more accurate predictions. Therefore, we hope that our method will further facilitate the analysis of novel variants and consequently, the molecular diagnosis of FH. This is particularly important for populations that have not been extensively studied, such as the Greek population. As the number of Greek FH patients examined by whole-gene sequencing increases, more novel variants are expected to be defined, and we anticipate that our in silico approach will be advantageous in predicting their pathogenicity. Nevertheless, it must be emphasized that all in silico predictions give an approximation for pathogenicity and cannot replace experimental validation. It is however, a strong basis for further, focused experimental research into the role of these variants in the genetic cause of FH.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information

We are grateful to the University of Athens for funding this project. Many thanks to Anastasia Skouma, Anastasia Tsaroumi, and Aikaterini Michalopoulou for blood collection.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information
  • Abifadel, M., Varret, M., Rabès, J. P., Allard, D., Ouguerram, K., Devillers, M., Cruaud, C., Benjannet, S., Wickham, L., Erlich, D., Derré, A., Villéger, L., Farnier, M., Beucler, I., Bruckert, E., Chambaz, J., Chanu, B., Lecerf, J. M., Luc, G., Moulin, P., Weissenbach, J., Prat, A., Krempf, M., Junien, C., Seidah, N. G. & Boileau, C. (2003) Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet 34, 154156.
  • Assouline, L., Leitersdorf, E., Lambert, M., Reshef, A., Feoli-Fonseca, J. C. & Levy, E. (1997) Identification of two novel LDL receptor gene defects in French-Canadian pediatric population: mutational analysis and biochemical studies. Hum Mutat 9, 555562.
  • Bao, L., Zhou, M. & Cui, Y. (2005) nsSNPAnalyzer: identifying disease-associated nonsynonymous single nucleotide polymorphisms. Nucleic Acids Res 33, W480W482.
  • Brunak, S., Engelbrecht, J. & Knudsen, S. (1991) Prediction of Human mRNA Donor and Acceptor Sites from the DNA Sequence. J Mol Biol 220, 4965.
  • Capriotti, E., Fariselli, P. & Casadio, R. (2005) I-Mutant2.0: predicting stability changes upon mutation from the protein sequence or structure. Nucleic Acids Res 1, W306W310.
  • Cheng, J., Randall, A. & Baldi, P. (2006) Prediction of protein stability changes for single-site mutations using support vector machines. Proteins 1, 11251132.
  • Damgaard, D., Larsen, M. L., Nissen, P. H., Jensen, J. M., Jensen, H. K., Soerensen, V. R., Jensen, L. G. & Faergeman, O. (2005) The relationship of molecular genetic to clinical diagnosis of familial hypercholesterolemia in a Danish population. Atherosclerosis 180, 155160.
  • Davis, C. G., Goldstein, J. L., Südhof, T. C., Anderson, R. G., Russell, D. W. & Brown, M. S. (1987) Acid-dependent ligand dissociation and recycling of LDL receptor mediated by growth factor homology region. Nature 326, 760765.
  • Dedoussis, G. V., Genschel, J., Bochow, B., Pitsavos, C., Skoumas, J., Prassa, M., Lkhagvasuren, S., Toutouzas, P., Vogt, A., Kassner, U., Thomas, H. P. & Schmidt, H. (2004a) Molecular characterization of familial hypercholesterolemia in German and Greek patients. Hum Mutat 23, 285286.
  • Dedoussis, G. V., Pitsavos, C., Kelberman, D., Skoumas, J., Prassa, M. E., Choumerianou, D. M., Stefanadis, C., Humphries, S. E. & Toutouzas, P. (2006) FH-Pyrgos: a novel mutation in the promoter (-45delT) of the low-density lipoprotein receptor gene associated with familial hypercholesterolemia. Clin Genet 64, 414419.
  • Dedoussis, G. V., Skoumas, J., Pitsavos, C., Choumerianou, D. M., Genschel, J., Schmidt, H. & Stefanadis, C. (2004b) FH clinical phenotype in Greek patients with LDL-R defective vs. negative mutations. Eur J Clin Invest 34, 402409.
  • Dehouck, Y., Grosfils, A., Folch, B., Gilis, D., Bogaerts, P. & Rooman, M. (2009) Fast and accurate predictions of protein stability changes upon mutations using statistical potentials and neural networks: PoPMuSiC-2.0. Bioinformatics 25, 25372543.
  • den Dunnen, J. T. & Antonarakis, S. E. (2000) Mutation nomenclature extensions and suggestions to describe complex mutations: a discussion. Hum Mutat 15, 712.
  • Desmet, F. O., Hamroun, D., Lalande, M., Collod-Béroud, G., Claustres, M. & Béroud, C. (2009) Human Splicing Finder: an online bioinformatics tool to predict splicing signals. Nucleic Acids Res 37, e67.
  • Diakou, M., Miltiadous, G., Xenophontos, S. L., Manoli, P., Cariolou, M. A. & Elisaf, M. (2011) Spectrum of LDLR gene mutations, including a novel mutation causing familial hypercholesterolaemia, in North-western Greece. Eur J Intern Med 22, e5559.
  • Esser, V., Limbird, L. E., Brown, M. S., Goldstein, J. L. & Russell, D. W. (1988) Mutational analysis of the ligand binding domain of the low density lipoprotein receptor. J Biol Chem 263, 1328213290.
  • Fahed, A. C. & Nemer, G. M. (2011) Familial hypercholesterolemia: the lipids or the genes? Nutr Metab 8, 2334.
  • Ferrè, F. & Clote, P. (2005) DiANNA: a web server for disulfide connectivity prediction. Nucleic Acids Res 33, W230232.
  • Ferrer-Costa, C., Gelpí, J. L., Zamakola, L., Parraga, I., de la Cruz, X. & Orozco, M. (2005) PMUT: a web-based tool for the annotation of pathological mutations on proteins. Bioinformatics 21, 31763178.
  • García-García, A. B., Real, J. T., Puig, O., Cebolla, E., Marín-García, P., Martínez Ferrandis, J. I., García-Sogo, M., Civera, M., Ascaso, J. F., Carmena, R., Armengod, M. E. & Chaves, F. J. (2001) Molecular genetics of familial hypercholesterolemia in Spain: Ten novel LDLR mutations and population analysis. Hum Mutat 18, 458459.
  • Glynou, K., Laios, E., Drogari, E. & Tsaoussis, V. (2008) Development of a universal chemiluminometric genotyping method for high-throughput detection of 7 LDLR gene mutations in Greek population. Clin Biochem 41, 335342.
  • González-Pérez, A. & López-Bigas, N. (2011) Improving the assessment of the outcome of nonsynonymous SNVs with a consensus deleteriousness score, Condel. Am J Hum Genet 88, 440449.
  • Hobbs, H. H., Brown, M. S. & Goldstein, J. L. (1992) Molecular genetics of the LDL receptor gene in familial hypercholesterolemia. Hum Mutat 1, 445466.
  • Innerarity, T. L., Mahley, R. W., Weisgraber, K. H., Bersot, T. P., Krauss, R. M., Vega, G. L., Grundy, S. M., Friedl, W., Davignon, J. & McCarthy, B. J. (1990) Familial defective apolipoprotein B-100: a mutation of apolipoprotein B that causes hypercholesterolemia. J Lipid Res 31, 13371349.
  • Jeon, H. & Blacklow, S. C. (2005) Structure and physiologic function of the low-density lipoprotein receptor. Annu Rev Biochem 74, 535562.
  • Jeon, H., Meng, W., Takagi, J., Eck, M. J., Springer, T. A. & Blacklow, S. C. (2001) Implications for familial hypercholesterolemia from the structure of the LDL receptor YWTD-EGF domain pair. Nat Struct Biol 8, 499504.
  • Kessling, A. M., Seed, M., Taylor, R., Wynn, V. & Humphries, S. E. (1990) Rising cholesterol levels in children with familial hypercholesterolaemia. Biomed Pharmacother 44, 373379.
  • Kotze, M. J., de Villiers, J. N., Loubser, O., Thiart, R., Scholtz, C. L. & Raal, F. J. (1997) A double mutant LDL receptor allele in a Cypriot family with heterozygous familial hypercholesterolemia. Hum Genet 100, 101103.
  • Laios, E. & Drogari, E. (2006) Analysis of LDLR mutations in familial hypercholesterolemia patients in Greece by use of the NanoChip microelectronic array technology. Clin Chim Acta 374, 9399.
  • Leigh, S. E., Foster, A. H., Whittall, R. A., Hubbart, C. S. & Humphries, S. E. (2008) Update and analysis of the University College London low density lipoprotein receptor familial hypercholesterolemia database. Ann Hum Genet 72, 485498.
  • Marks, D., Thorogood, M., Neil, H. A. & Humphries, S. E. (2003) A review on the diagnosis, natural history, and treatment of familial hypercholesterolemia. Atherosclerosis 168, 114.
  • Mavroidis, N., Traeger-Synodinos, J., Kanavakis, E., Drogari, E., Matsaniotis, N., Humphries, S. E., Day, I. N. & Kattamis, C. (1997) A high incidence of mutations in exon 6 of the low-density lipoprotein receptor gene in Greek familial hypercholesterolemia patients, including a novel mutation. Hum Mutat 9, 274276.
  • Mi, H., Guo, N., Kejariwal, A. & Thomas, P. D. (2007) PANTHER version 6: protein sequence and function evolution data with expanded representation of biological pathways. Nucleic Acids Res 35, D247-D252.
  • Miltiadous, G., Elisaf, M., Bairaktari, H., Xenophontos, S. L., Manoli, P. & Cariolou, M. A. (2001) Characterization and geographic distribution of the low density lipoprotein receptor (LDLR) gene mutations in northwestern Greece. Hum Mutat 17, 432433.
  • Ng, P. C. & Henikoff, S. (2003) SIFT: predicting amino acid changes that affect protein function. Nucleic Acids Res 31, 38123814.
  • Ramensky, V., Bork, P. & Sunyaev, S. (2002) Human non-synonymous SNPs: server and survey. Nucleic Acids Res 30, 38943900.
  • Reese, M. G. (2001) Application of a time-delay neural network to promoter annotation in the Drosophila melanogaster genome. Comput Chem 26, 5156.
  • Reva, B., Antipin, Y. & Sander, C. (2011) Predicting the Functional Impact of Protein Mutations: Application to Cancer Genomics. Nucleic Acids Res 39, e118.
  • Schwarz, J. M., Rödelsperger, C., Schuelke, M. & Seelow, D. (2010) MutationTaster evaluates disease-causing potential of sequence alterations. Nat Methods 7, 575576.
  • Soutar, A. K. & Naoumova, R. P. (2007) Mechanisms of disease: genetic causes of familial hypercholesterolemia. Nat Clin Pract Cardiovasc Med 4, 214225.
  • Talmud, P. J., Shah, S., Whittall, R., Futema, M., Howard, P., Cooper, J. A., Harrison, S. C., Li, K., Drenos, F., Karpe, F., Neil, H. A., Descamps, O. S., Langenberg, C., Lench, N., Kivimaki, M., Whittaker, J., Hingorani, A. D., Kumari, M. & Humphries, S. E. (2013) Use of low-density lipoprotein cholesterol gene score to distinguish patients with polygenic and monogenic familial hypercholesterolaemia: a case-control study. Lancet 381, 12931301.
  • Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 46734680.
  • Thusberg, J. & Vihinen, M. (2009) Pathogenic or not? And if so, then how? Studying the effects of missense mutations using bioinformatics methods. Hum Mutat 30:703714.
  • Traeger-Synodinos, J., Mavroidis, N., Kanavakis, E., Drogari, E., Humphries, S. E., Day, I. N., Kattamis, C. & Matsaniotis, N. (1998) Analysis of low density lipoprotein receptor gene mutations and microsatellite haplotypes in Greek FH heterozygous children: six independent ancestors account for 60% of probands. Hum Genet 102, 343347.
  • Usifo, E., Leigh, S. E., Whittall, R. A., Lench, N., Taylor, A., Yeats, C., Orengo, C.A., Martin, A. C., Celli, J. & Humphries, S. E. (2012) Low-density lipoprotein receptor gene familial hypercholesterolemia variant database: update and pathological assessment. Ann Hum Genet 76, 387401.
  • Venselaar, H., Te Beek, T. A., Kuipers, R. K., Hekkelman, M. L. & Vriend, G. (2010) Protein structure analysis of mutations causing inheritable diseases. An e-Science approach with life scientist friendly interfaces. BMC Bioinfo 11, 548557.
  • Webb, J. C., Sun, X. M., McCarthy, S. N., Neuwirth, C., Thompson, G. R., Knight, B. L. & Soutar, A. K. (1996) Characterization of mutations in the low density lipoprotein (LDL)-receptor gene in patients with homozygous familial hypercholesterolemia, and frequency of these mutations in FH patients in the United Kingdom. J Lipid Res 37: 368381.
  • Xenophontos, S. L., Pierides, A., Demetriou, K., Avraamides, P., Manoli, P., Ayrton, N., Skordis, N., Anastasiadou, V., Miltiadous, G. & Cariolou, M. A. (2000) Geographical clustering of low density lipoprotein receptor gene mutations (C292X; Q363X; D365E & C660X) in Cyprus. Hum Mutat 15, 380384.
  • Yin, S., Ding, F. & Dokholyan, N. V. (2007) Eris: An automated estimator of protein stability. Nat Methods 4, 466467.
  • Yue, P. & Moult, J. (2006) Identification and analysis of deleterious human SNPs. J Mol Biol 356, 12631274.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Conflict of Interest
  9. References
  10. Supporting Information

Disclaimer: Supplementary materials have been peer-reviewed but not copyedited.

FilenameFormatSizeDescription
ahg12032-sup-0001-TableS1.doc44KTable S1: Primers and PCR conditions used for PCR amplification of LDLR.
ahg12032-sup-0002-TableS2.xls33KTable S2: Family history of the index cases carrying the novel LDLR variants.
ahg12032-sup-0003-TableS3.xls32KTable S3: Results of in silico analysis of novel LDLR variants.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.